Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Population patterns of hair zinc, dietary and socio-demographic determinants Vaghri, Ziba 2008

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2008_spring_vaghri_ziba.pdf [ 11.63MB ]
Metadata
JSON: 24-1.0066609.json
JSON-LD: 24-1.0066609-ld.json
RDF/XML (Pretty): 24-1.0066609-rdf.xml
RDF/JSON: 24-1.0066609-rdf.json
Turtle: 24-1.0066609-turtle.txt
N-Triples: 24-1.0066609-rdf-ntriples.txt
Original Record: 24-1.0066609-source.json
Full Text
24-1.0066609-fulltext.txt
Citation
24-1.0066609.ris

Full Text

POPULATION PATTERNS OF HAIR ZINC, DIETARY AND SOCIO-DEMOGRAPHIC DETERMINANTS by Ziba Vaghri BN ., Imperial University of Medical Sciences, Tehran, Iran 1985 . M.Sc ., University of Kashmir, India 1987 M.Sc ., University of British Columbia, 2000  THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY  in  THE FACULTY OF GRADUATE STUDIES (Human Nutrition)  THE UNIVERSITY OF BRITISH COLUMBIA  January 2008 OZiba Vaghri, 2008  ABSTRACT Marginal zinc deficiency (MZD) exists in children of industrialized societies and can impair growth and development . Presently there are no data available on its global prevalence. It is believed that MZD is one of the most common hidden deficiencies throughout the world . This is partly because of the lack of sensitivity and specificity of serum zinc, the most commonly used biomarker of zinc status, to detect MZD . This deficiency in children is always accompanied by a decrease in hair zinc . Although in research settings hair zinc is a recognized biomarker of MZD in children, health practitioners do not presently use it. These cross-sectional studies were designed to examine the hair zinc status of preschoolers in Vancouver . They also aimed at exploring some dietary and non-dietary factors associated with hair zinc status in an attempt to construct and validate a screening tool for detection of MZD. Our first study indicated a mean hair zinc of 75±30 pg/g, with 46% below the cutoff (<70µg/g) for a group (n=87) of low-income preschoolers (Chapter II) . Among these children we observed negative associations between the hair zinc and consumption of dairy (R2 =0.09, P=0 .01) and milk (R2 =0.08, P=0 .01), being described as "often sick" (R2 =0.55, P=0 .00) and "eating unhealthy" (R2 =0.16 P=0 .00), and prolonged breastfeeding (R2 =0.11, P=0 .01). Our citywide survey (n=719) indicated a mean hair zinc of 116±43 µg/g with 17% below the cutoff (Chapter III) . Logistic regression analysis indicated sex, age, maternal education, the number of adults at home, consumption frequency of milk, "scores of activity level", "being described as frequently sick" and "taking supplements containing  ii  iron" as the significant predictors of hair zinc status . However, the final model had 16% sensitivity while having 98 .3% specificity, indicating its lack of usefulness as a screening tool. Our study provides important information on the hair zinc status of Vancouver preschoolers . Although we did not accomplish our primary goal of constructing and validating a screening tool, we did identify some factors in children and their environment associated with hair zinc, which may help in better understanding of hair zinc as a biomarker of MZD .  ;;  i  TABLE OF CONTENTS ABSTRACT TABLE OF CONTENTS  iv  LIST OF TABLES  xii  LIST OF FIGURES  xiv  PREFACE  xv  ACKNOWLEDGMENTS  xvi  DEDICATION CO-AUTHORSHIP STATEMENT  xx  CHAPTER I : INTRODUCTION  1  I.1. BACKGROUND  2  I.2. LITERATURE REVIEW  4  I.2.I. Population health  4  I.2 .II. Marginal zinc deficiency (MZD)  7  I.2 . II.A. The prevalence  9  I.2.II.B. Signs and symptoms  10  I.2 . H. B. a. Serum/plasma zinc  13  I.2.II.B . b . Hair zinc level  15  I.2.II .C . Population health approach to MZD  23  I.2.III . Factors associated with zinc deficiency  26  I.2 .III.1 . Dietary factors associated with zinc deficiency  26  I .2. III .2 . Socio-demographic factors associated with zinc deficiency/status  30  iv  I.2.III.3 . Other factors associated with zinc deficiency/status  33  I.3. The present knowledge gap  36  I.4. Purpose of the study  38  I.S. Objectives  38  I.S .I. Objectives for chapter II  38  I.5 .II . Objectives for chapter III  39  I .6. Bibliography  41  CHAPTER II 52 HAIR ZINC STATUS AND DETERMINANTS OF LOW HAIR ZINC AMONG A GROUP OF LOW-INCOME PRESCHOOLERS OF VANCOUVER 52 11.1 . INTRODUCTION  52  II.2 . METHODS AND MATERIALS  56  II.2.a. Study design  56  II.2.b. Subject recruitment  57  II .2.c. Inclusion criteria  57  II.2.d. Nutrition Clinics  58  II .2.e. Socio-demographic and socio-economic questionnaire  58  II.2.f. Biochemical data  59  II.2.f.A. Collection of hair samples for zinc analysis  59  II .2 .f. B. Analyses of hair samples  60  II.2.g. Anthropometrics  60  II .2 .h. Dietary data ; collection and processing  61  II .2.h .A. Data collection  61  II.2.h .A.a. Food frequency questionnaire, FFQ  61  v  II.2 .h.A.b . Data on other nutrition and health related variables  63  II.2.h.B . Processing of dietary data  63  II.2.h.B.a. Food Groups Assignment  63  II.2.h.B.a.Estimating the number of servings  65  II.2 .i. Statistical Analysis  65  11 .3 . RESULTS  67  II.3 .a. Study population  67  II.3 .b. Hair zinc status of the study population  69  II. 3 .c. Associations of the 5 main socio-demographic and socio-economic variables with hair zinc 72 II .3 .d. Associations of some dietary factors with hair zinc 75 11.3.e. Associations of some health and behavior related variables with hair zinc . . . . 78 II .3 .f. Contribution of all variables associating significantly with hair zinc to the overall variability in hair zinc of the inner city children  82  II.4. DISCUSSION  84  II.5. BIBLIOGRAPHY  91  CHAPTER III 98 HAIR ZINC STATUS AND DETERMINANTS OF HAIR ZINC OF VANCOUVER PRESCHOOLERS AND THE USEFULNESS OF THESE DETERMINANTS AS A SCREENING TOOL FOR MARGINAL ZINC DEFICIENCY 98 III . 1 . INTRODUCTION  99  I1I.2. METHODS AND MATERIALS  104  III.2.a.Study design  104  III.2 .b. Contacting the potential centers to participate  105  III.2.c . Meeting with centers prior to the commencement of the survey  106  11I.2. d. Inclusion criteria  107  vi  11I.2. e. Survey Tool  107  III .2.e.A. Development of the survey questionnaire  107  III.2 .e.B . Components of the survey questionnaire  107 108  111.2 . e . B.1 . Socio-Demographic Information  II1.2.e .B.2. Information on child's breast-feeding history and perceived health status 108  III.2. e. B . 3 . Information on supplement intake  109  111.2 . e. B. 4 . Information on children 's eating behavior  110  III.2 . e . B. S. Food Security Questions  111  111.2. e . B. 6. Behavioral Information  112 113  III.2.e.C . Verifying the face validity of the questionnaire III.2.f. The survey sequence of events  114  III .2 .g. Sample Size Calculation  118  III.2 .h. Pre-testing the survey  122  III.2.i . Reliability of the questionnaire  122  III.2.j . Incentive  123  III.2 .k . "Thank you" letters  123  1II.2 .1. Educational pamphlets  123  III.2.m. Biochemical data  124  11I.2.n . Anthropometrics  125  III.2.o. Statistical analyses  125  I1I .3 . RESULTS  128  III.3 . a . Survey population  128  III. 3 .b. Hair zinc status of the survey population  130  vii  III .3 .c. Associations of the 5 main socio-demographic and socio-economic variables with hair zinc 135 III.3.d. Association of some family characteristics with hair zinc of the children  145  III.3.e. Association of eating behaviors with hair zinc of the children  149  III.3 .f. Associations of breast-feeding, parental perception of children's health status and taking supplements containing iron with hair zinc of the children 153 III .3 .g. Contribution of all variables associating significantly with hair zinc, to the overall variability in hair zinc 159 III.3 .h. Exploring the usefulness of the model/questionnaire produced by these components in predicting hair zinc status  161  III.4. DISCUSSION III.5.BIBLIOGRAPHY  166 188  CHAPTER IV : CONCLUSION  203  IV.1. GENERAL CONCLUSIONS  204  IV.2. STRENGTHS AND LIMITATIONS OF THE STUDY  210  IV.4.BIBLIOGRAPHY  215  APPENDICES Appendix II-1 : Informed consent form for recruitment questionnaire  220  Appendix II-2 : Recruitment questionnaire for study of low-income preschoolers  221  Appendix II-3 : Add placed in the bulletin boards of the community centers inviting eligible families to attend the nutrition clinics 224 Appendix II-4 : Informed consent form for attending clinics  226  Appendix II-5 : passports used in Nutrition Clinics  228  Appendix II-6 : An excerpt from the socio-demographic questionnaire of the study of "Food insecurity in inner city children", which collected data on  vii  maternal education and family income of the study participants  229  Appendix.Il-7: Measurement sheet used in Nutrition Clinics  230  Appendix II-8 : Food frequency questionnaire used in study of low-income preschoolers  231  Appendix II-9 . Comparison of the daily food intake of boys and girls of the study . . .250 Appendix III-1 : Certificate of ethics approval for "Population patterns of hair zinc  251  Appendix III-2 : Neighborhood maps produced by GIS unit of HELP to facilitate the planning of the recruitment strategy  252  Appendix III-3 : Letter of support from the Executive Director of West Coast Child Care Resource CentrE  253  Appendix III-4 : letter from the principal investigator of the project  255  Appendix III-5 : The initial letter of contact for survey participants  257  Appendix III-6 : Recruitment form for participants of focus groups  259  Appendix. III-7: Information sheet for recruited focus group participants  260  Appendix III-8: Informed consent form for focus group participants  261  Appendix III-9 : Survey questionnaire  262  Appendix III-10 : Survey questionnaire translated in Punjabi  268  Appendix III-11 : Survey questionnaire translated in Cantonese  274  Appendix III -12 : Subject's information letter  280  Appendix III-13 : Informed consent form for survey participants  283  ix  Appendix III-14 : Log sheet of returned surveys  284  Appendix III-15 : Reminder letters  285  Appendix III-16: Notice posted in the centers announcing the day and time that the survey team will visit the center for sample collection and height and weight measurement of the consented children  286  Appendix III-17: Cover letter for reliability package  287  Appendix III-18 : Letter to the winners of baskets  288  Appendix III-19: pictures of the winners of "baskets  289  Appendix III-20: Thank you letter for survey participants  290  Appendix III-21 : Thank you letter for participating centers  291  Appendix III-22 : Thank you letter for West Coast Family Services  292  Appendix III-23: The educational pamphlet  293  Appendix III-24: Log sheet of the hair samples delivered to the laboratory  294  Appendix III -25 : Log sheet of height and weight  295  Appendix III-26 : Moments captured ; Nutrition clinics and survey  296  Appendix III-27: Hair zinc level of 5 main ethnicities of the study along with the sex and age breakdown of these ethnic groups  298  Appendix III-28 : Regression analyses of "taking supplements containing iron" with eating behaviors  299  Appendix III-29: Comparison of the indices of growth (HAZ and WHZ) and nutrition (WAZ) of low hair zinc and normal hair zinc younger (<4 years old) children  300  Appendix III-30 : Mean ± SD of the daily intake frequency of some main food groups of survey children and the three main ethnic groups  301  Appendix III-31: Distribution of supplement takers among the survey participants (bottom pie) and distribution of iron supplement takers among the supplement takers of the survey  302  xi  LIST OF TABLES Table II.1 . Description of study population (n=87)  68  Table II.II. Socio-demographic and socio-economic description of study population and hair zinc levels and the occurrence of low hair zinc based on these variables (n=87)  72  Table II.III. Relationship between hair zinc and demographic factors (n=87)  73  Table II.IV. Average (±SD) of daily intakes of different food groups (servings/day), and regression analyses (both unadjusted and Adjusted) of these intakes with hair zinc (µg /g) of the study population (n=74)  76  Table II.V. Relationship between hair zinc and some nutrition and health related variables as well as socio-demographic and anthropometric variables among study population  79  Table II.VI. Step-wise linear regression analyses of hair zinc and the variables significantly associating with it  82  Table.III.I : A summary of the estimated number of survey packages needed to be given out in each neighborhood  121  Table III. II. Socio-demographic description of study population (N=719)  129  Table III . III. Mean (±SD) hair zinc, and rate of low hair zinc and the contribution of each neighborhood to survey participants  134  Table III.IV. A socio-demographic and socio-economic description of the study population and the hair zinc levels (X±SD) and the occurrence of low hair zinc based on these variables among them (n=719)  137  Table III .V. Influence of age on hair zinc (µg/g) of boys and girls of the survey (n=705)  140  Table III.VI. Relationship between hair zinc and the 5 main socio-demographic and socio-economic factors in the study population (n=719)  144  Table III .VII. Description of some family characteristics as well as the hair  xii  zinc level and occurrence of low hair zinc based on these characteristics (n=719)  146  Table III .VIII .. Relationship of hair zinc with some family characteristics  148  Table IXA . Comparison of the daily consumption frequency of some main food groups of low hair zinc and normal hair zinc children (n=719)  150  Table III.IX.B. Description of some food behaviors and the hair zinc level and occurrence of low hair zinc based on these behaviors (n=719)  151  Table III.X. Regression analyses of hair zinc (µg /g) with some eating behavior (n=705)  152  Table III.XI.A Description of some health behaviors and the hair zinc level and occurrence of low hair zinc based on these behaviors  154  Table III .XIB. Comparison of the scores of attentional focusing and activity level of low hair zinc and normal hair zinc children=719  155  Table III.XII. Regression analyses of hair zinc (µg /g) with some health behaviors  158  Table III.XIII. Step-wise linear regression analyses of hair zinc and the variables significantly associating with it  160  Table III.XIV. Logistic regression analysis of all significant correlates of low hair zinc (n=719)  163  Table III .XV.A. Classification of the survey children based on their hair zinc Status without the use of the model produced by logistic regression analysis  164  Table III.XV.B. Classification of the survey children based on their hair zinc status with the use of the model produced by logistic regression analysis  164  xii  LIST OF FIGURES Figure I.I. Steps in development of zinc deficiency (constructed based on King, 1990 andPrasad,198 8)  8  Figure II.I. Frequency distribution and descriptive statistics for the hair zinc level of the study population  70  Figure II.II. Hair zinc level of three groups of children, categorized based on length of exclusive breast-feeding    80  Figure III.I. Schematic presentation of survey waves and their time-lines  117  Figure III.II. Frequency distribution and descriptive statistics of the hair zinc Level of study population  131  Figure III.III. Comparison of hair zinc distribution of the present study with that of low -income inner city preschoolers in Vancouver  132  Figure III. IV. Scatter plot of hair zinc against age of the children (n=705)  139  Figure III.V. Comparison of the occurrence of low hair zinc in different age categories of girls (the upper line) and boys (the lower line)  142  Figure III.VI. Receiver Operating Characteristic (ROC) curve for dietary and social predictors to detect children with low hair zinc  165  xiv  PREFACE This thesis has been prepared using a manuscript-format described by the Faculty of Graduate Studies of the University of British Columbia University .Therefore chapters II and III are elaborated forms of the manuscripts to be submitted to relevant scientific journals. Although chapters are arranged in chronological order of the data collection, each chapter stands on its own and can be read in any order . The relevant forms and documents of each study are presented as appendices .  xv  ACKNOWLEDGEMENTS I like to start by acknowledging the great contribution of my research supervisor, Dr. Clyde Hertzman whose wealth of knowledge and passion for the betterment of children's life inspired and then supported my own research ideas and steered them to completion. I am also very grateful to the contribution of Dr . Susan Barr, my graduate advisor and a member of my supervisory committee, whose insightful suggestions, careful critiques and perceptive advice contributed to and greatly enhanced the quality of this work. My deep gratitude goes to Dr . Hubert Wong, a member of my supervisory committee, whose statistical knowledge and kind and patient teaching sessions always managed to show me what I should do next. Special thanks also go to Dr . Gwen Chapman who over the years taught me, by example, how to keep a balance between professionalism and compassion. I am grateful for my family ; for Maman, Agajan, the greatest and most caring parents in the world, my sisters, Simin, Nastaran, my sister-in-law Ladan and my brother, Ali, who were always ready to help and whose love and support pulled me through some very difficult parts of this journey. I am thankful for all the kind people encountered along the way, who made an impression on my work or me somehow, including the staff and fellow graduate students of FNH, friends at the BC Research Center for Women and Children's Health and the Human Early Learning Partnership (HELP) .  xvi  I feel blessed for the continuous support and encouragement that I received from two precious friends, Reverend Peter Niblock, and Dr . Pauline Coti, who never stopped believing in me and offered to help even before I ask for it. Finally, I am overwhelmed by the love I received from my children; my son, Arash, who during the final year of the program almost weekly had to refuel me by saying; "Keep going Mom, keep going, you are almost there!" and my daughter, Avishan, who, over the years, sang with me every morning in the car as we drove to daycare, elementary, and then high school. "Inch by inch row by row We are gonna make this garden grow all it takes is work and hope"  Thank you all!  In remembrance of Gholam-Reza Ghorbani his life and his legacy  xviii  My years in the doctoral program were a long journey that produced above and beyond this thesis . Difficult life trials were blissfully followed by great triumphs and invaluable life lessons. As the absence of light creates darkness and the absence of heat creates cold, lack of responsibility and respect in educational interactions create darkness and cold in which no effective education can possibly occur! This work is dedicated to all great educators who understand the enormity of their responsibility and the significance of their roles in training students and shaping their lives. Those who put caring and respect as essential and ever-present elements in their interactions, and in doing so, they set us free Since "in knowledge we find freedom"!  xix  CO-AUTHORSHIP STATEMENT Chapters II and III are being developed as manuscripts . For each manuscript, I have identified research questions, conceived of the study design, recruited all participants, completed all data collection and managed, planned and conducted data analyses. The co-authors will be those who make a significant contribution to either the study design and/or implementation, stimulation of the discussion of the results, or editorial process of the manuscripts .  xx  CHAPTER I INTRODUCTION  t  I.1. BACKGROUND It is evident that good health in the adult is predicated on good health in the child. Many of the key determinants governing health have their roots in the biological and social experiences that span childhood . A healthy child population is not only one of the indicators of a society's present health, but it may also be an indicator of that society's future well being and productivity. The health of a child derives from the influence of genes, `health behavior', ecology, social and societal characteristics, and medical care (Brownell et al, 2002). Nutrition as a part of `health behavior' is one of the key elements to sound childhood health. Nutritional inadequacy can result in deficiencies of both macro- and micronutrients . Some of these deficiencies are known to have an adverse effect on growth and development . The preschool period is a crucial stage of intense and ongoing growth and development. For this reason a preschooler is highly vulnerable to the impact of these deficiencies . Zinc is one such essential micronutrient. In its marginal form, zinc deficiency has been documented in many industrialized societies (Hambidge et al, 1972, Walravens et al, 1983, Nakamura et al, 1993, Buzina et al, 1980 and Chakar et al, 1993), including Canada (Smit Vanderkooy & Gibson, 1987 and Gibson et al, 1989-b) . The nature of the diets associated with marginal zinc deficiency (MZD) suggests that it may be more common than is thought among some vulnerable age groups in industrialized countries . Durung this stage, zinc deficiency can easily go undiagnosed .  2  Like many efficiencies in their early stages, marginal zinc deficiency has virtually no specific clinical symptoms . At the biochemical level, however, though it does not result in any measurable change in the main biomarker of zinc status i .e. the serum zinc, it does leave some clues that can be a helpful tool for its proper diagnosis . Marginal zinc deficiency, in children, is almost always accompanied by a decline in the hair zinc level. Hair zinc is believed to reflect the body's store of this element (Gibson, 1989). Evidence does support the view that hair zinc concentration can be an index of MZD in children, at least in developed countries where widespread protein-energy-malnutrition does not exist . The suitability and usefulness of hair zinc in the assessment of zinc status and the detection of deficiency at an early stage has been demonstrated by the growth response of affected children (low hair zinc children) to zinc supplementation . In addition, being a convenient and non-invasive test, it has this additional advantage in assessing MZD in the pediatric age group. As a part of pediatric health promotion, management of MZD can be approached through a population health approach . In this approach knowledge of the determinants of this deficiency is pivotal . These can include any factor within a broad range from environmental and health associated socio-demographic factors to interpersonal characteristics of the affected individuals and populations . Awareness of the determinants of a disease or condition is a pre-requisite to any successful intervention and can be gained from the gathered evidence of epidemiological studies and population surveys and large-scale studies. Studies of preschoolers have documented associations between low hair zinc and some dietary factors such as the intake of bioavailable zinc and the intake of enhancers 3  and inhibitors of zinc absorption (Smit Vanderkooy & Gibson, 1987). It has been further documented that certain socioeconomic factors may be associated with MZD/low hair zinc (Hambidge et al, 1976, Walravens et al, 1983 and Chakar et al, 1993). This study was the first to examine the hair zinc status of a large sample of Canadian preschoolers . It was also the first attempt to develop and validate a screening questionnaire for detection of low hair zinc/MZD. The information collected through this thesis will enhance our understanding of the risk factors associated with low hair zinc status in the preschool population, which could in turn serve in the identification of priority groups and at-risk sectors of the population. Prior to presenting the study and its findings, the relevant literature will be reviewed in an attempt to present the current state of knowledge and those areas where there is still a knowledge gap . The literature review will begin with a brief introduction to the population health approach as a framework for the study . Subsequently marginal zinc deficiency, its prevalence, diagnosis, signs and symptoms will be reviewed . Hair zinc as a cardinal biomarker of this deficiency will then be examined . The literature review will end by highlighting some of the gaps in our present knowledge. 1.2. LITERATURE REVIEW 1.2.1. Population health The concept of population health in Canada began in the 1970s . Perhaps the point of genesis of population health in Canada was the federal government's White Paper, proposing that changes in lifestyles or social and physical environments would likely lead to more improvement in health than would be achieved by spending more money on existing health care delivery systems (Government of Canada, 1974) . 4  Population health is defined as the health, well-being, and functioning of a clearly defined population (Kindig & Stoddart 2003) . The defining line can be based on locality, biological criteria such as age, social criteria such as socio-economic status, or cultural criteria. Population health is an approach to health that aims to improve the health of the entire population and to reduce health inequities among population groups . In this approach population is not just a group of individuals, but also the trends and themes among them . Therefore, " . . . the outcome and benefit of such an approach extends beyond improved population health outcomes to include a sustainable and integrated health system, increased national growth and productivity and strengthened social structure" (Public Health Agency of Canada, 2002) . The population health approach to health has some central elements on which it is founded and the knowledge of these elements is essential to its understanding. One of the main pillars of population health is a fundamental belief that there are factors outside individuals and the health care system that can affect health significantly. Therefore, the entire range of individual and collective factors and the interaction among these factors are to be considered if one intends to address health in a comprehensive and wholesome manner. These factors, referred to as "determinants of health", are the key elements of the population health approach. Another important notion in population health is that the earlier in the causal stream action is taken, the 'greater is the potential for population health gain . Therefore, "primary prevention" is the preferred point of intervention . To embark on this "primary prevention" task, a thorough knowledge of all the factors involved in, leading to and  5  influencing the morbidity of the disease is essential . In short, primary prevention is a socio-environmental approach, confronting the root cause of disease. The primary prevention approach can be adopted within many intervention frameworks identified for population health improvement ."Improved child development" has been acknowledged as one of these frameworks and has gained a great deal of attention from both the research and policy maker communities within the last decade . A national Children's Agenda was agreed to in the late 1990s in Canada . Since then some impressive steps have been taken towards the development of comprehensive strategies for optimum early childhood development across Canada, among them the recent initiative of the Early Development Index (EDI), which, in British Columbia, measures kindergarten children's readiness for school in 5 areas of child development that have a long-term impact on health, well-being and school success (Hertzman et al, 2004) . The findings of this province-wide study have resulted in the development of BC Child Atlas (HELP, 2007) . BC Child Atlas is an invaluable guide for further planning for population health promotion within the framework of child development and via a primary prevention approach. Apart from environmental factors such as community resources, early child development can also be affected by nutrition . Nutritional inadequacies in childhood can give rise to macronutrient and micronutrient deficiencies . Although the former are not frequent among the affluent and industrialized nations, the latter are a valid pediatric health concern of these nations . Among the micronutrient deficiencies of concern in North America is marginal zinc deficiency (MZD) . In the following sections this  6  deficiency, its prevalence, signs and symptoms and potential biomarkers will be reviewed. 1.2.I L Marginal zinc deficiency (MZD) Zinc deficiency has been observed across a wide spectrum from marginal and mild to advanced and severe (Prasad, 1985) . Clinical signs of marginal zinc deficiency in childhood include impaired growth, poor appetite and diminished taste acuity (Gibson, 1989), none of which is specific to zinc deficiency per se . These symptoms, however, are usually accompanied by a decline in hair zinc in the case of MZD . Marginal zinc deficiency can best be detected through a positive growth response in supplementation trials (Prasad, 1988) . Figure I.I is a schematic presentation of the three stages of zinc deficiency and their accompanying clinical and biochemical signs and symptoms .  7  Adaptation tnaka~tr and  stage apt' deficiency  accompanying blo chat-deal  and clinical svmotoms   Inadequate dietary auppl,  Reduced growth or zinc QKcrQ*IQrr  -7;4.vid iixxue  yiit~.r  onxa. ry "nova  dietary n horbade extended  M gib 111,4 1iiii al 'flair fro ni E ZP C..tn.ral lixx*a. dy*fain clIon  EZP : Exchangeable Zinc Pool In marginally zinc deficient individuals, the deficiency can advance to a severe stage without going through a moderate zinc deficiency if the zinc need increases drastically (e .g . in case of injury, burn and trauma) Figure 1 .I. Steps in development of zinc deficiency (constructed based on King, 1990, and Prasad, 1988)  8  As Figure I.I depicts, marginal zinc deficiency is a start point that can progress to the more serious stages of deficiency i .e . moderate and severe zinc deficiency. A moderate level of zinc deficiency has been reported in a variety of conditions as a secondary/conditional deficiency due to chronic diseases like cystic fibrosis and pancreatic insufficiency . Clinical manifestations of a moderate level of zinc deficiency include growth retardation, rough skin, poor appetite, hypogonadism in males, delayed wound healing, lethargy, abnormal neurosensory changes and compromised cellmediated immune dysfunction (Prasad, 1988). Chronic or severe deficiency is often accompanied by severe malnutrition and will exhibit clear signs of lowered immunity, halted or delayed puberty, stunting, dermatitis, alopecia, diarrhea, emotional disorder, and weight loss (Prasad, 1988) . With this level of deficiency one often sees a pronounced decline in serum/plasma zinc. Reduction in hair zinc may or may not be present depending upon the absence or presence of protein energy malnutrition (PEM) (Gibson, 1989). I.2. II.A. The prevalence Marginal zinc deficiency was first observed in American preschoolers of lowincome families in Colorado (Hambidge et al, 1972) . Though at present there is no precise information as to the number of people affected by dietary zinc deficiency, from the Food and Agricultural Organization, FAO, and its national food balance data, the Stockholm Conference, 2000 reported an estimated 48% of the global population at risk of zinc deficiency (Stockholm conference, 2000) . While severe deficiency appears to be predominant among underdeveloped and developing countries (Hambidge & Krebs, 2001), MZD is seen in developing and even developed countries such as the U .S. 9  (Hambidge et al, 1972 and Walravens et al, 1983), Canada (Smit Vanderkooy & Gibson,1987 and Gibson et al, 1989-b), Japan (Nakamura et al, 1993), and France (Chakar et al, 1993). In North America, the first study to report the existence of marginal growthlimiting zinc deficiency found low hair zinc levels in children to be associated with impaired taste acuity, poor appetite and low growth percentiles (Hambidge et al, 1972). Studies in the 1970s and 1980s confirmed this deficiency in apparently healthy infants, toddlers, and preschoolers even after routine fortification (in the US during the 1970s) of infant formulas with zinc (Hambidge et al, 1972, Hambidge et al, 1976, and Walravens et al, 1983) . Some researchers from Eastern Canada (Smit Vanderkooy & Gibson, 1987 and Gibson et al, 1989-b) have also described a growth-limiting marginal zinc deficiency in preschoolers of Ontario. Smit Vanderkooy & Gibson (1987) examined food consumption patterns and nutrient intakes of children aged 4-5 years in relation to their zinc status and suspected the existence of a marginal zinc deficiency manifested as low hair zinc (<70µg/g) and slowed physical growth . The occurrence of low hair zinc among these 107 study subjects was estimated at 21 .3% for boys and 4 .7% for girls . The subsequent study of this research team confirmed this deficiency among some Ontario children through the growth response of a selected (height < 15 th and mid-parent height >25 th percentiles) group of 5-7 year old boys to a zinc supplementation program (Gibson et al, 1989-b). No further study has been conducted on zinc status; of this age group in Canada. I.2.II.B. Signs and symptoms Marginal nutritional zinc deficiency is clinically characterized by the slowing of physical growth, poor appetite, and diminished taste acuity (hypogeusia) (Gibson, 1989).  to  These clinical signs are non-specific and often go unnoticed . For instance, loss of appetite can be the ramification of many other physiological, psychological and emotional problems . Similarly hypogeusia, impaired taste acuity, can be the result of a number of neurological or connective tissue disorders. Slowing of physical growth, although often an accompanying sign of MZD, cannot be used as a reliable diagnostic tool by itself for various reasons . Firstly, the slowed physical growth at the marginal level of deficiency is not always a visually apparent and easily detectable sign . A health care professional has to obtain the anthropometric data and convert these to height-for-age, weight-for-age, and weight-forheight Z scores in order to test their normalcy, using the appropriate reference tables (Centers for Disease Control and Prevention, 2002) . Secondly, as mentioned earlier, slowed physical growth is not specific to zinc deficiency and can be the consequence of other micro-nutrients deficiencies such as iron, iodine and vitamin A . Thirdly, slowed physical growth does not always manifest itself in measurable changes in Z scores or other anthropometric indices and this feature of MZD can therefore remain un-detectable for some children . Height achieved is the result of genetic endowment and macro as well as micronutrient availability during the growth period . Slowed growth may result in anthropometric measurements that are still within the expected values of child's age group, simply because the child had a larger genetic allowance. Therefore, it is reasonable to conclude that none of the clinical features are specific and as a result none can be the diagnostic tool by itself.  11  As for the biochemical features of marginal zinc deficiency, the situation is somewhat different . One such feature is diminished hair zinc content . Animal studies reveal that the uptake of zinc by hair is slow (Deeming and Weber, 1977) and may be impaired preferentially if the amount of zinc absorbed is decreased . Therefore, it is possible for hair zinc to be markedly depressed while more vital tissues continue to receive an adequate supply, with few, if any, other symptoms of zinc deficiency apparent (Hambidge, 1980). This physiology points to a superior sensitivity, as well as a specificity, of hair zinc as an index of body zinc status . Evidence suggests that hair zinc concentration could be an indicator of chronic marginal zinc deficiency during childhood when the confounding effect of severe protein-energy malnutrition is absent (Hambidge et al, 1972, Smit Vanderkooy & Gibson, 1987 and Gibson et al, 1989-b) . Hair zinc, therefore, cannot be used in cases of very severe malnutrition and/or severe zinc deficiency when the rate of growth of the hair shaft is diminished . Under these circumstances the hair zinc level may be normal or even higher than normal (Hambidge, 1980 and Gibson, 1989) . For these reasons, diminished hair zinc in children, in the absence of PEM, has been viewed as the cardinal feature of marginal zinc deficiency (Gibson, 1989). Serum zinc, the most widely used biomarker of zinc deficiency is not a sensitive tool for the diagnosis of MZD . At this level of deficiency the homeostatic mechanisms of the body are fully functional and active and prevent any drop in the serum zinc . Below is a brief review of these 2 biomarkers and the degree of their usefulness in the diagnosis of marginal zinc deficiency.  12  I.2. ILB.a.Serum/plasma zinc  Serum/plasma zinc is the most common marker of zinc status . Zinc in plasma is in extremely low concentrations and is tightly regulated and maintained within a normal range, despite a wide range of change in the dietary zinc intake (Hambidge, 1980) . This makes the plasma zinc an insensitive marker of dietary zinc deficiency . Despite the tight regulation/protection of this marker against dietary fluctuations, it is also easily affected by many other factors such as diurnal rhythm, stresses, infection, fasting/non-fasting, and plasma protein levels (Guillard et al, 1979) . Therefore, it is reasonable to conclude that serum zinc is neither specific nor sensitive as a biomarker of zinc status during the early stage of zinc deficiency. Despite this lack, a growing body of evidence supports its usefulness as an indicator of population zinc status (Bahl, Chaudhuri, & Pathak, 1994 and Brown, Peerson, & Allen, 1998) . The risk of zinc deficiency in a population can be estimated by comparing serum zinc data of that population with statistically defined cutoffs from a presumably healthy population (Hotz et al, 2003) . In addition, measurable differences in serum zinc concentration have been reported to occur in various populations due to changes in dietary zinc intakes and clinical conditions associated with zinc deficiency (Hunt, Mathys, & Johnson, 1998, and Srikumar et al, 1992) . Therefore, serum zinc can elucidate how a population zinc status (mean serum zinc, distribution, percentiles . . .) does compare with reference to a healthy population . It can also act as a reference for a given population itself and help to monitor the changes in zinc nutriture experienced over time. Low serum zinc concentrations have been reported in Canada in young women with low serum ferritin values (Donovan & Gibson, 1995), USA (Yokoi, Alcock, &  13  Sandstead, 1994), and New Zealand, (Gibson, Heath, & Ferguson, 2002) . Because the dietary factors associated with the etiology of iron deficiency also induce zinc deficiency, it has been reasoned that the prevalence of sub-optimal zinc status is likely to be comparable to that of iron deficiency (Gibson, Heath, & Ferguson 2002) . This has been supported by the co-morbidity of zinc and iron deficiency among many populations of underdeveloped societies, among them Middle Eastern adolescents . The zinc content of the diet of these individuals (with zero to very little animal meat) was a few times greater than the recommended nutrient intake (RNI) at that time (-30mg/day), but due to high levels of zinc and iron inhibitors in their diet they failed to meet their needs and were severely zinc and iron deficient (Prasad, 1985) . Whether or not this co-morbidity of iron and zinc deficiency applies to developed societies such as ours where large-scale ironfortification of cereals and grain products are in place, remains to be explored . No studies to date have been carried out to evaluate the impact of fortification of cereals with iron on the iron status of Canadians . However, in theory, the likelihood of being iron adequate while zinc deficient in a diet with a low intake of meat, in light of the iron fortification of cereals, could be entertained. The serum zinc level appears to be a reflection of dietary zinc only in extreme cases . It drops profoundly in persons with severe zinc deficiency when the intake is so low that homeostasis cannot be established without the use of zinc from the exchangeable zinc pool, EZP, (of which plasma/serum is a component) (King, 1990) . In view of the degree of protection conferred upon this zinc compartment in the human body (serum/plasma), it is not difficult to see why it is considered a very poor and misleading measure of marginal zinc deficiency (King, 1990 and Gibson, 1989) . 14  I.2.II.B.b . Hair zinc level  Trace elements get incorporated in the hair during its development and are derived from the matrix and connective tissue papillae (with blood and lymph vessels) and sebaceous glands . As the extruded hair approaches the skin surface, its outer layer becomes hardened and relatively impermeable ; therefore, isolating the hair from the body's continuing metabolic activities (Hopps, 1977) . For this reason zinc in the hair is not affected by the factors that affect serum zinc such as circadian rhythm . It is also not subject to a rigid regulation in response to fluctuations in dietary zinc . Once the metal becomes incorporated into the hair during hair growth, it is no longer in equilibrium with the body and is thus not affected by on-going metabolic events (Assarian & Oberleas, 1977) . Hair zinc reflects body stores of this element (Dorea & Paine, 1988) . Indeed, many researchers have discussed the suitability of this biological sample as an index of zinc status . As an index of zinc status, hair zinc has gained popularity over the last few decades . Perhaps this, in part, is due to the ease of sample collection as well as sample storage and handling . This is a particularly important factor when sampling large populations and in young children whose pain and discomfort experienced in an invasive sample collection method such as venipuncture may defer parents from participating in studies or conducting a regular follow-up on a symptom-free child. II.B.b.1 . Advantages and Shortcomings of hair zinc as a biomarker Hair as biopsy material offers many advantages such as ease of sample collection, handling and storage . In addition, the zinc in hair is much more concentrated than in other biological samples such as blood and urine . This higher concentration facilitates the analysis (Gibson, 1989) . 15  On the down side though, hair mineral content may be subject adventitiously to a variety of trace elements from environmental sources . However, it has been shown that washing procedures can overcome the effects of these exogenous materials and, if not to eliminate them completely, at least to minimize them (Bate, 1966) . Other environmental factors such as geography and season may also affect the hair zinc content . This creates problems in terms of interpreting and comparing the results of various studies that are done in different places or different seasons at the same location. A seasonal variation in hair zinc content has been reported by some Canadian scientists (Gibson et al, 1989-a) . Gibson and co-workers have studied groups of Canadian and Malawian children on a plant-based diet during 3 survey periods, and reported an inverse relationship between daytime light and hair zinc content . It appears that hair zinc has its highest levels in the seasons with the shortest daylight hours (December/January in Canada and July/August in Malawi) . It can be argued that the observed seasonal changes in hair zinc may originate from the seasonal changes in the rate of growth since seasonal changes in the growth of children are well documented (Marshall, 1975 and Marshall & Swan, 1971), and a positive relationship between hair zinc concentrations and growth has been demonstrated in some studies (Smit Vanderkooy & Gibson, 1989 and Hambidge et al, 1972). However, a similar trend in the hair zinc of American adults (Greger et al, 1978) and the absence of seasonal change in the serum copper content of Canadian children (Gibson et al, 1989-a) all indicate that the rate of general growth and hair growth are not the underlying reason for the seasonal changes observed in hair zinc concentration.  16  Based on the above-mentioned study Gibson and coworkers (1989-b) have proposed the use of 2 different cutoffs for children's hair zinc; 1 .68 µmol/g (110 µg/g) and 1 .07 µmol/g (70 µg/g) for seasons with shortest and longest daylight, respectively (Gibson et al, 1989-a) . It is important to note that the 70 µg/g cutoff appears to be universal. A search in pub med (http ://www.ncbi .nlm.nih.gov/sites/entrez) using the key words of "hair zinc", "zinc deficiency", and "children", reveals that except for this research group (Gibson and coworkers) no other researcher has used the winter cutoff. Geographic variations in the level of trace elements have been shown to exist. Gibson and DeWolfe (1979) have shown that hair zinc levels in full-term neonates from Cincinnati were comparable to those seen in similar subjects in Halifax, Nova Scotia, and the hair zinc levels of both of these groups of infants were higher than those found in comparable subjects from Denver (Hambidge et al, 1972) . Some have proposed that this difference arises from differences in the degree of environmental pollution (Zachwieja et al, 1995) . For instance, in Canada, a regional difference in the trace metal content of ground water has been documented (Gibson et al, 1987) . Whatever the underlying reason, it has been argued that the hair zinc of infants from 2 different geographic locations can only be compared if their identical hair growth rate has been established (Hambidge, 1980). The normal hair growth rate for infants has been reported at 0 .2 mm/day for scalp hair, which increases to 0 .3-0.5 mm/day during the post infancy period (Gibson, 1989). Accurate and well-designed studies addressing the geographic variations in older children do not, as yet, exist. Besides these environmental factors some individual characteristics such as hair color/pigmentation (Sky-Peck, 1990), sex (Smit Vanderkooy & Gibson 1987 and Esteban 17  et al, 1999), location of collected hair samples (Paschal et al, 1989) have been demonstrated to affect hair concentrations of several minerals including zinc. In general, sex is believed to affect hair metals due to the presence or absence of sex-linked hair treatment activities (e .g . coloring, permanent) (Esteban et al, 1999). While this may be part of the reason for the higher hair zinc of females, the higher physiological need of males for zinc (Castillo-Duran et al, 1994) should not be overlooked . A higher physiologic requirement may lead to the provision of most of the available zinc to the tissues having zinc-dependent activities and may result in a very small amount left to be channeled to the hair follicles, where the element (Zn) will stay in a storage system that is no longer available to the body . In any event, sex differences have been observed both in terms of hair zinc content and the occurrence of low hair zinc (Smit Vanderkooy & Gibson, 1987, Hambidge et al, 1980) . Nevertheless, whether these differences originate from endogenous reasons (higher needs for zinc) or exogenous ones (such as differences in grooming habits) they have not been properly investigated . In view of the fact that grooming habits are almost the same for both sexes when it comes to younger children, and also in light of the fact that sex differences reported in the 3 aforementioned studies (Smit Vanderkooy & Gibson, 1987, Gibson et al, 1989-b and Hambidge et al, 1976) were all in young children (<9 years old), the higher physiological needs may appear a more convincing explanation for the observed sex-based differences in hair zinc content. However, the existmce of few studies reporting a comparable hair zinc for the 2 sexes (Sakai, Warishi, and Nikiyama, 2000), or higher hair zinc for male children (Zachiwieja et al, 1995) may indicate the involvement of some other factors besides the physiology.  18  The amount of hair pigmentation has also been argued as a factor that may affect the hair zinc content . However, well-designed studies investigating the effect of pigmentation while controlling all other factors have yet to emerge . In a study of zinc deficiency among short Iranian teenagers, a post hoc analysis compared the hair zinc of these teenagers (all brunets) with their American counterparts (all blonds) and revealed no significant difference (Prasad, 1991) . This study, however, was a cross cultural comparison and other factors which may affect hair zinc content such as geography, ethnicity, and habitual diet, were not taken into account . At the organ level, however, Czechoslovakian investigators have documented a 5 times higher concentration of zinc in the pigment-containing organelle of hair, the melanosomes, to that of the hair shaft as a whole (Borovansky, Horckcko & Duchon, 1976). II.B.b.2 . Significance of low hair zinc in individuals and in a population The validity of hair zinc as a diagnostic tool of MZD in children as noted above has been demonstrated by the growth response of affected children to zinc supplementation (Gibson et al, 1989-a, Walravens et al, 1983 and Hambidge et al, 1972). In some of these studies, but not all, hair zinc has also increased in response to the supplementation . The discrepancies encountered may be due to variations in dose and duration of the zinc supplementation period and confounding effects of the season on hair zinc concentrations . Periods of 6 weeks or less are too short for a response in hair zinc (Gibson, 1989). Although the suitability and usefulness of hair zinc in the assessment of zinc status and detection of MZD in children of developed countries has been documented by zinc researchers (Gibson, 1989, Laker, 1982 and Van Wouwe & Van den Hamer 1985),  19  the value of hair zinc as an index of sub-optimal zinc status in adults is uncertain and controversial. In children, hair zinc correlates with many other indices of zinc deficiency, but not with serum zinc . The lack of relationship between the hair and serum zinc concentrations reported by studies of marginal zinc deficiency is expected and does not challenge the potential value of hair zinc in determination of zinc deficiency . In fact, it serves to underline that a low hair zinc value has a significance that is different from that of hypozincemia (Hambidge 1980), seen only in cases of severe zinc deficiency. While one needs to be very careful in interpreting results of hair zinc coming from isolated samples, an epidemiological use of samples obtained from a wide population may provide worthwhile information to the clinicians . At the population level, hair zinc seems to correlate with the intake of bioavailable zinc as shown in studies of zinc in populations with high dietary phytate intake (Gibson & Huddle, 1998), or those with little meat intake (Smit, Vanderkooy & Gibson, 1987) . Whether a change in dietary patterns brings a change in hair zinc in a given population or not, has not yet been investigated. All in all, it is believed that, if used in conjunction with other clinical and biochemical information, hair analyses may become a useful screening tool for determination of nutritional disorders including marginal zinc deficiency in individuals (Sky-Peck, 1990) . At the population level, analysis of recently grown hair for zinc content provides a biomarker of overall zinc status and could be useful in the assessment of the zinc status of groups of individuals and large populations (Contiero & Folin, 1994).  