UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Prevalence of iron-deficiency anaemia and low iron status and feeding practices among 9 months old infants… Lwanga, Dorcas Namubiru 1996

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

Item Metadata

Download

Media
831-ubc_1996-0148.pdf [ 6.38MB ]
Metadata
JSON: 831-1.0087071.json
JSON-LD: 831-1.0087071-ld.json
RDF/XML (Pretty): 831-1.0087071-rdf.xml
RDF/JSON: 831-1.0087071-rdf.json
Turtle: 831-1.0087071-turtle.txt
N-Triples: 831-1.0087071-rdf-ntriples.txt
Original Record: 831-1.0087071-source.json
Full Text
831-1.0087071-fulltext.txt
Citation
831-1.0087071.ris

Full Text

PREVALENCE OF IRON-DEFICIENCY ANAEMIA AND LOW IRON S T A T U S , AND FEEDING PRACTICES A M O N G 9 MONTH OLD INFANTS IN V A N C O U V E R . By DORCAS NAMUBIRU L W A N G A B . A . S c , The University of Guelph, 1992 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE in THE FACULTY OF GRADUATE STUDIES (School of Family and Nutritional Sciences) We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA J A N U A R Y , 1996 © DORCAS NAMUBIRU L W A N G A , 1996 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of SCHOOL O? t^tMuj /frop tvuT£<TioroM SaierocsS. The University of British Columbia Vancouver, Canada Date HMCH DE-6 (2/88) A B S T R A C T ABSTRACT Iron-deficiency anaemia (IDA) is considered to be the most common nutritional deficiency among infants and children worldwide. The consequences of IDA on the developing central nervous system may be irreversible; these may include delayed mental and motor development, and reduced school performance. Infants are particularly prone to develop IDA if given foods low in iron content or foods that contain iron of low bioavailability. Infants from low socioeconomic background and Asian, Black or Hispanic infants, are believed to be at higher risk for IDA than White infants. There are no published studies on iron status in relation to feeding history from British Columbia or other parts of Canada. The purpose of this study was to determine the prevalence of IDA and low iron status in 9 month old infants in Vancouver, British Columbia, and to determine which infants are at highest risk for IDA and low iron status based on their feeding history, economic and ethnic background. Infants who could participate in the study were identified from birth and death lists provided by the Vancouver Public Health Department. Eligibility criteria were that the infant was full term (gestational age a: 37-<42 weeks) with a birth weight of 2500-4500g and born between January 1st, and March 2nd, 1993, or between June 4th and August 7th, 1993 to parents resident in Vancouver, with an address to enable contact. Initial contact with parents of all eligible infants (n = 1813) was made through a letter. A subsequent telephone call was made to the parents to describe the study protocol and to arrange an appointment for parents interested in participating with their infant. All the appointments were made to coincide with the time when the infant would be 39 ± 1 weeks old. At the clinic appointment, blood samples were collected from the infants, parents completed questionnaires regarding their family background and their infant's nutritional history from birth to 9 months of age, and a visual ii A B S T R A C T recognition memory test (Fagan Test of Infant Intelligence) was administered to the infant. Because many tests used to diagnose iron-deficiency anaemia lack specificity, several tests were used in combination. IDA was classified as a Hgb ^101 g/L or Hgb <110 g/L with 2 or 3 abnormal biochemical tests from serum ferritin :S 10 / /g /L , total iron binding capacity > 60 //mol/L, and zinc protoporphyrin >70 //mol ZPP/mol heme. Low iron status was classified as a serum ferritin ^ 1 0 / /g/L without iron-deficiency anaemia. Four hundred and thirty four (434) 9 month old infants and their parents participated in the study, representing 23.9% of all the eligible infants. The prevalence of IDA and low iron status was 6.9% and 24.4%, respectively. A statistically significant association (p<0.0001) was found between the infants' iron status and the duration of breast-feeding. IDA and low iron status was found in 15.2% and 30.4%, respectively, in infants who had received breast-milk as their main source of milk for more than 8 months. In contrast, the prevalence of IDA and low iron status was 1.5% and 10.3%, respectively, among the group of infants who were never breast-fed. Feeding low iron milk (cows' milk, low iron infant formula or goat's milk) also showed a statistically significant (p < 0.05) association with the iron status of the infants. No statistically significant association was found between iron status and the age of introduction of specific solid foods (iron-fortified infant cereals, fruits, vegetables, legumes, egg yolk, tofu, meat, chicken, or fish), or fruit juice. No statistically significant association was found between the annual family income and iron status of the infants, when considering either the entire group of infants, or the infants from two parent households. However, a statistically significant association (p< 0.05) was found between iron status and family income for the group of infants (n = 30) from one parent families. Of note, only 7 of these infants were from families with an annual income of iii A B S T R A C T > $20,000. The low number of infants in this group limits the ability to predict the true prevalence of IDA and low iron status, or the association with income in the single parent families. No statistically significant association was found between the infants' iron status and the mothers' level of education. A higher prevalence of iron-deficiency anaemia and low iron status was found in infants of mothers born in Canada compared to infants of mothers not born in Canada. The difference was statistically significant (p<0.05) suggesting a higher risk for iron-deficiency anaemia and low iron status among infants' whose mother had been born in Canada. No statistically significant association was found between the infants' iron status and the number of years an immigrant mother had resided in Canada. Ethnic background as reported by the mother was significantly associated (p<0.05) with the iron status of the infants. Specifically, the prevalence of IDA was higher among infants of European and Canadian parentage than among infants of East Indian and Chinese parentage. No statistically significant differences were found between the scores on the visual recognition memory test of the infants with iron-deficiency anaemia, low iron status or normal iron status. In conclusion, the prevalence of iron-deficiency anaemia (6.9%) and low iron status (24.4%) among otherwise healthy 9 month old infants who participated in this study suggests the need to develop strategies for the prevention of iron-deficiency anaemia, or for early detection and treatment. The results of this study show that the infants at highest risk for iron-deficiency anaemia and low iron status in Vancouver, when defined by feeding history, are infants with a history of breast-feeding as the main source of milk for more than 8 months, and infants bottle-fed low iron milk (cows' milk, goat's milk or low iron infant formula). When defined by economic and ethnic background the infants at highest risk for iron-deficiency anaemia are infants of Caucasian mothers (European or Canadian). iv TABLE OF CONTENTS TABLE OF CONTENTS ABSTRACT ii ACKNOWLEDGEMENT xiii 1. INTRODUCTION 1 1.1. BACKGROUND 1 1.2. PURPOSE OF THE STUDY 3 1.3. STUDY OBJECTIVES 4 1.4. HYPOTHESES 5 2. REVIEW OF LITERATURE 6 2.1. IRON METABOLISM 6 2.1.1. Heme-containing proteins 6 2.1.2. Non-heme proteins 7 2.1.2.1. Storage proteins: Ferritin and hemosiderin 7 2.1.2.2. Transport protein: Transferrin . 