20  II.B.b.3 . Issues around hair sample collection Hair sample collection is easy and except for some rare occasions when some cultural taboos or individual beliefs militate against individuals having their hair touched by a stranger (Personal observation, March 23, 2004), the compliance for donating a sample is generally high . As much as the ease of sample collection has contributed to the popularity of hair as a biological sample for assessment of zinc status, the significance of a proper sample collection should not be overlooked . Aside from potential cultural barriers to sample collection, the lack of a standard protocol for sample collection can also be a cause of potential problems. At present the most reliable protocol appears to be one proposed by the International Atomic Energy Agency (IAEA) (Dippisch et al, 1999) and adopted by the Center of Disease Control (CDC) and the US Environmental Protection Agency (EPA). However, within the research setting, depending on the analytical methods used by the analyzing laboratories, wide variations of this protocol have been used. The 3 important factors in hair sample collection are 1) Sites of sample collection, 2) the size of the sample (mg) and 3) the length of the samples . Below is a brief review of these criteria and the possible modifications that they have been subjected to based on research resources and practicalities. 1 . The sites of sample collection : Interpretation of analytical data is dependent on the assumption of a normal rate of hair growth, approximately 1 cm/month . At any given time some scalp hairs are in a resting (telogen) phase (Hambidge, 1980) . Therefore, it is crucial to obtain a sample from various sites on the scalp . If analyses are limited to very few hair shafts, the chance selection of hairs in the resting phase may provide misleading 21  information . This criterion appears to be one that every research paper has provided assurance on in their sample collection procedure . The sample should be taken in small portions from at least 3-4 different locations . The recommended areas for collection are the nape (bottom) of the neck, the top portion of the neck, and the side of the scalp. 2. Sample size: Although the protocol adopted by the CDC defines an acceptable sample size as one weighing 500-1000 mg, in practice the sample size used does not always fall within this range . At present the sample size reported in the zinc literature varies within a larger range of 20-1000 mg . Part of this variation can come from the analytical techniques used for a given study . For instance, a study conducted to compare the recovery rate and detection limit using the spectrofluorophotometric method and atomic absorption spectrophotometry method revealed that while both can detect greater than 90% of the added zinc, the former has a lower detection limit and can be used confidently with much smaller samples (Kazi et al, 2001). 3. Length of the sample : The length of the collected hair should be the 2-3cm closest to the scalp which contains the newest growth . As the hair ages and grows away from the scalp the protein structures can "unwrap" and elements from external sources can be bound to the hair structure (Mertz, 1975) . Assuming a normal rate of hair growth, the zinc level in the proximal 1-2 cm of hair reflect the zinc uptake by the follicles during the period 4-8 weeks prior to the sample collection (Gibson, 1989) . It has been shown that the zinc content shows wide variations throughout the hair strand (Sky-Peck, 1990) . The protocol adopted by CDC recommends 2-3 cm from the scalp as the preferred sample length, the remainder being discarded .  22  I.2.II.C . Population health approach to MZD As noted, diagnosis of marginal zinc deficiency is hampered by the lack of a single, specific and sensitive biochemical index of zinc status . A large number of indices have been proposed, but many have problems affecting their use and interpretation . The most reliable method for diagnosing marginal zinc deficiency at present is a positive response to zinc supplementation . But such an approach is time consuming and cumbersome . Above all, the likelihood of overloading some individuals at the higher end of zinc intake is a concern that should be given serious thought . Potential hazards resulting from excessive intake of zinc makes the general supplementation of a population an impractical solution. In a large-scale zinc supplementation program a concern for children who are taking adequate zinc is therefore a valid concern . The Tolerable Upper Intake Level (UL) of zinc (combined food and supplements) has been set by the Food and Nutrition Board of the Institute of Medicine at 7mg/day and 12 mg/day for preschoolers of 1-3 and 4-8 years of age, respectively (Institute of Medicine, 2000) . Due to the lack of data from children, the UL for children were extrapolated from adults . The UL has been defined as the highest level of daily nutrient intake that is likely to pose no risk of adverse health effects for almost all individuals . It is advisable not to exceed this level, even though exceeding it would not be harmful for all children. In recent years some researchers have challenged this upper level . Analyses of data from the 4 th Total Diet Study (TDS) of Health Canada, conducted from 1992-1999 in 8 cities : Vancouver, Whitehorse, Calgary, Winnipeg, Ottawa, Toronto, Montreal, and Halifax, has revealed that for infants and young children 2 months to 3 years of age,  23  mean zinc intake estimates exceeded the zinc UL, which is based on the adverse effect of zinc on copper metabolism (i .e., reduced copper status) (Cockell et al, 2003) . Analyses of the dietary intake data obtained during the 1994-1996 and 1998 Continuing Survey of Food Intakes by Individuals (CSFII) have also documented an intake much higher than the recommended levels among most US preschoolers . The intake of many of these children has been estimated to be larger than their respective UL (Arsenault & Brown, 2003). Perhaps due to the bioavailability issue these high levels of dietary zinc intake do not seem to have posed a health problem for these children . However, if zinc intake continues to be at the higher end on a regular basis and through supplementation, the amount of zinc consumed by children may become excessive and subject the child to excessive intake risks. While there have been no recent reports of zinc toxicity in Canadian or US children, and it is unlikely that zinc intake from food is high enough to have a negative effect on health status, this is not the case for supplemented zinc . Until such a time that the controversy around the new Dietary Reference Intake, DRI, of zinc is addressed and resolved in a scientific manner, it is only reasonable to comply with the current guidelines and avoid routine excess intake. The adverse effect of the excess intake of zinc on the bio avaiability and homeostasis of other essential trace elements has been documented . It has been shown to affect adversely the balance of copper and iron . Studies have reported copper deficiency anemia and decreased serum copper and ceruloplasmain (Parasad et al, 1978), decreased erythrocyte copper-zinc superoxide (Fisher et al, 1975) and increased copper excretion (Festa et al, 1985) as undesired side effects of increased zinc intake on copper 24  metabolism . Zinc sulfate supplementation studies of women with low iron stores have revealed amelioration in zinc indices while inducing a cellular iron deficiency and further exacerbation of iron status (Donangel, Woodhouse, & King, 2002) . Alterations in lipoprotein metabolism have also been reported with excessive zinc intake (Chandra, 1984) . The usefulness of general supplementation of a population becomes particularly questionable when the gravity of some of the potential adverse effects associated with a chronic intake of excess zinc, such as suppression of the immune response and reduction of HDL cholesterol (Chandra, 1984), are considered . This line of reasoning is not meant to discredit the therapeutic value of supplementation, but to point to the importance of identifying the right subset of the population for this therapeutic approach . Identifying the sector of population eligible for supplementation will not only reduce the cost of the program, but will also circumvent the unwanted ramifications of large-scale supplementation which would include children who are at the higher end of zinc intake. Therefore, when it comes to populations, the need to be able to identify the subset of populations that are at risk appears to be a rational pre-requisite to any widespread program such as supplementation. To identify the at-risk sectors of population the first and most important step is to identify the risk or predisposing factors of MZD . These factors when compiled can be used to construct a questionnaire that can serve as a screening tool to detect at-risk individuals and to detect the deficiency early on . Individual studies of zinc deficiency have documented associations between various socio-demographic and/or dietary factors  25  and zinc deficiency/status in general . The following sections are a brief review of these factors. 1.2.111 . Factors associated with zinc deficiency Studies of zinc deficiency have documented associations between zinc deficiency and various dietary and non-dietary factors . These are discussed under 3 subheadings: 1) Dietary factors associated with zinc deficiency/status 2) Socio-demographic factors associated with zinc deficiency/status 3) Other factors associated with zinc deficiency/status I.2.II1.1. Dietary factors associated with zinc deficiency After a meal, digestive processes release zinc from the food components . This zinc, along with endogenously secreted zinc, is present within the intraluminal compartment as free zinc . As this free zinc moves, it forms complexes with ligands and with other food components . This process of complex formation can either enhance or hinder the process of absorption . Below is a brief overview of these food components and the nature of their effect on overall zinc absorption. Protein :  The amount of protein in a meal is positively correlated with zinc absorption. When compiling results from several studies with humans to whom various protein sources and, amounts had been administered, fractional zinc absorption increased in linear fashion with increasing protein content (Sandstrom & Cederblad 1980) . The type of protein in a meal will also affect zinc bioavailability . Animal protein has been shown to counteract the inhibitory effect of phytate on zinc absorption from single meals  26  (Sandstrom & Cederblad 1980) . It has been speculated that this may be due to amino acids released from the protein that keep the zinc in solution and not a unique effect of animal protein per se . Either way the evidence supports the view that the presence of animal proteins affects the zinc status of humans positively . Study of the dietary intake of Canadian preschool children revealed an association between the lower consumption of animal proteins and the existence of suboptimal zinc status (Smit Vanderkooy & Gibson, 1987). Phytate :  The finding of zinc deficiency in human subjects on a high phytate diet in the Middle East suggested that phytate could adversely affect zinc status in humans (Halsted et al, 1972) . Inositol hexaphosphates (I6P) or phytate and inositolpentaphosphates (I5P) bind zinc and reduce its absorption (Lonnerdal et al, 1989) . Phytate is found in varying amounts in plant products, with grains and legumes having especially high levels . The phosphate groups in inositol hexaphosphate can form strong and insoluble complexes with cations such as zinc, and because the gastrointestinal tract of higher species lacks any significant phytase activity, phytate-bound minerals will be excreted in the stool. Among the inhibitors of zinc absorption, phytate has been recognized as the most potent inhibitor and has been studied extensively . A direct association between the amount of phytate in the diet and occurrence of zinc deficiency has been documented in many populations (Prasad, 1991). Various food-processing procedures such as soaking, fermenting and germinating, have been shown to reduce the phytate content of the food (Gibson et al, 2003) . Among  27  them, perhaps due to the large consumption of bread in many societies, fermentation has attracted a great deal of attention. Phytate levels decrease during baking . It is possible to control the extent of this decrease and thus create bread with higher mineral bioavaility . Normally, to reduce the phytate content of the bread to a level considered not to mal-affect mineral absorption, without adding phytase, it would be necessary to increase the activity of the naturally occurring phytase in the flour or that in the yeast. During the traditional transformation of flour into bread, reduction in phytate content as a consequence of the activity of native phytase usually does not greatly improve mineral bioavailability. Although during traditional bread making the phytate (IP6) content of the bread is decreased considerably, inisitol pentophosphate, IP5 will still be present in the media. It has been shown that IP6 and IP5 are the strongest inhibitors of zinc absorption (Sandstorm & Sandberg, 1992) and need to be removed almost completely in order to increase absorption of this mineral . In addition, although IP4 and IP3 in isolation are shown to have no inhibitory effect on zinc absorption (Sandberg et al, 1999), when present in a mixture of inisitol phosphates, IP3 and IP4 may also contribute to a negative effect on bioavaialaility of minerals. This effect has been clearly shown for iron (Brune et al, 1992) . This is accomplished through interactions with the more phosphorylated inositol phosphates (Sandberg et al, 1999) . This is probably also true for zinc absorption as a strong negative correlation was found between zinc absorption and the sum of native IP3 through IP6 from cereal meals (Sandberg, 1991). As for the reduction of phytate due to the phytase activity of the leavening agent, various bread-making procedures using different leavening agents have been designed to 28  improve the nutritive value of the bread . For instance, sourdough is a symbiotic culture of lactobacilli and yeasts used to leaven bread . Studies suggest that sourdough fermentation is the best way to hydrolyze phytate and to improve zinc assimilation . Sourdough bread contained only about 30% and 61% of the phytate content of the unprocessed flour and yeast bread, respectively . The traditional bread-making agents are more stable in the physiological conditions of the intestine (Fredrikson et al, 2002) and as a result more resistant to undergoing further chemical reactions. Fiber :  Fiber is often implied as having a negative effect on zinc absorption . However, this is usually due to the fact that most fiber-containing foods also contain phytate. Studies have shown that experimental reduction of the phytate content of a high fiber bread considerably increased zinc absorption to a degree similar to that from white bread (low fiber), suggesting that fiber in itself has little or no effect on zinc absorption (Navert' Sandstrom & Cederblad, 1985) . In agreement with this finding, studies on isolated fiber components such as cellulose have shown no significant inhibitory effect on zinc absorption (Turnlund et al, 1984). Calcium: The inhibitory effect of calcium on zinc absorption is not that straightforward. Some argue that it is unlikely that calcium per se has a negative effect on zinc absorption (Lonnerdal et al, 1984) . Studies on long-term use of calcium supplements have shown no adverse effect from calcium on zinc status ; Gambian women who were given 1000 mg calcium/d had plasma zinc concentrations similar to those of unsupplemented women (Yan et al, 1996) . 29  Calcium potentiates the complex formation of zinc with phytate (Oberleas, Muhrer and O'Dell, 1966) and hinders its absorption through this avenue . It has been suggested that the ratio formula [Ca] x ([phytate]/[Zn]) can be used as a predictor of zinc bioavailability (Gibson, Smit Vanderkooy, & Thompson, 1991) . In rodents the phytate zinc ratio of <6 was found to be optimum for the growth of rats fed a diet containing 1 .6 percent calcium . However, this ratio had to be lowered if the dietary calcium increased (Oberleas, Muhrer, and O'Dell, 1966) . Study of the dietary phytate content of Canadian preschoolers showed that dietary [Ca x (phytate/Zn)] millimolar ratios, expressed per 1000 Kcal, correlated with the zinc nutriture of the children . Children with hair zinc below the cutoff had diets with a significantly higher ratio of [Ca x (phytate/Zn)], than children who had hair zinc above the cutoff value (Gibson, Smit Vanderkooy, & Thompson, 1991). I.2. III.2. Socio-demographic factors associated with zinc deficiency/status Family income:  In general the health-wealth relationship is a well-established and consistent phenomenon. Differences in per capita income (a measure of wealth) appear to be closely and positively related to the length and healthfulness of life . The pattern is not the simple difference between the healthy rich and unhealthy poor. It is present in a "gradient" manner and applies to each and every class within the population i .e. the health status of each class within the population appears to be better than the classes below and worse than the classes above (Hertzman, 1998) . This unique and remarkably persistent influence of wealth on health is played through various avenues, nutrition being one of them.  30  Undernutrition or malnutrition often occurs in the context of poverty and predisposes the individual to a wide array of health problems . For instance, the best sources of dietary zinc (i .e. animal proteins) are the most expensive food items ; therefore, those subsisting on low-income diets are likely to lack or have small amounts of these rich sources of zinc in their diet and as a result be at higher risk of developing marginal zinc status (Sandstead, 1973) . Children with a physiologic need to have a positive zinc balance are more vulnerable in this sense . Studies have shown significantly lower (yet normal) serum and hair zinc means for children of low-income families than those of middle income in the U .S. (Hambidge et al, 1976) and France (Chakar et al, 1993). Associations between poverty and low serum zinc have also been reported among the participants of NHANES II (Pilch & Senti, 1985). sex  Sex differences have also been discussed in the zinc literature . In general, starting from early childhood, males have been reported to have higher mean serum zinc levels than females . The difference in serum zinc between sexes grows as the child does and peaks at 20-44 years. Despite this higher serum zinc, when it comes to MZD, evidence suggests that boys may be at higher risk than the girls of the same age . (Walravens et al, 1983 and Smit Vanderkooy & Gibson 1987) . The Canadian study of zinc nutriture in preschoolers of Ontario reported a significantly higher rate of occurrence of low hair zinc (< 70µg/g) for boys (21% and 5% for boys and girls, respectively, P=0 .00) (Smit Vander kooy & Gibson, 1987). Sex has also been reported as an important factor affecting the hair zinc content of both children and adults (Creason et al, 1975, Gibson & DeWolfe, 1980, and Smit 31  Vanderkooy & Gibson, 1987) . A higher physiological requirement in males has often been speculated as the underlying reason for their lower hair zinc . Factors such as lower absorption of dietary zinc in males (Turk, 1977) and their higher obligatory losses (Hunt et al, 1992 and Lee et al 1993) have been presented as the underlying reasons for this higher need . Although available data for adults adequately support a higher physiological requirement for males, the relevant data for children is missing . For this reason the physiological requirement of male and female preschoolers (computed by extrapolation from infants) have been estimated the same for male and female preschoolers (Institute of Medicine, 2002). Age :  The preschool period comprises a large portion of childhood . Younger preschoolers (24-36 months) may still have not recovered from the unfavorable condition of in-utero period (such as mother's zinc deficiency during pregnancy), at birth (such as prematurity) or infancy (such as being breast-fed by a mother with low zinc status) . Age differences in the hair zinc level of preschoolers (younger preschoolers having lower hair zinc level than the older ones) have been shown in different studies (Klevay, 1970, Hambidge et al, 1972), however, it is not yet clear whether this is a normal physiological decline similar to what is experienced in the hemoglobin level during post infancy (Irwin and Kitchener, 2001), or originating from environmental factors such as changes in nutrition. Maternal education:  Maternal education may also affect the nutriture of zinc . In general, increased maternal education can positively affect the food choices made, and methods of food 32  preparation and preservation positively . In addition, rising maternal education could elevate the economic status and as a result the buying power of the family . It is important to note that this positive correlation between family income and maternal education may not be true in certain population pockets . For instance, in societies such as Canada where the former education of many immigrants is not recognized, higher education and training does not necessarily translate into a better paying job . In such populations the effects of maternal education on a child's health is likely to be distorted. Ethnicity:  Analysis of NHANES II has revealed lower serum zinc for African Americans compared to Caucasians (Pilch & Senti, 1985) . Although the difference was not great, it can be indicative of a higher risk at the population level and further nutritional assessment of this group in respect to its zinc nutriture would be advisable. I .2 .III .3 . Other factors associated with zinc deficiency/status Iron supplementation  At levels of essential micronutrients present in foods, most micronutrients appear to use specific absorptive mechanisms and they don't seem to be at risk of being malaffected by the interaction of other micronutrients . However, in aqueous solutions and/or at higher intake levels, competition between elements with similar chemical characteristics and uptake by non-regulated processes can take place. Elements with similar physiochemical properties may interact with each other at the level of absorption . Iron and zinc share a part of their absorptive pathway. Iron absorption involves two carriers, one involved in the uptake of iron across the microvillus membrane of the enterocyte and the other in its transfer to the plasma at the basolateral 33  surface. The uptake phase is thought to involve divalent metal transporter-1 (DMT1). This transporter is a carrier for zinc as well . Therefore, at the point of transfer from brush border into the enterocytes, if concentration of anyone of these two metals increased the absorption of the other will be affected in a negative manner as a result of the unavailability of the carrier, DMT1. These interactions have clearly been demonstrated in experimental absorption studies (Pedrosa & Cozzolino, 1993, and Isfaoun, 1997) and to some extent have been confirmed in supplementation studies (Hambidge et al, 1987) . Some investigators (Hambidge et al, 1987) have reported a decline in serum zinc during the course of iron therapy . Although most studies have demonstrated iron and zinc interactions with amounts of supplemental iron over 30 mg/d, multivitamins containing as little iron as 18 mg/d have also been found to lower plasma zinc concentrations in pregnant teens (Dawson et al. 1989). Furthermore, alterations in plasma zinc concentrations during pregnancy may be evident following only 1 week of iron supplementation (20 mg/d) (Hambidge et al . 1987) . Studies of zinc status in pregnant women immediately before iron therapy and at 1 and 4 weeks thereafter indicated a decline in plasma zinc from baseline in just 1 week . The decline remained statistically significant even after adjustment for the expected physiologic decline over the same interval of gestation . Similar effects of iron therapy on zinc status (serum zinc) have been confirmed by some (Bloxham et al, 1989 and O'Brien et al, 2000) but not all (Yip et al, 1985) investigators. Early childhood experiences  Early experiences affecting zinc balance such as prematurity and being born from a zinc deficient mother may also affect the zinc status of young children . Some studies 34  suggest that prematurely born infants are at higher risk for inadequate intake of dietary zinc (Friel et al, 1985) . Premature infants are not only born with a negative zinc balance, but also, due to a susceptibility to after-birth infections, have a higher zinc demand . Such children starting life with a negative zinc balance and having a high zinc demand due to the high rate of growth (a highly zinc-dependent physiologic event) may not have a chance to correct the deficit and may continue carrying it to the later stages of life. Breast-feeding:  Breast-feeding and its duration have been shown to affect children's zinc status. The bioavailability of zinc from human milk is significantly higher than that of cow milk (Lonnerdale, Keen &Hurley, 1981) and soy-based formulas (Sandstrom, Cederblad & Lonnerdal, 1983), due partly to the presence of zinc-binding ligands facilitating zinc absorption (Duncan & Hurley, 1978) . Some studies (Walravens et al, 1992), including a Canadian study (MacDonald, Gibson & Miles, 1982), have reported the most favorable zinc status in breast-fed compared to formula-fed infants, even those fed with zincsupplemented formula. However, there is an unusually sharp physiological drop in the zinc content of human milk after birth (Krebs & Hambidge, 1986) making this food unable to provide an infant with adequate zinc beyond 6 months . Growth-limiting zinc deficiency has been shown in infants principally fed with human milk even after the age of 4 months (Walravens et al," 1992). Erratic eating behaviors:  Erratic appetite and seemingly problematic eating behaviors such as being a picky eater, and lack of variety in the diet are some of the features common to young children. 35  Being a picky eater is part of what it means to be a young child and particularly a toddler. They may eat only fruits one day, and vegetables the next, and peanut butter and jelly sandwiches for breakfast, lunch and dinner for few days at a time . These behaviors can easily affect the quantity and quality of the food consumed by children . In fact it is believed that they may serve as a means for adjusting the food intake to meet the relatively decreased demand during this stage of life . Preschoolers gain weight more slowly compared to the prior stage of life, so they need relatively less food, and perhaps nature gets this accomplished through these eating behaviors (Sears, 2006) . This appears to be problematic, and often it really isn't . As expected and often transient as these behaviors are, they may result in nutritional deficiencies when they become chronic . Lack of variety in the diet and elimination of some foods or food groups can adversely affect the micronutrient balance in general, with some nutrients being at higher risk than the others . Zinc with its high requirement during this stage, is an example of such a nutrient. It is possible for zinc balance to be disrupted if these behaviors persist for a prolonged period of time. 1.3. The present knowledge gap After recognition of MZD in 1970's, 2 studies addressed this deficiency among Canadian preschoolers . The first study (Smit Vanderkooy & Gibson, 1987) documented the existence of potential MZD, while the second (Gibson et al, 1989-b) confirmed its existence among a group of preschoolers followi :ig a zinc supplementation study . Since then, there has never been another report of zinc status in Canadian preschoolers. Despite the lack of any further work with this age group in Canada, combined evidence from the children of other industrial societies indicate that MZD is common  36  among the children of developed societies (Hambidge et al, 1972, Walravens et al, 1983, Nakamura et al, 1993, Buzina et al, 1980 and Chakar et al, 1993) . Many factors may contribute to this high prevalence ranging from human physiology (such as the sensitivity of the human body to the shortage of this element and the inefficiency of the human homeostasis for zinc balance) to environmental (such as the changes in the intake of certain food groups, such as red meat) to the inadequacy of the diagnostic tools. The inadequacy of diagnostic tools is perhaps one of the main contributors to the high prevalence of MZD in developed countries . Serum zinc, the most commonly used biomarker of zinc status, is not capable of diagnosing the deficiency during its very early stages when the deficiency is only marginal . At present the only certain way of diagnosing MZD, in research as well as among practitioners, is by the growth response observed in the child following a period of zinc supplementation. On the other hand, studies have shown a consistent decrease in hair zinc of children in whom MZD has later been confirmed through their positive growth response to a course of zinc supplementation (Hambidge et al, 1976, Walravens et al, 1983, and Gibson et al, 1989) . However, hair zinc is not used or recognized by practitioners . Under these circumstances, an awareness of the risk factors of the deficiency and factors associated with it can be invaluable . These factors when compiled can be used to construct the profile of at-risk individuals and may, therefore, have some usefulness as a screening tool. This study was designed to fill the gaps in our knowledge about the present zinc status, (as indicated by hair zinc) of healthy free-living Canadian preschoolers . It also aimed to construct and validate a screening tool for low hair zinc/MZD among children  37  by identifying and compiling a list of some dietary and social factors associated with hair zinc. 1.4. Purpose of the study This was an exploratory study and its chief purpose was to examine the hair zinc status of preschoolers of Vancouver, and to construct and validate a short and easy-toadminister screening questionnaire that could be used as an upstream preventative measure to predict low hair zinc/MZD among preschoolers . In doing so, we planned to investigate both the dietary and socio-demographic population patterns of hair zinc for this age group and to understand better the eating patterns, social settings and factors in family environments that associate with the hair zinc level of the preschoolers. L5. Objectives 1.5.1 . Objectives for chapter 11 In a group of conveniently sampled low-income preschoolers from 4 inner-city neighborhoods: 1) To determine mean hair zinc levels and the prevalence of low hair zinc. 2) To explore the relationship between the 5 main socioeconomic and demographic factors (age, sex, ethnicity, family income and maternal education) and hair zinc levels, both individually as well as collectively. 3) To investigate the relationship of consumption of some food groups ("dairy", "meat, fish and poultry" and "cereals and grains") as well as some nutrition, health related variables (as perceived by the caregivers) with hair zinc levels, both univariately and after adjustment for demographic and socio-economic factors. 4) To assess the collective contribution of all of the above-mentioned factors (factors 38  significantly associated with hair zinc) for explaining the variability in hair zinc. 1.5.11. Objectives for chapter III In a large representative sample of preschoolers from the city of Vancouver; 1. To assess the hair zinc status and the extent of low hair zinc in Vancouver preschoolers through a large and representative sample from the city. 2. To explore the relationship between the 5 main demographic and socioeconomic factors (age, sex, ethnicity, family income and maternal education) and hair zinc levels. 3. To examine the relationship between some additional family characteristics (family size, composition, food security and paternal education) and hair zinc levels. 4. To investigate the relationship between some eating behaviors such as the intake frequency of meat, dairy, cereals and grains, as well as variables such as "being concerned about child's eating", "adequacy of child's eating", "pickiness", and "unhealthy eating" with the hair zinc level of children. 5. To examine the relationship of parental perceptions of some health-related variables (such as overall health, frequency of a child being sick, breast-feeding pattern, and intake of supplements containing iron), and behavior related variables (such as activity level and attentional focusing scores) with the hair zinc of children. 6. To examine the relationship between the indices of growth (height-for-age Z scores and weight-for-height Z scores) and nutrition (weight-for-age Z scores) with the hair zinc of children. 7. To identify which of the variables significantly associated with hair zinc can predict the hair zinc levels. 39  8. To find out which of the variables significantly associated with low hair zinc status can predict hair zinc status (i.e. low hair zinc). 9. To assess the collective usefulness of the variables identified under objective 8 and the constructed questionnaire based on this model, for predicting the hair zinc status of children by exploring the classification tables and the Receiver Operating Characteristic (ROC) curve of the predictive model .  40  I .6. Bibliography Arsenault JE, Brown KH . Zinc intake of US preschool children exceeds new dietary reference intakes. Am J Clin Nutr 2003 ; 78: 1011-7.  Assarian GS, Oberleas D . Effect of washing procedures on trace element content of hair . Clin Chem 1977; 23 : 1771-2.  Bahl L, Chaudhuri LS, Pathak RM . Study of serum zinc in neonates and their mothers in Shimla hills (Himachal Pradesh) . Indian J Pediatr 1994 ; 61 : 571-5.  Bate LC. Adsorption and elution of trace elements on human hair. Int J Appl Radiat Isot 1966 ; 17 : 41723.  BC Ministry of Health Services . British Columbia Nutrition Survey, BCNS. Report on food groups. 2004. Retrieved May 22, 2007 from http://www.hlth.gov.bc.ca/preventlnutrition/pdf/fgreport.pdf  Bloxam DL, Williams NR, Waskett RJ, Pattinson-Green PM, Moraji Y, Stewart SG . Maternal zinc during oral iron supplementation in pregnancy . Clin Sci 1989 ; 76: 59-65.  Borovansky J, Horcicko J, Duchon J . The hair melanosome : another tissue reservoir of zinc. Physiol Bohemoslov 1976; 25 : 87-91.  Brown KH, Peerson JM, Allen LH. Effect of zinc supplementation on children's growth : a metaanalysis of intervention trials. Bibl Nutr Dieta 1998; 54: 76-83.  Brownell M, Mayer T, Martens PJ, Kozyrskyj A, Fergusson P, Bodnarchuk J, Derksen S, Friesen D, Walld R. Using a population-based health information system to study child health . Can J Public Health 2002 ; 93 : S9-14.  Brune M, Rossander-Hulten L, Hallberg L, Gleerup A, Sandberg AS . Iron absorption from bread in humans : inhibiting effects of cereal fiber, phytate and inositol phosphates with different numbers of phosphate groups. J Nutr 1992 ; 122 : 442-9. 41  Castillo-Duran C, Garcia H, Vengas P, Torre L . Zinc supplementation. Acta Paediatr 1994 ; 83 : 833-7.  Center for Disease Control . CDC growth charts. 2000 . Retrieved October 11, 2004 from http://www. cdc . gov/nchs/about/maj or/nhanes/growthcharts/datafiles .htm  Chakar A, Mokni R, Walravens PA, Chappuis P, Bleiberg-Daniel F, Mahu JL, Lemonnier D. Plasma zinc and copper in Paris area preschool children with growth impairment. Biol Trace Elem Res 1993; 38 : 97-106.  Chandra RK . Excessive intake of zinc impairs immune responses . JAMA 1984 ; 252:1443-6. Cockell KA, Xu-Liang, Cao XL, McKenzie A, Pepper K, Dabeka R . (2003) . Food Directorate . Health Canada. (Document Reproduction Service No. K1A OL2).  Contiero E, Folin M . Trace elements nutritional status . Use of hair as a diagnostic tool . Biol Trace Elem Res 1994 ; 40 : 151-60.  Creason JP, Hinners TA, Bumgarner JE, Pinkerton C . Trace elements in hair as related to exposure in metropolitan New York USA . Clin Chem 1975 ; 21 : 603 Dawson EB, Albers J, McGanity WJ . Serum zinc changes due to iron supplementation in teen-age pregnancy . Am J Clin Nutr 1989 ; 50: 848-52.  Deeming SB, Weber CW . Evaluation of hair analysis for determination of zinc status using rats . Am J Clin Nutr 1977 ; 30: 2047-52.  Deppisch LM, Centeno JA, Gemmel DJ, Tones NL . Andrew Jackson's exposure to mercury and lead: poisoned president? JAMA 1999 ; 282 : 569-71.  Donangelo CM, Woodhouse LR, King SM, Viteri FE, King JC . Supplemental zinc lowers measures of iron status in young women with low iron reserves . J Nutr 2002 ; 132 : 1860-4 . 42  Donovan UM, Gibson RS . Iron and zinc status of young women aged 14 to 19 years consuming vegetarian and omnivorous diets . J Am Coll Nutr 1995 ; 14 : 463-72.  Dorea JG, Paine PA . Making sense of the hair zinc literature : where do we go from here? Arch Latino Am Nutr 1988 ; 38 : 93-112.  Duncan JR, Hurley LS . Intestinal absorption of zinc : a role for a zinc-binding ligand in milk. Am J Physiol 1978 ; 235 : E556-9. Esteban E, Rubin CH, Jones RL, Noonan G . Hair and blood as substrates for screening children for lead poisoning . Arch Environ Health 1999 ; 54: 436-40. Festa MD, Anderson HL, Dowdy RP, Ellersieck MR . Effect of zinc intake on copper excretion and retention in men . Am J Clin Nutr 1985 ; 41 : 285-92.  Fisher GL . Function and homeostasis of copper and zinc in mammals . Sci Total Environ 1975 ; 4: 373412.  Fredrikson M, Andlid T, Haikara A, Sandberg AS . Phytate degradation by microorganisms in synthetic media and pea flour. J Appl Microbiol 2002 ; 93 : 197-204.  Friel JK, Gibson RS, Kawash GF, Watts J . Dietary zinc intake and growth during infancy . J Pediatr Gastroenterol Nutr 1985 ; 4 : 746-51.  Gibson RS, DeWolfe MS . The zinc, copper, manganese, vanadium, and iodine content of hair from 38 Canadian neonates. Pediatr Res 1979 ; 13 : 959-62.  Gibson RS, Ferguson EF, Vanderkooy PD, MacDonald AC . Seasonal variations in hair zinc concentrations in Canadian and African children. Sci Total Environ 1989-a ; 84: 291-8.  43  Gibson RS, Heath AL, Ferguson EL . Risk of suboptimal iron and zinc nutriture among adolescent girls in Australia and New Zealand : causes, consequences, and solutions . Asia Pac J Clin Nutr 2002 ; 11: S543-52.  Gibson RS, Huddle JM . Suboptimal zinc status in pregnant Malawian women : its association with low intakes of poorly available zinc, frequent reproductive cycling, and malaria . Am J Clin Nutr 1998 ; 67: 702-9.  Gibson RS, Smit Vanderkooy PD, Thompson L . Dietary phytate x calcium/zinc millimolar ratios and zinc nutriture in some Ontario preschool children . Biol Trace Elem Res 1991 ; 30: 87-94.  Gibson RS, Vanderkooy PD, MacDonald AC, Goldman A, Ryan BA, Berry MA . Growth-limiting, mild zinc-deficiency syndrome in some southern Ontario boys with low height percentiles . Am J Clin Nutr 1989-b ; 49 : 1266-73.  Gibson RS, Vanderkooy PS, McLennan CE, Mercer NM . Contribution of tap water to mineral intakes of Canadian preschool children . Arch Environ Health 1987 ; 42: 165-9. Gibson RS . Assessment of trace element status in humans. Prog Food Nutr Sci 1989 ; 13 : 67-111. Gibson RS, Yeudall F, Drost N, Mtitimuni BM, Cullinan TR . Experiences of a community-based dietary intervention to enhance micronutrient adequacy of diets low in animal source foods and high in phytate: a case study in rural Malawian children . J Nutr 2003 ; 133 : 3992S-9S. Government of Canada . New perspective on the Health of Canadians, 1974 . Retrieved July 12, 2005 from http://www.phac-aspc. gc . ca/ph-sp/phdd/pdf/perspective .pdf.  Greger JL, Higgins MM, Abernathy RP, Kirksey A, DeCorso MB, Baligar P . Nutritional status of adolescent girls in regard to zinc, copper, and iron . Am J Clin Nutr 1978 ; 31 : 269-75 .  44  Guillard 0, Piriou A, Gombert J, Reiss D . Diurnal . variations of zinc, copper and magnesium in the serum of normal fasting adults. Biomedicine 1979 ; 31 : 193-4.  Halsted JA, Ronaghy HA, Abadi P, Haghshenass M, Amirhakemi GH, Barakat RM, Reinhold JG . Zinc deficiency in man . The Shiraz experiment . Am J Med 1972 ; 53 : 277-84. Hambidge KM, Hambidge C, Jacobs M, Baum JD . Low levels of zinc in hair, anorexia, poor growth, and hypogeusia in children . Pediatr Res 1972 ; 6 : 868-74. Hambidge KM, Krebs NF, Sibley L . Acute effects of iron therapy on zinc status during pregnancy. Obstet Gynecol 1987 ; 70 : 593-6.  Hambidge KM, Walravens PA, Brown RM, Webster J, White S, Anthony M, Roth ML . Zinc nutrition of preschool children in the Denver Head Start program . Am J Clin Nutr 1976 ; 29 : 734-8. Hambidge M, Krebs NF . Interrelationships of key variables of human zinc homeostasis : relevance to dietary zinc requirements . Annu Rev Nutr 2001 ; 21 : 429-52. Hambidge M, Krebs N . Assessment of zinc status in man . Indian J Pediatr 1995 ; 62: 169-180. Hertzman C . Making Early Childhood Development a Priority : Lessons from Vancouver Canadian Center for Policy Alternatives, 2004 . Retrieved March 22, 2005 from http://www . earlylearning .ubc. ca/documents/Making%20ECD%20a%20Priority .pdf Hertzman C . The case for child development as a determinant of health . Can J Public Health 1998 ; 89: S 14-9, S16-21.  Hopps HC . The biologic bases for using hair and nail for analyses of trace elements . Sci Total Environ 1977 ; 7 : 71-89.  Hotz C, Peerson JM, Brown KH . Suggested lower cutoffs of serum zinc concentrations for assessing 45  zinc status : reanalysis of the second National Health and Nutrition Examination Survey data (19761980). Am J Clin Nutr 2003 ; 78 : 756-64.  Human Early Learning Partnership . BC Child Atlas . 2006. Retrieved May 25, 2007 from http://www.help.ubc .ca/bc-atlas-child-development .htm. Hunt CD, Johnson PE, Herbel J, Mullen LK . Effects of dietary zinc depletion on seminal volume and zinc loss, serum testosterone concentration, and sperm morphology in young men . Am J Clin Nutr 1992; 56 : 148-57.  Hunt JR, Matthys LA, Johnson LK . Zinc absorption, mineral balance, and blood lipids in women consuming controlled lacto ovo-vegetarian and omnivorous diets for 8 weeks . Am J Clin Nutr 1998; 67: 421-30. Institute of Medicine . Dietary Reference Intakes for Vitamin A,Vitamin K, Arsenic, Boron, Chromium, Copper, Iodine, Iron, Manganese, Molybdenum, Nickel, Silicon, Vanadium, and Zinc . Washington DC: National Academy Press . 2000. International Zinc Association, IZA . Zinc and Human Health . 2000 . Retrieved July 27, 2005 from http ://www.zinc-health .org/Health/Conf-22 .cfm  Irwin JJ, Kitchener JT . Anemia in children. Am Fam Physician 2001 ; 64 : 1379-86. Isfaoun A, Bureau F, Mouly-Boudey M, Drosdowsky M, Arhan P, Bougle D . Relationships between iron and zinc metabolism : predictive value of digestive absorption on tissue storage . J Trace Elem Med Biol 1997 ; 11 :23-7. Kazi . S, Ali SS, Jakharani MA, Kazi TG, Kazi GH . Spectro-fluorophotometric study of zinc level in biological samples and its comparison by atomic absorption spectrophotometry . Sci 2001 ; 2 : 78-80. Kindig D, Stoddart G . What is population health? Am J Public Health 2003 ; 93 : 380-3 . 46  King JC, Hambidge KM, Westcott JL, Kern DL, Marshall G . Daily variation in plasma zinc concentrations in women fed meals at six-hour intervals . J Nutr 1994 ; 124 : 508-516.  King JC. Assessment of zinc status . J Nutr 1990 ; 11 : 1474-9. Klevay LM . Hair as a biopsy material. I. Assessment of zinc nutrition . Nutr 1970 ; 23 : 284-9. Krebs NF, Hambidge KM . Zinc requirements and zinc intakes of breast-fed infants. Am J Clin Nutr 1986 ; 43 : 288-92. Laker M . On determining trace element levels in man : the uses of blood and hair. Lancet 1982 ; 31: 260-2.  Lee DY, Prasad AS, Hydrick-Adair C, Brewer G, Johnson PE .Homeostasis of zinc in marginal human zinc deficiency : role of absorption and endogenous excretion of zinc . J Lab Clin Med .1993 ; 122 : 54956.  Lonnerdal B, Cederblad A, Davidsson L, Sandstrom B . The effect of individual components of soy formula and cows' milk formula on zinc bioavailability . Am J Clin Nutr 1984 ; 40: 1064-70. Lonnerdale B, Keen CL, Hurley LS . Manganese binding proteins in human and cow's milk . Am J Clin Nutr 1985 ; 41 : 550-9.  Lonnerdal B, Sandberg AS, Sandstrom B, Kunz C . Inhibitory effects of phytic acid and other inositol phosphates on zinc and calcium absorption in suckling rats . J Nutr 1989; 119 : 211-4. MacDonald LD, Gibson RS, Miles JF . Changes in hair zinc and copper concentrations of breast fed and bottle fed infants during the first six months . Acta Paediatr Scand 1982 ; 71 : 785-9.  Manary MJ, Hotz C, Krebs NF, Gibson RS, Westcott JE, Broadhead RL, Hambidge KM . Zinc  47  homeostasis in Malawian children consuming a high-phytate, maize-based diet . Am J Clin Nutr 2002; 75 : 1057-61. Marshall WA . Growth and sexual maturation in normal puberty . Clin Endocrinol Metab 1975 ; 4 : 3-25. Marshall WA, Swan AV . Seasonal variation in growth rates of normal and blind children . Hum Biol 1971 ; 43 : 502-16.  Mertz W. Trace-element nutrition in health and disease : contributions and problems of analysis . Clin Chem 1975 ; 21 : 468-75.  Nakamura T, Nishiyama S, Futagoishi-Suginohara Y, Matsuda I, Higashi A . Mild to moderate zinc deficiency in short children : effect of zinc supplementation on linear growth velocity . J Pediatr 1993; 123 : 65-9.  Navert B, Sandstrom B, Cederblad A. Reduction of the phytate content of bran by leavening in bread and its effect on zinc absorption in man . Br J Nutr 1985 ; 53 : 47-53.  Oberleas D, Muhrer ME, O'Dell BL . Dietary metal-complexing agents and zinc availability in the rat . J Nutr 1966 ; 90: 56-62.  O'Brien KO, Zavaleta N, Caulfield LE, Wen J, Abrams SA . Prenatal iron supplements impair zinc absorption in pregnant Peruvian women . J Nutr 2000; 130 : 2251-5.  Paschal DC, DiPietro ES, Phillips DL, Gunter EW . Age dependence of metals in hair in a selected U .S. population. Environ Res 1989 ; 48 : 17-28.  Pedrosa LF, Cozzolino SM . Iron supplementation effects on zinc bioavailability in a regional diet of northeastern Brazil . Rev Saude Publica 1993 ; 27 : 266-70. Pilch SM, Senti FR . Analysis of zinc data from the second National Health and Nutrition Examination Survey (NHANES II) . J Nutr 1985 ; 115 : 1393-7. 48  Prasad A . Recent developments in the diagnosis of zinc deficiency in man . In: Alan A, ed . Clinical applications of recent advances in zinc metabolism .  New York : Liss Inc . 1982 :141-149.  Prasad AS . Discovery of human zinc deficiency and studies in an experimental human model. Am J Clin Nutr 1991 ; 53 : 403-12.  Prasad AS, Brewer GJ, Schoomaker EB, Rabbani P . Hypocupremia induced by zinc therapy in adults. JAMA 1978 ; 240 : 2166-8.  Prasad AS . Clinical and biochemical manifestation of zinc deficiency in human subjects . J Pharmacol 1985; 16 : 344-52.  Prasad AS . Zinc in growth and development and spectrum of human zinc deficiency . J Am Coll Nutr 1988; 7 : 377-84.  Public Health Agency of Canada . Population Health . 2002.Retrieved July 2, 2005 from www.phac-aspc.gc.ca/ph-sp/phdd/approach/approach .html Sakai T, Warishi M, Nishiyama K . Changes in trace element concentrations in hair of growing children. Biol Trace Elem Res 2000 ; 77: 43-51. Sandberg AS . The effect of food processing on phytate hydrolysis and availability of iron and zinc. Adv Exp Med Biol 1991 ; 289 : 499-508.  Sandberg AS, Brune M, Carlsson NG, Hallberg L, Skoglund E, Rossander-Hulthen L . Inositol phosphates with different numbers of phosphate groups influence iron absorption in humans . Am J Clin Nutr 1999 ; 70 : 240-6.  Sandstead HH . Zinc nutrition in the United States . Am J Clin Nutr 1973 ; 26: 1251-60 .  49  Sandstrom B, Cederblad A . Zinc absorption from composite meals . II . Influence of the main protein source. Am J Clin Nutr 1980 ; 33 : 1778-83. Sandstrom B, Sandberg AS . Inhibitory effects of isolated inositol phosphates on zinc absorption in humans . J Trace Elem Electrolytes Health Dis 1992 ; 6 : 99-103. Sears, W. Feeding infants and toddlers . 2006 . Retrieved August 14, 2007 from http ://vvww.askdrsears .com/html/3/TO30800 .asp Sky-Peck HH . Distribution of trace elements in human hair . Clin Physiol Biochem 1990 ; 8 : 70-80. Smit Vanderkooy PD, Gibson RS . Food consumption patterns of Canadian preschool children in relation to zinc and growth status. Am J Clin Nutr 1987 ; 45 : 609-16. Srikumar TS, Johansson GK, Ockerman PA, Gustafsson JA, Akesson B . Trace element status in healthy subjects switching from a mixed to a lacto-vegetarian diet for 12 months . Am J Clin Nutr 1992 ; 55 : 885-90. Turk DE . Effect of sex steroids on zinc absorption . Fed Proc 1977 ; 36:1100. Turnlund JR, King JC, Keyes WR, Gong B, Michel MC . A stable isotope study of zinc absorption in young men : effects of phytate and alpha-cellulose . Am J Clin Nutr 1984 ; 40 : 1071-7. Van Wouwe JP, van den Hamer CJ . Hair zinc in infancy and childhood . Sci Total Environ 1985 ; 42: 149-55.  Walravens PA, Chakar A, Mokni R, Denise J, Lemonnier D . Zinc supplements in breastfed infants. Lancet 1992 ; 9: 683-5.  Walravens PA, Krebs NF, Hambidge KM . Linear growth of low income preschool children receiving a zinc supplement. Am J Clin Nutr 1983 ; 38: 195-201 . 50  Yan L, Prentice A, Dibba B, Jarjou LM, Stirling DM, Fairweather-Tait S . The effect of long-term calcium supplementation on indices of iron, zinc and magnesium status in lactating Gambian women. Br J Nutr 1996; 76: 821-31.  Yokoi K, Alcock NW, Sandstead HH . Iron and zinc nutriture of pre-menopausal women: associations of diet with serum ferritin and plasma zinc disappearance and of serum ferritin with plasma zinc and plasma zinc disappearance . J Lab Clin Med 1994 ; 124 : 852-61.  Zachwieja Z, Chlopicka J, Schlegel-Zawadzka M, Zagrodzki P, Wypchlo J, Krosniak M . Evaluation of zinc content in children's hair . Biol Trace Elem Res 1995 ; 47: 141-5 .  51  CHAPTER II 1  HAIR ZINC STATUS AND DETERMINANTS OF LOW HAIR ZINC AMONG A GROUP OF LOW-INCOME PRESCHOOLERS OF VANCOUVER  I A version of this chapter has been developed for publication . Vaghri , Z, et al (2008) . 1-lair zinc of a group of low income preschoolers of Vancouver .  52  II.1 . INTRODUCTION Studies of zinc deficiency in industrialized societies (Hambidge et al, 1972, Buzina et al, 1980, Walravens et al, 1983, Nakimura et al, 1993, and Chakar et al, 1993), including Canada (Smit Vanderkooy & Gibson, 1987 and Gibson et al, 1989-b), have reported a consistent diminution in the hair zinc level of children with marginal zinc deficiency, MZD . Marginal zinc deficiency is encountered when dietary zinc intake, over time, is marginally short of meeting the body's zinc needs . This deficiency has been described as having effects such as slight growth retardation, poor appetite, impaired taste acuity (Hambidge et al, 1972), and male oligospermia (Polysangam et al, 1997), none of which are specific clinical symptoms. Diagnosis of MZD is also hindered by the fact that serum zinc, the most commonly used biomarker of zinc status, is neither sensitive nor specific enough to detect this deficiency . In part due to this, there is at present no data available on the prevalence of MZD. However, MZD is believed to be the most common type of zinc deficiency world-wide (Endre, Beck & Prasad, 1990) . Over the last few decades, hair zinc has been discussed frequently and often used as the biomarker of this deficiency . It appears to have both the sensitivity (Hambidge, 1980) and specificity (Hopps, 1977) required as a biomarker of zinc status during the very early stages of the deficiency. Zinc is incorporated in hair during its development and is derived from the matrix of connective tissue papillae (with blood and lymph vessels) and sebaceous glands (Hopps, 1977) . The uptake of zinc by the hair is quite slow and may be hindered preferentially if the body's zinc supply is decreased . In these circumstances the more important zinc dependent organs and machineries continue to receive an adequate supply 53  without any symptom of zinc deficiency becoming apparent (Hambidge, 1980) while the concentration of zinc in the hair is reduced . It is this special physiology that confers sensitivity to hair zinc as a biomarker of marginal zinc deficiency. Hair zinc as a biomarker of MZD is also specific . This important feature derives from the anatomy of hair . During the hair's growth the metal is incorporated in the hair. Once the follicle reaches the skin surface and buds out, it is no longer in equilibrium with the body (Assarian & Oberleas, 1977) . As the extruded hair approaches the skin surface, its outer layer hardens and becomes relatively impermeable, thus isolating the hair and its contents from the body's continuing metabolic activities (Hopps, 1977) . Due to this characteristic, hair zinc is trapped and no longer available to the body for redistribution to zinc dependent organs/processes in a time of need . Nor is it under tight homeostatic regulation as is serum zinc . All in all, hair zinc concentration is believed to be a good index of MZD in children, at least in developed countries where widespread proteinenergy-malnutrition does not exist (Gibson, 1989). Hair zinc levels in humans reflect dietary intake of this element for the 4-8 weeks preceding the sample collection (Hambidge et al, 1972) . Studies of MZD in children (Smit Vanderkooy & Gibson, 1987, Gibson et al, 1989-b and Krebs, Hambidge & Walravens, 1984) suggest that in developed countries low dietary zinc intake is not the only cause of MZD . In developed countries changes in food consumption patterns can result in this deficiency, while the amount of dietary zinc (the zinc ingested through the food) is seemingly adequate . Studies in Canada (Gibson et al, 1989) and elsewhere (Hambidge et al, 1972) have shown a positive correlation between the intake of bioavailable zinc from animal protein and hair zinc status . Animal protein - red meat in 54  particular - is not only an abundant source of zinc (Health Canada, 1999) but also has been shown, even from a single meal, to counteract the effect of zinc absorption inhibitors (Sandstrom & Cederblad, 1980). In contrast, intake of some plant foods has shown a negative correlation with hair zinc (Smit Vanderkooy & Gibson, 1987) . It is generally believed that changes in dietary patterns may, in fact, be the principal underlying reason for the widespread occurrence of this deficiency in developed countries (Gibson et al, 2001) and people with particular food choices/patterns may therefore be at higher risk for this deficiency. Food choices are made under the influence of many factors . Among them income, a strong determinant of health, affects food choices in several ways . With the best sources of dietary zinc located in the most expensive foods, those subsisting on low income are at particular risk of inadequate zinc nutrition (Sandstead, 1973) . Overall, socio-economic factors do leave their footprint in most nutritional deficiencies . Differences in zinc deficiency based on the socio-economic factors, whether mediated by food choices and intake or independent of them, are noteworthy. Information available on the zinc status of Canadian preschoolers is scarce with no information at all on the zinc status of low-income preschoolers . However, it has been reported that low-income children are at higher risk for some nutritional deficiencies (Silva et al, 2001, and Skalicky et al, 2006) . We, therefore, designed this study to investigate the hair zinc status of a group of low-income preschoolers from the Vancouver inner city . The specific objectives of this study were: 1) To assess the hair zinc status of a group of conveniently sampled low-income preschoolers from 4 inner city neighborhoods . 55  2) To explore the relationship between the 5 main socio-demographic and socioeconomic factors (age, sex, ethnicity, family income and maternal education) and hair zinc levels, both individually as well as collectively. 3) To examine the existence of any association between hair zinc and the anthropometric Z scores (as indices of growth and nutrition) in an attempt to verify any functional ramifications (effects on vertical as well as ponderal growth) of low hair zinc. 4) To investigate the relationship of consumption of the main food groups as well as some nutrition, health and behavior related variables, with hair zinc levels, both univariately and after adjustment for socio-demographic and socio-economic factors. 5) To assess the collective contribution of all of the above-mentioned factors (factors significantly associated with hair zinc) for explaining the variability in hair zinc. 11 .2. METHODS AND MATERIALS II .2 .a. Study design This study was a cross-sectional exploratory study carried out with a group of conveniently sampled low income preschoolers in Vancouver's inner city during February-March of 2004 . To ensure the low-income status of the study, participants living in 4 low-income neighborhoods (Strathcona, Mount Pleasant, GrandviewWoodlands and Hastings-Sunrise) located for the most part in East Vancouver were selected for recruitment . This is a multicultural region where the income of at least 30% of the families fell below the Statistics Canada low-income cutoff in 2001 (Canadian Council on Social Development, 2001) . All 4 neighborhoods have also been identified as being highly vulnerable regarding early child development (Hertzman et al, 2002). Through field visits and consultations with community workers, 2 large community 56  centers and/or neighborhood houses from each neighborhood, that provided a wide variety of programs for families with preschoolers, were contacted and selected to participate in the study by a) allowing us to recruit from their programs and b) hosting a nutrition clinic for the recruited subjects . Of the 8 centers contacted, 2 did not respond to our initial and follow-up contacts . Two centers refused to participate, 1 because it was already engaged in another ongoing study, and the other because its board refused for unexplained reasons . The 4 remaining centers did respond to our request and agreed to participate in the study . The study was approved by the Human Ethics Committees of the University of British Columbia (Appendix I-1) and the BC Children's and Women's Health Center . Informed consent was obtained from the parents. II.2 .b. Subject recruitment Through the programs offered in the centers, parents/caregivers of apparently healthy 24-71 months old preschoolers were approached either individually or in groups. The study purpose and protocol were explained to those parents wishing to participate in the study . They were then requested to fill out a 9-item recruitment questionnaire (Appendix I1-2) . From a list provided at the bottom of the recruitment sheet they were also asked to check a preferred clinic date and a morning or afternoon appointment . In addition, ads were placed at the centers on various bulletin boards inviting families to attend the nutrition clinics (with recruitment to take place at the clinics) (Appendix II-3). A total of 256 families were thus recruited. II.2.c . Inclusion criteria The inclusion criteria for this study were ; 1) being healthy (not diagnosed with any major or chronic illness), 2) being between 24-71 months of age, 3) residing in 57  Vancouver, 4) not having a sibling already recruited for the study, and 5) the primary care giver being fluent in English, Mandarin or Cantonese. II.2 .d. Nutrition Clinics Six clinics from 9am to 5pm were held at Britannia Community Center (in Grandview Woodlands), Kiwassa Neighborhood House (in Hasting Sunrise), Mount Pleasant Neighborhood House (in Mount Pleasant) and Strathcona Community Center (in Strathcona) . Britannia and Strathcona each hosted 2 clinics. The 250 recruited families were contacted by phone and an appointment was confirmed for their preferred clinic day 2-4 days prior to the clinic day . Out of 250 calls made 207 accepted our invitation (acceptance rate 83%) and 144 attended the clinics (attendance rate 70%) . The study purpose and protocol were re-explained to the parents, and another informed consent form (for participating in the clinics) was signed (Appendix II-4) . Socio-demographic, socio-economic and dietary data were obtained, children's hair samples for zinc analysis were collected and their heights and weights were measured . At the time of admission, each child was given a colorful pictorial passport (Appendix II-5), which received a stamp at every station that the child attended. Upon inspection of his/her passport at the end of the clinic, to ensure that every station had been attended, all children with completed passports received a gift certificate as an incentive. IL2 .e. Socia-demographic and socio-economic questionnaire The information on sex, age (date of birth) and ethnicity was collected through the recruitment questionnaire (appendix II-2) . Due to a predominance of Caucasian and Chinese, and small and scattered presence of all other ethnic groups, we categorized the 58  ethnicity of our study participants into 3 categories of Caucasian, Chinese and "other ethnicities". The data on maternal education and family income were collected through the socio-demographic and socio-economic questionnaire used in a sister study (food insecurity among inner city children) conducted side by side in our nutrition clinics . This questionnaire (Appendix II-6) collected information on maternal education as well as family size and income . Since the children were recruited from the very low-income part of the city, in order to measure the extent of poverty, the income groupings used were based on the gradation of Statistics Canada's low-income cut-off (LICO), commonly considered to be Canada's poverty line. According to this scaling, households that spend disproportionate amounts of their pre-tax income on food, clothing and shelter- 20% above the average family- are considered low income . Low-income cut-offs vary by family size as well as the population of community of residence . Income levels were calculated after collecting information on these other variables and using the appropriate t II.2.f. Biochemical data II.2.f.A. Collection of hair samples for zinc analysis For this purpose, stainless steel scissors dipped in methanol, as recommended and provided by the trace element analysis department of BC Children's Hospital, BCCH, were used . A brief simulation session of sample collection was set up with the attendance of the research student, the person conducting the analyses and the person assigned to the clinic's sample collection station . After the mineral analyst's demonstration, a simulated hair sample collection was performed on one of the session's attendants using the demonstrated protocol and the critical points of the procedure were discussed . Hair  59  samples were cut, as close to the scalp as possible, at 3-4 locations at the back of the head . Only the first 1-2 cm proximal to the scalp was kept . The collected samples were placed in pre-labeled zip-lock bags and delivered to the "Nutrition Research Program" laboratories . There, the samples were thoroughly inspected for sample size adequacy and sample length, and their weights recorded . Subjects with an inadequate sample size or length were also recorded (These samples, although analyzed and their results reported, were later excluded from statistical analyses) . The samples were stored overnight and delivered to the BCCH laboratory the following day. Only 131 of the 144 children of the clinics provided a hair sample, out of which only 87 (66%) samples with acceptable weight and lengths were used in data analyses. Analyses of variances of age, sex, ethnicity, family income and maternal education between children with adequate (n=87) and inadequate (n=44) hair samples did not indicate any statistically significant differences. II.2.f. B . Analyses of hair samples The collected hair samples were processed (Puchyr et al, 1998), acid washed, digested and analyzed for zinc content by Inductively Coupled Plasma Mass Spectrometry, ICP-MS (Lockitch et al, 2005) . The samples were run in singles (due to the inadequacy of samples to run duplicates) and standard reference materials were purchased and used during the analyses . The in-house reference (Quebec hair sample) for the Mineral Analysis Division of BC Children Hospital was also used with every run. The most commonly used cut-off of 70 µg/g, and defined as 3 SD below the adult mean (Hambidge et al, 1972), was used for this study .  60  II.2.g. Anthropometrics Trained research personnel measured the children's height and weight . Their standing heights were measured to the nearest 1/8 th of an inch with a portable direct reading stadiometer, the subjects being shoeless . Heights were converted to metric units later. Body weight was measured to the nearest 0 .1 kg using a lithium electronic scale (Taylor Precision Products, L .P, Model 7300) with the scale placed on hard floor and children shoeless and wearing light indoor clothing . All the readings were done in multiples until 2 consecutive numbers were identical or not differing by more than 0 .5 unit of measurement . When 2 consecutive readings were not identical, the average of the 2 readings was recorded . In addition we obtained at the clinics the self-reported heights of both biological parents (Appendix 1I-8). The growth charts of the Center of Disease Control, CDC (Center of Disease Control, 2002) were used to calculate the weight for age (WAZ) z scores of the children. Because there is a significant genetic component in the attainment of height, we used the mid-parent height ([mother's height + father's height]/2) and parent-specific adjustments for stature of the children were carried out (Himes et al 1985) . The resultant adjusted heights were used in the calculation of height-for-age (HAZ) and weight-for-height (WHZ) scores from the growth charts of CDC. II.2.h. Dietary data ; collection and processing II.2 .h .A. Data collection II.2.h.A.a. Food frequency questionnaire, FFQ The FFQ was developed combining 2 FFQs used in previous studies (Williams, 2000 and Innis et al, 1997) . Adjustments were made to the portion sizes to make it 61  applicable to preschoolers, and some new food items were added to make it more relevant to the multiethnic makeup of the preschoolers of the study . The questionnaire underwent repeated and extensive reviews by 2 practicing dietitians in the presence of all the research staff due to be trained in its use . The final versions used in the clinics contained 177 items (Appendix II-7) . Because the FFQ was not validated for data collection on children's zinc intake, these data were not used quantitatively for zinc or any of the nutrients. We compared our mean (+SE) energy intake (1800+93 kcal for children <4 years old and 1900+107 kcal for children>4 years old) with those reported for low income American preschoolers of the Continuing Survey of Food Intakes by Individuals data, CSFII, 1994-1998 (1481+39 for children 2-3 years old, and 1742+64 for children 4-8 years old) (Knol, Haughton & Fitzhugh, 2005) . While the 2 older age groups had a comparable total calorie, those of our younger children were higher than the younger children of CSFII . We reasoned that the presence of much older children in the older age group of CSFII (2-8 years old in CSFII versus 2-6 years old in our study) had resulted in a mean high enough to be comparable with the mean of our older age group . Overall, we concluded, there was a significantly higher mean for caloric intake for children of our study compared to those of CSFII and an overestimation in our dietary tool . Such overestimation of dietary intake through a food frequency questionnaire (FFQ) is a common and recognized problem. (Treiber et al, 1990 and Stein et al, 1992) that can mask the relationships between food intake and disease . To avoid this, at the time of data analyses, we entered the total energy intake of children as a covariate in all regression analyses pertaining to food intake .  62  Dietitians fluent in Mandarin or Cantonese administered the FFQ in these languages when needed (Canadian Council on Social development, 2001). II.2.h .A .b. Data on other nutrition and health related variables These data were collected through both the recruitment questionnaire and the FFQ . Nutrition and health-related data collected through the former were: 1)Two questions that assessed the parents' perception of their child's eating behavior by first asking if they are concerned about their children's eating in general, and then providing 10 options describing the common eating characteristic/problems of young children (such as picky eater and not eating enough) as identified in our meetings with community dieticians, from which they were to choose one or more that best fit . Space was also provided in this first question for concerned parents to write about any additional eating behavior/problem in their children not provided as an option. 2) Two questions asking parents first whether their children are frequently sick and then asking them to describe their general health from 3 options : "very healthy", "average health", "not so healthy". In addition, 2 questions appended to the FFQ collected information on breastfeeding history . Each study participant had the same code on all questionnaires (the recruitment questionnaire, FFQ . . .) . These identified the respective questionnaires and were then used in transcribing all relevant data for further analyses. II.2.h.B. Processing of dietary data II.2.h.B .a. Food Groups Assignment All of the foods recorded in the FFQ were extracted to make a `Master Food List'. Those in the master food list generally belonged to 1 of the 2 main categories of "basic 63  food" such as chicken or "combination food" such as shepherds' pie . All the items in the basic food category were assigned a code corresponding to the food groups they belong to based on the four food groups defined by Eating Well with Canada's Food Guide for Healthy Eating, CFGHE ('M' for meat and meat alternatives, 'F&V' for fruits and vegetables, `D' for dairy, `C' for cereals and grain products) . For example, bagels were given a `C' code, while cheese strings and mangos were given codes `D' and `F', respectively. A number of frequently reported foods in the FFQ of children, such as popcorn, ice cream and pretzels, which, due to their excessive fat, sugar or salt content, were not considered part of the aforementioned 4 main food groups, were assigned a code `O' standing for `Other Food Group' . In deciding what belongs to this list, the methodology section of the recent BC nutrition survey was consulted and their list was referred to when needed (BC Nutrition Survey, 2002). The assignment of food group codes to combination foods like commonly reported items such as pizza or macaroni and cheese was done through consultations with 2 practicing dietitians who reviewed the recipes of these foods as required . A list of combination foods together with available recipes was then given these 2 dietitians . They were asked to review the list and decide on the percentage breakdown of their ingredients . The research student made a similar list as well . The 3 lists were then reviewed and cross-matched by the research student . The items having 2 or 3 different codes were extracted and a mini-list was developed . The student met with the 2 dietitians and this mini-list of combination foods was reviewed and discussed until consensus was reached regarding the codes to be used and the percentages assigned to their components. 64  II.2.h.B.a.Estimating the number of servings Once the master-sheet of the food codes was completed, the food list for each child was entered into the food analysis software, ESHA, (Esha Food Analysis Software, version 8 .4.0) and printed . The printed sheets were coded manually as described and the number of servings/day for each food group was calculated . To do this, the method of Junkins and Laffey, 2003, used in the recent BC Nutrition Survey, BCNS, was adopted. For all food groups, the number of servings was calculated directly by dividing the total amount consumed (grams, ounces, milliliters etc.) by the serving size of that food group for that age. For instance, if the total bread intake of a 5 year old child was 3 .5 slices /day, then since 1 slice of bread is defined as 1 serving, the total bread consumed by this child was calculated as 3 .5/1=3 .5 servings/day. 11 .2 .i. Statistical Analysis The data were analyzed using Statistical Package for Social Science, SPSS, (version 13 .0, 2005). Using this software, descriptive statistics (such as mean and standard deviation) were computed for all variables . The primary outcome variable in this study was the hair zinc level . These outcomes were explored in both continuous (hair zinc level in µg zinc /g of hair) and binary (low hair zinc i .e. hair zinc<70 µg/g and normal hair zinc) forms. Five sets of linear regression models were used to analyze hair zinc level as a continuous outcome . First, the association between hair zinc level and each sociodemographic or socio-economic variable was assessed . Second, the association between hair zinc level and each dietary variable was assessed, both univariately and after adjustment for the socio-demographic and socio-economic variables was carried out. 65  Third, the association between hair zinc level and each eating behavior was assessed, both univariately and after adjustment for the socio-demographic and socio-economic variables . Fourth, the association between hair zinc level and each anthropometric Z score was assessed, both univariately and after adjustment for income, ethnicity, and maternal education was carried out . Finally, all the variables in the above-mentioned regression analyses, that revealed a significant association with hair zinc, were inserted in a forward stepwise linear regression analysis with hair zinc as the dependent variable. The forward stepwise regression analysis was carried out in order to identify the variables that could explain some of the variability in children's hair zinc . The inclusion criterion for variables to be inserted in this analysis was having a P value of <0 .05 in their individual adjusted linear regression analyses with hair zinc as the outcome variable . For this regression analyses the collinearity of the variables inserted into the model was verified . Collinearity between two components of a regression model could "blow up" the standard deviation and result in elimination of one of the variables from the final model produced. When the outcome variable was continuous and the explanatory variable of interest was categorical, differences between groups were verified using one-way analysis of variances (ANOVA) followed by Tukey's post hoc analyses, if applicable. When the outcome variable and the explanatory variable of interest were both in categorical form, the differences were examined using Chi square analysis (e.g. the difference in the occurrence of low hair zinc between boys and girls) . Adjustments in chi square  analyses were carried out when the differences between more than 2 groups were  explored . For all the analyses, the level of significance was set at P<0 .05 .  66  Although considering the power of this study (n=87) conducting multiple analyses may pose a threat to the validity of the analyses, it is important to bear in mind that the mere purpose of these analyses was to identify as many variables (variables showing significant associations with hair zinc) as possible, as components of the questionnaire that was going to be constructed in the second study. 11.3 . RESULTS II.3.a.Study population A description of the study population is given in Table II .I . There were 87 children in the sample with the 2 sexes roughly equal in representation . Because CFGHE provides recommendations based on age breakdown of children<4 years old and children>4 years old, age 4 was arbitrarily taken as the dividing line to categorize children into 2 groups of younger (<4 years old) and older (>4 years old) preschoolers. The younger and older children, as well, were equally represented. The families had a multi-ethnic make-up with a majority of the children having Chinese or Caucasian ethnicity . More than half of the study families (57%) were below the poverty line defined by Statistics Canada . About 36% of our study mothers did not have post secondary education. As a population the mean (+SD) of HAZ, WAZ, and WHZ scores were within the range defined for a population with normal growth distribution . Overweight (WAZ>2), underweight (WAZ<-2), stunting (HAZ<-2) (Walravens, Krebs & Hambidge, 1983, and Gibson et al, 1989) and slowed-growth (HAZ<-1) (Gibson et al, 1989) were not observed at more than the expected baseline .  67  Table II.1 . Description of study population (n=87). Socio-demographic variables Sex: Boys Girls Not recorded Age: < 4 years old > 4 years old Not recorded Ethnicity: Caucasian Chinese Others Not recorded  Income: <50% LICO * 50%- <100% LICO 100-150% LICO >150%LICO Not recorded Maternal education: No post secondary education With post secondary education Not recorded Anthropometrics f : Height for age z scores ** Weight for age z scores Weight for height z scores  Number (%) 43(49) 37(43) 7(8) 41(47) 39(45) 7(8) 27(31) 35(40) 21(24) 4(5) 27(31) 23(26) 15(17) 14(16) 8(10) 31(36) 48(55) 8(9) Mean ± SD -0.15±1 .7 0.20±1 .2 0 .27±1 .3  *LICO=Low-income cutoff .  t For anthropometrics the sample sizes were 58, 61 and 69, respectively. * Three subjects with HAZ scores falling outside the criterion of -3 .5< Z scores<3 .5 (Fallon and Spada, 1997) were considered outliers and were removed from the database .  68  II.3 .b. Hair zinc status of the study population Hair zinc of the survey population was normally distributed with a mean (±SD) of 75±30 µg/g, while their hair zinc varied within a wide range of 10-160 µg/g (Figure II.I). The study children also had a high incidence of low hair zinc levels (hair zinc<70 µg/g), as illustrated by the shaded area of Figure II.I. The fraction of children with hair zinc below the cutoff was 43% (n=40) of the study population .  69    12 .5 -  10D-  7 .5  2 .0 -  Mc an—75  2  0 .0 0.00  25 ' 00  50 00  75 .00  1 00 .00  1 25 .00  r 150 .00  	L  SIB=3U 14=87  Hair Zinc (uglg)  The shaded area demonstrates the children with hair zinc below the cutoff of 70 pg/g.  Figure ILI. Frequency distribution and descriptive statistics for the hair zinc level of the study population.  70  II. 3.c . Associations of the 5 main socio-demographic and socio-economic variables with hair zinc Table II.II summarizes the hair zinc level together with the occurrence of low hair zinc based on the 5 main socio-demographic and socio-economic variables (sex, age, ethnicity, income, and maternal education) . As shown, there was no significant difference in the hair zinc of the children based on sex, age, ethnicity, family income and maternal education. The occurrence of low hair zinc, however, was significantly higher among younger preschoolers (children<4 years old) . The mean hair zinc level for younger preschoolers was also slightly lower than the older group, although this difference did not reach statistical significance. To explore any relationship that socio-demographic and socio-economic factors may have with hair zinc, we ran the linear regression analyses of age, sex, family income, maternal education, and ethnicity with hair zinc as the dependent variable (Table II.III). There were no statistically significant associations between hair zinc levels and anyone of these variables . When all 5 variables were inserted in a linear regression model simultaneously, the model produced lacked statistical significance (R 2 =0.08, P=0 .28).  71  Table 11.11. Socio-demographic and socio-economic description of study population and hair zinc levels and the occurrence of low hair zinc based on these variables (n=87). Socio-demographic variables  Number (%))  Hair Zinc (l.tg/g) Mean ±SD F  43(49)  80±31  p  Low hair zinc* t # (%)  x  p  0.96  0.33  3 .8  0.04  3 .5  0.17  0 .16  0 .69  0 .31  0 .58  Sex Boys  16(30) 2.5  Girls  37(43)  Missing Age  7(8)  < 4 years  41(47)  0.12 19(50)  69+31  73+34  22(55) 0 .28  >4 years Missing Ethnicity  39(45) 7(8)  76±24  Caucasian Chinese Others  27(31) 35(40) 21(24)  74+30 69±28 8532  Missing  4(5)  0.60 13(33)  1 .9  0.15  13(52) 17(50) 5(26)  Income <LICO**  50(57)  23(46)  73+29 0.26  >LICO  29(33)  Missing  8(10)  0.61  77±31  12(41)  12(41)  Maternal education No post secondary education  31(36)  76±28  With post secondary education  48(55)  73+31  Missing  8(9)  0.13  0 .73 23(48)  Low hair zinc=Hair zinc <70 pg/g. Percentages of children of that category who had hair zinc data available. **LICO=Low income cut off .  72  Table II.III. Relationship between hair zinc and demographic factors (n=87). Variable  Bt (95% CI)  R2*  p  Age (months)  0 .1(-0 .5, 0.6)  0 .00  0 .84  Sex  -10(-23, 2 .7)  0 .03  0.12  Ethnicity Caucasian Chinese Others  0.02  0.25  -12(-29, 6) -16(-33, 0 .4)  Maternal education  -2 .5 (-17, 12)  0.00  0 .72  Family income <50% LICO $ 50%- <100% LICO 100-150% LICO >150%LICO  0.00  0.61  -41(-24, 16) -0.9(-21, 19) -1 .7(-21, 24)  tBs are the unstandardized regression coefficients. Values indicate R2 of individual variables with hair zinc (unadjusted regression analyses) When all 5 variables were inserted in the regression model, simultaneously ; R2 =0 .08, P=0 .28) .  73  II.3.d. Associations of some dietary factors with hair zinc Table II.IV displays the mean (+SD) of the number of servings of the main food groups defined by CFGHE and some of their subgroups . For all main food groups the population mean met the number of servings recommended by CFGHE for preschoolers (For <4 years old the recommendations are; 1,2,3, and 4 serving/d "Meat and Alternatives", "Dairy", "Cereals and Grains", and "Fruits and Vegetables", respectively. For >4 years old the recommendations are ; 1, 2, 4, and 5 servings/d of "Meat and Alternatives", "Dairy", "Cereals and Grains" and "Fruits and Vegetables", respectively (Health Canada, 2007). The "whole grain" and "fruits and vegetables" were the 2 most poorly met food groups with 91% (n=75) and 54% (n=40) of the children not meeting their recommended number of servings, respectively (Table II .IV). The dairy group, on the other hand, was one for which most of the children met their recommendations . Only 18% (n=13) failed to meet the recommendations for dairy, while 53% (n=39) of children exceeded the recommendation of CFGHE by more than one serving per day . This high consumption of dairy mainly stemmed from the high milk consumption, which reached as high as 6 servings of milk per day for some children. There were no significant differences between the mean intakes of boys and girls in any of the food groups (Appendix II-9) . Also the percentage of girls and boys exceeding recommendations for dairy consumption were not significantly different (52% and 53%, respectively, (x2 =0 .001, df=1, p=0 .97).  74  Table II.IV also shows the results of regression analyses of hair zinc and intake of the various food groups . Total daily energy was entered as a covariate in regression analyses of all food groups. As shown, there was an inverse association between the daily consumption of dairy, as well as milk consumption, with the children's hair zinc . These associations remained significant even after adjustment for age, sex, family income, maternal education and ethnicity .  75    Table II.IV. Average (+SD) of daily intake of different food groups (servings/day), and regression analyses (both unadjusted and adjusted) of these intakes with hair zinc (µg /g) of the study population (n=74).  	P Food group  Intake Serving/day (Mean+ SD)  13 * (95% Cl)  Unadjusted K.Z  Adjusted T R `3 P  Meat and alternatives  2.9 ± 1 .7 18(24%) '  3 .1(-7 .9, 1 .7)  0 .02  0 .20  0 .09  0.28  Dairy  3 .9 ± 2.2 13(18%)  -4 .5(-7.9, -1 .1)  0.09  0.01  0.16  0.02  Milk  2.1+1 .6  -5 .2(-9 .2, -1 .2)  0 .08  0.01  0.26  0.01  Fruits & vegetables  7.2 + 5 .3 40(54%)  -0 .1(-1 .5, 1 .7)  0.01  0.93  0.08  0.48  Cereals and Grains  6 .9 + 3 .6 22(30%)  -1 .1(-3 .4, 1 .1)  0.01  0.31  0 .08  0 .59  Whole grain  1 .4 +2.0 75(9i%) +  -3 .3(-7 .1 . 0 .4)  0 .04  0 .08  0 .10  0.25  Non-whole grain  5 .5 + 2 .8  0.0(-2 .7, 2.7)  0.00  0 .98  0.08  0.80  Bs are the unstandardized rearession coefficients . t Adjusted regression for all variables has been carried out using, age, sex, ethnicity, income, maternal education and total calories as covariates. '+ R 2 indicates the overall R 2 of the model produced with that variable and the 5 main socio-demographic variables (age, sex, ethnicity, maternal education and family income) in it . The R 2 for socio-demographic variables alone was 0 .08. ff Number (%) indicates the number and percentage of children whose intake (of that food group) did not meet the recommendations of CFGHE. ***There are no recommendations for milk in CFGHE. Number (%) under whole grain indicates the number and proportion of children whose intake of whole grain did not meet the recommendation of CFGHE, which recommends 50% of cereal and grains consumed should come from the whole grain products (Health Canada, 2007) .  76  II.3.e. Associations of some health and behavior related variables with hair zinc As Table II .V. displays, prolonged breast-feeding was associated with low hair zinc while breast-feeding per se was associated with increased hair zinc . Both these associations remained significant even after the adjustments for socio-demographic and socio-economic variables were carried out . Being described by the caregiver as either "often sick" or "eats unhealthy," were negatively associated with hair zinc . These associations were independent of the socio-demographic and socio-economic factors. Neither being described as "not eating enough" nor "being in a single parent household" revealed any significant association with the children's hair zinc. Regression analyses indicated no significant associations between hair zinc and HAZ and WHZ scores . Weight-for-age Z scores on the other hand revealed a positive association with hair zinc . However, when the regression was adjusted for sociodemographic factors, the association was not significant anymore . When the children were also categorized based on sex and/or hair zinc status (separately, then together) there was no significant association between hair zinc and any one of the anthropometrics Z scores of boys, girls, all children with low hair zinc, all children with normal hair zinc, low hair zinc boys or low hair zinc girls (data not shown). Among these preschoolers, prolonged "exclusive" breastfeeding was common and 52% of breast fed children were reported to have been breast fed exclusively for >6 months (Figure . I1 .1!). We do not have any explanation for this high rate of prolonged "exclusive" breast feeding among this group of low-income children . However, in light of the fact that most of our Chinese caregivers filled out FFQ through an interpreter, we suspect that the concept of "exclusive breast feeding" may not have been communicated 77  to the caregivers properly and caregivers may have reported the duration of breast feeding rather than that of exclusive breast feeding. Figure 11.II. also displays that children fed for >6 months had hair zinc lower than those fed for <6 months (70 .2+28 and 89 .7+28, respectively, p= 0 .02) and comparable to those never breast-fed (70 .2+28 and 50 .3+22, respectively, p=0 .30), as indicated by one-way analysis of variance . ANOVA followed by Tukey's post hoc analyses .  78    Table II.V. Relationship between hair zinc and some nutrition and health related variables among study population. Variable  Unadjusted R  Having been ** breastfed (n=66 . 10)  0.07  Prolonged tt breast feeding (n=41, 26)  B* (95% CI)  Adjusted t R`  0 .02  21(3 .1, 39)  0.20  0.02  0.11  0.01  -19(-33, -5 .6)  0 .27  0.01  Being described as "often sick. (n=15, 52)  0.55  0.00  -43(-36, -52)  0.58  0.00  Does not eat enough (n=23, 42)  0.02  0 .29  6 .8(-5 .8, 19)  0 .14  0 .20  Being described as" unhealthy eater"(n=8, 57)  0.16  0.00  -30(-13, -47)  0.20  0.01  Being in a single parent household (n=19 . 64)  0.03  0.13  1 1(-3 .6, 27)  0.17  0.07  Height for age Z scores  0 .00  0.83  0 .53(-4 .1, 5 .1)  0.16  0.98  Weight for age Z scores  0 .06  0 .04  6.5(0.2, 13)  0 .17  0 .08  Weight for height Z scores  0.00  0.66  -1 .3(-7 .2 .4.6)  0.16  0.15  P  P  Bs are the unstandardized regression coefficients.  tAdjusted regression for all variables (except for anthropometric variables) has been carried out using, age, sex, ethnicity, income and maternal education as covariates. For anthropometric variables the adjustments were carried out using ethnicity, family income and maternal education as covariates . The R 2 for socio-demographic variables alone was 0 .08. **The first and second numbers in brackets indicate number of children who had and did not have the variable in question, respectively . For example, the first number in the first row of the table, (66), is the number of the children who had been breast-fed, the second number, (10), is the number of children who had not been breast-fed. tt The term "prolonged" is used to indicate "exclusive" breast-feeding beyond 6 months . 79    Not breastfed  Breastfed=<6 months  Breastfed>6 months  Analysis of variance followed by Tukey's post-hoc analyses Indicated significant differences between this category and the other 2 categories of children (P< 0 .05).  Figure 11 .II. Hair zinc level of 3 groups of children, categorized based on length of "exclusive" breast-feeding .  80  II.3.f. Contribution of all variables associating significantly with hair zinc to the overall variability in hair zinc of the inner city children Inclusion of all dietary and non-dietary variables that had shown a significant association with hair zinc in a forward stepwise linear regression analysis revealed that being described as "often sick" and "having been breast-fed" were the only 2 significant predictors of hair zinc, with food intake no longer a significant component in this model (Table II .VI). In addition to these 2 variables, other variables that (in univariate regression analyses) had already shown a significant association with hair zinc were entered in the above-mentioned stepwise regression model . These variables were "daily consumption of dairy", "daily consumption of milk", "prolonged breast-feeding", and being described as "eats unhealthy" . Nevertheless, these variables did not seem to be significant predictors of hair zinc level, as they were eliminated from the final model produced by stepwise regression analyses . The overall model accounted for only 5% of the variation in hair zinc (R 2 =0.05, P=0 .00).  81  Table II.VI. Stepwise linear regression analyses of hair zinc and the variables significantly associating with it. SE  Variables  P  R2 change  P  Constant  -22  15  Often sick  36  5 .0  0 .7  0 .46  0 .00  Breastfeeding  16  6.9  0.2  0.04  0.03  0.15  *Bs are unstandardized co-efficients. For the whole model ; R 2 =0 .50, P=0.00 .  82  II.4. DISCUSSION Results of this study revealed a very low mean of hair zinc for this group of lowincome preschoolers . The occurrence of low hair zinc (<70 µg/g) was widespread (46%), and the mean of the hair zinc was significantly lower than the previously reported (Smit Vanderkooy & Gibson, 1987) mean for middle-income Canadian preschoolers (75±30 gg/g versus 116+34, respectively, p=0 .00) (Figure II.I). We did not observe any sex or age-based differences in the mean hair zinc level or any differences in the occurrence of low hair zinc between the 2 sexes (Table II.II). On the other hand, our population mean was comparable with the levels reported for 2 groups of low-income American children who were selected based on their lower growth percentiles (height-for-age<10 th percentile) (Hambidge et al, 1976, Walravens et al, 1983) . This low population mean, along with the fact that 46% of our children had hair zinc <70 µg/g, made the overall zinc status of this population a concern . Our children were not recruited based on any inclusion criteria related specifically to suspected zinc deficiency, such as short stature, low zinc intake, or slow growth velocity, and they were a group of apparently healthy low-income preschoolers. The only common factor between our study children and the children of the 2 aforementioned studies, in whom the existence of zinc deficiency was confirmed following the growth response of the children to a period of zinc supplementation, was their socio-economic status . This not only indicates that low hair zinc might be common in groups of low-income children, but also suggests a potential zinc deficiency in our study children. Despite the comparability of the hair zinc of our children with 2 groups of short low-income children (Hambidge et al, 1976, and Walravens, Krebs & Hambidge, 1983), 83  the mean + SD of height-for-age Z score of our study population was normal (Table II. I) and comparable to that of low-income children of NHANES II (-0 .19 and -0.15, respectively) (Yip, Scanlon & Trowbridge, 1993), as was the nutritional status of the study population (indicated by children's WAZ and WHZ scores). Income is a strong and consistent determinant of health and indeed is first among the 12 identified in Health Canada's model of population health. Low-income children are known to be at higher risk for nutritional deficiencies . Perhaps this is, in part, due to some characteristic eating patterns of low-income children that predispose them to a given deficiency more than the children of a higher income group . One eating pattern characteristic of our children was a high consumption of dairy . Milk drinks were a large part of this high consumption . Indeed, drinking large amounts of milk was common among these children, reaching as high as 6 glasses per day for some children . Milk and milk products, although the second major source of energy for young North American children, are one of the food groups where children's intake often does not meet the recommended levels . Cycle 2 .2 of the Canadian Community Health Survey, CCHS, the most recent survey of Canadians' eating habits, indicates that dairy products are generally under-consumed and more than one-third (33%) of the children in the 2-9 years age group did not have the recommended 2 servings of milk products a day, based on the data collected by the 24-hours food recall (Health Canada, 2004) . For our study children this proportion was much lower (18% versus 33%), while a large portion of our children (53%) exceeded the recommended 2 servings/day by more than 1 serving a day (Table II.IV) . The results of the recent nationwide survey also indicate that the 2-9 years age group (of mixed income) was the only age group in which the average number of 84  servings for dairy was met and was actually slightly higher than the present recommendations and minimum recommendation of CFGHE (2 .31 vs . 2). Perhaps this age propensity for "dairy" combined with the relative affordability of dairy (Shanklin & Wie, 2001) is one of the factors contributing to the observed high dairy consumption among this group of low-income children. In our study high dairy intake also revealed a significant negative association with the hair zinc of children (Table I1 .111). Dairy products are high in calcium and some studies have shown that excessive calcium intake can interfere with zinc absorption through the capacity of calcium to potentiate complex formation between zinc and ligands present in the diet, such as phytate (Oberleas, Muhrer & O'Dell, 1966) . This is consistent with 2 studies of Canadian children . The first showed an inverse relationship between calcium intake and hair zinc (Smit Vanderkooy & Gibson, 1987), while the second showed that dietary [Ca x (phytate/Zn)] millimolar ratios were inversely associated with the zinc status (Gibson et al, 1991) . Consistent with this inverse association, we also observed a significant difference between the hair zinc of children who had a dairy intake > 3 servings/day (exceeding the present recommendations by more than 1 serving and exceeding the upper limit of previous recommendations by CFGHE) and those who did not exceed 3 servings/day (68+27 and 82+30 gg/g, respectively, p=0 .04). This finding confirms the previously reported negative association of excess calcium with zinc nutriture . It also extends this finding from the level of the mineral-mineral to the food-mineral level. Another common food feature among our study children was prolonged "exclusive" breast-feeding (feeding child with breast milk only) . Data on the effect of 85  income on breast-feeding are inconsistent (Dennis, 2002 and Dubois & Girard, 2003). Dubois & Girard's (2003) analysis of the Longitudinal Study of Child Development in Quebec (ELDEQ ; 1998-2002, n=2,223), clearly indicates that the mother's age, followed by her education, are the 2 strongest determinants of "exclusive" breast-feeding at 4 months, and shows that these 2 factors positively affect the length of breast feeding and over-ride the effects of other factors governing family environment including family income . However, this study and most of the literature on social determinants of "exclusive" breast-feeding focus on feeding up to 6 months, which is the focus of the public health recommendations (WHO, 2008) . We do not have an evidence-based explanation in our study for what factor(s) contributed to this characteristic of "exclusive" breast-feeding past 6 months. Whatever the underlying reason for extending the duration of breast-feeding, the study did show a negative association with children's hair zinc (Table II .V), while breast-feeding per se was positively associated with hair zinc . Furthermore, the children who were `exclusively' breast-fed for >6 months had significantly lower hair zinc than those who were fed for less than this period (70 .2±28 versus 89 .7±28 µg/g, respectively, p=0.01) (Figure II.II) . The World Health Organization (WHO, 2008) recommends exclusive breast-feeding up to 6 months . This is based on some data suggesting potential risks with `exclusive' breast-feeding beyond 6 months, including growth faltering and deficiency of some micronutrients . The bioavailability of zinc from human rrilk is significantly high (Lonnerdal, Keen & Hurley, 1981 and Sandstrom, Cederblad & Lonnerdale, 1983), however, the zinc content of milk undergoes a sharp decline over the first 6 postpartum months (sometimes dropping to as low as 25% of the initial supply)  86  (Krebs & Hambidge, 1986) . Beyond 6 months, breast milk is not able to supply an infant with adequate zinc . At this point the introduction of other foods, particularly animal protein and cereals are needed to meet the high zinc requirement of a rapidly growing infant . Growth-limiting zinc deficiency has been shown in infants principally fed with human milk even after the age of 4 months (Walravens et al, 1992) . Although hair zinc does reveal the zinc status over the previous few months, chronic mild zinc deficiency in childhood has hair zinc concentrations that are consistently low . It is possible that some of these children had low hair zinc from early infancy and, given their high physiological requirement during the time of accelerated growth, they never recovered from it. Whether as an indicator of a chronic sub-optimal zinc status, inherited from the early years of life, or as a sub-optimal zinc status resulting from the dietary insufficiencies of later years, this low hair zinc may have some negative impact on the general health status of children . The adverse effect of zinc deficiency on various aspects of the immune system has been documented (Shankar & Prasad, 1998) . This wellrecognized adverse effect was reflected in our study in the strong and significant negative association between being described as "often sick" and the hair zinc (R 2 =0.58, P=0 .00) (Table II .V). As an acute phase response element, plasma zinc is redistributed to other tissues in response to acute infection/inflammation (Prasad, 1998) . Conceivably, reduced hair zinc can be the net result of an ongoing decline in the plasma zinc of children who are "often sick" . Conversely, a lower zinc status, as indicated by lower hair zinc, can interfere with the function of the immune system, a highly zinc-dependent system, and be the cause of being "often sick" . Other studies of preschoolers have indicated lower hair  87  zinc among children who are frequently sick with upper respiratory tract infections (Von Wouwe, Wolff & Van Gelderen, 1986 and Lombeck et al, 1988). In this study there was no evidence of any association between hair zinc and growth status (Table II.V), an association shown in some studies (Hambidge 1972, Smit Vanderkooy & Gibson, 1987 and Gibson et al, 1989) . Although the anthropometric data did not support the existence of any overt growth retardation, this does not eliminate the possibility of the existence of MZD among our study children. Children with marginally low hair zinc and normal growth parameters have been shown to exhibit a differential growth response (compared to growth in the placebo group) when supplemented with zinc (Ruz et al, 1997). When all of the variables that revealed significant association with hair zinc were inserted in a stepwise linear regression analysis, being described as "often sick" and "having been breast-fed" remained significant components of the final model as the only two variables that could explain part of the variability in children's hair zinc (Table II.VI). The collective model could account for only 5% of the variability in hair zinc level (R2 =0.05, p=0.00). Overall, our results indicate widespread low hair zinc among this group of lowincome children . The biochemical evidence, combined with some predisposing dietary factors, (i.e. the excessive calcium intake), and some indirect evidence of impaired immunity (i .e. often sick with flu and common infections), suggests the likely existence of MZD among these children, even though their growth parameters did not provide evidence to support this . This assessment can only be confirmed or ruled out by carefully designed supplementation studies . 88  This study has some limitations that should be acknowledged . First, the food frequency questionnaire was not validated to collect data on zinc consumption . We thus refrained from exploring the zinc consumption of children based on the nutrient analysis of the FFQ. In addition, all of our data, including the dietary data, were self-reported. Unlike adults who are known to under-report their food intake, caregivers over-report the children's intake (Devaney et al, 2004) . While part of the reason for this over reporting may be social desirability, part of it may also arise from over- estimating the serving size. Research has shown that most people are unaware of what constitutes an appropriate portion size (Young & Nestle, 1998) . We had tried to minimize this shortcoming of the FFQ by using plastic food models and, to increase the accuracy of our dietary data, we collected it from a food frequency questionnaire completed through an interview administered by trained dietitians and research personnel rather than by parents. Secondly, our study sample was small and we could not explore any age and/or sex based differences. Thirdly, selection bias may have limited the generalization of our study's findings. Our children were a group of low-income children recruited in a conveniently from selected community centers of selected neighborhoods . Therefore, our findings may not be generalizable to all low-income preschoolers of the city. Lastly, as in any cross-sectional study, our findings only indicate associations and the observed relationships cannot be taken as a cause and effect relationship or lack of it. This study provides some important reference values on hair zinc level of low-income Canadian preschoolers . Our findings also add to the present literature by confirming an association of hair zinc with some previously reported variables and identifying its 89  association with some new variables . These findings need to be confirmed with larger studies of apparently healthy low-income preschoolers .  90  IL5. BIBLIOGRAPHY Assarian GS, Oberleas D. Effect of washing procedures on trace element content of hair. Clin Chim 1977 ; 23 :1771-2.  BC Nutrition Survey. Report on food groups .2004. Retrieved May 29, 2005 from http://www.healthservices . gov.bc.ca/prevent/nutrition/pdf/fgreport .pdf  Buzina R, Jusic M, Sapunar J, Milanovic N . Zinc nutrition and taste acuity in school children with impaired growth . Am J Clin Nutr 1980 ; 33 : 2262-7. Canadian Council on Social Development . Defining and Re-defining Poverty : A CCSD Perspective . 2001 . Retrieved February 5, 2007 from http ://ccsd .ca/pubs/2001/povertypp .htm  Center for Disease Control . CDC growth charts . 2002 . Retrieved August 11, 2005 from http : //www. cdc. go v/nchs/about/major/nhanes/gro wthcharts/datafil e s . htm  Chakar A, Mokni R, Walravens PA, Chappuis P, Bleiberg-Daniel F, Mahu JL, Lemonnier D . Plasma zinc and copper in Paris area preschool children with growth impairment. Biol Trace Elem Res 1993 ; 38 : 97-106.  Dennis CL . Breastfeeding initiation and duration : a 1990-2000-literature review . J Obstet Gynecol Neonatal Nurs 2002 ; 31 : 12-32.  Devaney L, Kalb R, Briefel T, Zavitsky-Novak N, Ziegler P . Feeding Infants and Toddlers Study . Overview of the study design . J Am Diet Assn 2004 ; 104 : S8-S13.  Dubois L, Girard M . Social determinants of initiation, duration and exclusivity of breastfeeding at the population level : the results of the Longitudinal Study of Child Development in Quebec (ELDEQ 1998-2002) . Can J Public Health 2003 ; 94 : 300-5. 91  Endre L, Beck FWJ, Prasad AS . The role of zinc in human health . Trace Elements Exp Med 1990; 3 : 337-75. Fallon A, Spada C . Detection and accommodation of outliers in normally distributed data sets, 1997 . Retrieved October 3, 2007 from http : //www. cee. vt . edu/ewr/environmental/teach/smprimer/outlier/outlier . html  Gibson RS, Heath AL, Limbaga ML, Prosser N . Are changes in food consumption patterns associated with lower biochemical zinc status among women from Dunedin, New Zealand? Br J Nutr 2001 ; 86: 71-80. Gibson RS, Smit Vanderkooy PD, Thompson L . Dietary phytate x calcium/zinc millimolar ratios and zinc nutriture in some Ontario preschool children . Biol Trace Elem Res 1991 ; 30: 87-94.  Gibson RS, Vanderkooy PD, MacDonald AC, Goldman A, Ryan BA, Berry M . A growth-limiting, mild zinc-deficiency syndrome in some southern Ontario boys with low height percentiles . Am J Clin Nutr 1989 ; 49 : 1266-73. Gibson RS . Assessment of trace element status in humans. Prog Food Nutr Sci 1989; 13 : 67-111.  Hambidge KM . Hair analysis . Pediatr Clin North Am 1980 ; 27: 855-60. Hambidge KM, Hambidge C, Jacobs M, Baum JD . Low levels of zinc in hair, anorexia, poor growth, and hypogeusia in children . Pediatr Res 1972 ; 6: 868-74. Hambidge KM, Walravens PA, Brown RM, Webster J, White S, Anthony M, Roth ML . Zinc nutrition of preschool children in the Denver Head Start program . Am J Clin Nutr 1976 ; 29: 734-8. 92  Health Canada, Canada's Food Guide for Healthy Eating, 2007 . Retrieved March 4, 2007 from http://www .hc-sc. gc .ca/fn-an/food-guide-aliment/myguide-monguide/index_e .htmlHealth Canada, Nutrient value of some common foods,1999 . Retrieved December 12, 2006 from http://www .hc-sc.gc .ca/fn-an/alt formats/hpfb-dgpsa/pdf/nutrition/nvscf-vngau_e .pdf Health Canada . Canadian Community Health Survey : Overview of Canadians' eating habits, 2004 . Retrieved January 16, 2007 from http://www .statcan.ca/Daily/English/060706/dO60706b .htm Hertzman C, McLean S, Kohen D, Dunn J, Evans T . Early development in Vancouver : Report of the Community Asset Mapping Project (CAMP) . Retrieved October 19, 2006 from http:I/www.help.ubc .ca/pubMaps/camp/vancouverreport .pdf Himes JH, Roche AF, Thissen D, Moore WM . Parent-specific adjustments for evaluation of recumbent length and stature of children . pediatr 1985 ; 75: 304-13. Hopps HC . The biologic bases for using hair and nail for analyses of trace elements. Sci Total Environ . 1977 ; 7: 71-89.  Innis SM, Nelson CM, Wadsworth LD, MacLaren IA, Lwanga D . Incidence of irondeficiency anemia and depleted iron stores among nine-month-old infants in Vancouver, Canada . Can J Public Health 1997 ; 88 : 80-4. Junkins B, Laffey P. (2003) . B .C. Nutrition Survey Canada Food Guide Tables E451313-011 CFG-V 1 . Biostatistics and Computer Applications Division, Food Directorate, Health Canada . 93  Knol LL, Haughton B, Fitzhugh EC . Dietary patterns of young, low-income US children . J Am Diet Assoc 2005 ; 105 : 1765-73. Krebs NF, Hambidge KM, Walravens PA . Increased food intake of young children receiving a zinc supplement . Am J Dis Child 1984 ; 138 : 270-3. Krebs NF, Hambidge KM . Zinc requirements and zinc intakes of breast-fed infants. Am J Clin Nutr 1986 ; 43 : 288-92. Lockitch, G ., Mattman, A ., Palaty J. (2005) . "Trace Metals analyses . American Association of Clinical Chemists Annual Meeting, Orlando FL, July 27 (2005). Abstract obtained from the Author (JP)  Lombeck I, Wilhelm L, Hafner D, Roloff K, Ohnesorge KL . Hair zinc of young children from rural and urban areas in North Rhine-Westphalia, Federal Republic of Germany . Eur J Pediatr 1988 ; 147 : 179-83. Lonnerdal B, Keen CL, Hurley LS . Iron, copper, zinc, and manganese in milk . Annu Rev Nutr 1981 ; 1 : 149-74.  Nakamura T, Nishiyama S, Futagoishi-Suginohara Y, Matsuda I, Higashi A . Mild to moderate zinc deficiency in short children : effect of zinc supplementation on linear growth velocity . J Pediatr 1993 ; 123 : 65-9. Oberleas D, Muhrer ME, O'Dell BL . Dietary metal-complexing agents and zinc availability in the rat . J Nutr 1966 ; 90 : 56-62. Pilch SM, Senti FR . Analysis of zinc data from the second National Health and Nutrition Examination Survey (NHANES II) . J Nutr 1985 ; 115 : 1393-7 .  94  Ploysangam A, Falciglia GA, Brehm BJ . Effect of marginal zinc deficiency on human growth and development. J Trop Pediatr 1997 ; 43 : 192-8.  Prasad AS . Zinc and immunity . Mol Cell Biochem 1998 ; 188 : 63-9.  Puchyr RF, Bass DA, Gajewski R, Calvin M, Marquardt W, Urek K, Druvan ME, Quig. Preparation of hair for measurement of elements by inductively coupled plasma-mass spectrometry (ICP-MS) . Biol Tr Elem Res 1998 ; 62 : 167-82.  Ruz M, Castillo-Duran C, Lara X, Codoceo J, Rebolledo A, Atalah E . A 14-month zinc-supplementation trial in apparently healthy Chilean preschool children . Am J Clin Nutr 1997 ; 66 : 1406-13.  Sandstead HH . Zinc nutrition in the United States . Am J Clin Nutr 1973 ; 26: 125160.  Sandstrom B, Cederblad A, Lonnerdal B . Zinc absorption from human milk, cow's milk, and infant formulas . Am J Dis Child 1983 ; 137 : 726-9.  Sandstrom B, Cederblad A . Zinc absorption from composite meals . II. Influence of the main protein source . Am J Clin Nutr 1980 ; 33 : 1778-83.  Shankar AH, Prasad AS . Zinc and immune function : the biological basis of altered resistance to infection . Am J Clin Nutr 1998 ; 68 : 447S-463S. Shanklin CW, Wie S . Nutrient contribution per 100 kcal and per penny for the 5 meal components in school lunch : entree, milk, vegetable/fruit, bread/grain, and miscellaneous . J Am Diet Assoc 2001 ; 101 : 1358-61 .  95  Silva A, Whitman S, Margellos H, Ansell D . Evaluating Chicago's success in reaching the Healthy People 2000 goal of reducing health disparities . Public Health Rep 2001 ; 116: 484-94.  Skalicky A, Meyers AF, Adams WG, Yang Z, Cook JT, Frank DA . Child food insecurity and iron deficiency anemia in low-income infants and toddlers in the United States . Matern Child Health J 2006 ; 10 : 177-85.  Smit Vanderkooy PD, Gibson RS . Food consumption patterns of Canadian preschool children in relation to zinc and growth status . Am J Clin Nutr 1987 ; 45 : 609-16. Stein AD, Shea S, Basch CE, Contento IR, Zybert P . Consistency of the Willett semi quantitative food frequency questionnaire and 24-hour dietary recalls in estimating nutrient intakes of preschool children . Am J Epidemiol 1992 ; 135 : 667-77.  Treiber FA, Leonard SB, Frank G, Musante L, Davis H, Strong WB, Levy M . Dietary assessment instruments for preschool children : reliability of parental responses to the 24-hour recall and a food frequency questionnaire . J Am Diet Assoc 1990 ; 90 : 81420. Van Wouwe JP, de Wolff FA, Van Gelderen HH . Zinc in hair and urine of pediatric patients. Clin Chim Acta 1986; 155 : 77-82.  Walravens PA, Chakar A, Mokni R, Denise J, Lemonnier D . Zinc supplements in breast fed infants. Lancet 1992 ; 340 : 683-5.  Walravens PA, Krebs NF, Hambidge KM . Linear growth of low-income preschool children receiving a zinc supplement . Am J Clin Nutr 1983 ; 38: 195-201.  Williams PL, Iron status among infants 8-26 months of age in Vancouver and sociocultural/dietary predictors of risk of iron deficiency anemia . Doctoral dissertation in 96  the school of Family and Nutritional Sciences . The University of British Columbia: Vancouver, 2000.  World Health Organization. Global Strategy for Infant and Young Child Feeding,. 2008 . Retrieved January 28, 2008 from http ://www.who.int/topics/breastfeeding/enl  Yip R, Scanlon K, Trowbridge F. Trends and patterns in height and weight status of low-income U .S . children . Crit Rev Food Sci Nutr 1993 ; 33 : 409-21. Young LR, Nestle M . Variation in perceptions of a "medium" food portion: Implications for dietary guidance . J Am Diet Assoc 1998 ; 98 : 458-9  97  CHAPTER III2 HAIR ZINC STATUS AND DETERMINANTS OF HAIR ZINC OF VANCOUVER PRESCHOOLERS AND THE USEFULNESS OF THESE DETERMINANTS AS A SCREENING TOOL FOR MARGINAL ZINC DEFICIENCY  2A version of this chapter has been developed for publication, Vaghri, Z . Barr, S . Wong, H . Chapman, G. and Hertzman, C . (2008) .  98  III. 1 . INTRODUCTION Marginal zinc deficiency (MZD) is the very early stage in the spectrum of zinc deficiency when the body's zinc supply is chronically and marginally short of meeting the physiologic needs (Prasad, 1988) . During this stage most of the clinical signs and definite diagnostic symptoms of zinc deficiency such as dermal lesions, alopecia and severe growth retardation, are absent . The few symptoms present in MZD are either subtle like stunted growth, or unspecific like loss of appetite, which can be attributed to many other health problems. Due to this lack of clear and specific signs and symptoms the deficiency can easily go unnoticed by caregivers and consequently remain undiagnosed by healthcare professionals . Though at present there is no precise information as to the number of people affected, judging from the predisposing dietary factors of MZD, it is believed that this deficiency is the most common form of zinc deficiency in developed countries and worldwide (Prentice, 1993). Marginal zinc deficiency is a particular concern when it comes to children . Its adverse effect on growth and development, subtle as it is, has been well documented (Hambidge et al, 1972, Walravens, Krebs & Hambidge, 1983, Gibson et al, 1989-a and Ploysangam, Falciglia & Brehm, 1997) . It is possible that these ill effects are not confined to childhood, since the effect of early childhood health experiences in later life, on well being, coping skills and competence, is very powerful (Currie & Madrian, 1999). A healthy child population not only speaks highly of the present health status of its society, but also indicates a reduced cost of future health care, increased productivity and an enhanced economy (Sardell, 1990) . Because of the adverse effect of MZD on 99  children's health and well being, and the impact of children's well being on society, MZD is, indeed, an issue of significance for a healthy society. Canadians are among the healthiest of the world's societies (WHO, 1999) . Yet here, in a country with high standards of living, major health disparities still exist . These inequities create a gap within the population and affect the health status of the whole country . The larger the gap, the lower the health status of the overall population (Wilkinson, 1996 and Wilkinson & Marmot, 1998) . In order to reduce health inequities and thus improve further the health status of all Canadians, Health Canada has identified and adopted the "population health approach" as its approach to policy and program development for public health (Public Health Agency, 2002). The population health approach, as the phrase suggests, addresses the health status of the population, as opposed to that of individuals . Since our understanding of what makes and keeps people healthy is continuously evolving, this approach to evaluating health status considers the entire range of both individual and collective factors affecting a population . These elements, 1) focusing on the population and 2) addressing a variety of health determinants, are 2 out of the 8 key elements upon which the population health approach is based (Public Health Agency, 2002). The knowledge of determinants (both social and environmental) of a given health issue can aid population health promotion in 2 important ways . It can be useful in taking measures that can inhibit the emergence and establishment of environmental, economic, social and behavioral conditions known to increase the risk of disease . It can also be useful in constructing the profile of the at-risk individuals/groups which can, subsequently, serve as a detection tool in the hands of health care professionals . While ioo  the former is known as primary prevention and the latter as secondary prevention, they are both considered effective upstream interventions for public health promotion. Upstream intervention, which is central to the population health approach (Federal/Provincial/Territorial Advisory Committee on Population Health, 1999), can aid in directing investments to those areas that have the greatest potential for the enhancement of health . It can also help in maximizing that potential vis a vis cost-saving concerns . The population health approach is grounded in the conviction that the earlier in the causal stream action is taken, the greater the likelihood for gain in population health. With this in mind, the most effective interventions may be those that aim not just at the onset of disease but also at the conditions that pertain prior to the onset. An essential requirement for such intervention is a detection tool . It can often be a biological marker sufficiently sensitive and reliable to detect the disease at an early stage. Such a biomarker should also be one widely recognized and in common use by practitioners. The collective results of supplementation studies confirming the existence of MZD in children of various developed countries suggest that, where protein energy malnutrition does not exist, hair zinc is a good indicator of this deficiency among children (Gibson, 1989) . It is sensitive enough for detection even in the very early stage of zinc deficiency (Hambidge, 1980) . In addition, owing to the structural characteristics of a hair follicle which isolates the element within itself, hair zinc is immune to fluctuations resulting from other physiological events in the body (Hopps, 1977). However, despite its sensitivity, specificity, and relatively prevalent use in research, health practitioners seldom use this biomarker . 101  Alternatively, in the absence of a single sensitive and specific biomarker that is also commonly recognized and widely used, the profile of an at-risk individual can be of great help . Such a profile, constructed from the determinants of that health issue/disease, can be used to construct a questionnaire which when validated can become a screening tool . However, an indicator of the disease is needed in order to identify its determinants and their interactions. Hair zinc can fulfill this role by acting as an indicator of zinc status in studying the determinants of MZD. At present our knowledge of the determinants of MZD is very limited . In Canada the last study conducted on the zinc status of a group of healthy Canadian preschoolers dates back to the 1980s (Smit Vander Kooy & Gibson, 1987) . This deficiency (MZD) was later confirmed through a zinc-supplementation study (Gibson et al, 1989-a) . In more recent years our own study of a group of low-income preschoolers of Vancouver indicated widespread low hair zinc while identifying a few potential determinants of low hair zinc status (Chapter II of this thesis). This has generated in us concern as well as curiosity regarding the overall zinc status of Vancouver preschoolers . Combined, they posed the need for a new and large-scale study that would address these issues. Therefore, with these knowledge gaps in mind and using hair zinc as the biomarker of zinc status, this study was undertaken to examine the zinc status of Vancouver preschoolers through a large and representative sample . In addition, we planned to explore and identify the determinants of sub-optimal zinc status in an attempt to construct a short and simple questionnaire that could serve as a screening tool for low hair zinc/MZD . We intended, then, to validate the developed questionnaire. The specific objectives of this study were : 102  1. To assess the hair zinc status and the extent of low hair zinc in Vancouver preschoolers through a large and representative sample from the city. 2. To explore the relationship between the 5 main demographic and socioeconomic factors (age, sex, ethnicity, family income and maternal education) and hair zinc levels. 3. To examine the relationship between some additional family characteristics (family size, and composition, food security and paternal education) and hair zinc levels. 4. To investigate the relationship between some eating behaviors such as the intake frequency of meat, dairy, cereals and grains, as well as variables such as "being concerned about child's eating", "adequacy of child's eating", "pickiness", and "unhealthy eating" with the children's hair zinc level. 5. To examine the relationship of parental perceptions of some health-related variables (such as overall health, frequency of a child being sick, breast-feeding pattern, and intake of supplements containing iron), and behavior-related variables (such as the scores of activity level and attentional focusing) with the hair zinc of children. 6. To examine the relationship between the indices of growth (HAZ and WHZ scores) and nutrition (WHZ scores) with the children's hair zinc. 7. To determine which of the variables significantly associated with hair zinc, can predict the hair zinc levels. 8. To verify which of the variables associating significantly with low hair zinc status can predict hair zinc status (i .e . low hair zinc). 9. To assess the collective usefulness of the variables identified under objective 8 (and the questionnaire constructed based on this model), in predicting the hair zinc status of 103  children by exploring the classification tables and the Receiver's Operating Characteristic (ROC) curve of the model. It was our belief that the outcome of this study will make an important contribution to the MZD literature by expanding our understanding of the factors in children and their environment and social setting that are associated with (and may therefore influence) their hair zinc status . It will also provide information on the hair zinc/zinc status of the city's preschoolers while making some reference values available for future researchers working on MZD in apparently healthy Canadian preschoolers and preschoolers of industrialized societies in general. I11.2. METHODS AND MATERIALS [II.2.a.Study design The study, which was approved by the Human Ethics Committees of the University of British Columbia (Appendix III-1), was designed as a citywide survey representing all 23 neighborhoods of the city of Vancouver . Daycare and preschool centers were planned as the sources for our recruitment. Through the cooperation of the West Coast Childcare Resource Center we obtained the list of all licensed daycares and preschool centers in all 23 neighborhoods of the city . From this database, depending on the number of subjects needed to represent each neighborhood (more information on our sampling procedure is provided later), we selected a few daycares and/or preschool centers as potential contact centers . In studying data from the 2001 census, it was apparent that there is a wide variation in the socio-economic status (SES) even within the individual neighborhoods . In order to have a fair representation from each neighborhood, they were broken down into smaller units referred to as dissemination areas (DA) (the 104  basic building block of Canada's census geography composed of one or more neighboring blocks having a population of 400-700 people) . The Geographic Information System (GIS) of the Human Early Learning Partnership (HELP) produced maps of each neighborhood with their DAs color-coded indicating the SES status of that DA . The potential participating centers (daycare and preschool centers) were identified on the map of each neighborhood . Care was taken to have a fair representation of each color (each SES) when choosing the centers to participate . A sample neighborhood map is enclosed (Appendix III-2). I11 .2.b. Contacting the potential centers to participate Once the list of potential centers was completed, they were contacted by telephone . During the initial call, the research student, after introducing herself and her affiliation with the University of British Columbia (UBC) explained the reason for her call. Then, the survey, the people, and the institutions involved (UBC and HELP) were introduced . At this point the research student asked for the name of the person in charge of the center and whether the center would be willing to participate, so that a package could be put in the mail for them . Some centers (n=13) expressed an unwillingness to participate for varying reasons . Some, whose boards were comprised solely of parents, expressed a general lack of interest in participating in research . Others indicated that the survey would coincide, and therefore conflict, with other activities already scheduled, such as religious or ethnic holidays . With each refusal a new center, within the same neighborhood DA, was selected for contact . It was also during these phone calls that inquiry was made as to the enrolment number, and the data already provided by the West Coast Child Care Resource Center was then updated . 105  In total, 55 centers agreed to participate, while 13 refused (55/68=81% response rate) . Following this initial contact, the survey information package was mailed to those centers that had expressed an interest in participating . It contained a supporting letter from the Executive Director of West Coast Child Care Resource Centre (Appendix III-3), a letter from the principal investigator of the project (Appendix III-4) and a schematic presentation outlining the sequence of the survey events together with their allotted time intervals (presented as Figure I11.1 in the upcoming sections). I1I.2.c . Meeting with centers prior to the commencement of the survey The survey was planned in 4 waves, each starting and ending in a staggered manner. Waves 1, 2, 3, and 4 covered 6, 5, 5, and 7 neighborhoods of the city, respectively . Second phone calls were made to the participating centers 2 weeks prior to the survey start requesting a meeting with as many childcare workers as possible. During this visit: - The survey and the survey team were introduced to the childcare professionals. - The time line for the survey and the activities involved were discussed. - The significance of their role as the middle person and their impact on the response rate was acknowledged and their cooperation requested. - Their questions were answered and their concerns received . As a result, some centerspecific changes in the survey process were made when necessary. -Contact information of the survey team was provided and communication of the centers with the survey team throughout the process was encouraged .  106  - The "initial letter of contact" (Appendix III-5) was given out for distribution in a week's time. III .2.d. inclusion criteria The inclusion criteria for our survey were : 1) residency in the city of Vancouver, 2) being healthy (not diagnosed with any major or chronic illness), and 3) being between 24-71 months old. III.2.e. Survey Tool The survey tool was a 41- item questionnaire covering 4 subjects : the child's background, family background, child's eating habits, and child's behavior. The following 3 sections outline the process of development and validation of this questionnaire and describe its components. III.2.e.A. Development of the survey questionnaire From a thorough search of the literature, all of the variables that have been shown to correlate with the hair zinc of children were identified . The shortest, clearest version of the question for each relevant variable was then synthesized from existing items . In order to elicit a specific response, (Czaja & Blair, 1996) and also to attain a higher response rate (Passmore et al, 2002), the respondents were given specific response categories to choose from . After constructing a question for each of the variables, all of the questions were then compiled to create the survey instrument. III.2.e.B . Components of the survey questionnaire The following sections will explain the different areas on which the survey questionnaire collected information (in the order that they appear in the questionnaire), and will describe the questions designed for each area . 107  111.2.e.B.1. Socio-Demographic Information  This information was collected through 2 groups of questions arranged under 2 headings: the child's background and family background . Under these headings we collected information on the child's sex, date and place of birth, ethnicity of parents, maternal and paternal education, family income, family size and composition (number of adults and number of children living in the house). Since in our previous study in Vancouver inner city (chapter II of this thesis) we had noticed a significantly different hair zinc level and status in children < 4 and those >4 years, we used this age breakdown for the initial categorization of the subjects of this study as well. The most recent census has reported a median income of $39,400 for the city of Vancouver (Statistics Canada, 2005) . We categorized the income level into 4 categories (increments of 20,000) with 2 below and 2 above this median. III.2.e.B.2. Information on child's breast-feeding history and perceived health status  We asked parents to provide some information on their breast-feeding history through 3 questions; 1) The first question asked them whether or not the child was breast-fed and, if so, to provide the duration of breast feeding by filling in the numbers in 2 blanks (from the time she/he was  months to the time she/he was  months old).  2) The second question asked about whether or not the child received infant formula and, if so, to provide the length of time that formula was fed by filling in the numbers in 2 blanks (from the time she/he was  months to the time she/he  was months old) . 108  3) The third question asked about the age of weaning by requesting the parents to fill in the weaning age in the blank provided (when she/he was  months old).  Using the 3 above-mentioned questions we figured out the duration of "exclusive" breast-feeding. We also requested the caregivers to answer 2 further questions ; one asked parents to describe their child's health using 1 of the 4 adjectives : very good, good, fair and poor. The second asked them to record the frequency (times per month or year) of their children being sick with flu, fever, cold and other common infections . We used both these questions as indices of the children's overall health status and, ultimately as, proxies of their immune system function. III.2.e.B.3. Information on supplement intake  In order to obtain information on the use of supplements, as to their brand name and frequency of use, a 3 part question was designed and placed under the first category of the survey questionnaire, namely the "child's background". 1)The first of the 3 parts asked whether or not the child took or takes any supplement(s). 2) The second part inquired as to the period of time over which the supplement had been consumed (e .g. from the time that child was 18 months up until she was 4 years old). 3) The third part asked for the name(s) of the supplement(s) given to the child. In cases where multi-mineral supplements were reported, we constructed a list and visited local pharmacies and grocery stores to examine the supplement shelves so as to verify if and how much iron and zinc they contained .  109  III.2.e.B.4. Information on children's eating behavior  This information was collected under the heading of "Child's Eating Habits" . In an effort to obtain a high response rate, we, in general, used questions that were straightforward and easy to answer without requiring a lot of time commitment . This basic principle applied to our dietary questions as well . We did not use a comprehensive assessment of diet . Under the heading of "child's eating habits", there were 5 major sections each of which contained multiple (between 1-10) questions to collect information on the following topics; • The first section contained 7 questions that assessed the parents' perception of their child's eating behavior. It commenced by asking the parents if they are concerned about their children's eating in general, then provided 7 stated options from which to choose . These options were some of the common eating characteristics/problems of young children (such as being a picky eater, not eating enough etc .) identified in our meetings with community dietitians . The parents could select more than one behavior . Space was also provided in this first question for concerned parents to write about any additional eating behavior/problem in their children which was not provided as an option. • The second section was a single question collecting information on the consumption frequency of fluids (everything but water) intake . Although no study of zinc has assessed any associations between poor zinc nutriture and fluid consumption, such an association has been documented for some other micronutrients (Guenther, 1986 and Williams, 2000) . Estimation of fluid consumption was done to enable us to explore these associations . 110  •  The third section comprised of 6 questions, collected information on the consumption frequency of meat, fish and poultry.  •  The fourth section comprised of 10 questions, collected information on the consumption frequency of cereals and grains.  • The fifth section comprised of 3 questions, collected information on the consumption frequency of dairy products. The 3rd, 4th, and 5th sections were designed to obtain information on the bioavailable sources of zinc i .e . meat, fish and poultry, and sources of inhibitors of zinc absorption i .e. dairy as a major source of calcium and cereals and grains as major sources of phytate. These 3 sections had asked questions on the usual consumption frequency of a variety of foods belonging to these food groups . The questions provided response options in the form of boxes (consumed daily/weekly/monthly) to check and blanks to fill in (once, twice, 3 times . . .). Portion size was not ascertained . As with the other parts of the survey questionnaire, the dietary section was subjected to multiple revisions and changes based on the outcome of our multiple focus groups. III.2.e.B.5. Food Security Questions  This section comprised 3 pre-validated questions as to how often the family members experienced some aspect of hunger and/or concern about food, in order to measure food security in the survey families . They were adapted from the Cornell/Radimer food insecurity questionnaire (Olson et al, 1997) . These pre-validated questions are highly sensitive, widely used, and likely to identify families affected by food insecurity and hunger . The National Population Health Survey, NPHS, (1998-1999) also used these questions . Similar to the scoring system used in NPHS, the respondents in 111  our survey were considered to be living in a food insecure household if they responded "Yes" (either sometimes or often) to at least 1 of the following 3 question/statements (Che & Chen, 2001); In the past 12 months: Q1: Were you and others worried that food would run out before you got money to buy more? Q2: The food that you and others bought didn't last and there wasn't any money to get more. Q3: You and others, because of not having money, could not eat balanced meals. Three options were provided in choosing a response : "Often", "Sometimes", or "Never". III.2.e.B.6. Behavioral Information  The Child Behavior Questionnaire, CBQ, is a behavior checklist developed by Dr. Mary Rothbart, University of Oregon (Rothbart, 1996) . It is a highly differentiated assessment of temperament in early to middle childhood . In its original form, this tool measures 12 domains of behavior through 195 questions . The original CBQ was shortened in 2003 to measure the same domains of behavior through a reduced number of questions (n=85) (Putnam & Rothbart, 2003) . Through a series of email communications with Dr . Rothbart and Samuel Putnam, a collaborating scientist who has had a key role in development of this tool, we took this version and extracted the questions pertaining to 2 behavior domains of interest in zinc deficiency : activity level and attentional focusing. The first score is the measurement of the level of gross motor activity including rate and extent of locomotion . The other score is defined as the measure of the tendency to maintain attentional focus upon task-related channels (Rothbart et al, 2001) . 112  The questions addressing these 2 domains were further modified and shortened (by Dr . Rothbart's laboratory) in order to be more suitable for a survey . The final shortened questionnaire comprised of 13 questions measuring activity level (7 questions) and attentional focusing (6 questions). As with the original tool, the questions measuring the same behavioral domains were separated from each other by a question measuring another behavior in order to prevent parents from getting into a mental groove that could potentially affect the validity of their responses. III.2.e.C. Verifying the face validity of the questionnaire We assessed the face validity of the survey questionnaire through 5 focus groups set up in Hastings Sunrise, University Lands, Kensington Cedar Cottage, Strathcona, and Downtown. These neighborhoods were chosen because collectively they included every level of literacy in the city with Hasting Sunrise and University Lands at the low and high ends of the literacy spectrum, respectively . Each group included 6 parents . After each group met, a modified version of the questionnaire was developed and tested in the next focus group . This was repeated until no new information emerged. Efforts were made to include parents of different ethnicities in each group. Participants were recruited through day-care centers and personal contacts known to the research student . The snowballing method was also used . At the time of recruitment, the research student explained to the potential participants both the study and the organizations involved (UBC and HELP) . Individuals willing to participate in a focus group were asked to provide contact information by filling out a form (Appendix III-6). An information sheet containing the day and time of the focus group, along with the 113  address of the focus group location, and the research student's phone number was provided to all recruited participants (Appendix . III-7). The recruited parents were called the day before to remind them of their appointment. The focus groups began with the research student, as the group facilitator, reexplaining the purpose of the focus group and assuring the group of the confidentiality of the information and asking the participants to sign an informed consent form (Appendix III-8) . The parents were reimbursed for their transportation expenses (when applicable) and a gift certificate was presented to them at the end of each group . The outcome of the final focus group was sent to the supervisory committee for their final comments and then the final questionnaire (Appendix III-9) was given to 2 translators for translation into Punjabi and Cantonese . These questionnaires and accessory forms were then given to 2 other translators to translate them back to English . The research student reviewed the back-translated forms and met with the first set of translators to go over the problematic areas of the first translations . After a few rounds the final translated forms were produced (Appendix III-10 and 11 for the Punjabi and Cantonese survey questionnaires, respectively). 111.2 .E The survey sequence of events The survey began by sending home the "initial letter of contact" . One to 2 weeks after sending this letter to the parents, the centers distributed the "survey packages". These contained : "subject's information letter" (Appendix III-12) reiterating the information provided in the "initial letter of contact", a "consent form" (Appendix III13), and the "survey questionnaire" . A cardboard box labeled "UBC Survey" on all 4 114  sides was placed at the "sign in/sign out" station at each center as a continuing reminder to the parents to drop off their completed survey questionnaires along with their consent forms . A log sheet was provided beside the box for parents to sign their study code (printed on each envelope) together with the date of their survey return (Appendix III14). One week after distribution of the survey packages, a "reminder letter" (Appendix III-15) was sent to the homes of the children encouraging and reminding the caregivers to return their completed surveys . The survey forms continued to be received during the second week . It was also during this week that the survey team and the daycare authorities tried to set up a time convenient for both parties to visit the centers in order to take anthropometric measures and collect the hair samples of the children for whom a consent form had been received . Earlier in the week of our visit, a sign was posted at various locations in the centers in order to announce our arrival and ensure the attendance of the participating children on that day (Appendix III-16). Figure III.I is a schematic summary of this sequence of events . As shown, upon completion of the first 4 waves, a fifth wave, called wave-C, was designed to re-initiate the survey in a few neighborhoods where the total number of families responding to the survey did not meet the target numbers . Wave-C completed the survey in the 5 neighborhoods of Kensington Cedar Cottage, Hastings Sunrise, Grandview Woodlands, Marpole and Riley Park. The recruitment method for some wave 1- 4 neighborhoods and most of the waveC neighborhoods was different from our usual method . For them, where the rate of response was low and did not meet the target number of that neighborhood (KCC, SS, 115  GW, RP, ML, KO, HS, OR,), we contacted the coordinators of parent-child programs in community centers of those neighborhoods and requested their permission to drop in . We attended coffee-time and playtime programs where recruitment, survey questionnaire completion and sample collection were all accomplished together . We recruited 121 (17%) survey participants in this way . Through a series of one-way analyses of variance, ANOVA, we then compared the SES (age, sex, ethnicity, maternal education and family income) of these children and the children recruited by our usual recruitment method . No significant differences were observed .  116    Feb/2T Mar/4  Mar/29  A1)rA3  Apr/17  Mav/01  Mavl15  JunI05  L  Jun/A9 i Wave 1 : DT, WE, HS, SA, SY . UL  Visiting Centers Wave 2 : GW, RC, KCC, WP, MP Vsiiing  Coen Wave 3 : VF, KY, SS, SC, OR  4 Virg  Wave 4 : AR, DR, RP, KO. KL, ML, FV  SY Plg  PNL : Parents PbtlE'irrdion Lette s SY Pk S"tavey Package  Visiting Canters  Vining  Wave C : K.CC, ML, RP, GW, HS  Caters  DT= Downtown, We=West end, HS=Hasting Sunrise, SA=Strathcona, SY=Shaughnessy, UL=University Lands, GW=Grandview Woodlands, RC=Renfrew Collingwood, KCC=Kensington Cedar Cottage, WP=West Point Grey, MP=Mount Pleasant, VF=Victoria Fraserview, KY=Killarney, SS=Sunset, SC=South Cambie, OR=Oakridge, AR=Arbutus Ridge, DR=Dunbar, RP=Riley Park, KO=Kitsilano, KI=Kerrisdale, ML=Marpole, FV=Fariview,  Figure III.I. Schematic presentation of survey waves and their time-lines .  117  III .2 .g. Sample Size Calculation Initially we had planned to divide the returned surveys, randomly, into 2 equal batches to be used as follows: 1) One of the batches was to be used for exploring the socioeconomic, sociodemographic and food, health and behavior-related variables having significant association with low hair zinc . Then using the logistic regression analysis, the variables predictive of hair zinc were to be identified and used to construct a mini-questionnaire. 2) The second batch was to be used to validate this questionnaire. However, upon completion of the survey, our preliminary analyses of the first batch of surveys indicated that the R2 values of the variables significantly associating with hair zinc were too small to construct a useful questionnaire . It was our understanding that any questionnaire constructed based on these variables, would not have great predictive power and as such would not be of much use in practice . Therefore, the plan was modified and all of the returned surveys were used as one batch to increase the power of the study to explore as many variables as possible . The following is a description of the sample calculation for the initial plan. Based on the total number of variables that we had included in our survey questionnaire, we predicted that the likely number of variables for inclusion in the miniquestionnaire would not exceed 10 . Using this hypothetical maximum number, the following reasoning was used in estimating the sample size and the number of surveys required for distribution .  118  •  As a general rule of thumb 10 affected subjects (i .e . subjects with low hair zinc) are considered an adequate size to test each variable through regression analysis (Van Belle, 2002) . Therefore, for 10 variables to be tested, 100 affected children were going to be required.  •  Considering 100 subjects for development of a questionnaire, and 100 for validation of the questionnaire, 200 affected subjects was required to complete both tasks.  •  Results of our previous work in the city (Chapter II of this thesis) had indicated an occurrence rate of low hair zinc at 46%, while another Canadian study of this age group has reported 24% (Smit Vanderkooy & Gibson, 1987) . Taking the midpoint as the expected rate of occurrence gave us 35% . [46+24]/2= 35%.  •  Assuming, then, a 35% occurrence rate, in order to have 200 affected children we needed 572 children overall . [200x100]/35=572  •  Assuming 50% response rate, to have 572 returned required that 1144 be sent out.  [572 x100]/50=1144 •  Provision of 12% additional subjects to compensate for the incomplete, lost surveys and etc, gave;  [1144x12]/100=138 Therefore the required number were : 1144+138 = 1282 Rounding upward, we needed to distribute 1300 surveys . Table III.I is the summary of the number of surveys to be given out in each one of the 23 neighborhoods. Overall, 2550 survey packages were sent out and 772 were returned (30% response) . Out  119  of the returned surveys 719 had a competed survey questionnaire and hair sample result (attrition rate of 7%) .  120  Table.I1I.I: A summary of the estimated number of survey packages needed to be given out in each neighborhood .  Neighborhood Shaughnessy South Cambie Strathcona Oakridge Arbutus-Ridge Kerridsale West point Grey University lands West End Marpole Downtown Fairview Dunbar Victoria-Fraser view Riley park Killarney Mount Pleasant Kitsilano Grandview-Woodlands Hastings-Sunrise Sunset Renfrew-Collingswood Kensington Cedar Cottage Total Vancouver  Number of preschoolers in the neighborhood 270 310 345 375 410 420 510 515 655 810 835 955 995 1105 1200 1215 1260 1355 1525 1875 2285 2470 2670 24365  Percentage of Vancouver preschoolers 1 .11 1 .27 1 .42 1 .54 1 .68 1 .72 2 .09 2.11 2.69 3 .32 3 .43 3 .92 4.08 4.54 4 .93 4.99 5.17 5.56 6.26 7.70 9 .38 10.14 10.96 100  Number of surveys to be given 14 17 18 20 22 22 27 27 35 43 45 51 53 59 64 65 67 72 81 100 122 132 142 1300  121  III.2.h. Pre-testing the survey After all the survey components were in place, and prior to the survey commencement, the questionnaire, as well as the designed plans and procedures, were tested in 1 of the daycares of the Shaughnessy neighborhood . The survey began and ended as outlined . Based on the feedback, and our own observations, a few minor changes were made . Out of 50 surveys given out, only 10 were returned (20% response) giving us a first indication that our expected 50% response rate for the sample size calculation was optimistic . We then decided (depending on the size of the required sample from each neighborhood) to add from 1 to 4 centers to the centers list. IIL2,i . Reliability of the questionnaire Having the questionnaire completed by 10% (n=72) of the survey respondents at 2 different points in time allowed assessment of the reliability of the questionnaire . The second questionnaire, relying on the log sheet and the return of the first questionnaire, was given out in such a way that the subjects had 3 weeks between filling out the 2 questionnaires . The following steps were taken in order to complete this part of the study. Seventy-two subjects out of the survey participants were randomly sampled and the reliability package was given to them through the centers . This package contained: another copy of the questionnaire, a covering letter (Appendix III-17) and a gift certificate . The return from the reliability study was very high . To get the desired number of 72 subjects we had to send out 80 reliability packages only (return rate 90%) . The results of the reliability test for different components of the survey questionnaire will be presented in their respective result sections . 122  1II.2. j. Incentive We included small incentives for participants at various points of the survey . Gift certificates (in the value of $15) were provided for the participants of the focus groups and the reliability study . In addition, at the end of the survey, a raffle of 35 baskets was held (one for every 22 surveys returned thus giving a 5% chance of winning) with items relevant to preschoolers and caregivers . They were delivered to the winners' respective centers, along with a letter that requested them to call the survey line and confirm the receipt of the basket (Appendix III-18) . A few photographs of the winners are enclosed (Appendix III-19). IIL2.k. "Thank you" letters The gratitude of the research team was expressed to all of the participating parents through a thank you letter given to all participating children on the day of the sample collection (Appendix III-20), along with a sticker given to each child . Upon completion of the survey, a thank you card was mailed to all participating centers (Appendix III-21) as well as to West Coast Family Services (Appendix III-22). 111.2.1. Educational pamphlets To express further appreciation to the survey participants, an educational pamphlet on zinc, its dietary sources and the amounts required for children of different age groups, was prepared and supplied to all participating centers to be made available to their families . The on-line Practice-based Evidence in Nutrition, PEN, developed by Dietitians of Canada (Dietitians of Canada, 2006) was consulted for preparation of this pamphlet . A draft of this pamphlet was edited by Karol Traviss, the dietetic education coordinator, and Dr . Susan Barr, a member of the supervisory committee, and was then 123  put through 3 focus groups held in Kiwassa Neighborhood House and Marpole Family Place where educational attainment is generally low . Through these focus groups the level of literacy and ease of comprehension of the pamphlet were tested and the necessary adjustments made prior to printing the final product (Appendix III-23). I11 .2.m . Biochemical data A pair of stainless steel scissors wiped with ethanol swabs (between subjects) to avoid cross contamination was used to cut the hair samples . Following the protocol proposed by the International Atomic Energy Agency, IAEA, (Deppisch et al, 1999), and adopted by the Center of Disease Control, CDC, and the US Environmental Protection Agency, EPA, they were cut from 3-4 locations at the back of the head and as close to the scalp as possible . Only the first 1-2 centimeters proximal to the scalp were taken, the remainder being discarded . The collected samples were each placed in a pre-labeled coinenvelope . They were all inspected and cross-matched with a list of the consented children of the center visited that day . Samples were delivered to JR Labs in Burnaby on a weekly basis along with a log sheet listing all samples with their study code (Appendix III-24). The samples were processed (Puchyr et al, 1998) and analyzed for zinc content by inductively coupled plasma mass spectrometry, ICP-MS (EPA, 1994) . They were analyzed as single samples with a duplicate after every 10 samples . The within assay CV for this method was 3 .3% at a mean of 157 .3 µg zinc/g hair, while the between assay CV was 6.0% at a mean of 157 .7 gg zinc/g hair. The cutoff of 70 µg/g, the one most commonly used cutoff which has been proposed by Hambidge (1972) and defined as 3 SD below the adult mean, was used to define low hair zinc in this study . 124  III.2.n. Anthropometrics Trained research personnel measured the children's height and weight and recorded them on a log sheet prepared in advance (Appendix III-25) . A measuring rod, aligned and secured on any wall in the room that was flush with an uncarpeted floor, was employed to measure the height as each child stood shoeless with back to the wall, feet together and looking straight ahead . Heights were measured to the nearest 0 .1 cm . Body weight was measured to the nearest 0 .1 kg using a lithium electronic scale (Taylor Precision Products, L .P, Model 7300) with the scale placed on a hard floor and children standing on it shoeless and wearing light indoor clothing only . All the readings were done in multiples until 2 consecutive numbers were identical or not different by more than 0 .5 unit of measurement . When 2 consecutive readings were not identical, the average of the 2 readings was recorded. The reference tables of the Center for Disease Control (CDC, 2000) were used to calculate the weight for age (WAZ), height-for-age (HAZ) and weight-for-height (WHZ) z scores of all children. LII.2.o . Statistical analyses The data were analyzed using the Statistical Package for Social Science, SPSS, (version 13 .0, 2005) . Using this software, descriptive statistics (such as mean and standard deviation) were computed for all variables. In order to assess the relationship between each of the variables of interest (5 main socioeconomic and socio-demographic variables, some family characteristics, frequency of the intake of some food groups, parental perceptions of some nutrition related, health-related and behavior-related variables and intake of iron supplements) and 125  hair zinc levels, univariate regression analyses were carried out with hair zinc as the outcome variable. In order to examine how the variables associating significantly with hair zinc collectively predicted hair zinc, a forward stepwise linear regression analysis was carried out with hair zinc level as the outcome variable and all of the variables that were significantly associated with hair zinc level as predictors . The inclusion criterion for predictors in this analysis was having p<0 .05 in the adjusted regression analysis. In order to identify the variables that may have usefulness in predicting the hair zinc status of children, and as a result the construction of a screening questionnaire, we set up a logistic regression analysis with hair zinc (binary data) as the independent variable and all of the variables that associated significantly with low hair zinc as predictive variables. The inclusion criterion for this analysis was having p<0 .05 in the chi square analysis of the occurrence of low hair zinc, when children were classified based  on a given variable (for example being described as "unhealthy eater" was included in this analysis because the occurrence of low hair zinc in children who were described as "unhealthy eater" was significantly higher than that in children who were not described so). In cases of the variables pertaining to the consumption frequency of food groups as well as the scores of 2 behavior domains, (which were continuous data), variables were inserted into the logistic regression analysis only if the difference in the consumption frequency of a given food or the score of a given behavior of the low hair zinc and normal hair zinc children was significant (p<0.05) . For both these regression analyses (the step-wise and logistic regression analyses), the collinearity of the variables inserted into the models were verified . Collinearity between two components of a regression 126  model could "blow up" the standard deviation and result in elimination of one of the two variables from the final model produced. In order to examine the utility of the logistic regression model produced in predicting the hair zinc status of children, the classification table and the Receiver Operating Characteristic (ROC) curve for the fitted logistic regression model was constructed. The differences between the hair zinc levels of different sexes and age groups were verified using ANOVA, followed by Tukey's post hoc analyses, when applicable. When the outcome variable and the variable of interest were both in categorical form, the differences were examined using Chi square analysis . The groups with different occurrence of low hair zinc were differentiated using continuity corrected chi-square (adjusted chi square) (e.g . difference in the occurrence of low hair zinc between different ethnic groups) . The chance of having low hair zinc for children in different categories (such as for boys and girls) was computed using risk analysis statistics . For all analyses the level of significance was set at <0 .05.  127  111.3. RESULTS III .3.a. Survey population A description of the study population has been given in Table III .II. As shown, boys and girls were equally represented, while the representation of younger (<4 years old) preschoolers was slightly less than the older (>4 years old) ones. As Table III .II indicates children with Caucasian (42%), Chinese (31%) and East Indian (8%) ancestry made up the 3 main ethnicities of the survey . More than half of the survey mothers (58%) held a college or university degree, while 7% of them had not completed high school . Income of the survey families varied within the range of <$20,000->$60,000/year with the largest portion of the families at >$60,000 and 11% at <$20,000/year. Anthropometric data indicated normal growth and nutritional status for the survey population . About 50% of HAZ and 51% of WAZ scores of the study population were between -0.67 and 0 .67, corresponding to the 25 th and 75 th percentiles. The rate of stunting (HAZ<-2) and malnutrition (WAZ<-2) (Walravens, Krebs & Hambidge, 1983 and Gibson et al, 1989), at 2% (n=13) and 2 .3% (n=16), respectively, was not higher than the expected baseline .  128    Table III. II. Socio-demographic description of study population (N=719). Number (%)* Socio-demographic variables Sex Boys 360 (50) Girls 359 (50) Age < 4 years 305 (42) >4 years 400 (56) Missing  14 (2)  Ethnicity Caucasian Chinese East Indian Other ethnicities t  305 (42) 223 (31) 54 (8) 134 (19)  Missing  3 (0.0)  Maternal education <High school High school Some college College or university degree Missing  8 (1)  Income <20,000/year 20-39,000/year 40-60,000/year >60,000/year  67 (9) 120 (17) 94 (13) 310 (43)  Missing  	Mean Anthropometrics ** HAZ scores WAZ scores WHZ scores  49 (7) 72 (10) 170 (24) 420 (58)  128 (18)  ±SD 0.09 + 1 .0 0.24 ± 1 .2 0.25 ± 1 .6  Indicate percentages of the total survey population  OOther ethnicities : 32 Filipinos, 19 Japanese, 18 identified themselves as Asians, 11 ..Vietnamese, 10 Latin, 7 Korean, 6 Aboriginal, 5 Middle Eastern, and 26 miscellaneous. Only 662 children had anthropometric data. HAZ: Height-for-age Z scores WAZ : Weight-for-age Z scores WHZ : weight-for-height Z scores  129  IIL 3.b . Hair zinc status of the survey population Hair zinc of the survey population was normally distributed with a mean (+SD) of 116+43 µg/g, while the hair zinc of the children varied within a wide range of 17-297 tg/g (Figure III.II). Figure III.III is a schematic comparison of the hair zinc distribution of the survey with that of the inner city children (Chapter II of the thesis) . The mean (+SD) of the survey population was significantly higher than that of the inner city children and corresponded to the 90 th percentile in the hair zinc distribution curve of the inner city children .  130    6o —  20 —  Mean=116 SD=43 N=719  0 .00  50 .00  100 .00  150 .00  200 .00  250 .00  300 .00  Hair Zinc (uglg)  The shaded area depicts the fraction of population with hair zinc below cutoff value of 70 pg/g.  Figure I11.II. Frequency distribution and descriptive statistics of the hair zinc level of the study population .  131    60-  Survey Children  Inner city Children  I 0 .00  50 .00  100 .00  T 150 .00  1 200 .00  250 .00  300 .00  Hair Zinc (uglg)  The arrow demonstrates the mean of survey corresponding to the 90th percentile of inner city children.  Figure III.III. Comparison of hair zinc distribution of the present study with that of low -income inner city preschoolers in Vancouver.  132  The overall prevalence of low hair zinc was 17% (121/719) (Table III.III) . As displayed, there were inter-neighborhood differences in the mean (±SD) of hair zinc, and the occurrence of low hair zinc . However, due to the lack of statistical power, we refrained from any further analysis of data at the neighborhood level . There were also differences in the response rate of neighborhoods, in some cases as much as 10-fold. Some of the possible contributing factors to this fluctuation in response rate could be education level of the caregivers, centers' and parents' previous experience with research, and their previous knowledge of or exposure to HELP, one of the two key institutions involved in the survey .  133      0  N  x  n  O  0  0  ^  Q C/]  CL)  '0 O 0  N N  O O  n  00 O\  N  M  ~to)  2  e  ly oo _ M  o0  00 O  N c  00  71-  d••  11'  00 ~-  0  O  CV  71-  N OCN  Cs-  N  N Os O VO  -. N M N N  N a1 C  NC N  c  00 M c )  ,--n  '•1 -  + b `-' N M O N '" N j g- c -• O II^ II oo II C7  •- a) C) 0 •a  N a\ oo  --  N  a) ~II II I) II  ^ 00  •-- N 71- 00 00 O1 M VD 01 N- --. \ -  ^  s~  -" M >~ II  '~ O  a)  ^  ^  II MImoN 7)  II  C II 0  O")  M M O a1 N 71- N 2  II  ~^`O'3 II  i  O -.,  N  M .--~  N00 ~ ~-  N  -H +I -H +I -H +I -H -F{ -H -H N M 00 00 00  N~-  -H -H -H -H -H -H -H -H N O\ ^" N N_ M -  00  M v)  O  d  N  N --~ VD 00 O\ 71- N 00 00 01 00 N vn 71' op N 0 01 N M M N N N -+ M N M M N --+ M  .' .  (NICV~  +I -H -H -FI  00  M •-,  a)i v  M -- ~n  CNI M (T CV  O1 O 0 N 00 N  v) N ~O  a)  V'1  II O  a)  U  (2 '  cd  II  o ^N  a~i  0 O C CI)  N v  III.3.c . Associations of the 5 main socio-demographic and socio-economic variables with hair zinc Table III .IV summarizes the hair zinc level together with the occurrence of low hair zinc based on the 5 socio-demographic and socio-economic variables . As shown, there was a significant difference in the hair zinc of the children as well as the occurrence of low hair zinc based on age, ethnicity, and maternal education . Sex-based differences were also observed in the occurrence of low hair zinc . However, the mean hair zinc for the 2 sexes was comparable (118+39 and 113 +47 µg/g for boys and girls, respectively, p=0 .12) . The age-based differences observed in the hair zinc level and the occurrence of the low hair zinc will be elaborated upon later. As Table III.IV indicates the rate of low hair zinc among different ethnicities varied within a range of 0% for East Indian children to 26% for Chinese children, while the Caucasian children had a rate in between (13%) . The hair color was not associated with the zinc level (R2 =0 .00, p=0 .22) and the hair zinc level of dark (black and brown) and light (blond) hair samples were statistically comparable (116+44 versus 112+40 gg/g for dark and light hair samples, respectively, p=0 .22) . There were no significant differences in the ratio of different age or/sex groups among these 3 ethnicities (Appendix III-27). The hair zinc level as well as the occurrence of low hair zinc in children of the 2 lowest categories of maternal education was significantly different from that of the children in the 2 highest categories . The income level was not associated with the hair zinc level or the occurrence of low hair zinc among the survey children . 135  The results of the reliability test of the survey tool indicated substantial to perfect agreement (Saw & Ng, 2001) between the socio-economic and socio-demographic data collected at 2 points of time (R2 varying within the range of 0 .67-1 .00, for income, maternal education, date of birth, ethnicity, and sex) .  136    Table III.IV. A socio-demographic and socio-economic description of the study population and the hair zinc levels (X±SD) and the occurrence of low hair zinc based on these variables among them (n=719). Variables Sex Boys  n # (%) 360 (50)  Hair Zinc (µg/g) Mean +SD F 118 ± 39  Missing  < 4 years  359 (50)  Missing  0.12  305(43) 400(57)  0.00  124 ± 40 b  42(11) Y  41(13)x 58 (26) '  East Indian Other ethnicities t  54 (8) 134 (19)  154 ± 37° 117 ± 41 a  0 (0) z 21(16) "  53 (7) 72 (10)  96 ± 49 a 104 ± 45 a  16(30) " 22(30) x  170 (24) 416 (59)  120 ± 43 b 118 ± 41 6  20(12) Y 61(15) Y  113 ± 50 113 ± 42  15 (22) 23 (19)  16  Income <20,000/year 20-40,000/year  Missing  27  0 .00  0.00  21  0.00  4 .0  0 .26  8  67 (11) 120 (20)  0 .9 40-60,000/year >60,000/year  0 .00  3  7.0  Missing  0 .00  14  116 f 41 a 104 ± 43 b  Some college College or university degree  27  77(25) "  104 ± 45 a  305 (42) 223 (31)  Maternal education <High school High school  0.00  0  Ethnicity Caucasian Chinese  Missing  9.7 76 (21) Y  113 ± 47  39 >4 years  P  45 (12) " 2.5  Girls  p  Low hair zinc . # (%) X2  94 (16) 310 (53)  121 ± 46 118 f 41  0.43 14 (15) 43 (14)  128  * Low hair , zinc=Hair zinc<70pg/g. t Other ethnicities : 32 Filipinos, 19 Japanese, 18 self-identified as Asians, 11 Vietnamese, 10 Latin, 7 Korean, 6 Aboriginal, 5 Middle Eastern and 26 miscellaneous. a,b,c : Indicate statistically significant differences in the hair zinc level of different categories of a given variable, verified by one-way ANOVA followed by Tukey's post hoc analysis, if applicable. x,y,z : Indicate statistically significant differences in the occurrence of low hair zinc in different categories of a given variable, verified by adjusted Chi square analysis, if applicable . 137  There were 705 children who had both demographic information (full birth date and sex) and biochemical data available . The hair zinc mean±SD was 115+43 µg/g. The scatter plot of children's hair zinc against their age indicated a trend of increasing hair zinc with increasing age (Figure III .IV) . The sample collection time (March-June) was not associated with the hair zinc level (R2 =0 .01, p=0 .07) (no trend was observed between the hair zinc level and the month of year that sample was collected). To explore these differences further, the children's age was categorized in increments of 12 months . We observed a significant increase in hair zinc upon completion of the fourth year of life (from 107+42 µg/g for children in 36-<48 months category to 125+40 µg/g, for children in 48-<60 months category) and remaining high thereafter (Table III.V). This trend applied to both sexes, individually and collectively. The sex-by-age interaction on hair zinc level was not significant as verified by ANOVA (R2 =0 .05, P=0 .31).  138    0  24  48 Age (month)  72  96  Figure III. IV. Scatter plot of hair zinc against age of the children (n=705) .  139  Table III.V . Influence of age on hair zinc (µg/g) of boys and girls of the survey (n =705).  Age (months)  N  Boys Hair zinc (µg/g)  n  Girls Hair zinc 0µg/g)  n  Hair zinc (µg/g)  24-<36  103  98+49'  43  105+46 "  60  93+52  36-<48  203  107+42 "  106  107+39 "  97  107+46 "y  48-<60  234  125+40y  116  125+37 "  118  124+43 z  60-<72  165  123±40"  88  127+34 y  77  119+45 Z"  All  705  115+43  353  118+39  352  113+47  In total 705 children had both date of birth and sex recorded and hair zinc data available. There were no sex-based differences in hair zinc level overall or by age group. Superscripts x, y, z . . . are used to indicate statistically significant differences between the consecutive values in a given column. Values in a column not having a common superscript are statistically different (P<0 .05).  140  In agreement with the changes in the mean of hair zinc with age, for both sexes, a significant decrease was observed in the occurrence of low hair zinc with age . In both sexes, a significant decrease was observed in going from children in the age category of 36-<48 to the category of 48-<60 months of age, as verified by Chi square analyses (for boys: 19% versus 8%, x2 =5 .2, df=1, p=0.02 and for girls : 25% versus 11%, x2 =7.0, df=l,p=0 .01) (Figure III.V). The occurrence of low hair zinc appeared to plateau after completion of the 4th year of life in both sexes. In addition, as represented by the 2 bars in Figure III .V, the overall rate of low hair zinc was significantly higher in girls than in boys (21% and 12%, respectively, x 2 =8.6, df=l, p=0.00). The sex-by-age interaction on occurrence of low hair zinc was not significant as verified by ANOVA (R2 =0.05, P=0 .31). The risk analysis of having low hair zinc for the 2 sexes indicated a significantly higher risk for girls versus boys (OR=1 .9, 95%CI [1 .2, 2.8]) and for children <4years versus those >4 years (OR=2 .9, 95%CI [1 .9, 4.3]) .  141    45 -  - 35  24-x36  36-<48  48-<60  60-<72  Boys Girls  Indicates significantly different occurrence of low hair zinc for two sexes (X2 =8 .6, df=l, p=0 .00)  Figure qI.V. Comparison of the occurrence of low hair zinc in different age categories of girls (the upper line) and boys (the lower line) .  142  The associations between hair zinc level and each of the 5 main demographic or socioeconomic variables (age, sex, ethnicity, maternal education, and family income) were assessed through univariate linear regression analyses . Table III.VI summarizes the results of these analyses . As shown, age, ethnicity and maternal education indicated significant associations with the children's hair zinc level . These associations remained significant even when all 5 socio-demographic and socio-economic variables were inserted in the regression model simultaneously . The overall model accounted for 12% of the variability in hair zinc (R2=0 .12, p=0 .00) .  143  Table III .VI . Relationship between hair zinc and the 5 main socio-demographic and socio-economic factors in the study population (n=719).  Variable  B " (95% CI)  R2  P  Age (months)  0.9 (0.6, 1 .2)  0.06  0.00  Sex  -5.1 (-11, 1 .3)  0.00  0.12  Ethnicity Caucasian Chinese East Indian Other ethnicities  0 .08  0 .00  116 (112,121) 104 (99-110) 154 (143-165) 117 (110-124) 0 .02  0.00  0.01  0.42  Maternal education <High school High school Some college Held a degree  Family income <20,000/year 20-40,000/year 40-60,000/year >60,000/year  96 (85, 108) 104 (94, 114) 120 (113, 126) 118 (114, 123)  113 (102, 123) 113 (105, 121) 121 (112, 130) 118 (113, 123)  *Bs are unstandardized coefficients. When all 5 variables were inserted in the regression model, simultaneously; age, ethnicity and maternal education remained as significant variables of this model . For the overall model R 2 =0 .12, P=0 .00 .  144  III.3 .d. Association of some family characteristics with hair zinc of the children Table III. VII presents data on some family characteristics and the hair zinc level and occurrence of low hair zinc based on these characteristics . About 14% of the survey families were food insecure (as assessed by a "yes" response to any of the 3 questions). As the table indicates, most of the fathers (59%) were university graduates while 8% had not completed high school . About 72% of the survey families were 2 adult households, while 10% of the survey children lived in a lone-parent household and 18% in households with more than 2 adults . The majority of survey families had 2-4 members . About 24% of survey families had between 5-7 members, and a small fraction (2%) were larger families having >8 members . Significant differences in the hair zinc level were apparent based on "being in a single parent household", "number of adults at home" and "family size". Differences in the occurrence of low hair zinc were observed based on the number of adults at home and being in a single parent household, only. The results of the reliability test of the survey tool for collecting data on family characteristics indicated a good agreement between the 2 sets of data collected by the questionnaire at 2 points of time (R 2 varying between the range of 0 .64-0 .83).  145  Table III.VII. Description of some family characteristics as well as the hair zinc level and occurrence of low hair zinc based on these characteristics (n=719). Variables  n # (%)  Ilair Zinc (µg/g) Mean =SD F  p  Low hair zinc # (%) y'-  p  Food Insecurity Food secure 575(86)  114+44  Food insecure Missing  Father's education <High school High school  94(16) 0.1  0.80  92(14) 52  115+42  20(22)  52(8) 78(11)  104 ± 45 115+ 47  12(23) 17(22) 1 .4  Some college Holds degree(s)  157(22) 418(59)  Missing  14  0.24  117+43 116+ 43  1 .6  0.20  3 .7  0 .30  5 .9  0.02  18  0.00  1 .1  0 .59  22(14) 69(16)  Lone parent household Yes 70(10)  103+45'  No Missing  Number of adults at home 1 adult 2 adults 3 --4 adults 5-6 adults Missing  Family size 2-4 person 5-7 person >8 person Missing  19(27)" 6.0  0.02  649(90) 0  1.17+43 b  102(16)`  70(10) 522(72)  103±43 a 114+43 `'  19(27)" 92(18)` 5 .5  108(15) 19(3) 0  128+37 1' 124±64A  528(74) 172(24) 13(2) 6  113±42 a 120±44a 152±56b  0.00 5(5)` 5(26) ')  6.5  0 .00  92(17) 27(17) 1(13)  Low hair zinc=Hair zinc570pg/g. a,b,c : Indicate statistically significant differences in the hair zinc level of different categories of a given variable, verified by one-way ANOVA followed by Tukey's post hoc analysis, if applicable. x,y,z : Indicate statistically significant differences in the occurrence of low hair zinc in different categories of a given variable, verified by adjusted Chi square analysis, if applicable .  146  To explore the associations of family characteristics with the children's hair zinc, linear regression analyses were carried out between the hair zinc level and these variables. As Table III .VIII shows, "number of adults at home" "being in a lone parent household" and "family size" had significant associations with hair zinc . These variables accounted for a very small proportion of variance (accounting for 3%, 2% and 3% of the variance, respectively) . When the adjustments for the 5 main socio-demographic and socio-economic factors were carried out, "being in a lone parent household" no longer indicated a significant association with hair zinc. Educational attainments of the father and food security of the family were not associated with the children's hair zinc .  147  Table III.VIII .. Relationship of hair zinc with some family characteristics.  Variables  Father's education <High school High school Some college Held a degree  Unadjusted Bt(95% CI) R2  P  Adjusted* P RZ;fi  Mean (95% CI) 104 (92, 115) 115 (106, 125) 117 (110, 124) 116 (112, 120)  0.0  0.24  1 .2 (-8 .1, 11)  0 .00  0.80  0.10  0.90  -13 (-24, -2 .6)  0.01  0.01  0.12  0.07  Number of adults at homett (per each adult)  7.3 (3 .6, 11)  0.02  0.00  0.13  0 .05  Family size tt (per each person in the family)  4.3 (1 .9, 6.8)  0.02  0.00  0.13  0.02  Food security  Being in a lone parent household  0.12  0 .12  * Adjusted regression analyses have been carried out using age, sex, ethnicity, maternal education and family income as covariates. tBs are unstandardized coefficients.  ** R 2 represent the R 2 s of adjusting covariates (R 2 =0 .12) combined with R 2 of the new variable in the adjusted regression analyses. ttT wo variables of "number of adults at home" and "family size" were the only continuous variables . The 2 variables of food security and being in a lone-parent household were categorical data in a binary form (yes/No).  148  III.3.e . Association of eating behaviors with hair zinc of the children We collected data on the consumption frequency of some main food groups . We also explored the differences in the intake frequency of these food groups based on hair zinc status (Table IX.A). Consumption frequency of dairy, milk, cereals and grains, and whole grains of low hair zinc children were significantly higher than that of normal hair zinc children. We collected data on the caregivers' perception of the children's eating behavior. Table III .IX.B. displays the hair zinc level and occurrence of low hair zinc based on them . Our data indicated differences in hair zinc as well as the occurrence of low hair zinc based on being described as "eating unhealthy" and "not eating enough". To explore this further, we examined the associations of hair zinc with these variables and with the intake frequency of some food groups through linear regression analyses (Table III.X). Our data indicated inverse associations between hair zinc and the intake frequency of "dairy" and "milk" . In addition, being described as "eating unhealthy" and "not eating enough" had significant negative associations with hair zinc as well . In general, there was a moderate to substantial agreement (Saw & Ng, 2001) between the 2 sets of data collected on the consumption of food groups with R 2 varying within the range of 0 .67-0.77.  149  Table III. IXA. Comparison of the daily consumption frequency of some main food groups of low hair zinc and normal hair zinc children (n=719) Food groups  Low hair zinc children * (n=121) 1 .5+1 .1  Normal hair zinc children (n=598) 1 .4 ± 1 .1  0 .42  Dairy  3 .7+1 .5  3 .3+1 .6  0.01  Milk  2 .3±1 .2  1 .9±1 .1  0.00  Cereals and grains  3.6+1 .6  3 .1+1 .4  0 .01  Whole grains  1 .7+1 .6  1 .5+1 .2  0.02  Fluids  3 .3+1 .5  3 .4±1 .4  0.45  Meat, fish, and poultry  *  P  Values indicate number of servings/day .  150  Table III .IX.B. Description of some food behaviors and the hair zinc level and occurrence of low hair zinc based on these behaviors (n=719). N # (%)  Hair Zinc (µg/g) Mean +SD F  p  Low hair zinc* # (%) p  Concerned about child's eating; Yes  360(51)  114 ± 44  66(18) 1 .2  No  352(49)  117 ± 42  79(11)  103 f44 a  0.28  1 .2  0.28  3 .3  0 .01  0.1  0.08  7.8  0.01  55(15)  Described as eating unhealthy Yes  19(24)' 7.9  No  640(89)  117 f 43 b  223(32)  115 ± 43  0.01 102(16)Y  Described as picky eater Yes  38(13) 0.1  No  0.73  486(68)  116 f 43  83(15)  Yes  157(22)  109 f 48 a  38(24) x  No  562(88)  117 f 42 b  Described as not eating enough  5.1  0.02 83(15)Y  Low hair zinc=Hair zinc<70pg/g. a,b,c : Indicate statistically significant differences in the hair zinc level of different categories of a given variable, verified by one-way ANOVA followed by Tukey's post hoc analysis, if applicable. x,y,z : Indicate statistically significant differences in the occurrence of low hair zinc in different categories of a given variable, verified by adjusted Chi square analysis, if applicable .  151  Table III.X. Regression analyses of hair zinc (µg /g) with some eating behavior (n=705). Unadjusted  Adjusted  Food groups  B t (95% CI)  R2  P  R2**  P  Caregiver being concerned about child's eating  3 .5(-2 .8, 9.8)  0.00  0.28  0.12  0 .70  Being described as eating unhealthy  14(4 .3, 24)  0.01  0.01  0 .13  0.04  Being described as picky eater  1 .2(-5 .6, 8 .0)  0.00  0.73  0.12  0 .69  Being described as not eating enough  8.8(1 .1, 16)  0.01  0.02  0.13  0.04  -1 .6(-4 .6, 1 .3)  0 .00  0.29  0 .12  0.67  Consumption frequency of; Dairy  -3 .6(-5 .6, -1 .6)  0.02  0 .00  0.14  0 .00  Milk  -7.2(-10, -4.5)  0.04  0.00 0.14  0.00  0.12  0.06  0.12  0.08  Consumption frequency of meat, fish, and poultry  Consumption frequency of; Cereals& Grains  -2.9(-5 .0, -0.8)  0.01  0 .01  Whole grains & cereals  -2.0(-4 .5, 0 .4)  0.00  0.10  * Adjustments have been carried out using age, sex, ethnicity, maternal education and family income as covariates. t Bs are unstandardized coefficients' **R2 represent the R 2 s of adjusting covariates (R 2 =0 .12) combined with R 2 of the new variable in the analysis .  152  III.3.f. Associations of breast-feeding, parental perception of children's health status and taking supplements containing iron with hair zinc of the children Table III .XI.A displays descriptive data on the children's health and health behaviors (as perceived by the caregivers) and the hair zinc level and occurrence of low hair zinc based on these behaviors . As shown, children described as being frequently sick and having poor health had significantly lower hair zinc level and higher occurrence of low hair zinc than their counterparts. No differences in the hair zinc level or the occurrence of low hair zinc was apparent based on the breast-feeding pattern. Taking supplement(s) was common among our survey children . About 37% (240/671) of parents reported giving >1 supplement(s) to their children (Appendix III31). About 29% (197/671) of the survey children were taking supplements containing iron. The children taking supplements containing iron not only had significantly lower hair zinc ; they also revealed a higher occurrence of low hair zinc. Since our data on scores of attentional focusing and activity and the anthropometric Z scores were all in the form of continuous data, we could not include these variables in Table XI .A. However, we compared the scores of the 2 behavior domains of CBQ and the anthropometric Z scores of the 2 groups of children with low and normal hair zinc level through one-way analyses of variance, ANOVA (Table III.XI.B). Our results indicated a significantly lower score of activity level for low hair zinc children.  153  Table III .XI.A Description of some health behaviors and the hair zinc level and occurrence of low hair zinc based on these behaviors. Hair Zinc Low hair zinc * n a # (%) Mean +SD F p Described as frequently sick (µg/g)  x  p  # (%)  Yes  32(5)  12(38)"  95 ± 50 a 7.5  b  No Health status described poor Yes  686(95)  117 + 43  57(8)  103 ± 53 a  No Breastfeeding  662(92)  117 ± 42b  79 409 230  111+42 117 + 45 114 + 42  197(29)  112 + 44 a  Yes  474(71)  119 + 42 b  0.01  9.6  0.00  0.1  0.95  3 .9  0 .03  18(32)x 0.02 103(16) Y  1 .0  0 .38  13(17) 68(17) 40(17)  39(20)' 4.1  No  10 109(16) Y  5 .6  Not fed Fed< 6 months Fed>6 months Taking supplements containing iron  0.01  0 .04 65(14)Y  Low hair zinc=Hair zinc<70pg/g. a,b,c : are used to indicate statistically significant differences in the hair zinc level of different subcategories of a variable, as verified by one way ANOVA followed by Tukey's post hoc analysis, if applicable. x,y,z : are used to indicate statistically significant differences in the occurrence of low hair zinc in different subcategories of a variable, as verified by Chi square analysis .  154  Table III .XIB . Comparison of the scores of attentional focusing and activity level of low hair zinc and normal hair zinc children=719 Variable  Low hair zinc children (n=121)  Normal hair zinc children (n=598)  P  28+6 30+7  29+6 33+6  0.26 0 .00  -0 .03+0 .9 0.21+1 .5 0.20+2 .0  0.08+0 .9 0.24+1 .1 0.26+1 .3  0 .30 0 .27 0.65  CBQ behavioral domains (n=121, 598) Attentional focusing Activity level Anthropometric Z scpres (n=110, 554) HAZ WAZ WHZ CBQ : Child Behavior Questionnaire * Low hair zinc= Hair zinc <70pg/g. HAZ : Height-for-age Z scores WAZ : Weight-for-age Z scores WHZ : Weight-for-height Z scores  155  We explored associations between the children's hair zinc and all of the abovementioned variables . We observed that being described as "having poor health" and "being frequently sick" as well as "taking supplements containing iron" had negative associations with the children's hair zinc . (Table III .XII). When the analysis was adjusted for socio-demographic variables, however, only the association of "taking supplements containing iron" with hair zinc remained significant. Our data also indicated that the scores of "activity level", but not the "attentional focusing", associated with hair zinc in a positive manner. The internal consistency of the short CBQ, verified by Dr . Rothbart's research laboratory was a = 0.71 and 0.81 for activity level and attentional focusing, respectively (Putnam, personal email communications), while the results of our own reliability test indicated substantial agreements between the data collected on the children's behavior at 2 points of time (R2 =0.68 and 0 .72, for" activity level" and "attentional focusing", respectively). There were also substantial agreements between the 2 sets of data collected on the parents' perception of their child's health status with R 2 varying within the range 0 .620 .85 . When we explored the existence of any possible association between "taking supplements containing iron" and eating behaviors (Appendix III-28) of the children we noticed a significant association between "taking supplements containing iron" and 3 variables of "caregiver being concerned about child's eating", being described as "not eating enough" and being described as "picky eater" among which being described as "not eating enough" had already been shown associating significantly with hair zinc . Risk 156  analysis indicated that the children described as "not eating enough" were 40% more likely to be given supplements containing iron (f =2.9, OR=1 .4, 95%CI=[1 .1, 2.1).  157  Table III.XII. Regression analyses of hair zinc (µg /g) with breast-feeding, and some health-related and behavior-related variables. Unadjusted  Adjusted *  Variables  B t (95% CI)  R2  P  R 2**  P  Described as frequently sick  21(6.1, 37)  0.01  0.01  0.12  0.07  -14(-26, -2 .4)  0 .01  0 .02  0 .12  0 .08  Breastfeeding:  Mean (95% CI)  0.00  0.38  0 .12  0.06  Not fed Fed<6 months Fed>6 months  111(101, 121) 117(113, 121) 114(108, 119)  Described as having poor health  Scores of CBQ: Attentional focusing Activity level  88 (85, 92) 3 .5 (3 .4, 3 .6)  0.00 0 .02  0.44 0.00  0.12 0.13  0.29 0.05  Taking supplements containing iron  119(115, 123)  0.01  0 .04  0.13  0 .01  -0.5(-3 .9, 2.9) 0.8(-1 .7, 3.3) 0.6(-1 .7, 3 .3)  0 .00 0 .00 0.00  0.79 0.54 0.63  0.09 0.09 0.09  0.92 0.19 0 .27  Anthropometric Z scores: HAZ WAZ WHZ  * Adjustments have been carried out using age, sex, ethnicity, maternal education and family income as covariates. t Bs are unstandardized co-efficients. . .R2 represent the R 2 s of adjusting covariates (R 2 =0 .12) combined with R 2 of the new variable in the analysis. Data on anthropometric Z scores and scores of attentional focusing and activity level were in the form of continuous data. CBQ denotes Child Behavior Questionnaire, while HAZ, WAZ, and WHZ denote Height-for-age Z scores, Weight-for-age Z scores, and Weight-for-height Z scores, respectively .  158  IIL3.g. Contribution of all variables associating significantly with hair zinc, to the overall variability in hair zinc We set up a stepwise linear regression analysis with all of the previously mentioned variables that had shown a significant association with hair zinc . The inclusion criteria for entering this analysis was having P<0 .05 in adjusted regression analyses of that variable (independent variable) with hair zinc (dependent variable). The result of this analysis revealed that age, maternal education, family size, frequency of daily milk consumption, being described as "eating unhealthy", scores of activity level, and taking supplements containing iron were the significant predictors of hair zinc. The model produced by these variables could account for about 15% of the variability in hair zinc (Table III.XIII). In addition to these variables, others entered in the stepwise regression analysis were ethnicity, "number of adults at home", being described as "not eating enough", and "frequency of dairy consumption" . Nevertheless, these variables did not seem to be significant independent predictors of hair zinc level, as they were eliminated from the final model produced. One likely reason for the elimination of "number of adults at home" from the final model could be its strong correlation with "family size" (r =0 .51, p=0 .00). Likewise the strong correlation of "frequency of dairy consumption" and the "frequency of milk consumption" (r =0 .62, p=0 .00) could be a possible reason for elimination of the "frequency of dairy consumption" from the final model .  159  Table III.XIII. Stepwise linear regression analyses of hair zinc and the variables significantly associating with it. R2 change  P  ---  -----  0 .13  0.1  0.2  0 .05  0.00  6.8  1 .8  0 .2  0.03  0.00  Family size  4 .7  1 .2  0.1  0.02  0.00  Frequency of daily milk consumption  -4 .6  1 .4  -0.1  0.02  0 .00  Being described as eating unhealthy  -16  5 .3  0 .1  0.01  0.01  Score of activity level  0.8  0.3  0.1  0.01  0.00  Taking supplements containing iron  -8.5  3 .5  0.1  0 .01  0.02  Predictors in the model  B*  SE  Constant  -27  18  Age  0.8  Maternal education  * B indicates unstandardized coefficient. For the whole model ; R2 =0 .15, P=0 .00. R2 change : Change in the R 2 of the model as the result of the insertion of that variable into the model Other variables that were inserted in the analysis but did not become a component of the final model produced were ethnicity, "number of adults at home, being described as "not eating enough", "frequency of daily dairy consumption" .  160  III.3.h . Exploring the usefulness of the model/questionnaire produced by these components in predicting hair zinc status To produce a model that may have usefulness in predicting the hair zinc status of children, we set up a forward logistic regression analysis of hair zinc (binary outcome) with all of the variables significantly associating with low hair zinc status . The variables entering the analysis were : sex, age, ethnicity, maternal education, number of adults at home, being in a lone-parent household, described as "not eating enough", described as "eating unhealthy", consumption frequency of cereals, consumption frequency of whole grains, consumption frequency of dairy, consumption frequency of milk, scores of activity level, described as "having poor health", described as "being frequently sick" and taking supplements containing iron . Table III.XIV displays the results of this logistic regression analysis . In this model, sex, age, maternal education, number of adults at home, described as "not eating enough", consumption frequency of milk, scores of acitivity level, described as "being frequently sick", and taking supplements containing iron appeared as statistically significant predictors of low hair zinc. To test the utility of this model in predicting the hair zinc status we constructed the classification table for classifying the study children based on their hair zinc status, when using (Table III.XV.B) and not using (Tables III .XV.A) this model . As shown, the model generates only a very small improvement in the overall percentage of children classified correctly (84 .4% when not using the model versus 85 .5% when using the model) and the model could only detect 14 of the 99 low hair zinc children . This finding i.e . the lack of the model's practical efficacy is further supported in the Receiver Operating Characteristic (ROC) curve for the fitted logistic regression model . We 161  conducted ROC analysis as a way of examining the detection power of the model . As Figure III .IV presents, the model produced did not have much practical utility as the area under the curve, AUC, was only 0 .72 . This area measures discrimination, that is, the ability of the model to correctly classify those with and without the disease . The value of 0.72 indicates that such model has only fair discrimination ability . For clinical purposes and screening tools, a model with AUC greater than 0 .80 is desirable .  162  Table III .XIV. Logistic regression analysis of all factors associated significantly with low hair zinc (n=719). Variables Constant Sex Age Maternal education Number of adults at home Described as "not eating enough" Consumption frequency of milk Score of activity level Described as "being frequently sick" Taking supplements containing iron  B* 0.1  S .E 1.0  -2.9  P 0.01  1 .7  0 .2  0.5  0.04  1 .1  0.1  0.1  0.00  1 .3  0 .1  0.3  0 .01  1 .4  0.2  0.3  0.04  0.6  0 .3  -0 .6  0 .04  0.8  0.1  -0.2  0 .03  1 .1  0 .0  0.0  0 .02  0.3  0.5  -1 .1  0.03  0 .6  0.3  -0.6  0 .02  13  * Bs are standardized coefficients. Other variables that were inserted in the analysis but did not become a component of the final model were: ethnicity, being in a lone-parent household, described as "eating unhealthy", consumption frequency of cereals, consumption frequency of whole grains, consumption frequency of dairy, and described as "having poor health ."  163  Table III.XV.A. Classification of the survey children based on their hair zinc status without the use of the model produced by logistic regression analysis. Observed  Predicted  Hair zinc status Low Normal Low Hair zinc status Normal  0  0  100  541  0.00  100  Percentage Correct  84 .4  Table III.XV.B. Classification of the survey children based on their hair zinc status with the use of the model produced by logistic regression analysis . Predicted  Observed Hair zinc status Low Normal  Low Hair zinc status Normal Overall Percentage  16  9  84  532  16  98.3  Percentage Correct  85 .5  164    1 .o-  0 .8-  0.4-  0 .2-  0 .0 0 .0  0 .2  I  I  I  I  0.4  0 .6  0 .8  1 .0  1 - Specificity Predictive probability (Area under the curve)=0 .72.  Figure III.VI . Receiver Operating Characteristic (ROC) curve of the final regression model to detect children with low hair zinc.  165  III.4. DISCUSSION Our study explored the hair zinc status of Vancouver preschoolers by taking a 3% sample of this age group (n=719), with representation of the main ethnicities comparable to percentages reported for the city (Statistics Canada, 2001) (Table III.II). The study findings indicated that the average hair zinc in this population, together with the occurrence of low hair zinc (<70 µg/g) (Figure III.II), were comparable with the only available report for a group of apparently healthy Canadian preschoolers (Smit VanderKooy & Gibson, 1987) as well as with reports on children of other industrialized societies (Hambidge et al, 1976, and Lombeck et al, 1988). Overall, our analyses indicated sex, age, maternal education, "the number of adults at home", being described as "not eating enough", consumption frequency of milk, scores of activity level, being described as "frequently sick", and taking supplements containing iron as the significant predictors of hair zinc status in the final logistic regression model produced (Table III.XIV). However, the predictive power of this model (and consequently any screening questionnaire constructed based on this model) was not strong enough to be of any practical use . As a result, the primary goal in this study i .e. constructing and validating a small and simple screening questionnaire could not be achieved . However, in working through the process that built up the final logistic regression model, we identified a number of factors in the children, their family environment and social setting that were significantly associated with hair zinc and hair zinc status and may potentially affect their zinc status. We observed that the two sexes had a comparable hair zinc mean while having a significantly different occurrence of low hair zinc (Table IILIV) . The occurrence of low 166  hair zinc among girls was so much higher that "being female" conferred a 90% higher risk on children . We could not explain this finding given that the majority of evidence in the zinc literature indicates male children as the more vulnerable sex to low hair zinc (Lambed( et al, 1988, Smit Vanderkooy & Gibson, 1987) and having a better response to zinc supplementation (Walravens & Hambidge, 1976, Hambidge et al ., 1979, Heinersdorff & Tylor, 1979, Hambidge et al, 1979 and Walravens et al, 1983) . The literature speculates on the higher physiological need in male as the underlying reason for these observed differences . However, in light of our finding and the findings of Zachiwieja and co-workers (1995) that have shown a higher hair zinc for boys, it is reasonable to believe that there may be other factors besides physiology at work here, factors that, unlike the physiology of male humans, differ for male children of different environments/studies . One such factor could be the sex-based differences in food intake that has been documented in some Canadian studies (Smit Vanderkooy & Gibson, 1987, Nakano, et al, 2005, and Kuhnlein et al, 2007) and in studies of other industrial societies (Ganjii, iIampl, & Betts, 2003, Glynn et al, 2005, and 'I'ouvier et al, 2006) as well . In our study, due to the lack of quantitative dietary data Ive could not verify the existence of such differences as an explanation of the observed differences in the hair zinc status of our survey boys and girls.  Our data also indicated an increase in hair zinc as the age of the children increased (Figure III .IV). 'When we categorized children into finer age categories (12 month increments) we noticed that the children's hair zinc increased from 107 to 125 µg/g at age 4 regardless of sex (Table III.V). Accordingly, again regardless of sex, the occurrence of low hair zinc declined with advancing age (Figure III .V). Zinc literature 167  indicates a change in hair zinc concentration with age (Klevay, 1970, Hambidge et al, 1972, Lombeck et al, 1988, Van Wouwe, 1995 and Meng, 1998) . While there seems to be consensus on the decrease in hair zinc during infancy and then an increase during postinfancy, information on the nadir and zenith of this change is inconsistent . A result similar to ours has been reported in the cross-sectional study of Hambidge and coworkers (1972), which showed that the hair zinc for neonates was closely comparable with that of young adults, while for infants it was lower than that of young adults . These levels remained low until the children were 4 years of age, and then rose, whereas others (Klevay, 1970, Van Wouwe, 1995 and Sakai et al, 2000) have indicated a different age for this rise in hair zinc . These studies, however, either did not look at the preschool years individually, thus keeping all 0-5 years in one group (Klevay, 1970 and Van Wouwe, 1995), or had only a small sample size for each age segment (Sakai et al, 2000) . Our study is the first sufficiently large study of this age group that has been able to sample an adequate number of children from each age cluster (2, 3, 4, 5, years old) and has explored their trends and differences . This large sample size, coupled with a superior method of hair zinc analysis ICP-MS, which has improved dramatically the measurement of trace elements in hair (Bass et al, 2001, and Miekeley et al, 1998), confers a high reliability on our results. This age-based difference in hair zinc deserves further investigation . If this low hair zinc has no functional/clinical ramification, it may be a normal physiological level . If so, it may necessitate the validation of the commonly used hair zinc cutoff (<70 µg/g) in younger preschoolers and may challenge its usefulness as a biomarker of MZD among younger preschoolers . Hambidge and co-workers (1972), in their study of preschoolers, 168  documented associations between hair zinc and some clinical signs of zinc deficiency such as anorexia and poor growth among apparently healthy younger children (<4 years of age) . However, this association was evident for children who had a hair zinc level below 30 µg/g, and was not apparent in children (below the age of 4) whose hair zinc was 30-70 µg/g . This observation clearly undermines the appropriateness of using the cutoff of 70 µg/g, for younger children . However, since that study (perhaps owing to the absence of any further research on this issue) researchers continue to include these 2 age groups in the same study using the same cutoff for both younger and older children (Hambidge et al, 1976, Walravens, Krebs & Hambidge, 1983, Lombeck et al 1988 and Van Wouwe, 1995). In our study, we did not observe any differences in the indices of growth or nutrition between younger low hair zinc children (based on the cutoff of 30 µg/g) and their counterparts with normal hair zinc (Appendix III-29) . Further, the lack of any associations between hair zinc status and the indices of growth and nutrition (the anthropometric Z scores) persisted, even when the hair zinc status of the younger children was assigned using the cutoff of 30 gg/g . Clearly some additional studies are warranted to further investigate this important issue. Our survey also indicated differences in the hair zinc level as well as the occurrence of low hair zinc among children of different ethnicities (Table III.IV). The differences in the rate of low hair zinc varied largely, East Indian and Chinese children being at the lower and upper end of this range, respectively, this in light of the fact that there were no significant differences in the distribution of vulnerable sex and age group among the three main ethnic groups of the survey (Appendix III-27) . 169  Caucasian children were positioned in the middle with a 13% occurrence of low hair zinc, significantly lower than the rate of 24% reported for the white middle income Canadian preschoolers of Smit Vanderkooy & Gibson (1987) . Chinese survey children, on the other hand, had the lowest mean of hair zinc and the highest occurrence of low hair zinc, which resulted in more than a quarter of them showing hair zinc below the cutoff. Zinc deficiency has been a concern among Chinese in earlier years (Eggleton, 1940) and more recently too (Chen, 1985 and Sheng et al, 2006) . Our dietary data (Appendix III-30) did not indicate any differences among ethnicities that might explain this high occurrence of low hair zinc among our Chinese children . In addition, our data measured the frequency of intake and not the total quantity of daily intake . For these reasons we were not able to link our biochemical data (the hair zinc measurements) to the dietary data . However, the general nature of the Chinese diet (a cereal and plant-based diet with small amounts of meat) may help to explain this finding . The 2 food groups "cereals and grains" and "fruits and vegetables" are often sources of zinc inhibitors such as phytate and fiber . Therefore, the status of zinc in the Chinese may be sub-optimal, due not only to a low supply of zinc, but also inhibition of whatever zinc is present in the diet. Many Chinese immigrants do not stop eating Chinese food, but rather layer Western foods and eating behavior over their stable Chinese pattern (Satia et al, 2000 and RovilleSausse, 2005) . In addition, a comparison of the mean hair zinc of our Chinese children with the previously reported values for Chinese children living in China (104±43 and 100+41, respectively) (Chen et al, 1985) may also be supporting evidence that the diet of our Chinese children may have been a traditional Chinese diet .  170  Unlike this finding, our data on the hair zinc status of our East Indian children were somewhat unexpected . We observed no low hair zinc child among our children with East Indian ethnicity . The published literature indicates the existence of widespread mild (Agte, Chiplonkar & Tarwadi, 2005) as well as severe (Chiplonkar et al, 2004, Bhandari et al, 2002 and Sazawal et al, 1996) zinc deficiency in East Indians living in India . There are no published data on the zinc nutriture of East Indian immigrants living in North America or other industrialized societies . However, sub-optimal iron status has been documented in East Indian immigrant populations residing in Canada (Bindra & Gibson, 1986), the United Kingdom (Robertson et al, 1982), and the USA (Ganapathy & Dhanda, 1980) and these findings have been attributed to the retention of traditional East Indian eating patterns among East Indian immigrants . These patterns include a predominance of lacto-ovo-vegetarianism and daily consumption of unleavened chapatti bread made from whole-wheat flour with or without added bran (Bindra & Gibson, 1986), which is very high in phytate . Such a diet with a lack of bioavailable dietary zinc and a provision of large amounts of zinc inhibitors (phytates), could easily lead to an unfavorable zinc status as well . In our study, only 28% (n=15) of our East Indian children were lacto-ovovegetarian . As for the other component of the traditional East Indian diet (the daily consumption of large amounts of high phytate bread) due to the limitations of our dietary tool we cannot comment further. At this point we cannot be certain whether ethnicity per se is the explaining factor for these differences or other factors associated with ethnicity (such as ethnic food etc) are actually mediators . However, the fact that ethnicity did not remain as a significant component of our final stepwise or logistic regression models as the predictor of hair zinc 171  (Table III . XIII and Table III.XIV, respectively), suggests the likelihood of the latter rather than the former . Further studies with comprehensive dietary information are warranted in order to verify these findings and explore any possible dietary or other nondietary explanation for them. We observed a significant difference in the children's hair zinc based on maternal education (Table III.IV). There was no gradient effect (wherein the hair zinc status of children of each category of maternal education would have been better than the one above and worse than the one below it in the table) from maternal education and either the hair zinc level or the occurrence of low hair zinc . Both these variables, though, were significantly different in moving from children with maternal education < high school (the first 2 categories of maternal education) to those with maternal education >high school (the last 2 categories of maternal education) . Since dichotomizing data on a variable that has been measured by a continuous scale (i .e. the hair zinc) could potentially result in a loss of information and even increase a Type I error (Maxwell & Delaney, 1993), we decided to keep the information on hair zinc intact in its continuous form and explore associations between the hair zinc of children and the socio-economic and sociodemographic variables through regression analyses . The results of these analyses, while confirming the associations of age and ethnicity with hair zinc, indicated also a significant positive association between hair zinc and maternal education (Table III.VI). However, this variable (maternal education) could only explain 2% of the variance in hair zinc. Other research has indicated the positive effect of maternal education on social development, emotional well-being and the physical health of children (Zill et al, 1995 172  and Zill et al, 1996) . An avenue by which maternal education may affect a child's health is nutrition . Studies have shown more favorable eating patterns in children with higher maternal education (North & Emmet, 2000) . Some studies have actually gone beyond this and shown evidence in support of the thesis that, for young children, maternal education may be the most important factor determining the adequacy of their nutrition (Watt, Dykes & Sheiham, 2001) . One way through which the positive effect of maternal education on a child's nutrition can be played out is through an increase in family income, which, in turn, could result in higher buying power and increased accessibility of better food . However, this was not the likely path for the positive association of maternal education with the hair zinc level in our study, since income itself was not, as indicated, significantly associated with hair zinc (Table III.VI) . Further, the positive association of maternal education with hair zinc remained significant, even after adjustments for the 5 socio-demographic variables, including family income were carried out . Therefore, the observed positive association of maternal education with hair zinc appeared to be independent of its potential enhancing effect on a family's income. Nevertheless, increased parental education can have a positive effect on a child's nutrition in other ways. Research today does not support the contention that maternal education is just another surrogate for measuring social class or socio-economic status . In households with low maternal education it is not just the lack of financial resources that are at work but also a lack of knowledge coupled possibly with unhealthy parental food behaviors that pose a threat to a child's nutrition and health (Dubois, 2006) . Educated parents are likely to have some knowledge of childhood health and nutrition . In addition, a better-educated mother would be more likely to eat healthy herself (Basiotis et al, 2002) 173  and influence the eating patterns of her child by modeling healthy behavior and setting a positive example. The lack of any significant association between the hair zinc of children and their family income was an interesting finding of our study . Since our information on family income was the "recorded income" and not the "income per capita", all the abovementioned analyses were repeated with "total family members" inserted as the covariate. We did not observe any difference in the results . Data from the US (Silva, 2001 and Skalicky et al, 2006) present supporting evidence for the existence of an inverse relationship between family income and some nutritional deficiencies . Some studies from Quebec have shown family socio-economic status to be related to some elements in the quality of a preschoolers' diet (Dubois & Girard, 2001) . This lack of association between income and hair zinc level in our study, despite some pre-existing evidence on the association of income with nutritional deficiencies, was not an unexpected finding for a nutritional deficiency such as MZD . Research indicates that marginal zinc deficiency, unlike severe zinc deficiency, which usually co-occurs with caloric deficiency and hunger (Muller & Krawinkel, 2005), is usually but not always, the result of food consumption patterns rather than the quantity of food consumed (Gibson et al, 2001) . This deficiency, although it may occur in underdeveloped and developing societies as well, is more prevalent in industrialized societies (Gibson et al, 1989, Chakar et al, 1993 and Nakamura, 1993), where food insecurity, although not non-e'istent, is not as large a concern . In other words, while extreme low income and food insecurity may predispose a child to MZD, food security will not necessarily protect him/her . This point was reflected  174  in our data, when we observed no significant associations between the hair zinc level of children and their families' food security (Table III.VIII). The occurrence of food insecurity among our survey families was 14%, comparable to 10% reported by Statistics Canada (2001) for the Vancouver population. An additional explanation for the lack of association between food security and hair zinc may come from the fact that a food insecure family does not always equate to food insecure children (Che & Chen, 2001) . Some Canadian studies indicate that younger children seem to be protected from poor quality diets in households with limited resources to acquire food (Glanville & McIntyre, 2006, Radimer et al, 1992 and McIntyre et al, 2003) . Family food insecurity affects the status of food security of the children in some circumstances only . In support of the findings of these small scale studies, data from our recent National Population Health Survey, NPHS, 1998/1999 also indicate over half of the Canadian children in food insecure households were food secure, while (20%) of these children experienced food insecurity only to the level of worrying about food (marginal food security) and about 29% had actually experienced food insecurity to the point of compromising their diet (Che & Chen, 2001) . In addition, in the classification we used to categorize our families' food security status (Che & Chen, 2001), answering "yes" to anyone of the three questions would result in being classified as food insecure . A family, which is in this category due to the giving of a positive answer to the first question, has experienced food insecurity only to the extent of worrying about food . This degree of food insecurity may not lead to any actual change in food intake. We also observed a statistically significant difference between the mean (±SD) of the hair zinc of children from lone parent households and those who were not in a lone 175  parent household (Table III .VII), indicating that a child's nutrition may be affected by lone parenthood . Data analyses from the second National Health and Nutrition Examination Survey, NHANES II, and a comparison of the nutritional status of the preschoolers from intact and lone-parent families have provided some evidence linking lone parenthood with sub-optimal nutritional status (Bowering & Wynn, 1986) . The usual association of lone parenthood with poverty (Dubois et al . 2000 and Family Pediatrics, 2003) did not appear to be the likely explanation for this observation. Within lone parent households, there are other factors that may work adversely on a child's nutrition . Among them are the stress and anxiety level as speculated by Bowering & Wynn (1986) and perhaps the scarcity of time available for childcare . This shortage of time may have its deleterious effect on the quality of the care given to children including the nutritional care as in meal planning and allocating time for the proper preparation and cooking of nutritious meals . Congruent with this line of reasoning was a unique finding of a positive association between the children's hair zinc and "number of adults at home" (Table III .VII) . Although observing this latter association strengthened our speculation of "time scarcity" as the possible underlying reason for the negative association between hair zinc and "being in a lone parent household", the negative association of "being in lone-parent household" with hair zinc was no longer significant when we adjusted the regression analysis . Additionally, "being in lone-parent household" was not a component of our final stepwise or logistic regression model. The last observation of family characteristics was the positive association between children's hair zinc and family size (Table III .VII) . This led us to suspect that this positive association may have stemmed from an increase in the "number of adults at 176  home" . However, in the final stage of our statistical analyses, when we inserted "number of adults at home" and "family size" in the stepwise regression (Table III.XIII) analyses, only the "family size" remained as a significant component as the variable that could explain part of the variation in hair zinc level . On the other hand, "number of adults at home" was the significant component of the final model produced by logistic regression analysis (Table III.XIV), as one of the variables that could predict hair zinc status. Through collecting data on caregivers' perceptions of their children's eating behaviors, we observed differences in the hair zinc level and the occurrence of low hair zinc based on being described as "eating unhealthy" (Table III.IX.A) . This variable was also negatively associated with hair zinc of the children in linear regression analyses (Table III.X) . In our survey, similar to our study of the low-income preschoolers of Vancouver, the survey parents had not described what is perceived as "unhealthy eating". Whatever this behavior (so described) may have been for individual parents of this group of children, our analysis of this behavior showed an association with one common health outcome - having lower mean hair zinc . In our final analysis, however, being described as "eating unhealthy" was a significant component of the stepwise regression analysis (Table III .XIII), but not a significant component of the model produced by logistic regression analysis (Table III.XIV). Another parent-described eating behavior accompanied by significantly lower hair zinc level and nigher occurrence of low hair zinc among children was "not eating enough" . We asked parents whether or not they perceived their child as a "picky eater" or "not eating enough" . We used these 2 eating behaviors as 2 proxies for loss of appetite, a clinical symptom often described in MZD (Mangian, Lig, & Shay, 1997, and Mantzoros 177  et al, 1998) . Being described as "not eating enough" remained as a significant component of the final logistic regression model. Among other eating behaviors we observed significant negative association between the children's hair zinc and the consumption frequency of milk and dairy (Table III.X). Very frequent milk consumption was a common dietary theme among our preschoolers . Our previous study of low-income preschoolers of Vancouver (chapter II of this thesis) had indicated frequent consumption to the point of over-consumption of dairy and milk. However, the dietary tool of our survey, unlike that of the inner city study was not detailed enough to differentiate between repeated milk consumption constituted as over consumption or just frequent consumption of small portions . Heavy milk consumption (Lino et al, 1999 and Kranz et al, 2006) and milk being a favorite food among preschoolers (Skinner et al, 1999) have been reported in the literature . Perhaps its ease of consumption and familiarity with its taste from infancy, when combined with parental approval, are key factors leading to the tendency of this age group towards frequent milk consumption. On average young children consume only small amounts of food at meals . They, therefore, must have a nutrient-dense diet with variety in order to provide sufficient amounts of all nutrients for growth and development (Curran & Barnes, 2000) . Frequent consumption of milk, similar to any other fluid, can displace other foods from a child's diet including the zinc-rich foods . Frequent consumption of milk and dairy, if resulting in over consumption, can also interfere with zinc absorption . Dairy foods contain large amounts of calcium, an element that can potentiate the complex formation between zinc and phytate and thus lead to zinc excretion . The adverse effect of over-consumption of 178  calcium on zinc status has clearly been shown in animal studies (Oberleas, 1966) . Studies of Canadian children have also documented an inverse relationship between calcium intake and their hair zinc status (Smit Vanderkooy & Gibson, 1987 and Gibson et al, 1991) . It is important to note that in both our final stepwise regression analysis and logistic regression analysis the consumption frequency of milk remained as a significant component, indicating its predictive power for hair zinc level and hair zinc status, respectively . However, it contributed only 2% to the overall explained variance. From our information on parents' perceptions of their children's health and parents' report of their health behavior we learned that being described as "frequently sick" or "having poor health" was also inversely associated with hair zinc (Table III.XII). A strong inverse association was also observed between hair zinc and perceptions of parents as to their children's health as detailed in our previous work with low-income preschoolers of Vancouver (chapter II of this thesis) . It is known that the immune system is highly zinc-dependent (Ibs & Rink, 2003) such that when the zinc supply to the body is inadequate the function of this system may well be compromised with a subsequent weakening of the child's resistance to common infections . Conversely, since zinc is an element of "acute phase response" and plasma zinc is redistributed to other tissues during infection/inflammation (Prasad, 1981), reduced hair zinc could be the net result of the ongoing decline in the plasma zinc of the children who were described as "often sick" . Like any other cross-sectional study, all of our data were collected ~t one point of time and we cannot therefore be certain which one of these 2 events (the reduction in hair zinc and being frequently sick) occurred first . Regardless of the sequence of the events, the inverse association of hair zinc with frequency of common  179  infections has also been shown in studies of preschoolers in other developed countries (Von Wouwe et al, 1986 and Lombeck et al, 1988) . In our study, being described as "frequently sick" remained as a predictor of hair zinc status by virtue of its appearance in our final logistic regression model (Table III.XIV), only. In addition to the immune system, the zinc literature points also to the adverse effect of sub-optimal zinc status on other systems, among them the central nervous system and, as a result, the cognitive function . The results of data collected through CBQ on 2 behavioral domains of "attentional focusing" and "activity level" indicated that the scores of activity level were significantly lower for low hair zinc children compared to their normal hair zinc counterparts (Table III.XI.B). In addition, these scores were positively associated with the hair zinc level of the children (Table III.XII). Experimental animal studies (Golub et al, 1994 and 1996) and observational studies (Kirksey et al, 1994) and zinc supplementation trials (Ashworth et al, 1998, Bentley et al, 1997, Sazawal et al, 1996 and Sandstead et al, 1998) in humans have all documented the negative association of zinc deficiency with activity level and motor development . Our study provided an added evidence to this body of literature . This variable (the scores of activity level) remained as a significant component in both of our final regression models (Tables III.XVIII and III .XIV). One of the noteworthy observations in our study was the widespread consumption of supplements, a large portion of which contained iron (Appendix III-31) . Analyses of the hair zinc data based on consumption of supplements containing iron indicated a significantly lower hair zinc level and higher occurrence of low hair zinc among the children taking iron supplements (Table III .XI.A) . Iron is known to have an antagonistic 180  interaction with zinc, and hence the potential to lower the biochemical zinc status (Solomons, 1986) . The discovery of divalent metal transporter 1 (DMT1) (Gunshin et al, 1997) has provided an explanation for the competitive absorption of iron and zinc from the small intestine. Although the adverse effect of iron supplementation on zinc absorption has been documented (Valberg, Flanagan, & Chamberlain, 1984, Sandstrom et al, 1985 and Troost et al, 2003), its consequence on the overall zinc status of the body has been inconsistent (Solomons et al, 1981, Solomons et al, 1983, Meadows et al, 1983, Solomons, 1986 and Troost et al, 2003) . However, plasma Zn is tightly protected from the wide fluctuations experienced as a result of a broad change in zinc available to the body, and is thus maintained within a normal range (Hambidge & Krebs, 2001) . For this reason any conclusions regarding the effect of iron supplements on zinc status are difficult to make with any certainty when plasma zinc is taken as the biomarker of zinc status . Such results could simply reflect the net result of the redistribution of zinc within the body . However, as noted, zinc in a hair follicle is not subject to redistribution in response to the ongoing metabolic events in the body (Hopps, 1977) . Supporting evidence for this argument will be found in a study of New Zealand women where the plasma zinc concentrations were not significantly different for supplement takers while their mean hair zinc value was significantly lower (Gibson, 2001). Although we observed as well a significant negative association between taking supplements containing iron and hair zinc level in our linear regression analyses (both adjusted and unadjusted) (Table III .XII), it is also noteworthy that the amount of iron in our supplements was very small, lying in the range of 4-18 mg/day . Most studies have 181  demonstrated iron and Zn interactions when the amount of supplemental iron exceeds 30 mg/d . The smallest iron concentration known to affect zinc status is 18 mg/d (Dawson et al, 1989) . However, it is important to bear in mind that the subjects of this study were pregnant teens with a significantly high zinc requirement (Estimated Average Requirement of 10 mg/day for pregnancy, versus 4 mg/day for preschoolers) (Institute of Medicine, 2000) . Furthermore, the supplements containing iron at the higher end of the range (Flintstones Children's chewable Multivitamin/Multmineral Supplement which provides 18mg of iron/tablet) also contained 15mg of zinc in each tablet . To sum up, considering the low concentration of iron in most supplements and the simultaneous provision of zinc in those with higher iron concentration, it was improbable that these supplements would have an adverse effect on zinc status . We speculated that the observed association might have originated from another variable. In exploring our data further, we noticed that there was significant association between being described as "not eating enough" and taking supplements containing iron. In fact, the odds ratio of taking supplements containing iron among children who were described as "not eating enough" was 40% higher than the other children . It is possible that parents who had observed their children not eating enough administered these supplements as a way of safeguarding their health . However, taking supplements containing iron remained as the significant component of both our final regression models, while "eating unhealthy" ,was only a significant component of the logistic regression model produced (Table III .XIV). At this point and within the limitations of our data, we cannot ascertain whether the observed negative association was the result of a mineral/mineral interaction or the 182  artifact of some other middle variable . However, the antagonistic effect of iron supplementation on zinc status is a concern that warrants further investigation . Although the Canadian Pediatrics Society (CPS) and the Dietitians of Canada recommend the intake of a wide variety of foods as a way of obtaining adequate vitamins and minerals rather than the intake of supplements (CPS, 2007), the intake of iron supplements for some children is crucial . Iron deficiency is a valid pediatric health concern even in developed countries (McCann & Ames, 2007), and despite global initiatives such as food fortification and enrichment, a large number of children in the US (CDC, 2002) as well as in Canada (Williams, 2000) still fail to meet their needs for this essential mineral . In view of the significance of zinc and iron for the human body and the consequences of their deficiencies on children's health and well being, further research on the possible interaction of these 2 minerals on each other is clearly warranted . This issue has a particular significance in a society such as ours where, based on our recent provincial survey, almost 1/3 of adults take supplements (Barr, 2004), and according to our survey over 1/3 of parents give supplements to their children (Appendix III-31). The results of the stepwise regression analysis of hair zinc together with all of the variables in this study indicated age, maternal education, family size, consumption frequency of milk, being described as "eating unhealthy", scores of activity level, and taking supplements containing iron as the variables that could explain part of the variability in hair zinc level (Table III.XIII). Similarly, subsequent to the identification of all of the dietary and non-dietary factors associated with low hair zinc, we constructed a logistic regression model with hair zinc (low versus normal) and all these variables . As displayed in Table III .XIV sex, age, 183  maternal education, number of adults at home, being described as "not eating enough", consumption frequency of milk, scores of activity level, described as "being frequently sick", and taking supplements containing iron were the significant components of the final model produced. However, as expected from the results of the regression analyses (the small R2 value of these analyses) the collective model did not provide adequate predictive power . This was clearly reflected in the small improvement in the percentage of children classified correctly as normal/low hair zinc when this model was used (85 .5% versus 84 .4% with and without the use of model, respectively) (Table III.XV.B and Table III .XV.A) . In other words, although this screening questionnaire identified most of the normal hair zinc children correctly (98 .3% specificity), it managed to identify only 16 of the 100 low hair zinc children (16% sensitivity), thereby failing to fulfill the function of a first-stage screening tool to reduce the number of children needed to undergo subsequent exploratory and diagnostic tests . The positive predictive value of this model was 64% (16/25) while its negative predictive value was 86% (532/616) (Table III.XV.A). In total, only about 15% of the variability in hair zinc could be explained by the variables assessed in this study . The rest of the variability (85%) in hair zinc stems from factors not addressed in this study . Our study questions concerned risk factors that have been commonly acknowledged as risk factors of zinc deficiency in both marginal and severe forms. The predictive probability of a model based on such questions was calculated at 0 .72 as indicated by the area under the curve of ROC (Figure IILVI). Obviously other factors that were not included in this study may be more predictive of low hair zinc and MZD .  184  The large variation in zinc absorption due to the presence or absence of other dietary and non-dietary enhancers and inhibitors such as the status of minerals like iron and lead may constitute some of these factors . In addition zinc is one of the most omnipresent and multifunctional minerals in the human body . Therefore many physiological and pathological events occurring in the body may affect its balance and status and outweigh the cumulative effect of some of the previously identified factors that were also explored in this study. The zinc literature does not contain any previous studies attempting to construct a screening questionnaire and examine its utility . However, such attempts have been made for screening iron deficiency (Boutry & Needlman, 1996, Bogen et al, 2000, and Williams, 2000) . These questionnaires are reported to have a high sensitivity (85%-95%) and a modest specificity (15%-49%) . A simple dietary check list constructed by Williams (2000) based on the assessment of 152 children (8-26 months old) in the city of Vancouver in an attempt to serve as a first stage screening tool of iron deficiency anemia had 87% sensitivity and 49% specificity. Although we could not reach the primary goal of this study in constructing and then validating a short questionnaire that could serve as a screening tool for marginal zinc deficiency, this study did contribute in several ways to the zinc literature . It updated our information on the hair zinc status and the prevalence of low hair zinc in an apparently healthy population of Canadian preschoolers 2 decades after the last Canadian study of this age group was published . The data also provided some important reference values for healthy children of North America and perhaps the children of industrialized societies in general . This study raised some key questions for zinc research such as the relationship 185  between hair zinc and age, and set the stage for future studies to further investigate this imperative issue . Finally, this study identified and drew to our attention factors in a child's social setting and family environment, which may predispose her/him to a suboptimal hair zinc status. However, as with other studies, ours has limitations, which should be kept in mind when generalizing its findings . One such limitation arises from our sampling method, which may have conferred "selection bias" to the findings . By using the preschool and daycare centers as a major source of recruitment, we may have excluded children of stay-home mothers who are not the consumers of these services . However, our recruitment of some children (n=121/719=17%) through drop-in sessions at community centers, where parents (stay-home parents) and children are engaged together in various activities, may have attenuated this bias. Another factor that may have introduced some bias to our study is the selection bias due to non-response . The overall response rate of our survey was 30% with 70% of the parents receiving our survey package choosing not to respond . Non-response in survey research can pose a threat to the generalization of results, if respondents and nonrespondents differ systematically (Barriball & While, 1999) . We did not have any way of comparing the characteristics of non-respondents with respondents . However, in general, it has been shown that non-respondents have a lower socioeconomic (Groves & Couper, 1992, Mishra et al, 1993 and Goyder, Warriner, & Miller, 2002), worse health (Jackson, Chambless & Yang, 1996 and Shahar, Folsom & Jackson, 1996) and a better selfreported health (Mishra et al, 1993) status than respondents . One possible implication of  186  this error on the findings of our study would be under-estimation of the prevalence of low hair zinc due to the participation of the healthier fraction of the population. One last methodological shortcoming of our study was our survey questionnaire not offering any visual aid for the serving sizes of different food groups to the participants . Such an aid would have enabled us to have some idea about the overall daily intake of food groups and compare the intake of our participants to the recommendations made by Eating Well with Canada's Food Guide for Healthy Eating, CFGHE (Health Canada, 2007) and then draw some valid and meaningful conclusions. Finally, it is important to keep in mind that this was a cross-sectional survey with a limitation inherent in all cross-sectional studies . All of the relationships observed were those of association . As a result, we did not and cannot establish the existence of any causal relationship, or lack of it .  187  III.5.BIBLIOGRAPHY Agte VV, Chiplonkar SA, Tarwadi KV . Factors influencing zinc status of apparently healthy indians . J Am Coll Nutr 2005 ; 24 : 334-41. Ashworth A, Morris SS, Lira PI, Grantham-McGregor SM . Zinc supplementation, mental development and behavior in low birth weight term infants in northeast Brazil . Eur J Clin Nutr 1998 ; 52 : 223-7.  Barr, SI . (2004) . British Columbia Nutrition Survey, Report on Supplements . BC. Ministry of Health Planning. Retrieved July 20, 2005 from http ://www.healthservices . goy. bcbc .ca/prevent/nutrition/pdf/nutrition_supp .pdf Barriball KL, While AE . Non-response in survey research : a methodological discussion and development of an explanatory model . J Adv Nurs 1999; 30 : 677-86. Basiotis, P .P., Carlson, A., Gerrior, S .A ., Juan, WY,& Lino, M . (2002) . The Healthy Eating Index : 1999-2000. U .S .Department of Agriculture, Center for Nutrition Policy and  Promotion.  Bass DA, Hickock D, Quig D, Urek K . Trace element analysis in hair : factors determining accuracy, precision, and reliability . Altern Med Rev 2001 ; 6 : 472-81.  Bentley ME, Caulfield LE, Ram M, Santizo MC, Hurtado E, Rivera JA, Ruel MT, Brown KH . Zinc supplementation affects the activity patterns of rural Guatemalan infants . J Nutr 1997 ; 127 : 1333-8.  Bhandari N, Bahl R, Taneja S, Strand T, Molbak K, Ulvik RJ, Sommerfelt H, Bhan MK. Substantial reduction in severe diarrheal morbidity by daily zinc supplementation in young North Indian children. Pediatrics 2002 ; 109 : e86.  188  Bindra GS, Gibson RS . Iron status of predominantly lacto-ovo vegetarian East Indian immigrants to Canada : a model approach . Am J Clin Nutr 1986 ; 44: 643-52.  Bogen DL, Duggan AK, Dover GJ, Wilson MH . Screening for iron deficiency anemia by dietary history in a high-risk population . Pediatrics, 2000 ; 105 : 1254-9.  Bowering J, Wynn ' RL . Nutritional Status o f preschool children from intact and divorced families. Fam Consum Sci Res J 1986 ; 15 : 132-40.  Boutry M, Needlman R . Use of diet history in the screening of iron deficiency . Peds 1996 ; 98 : 1138-42.  Canadian Pediatrics Society, CPS . Promoting healthy eating habits . 2007 . Retrieved February 11, 2007 from http://www.caringforkids.cps .ca/eating/HeathyEating .htm Center for Disease Control . Iron deficiency in the United States, 1999—2000 ; Morbidity and Mortality . Wkly Rep 2002 ; 51 : 897—9.  Center for Disease Control . CDC growth charts. 2000 . Retrieved October 11, 2006 from http://www.cdc . gov/nchs/about/maj or/nhanes/growthcharts/datafiles .htm  Chakar A, Mokni R, Walravens PA, Chappuis P, Bleiberg-Daniel F, Mahu JL, Lemonnier D. Plasma zinc and copper in Paris area preschool children with growth impairment. Biol Trace Elem Res 1993 ; 38 : 97-106.  Che J, Chen J . Food insecurity in Canadian households . Health Rep 2001 ; 12 : 11-22 .  189  Chen XC, Yin TA, He JS, Ma QY, Han ZM. Lower levels of zinc in hair and blood, pica anorexia, and poor growth in Chinese preschool children . Am J Clin Nutr 1985 ; 42: 694700. Chiplonkar SA, Agte VV, Tarwadi KV, Paknikar KM, Diwate UP . Micronutrient deficiencies as predisposing factors for hypertension in lacto-vegetarian Indian adults J Am Coll Nutr 2004 ; 23 : 239-47.  Curran JS, Barness, LA . Nutritional requirements . In: Behrman RE, Kliegman RM, Jenson HB, eds . Textbook of pediatrics. Philadelphia: WB Saunders Company, 2000: 138-142.  Currie J, Madrian B . Health, health insurance and the labor market . In: Card D, . Ashenfelter 0, eds . Handbook of Labor Economics New York: WB Saunders Company, 1999 : 3309-407.  Czaja R, Blair J . Designing surveys : a guide to decisions and procedures . Washington DC : Pine Forge Press. 1996.  Dawson EB, Albers J, McGanity WJ . Serum zinc changes due to iron supplementation in teen-age pregnancy . Am J Clin Nutr 1989 ; 50: 848-52.  Deppisch LM, Centeno JA, Gemmel DJ, Tones NL . Andrew Jackson's exposure to mercury and lead: poisoned president? JAMA 1999 ; 282 : 569-71. Dietetics Canada . Practice-Based Evidence in Nutrition, PEN . 2006 . Retrieved: October18, 2006 from http://www.dieteticsatwork.com/pen/KnowledgePathway .asp?kpid=3 66>  Dubois L, Bedard B, Girard M, Bertrand M, Hamelin AM . Alimentation : perceptions in: pratique et insecurite alimentaire, in enquete sociale et du sante 1998 . Quebec ; Institute de la statistique du Quebec, 2000 . 190  Dubois L, Girard M . Social determinants of initiation, duration and exclusivity of breastfeeding at the population level : the results of the Longitudinal Study of Child Development in Quebec (ELDEQ 1998-2002) . Can J Public Health 2003 ; 94: 300-5.  Dubois L, Girard M . Social position and nutrition : a gradient relationship in Canada and the USA . Eur J Clin Nutr 2001 ; 55: 366-73.  Dubois, L . Food, Nutrition, and Population Health : From scarcity to social inequalities. In : Heymann J, Hertzman C, Barer ML, Evans RG, eds . Healthier Societies . From Analysis to Action . New York : Oxford University Press, 2006 : 417-31.  Eggleton WG . The zinc and copper contents of the organs and tissues of Chinese subjects . Biochem J 1940 ; 34 : 991-7. Environmental Protection Agency of US . Metals by Inductively Coupled Plasma / Mass Spectrometry . Method 3125 . NY: EPA . 1994.  Federal/Provincial/Territorial Advisory Committee on Population Health, Intersectoral Action . Towards Population Health, Ottawa . 1999. Ganapathy SN, Dhanda S . Protein and iron nutrition in lacto-ova vegetarian Indo-Aryan United States residents . Ind J Nutr Dietet 1980 ; 17: 45-52.  Gibson RS, Heath AL, Limbaga ML, Prosser N, Skeaff CM . Are changes in food consumption patterns associated with lower biochemical zinc status among women from Dunedin, New Zealand? Br J Nutr 2001 ; 86: 71-80.  Ganji V, Hampl JS, Betts NM . Race-, gender- and age-specific differences in dietary micronutrient intakes of US children . Int J Food Sci Nutr 2003 ; 54 : 485-90.  Gibson RS, Smit Vanderkooy PD, Thompson L . Dietary phytate x calcium/zinc mill molar ratios and zinc nutriture in some Ontario preschool children . Biol Trace Elem Res 1991 ; 30: 87-94.  Gibson RS, Vanderkooy PD, MacDonald AC, Goldman A, Ryan BA, Berry MA. Growth-limiting, mild zinc-deficiency syndrome in some southern Ontario boys with low height percentiles. Am J Clin Nutr 1989 ; 49: 1266-73. Gibson RS . Assessment of trace element status in human . Prog Food Nutr Sci 1989 : 13: 67-111.  Glanville NT, McIntyre L . Diet quality of Atlantic families headed by single mothers. Can J Diet Pract Res 2006 ; 67: 28-35.  Glynn L, Emmett P, Rogers I; ALSPAC Study Team . Food and nutrient intakes of a population sample of 7-year-old children in the south-west of England in 1999/2000 what difference does gender make? J Hum Nutr Diet . 2005 Feb ;18(1) :7-19 ; quiz 21-3.  Golub MS, Takeuchi PT, Keen CL, Gershwin ME, Hendrickx AG, Lonnerdal B. Modulation of behavioral performance of prepubertal monkeys by moderate dietary zinc deprivation . Am J Clin Nutr 1994 ; 60: 238-43.  Golub MS, Takeuchi PT, Keen CL, Hendrickx AG, Gershwin ME . Activity and attention in zinc-deprived adolescent monkeys . Am J Clin Nutr 1996 ; 64 : 908-15.  Goyder J, Warriner K, Miller S . Evaluating Socioeconomic Status (SES) bias in survey nonresponse . J Stat 2002 ; 18 : 1-11 .  192  Groves RM, Couper MP . Correlates of non response in personal visit surveys . Am Stat Assoc Proc Sect Survey Res Meth 1992 ; 13 : 102-11. Guenther PM . Beverages in the diets of American teenagers . J Am Diet Assoc 1986 ; 86: 493-9.  Gunshin H, Mackenzie B, Berger UV, Gunshin Y, Romero MF, Boron WF, Nussberger S, Gollan JL, Hediger MA . Cloning and characterization of a mammalian proton-coupled metal-ion transporter . Nature 1997 ; 388: 482-8. Hambidge KM . Hair analysis . Pediatr Clin North Am 1980 ; 27 : 855-60. Hambidge KM, Krebs NF . Interrelationships of key variables of human zinc homeostasis: relevance to dietary zinc requirements . Annu Rev Nutr 2001 ; 21 : 429-52. Hambidge KM, Chavez MN, Brown RM, Walravens PA . Zinc nutritional status of young middle-income children and effects of consuming zinc-fortified breakfast cereals . Am J Clin Nutr 1979 ; 32 : 2532-9.  Hambidge KM, Hambidge C, Jacobs M, Baum JD . Low levels of zinc in hair, anorexia, poor growth, and hypogeusia in children . Pediatr Res 1972 ; 6 : 868-74.  Hambidge KM, Walravens PA, Webster J, White S, Anthony M, Roth ML . Zinc nutrition of preschool children in the Denver Head Start program . Am J Clin Nutr 1976 ; 29: 734-8. Health Canada, Canada's Food Guide for Healthy Eating, 2007 . Retrieved March 14, 2007 from http://www. he-sc. gc . ca/fn-an/food-guide-aliment/myguide-monguide/index_e . htmlHeinersdorff N, Tylor TG. Concentration of zinc in the hair of school children . Arch Dis Child 1979 ; 54: 958-60. 193  Hopps HC . The biologic bases for using hair and nail for analyses of trace elements . Sci Total Environ 1977 ; 7 : 71-89.  Ibs KH, Rink L . Zinc-altered immune function . J Nutr 2003 ; 133 : 1452S-1456S. Institute of Medicine . Dietary Reference Intakes for Vitamin A,Vitamin K, Arsenic, Boron, Chromium, Copper, Iodine, Iron, Manganese, Molybdenum, Nickel, Silicon, Vanadium, and Zinc . Washington DC : National Academy Press . 2000. Jackson R, Chambless LE, Yang K . Differences between respondents and nonrespondents in a multicenter community-based study varies by gender ethnicity . The Atherosclerosis Risk in Communities (ARIC) Study Investigators . J Clin Epidemiol 1996; 49 : 1441-1446.  Kirksey A, Wachs TD, Yunis F, Srinath U, Rahmanifar A, McCabe GP, Galal OM, Harrison GG, Jerome NW . Relation of maternal zinc nutriture to pregnancy outcome and infant development in an Egyptian village . Am J Clin Nutr 1994 ; 60 : 782-92. Klevay LM . Hair as a biopsy material . I . Assessment of zinc nutrition . Nutr 1970 ; 23: 284-9.  Kranz S, Hartman T, Siega-Riz AM, Herring AH . A diet quality index for American preschoolers based on current dietary intake recommendations and an indicator of energy balance . J Am Diet Assoc 2006 ; 106 : 1594-604. Kuhnlein HV, Receveur 0, Soueida R, Berti PR . Unique patterns of dietary adequacy in three cultures of Canadian Arctic indigenous peoples . Public Health Nutr 2007 ; 5 :1-12. Lino M, Greeior S, Basiotis P, Anand RS . Reportcard on the diet quality of children . Fam Econ Nutr Revies1999 ; 12 : 78-80.  194  Lombeck I, Wilhelm L, Hafner D, Roloff K, Ohnesorge KL . Hair zinc of young children from rural and urban areas in North Rhine-Westphalia, Federal Republic of Germany. Eur J Pediatr 1988 ; 147: 179-83.  Liptrap DO, Miller ER, Ullrey DE, Whitenack DL, Schoepke BL, Luecke RW . Sex influence on the zinc requirement of developing swine . J Anim Sci 1970 ; 30: 736-41. Mangian HF, Paul GL, Shay NF : Blood leptin levels are reduced during zinc deficiency induced anorexia . FASEB 1997 ; 11 : A1124.  Mantzoros CS, Prasad AS, Frances M, Beck WJ, Grabowski S, Kaplan J, Adair C, Brewer GJ . Zinc may regulate serum leptin concentrations in humans J Am Coll Nutr 1998; 17 : 270-5.  Maxwell SE, Delaney FED . Bivariate median splits and spurious statistical significance. Psychol Bull 1993 ; 113 : 181-90. McCann JC, Ames BN . An overview of evidence for a causal relation between iron deficiency during development and deficits in cognitive or behavioral function . Am J Clin Nutr 2007 ; 85 : 931-45.  McIntyre L, Glanville NT, Raine KD, Dayle JB, Anderson B, Battaglia N . Do lowincome lone mothers compromise their nutrition to feed their children? CMAJ 2003 ; 168: 686-91.  Meadows NJ, Grainger SL, Ruse W, Keeling PWN, Thompson RPH . Oral iron and the bioavailability of zinc . Br Med J 1983 ; 287 : 1013-4. Meng Z . Age- and sex-related differences in zinc and lead levels in human hair . Biol Trace Elem Res 1998 ; 61 : 79-87.  195  Miekeley N, Dias Carneiro MT, da Silveira CL . How reliable is human hair reference intervals for trace elements? Sci Total Environ 1998 ; 218 : 9-17. Mishra SI, Dooley D, Catalano R, Serxner S . Telephone health surveys : potential bias from non-completion . Am J Public Health 1993 ; 83 : 94-9. Muller 0, Krawinkel M . Malnutrition and health in developing countries . CMAJ 2005; 173 : 279-86.  Nakamura T. Nishiyama S . Futagoishi-Suginohara Y, Matsuda I . Higashi A. Mild to moderate zinc deficiency in short children : effect of zinc supplementation on linear growth velocity . J Pediatr 1993 : 123 : 65-69. Nakano T, Fediuk K . Kassi N, Egeland GM, Kuhnlein HV . Dietary nutrients and anthropometry of Dene/Metis and Yukon children . Int J Circumpolar Health . 2005 Apr;64(2):147-56.  North K, Emmett P . Multivariate analysis of diet among three-year-old children and associations with socio-demographic characteristics . The Avon Longitudinal Study of Pregnancy and Childhood (ALSPAC) Study Team . Eur J Clin Nutr 2000 ; 54 : 73-80. Oberleas D, Muhrer ME, O'Dell BL . Dietary metal-complexing agents and zinc availability in the rat . J Nutr 1966 ; 90: 56-62.  Olson C, Rauschenbach BS, Frangillio EA Jr, Kendall A . Factors contributing to food insecurity in a rural upstate New York county . Fam Econ Nutr Rev 1997 ; 10: 2-17. Passmore C, Dobbie AE, Parchman M, Tysinger J . Guidelines for constructing a survey. Fam Med 2002 ; 34 : 281-6.  196  Ploysangam A, Falciglia GA, Brehm BJ . Effect of marginal zinc deficiency on human growth and development . J Trop Pediatr 1997 ; 43 : 192-8. Prasad AS . Zinc deficiency in human subjects . Prog Clin Biol Res 1981 ; 77: 165-77. Prasad A . Zinc in growth and development and spectrum of human zinc deficiency . J Am Coll Nutr 1988 ; 7 : 377-84. Prentice A . Does mild zinc deficiency contributes to poor growth performance? Nutr Rev 1993; 51 : 268-70.  Public Health Agency of Canada . Population Health . 2002 .Retrieved July 2, 2005 from www.phac-aspc.gc.ca/ph-sp/phdd/approach/approach .html Puchyr RF, Bass DA, Gajewski R, Calvin M, Marquardt W, Urek K, Druvan ME, Quig D . Preparation of Hair for Measurement of Elements by Inductively Coupled PlasmaMass Spectrometry (ICP-MS) . Biol Tr Elem Res 1998 ; 62 : 167-82. Putnam SP, Rothbart MK . Development of short and very short forms of the Children's Behavior Questionnaire . J person assess 2003 ; 87: 103-13.  Radimer KL, Olson CM, Greene JC, Campbell, CC, Habicht J . Understanding hunger and developing indicators to assess it in women and children . J Nutr Ed 1992 ; 24: S3644.  Robertson I, Glekin BM, Henderson JB, McIntosh WB, Lakhani A, Dunnigan MG. Nutritional deficiencies among ethnic minorities in the United Kingdom . Proc Nutr Soc 1982; 41: 243-56 .  197  Rothbart MK, Ahadi SA, Hershey KL, Fisher P . Investigations of temperament at the three to seven years the children's behavior questionnaire . Child Dev 2001 ; 72 : 1394408.  Roville-Sausse FN . Westernization of the nutritional pattern of Chinese children living in France. Pub Health 2005 ; 119 : 726-33.  Sakai T, Warishi M, Nishiyama K. Changes in trace element concentrations in hair of growing children. Biol Trace Elem Res 2000 ; 77 : 43-51.  Sandstead HH, Penland JG, Alcock NW, Dayal HH, Chen XC, Li JS, Zhao F, Yang JJ. Effects of repletion with zinc and other micronutrients on neuropsychological performance and growth of Chinese children. Am J Clin Nutr 1998 ; 68: 470S-75S.  Sandstrom B, Davidsson L, Cederblad A, Lonnerdal B . Oral iron, dietary ligands and zinc absorption . J Nutr 1985 ; 115 : 411-14.  Sardell A. Child health policy in the US : the paradox of consensus . J Health Polit Policy Law 1990; 15 : 271-304.  Satia JA, Patterson RE, Taylor VM, Cheney CL, Shiu-Thornton S, Chitnarong K, Kristal AR. Use of qualitative methods to study diet, acculturation, and health in ChineseAmerican women . J Am Diet Assoc 2000; 100 : 934-40.  Saw SM, Ng TP . The design and assessment of questionnaires in clinical research Singapore Med J 2001 ; 42: 131-5.  Sazawal S, Bentley M, Black RE, Dhingra P, George S, Bhan MK . Effect of zinc supplementation on observed activity in low socioeconomic Indian preschool children. Pediatr 1996; 98 : 1132-7.  198  Shahar E, Folsom AR, Jackson R . The effect of non-response on prevalence estimates for a referent population : insights from a population-based cohort study . Ann Epidemiol 1996; 6 : 498-506.  Sheng XY, Hambidge KM, Zhu XX, Ni JX, Bailey KB, Gibson RS, Krebs NF . Major variables of zinc homeostasis in Chinese toddlers . Am J Clin Nutr 2006 ; 84: 389-94. Silva A, Whitman S, Margellos H, Ansell D . Evaluating Chicago's success in reaching the Healthy People 2000 goal of reducing health disparities . Public Health Rep 2001; 116 : 484-94.  Skalicky A, Meyers AF, Adams WG, Yang Z, Cook JT, Frank DA . Child food insecurity and iron deficiency anemia in low-income infants and toddlers in the United States. Matern Child Health J 2006 ; 10 : 177-85.  Skinner JD, Carruth BR, Houck KS, Bounds W, Morris M, Cox DR, Moran J 3rd, Coletta F . Longitudinal study of nutrient and food intakes of white preschool children aged 24 to 60 months . J Am Diet Assoc 1999 ; 99 : 1514-21. Smit Vanderkooy PD, Gibson RS . Food consumption patterns of Canadian preschool children in relation to zinc and growth status . Am J Clin Nutr 1987 ; 45 : 609-16. Solomons NW, Jacob RA . Studies on the bioavailability of zinc in humans : effects of heme and nonheme iron on the absorption of zinc . Am J Clin Nutr 1981 ; 34 : 475-82. Solomons NW, Pinedea 0, Viteri F, Sandstead H . Studies on the bioavailability of zinc in humans : mechanism of the intestinal interaction of nonheme iron and zinc . J Nutr 1983; 113 : 337-49.  Solomons NW . Competitive interaction of iron and zinc in the diet : consequences for human nutrition . J Nuts- 19W. 116 . 927-35 . 199  Statistics Canada. 2001 Community Profiles : Vancouver Census Metropolitan Area. Ottawa :Statistics Canada, 2001 . Retrieved April 11, 2005 from http ://www 12 . state an .ca/english/profil01 /CPO 1 /Index .cfm?Lang=E Statistics Canada, 2005 . Retrieved August 20, 2007 from http://www .statcan .ca/english/research/89-6 1 3 -MIE/20060 10/tables/table3 . htm Statistics Canada. Food consumption highlights . 2007. Retrieved May 22, 2007 from http://www. statcan .ca/english/ads/23F0001 XCB/highlight .htm Swenerton H, Hurley LS . Severe zinc deficiency in male and female rats . J Nutr 1968; 95 : 8-18.  Touvier M, Lioret S, Vanrullen I, Bocle JC, Boutron-Ruault MC, Berta JL, Volatier JL .Vitamin and mineral inadequacy in the French population : estimation and application for the optimization of food fortification .Int J Vitam Nutr Res . 2006 Nov ;76(6) :343-51. Troost FJ . Brummer RJ, Dainty JR, HoogewerfT JA . Bull VJ, Saris WH . Iron supplements inhibit zinc but not copper absorption in vivo in ileostomy subjects . Am J Clin Nutr 2003 ; 7 : 1018-23. University of Oregon . Mary Rothbart's the instrument . 1996 . Retrieved Oct 11, 2005 from http : //www . bowdoin . edu/— sputnam/rothbart-temperament-questionnaires/i nstrumentdescriptions/chi ldrens-behavior-questionnaire . html  Valberg LS, Flanagan PR, Chamberlain MJ . Effects of iron, tin, and copper on zinc absorption in humans . Am J Clin Nutr 1984 ; 40 : 536-41 .  200  Van Belle. Statistical rules of thumb . Washinton DC : John Willey & Sons publication, 2002.  Van Wouwe JP, De Wolff FA, Van Gelderen HH . Zinc in hair and urine of pediatric patients . Clin Chim Acta 1986 ; 155 : 77-82. Walravens PA, Hambidge KM . Growth of infants fed zinc-supplemented formula. Am J Clin Nutr 1976 ; 29: 1114-21.  Walravens PA, Krebs NF, Hambidge KM . Linear growth of low income preschool children receiving a zinc supplement . Am J Clin Nutr 1983 ; 38: 195-201.  Watt RG, Dykes J, Sheiham A . Socio-economic determinants of selected dietary indicators in British pre-school children . Public Health Nutr 2001 ; 4 : 1229-33.  Wilkinson, R . Unhealthy Societies : The Afflictions of Inequality . New York : Routledge, 1996.  Wilkinson, R ., & Marmot, M . Social Determinants of Health : The Solid Facts. Copenhagen : WHO, 1998.  Williams PL, Iron status among infants 8-26 months of age in Vancouver and sociocultural/dietary predictors of risk of iron deficiency anemia . Doctoral dissertation in the school of Family and Nutritional Sciences . The University of British Columbia: Vancouver, 2000. World Health Organization . Targets for health for all . Report of WHO region:( Office for Europe . Copenhagen: WHO, 1991. Zachwieja Z, Chlopicka J, Schlege-Zawadzka M, Zagrodzk P, Wypchlo J, Krosniak M. Evaluation of zinc content in children's hair . Biol Trace Elem Res . 1995 ; 47: 141-5 .  Zill N. Parental schooling and children's health . Public Health Rep 1996 ; 111 : 34-43.  Zill N, Coffins M, West J, Germino-Hausken E . Approaching kindergarten : a look at preschoolers in the U.S.  Washington, DC : Department of Education, National Center for  Education Statistics, 1995 .  202  CHAPTER IV CONCLUSION  203  IVA. GENERAL CONCLUSIONS This study, being the first conducted in the last 20 years on the zinc status of Canadian preschoolers, provides important hair zinc reference values for healthy Canadian preschoolers . Due to its relatively large sample size and with the participants being healthy free-living children, these values may also serve as a reference for healthy preschoolers of industrialized societies . Our study makes, as well, a significant contribution to our understanding of hair zinc and factors associating with it. First and foremost, we observed the occurrence of low hair zinc in 46% of our low-income inner city children (Chapter II), and in 17% of our survey participants (a representative sample of the whole city) (Chapter III) . The much higher population mean of the survey children compared to that of the low-income inner city children when combined with the much lower incidence of low hair zinc for the survey children (17% versus 46% in the inner city sample) conveyed a relatively better picture of zinc status for the preschoolers of the city . In addition, this comparison reiterated the well-recognized fact that children of low-income families are at higher risk of sub-optimal nutritional status . Radical differences (similar to ours) in the occurrence of low hair zinc between low-income (Hambidge et al, 1972) and middle income (Hambidge et al 1976) children of the same society have also been shown in American preschoolers (3-5 years) (48% versus 19%, respectively) . The comparability of these numbers with ours (46% versus 17%, respectively) as well as the similarity of the differences observed between the 2 income classes of these 2 sets of data, are quite remarkable . The higher occurrence of low hair zinc among low-income inner city preschoolers compared to the survey (the overall city) was not unexpected . Nevertheless, the extent of the occurrence of low hair 204  zinc (46%) was alarming. Almost 1 out of every 2 apparently healthy low-income preschoolers from inner city of Vancouver, a major metropolitan center in one of the healthiest and most affluent countries of the world (WHO, 1991), had low hair zinc and, potentially, MZD. It is evidence like this that puts the assertion of marginal zinc deficiency as the most common hidden health problem in children (Hakim et al, 2006) into context. Although the occurrence of low hair zinc was significantly higher among the lowincome and potentially food insecure inner city children (46% versus 17%, x 2=34 df 1, p=0.00), there was no association between income and hair zinc within either of the samples . Taken together, our results indicated that the recognized "gradient" effect of income on health status (Hertzman, 1998) might not govern hair zinc status (MZD) . All socio-economic classes are at risk for the occurrence of low hair zinc and the differences only become measurable in extreme conditions . Comparing the results of our survey combined with that of our inner city study (Chapter II) and the study of Broughton et al (2006) which explored food security among Vancouver inner city families (the same cohort used in chapter II of this thesis) provides some support for this statement. Broughton and co-workers estimated 50% food insecurity for that group of inner city families of Vancouver while we observed the occurrence of low hair zinc for the same cohort at 46% . Yet in our survey we observed the rate of food insecurity and low hair zinc at 14% and 17%, respectively. Looking at these 2 studios may provide a support to the statement that the impact of a decline in income level on hair zinc becomes measurable only when the decline is severe and the food security of the child is jeopardized . 205  Looking at the results of two studies also indicated that while low income children of first study had very low mean hair zinc (73µg/g), the low income children of survey had much higher mean hair zinc as a group (113 µg/g) . There are few possible explanations for this observation . First, this could be the result of some differences in the living environment of children at the neighborhood level . The available research suggests that some favorable factors at neighborhood level could attenuate some of the negative effects of low SES on children (BC child atlas, 2005). Some examples of these factors, relevant to the issue of discussion, could be having community centers offering nutrition and health classes to caregivers, having food banks or other initiatives promoting early child development. Alternatively, the observed differences in the mean hair zinc of inner city children and low income children of survey could be the result of a measurement error, as the samples of two studies were analyzed in two different laboratories. While considering the ill effects of poverty on health, it is reasonable to assume that the observed differences are real. But it is also possible that some differences in the design and methodology of the 2 studies may have produced the observed drastic differences. The methodological differences between these studies include the fact that the hair zinc analyses were carried out in different laboratories . In addition, the amount of hair samples collected in inner city study were just enough to carry out single analyses, whereas the samples of our survey were analyzed as single samples with a duplicate after every 10 samples. However, the small coefficient of variation of the survey samples (3 .3% for within assay and 6 .0% for between assay) may indicate that the results of single analyses may still be reliable . 206  In both of our studies we observed significant differences in hair zinc as well as the occurrence of low hair zinc between children younger than 4 years and those who were 4 and over . This finding raises a fundamental question : Could this lower hair zinc in younger preschoolers be a normal physiological level independent of diet or any other environmental factor? This is an important question to address because, if it is, this will undermine the usefulness of having the same cutoff for younger children . While this critical question remains to be answered, perhaps comparing the results of studies with different age groups may not be advisable. Furthermore, we observed differences in the occurrence of low hair zinc based on sex . Unlike the conclusion in most zinc literature (Hambidge et al, 1976, Gibson & DeWolfe, 1979, Smit Vanderkooy & Gibson, 1987 and Sky-Peck, 1990) indicating male children as the sex at higher risk for low hair zinc, our data showed the female preschoolers as the more vulnerable sex (a higher occurrence of low hair zinc for girls but a comparable mean hair zinc for the 2 sexes) . Previous studies speculated a greater physiological need in males based on evidence from animal studies (Liptrap et al, 1970 and Swenerton & Hurley, 1968) as the reason for the observed differences . However, our finding, combined with the findings of other studies similar to ours, which have not found a significant difference in hair zinc between the 2 sexes (Sakai, Warishi & Nikiyama, 2000) or have actually observed higher hair zinc for male children (Zachiwieja et al, 1995) points to the inadequacy of physiology by itself as the explanatory factor for this observation. Our finding that Chinese children have a higher susceptibility for low hair zinc was not unexpected in view of the nature of a traditional Chinese diet, which was 207  supposedly consumed by these children . However, we also made the unanticipated observation that not a single one of our fifty-four East Indian children was low in hair zinc. Like sex differences in our study, differences in the hair zinc level and the occurrence of low hair zinc among different ethnic groups could not be explained based on the limited dietary information gathered from the children . Whether these differences were due to the eating patterns of the ethnic groups and our dietary questionnaire was not sensitive enough to capture them, or due to some inherent factors in each ethnicity, we cannot answer. Our study also indicated an inverse association between taking supplements containing iron and hair zinc . When we explored the relationship between taking supplements containing iron and other eating behaviors, we observed positive associations between "taking supplements containing iron" and "caregiver being concerned about child's nutrition", being described as "picky eater" and being described as "not eating enough" . Taken together, we speculated that the negative association of hair zinc with taking supplements containing iron might have stemmed from the fact that most of the iron takers were children whose nutrition was a concern of the caregivers, and children who had been described as a "picky eater" or "not eating enough" . Due to perceiving child with these eating characteristics, therefore, the caregivers had begun giving them these supplements (in order to provide the nutrients that they thought the children were not getting through their diet) . However, in our final logistic regression model, both variables of "taking supplements containing iron" and being described as "not eating enough" remained as significant (albeit very weak) predictors of hair zinc status . 208  While the literature has documented the ill effect of iron supplements on zinc absorption as well as zinc status, we refrained from drawing such a conclusion based on our data due to the very small concentration of iron taken by our children (4-18mg/day), as well as the fact that iron was often con-consumed with zinc . On the other hand this study is the first population-based study to document such an association . Therefore, and in view of the significance of the iron and zinc balance in human health, it is reasonable to raise this issue as another question generated by this study that requires to be addressed properly. Finally, we showed associations of hair zinc with some health and behavior related variables. In both groups of children (the low income children of inner city as well as the survey children), there were negative associations between being described as "frequently sick" and the children's hair zinc. In our survey children we also noticed a positive association between their hair zinc and the activity level score . Evidence from human studies, though scarce, does refer to decreased activity level as a consequence of zinc deficiency . The literature does not offer a plausible explanation for this finding. In this study there was also no evidence of any association between hair zinc and growth status, an association shown in many, but not all, studies of MZD . Although the anthropometric data did not support the existence of any overt growth retardation, this does not eliminate the possibility of the existence of MZD among our study for the following reasons. First, children with marginally low hair zinc and normal growth parameters have been shown to exhibit a differential growth response (compared to growth in the placebo group) when supplemented with zinc (Ruz et al, 1997) . Since zinc has no 209  pharmacological growth promoting effect and it results in growth only when the growth had been hindered due to the shortage of zinc in the human body, this growth response indicates a pre-existing zinc deficiency . Therefore, in our study, while the existence of measurable anthropometric evidence could support the diagnosis of MZD, the absence of such data does not eliminate this diagnosis. Second, there was a suggestion of a compromised immune system among some of these children . The low hair zinc children were more often described as having poor health and being frequently sick . The literature has also documented a more frequent occurrence of cold and common infections among children with MZD. Third, we observed a significantly lower activity level score among our low hair zinc children. This reduced activity level has also been recognized in children with suboptimal zinc nutriture. This evidence taken together, we believe that at least a fraction of these low hair zinc children were potentially marginally zinc deficient. At this point and within the constraints of our data, this is the extent of the conclusion we can draw . The confirmation or elimination of this possibility (the existence of MZD among these children) can only be substantiated through a zinc supplementation study. IV.2. STRENGTHS AND LIMITATIONS OF THE STUDY Our citywide survey (chapter III) was the first large study of zinc status in Canada. There are a few characteristics in our sampling (and methodology in general) that could have improved the quality of our data and increased the credibility of our findings .  210  We invested a great deal of care and deliberation in our sampling strategy . The significance of this survey is not just in its large sample (n=719), but also in the representativeness of this sample . Through the use of information from the 2001 census (Statistics Canada, 2001), and employing a stratified sampling method (details in chapter III) we ensured that every neighborhood in the city of Vancouver was adequately represented . In addition by means of useful maps produced by the combined effort of the UBC data library and the Geographic Information System, GIS, of the Human Early Learning Partnership (HELP) (Appendix III-2) we ensured that the intra-neighborhood income differences were carefully considered and properly reflected in the subjects recruited from each neighborhood . We believe the survey had a very fair representation of children of all neighborhoods and all income classes within them. The other area in the methodology of our citywide survey in which a great deal of care was invested was in the hair sample collection itself. It has been shown that the collection of a hair sample is a critical factor in the credibility of the hair zinc results . In order to ensure strict adherence to one protocol, and in an attempt to minimize the interexaminer error (a bias arising as a result of subtle differences in the methods of different examiners) all of our samples were collected by the research student only . Although this extended the sample collection over a longer time period (March-June), it clearly assured the reliability of the sample collection. Finally, the large sample size of the survey and the fact that all the participants were apparently healthy free-living children may permit the generalization of our findings to many North American preschool populations with similar ethnic make up .  211  As with any study, ours had its limitations . First, we recruited most of our survey children (83%) through the preschool and daycare centers of the city, and only 15% (Personal communication with Ms . Diane Liscumb, West Coast Family Services) of the families with preschoolers use these services . Using this database we excluded the children who are not enrolled in these services . It is reasonable to assume, therefore, that the children of stay-home mothers were excluded by our sampling method . A large portion of such parents could have been unemployed due to a lack of training, skill and/or education . We might, therefore, have skewed the maternal education of the study participants towards the higher end of the spectrum and thus introduced a selection bias into the study . Perhaps the high maternal education observed among a large proportion of our survey participants is indicative of this bias (58% versus 38% and 42% reported for adult females 20-34 and 35-44 years old, respectively) (Statistics Canada, 2001). Another factor to bear in mind when considering the extent to which our findings can be generalized is the moderate response rate of the survey . Our overall response rate was 30%, with the other 70% of the families receiving our survey package not replying. Non-response in a survey can pose a threat to the generalization of results if respondents and non-respondents differ systemically (Barriball & While, 1999) . We did not have any way of verifying the existence of any systemic differences between the respondents and non-respondents . The research evidence provides support for significant differences in the socio-economic (Groves & Couper, 1992 and Mishra et al 1993) as well as the health (Shahar, Folsom, & Jackson, 1996 and Mishra et al, 1993) status between respondents and non-respondents .  212  The third possible limitation of our study was that most of our data was parent reported . Therefore, the validity of the data depended on the truthfulness and memory of the responding caregiver/parent . The first factor, truthfulness, particularly, becomes an issue when it comes to reporting the food intake of the children . Despite the fact that the estimates of adult food and energy intake is often under-reported (Kraft & Dwyer, 1987) parents may over report their children's intake (Devaney et al, 2004) for the simple reason of social acceptability . In addition, it may be difficult for them to keep in mind that what is served and what was consumed are not often the same . Lastly, parents may vary in their idea of individual serving sizes (Young et al, 1998), and what is 1 serving size for one parent may not be the same for another. As for reliance of our data on a parent's memory, research in this area indicates that the caregiver's recall of events regarding their children can indeed be considered a reliable source and has demonstrated a very good correlation with some gold standards such as a pediatrician's record of events (Pless & Pless, 1995) . The high R2 values of the associations of the 2 sets of data collected during our reliability study are also testimony to this statement. One limitation specific to the study of the inner city was the small quantity of the hair samples which did not allow any analysis to be run in duplicate. Finally, as it is with any cross sectional study, our data indicates associations at the very best. Therefore, we cannot make any conclusion regarding the existence of any casual relationship between the variables studied or for that matter, or lack of it .  213  IV.3. FUTURE DIRECTIONS This work, while providing some useful information about the zinc status of Vancouver's preschoolers and hair zinc in general has generated some very important questions that need to be answered . One such is the hair zinc status of low-income preschoolers of the city (Chapter II) where we observed widespread low hair zinc among the inner-city children . Further studies are to be conducted to verify this in larger samples of low-income preschoolers . Carefully designed studies are also to be implemented to better understand the functional/clinical ramifications of low hair zinc for these children . For instance, longitudinal studies with serial anthropometric measurements and measure of growth velocity would enable one to understand the consequences of such low hair zinc on a child's physical growth, if any . Alternatively, supplementation studies would be able to determine if provision of zinc triggers any growth response in children, which would be confirmation for a pre-existing zinc deficiency. Our study indicated children younger than 4 years old as those at higher risk of low hair zinc . The hair zinc of these children was significantly lower than that of the older children . It is critical to understand whether this observation is simply the normal physiological status of this stage of life or a deviation from the normal range due to different environmental factors such as nutrition . An answer to this question is extremely important, since, as discussed, if this observation is simply the result of the normal physiology of this stage of life, there is a need for setting up a different cut off and validating it in a large sample of children of this age group .  214  Lastly, differences observed in the hair zinc level of the children of different ethnicities (Chapter III) warrant further research . Studies with a more representative and larger sample of Chinese as well as East Indian children are to be conducted to confirm these findings. Such studies should gather detailed and elaborate dietary data and information on family background in order to better understand the children's diet as well as their social setting . Such studies would enable us to determine if ethnicity per se or other ethnicity-related factors (for example ethnic food) is the underlying explanation for the observed differences in hair zinc . It is possible that there are lessons to learn from the ethnicities with unexpected rate of low hair zinc. All in all, this work has increased our understanding of the factors associated with the hair zinc of children and in doing so has brought into focus several important questions that may provide direction for future investigations which will further our knowledge of hair zinc and factors affecting it .  215  IV.4.BIBLIOGRAPHY Barriball KL, While AE . Non-response in survey research : a methodological discussion and development of an explanatory model . J Adv Nurs 1999; 30 : 677-86. Broughton MA, Janssen PS, Hertzman C . Innis SM, Frankish CJ . Predictors and outcomes of household food insecurity among inner city families with preschool children in Vancouver . Can J Public Health 2006 ; 97: 214-6.  Devaney L, Kalb R, Briefel T, Zavitsky-Novak N, Ziegler P . Feeding Infants and Toddlers Study Overview of the study design . J Am Diet Assn 2004 ; 104 : S8-S13. Gibson RS, DeWolfe MS . The zinc, copper, manganese, vanadium, and iodine content of hair from 38 Canadian neonates . Pediatr Res 1979 ; 13 : 959-62. Groves RM, Couper MP . Correlates of non-response in personal visit surveys . Am Stat Assoc Proc Sect Survey Res Meth 1992 ; 13 : 102-11.  Hakimi SM, Hashemi F, Valaeei N, Seyed-Masood K, Velayati AA, Boloursaz MR. The effect of supplemental zinc on the height and weight percentiles of children. Arch Iran Med 2006 ; 9: 148-52.  Hambidge KM, Hambidge C, Jacobs M, Baum JD . Low levels of zinc in hair, anorexia, poor growth, and hypogeusia in children . Pediatr Res 1972 ; 6 : 868-74.  Hambidge KM, Walravens PA, Brown RM, Webster J, White S, Anthony M, Roth ML. Zinc nutrition of preschool children in the Denver Head Start program . Am J Clin Nutr 1976; 29: 734-8.  Hertzman C . The case for child development as a determinant of health . Can J Public Health 1998; 89: S16-21 .  216  Krall EA, Dwyer JT . Validity of a food frequency questionnaire and a food diary in a short-term recall . J Am Diet Assoc 1987; 87: 1374-7.  Liptrap DO, Miller ER, Ulirey DE, Whitenack DL, Schoepke BL, Luecke RW . Sex influence on the zinc requirement of developing swine . J Anim Sci 1970 ; 30: 736-41.  Mishra SI, Dooley D, Catalano R, Serxner S . Telephone health surveys : potential bias from non-completion . Am J Public Health 1993 ; 83 : 94-9. Pless CE, Pless IB . How well they remember . The accuracy of parent reports . Arch Pediatr Adolesc Med 1995 ; 149 : 553-8. Ruz M, Castillo-Duran C, Lara X, Codoceo J, Rebolledo A, Atalah E . A 14-month zincsupplementation trial in apparently healthy Chilean preschool children . Am J Clin Nutr 1997; 66: 1406-13.  Sakai T, Warishi M, Nishiyama K . Changes in trace element concentrations in hair of growing children . Biol Trace Elem Res 2000 ; 77 : 43-51. Shahar E, Folsom AR, Jackson R . The effect of non-response on prevalence estimates for a referent population : insights from a population-based cohort study . Ann Epidemiol 1996; 6: 498-506.  Sky-Peck HH . Distribution of trace elements in human hair . Clin Physiol Biochem 1990; 8 : 70-80.  Smit Vanderkooy PD, Gibson RS . Food consumption patterns of Canadian preschool children in relation to zinc and growth status . Am J Clin Nutr 1987 ; 45 : 609-16.  217  Statistics Canada . 2001 Community Profiles :Vancouver Census Metropolitan Area. Ottawa : Statistics Canada, 2001 . Retrieved April 11, 2005 from http : //www 12 . statcan . ca/eng l i sh/profil O 1 /CPO 1 /Index . cfm?Lang=E Swenerton H, Hurley LS . Severe zinc deficiency in male and female rats . J Nutr 1968; 95 : 8-18.  World Health Organization Regional Office for Europe . (1991) . Targets for health for all. Copenhagen.  Young LR, Nestle M . Variation in perceptions of a "medium" food portion : Implications for dietary guidance . J Am Diet Assoc1998 ; 98 : 458-59.  Zachwieja Z, Chlopicka J, Schlegel-Zawadzka M, Zagrodzki P, Wypchlo J, Krosniak M. Evaluation of zinc content in children's hair . Biol Trace Elem Res 1995 ; 47: 141-5.  218  APPENDICES  219    Appendix II-l : Certificate of ethic approval for the inner city study  .  The University of British Columbia Office of Research Services Clinical Research Ethics Board Room 210, Research Pavilion, 625 West 10a Avenue, Vancouver, BC V5Z 11.A  PRNCPAL HVESTIGATOR  Notice OEPARTRENT of Ethical Review : Memo  Innis, S.M . pSNTUTEN(5) CO  • . ..  Paediatrics  WHERE RESEARCH Wal  aE  CARRIED OLT  BCCH Research Centre  INVESTIGATORS  sPONSORHO AGENCIES  —Mu Pt  January, 2004  BC Ministry of Children and Family Development  Cs-Morbidity of Iron Deficiency & Mi~yonutrieni {$E;hitdreo 18 Months- 5 r cars of Ages  Deficiency it Risk of  Ne ;  in  The UBC CREB has reviewed your proposed study, and has issued a Certificate of Approval. Please make sure the following items have been corrected/amended within 21 days . Ensure that any revisions to the consent form are made on the correct institutional letterhead and are either underlined or in BOLD text . A letter that explains how the requested changes have been addressed should accompany your response. If needed for Clinical Trials, please ensure that the Consent Form Version Date is updated to reflect any required revisions and that the new version date is clearly indicated in your response . An updated Certificate of Approval will be issued accordingly. All other responses will be acknowledged by the CREB Chair . You may submit the revisions by email to Catherine Sutherland at c.sutherland@ a ors.ubc.ca  Consent: I . Include page numbering. 2. Include an invitation to participate and an explanation as to why they are being invited. 3. Page 3 : Include the standard UBC CREB confidentiality statement as per UBC CREB Guidance Notes 39.7 .1 . Refer to http ;//www .ors.ubc .ca/et}tice/fomu/GNinitialavp .htm#Guide39.7 4. Page 2 : Change "Director of Research Services" to "Research Subject Information Line" . Phone number remains the same. 5. Include space for the printed names of all who sign the consent.  If you have any questions regarding these requirements, please call: Dr. Peter Loewen, Chair, 604-822-7985 Dr. Main Gagnon, Associate Chair, 604-875-3174 Dr. James McCormack, Associate Chair, 604-822-1710 Ma . Susan Clumick, Manager Ethical Reviews (ORS), 604-875-4149 For furtherinformation.refer to the UBC CREB's GuidanceNotesandPolicies at: www,ors.ubc.cs/ethics/hunan subjects.htni. Please send all correspondence to : Clinical Research Ethics Office, Room 210, Research Pavilion, 828 West 10 th Avenue. Vancouver. BC Canada V5Z 1L8  220    Appendix II-2.Recruitment questionnaire for the study of low-income preschoolers  Children's Eating and Nutrition Clinic Clinic Sign-Up Form Child's gender: Boy O Girl O Child's Date of birth  / /  Year/mo/day  1 . Are you concerned about your child's eating?  q Yes O No  Why are you concerned? Please check all your concerns. q Picky eater O Doesn't eat often enough Cl Won't try new foods D Eats too much O Always eats the same foods 0 Eats unhealthy q Eats too often O Not enough variety in diet 0 Doesn't eat enough q Other  2. Does your child take a vitamin/mineral supplement?  q Yes No Please list name of supplement(s) and how often taken.  How long has he/she been taking it?  221    OYes  No  To what food(s)?  4. Is your child often sick (flu, fever, diarrhea . . .)  O  Yes  O No  5 . How would you describe your child's health: O Very healthy El Average health a Not so healthy  6 . Does your family ever eat meat, chicken or fish? O yes O no  7. Ethnic Background  Mother's Ethnicity  Father's  Ethnicity  The time needed for this clinic will be about 60 minutes.  If you would like to participate in this study please write your name, address and phone number below.  O Yes/Maybe - I would like to take part in the research .  222    Name  Phone Number  Address  Street number and name  City  Province  Postal  Code  What time of day would you like your clinic appointment? Wednesday, March 3 0 Morning  O Afternoon  Specify time  Thursday, March 4 0 Morning  O Afternoon  Specify time  We will call you to confirm your appointment time .  223  Appendix II-3 : Add placed on the bulletin boards of the community centers inviting eligible families to attend the nutrition clinics  'VIr.-Yr ..* :.V  .' .7re  Children's Eating and Nutrition Our study allows parents of children aged 2 to 5 years to bring children to a clinic to find out about their nutritional health .  We are interested in finding out  what young children eat today and how the foods they eat affect their health and growth.  Who should think about coming to the clinic? q If your child is 2-5 years old and you live in Vancouver, please come.  What are the benefits of participating? q Learn about the health of your child. O Vitamin-mineral supplements for children, if a deficiency is found . O Grocery store gift certificate for parents.  What happens if my child and I are involved? q Your child will be weighed and measured. O A blood sample will be drawn by a pediatric phlebotomist (a blood drawing specialist from BC Children's Hospital) O You will be asked questions about your child .  224  Your information will be strictly confidential . No information that identifies you or your child will be given to anyone without your permission.  Where is the clinic? Strathcona Community Centre, 601 Keefer Street Britannia community centre 1661 Napier Street  When? Wednesday, March 3 and Thursday, March 4, 2004 at Stratchona AND Monday March the 29 th and Tuesday March the 30 th at Britannia q Clinic is open from 9am — 5 pm. q Your visit will take 30-60 minutes.  To find out more or to register : we will be at Grandview, 2075 Woodland Drive on Monday February 23 rd 4pm-6pm REFRESHMENTS WILL BE SERVED  225    Appendix II-4 : Informed Consent form for the study of low-income preschoolers CO-MORBIDITY OF IRON DEFICIENCY AND MICRONUTRIENT DEFICIENCY Informed Consent Principal Investigator  Dr. Sheila Innis Department of Pediatrics, Faculty of Medicine University of British Columbia  Emergency Phone Number (604) 875-2434 Site  B .C .'s Children's Hospital  Summary of Project Iron deficiency is often considered to be the most common nutritional deficiency in Canada and among children worldwide . Iron deficiency is an important health problem because it can lower children's activity and interaction with their environment . This can have important effects on learning . Many of the foods that are good sources of iron are also good sources of zinc . Zinc is important for growth, immune function and also has important effects on learning . We have recently found low zinc levels in about 4 in 10 preschool children who attended our nutrition research clinics . Zinc deficiency and iron deficiency are both easily treated . The purpose of this study is twofold . In the first part we will examine children's diets in a series of short questions and examine the relation of children's diets to the risk of iron and zinc deficiency . We would like to better understand how many children have zinc deficiency, and if zinc deficiency is due to inadequate zinc intake or if absorption is limited by large amounts of cereal fibers in children's diets.  I understand that by participating in this study I will: 1) 2) 3)  Allow my child's weight, height, arm circumference and triceps skin fold thickness to be measured. Provide information on my child's usual food intake, and complete a questionnaire on some of my child's behaviour. Allow a registered technologist to obtain a small blood sample from my child (about '/z teaspoon), and allow a small sample of my child's hair to be collected by myself or the technician present at the clinic.  The benefit that I will receive from this study is an assessment of my child's iron and zinc status . The participation of my child in this study will also contribute valuable information on how dietary intakes of preschool children are linked to iron and zinc 226    nutrition. My participation in this study will require about 60 minutes of my time . My name and my child's name will be treated confidentially by the use of code numbers and will not be used in any report of this study . I may refuse to participate or withdraw from this study at any time with no consequence to my child or me. Risks There are no known risks to participating in this study . A certified technologist or nurse will draw the small amount of blood . Minor discomfort and some temporary discoloration may occur at the site of the blood draw. If I have any questions about this study I may contact Ziba Vaghri at 604-875-2492 or Dr. Sheila Innis at 604-875-2431 at any time. Consent The objectives and procedures of this study have been explained to me to my satisfaction. I understand that I may withdraw from this study at any time . If I have any concerns about my treatment or rights in this study I may telephone the Director of Research Services at 604-822-8598 . I acknowledge that I have received a copy of the consent form for my own records. Compensation for Injury Signing this consent form in no way limits your legal rights against the sponsor, investigators, or anyone else. I  voluntarily give consent for my child (Parent, please print)  to participate in the study entitled (Please print) Co-Morbidity of Iron Deficiency and Micronutrient Deficiency, and Risk of cognitive Delays in Preschool Children 18 months - 5 years of Age The parent(s) and the investigator are satisfied that the information contained in this consent form has been explained and that all questions have been answered.  Signed  Date  Witness  Date  Investigator's Signature  Date  227  PASSPORT Welcome to the Adventures to Good Health! Travel and visit all four of our healthy friends and collect your prizes.  The Brave Bee  This is to certify that has completed the Adventures to Good Health!  Congratulations!  223  Appendix II-6 : An excerpt from the socio-demographic questionnaire of the study of "Food insecurity in inner city children", which collected data on maternal education and familyincome of the study participants _Q3 Please look at the card and estimate in which of the following groups your total household income from all sources falls? Tell me the Code letter of the range (a card showing a table containing income levels for different family sizes is shown to the clientwill be attached to the final thesis). Household Income  Code  1 . Response categories : <50% of LICO, 2, 50-100% of LICO 3 . 100-150% of LICO, 4, >150% of LICO  K P B N  ED_Q1 Which of the following best describes the highest level of education you have received: 1. less than high school completion 2. high school graduate 3. some college or university (including trade school) 4. college or university graduate  What is your postal code: How many people live in your household as of today? How many, children live in the your household?  229    Appendix II-7 : Measurement sheet used in the Nutrition Clinics  MEASUREMENT SHEET Boy q  1 . Your child is a :  Girl q  2. What is your child's birth date? Year/ Month/ day  /  /  3. What was your child's birth weight? kg  or   lbs  4 . Was anyone of the followings true about your pregnancy? He/she was born full term q  He/she was born prematurely q You gave birth to more than one child q    weeks of age  How many?  Babies  You were diagnosed with gestational diabetes q You took vitamin and nutrient supplements q You smoked  yes q  no q  rarely q  5. Anthropometric measurements:  Wt :  kg  Mom's height (self-reported) :  Ht :  cm  Dad's height (self-reported):  7. Does your child have any health condition or sickness that you are aware of?  q Yes  q No If yes, please specify: 8. Is your child taking any medication at present? q No  q Yes  230    Appendix.II-8: Food Frequency questionnaire used in the study of low income preschoolers  FOOD FREQUENCY QUESTIONNAIRE  NUTRITION  RESEARCH PROGRAM  Subject Code:  zv042 _  Location of Visit : Interviewer:  231    Item Name  Brand/Homemade  Amount  Frequency  DAIRY PRODUCTS Please tell me how much cow's milk your child drinks and what type of milk  1)  Milk (drinking), cow's  	circle  Homo, 2%, 1%, skim  Chocolate milk  q cup/ml  q Day  Hot chocolate  q cup/ml  0 Day  2)  Goat's milk  q cup/mI  q Day  3)  Soy milks/rice milks Please tell me how much cheese your child ea in meals or sandwiches, do not include chees pasta or on pizza  q cup/ml  q Day  q oz/g q cup q slice  q Day q Week  4) Cheese, hard cheeses cheddar, mozzarella, Swiss . Brie 5)  Cream cheese  q oz/g q tsp q Tsp  q Day q Week q Every 2 weeks  6)  Cottage cheese (1%, 2%, creamed or whole)  q oz/g q cup q piece  q Day q Week q Every 2 weeks  7)  Processed cheese slices  q Day q Brand  q Slices  q Week q Every 2 weeks  q oz/g  q Day  10) Feta cheese  q oz/g q Piece  q Day q Week q Every 2 weeks  1) Yogurt (see sample)  q oz/g q cup/ml  q Day q Week q Every 2 weeks  8)  Cheese Spreads (eg. Cheez Whiz, Country Crock  13)  Minigo (Yoplait fresh cheese product)  q oz/g q cup/ml  q Day q Week q Every 2 weeks  14)  Ice cream/ Frozen yogurt  q cup/ml  q Day q Week 232    Item Name  Brand/Homemade  Amount  Frequency q Every 2 weeks  15) Milkshakes/ Yogurt shakes  16) Sour cream, whipping cream    q Brand  q Cup/ml  q Day q Week q Every 2 weeks   q Cup/ml  q Day  q Homemade  17) Whipped Topping  q Week q Every 2 weeks   -cool Whip, Nutri Whip   q Oz/g   q Day  q tsp   q Week  -Whipping cream  q Tbsp  Other Dairy Productseggnog, Caresses, fresh cheese, cream substitutes  q oz/g q Cup/ml q Piece  q Every 2 weeks q Day q Week q Every 2 weeks   EGGS : Please tell me how many eggs your child eats, boiled, scrambled, fried, or in omelettes . Do not include eggs in baked dishes  1)  Eggs  2)  Egg yolk only  3)  q number  q Day  q number  q Day q Week q Every 2 weeks  q number  q Day q Week q Every 2 weeks  q oz/g  q Day  q tsp q Tbsp  q Week  q Brand  q Oz/g q tsp q Tbsp  q Day q Week q Every 2 weeks  q Brand q Homemade  q tsp q Tbsp  q Day q Week  Egg whites only  TABLE/COOKING FAT  1)  Margarine  Brand : Soft q Hard q  for spreading on breads, crackers or vegetables 3)  Butter for spreading on breads, crackers or vegetables  Please tell me about oils you use in cooking. 5)  Oils (example In cooking pancakes and frying foods  233    Item Name  Brand/Homemade  Amount  Frequency q Every 2 weeks  Please tell me about oils you use in salads : 6)  Salads, home made  Oil used :  7)  Salad dressing, purchased  OilNinegar q  10)  q Day q Week q Every 2 weeks  q tsp q Tbsp  q Day  Creamy q  Mayonnaise, Miracle Whip  8)  q tsp q Tbsp q Cup/ml  q Week q Every 2 weeks q tsp q Tbsp  q Day q Week q Every 2 weeks  q tsp  q Day  - tahini, nut butters  q Tbsp  q Week q Every 2 weeks  Other Table/Cooking Fats — cereal cream, sour cream  q tsp q Tbsp q Cup/ml  q Day q Week q Every 2 weeks  Peanut butter  Low fat q Regular q  Specify  BREADS/CEREALs : Please tell us about the type of bread your child eats 1)  Bread  Specify  (Including pita/bagels) corn, rye, sourdough, chapatti, roti  q Piece  q Day  q White q Whole wheat  q Week  q Multigrain  2)  q Rye q Sourdough q Corn  Buns/Rolls (Including hamburger/hot dog buns) include only buns for hamburgers and hotdogs made at home 3)  Tortillas  _ q Piece  q Day q Week  Flour  q 6"  q Day  q White  q 8"  q Week  q 12"  q every 2 Weeks  DO Not include tortillas in purchased fast foods  q Whole wheat  Tacos  q Corn q Soft Wheat (?) 234    Item Name  Brand/Homemade  Pita  Amount  Frequency  q White q Whole wheat  Bagels  q White q Whole wheat q Multigrain  q Day  Breadsticks/Croutons  q Piece  q Week  q Cup  q Day q Week  _ q Cup  q Day q Week  q Homemade 4)  Cereals, cold breakfast  e.g. corn flakes, rice krispies, corn pops 5)  Cereals, cooked  6)  Wheat germ  q Oats (porridge) q Cream of wheat q?  (Used in/on foods)  7)  q Brand q Homemade  Bran, on foods  q tsp  q Day  q Tsp  q Week q Every 2 weeks  _ q tsp  q Day  q Tbsp  q Week q Every 2 weeks   Milk added to cereals  q 1 cup milk  Specify  q 1/2 cup milk For each cup of cereal  q 1/4 cup milk   Sugar added to cereal  qNo milk  Specify  tsp sugar   For each cup of cereal .  Tbsp sugar q No sugar  Baked Goods Breakfast goods ; Pancakes/Waffles .	 Specify  Surup q No syrup  Tbsp tsp  q 2"  q Day  q 4"  q Week  q 5"  q Every 2 weeks  q 6"  Muffins, bran  Fruit muffins, blueberry, carrots,  q Small q regular q Large _ (Deli size) q Small  q Day q Week q Every 2 weeks q Day 235    Item Name  Brand/Homemade  Frequency  Amount  q Week q Every 2 weeks  O regular q Large (Deli size)  10)  O Day q Week q Every 2 weeks  q Piece  English Muffin  _ q piece  Scones  0 Day q Week q Every 2 weeks  q Brand q Homemade 11)  Tea biscuits  q Brand  q Piece  q Day q Week  q Homemade  q Every 2 weeks Pop tarts/ Toaster Pastries  q Brand  q Day q Week q Every 2 weeks  q Piece  12)  Donuts, Fritters  _ q Piece  CI Day q Week q Every 2 weeks  13)  Danish, Pastries  _ q Piece  q Day q Week q Every 2 weeks  14)  Croissants  15)  Cakes, white  q With icing q Without icing  16)  Cakes, chocolate  q Oz/g q Cup/ml q Piece  q Day q Week q Every 2 weeks  q 2" square  q Day  _ q 1/8 cake  q Week q Every 2 weeks  0 With icing  q Day  q Without icing  q Week q Every 2 weeks  Fruit tarts  q Day  O Slice (1/8")  q Week q Every 2 weeks  Custard tarts  q piece  q Day q Week 236    Item Name  Brand/Homemade  Amount  Frequency q Every 2 weeks  17)  Cookies  Specify  q Plain  q q q q 18)  Cinnamon buns  19)  Wagon wheels  20)  Instant noodle, pkg dry  21)  q piece  q Day  q Week q Every 2 weeks  Oat meal Chocolate Fruit-filled Cream-filled _ q 1/4 Piece q % piece  q Day q Week q Every 2 weeks  q 1 /4 piece q 1/2 piece  q Day q Week q Every 2 weeks  q Brand  q 1/4 pkg q 1 /2 pkg  q Day q Week q Every 2 weeks  q Brand q Homemade  q cup/ml q piece  Cooked noodle Please tell me about the cooked noodles that your child eats ; please include all the noodles that you use to make dishes Except for macaroni and cheese . Include spaghetti and lasagna and cannelloni (tell me about the sauces later)  q Day q Week q Every 2 weeks  22)  Spaghetti, boiled  q Cup  q Day q Week q Every 2 weeks  23)  Other noodles ; fettuccini, rigatoni  q Cup  q Day q Week q Every 2 weeks  24)  Rice noodles (Boiled)  25) Macaroni and cheese, Home-made and packaged (e .g . KD)  26)  Canned noodles, with tomato sauce e .g . alphagetti, with tomato and meat e .g . ravioli  q Cup  q q q q q  q Day q Week q Month q Day q Week q Every 2 weeks  Cup 1/4 pkg 1/2pkg 3/4pkg 1 pkg  q Cup  q Day q Week q Every 2 weeks 237    Item Name  Brand/Homemade  Amount  Frequency  Lasagna/cannelloni,	 with vegetables q Day  Lasagna/cannelloni  With meat  q Cup  q Week q Every 2 weeks  q Cup  q Day q Week q Every 2 weeks  Cooked rice  q White  q cup  q Day  q Brown q Wild  q Week q Every 2 weeks q Cup  Other grain products, couscous  q Day q Week q Every 2 weeks  q oz/g q piece  q Day q Week q Every 2 weeks  6) Lamb (including roast, chops, etc)  q oz/g q piece  q Day q Week q Every 2 weeks  7) Chicken, turkey or other poultry  q oz/g q piece  q Day q Week q Every 2 weeks  q oz/g q piece  q Day q Week q Every 2 weeks  q oz/g q piece  q Day q Week q Every 2 weeks  q oz/g q cup q piece  q Day q Week q Every 2 weeks  q oz/g q piece  q Day q Week q Every 2 weeks  8) Chicken nuggets/ strips  9) Wild game ( moose, deer, etc-fresh, frozen, dried) 10)  11)  Canned fish and Shellfish, example . Tuna, salmon,  Fresh and frozen fish and shell fish  q Brand q Homemade  f  sushi . . .  q Brand q Homemade  q Brand  238    Item Name 12)  13) b  14)  Brand/Homemade  Other seafood - scallop, clams - Lobsters, mussels, oysters Deli meats; bologna, salami, pepperoni, shell fish  Wieners, hot dogs, sausages  Amount  Frequency  q oz/g q piece  q Day q Week q Every 2 weeks  q oz/g  q Day  q slices  q Week q Every 2 weeks  q oz/g q number  q Day q Week q Every 2 weeks  MEATALTERNATI VES  1)  Firm or medium firm tofu or soybean curd  2)  3)  4)  q oz/g  q Day  q cup q piece  q Week q Every 2 weeks  q Brand q Homemade  q oz/g q cup q piece  q Day q Week q Every 2 weeks  q Brand q Homemade  q oz/g q cup q number  q Day q Week q Every 2 weeks  q Brand  q oz/g q cup q piece  q Day q Week q Every 2 weeks  q Brand  Soft or desert tofu  Soy wiener/ vegetarian wiener  Other meat replacements  COMBINATION DISHES CASSEROLS WITH MEAT, FISH AND POULTRY 1)  2)  3)  4)  5)  Mixed dishes prepared, with beef (e .g . shepherd's pie, pot pie, chili, stew) Mixed dishes made with fish (e .g ., tuna casserole)  Mixed dishes made with pork  Mixed dishes made with lamb  Pizza with meat  q Brand q Homemade  q Brand q Homemade  q Brand q Homemade  q Brand q Homemade  q oz/g q cup q Tbsp  q Day q Week q Every 2 weeks  q oz/g q cup  q Day q Week q Every 2 weeks  q oz/g q cup  q Day q Week q Every 2 weeks  q oz/g q cup  q Day q Week q Every 2 weeks  q slice  q Day 239    Item Name  Brand/Homemade  Amount  Frequency  q Brand q Homemade 6)  7)  Enchiladas and Taco with meat  q Week q Every 2 weeks q piece  q Day q Week q Every 2 weeks  q piece  q Day  q Brand q Homemade  Filled buns, baked or steamed, meat filled  8)  Luncheon Meats/ Spreads in sandwiches/subs  q Week q Every 2 weeks q Brand q Homemade  q oz/g q cup q piece  q Day q Week q Every 2 weeks  q piece  q Day q Week q Every 2 weeks  q piece  q Day q Week q Every 2 weeks  q piece q slice  q Day q Week q Every 2 weeks  q oz/g q cup q piece  q Day q Week q Every 2 weeks  q ozlg q cup q piece  q Day q Week q Every 2 weeks  COMBINATION DISHES WITH CHEESE 1)  2)  3)  4)  Enchiladas, cheese filled  Perogies (potato and cheese filled, or onion filled)  Pizza and pizza pockets with cheese and no meat  Quiche without meat (with cheese)  q Brand q Homemade  q Brand q Homemade  q Brand q Homemade  q Brand q Homemade  COMBINATION DISHES WITH VEGETABLES  1)  2)  cooked lentils, beans, or peas (e .g . lentil stew or soup)  q oz/g  q Day  q Brand q Homemade  q cup q piece  q Week q Every 2 weeks  q oz/g  q Day  q Brand q Homemade  q cup q piece  q Week q Every 2 weeks  Vegetarian pasta dishes  3)  q Brand q Homemade  Other mixed dishes  240    Item Name  Brand/Homemade  Amount  Frequency  SOUPS 1)  Broth Type e.g.  Vegetable  2)  3)  Broth type ; chicken, Beef and fish  q cup/ml  q Day q Week q Every 2 weeks  q cup/ml  El Day q Week q Every 2 weeks  q cup/ml  q Day q Week q Every 2 weeks  q cup/ml  q Day q Week q Every 2 weeks  q oz/g q cup/ml q piece  q Day q Week q Every 2 weeks  q Brand q Homemade  q Brand q Homemade  Cream-type soups q Brand q Homemade  4)  Noodle soups q Brand q Homemade  5)  Other types of soup q Brand q Homemade  VEGETABLE ; CANNED, FRESH OR FROZEN 1)  Broccoli  q oz/g q cup q piece  q Day q Week q Every 2 weeks  2)  Carrots  q oz/g q cup q piece  q Day q Week q Every 2 weeks  3)  Corn ; cream or niblets  q oz/g q cup q piece  q Day q Week q Every 2 weeks  4)  Green peas  q oz/g q cup  q Day q Week q Every 2 weeks  5)  Spinach, cooked  q oz/g q cup  q Day q Week q Every 2 weeks  6)  Green beans, string beans, yellow beans  q oz/g q cup q piece  q Day q Week q Every 2 weeks  7)  Potatoes ; mashed, baked, salad or boiled  q oz/g q cup  q Day q Week 241    Item Name  Brand/Homemade  Amount  Frequency  q piece  q Every 2 weeks  8)  French fries, home fries, Pan fries  9)  Squash, all types   q oz/g q cup q piece   q Day q Week q Every 2 weeks  10)  Cabbage   q oz/g q cup   q Day q Week q Every 2 weeks  11)  Brussel ; sprouts   q oz/g q cup q piece   q Day q Week q Every 2 weeks  12)  Celery   q oz/g q cup q sticks   q Day q Week q Every 2 weeks  13)  Chick peas   q oz/g q cup   q Day q Week q Every 2 weeks  14)  Lentils/split peas   q oz/g q cup   q Day q Week q Every 2 weeks  15)  Kidney beans  16)  tomato  17)  Lettuce  18)  Cucumber   q cup q slices   q Day q Week q Every 2 weeks  19)  Peppers   q cup q piece   q Day q Week q Every 2 weeks  20)  Other vegetables  q oz/g q cup q piece  q oz/g q cup   q Day q Week q Every 2 weeks    q cup q piece   q Day q Week q Every 2 weeks    q cup q leaves      q Day q Week q Every 2 weeks   q cup q piece  q Day q Week q Every 2 weeks   q Day q Week  242    Item Name  Brand/Homemade  Amount  Frequency q Every 2 weeks  FRUIT (CANNED, FRESH OR FROZEN) 1) Apples and apple  sauces  q cup q piece  2)  Bananas  q piece  q Day q Week q Every 2 weeks  3)  Oranges  q piece  q Day q Week q Every 2 weeks  4)  Pears, peaches, nectarines and plums  q piece  q Day    q Day q Week q Every 2 weeks   q Week q Every 2 weeks 5) Grapes  6)  7)  Raisins, prunes and other dried fruits  q cup q piece  Melon (eg . Cantaloupe, honeydew, watermelon)  8)  Lychee  9)  Strawberries  10)  Other berries      (blueberries, raspberries) 11)  q oz/g q cup q piece  q Brand  Fruit cocktail or fresh fruit salad  q Day q Week q Every 2 weeks   q Day q Week q Every 2 weeks   q oz/g  q Day  q piece  q Week q Every 2 weeks  q piece  q Day q Week q Every 2 weeks  q cup q piece  q Day q Week q Every 2 weeks  q cup q piece    q Day q Week q Every 2 weeks   q cup  q Day  q Brand q Homemade 12) Other fruits  q Week q Every 2 weeks q oz/g  q Day  243    Item Name  Brand/Homemade  Amount  Frequency  q cup q piece  q Week q Every 2 weeks  q cup/ml  q Day q Week q Every 2 weeks  BEVERAGES 1)  Pure orange Juice and  grapefruit juice 2) Apple juice  q cup/ml  3) Five alive  q cup/ml   q Day q Week q Every 2 weeks  4)  q cup/ml   q Day q Week q Every 2 weeks  Other fruit juices (eg . Grape, pear, pineapple, papaya, cranberry)  5) Prune juice  q Day q Week q Every 2 weeks   q cup/ml   q Day q Week q Every 2 weeks  q cup/ml   q Day q Week q Every 2 weeks  q Brand q Homemade 6) Tomato and mixed Vegetable juices (eg . V8 juice)  q Brand q Homemade  7) Carrot juice  q cup/ml   0 Day q Week q Every 2 weeks  q Brand q Homemade 8)  9)  Sweetened fruit drinks including crystals and boxed varieties (eg . Tang, Kool-Aid, Ribena)  q cup/ml  q Day q Week q Every 2 weeks  q Brand q Homemade  Pop (regular)  q cup/ml  q Day q Week q Every 2 weeks   q Brand q Homemade 10) Pop (diet)  q cup/ml   q Day q Week q Every 2 weeks  q oz/g q cup/ml q piece   q Day q Week q Every 2 weeks  q Brand q Homemade 11) Carbonated fruit drinks (eg . Koala Springs)  12) Tea  q Brand q Homemade  q cup/ml  q Day q Week   244    Item Name  Brand/Homemade  Amount  Frequency q Every 2 weeks  13)  Coffee  14)  Other beverages  q cup/ml  q Day q Week q Every 2 weeks  q cup/ml  q Day q Week q Every 2 weeks  q cup/ml  q Day q Week q Every 2 weeks  q cup/ml  q Day q Week q Every 2 weeks  q cup/ml q piece  q Day q Week q Every 2 weeks  q oz/g q cup/ml q piece  q Day q Week q Every 2 weeks  q Brand q Homemade  DESSERTS 1)  Custard q Brand q Homemade  2)  Pudding q Brand q Homemade  3)  Jello q Brand q Homemade  4)  Popsicle or Mr . Freezie q Brand q Homemade  SNACKS 1)  Plain or cheese crackers  (Ritz, cheese type and soda crackers) 2)  Wheat crackers  (Stone wheat, thins, Triscuits, wholegrain, soda crackers) 3)  4)  q piece  Potato chips, cheeses or tortilla chips  q Day q Week q Every 2 weeks  q Brand q Homemade q piece  q Day q Week El Every 2 weeks  q piece q pkg  q Day q Week q Every 2 weeks  q oz/g q cup  q Day q Week  q Brand q Homemade  q Brand q Homemade  Popcorn q Brand q Homemade  5)  Party snacks - Nuts & Bolts, pretzels - Crunch N Munch  q Brand q Homemade  q Every 2 weeks q cup  q Day  q piece  q Week q Every 2 weeks 245    Item Name  Brand/Homemade  Amount     Frequency  q cup q piece  q Day q Week q Every 2 weeks   q cup q piece   q Day q Week q Every 2 weeks   q Day q Week q Every 2 weeks  8)  Other nuts   q cup q piece  9)  Seeds (eg, sunflower seeds)   q cup q piece   q Day q Week q Every 2 weeks    q cup q piece   q Day q Week q Every 2 weeks   q Brand q Homemade  q oz/g q cup q piece  q Brand  q piece  q Brand q Homemade  q Tsp q tsp  10)  Other seeds  11)  Other snacks  q Day q Week q Every 2 weeks   JAMS, JELLIES,CANDIES  2) Jam and jellies on bread  3)  q Day q Week q Every 2 weeks   1) Candy  Chocolate bar  q Day q Week q Every 2 weeks  q piece  q Day q Week q Every 2 weeks  q oz/g q piece   q Brand  q Day q Week q Every 2 weeks  q oz/g q piece   q Brand  q Day q Week q Every 2 weeks  q Brand  q oz/g q piece  q Brand q Homemade 4)  5)  6)  Granola bar  Fruit roll up, fruit leather  Suckers lolly pops etc  q Day q Week q Every 2 weeks   CONDIMENTS  246    Item Name 1)  Brand/Homemade  Chatni q Brand q Homemade  3)  Other sauces q Brand q Homemade  4)  Frequency  q Tsp q tsp  q Day q Week q Every 2 weeks  Tomato ketchup q Brand q Homemade  2)  Amount  Other condiments q Brand q Homemade  q Tsp q tsp  q Day q Week q Every 2 weeks  q Tsp q tsp  q Day q Week q Every 2 weeks  q Tsp q tsp  CI Day q Week q Every 2 weeks  PURCHASED INFANT, JUNIOR AND TODDLER FOODS  1)  Cereal (eg . Rice, barley , oats or mixed)  2)  Cereals mixed with yogurt or/and fruits  3)  4)  5)  6)  7)  8)  Meat or poultry (eg, beef, pork, lamb, veal, ham, chicken or turkey)  q Tsp q tsp q jar  q Day q Week q Every 2 weeks  q Brand q Homemade  q Tsp q tsp q jar  q Day q Week q Every 2 weeks  q Brand q Homemade  q Tsp q tsp q jar  q Day q Week q Every 2 weeks  q Brand q Homemade  q Tsp q tsp q jar  q Day q Week q Every 2 weeks  q Brand q Homemade  q Tsp q tsp q jar  q Day q Week q Every 2 weeks  q Brand q Homemade  q Tsp q tsp q jar  q Day q Week q Every 2 weeks  q Brand q Homemade  q Tsp q tsp q jar  q Day q Week q Every 2 weeks  q Brand q Homemade  q Tsp q tsp q jar  q Day q Week q Every 2 weeks  q Brand  Liver  Meat or poultry with rice or noodle dinner  Vegetable and meat  Vegetables  Fruits  247    Item Name 9)  10)  Brand/Homemade  Prunes  12)  13)  q Tsp q tsp q jar  q Day q Week q Every 2 weeks  q Brand q Homemade  q Tsp q tsp q jar  q Day q Week q Every 2 weeks  q Brand q Homemade  q Tsp q tsp q jar  q Day q Week q Every 2 weeks  Fruit yogurt desserts  Custard or pudding  Other purchased baby foods  Frequency  q Brand q Homemade Fruit desserts  (eg . Tutti Frutti)  11)  Amount  q Tsp  q Day  q Brand q Homemade  q tsp q jar  q Week q Every 2 weeks  q Brand q Homemade  q Tsp q tsp q jar  q Day q Week q Every 2 weeks  1) Did you breast-feed your child? q Yes  q No  If yes, please specify (exclusively breast-fed);  1-2 months q months q  4-6 months q  >6-10 months q  12-18  2) Are you currently breast-feeding? q Yes Number of times per day  q No  3) Are you currently giving your child a commercial infant formula? q Yes (please go to question 4) q No (please go to questioti 5) 4) What brand types of formula do you Color of label much does your child Usually give your child? usually drink per day?  ( 1 )    How many times per day How or per week    times per q day q week 248    (2)  times per q day q week  (3)  times per q day q week  5) Do you usually give your child a vitamin/mineral supplement? q Yes (please go to Q5) q No 6) What brands/Types of Supplements? q day q week q month q day q week q month q day q week q month 7) Does your child takes a bottle? At night q Yes q NO  during day time naps q Yes q NO  during the day q Yes q NO  8) What is usually in your child's bottle? At night day time  during  Juice  Water  Other  249  Appendix 11-9 . Comparison of the daily intake (number of serving/ day) of main food groups and their sub categories for inner city boys and girls (n=82)  Food group  Servings/day (Boys, n=43)  Servings/day (Girls, n=35)  Meat and alternatives  2 .73±b1 .4  3 .12 + 1 .9  Cereals and Grains  7 .19 ± 3 .5  6.37 ± 3 .7  Whole wheat C&G  1 .33 ± 1 .9  1 .31 ± 2 .1  Non-whole wheat C&G  5 .85 ± 2 .9  5 .06 ± 2 .6  Dairy  3.62 ±2.0  3.92 ± 2.4  Fruits and Vegetables  7 .40 ± 5 .6  6.65 ± 5 .0  250    Appendix III-l : Certificate of ethic approval for "Population patterns of hair zinc" The University of British Columbia  Office of Research Services IIIr~W/A  Clinical Research Ethics Board Room 210, Research Pavilion, 828 West 10 th Avenue, Vancouver, BC V5Z 1L8 PRINCIPAL INVESTIGATOR  DEPARTMENT  Hertzman, C.  Health Care/Epidemiology  INSTITUTION(S) WHERE RESEARCH WILL BE CARRIED OUT :  NUMBER  CO5-0463  UBC Campus  CO-INVESTIGATORS:  Barr, Susan, Family & Nutr Sci ; Chapman, Gwenneth, Land & Food Systems ; Vaghri, Ziba, Land & Food Systems; Wong, Hubert, Health Care/Epidemiology SPONSORING AGENCIES :  TITLE :  Unfunded Research  Population Pattern of Hair Zinc ; Dietary and Sociodemographic Determinants  Notice of Ethical Review: Memo The UBC CREB has reviewed your proposed study, and has issued a Certificate of Approval. Please make sure the following items have been corrected/amended within 21 days . Ensure that any revisions to the consent form are made on the correct institutional letterhead and are either underlined or in BOLD text . A letter that explains how the requested changes have been addressed should accompany your response. If needed for Clinical Trials, please ensure that the Consent Form Version Date is updated to reflect any required revisions and that the new version date is clearly indicated in your response . An updated Certificate of Approval will be issued accordingly. All other responses will be acknowledged by the CREB Chair .  You may submit the revisions by email to Ms . Farin Ramji at farin .ramji(aors.ubc.ca. Thank you for your attention. Please provide the CREB a copy of the information pamphlet for approval once it has been prepared.  If you have any questions regarding these requirements, please call: Dr . Gail Beliward, Chair, 604-875-4111 ext . 62276 Dr . James McCormack, Associate Chair, 604-822-1710 Ms . Erin Skrapek, Manager, Clinical Ethical Reviews (ORS), 604-875-4149 For further information, refer to the UBC CREB's Guidance Notes and Policies at: www .ors .ubc .ca/ethics/human subjects .htm. Please send all correspondence to : Clinical Research Ethics Office, Room 210, Research  Pavilion, 828 West 10 th Avenue, Vancouver, BC Canada V5Z 1L  251    Appendix III-2 : Neighborhood Maps produced by the Geographic Information System, GIS, unit of the Human Early Learning Partnership, HELP, to facilitate planning the recruitment strategy  Average Mel Income (2000) by aenaue dl nrinatlon area. 1Lgnsln~ rs vl  RN"  0 Neighbo mood Bcunduiee  H  total Warne 2000 020065 -i1 4212-26686 72 -34070 64066 -,tmi  map uee forbidden for publlcetlon or dleseri*wtbn  251  Appendix III-3 : Letter of Support from the Executive Director of the West Coast Child Care Resource Center  3 rd  Floor, 210 W . Broadway Vancouver, BC V5Y 3W2  Ph : 604 .709 .5661 Fax : 604 .709 .5662 Toll-free : 1 .8 77 .262 .0022 Website : www .wstcoast .org  January 25, 2006 Dear Child Care Colleague: Westcoast Child Care Resource Centre is pleased to lend its support to an important research study regarding the dietary status of preschool-aged children in Vancouver . This research study is being conducted by Ziba Vaghri, a UBC Doctoral student under the direction of Dr . Clyde Hertzman, the Director of Human Early Learning Partnership (HELP). This study involves a city-wide survey of preschoolers to look into some of the dietary, health and socio-demographic factors related to zinc deficiency in young children . Zinc deficiency is quite common and can adversely affect school performance and a wide range of subsequent life events such as educational achievement, employment, income, and so on . . Previous work by UBC researchers has indicated that there are some factors that may place a child at a higher risk for this deficiency. The survey involves a questionnaire that will go to parents . Following completion of the questionnaire and with parental consent, researchers will come into your child care centre to take height and weight measurements of children as well as a small hair sample for the purpose of hair zinc analysis . This analysis can be an index of child's zinc status. You will be contacted in the near future for further discussion of this project, and how you can help parents become involved . Please feel free to direct any questions you might have regarding the study to the UBC researchers when they call or visit .  253  Westcoast Child Care Resource Centre asks you to consider participation in this research project . Such research is key to our understanding of the determinants of young children's health.  Sincerely'  Dianne Liscumb Executive Director Westcoast Child Care Resource Centre  254  Appendix III-4 : Letter from the principal investigator of the project  THE UNIVERSITY OF BRITISH COLUMBIA  Department of Health Care and Epidemi Faculty of Medicine Mather Building, 5804 Fairview Avenue Vancouver, B .C . V6T 1Z3 Tel : 604-822-2772 Fax : 604-822-4994 Website : www .healthcare .ubc.ca  Dear child educators; We are a group of researchers from the University of British Columbia planning to conduct a survey to investigate the "Population patterns of hair zinc in preschoolers" in the city of Vancouver . Your agency is key to helping us connect with parents of children 2-6 years old who may choose to participate in this study . Qualified professionals will treat participants and the information they provide with great care and confidentiality. Zinc is an essential micronutrient available from many sources including meat, dairy, cereal and grain products . Inadequate zinc intake can result in zinc deficiency . It has been shown that even minor zinc deficiency, termed marginal zinc deficiency (MZD), can interfere with children's normal growth and development . MZD can be easily prevented if adequate amounts of zinc-rich foods are included in a child's diet . When properly diagnosed, the correction of MZD is, also, relatively simple . Since the hair zinc content in children is considered an index of the body's zinc content, it can be used in detecting MZD . Children with MZD often have lower hair zinc content . The main purpose of this study is to investigate the hair zinc status of a representative group of Vancouver preschoolers . The second purpose of this survey is to examine and understand the dietary and environmental factors that could predispose children to low hair zinc . The information collected through this survey will increase our awareness on the at-risk groups of children in our communities. In few weeks, you will be provided with an "announcement letter" addressed to the parents of these children . This letter, which is to be sent home with every child, will notify parents about the upcoming survey package . Subsequently, two weeks after this 255  letter, survey packages will be supplied to your centers to be handed to the parents of children when they drop off or pick up their children. Few weeks after distributing the survey packages, we will need to come to your center (at a time that is convenient for your center) . During this visit we will collect the hair samples and measure the heights and weights of children whose parents have consented to this . We foresee that the survey will take 4-6 weeks. The information collected through this survey will better our understanding of zinc deficiency, one of the common and silent deficiencies of this age group . The success of this survey is largely dependent on the number of people participating in it. With this letter I am asking for your support in this important research . You can extend your support to this research by encouraging the parents to participate. Yours sincerely, Clyde Hertzman MD, M.Sc, FRCPC Director of Human Early Learning Partnership University of British Columbia  256  Appendix III-5 : Initial letter of contact for survey participants  THE UNIVERSITY OF BRITISH COLUM BIA Department of Health Care and Epidemi Faculty of Medicine Mather Building, 5804 Fairview Avenue Vancouver, B .C . V6T 1Z3 Tel : 604-822-2772 Fax : 604-822-4994 Website : www.healthcare .ubc .ca  Dear parents; In two weeks you will be receiving a survey package through your childcare center, which will invite you to participate in a survey to be undertaken by a group of researchers from the University of British Columbia . We are inviting you to participate in this survey .You have been selected because you have a child who is two years or older but not yet in school. This survey, entitled "Population Pattern of Low Hair Zinc in Preschoolers", is a part of the doctoral thesis ofa UBC student, Ziba Vaghri.It will be conducted through the University of British Columbia under the supervision ofmyself, Dr. Clyde Hertzman, a UBC professorand director of the Human Early Learning Partnership, HELP. The survey's purpose is to increase our understanding of zinc deficiency in preschool children both because it is a very common deficiency in childhood and also, one that to date remains largely unrecognized and is not easily detected. Your participation will involve answering a short questionnaire . It will take 10-15 minutes. You will also need to read and sign a consent form to give us permission to measure your child's weight and height and to take a sample of his/her hair . We ask that 257  the completed questionnaire and consent form be returned to your center at your early convenience. Taking a hair sample (taken at the center) is a simple procedure that will be explained to your child and will cause him/her no pain. By participating in this survey you will be doing your part to expand the present state of knowledge of zinc deficiency among preschool children . Your child's hair will be analyzed for its zinc content and the result of this analysis along with the information provided in the questionnaire will be used to better understand some possible patterns associated with this deficiency . This information will be treated with	 confidentiality. Sincerely, Clyde Hertzman MD, M .Sc, FRCPC Director of Human Early Learning Partnership University of British Columbia  258    Appendix III-6 : Recruitment form for the focus group participants  THE UNIVERSITY OF BRITISH COLUMBIA Food, Nutrition and health Faculty of Land and Food Systems 2205 East Mall Vancouver, BC, Canada V6T 1Z4 Phone : 604-822-2502 Fax : 604-822-5143  UBC  Dear parent/guardian We are planning a citywide survey of preschooler's health and nutrition . Before we initiate the survey we need to test our survey tool/questionnaire. Would you be willing to participate in a small group discussion that will help us to pretest our survey tool? In these groups you, along with 3-4 other mothers, will be asked to complete a questionnaire and take part in a subsequent group discussion . After completion of this session you will receive a $15 grocery certificate . The researchers in this study will facilitate the discussion groups. Would you like to participate in these focus groups? If you answered YES, please provide your name and phone number Phone Number  Name Ethnicity    age of the child  And email this form to; zibav@interchange .ubc .ca For your convenience he groups will be set up close to your daycare center. Do you need child minding? Yes q  No q 259  Appendix. III-7: Information sheet for recruited focus group participants  THE UNIVERSITY OF BRITISH COLUMBIA  Department of Health Care and Epidemiology Faculty of Medicine Mather Building, 5804 Fairview Avenue Vancouver, B .C . V6T 1Z3 Tel : 604-822-2772 Fax : 604-822-4994 Website : www .healthcare .ubc .ca  Dear focus group participant, You are scheduled to attend a focus group on January the  from  Am/Pm to  Am/Pm.  The focus groups will be held at:  If for any reason you are unable to attend please call Ziba (a, 604-812-1419to reschedule. We appreciate your support and look forward to seeing you again. Sincerely,  Ziba Vaghri Ph.D . Candidate University of British Columbia  260  Appendix III-8 : Informed Consent form for focus group participants  THE UNIVERSITY OF BRITISH COLUMBIA Department of Health Care and Epidemiology Faculty of Medicine Mather Building, 5804 Fairview Avenue Vancouver, B .C . V6T IZ3 Tel : 604-822-2772 Fax : 604-822-4994 Website : www .healthcare .ubc .ca  Informed consent:  I have accepted to participate in a focus group for the study of "population patterns of hair zinc" .Z..iba Vaghri, the student conducting the study, has explained to me, both, the purpose of this focus group and my role in it . I have read and understood the subject information and consent form. • • • • • • •  I have had sufficient time to consider the information provided and to ask for advice if necessary. I have had the opportunity to ask questions and have had satisfactory responses to my questions. I understand that all of the information collected will be kept confidential and that the result will only be used for scientific objectives. I understand that my participation in this focus group is voluntary and that I am completely free to refuse to participate or to withdraw from this study at any time I understand that I am not waiving any of my legal rights as a result of signing this consent form. I have read this form and I freely consent to participate in this focus group. I have been told that I will receive a dated and signed copy of this form.  Printed name of subject's Date legally acceptable representative  Printed name of principal investigator, designated representative Date  Signature  Signature  261    Appendix III-9 : Survey questionnaire THE UNIVERSITY OF BRITISH COLUMBIA Department of Health Care and Epidemiology, Faculty of Medicine Mather Building, 5804 Fairview Avenue Vancouver, B .C . V6T 1Z3  Tel: 604-822-2772 Fax : 604-822-4994 Website : www.healthcare .ubc .ca Date 2006/ /  I .CHILD'SBACKGROUND Gender : q Boy Date of Birth :  q Girl /  /  Day/ month / year Place of birth (country) : If not born in Canada, how long the child has been in Canada? years  and  months  1) Was your child fed with breast-milk? q No If yes, from the time she/he was  qYes  months to the time she/he was (Please write a number)  months old  (Please write a number) 2) Was your child fed with infant formula? q No If yes, from the time she/he was  qYes  months to the time she/he was (Please write a number)  months old  (Please write a number) 3) When did you introduce food to your child? When she/he was  months old. (Please write a number)  4) Does your child take or has taken a vitamin/mineral supplement? qNo q Yes If yes, from the time she/he was months to the time she/he was months old 262    (Please write a number) (Please write a number) 5) Please write the name of supplement(s) your child takes or took in the past (the brand name) 6) How would you describe your child's health? q Very good q Good q Fair  q Poor  7) How often does your child get sick with sick flu, fever, cold, diarrhea, cough etc? q Never q  II. CHILD'S EATING HABITS q Yes If "yes", why are you concerned? Please check all your concerns.  q No  1) Are you concerned about your child's eating? qPicky eater  q Does not eat enough  q Eats unhealthy foods  qEats too much  qAlways eats the same foods q Other (specify)  2) In total, how many glasses of liquid NOT COUNTING WATER, does your child drink in a day? (Milk, juices, soft drinks, BREAST MILK, and other drinks) qNone q One q two q three q four q five q more than five 3) Does your child eat the following food? Meat (beef, lamb, pork, veal, moose meat and . . .) q Yes q No Fish and shellfish (Includes crab, shrimps, lobsters, clams, scallops . . .) q Yes q No Poultry (Includes chicken, duck, goose, turkey, etc) q Yes How often? Meat  times  Fish and shellfish Poultry times  times  q No  qper day  OR  q per week OR qper month  [per day qper day  OR OR  El per week OR qper month q per week OR Oper month  4) How many times does your child eat the following? White bread  times  per  qday  q week  qmonth 263    per  q day  0 week  times per times per Brown and wild rice  q day q day  qweek qweek  Brown bread  times  White rice  White pasta (noodles, rice noodles . . .)  q month q month qmonth  times per [day per q day  Brown or whole-wheat pasta times Breakfast cereal (non-whole grain) times Breakfast cereal (whole grain) times Other grain products (like couscous, tortillas. . .) Other whole grain products) times  qweek q month q week q month  per q day q week qmonth per q day q week q month times per q day q week qmonth per qday qweek qmonth  5) How many times does your child eat/drink the following dairy products? Milk (NOT COUNTING SOY MILK) times qper day OR q per week Cheese times Yogurt and yogurt drinks (fruit or plain) 6. My child eats lunch ;  q per day times  q at home  OR Oper day  q per week OR q per week  qwith care giver  7 .If she/he does not eat at home, do you pack her/him a lunch? [Yes q No, the caregiver provides the lunch  III . FAMILY BACKGROUND 1)n What is the ethnic Background (example Chinese, First Nations, African American and... .) of child's biological parents? Mother Father 2) What is the height of child's biological parents? q feet/in Mother  qCentimeters qCentimeters  Father  3) What is the total number of your family members living together? Please indicate the relationship of each adult to this child (example ; one grandmother, one stepfather . . .).  q feet/in  Persons  4) What is the highest level of education completed? Child's father: Child's mother : q Grade qGrade 264    q High school diploma q College q University  q High school diploma q College q University  3) What is the total family income? (BEFORE TAX DEDUCTION) q Less than $10,000/year q $ 50,000-60,000/year q $ 10, 000-20,000/year 0 $ 60,000-70,000/year q $ 20,000-30,000/year q $ 30,000-40,000/year q $ 40,000-50,000/year  q More than 70,000/year q Do not wish to respond q On income assistance  4) In the past 12 months: 4.1 .Were you and others worried that food would run out before you got money to buy more? qOften true q sometimes true q never true q don't know 4.2.The food that you and others bought just didn't last, and there wasn't any money to get more? qOften true q sometimes true q never true q don't know 4.3 .You and others, because of not having money, could not eat balanced meals? qOften true q sometimes true q never true q don't know 5) Please record your postal code : V IV. CHILD BEHAVIOR QUESTIONNAIRE Instructions : Please read carefully before starting: We would like you to tell us what your child's reaction is likely to be in those situations. There are of course no "correct" ways of reacting ; children differ widely in their reactions, and it is these differences we are trying to learn about . Please be sure to circle a number or NA (not applied) for every item. My child: 1 . Seems always in a big hurry to get from one place to another. 1 2 3 4 5 6 7 Extremely Quite Slightly Neither Slightly Quite Extremely untrue untrue untrue true true true true nor false  NA  265    2. When practicing an activity, has a hard time keeping her/his mind on it 1 2 3 4 5 6 7 Extremely Quite Slightly Neither Slightly Quite Extremely untrue untrue untrue true true true true nor false 3. Tends to run rather than walk from room to room. 1 2 3 4 5 Extremely Quite Slightly Neither Slightly untrue untrue untrue true true nor false  6 Quite true  7 Extremely true  4. Will move from one task to another without completing any of them 1 2 3 4 5 6 7 Extremely Quite Slightly Neither Slightly Quite Extremely untrue untrue untrue true true true true nor false 5. When outside, often sits quietly. 1 2 3 4 Extremely Quite Slightly Neither untrue untrue untrue true nor false  5 Slightly true  6 Quite true  7 Extremely true  NA  NA  NA  NA  6. When drawing or coloring in a book, shows strong concentration 1 Extremely untrue  2 Quite untrue  3 Slightly untrue  4 Neither true nor false  5 Slightly true  6 Quite true  7 Extremely true  7. Moves about actively (runs, climbs, jumps) when playing in the house 1 2 3 4 5 6 7 Extremely Quite Slightly Neither Slightly Quite Extremely untrue untrue untrue true true true true nor false  NA  NA  8. When building or putting something together, becomes very involved in what she/he is doing, and works for long periods. 1 2 3 4 5 6 7 NA Extremely Quite Slightly Neither Slightly Quite Extremely untrue untrue untrue true true true true nor false 9. Prefers quiet activities to active games . 266    1 Extremely untrue  2 Quite untrue  3 Slightly untrue  4 Neither true nor false  5 Slightly true  6 Quite true  7 Extremely true  NA  5 6 7 Slightly Quite Extremely true true true  NA  10. Is easily distracted when listening to a story. 1 Extremely untrue  2 Quite untrue  4 3 Slightly Neither true untrue nor false  11 .Is full of energy, even in the evening 1 2 3 4 Slightly Neither Extremely Quite untrue untrue untrue true nor false  5 6 Slightly Quite true true  7 Extremely true  NA  12. Sometimes becomes absorbed in a picture book and looks at it for a long time. 1 2 3 4 5 6 7 NA Extremely Quite Slightly Neither Slightly Quite Extremely true untrue untrue untrue true true true nor false 13. Likes to sit quietly and watch people do things. 2 3 4 5 1 Extremely Quite Slightly Neither Slightly untrue true true untrue untrue nor false  6 Quite true  7 Extremely true  NA  Many thanks for participating in our survey!  267    	o  Appendix III-10 : Survey questionnaire translated into Punjabi  1 11 L' V 1\ 1 V 1V 0  111V  L  AIL  1 J.  1 V as  V  S aA v ♦,i ai J. as  Department of Health Care and Epidemiology Faculty of Medicine Mather Building, 5804 Fairview Avenue Vancouver, B.C. V6T 1Z3 Tel: 604-822-2M Fax: 604-822-4994 Website: www.healthcare .ubc .ca  vltn. zoos/ / 1 .0 Nitro rit Nan a gip rt f xeir w a~sx Par s (e) : * 3a r f fl;?r trefl` ►, fed wir if fen!T A3r r *tow aet? 9)  d st  A►  )r  $ar farm, fr?  T?5t  z) t In" *rut* an vs # fall o o I+ ;a) al str z rtl * er r Fft? ~#tval¢aam fin* Witfeufi r *ter a t t fra' ml* :rPa e tf rr a »r Fit? wain a) zit TA' z mar gia fea r fft? q t Baca af, 31' (fbr 'NM')  a ~r+ ndtk 3ar  o 	o 311 t 1 u) ttl Tft ` rapr Nar rar  *at, a► , ta *' 11'4` wit fit far  wit* f  if  xthit as  uPazt s r-a kr re few? vft  me.  268    	a-ga  ') at yaw gar  o a-di'  Pen at trfna atraslfl`a It`s t r lean t a-4 q a+ ~aa 3+ xalft Ohm fall)  xatf'r * a$  t) Tart Of *as fax gat Eit tt q age' waft  q s s Vail  10) VIN* IfVT a9r, a'ar a q avt r® 2sdr ( Rm. fear few  (  *we al'  fa'a t  )f- t  am. viva vtr  rhrreff) rift ma flag  o aft a*a a,a  q fit  m)  rdtEraf +r (w fatl)  !War 4 wax arl  ) wi tear f fa q fa fawn agar rear* »niter Aa r Y qo q 9 o q a  t zfl'r t(  ` twat) Tv,  r+r,  q8  Cu  q 1t+  a) al yaw ar,aT tat f % tY 'xrfW) ( If* (bt), i4'tr, Ow, after,  o of o Ise xft fAa (1ta, fit, *ma, aciR~r, B o of o zsl` *me t (foarrr, tar, zaaft,) ftt* mitt q tea few 1W o fear a Pas o fear walk firs sftz am a+a o fear fees Rea q fear ad' 1 q fiat Kett fag xal »0 Wm s fear fiats fW o firm ate fora o firer )rfth* tie varz t rer q  269    9)  * ~`dr  t fi  tr f wr  tft Wert  t?  q few few few q flea NIT o few xdtt fra daft ire o flea fart few o fiat z e* fww o f we' Wilt fora *r _ o wefeet .  o few fens few Ira 1W o fee xat f a yTm3*oft errofleafew fro q fewzefww q fewx firs fleet v rwwr _ rat o flee ft.). fire o fee ma fire o 1W x T Rrw  .  wt www r u ra o few ft**rr f r o few ITO r p hear x~flt3 far rrse gr slfisras (sr~rs )aArw Inv we) i Ina a fear f* Mr a f d tW o flea xrtt*' E* a1Ar ( eIT q feat f fww o fleet aar3 f4fs q fear xm ww o flea fir 1W q fiat us* 1W a iv Ism tite Awl (pp, Z 1c,,t...) fear }At fww wtww z itw a few s' fwv o f>ew wet l'Ea q fvdn,rtfta 1)  gars* 4v *t  furs Inr  zW fe& era 'r tt  gw(i gv wt fare) dta z# urRa I` a'sfa'an(  z j W)  fie  a  zra* z}.  o few fW fie  w*a *la r ~`6►  pfd gr  a fleet few fire via o heat few fW  t ir  q few wwa o Hear wet flaw a fleet vat fire.  e mar ow wrftmrr *w trtw  (tea flsr?)  Owe flan?)  2ww wig r aft fl' threw a'e zsrrs *wt t? zrm. o a+ o it *tom wow r 9)  we  3) ufi '  z  x *w z  Jrrlfwww, arzR  . )nar flra* zr w*aft R  fvla[er  zf! 3 (fttx aa+ fa titre fellp:es',  ...) fzrar  z) i.* ? woof tw z star m3* tit zta►:tt fard t a teem }r*s► q teem fire, (Wier ft) off` ize (Z~sa fit7) q R` ttzw  270    9)vu''  afar tat bra u4f  fez's  o fear few fare o few wwa fW o &rat HA* fare o flea fear 1W o fear wwt fora o firer flee  ra  Fit t  wqft lifer t?  fgt. T)rr  ra o far firs f o fear a 1W o fear xt fW that $ r r ra o f i a t fen fro o t s a r eZ3 fWa o fear > 	fire fire r r3 r 	re o fear fear fwir o fear eat fire © flea fW ref Pt r tit v'frz* VW o flew fen for o few vat a few fee .srsrt r ( ads) ra a fear fts 1W o fear eat fm  o flea Hat far v Err Newts (Aiwa fn. fa r std sf  wra o few km fps o fear ma fW q  .444  t err,  few wilt fww toA'aa r Eltwt *If Tarr  via o feet fen fW o flea wet fie a fear n t  rr5  aft  ofear fesf~aofearaa'3f gaafee r  tat fs:flint  * f~'dt wrw leer t?  r  o fleet few fl o few a ter ra o fear wet fie ra a fleas few fie  f lea wet we' fare o hear fen %v . v.& bra vdr wfa'arrt( sir # ( *up  r 2sf fare)  v twos mar +Per+ iw vie 4) am* fWa rI }YaT ter astir aI leer t wt (afsa fell) t. Ore Asa) 9) q  W M`e ariff v ier t 71P ewer w'+a o Zsaft' teens arara writ aft et its.   twt t't  fi.rgary fir,  GO Ire # wax v* .w+ar flier er ifi-Jt Maw aft t fair waft fee v0'a m erg.. .) fv3r Nr3T  it if ear tw w art )rf3r fir3r o tfcc d xr3*  Ora %I) a #  za  *vet lift t4 a teem Am,  (a fist) o #zhflza  271      ytrw  f~ f~a : flawasatrar  imp;  vfavf eft Ire *W aI  mr#I' vrg'f} rft 1 7fr ? fa fell al aver gen rr wr fav vv+ » feu wa r itt fir tiwrf ra tv E* ett far Wlrrr rrdi` w trl  farm* vat tat frsfuarf arrrf E'* 1  s fail flea tar Pt a'$ 7srff  a vie* t.. -vrrar wf  $ Lev* fro ave.  q) arinr Ift tear w eft gar* * viz ~(t ami t far ''er t f  s arm  z. rvdtazrrr  u. Wr 74'cft  4.t eft  ?)  Et  Re` five  area  p. limps Alt  s. af writ Is* amr  aYzr we yet*  *attar arse: t st f ` fir zr rv; arms' #v# sit rifts  9. fame. aura  ~. ar~sRarar~  wit  a. *sr emir  fits wit  u. tar sit  ~. a  a) far mat 9. flop, arcs  a it fro gal t aka  u . * ur mft  4. and ;it  a. fm 4r s a s u. ~a* ;0  a,  2.  2.  ardtf33'  i. weft ;I t  a. f y r  fir al'rr rsrlr pit  #rr a? #er ft. a. iv aaa  9 . fear rr(ft  e. o f Prat 10 vow  ftru arm rr  a, of wit rsf arrsa  ;artt  e. ?S a. 'v* arrw t vfs ni t  V. mu  u) `a+aa Iry f3 of 9.  raw atM3  u. > 3i srft  tace tar Miter t. 9 . rst eft rst area 2 . Q'ahacs a . iv* sow fangs wit Ito. airs eft` yot 4. wt sit  ' Tar fve Ida rrf * arae► '~ of art* fan re ffe'er asaref Amur_  NIIKW  ?.  anaRarrra  a . *vr ma.  8.  of A$t a+ wpm  272    ,)  wa t »r'ea use* J arrr facer tower, ewer »r3 z re*) 9. fa ra amra ? . zmflaar3 a. a+ atl WI wag L i aria U. r= . raft Aft '. f At fn. ifs 151lf" gala 1 gs wirer TO fbaa* arm ft ar waft err #~T 3 b a~3 ~'d aa< aa* afa'e ► firs am u. tar 'rift 9.  ?. mama €. a' t rat p. ffagrr  a. tar aura a. at at at vale t feu tn. aft` we  t) vf:~ flsas and* *at c'*' x#a ribs anlMi t alit fWi.rvr Iat'e' aver t. t.Nagar am z. arr:?arra a. tar arras e. a+ Rdt at amts u i ar Ret are xJt P. flow raft few arrr mfr  at  aaraft Baer Ow fps f atr-tea ara ors f'+e* J.. .. Nags a= 3. walla a. *3T area a. at ;Tdt at area u.tar Rift 4. ar t vdt p. Nauss IT* feu aft Mt Tr* 90) '  TO aTt ft ' hr *. s*H an- it alit a+rer. 1far;rr aura ?. agar a, tar area a. a+ mft at aura flea ara riff Iwtft u. tar t 4. aptt mft farrwrr mPr arzlt a` firvara aura U . afar r* 4z) ar t  t sa €.  t flame $ it 7D7' scar at flag s a*a *ma r *.  z. aratarra a. *3T area 8 . at mft at ma aft raft P. flaw ft $a ar* arfit` 1  9a) vIrr as * *vat orate► d ad tar t ' th aa* 3azr Wear t. a. at Aft a+ ataar % f a arra z. radiator a. ta r area flea arrr yat u. tar ATft 4. art a* z . flargs sit Rita fir  fir' * nit sga agw trnkry t  273    Appendix III-11 : Survey questionnaire translated into Cantonese Department of Health Care and Epido Faculty of Medicine Mather Building, 5804 Fairview Aver Vancouver, B .C. V6T 1Z3 Tel : 604-822-2772 Fax: 604-822-4994 Website: www.healthcare.ubc.ca Date 2006/ /  1.  : q  1  qq  lilt  > t  *bJLWI  /  /  ))a#Jr~4a  (H /A /i)  t1~(~~)~a~R  ~i  ) :  , *  A  1 .E~3kft~i~i~ ? 2.1tts/itkOM .  la  sSMAMAIN ? ? qI qFI  3.'( ff4OE-I1AA  OITA , /i it T~P`JL : 1~/1~ #1~A~1LTA ~ (iW ) 4 .1tt  TA Ail (#ilk )  e  lA11!<JLL  q q A, W  (  **  g  --TA Jc nav) TI  i? q q  TA Ail  /#tk  ovum  61a7e  A  1fflitt%allItsi? 0 o I  q q ilk, bk4tk/J 5  ?  TA )c  ovum  1gtlifiE.T113*11a p&S+1.f*e?  Atilt  TA  274    q A, JM1)u/*  9 >t3tX  'PEA warm  A .Tttit*tk)R AVIV  q M *  q 4~*  io ItMA.7. 0*84MI q AI> 4  q—flit  qfi*R*  — G, OWN, *R, RN, I k ?  I-2 Z)  qtA*O,Ao  iARA ,  3~ o *1 I M  q  1`A )c (xa*)  A/* (wriia )  ,  A/A  (walk* )  It. *bit, qq TA  A  1 : A ;e SAT l**l?  Dq  O11, t*t. ? ill*IEW  > PPfi*IN:  D  IICM# O§*ale qX~ (7i a ilt9~)  q RENikfk  ttt  .o RETRAIN**  D  q REM*O  NAi-4A .S.# 4$1#f;i .fit*1db T'Mt*? (N31,, AM, T*M10 4'1tP(soft drinks), *3 St M) 2:  q iR  q —  q q  q  E  R  q E  q  ~~  3 *MIA-7-uETP.1* 9Bui? Mt ( * SA, MA, '3` 7R  ca*%I,  PJ~,  RA 1 . ..)  0  1 q TA  AA*, ) q 1 q TA  OSIR?  275    oo1X 4  o*A  **?  t 1 ~aEOfkA  OM$  r  q lH  oo#~  oo45A  ~$  A  a* 1E1  q 0*I.JM  00WA    A  q *a  q o1&M  0 q*A  o* El  q o*I1  qq *A  UlQa t3lE  Y  M ±SEIX$1ffi  (A'$  A)  It 4P± SfIV, (1 q A PtS$$ASOi 5  o*I  Yee  A  o#i El  o q#J&NJ  qOSA  q *Q  00*$c  Oo  W'YIS), ' M*101111 .. .)  of&M  i 0*  q q* IILX4  A of n  A  q o*A  3 rtA *X at/ AlbFRJ MO i ?  f03 (T$M$N3)  MIR  %  o1 o*  ANI  iiRMIMIM i3'tf4 (*ROM) 	Y  oo* I  00*la  q0  *Mtq  e;  g *41  TtIza LF'1, "' ' /iMP 4Fil gi? 001 qq T, *1E*  F'E  6.*Al TA+ BE 7. 1'ftiI  o*LL  q  /$X9#  III  A/a  *SW*  1.TA .ttktl0ftali`i(Pn rl:IrlilA, t11A, *iNiA,  MIA ~1 . . ..)  41* 2.  t1k a1t*? it*   R, k 13IE, kit  iR ~jIE  276    q Xfi**k  Oily *  q  q X rtl* !►  i  0  OOiFp*k  OiC D  s t * aUIM0 (1 t)? 0 .,PT  q  $10,000/*  q $ 50,000-60,000/ if  q$ 10, 000-20,0004  q $ 60,000-70,0004  0$ 20,000-30,0004  q $ 70,0004  q$ 30,000-40,0004  O F WI I*  0$ 40,000-50,0004  0 D3  1&A#t1b  6 AA—1-=tA 6.1 ItIMIM.64*43i tetA' A gtItlfiN*•Ml!?  q *ut{iI 0AT 6.2 I t N T 6 9 *WET T O f.EH4 J,1` Wi * !? 0 `A q 'sctIAA q )A 6.3 ® J3 ' c, tiEM7 ►E1iiJI*I ? q J~ 0 *r=t*i AT  0fi9~ "i  0  0 fi991a'  q  7 iiiM4tt it lf*t,`fi i(postal code): V_  iv. Icb)Lff t Wi MI : ffN12M I4f (fii :  1140—t4lIE Mimi*  MA t''=M  277     1  2  3  4  5  6  f/#tq*** rPl*i3 3  4  5  6  *A XtT% *A t 4* * T*M TIM F 3. t7 1  NA  1>011 *M *  *A RITZ *A TIM TIM TIM 2. 7i (7 — 33MY,bRf, 1 2  7  7  NA  itZ MC A  1ME*R N11141JL* . 2  3  4  5  *A l!€Txt *A TIM TIM F*M *M tat*  6  7  NA  *%'  * *M *  4. **AA-#*4>ittAitM -4f* 1  2  3  4  A *A TIM TIM TIM  5  6  7  1 *M  *A  le  NA  J),IJ11 *  5. tPXel, XA*Mth&*. 1  2  g T* i  n  3  4  *A fi*!F T* tk3 *  5  6  7  *A  NA  1>1031 1M IM  6. At*T±111***ta4, l!'tlltt1faM11 gPt 1  2  3  4  5  6  *A *A TIM F* F*~F Wit* *M 7.  7  NA  *C  *  1='I IRC*RfI f3 f3 (,4R,11R,&t)  1 	IF iihv-J  2 3 4 5 L5 PFXv-J. *B  6 tee  7 4~  NA XII i  278    8. g 1  * gZ#MM t i(t-,g , ARIOPM1, f 2 3 4 5 6 4fl A  *A RITZ  9 . 43#I JYPRJt 1t, X$*%Pfl 1 2 3 4 4fl  TI 10. f *T1dt* 1 2 W'4  *A TIM  IM  Ai  7  NA  4X-4 JiM  A  .  5  6  TV *A t~1  3  4  5  6  *A N1TV  t13i  11.MttZ,auietOM WI. 1 2 3  4  5  itV  NA 1).011  I13  A  7  NA  'A  TIM *Ai* 1  12.*64itikikolM±4Mi` ti#-*111 . 1 2 3 4 5 *x.4 *fl A *A *A T T tit  7  6  A *A ATV *A  *  NA  ea St*A*i  TIM TIM T  T  7  4fl AJJ`~  t~3  TIM fi  ig Z  1A  -Kuffi.  JA  1  JM A  6  7  Ai A 1M JIM  NA J),U11  A  13.*AIMi& MA**1t 1  2  3  4  4Y-4 Al A *A IPA T TIM TIM TIFF tt i '  5  6  7  *A  JF  JM  NA  )),kOT A  ** fi t#1A1fit ]AAM 11 e  279  Appendix III-12 : Subject's Information letter  THE UNIVERSITY OF BRITISH COLUM BIA Department of Health Care and Epidemiology Faculty of Medicine Mather Building, 5804 Fairview Avenue Vancouver, B .C . V6T 1Z3 Tel : 604-822-2772 Fax : 604-822-4994 Website: www .healthcare .ubc .ca  SUBJECT INFORMATION AND CONSENT FORM FOR PARENTS/GUARDIANS Study title : The population pattern of low hair zinc in preschoolers  Principal investigator: Dr . Clyde Hertzman, MD, M .Sc, FRCPC Health Care and Epidemiology University of British Columbia Phone: (604) 822-3002 Fax: (604) 822-0640  Dear Parent/Guardian: You are being invited to take part in this research study because you have a preschooler 24 - 71 months old and are residing in the city of Vancouver. Your participation is entirely voluntary, so it is up to you whether or not you wish to take part in this study. Before you decide, it is important for you to know that your written consent is required, and to understand what the research involves . This consent form, which you are asked to sign if you wish to participate, will tell you about the study, why the research is being done and what is involved in participation . If you do decide to take part in this study, you are still free to withdraw at any time without giving any reason for your decision, and this will not affect you or your child in any way. Please take time to read the following information carefully and to discuss it with your family or friends before you decide. This project, comprising a part of the doctoral thesis of research student, Ziba Vaghri, is conducted through the University of British Columbia and in affiliation with the Human Early Learning Partnership and I, Dr . Clyde I lertzman a UBC professor, am the chief investigator of it. Zinc is an essential micronutrient available from many sources including meat, dairy, cereal and grain products . Inadequate zinc intake can result in zinc deficiency . It has been shown that even minor zinc deficiency, termed marginal zinc deficiency (MZD), can interfere with children's normal growth and development . MZD can be easily prevented if adequate amounts of zinc-rich foods are included in a child's diet . When properly diagnosed, the correction of MZD is, also, relatively simple . Since the  280  hair zinc content in children is considered an index of the body's zinc content, it can be used in detecting MZD. Children with M.ZD often have lower hair zinc content. The main purpose of this study is to investigate the hair zinc status of a representative group of Vancouver preschoolers . The second purpose of this survey is to examine and understand the dietary and environmental factors that could predispose children to low hair zinc . The information collected through this survey will increase our awareness on the at-risk groups of children in our communities. Please note that only one child from each household is eligible to participate. If you choose to participate, you will be invited to: 1. Read the consent form and sign it. 2. Answer the attached questionnaire, which will take 10-15 minutes . (You do not have to answer any questions on the questionnaire if you feel uncomfortable in answering). 3. Return the signed consent form together with the completed questionnaire to your childcare center. By signing you agree to participate in the study and thereby give us permission to measure your child's height and weight, and take a small hair sample . Upon receipt of your completed survey questionnaire and signed consent form, our research team will set up a day with daycare authorities to be present in the center and do the measurements and collect the hair samples of the children whose parents/guardians have consented . You will be notified about this date and if you wish to be present for the sample collection, you are welcome to do so, however; this is not a requirement. The procedure will be explained to your child and a tablespoonful of hair sample will be collected from the back of his/her head (3-4 different locations) while he/she is sitting on a chair . There is no pain or side effect for your child in this procedure. Your name will be entered in a raffle draw (with a 1/20 chance to win a gift basket with clothing and accessories useful to your child and yourself). There may or may not be direct benefits to your child from taking part in this study. But the information learned from this study may be used in the future to benefit children affected by marginal zinc deficiency. All of the information you provide, together with the result of your 281  child's hair analysis, will be confidential and will not be available to other research personnel . Your child will be given an identification number (the number at the top of this letter) and will be identified by this number only. All information will be kept in a locked filing cabinet in the principal investigator's office at the University of British Columbia . No information that discloses your identity will be released or published without your specific consent and no records, which identify you by name, or initials will be allowed to leave the Investigators' offices. Two weeks following the return of your completed questionnaire to the childcare center, 10% of the subjects will be selected randomly (like the flip of a coin) and contacted again to repeat the survey questionnaire . This repeated task would help us to evaluate our survey questionnaire and further examine its ability to collect the information it is designed to collect . Even if you decide to participate in the main survey, you can decide not to participate in this second part of the survey . However, a $15 grocery voucher will be sent to you with the second questionnaire. The research team will not share any part of the collected data or result of the hair analysis with the study parents/guardians individually. However, the results in the form of any future publications will be available to all through their daycare centers. If you have any concern about your rights as a research subject while participating in this study, contact the `Research Subject Information Line in the University of British Columbia Office of Research Services' at 604-8228598 . If you have any questions about this study before or during participation, you can contact Ziba Vaghri at (604) 812-1419 or Dr . Clyde Hertzman at (604) 822-3002 . Sincerely, Clyde Hertzman MD, M .Sc, FRCPC Director of Human Early Learning Partnership University of British Columbia  282    Appendix 1I1-13 : Informed Consent form for the survey participants Informed consent:  • •  "I have read and understood the subject information and consent form. I have had sufficient time to consider the information provided and to ask for advice if necessary. • I have had the opportunity to ask questions and have had satisfactory responses to my questions. • I understand that all of the information collected will be kept confidential and that the result will only be used for scientific objectives. • I understand that my participation in this study is voluntary and that I am completely free to refuse to participate or to withdraw from this study at any time • I understand that I am not waiving any of my legal rights as a result of signing this consent form. • I have read this form and I freely consent to participate in this study. • I have read and understood that by signing this consent I will also give my permission to the research team to measure my child's weight and height and collect a spoonful of hair sample from him/her • I have been told that I will receive a dated and signed copy of this form. • I have been told that I will not receive the result of hair analysis of my child, since this information is only valuable to the research.  Printed name of subject's Date legally acceptable representative  Signature  Printed _lame of witness Date  Signature  Printed name of principal investigator/ designated representative  Signature Date 283    Appendix III-14 : Log Sheet of returned surveys  UBC  Study code  Please sign the log sheet when you drop your survey into the box. THANK YOU FOR PARTICIPATING IN OUR SURVEY!  Date dropped  Comments  284  Appendix Ill-15 : Reminder Letters  THE UNIVERSITY OF BRITISH COLUMBIA  Department of Health Care and Epidemiology Faculty of Medicine Mather Building, 5804 Fairview Avenue Vancouver, B .C . V6T 1Z3 Tel : 604-822-2772 Fax : 604-822-4994 Website : www .healthcare .ubc .ca  Dear parent/guardian;  While ago you received a survey package through your childcare center . The purpose of the survey, entitled "Population Patterns of Hair Zinc in Preschoolers", is to collect some information that will help us better understand zinc deficiency which is a common and silent deficiency in childhood. To date we have not received your survey back . We are inviting you to take 10-15 to complete this survey and return it back to your center along with the signed consent form. In appreciation of the time and effort you invest in this study, your name will be entered in a raffle draw to win a prize. Sincerely, Clyde Hertzman MD, MSc, FRCPC Director of E••Iuman Early Learning Partnership University of British Columbia  285    Appendix III-16 : Notice posted in the centers announcing the day and time of the survey team visit for the hair sample collection and the height & weight measurements  THE UNIVERSITY OF BRITISH COLUMBIA  Department of Health Care and Epidemiology Faculty of Medicine Mather Building, 5804 Fairview Avenue Vancouver, B .C . V6T 1Z3 Tel : 604-822-2772 Fax : 604-822-4994 Website : www .healthcare .ubc .ca  ATTENTION PARENTS ON  WE WILL BE AT  YOUR CHILDCARE CENTER TO MEASURE YOUR CHILDREN AND COLLECT THEIR HAIR SAMPLES. PLEASE MAKE SURE YOUR CHILD IS IN THE CENTER IF HE/SHE IS A PARTICIPANT IN "CITYWIDE SURVEY OF PRESCHOOLERS" ."  286  Appendix III-17 : Covering letter for the Reliability Package  THE UNIVERSITY OF BRITISH COLUMBIA  Department of Health Care and Epidemi Faculty of Medicine Mather Building, 5804 Fairview Avenue Vancouver, B .C . V6T 1Z3 Tel : 604-822-2772 Fax : 604-822-4994 Website : www.healthcare .ubc.ca  Dear parent/guardian; A few weeks ago you participated in our survey entitled "Population Pattern of Low Hair Zinc in Preschoolers" . In the covering letter you were informed that some of the participants would be asked to fill out the same questionnaire again . This task is repeated for a specific purpose, one that enables us to verify that the questionnaire is a valid tool for the research involved.  With this letter we would like to inform you that you have been selected randomly (like the flip of the coin) to fill the questionnaire again . Please take another 10-15 minutes to answer the questions and return it to your childcare center at your early convenience . Please note that this time no hair sample will be collected from your child. As a token of our appreciation for your cooperation in this survey we have enclosed a grocery certificate in the value of $15 . Once again, we appreciate the time and effort you invest in this important matter and we look forward to receiving your second questionnaire. Sincerely, Clyde Hertzman MD, M .Sc, FRCPC Director of Human Early Learning Partnership University of British Columbia 287  Appendix 111-18 : Letter to the gift basket winners  UBC  UNIVERSITY OF BRITISH COLUMBIA "Citywide Survey of Preschoolers' Hair Zinc"  Dear parent, Congratulations! Upon completion of our citywide survey, we have drawn names of participants for gift baskets, and your name has been taken randomly as one of the winners of this draw.  Along with this letter you will receive a gift basket containing a few useful items for you and your little one.  In this basket, you will also find a gift certificate with a value of $20 from Safeway. It will be greatly appreciated if you call our survey line at 604-827-4525 and confirm the receipt of your basket.  Enjoy your gifts and once again we thank you for participating in our study!  With warm regards, Clyde Hertzman MD, M .Sc, FRCPC Director of Human Early Learning Partnership University of British Columbia  Ziba Vaghri . BN, M .Sc PhD candidate University of British Columbia Food, Nutrition and 1lealth  288  Appendix III-I9 : Pictures from one of the basket draw winners  289  Appendix III-20 : Thank you letter to the survey participants  THE UNIVERSITY OF BRITISH COLUMBIA  Department of Health Care and Epidemiology Faculty of Medicine Mather Building, 5804 Fairview Avenue Vancouver, B .C . V6T 1Z3 Tel : 604-822-2772 Fax : 604-822-4994 Website : www .healthcare.ubc .ca  Dear survey participant,  A few weeks ago you extended your kind support to us and participated in our survey entitled "Population Patterns of 1-lair Zinc in Preschoolers" . Your contribution to the study, along with that of the others, has provided valuable information for our research on hair zinc status of preschoolers in Vancouver . We expect this information to further our understanding of zinc deficiency among our children. With this letter we would like to acknowledge the significant contribution your participation has had on the success of our survey. Thank you! With warm regards and best wishes, Clyde Hertzman MD, M .Sc, FRCPC Director of Human Early Learning Partnership University of British Columbia Ziba Vaghri BN, M .Sc, PhD candidate Faculty of Land and Food Systems University of British Columbia  290    Thank yo  To Gail and the wonderful staff of the Pacific Spirit Day Care: Your cooperation and kind support of our survey helped greatly in bringing it to a successful completion. Thank you for participating in our survey! Thank you for making a difference! On behalf of the "UBC city-wide preschooler " and with warm regards, surve  291     Appendix III-22 : Thank you letter to West Coast Family Services THE UNIVERSFIX OF BIUT SH COLUMBL9 HELP Li'  within  nn int  the hacuiy  ceipt.'nary , ,a h inrtitlte r,1Graduate Studies  Clyde Hertzman Director 604 822 3002 Tel  604 .822.0640 Fax clyde.hertzman@ubc .ca  January 3, 2007 Dear Diane, In 2006 we launched and completed a citywide survey of preschoolers' hair zinc . Your agency was key in helping us connect with parents of children 2-6 years old . Your office extended kind support to this survey by endorsing it via a letter written to the childcare professionals in the preschool centers within the Registry as well as by introducing our survey to the readers of your monthly newsletter . The survey was completed successfully with the participation of over 60 centers and more than 800 preschoolers. We have developed an educational pamphlet that provides important information on zinc, its dietary sources, and ways to ensure its adequacy in children's diets . This pamphlet has been well received by the parents and caregivers of our focus groups set up to verify the ease of comprehension at level of literacy of the pamphlet. This pamphlet will he supplied to the participating centers (only) within the next few weeks . But as a token of our appreciation for your support of the survey, we wish . by this letter, to inform you that an electronic version of the pamphlet can be made available to you, should you wish to supply the other centers under the Registry's jurisdiction. Once again, thank you for your partnership in this important project! Yours sincerely, Clyde Hertzman MD, M .Sc, FRCPC Ziba Vaghri, BN, M .Sc, PhD Candidate Director of Human Early Learning Partnership University of British Columbia University of British Columbia Food, Nutrition, and Health 292  Appendix III-23 : The Educational Pamphlet     s  o  3 a -o  .O N  Vvo  .E  uv  Cs  N N  c -C a 0) U i 0) }q N II)• 0 V1 ~4d • 0 q 0)d L t q • N 13 L dCov N C 0)0 0) CL „p 0) • 01 NO OO 0C ..0 .,C E C 01 V • Ct) 13 01C 0 0d 01 . U 02 O T V •U 4' • O ••N C 0) N 0)} L. O 0 • Y• p 4'V O 3 0 0 C C =•1  N 5  E  	(  u)  7t- :E m o  (I)  O > co O' > co e.)  o  L ~ aa)) °– ~ i) :t'  H. .0 c c Q  U io  • • •  •  Q°' .L D  U N C O C  8 0 -0 ' co  o Via)  a) E 4- _J 0 > g  0  s  CN =p W N C • c ca w  I  a) E 0 I--  01  f C +- 4) ~' p  • t • .2)  a F Q v }i E a 3 1 .=:l  3  M  a n  N  A'  '7  0  0  C  -c  y  O-  -F A  D  	S    A 10  W  ~.  A C  A C -c  $-L  F+ N  v  a  7r  0 o  I--•  VI  0 3 O c  A  S  N  a  oo  (0 i , Ft' us S 3 0 3 a p  C  VI  y  `F  N  3  O  c  n  pp  W  A C 'Q  N  •  y  0  r . O  O`  hs  A l0  O c  O 3 A l0  ~-+  I-+  0a CT o0 sN o  a  O  3 a  O  `0  Q  3 A l0  -  N  O  ►-+  A  F+  0 (0  Cl. ((  °+  o _4.  0  y C A  3 y  U1  .P  N  N  N  W  A  O c  W  3 A  O  W  O  co  x..  0 10 i- 0_ -s 7c-  a  a`  a  o  x  OD  -c  A  F+  -t  -h  s s Co o c  S  3' o(0 N 3 3 A to  N  N  N  A  O  W  o  CL  a  -h C to (A  0  -s 0  OD  UI  N  A  O  CA)  0  CA)  0  O~  tO c , A 0 to IA 3 IA * v  Ul o  co  a a  0  n  -fh  3  ..+ (o  -4, 0 z  l0  3 La  n n  N  L0  (!1  .. 0  0  0 S C 0 <  W N ~+  to i  =to  -h 0 C O o_ N to  -h • o t  to lA  s to O 0  a  d  0  0 • a a  (0  0 0 a VI • 0 (  0 3  (0  $—&  Appendix III-24 : Log Sheet for the Hair Samples delivered to the laboratory  4  UBC  University of British Columbia Citywide survey of Preschoolers Monday June the 19th Batch #9 Total number of samples : 56 Subject code  Subject code  SS03001 SSO3002 SS03003 SS03004 SS03005 SS03006 SSO7OO4 SS07011 SS 07012 SS07016 G W04002 GWO4003 GW04005 GWO4006 G WO4008 G WO4OO9 GWO4O10 GWO6OO1 G W06002 G W06003 GWO6OO4 GWO6005 GW06006 RP-G-001 RP-G-002 RP-G-003 RP-G-004 RP-G-006 RP-G-007 RP-G-008 RP-G-009 RP-G-010 SS07011 SS 07012 SS0701 6  RP-BK-004 RP-BK-006 RP-BK-20 RP-BK-034 RP-BK-036 RP-BK-041 RP-BK-052 RP-BK-058 KLO1 017 SS-IR-001 SS-I R-003 SS-I R-004 SS-I R-005 SS-I R-006 SS-I R-007 SS-I R-008 SS-I R-009 SS-I R-011 SS-I R-012 SS-I R-013 SS-IR-014  Subject code  294    Appendix III-25 : Log Sheet for height and weight  UBC  UNIVERSITY OF BRITISH COLUMBIA Date: 	 Center: South Hill daycare (Branda) Name Weight (lb) Height (cm) Remarks Subject code  295  Appendix III-26 : Moments captured ; Nutrition clinics and survey  Spring of 2004, Nutrition Clinics, research student entertaining a study participant (up) and measuring her height (down)  296  Spring of 2006, survey anthropometrist in one of the participating preschool centers  297  Appendix 111 .27. Hair zinc level of 5 main ethnicities of the study along with the gender and age breakdown of these ethnic groups  Variables  Hair Zinc (µg/g)  Caucasian (n=302)  Chinese (n=219)  E. Indian (n=54)  116 ±41 a  104 ±43 6  154 ± 37°  163 (54) 140(46)  105(48) 117(52)  23(43) 31(57)  132 (44) 170(56)  103 (47) 116(53)  19 (35) 35(65)  Gender distribution #(%) Boys  Girls Age distribution #(%) < 48 months > 48 months  298  Appendix 111 .28. Regression analyses of "taking supplements containing iron" with eating behaviors Unadjusted Variables  Adjusted*  R2  P  B; (95% CI)  R2  p  Concerned about child's eating  0.02  0.00  0 .13(0.06, 0 .20)  0.03  0.00  Described as eating unhealthy  0 .01  0 .01  0.07(0.02, 0.12)  0.02  0.13  Described as picky eater  0.02  0.00  0 .13(0.06, 0.20)  0.05  0.00  Described as not eating enough  0.00  0 .01  0.06(-0 .01, 0 .13)  0.03  0.01  * Adjustments have been carried out using age, gender, ethnicity, maternal education and family income as covariates. $Bs are unstandardized co-efficients. &R2  represent the analysis .  R2 s  of adjusting covariates (R 2 =0 .12) combined with R 2 of the new variable in the  299  Appendix III-29 . Comparison of the indices of growth (l 1AZ and WHZ) and nutrition (WAZ) of low hair zinc and normal hair zinc younger (<4 years old) children Z scores HAZ  I.,ow hair zinc children (n=76) 0.03+0 .6  Normal hair zinc children (n=229 ) 0.16+1 .0  I' 0 .67  WAZ  0.60+2 .0  0 .451_1 .4  0.94  WHZ  0.61+2.5  0.41+1 .7  0.71  Low hair zinc : Hair zinc <70pg/g HAZ : Heigh-for-age Z scores WAZ : Weight-for-age Z scores WHZ : Weight-for-height Z scores  300  Appendix III-30 . Mean ± SD of the daily intake frequency of some main food groups of survey children and the three main ethnic groups. All (n=719) 1 .5 + 1 .1  Caucasian (n=305) 1 .9+0.7  Chinese (n=223) 1 .9+1 .2  East Indian (n=54) 1 .0+1 .2  Cereals and Grains  3 .5 + 1 .6  3 .1+1 .2  3.2+1 .6  3 .3+1 .7  Whole wheat C&G  1 .9 ± 1 .4  1 . .9+1 .2a*  1 .0+1 .2"  1 .6+1 .5'  Dairy  38+ 1 .6  3 .6+1 .5  3.1+1 .5  3 .4+2.0  Milk  2 .2 + 1 .2  2 .0±1 .1 "  1 .9+1 .1 0  1 .8+1 .3 a'  Meat & alternatives  Among the three ethnic groups values not sharing a common superscript (a,b, . .) are significantly different  301  0 supplement takers 0 Non-supplement takers  Appendix III-31 . Distribution of supplement takers among the survey participants (bottom pie) and distribution of iron supplement takers among the supplement takers of the survey (top pie) (n =671).  302  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0066609/manifest

Comment

Related Items