8 2.1.2.3. Transport protein: Lactoferrin 9 2.2. ABSORPTION AND REGULATION OF BODY IRON 10 2.2.1. Non-heme iron 10 2.2.2. Heme iron 11 2.2.3. Intestinal regulation of iron absorption 12 2.3. IRON STATUS IN INFANCY 13 2.3.1. Maternal-fetal relationships 13 2.3.2. Iron status of the infant after birth 14 2.3.3. Iron requirements of infancy 15 2.4. ASSESSMENT OF IRON STATUS 16 2.4.1. Stages of development of iron-deficiency 16 2.4.2. Laboratory measures of iron status among humans 19 2.4.2.1. Biochemical tests 19 2.4.2.1.1. Serum ferritin 19 2.4.2.1.2. Serum iron, total iron binding capacity and percent transferrin saturation 20 2.4.2.2. Hematological tests 21 2.4.2.2.1. Hemoglobin 21 2.4.2.2.2. Red cell indices 21 2.4.2.2.3. Zinc erythrocyte protoporphyrin 23 2.4.2.3. Diagnosis of iron-deficiency 25 v TABLE OF CONTENTS 2.5. IRON AND DEVELOPMENT 27 2.5.1. Location . 27 2.5.2. Function 28 2.4.3. Functional abnormalities associated with iron-deficiency 28 2.5.4. Association between iron status and cognitive development . . . 28 2.5.4.1. Evidence from animal studies 28 2.5.4.2. Evidence from human studies 29 2.5.5. Cognitive assessment 33 2.5.5.1. The visual recognition memory test (Fagan Test of Infant Intelligence) 33 2.6. IRON CONTENT IN DIETARY SOURCES OF IRON FOR THE INFANT . . . 36 2.6.1. Human milk 36 2.6.2. Cows' milk and low iron infant formulas 36 2.6.3. Fortified infant formulas 38 2.6.4. Solid foods . . . 40 2.6.4.1. Infant cereals 40 2.6.4.2. Other solids 40 2.7. INFANT FEEDING PRACTICES AND IRON-DEFICIENCY ANAEMIA 41 2.7.1. Breast-feeding 44 2.7.2. Cows' milk 45 2.7.3. Iron-fortified infant cereals 46 2.8. GUIDELINES FOR INFANT FEEDING 48 2.8.1. Canadian Paediatric Society Nutrition Committee 48 2.8.2. American Academy of Paediatrics Committee on Nutrition . . . . 49 2.8.3. The Vancouver Public Health Department 49 2.9. FAMILY INCOME, ETHNICITY AND IRON STATUS 50 2.9.1. Family Income 50 2.9.2. Ethnic background and prevalence of iron-deficiency anaemia . . 51 3. DESIGN AND METHODS 52 3.1. STUDY DESIGN 52 3.2. SUBJECTS : 52 3.2.1. Subject identification and selection criteria 52 3.2.2. Subject recruitment 52 3.3. CLINIC SCHEDULING 53 3.4. INFORMED CONSENT 54 3.5. ANTHROPOMETRIC MEASURES 54 vi TABLE OF CONTENTS 3.6. STUDY QUESTIONNAIRES 57 3.6.1. Developing the questionnaires 57 3.6.2. Validating the questionnaires 58 3.6.3. Completing the questionnaires 59 3.7. THE VISUAL RECOGNITION MEMORY TEST 59 3.8. HEMATOLOGY 61 3.8.1. Blood collection 61 3.8.2. Blood analyses 62 3.8.2.1. Analysis of hemoglobin 62 3.8.2.2. Analysis of zinc erythrocyte protoporphyrin 62 3.8.2.3. Analyses of ferritin, total serum iron and unsaturated iron binding capacity 63 3.8.2.4. Calculation of total serum iron, UIBC and TIBC 66 3.9. CLASSIFICATION OF IRON STATUS 67 3.10. DATA ANALYSIS 69 4. RESULTS 70 4.1. CLINIC ATTENDANCE 70 4.2. STUDY SAMPLE SIZE 74 4.3. DESCRIPTION OF THE STUDY POPULATION 75 4.3.1. Gender distribution 75 4.3.2. Parental age distribution 76 4.3.3. Marital status 79 4.3.4. Family income . 80 4.3.5. Parental education 83 4.3.6. Parental ethnic association 86 4.3.7. Immigrant and non-immigrant mothers 88 4.4. VISUAL RECOGNITION MEMORY TEST 90 4.5. FEEDING HISTORY 92 4.5.1. Duration of breast-feeding 92 4.5.2. Introduction of solid foods 94 4.5.3. Use of low iron milk 96 4.6. PREVALENCE OF IRON-DEFICIENCY ANAEMIA AND LOW IRON STATUS 98 vii TABLE OF CONTENTS 4.7. IRON STATUS AND ITS RELATION TO FEEDING PRACTICES 100 4.7.1. Relationship between iron status and duration of breast-feeding 100 4.7.2. Iron status in relation to use of low iron "milk" 102 4.7.3. Iron status in relation to the age of introduction of various foods 103 4.8. IRON STATUS AND SOCIOECONOMIC FACTORS . 105 4.8.1. Relationship between iron status and maternal education 105 4.8.2. Relationship between iron status and annual family income . . . . 107 4.9. IRON STATUS AND ETHNIC BACKGROUND 111 4.9.1. Infants iron status in relation to whether or not the mother was born in Canada 111 4.9.2. Relationship between iron status and number of years immigrant mother has been resident in Canada 111 4.9.3. Relationship between mothers' ethnic association and the infants' iron status 114 4.10. IRON STATUS AND PERFORMANCE ON THE VISUAL RECOGNITION MEMORY TEST 116 5. DISCUSSION . . 117 5.1. PREVALENCE OF IRON-DEFICIENCY ANAEMIA AND LOW IRON STATUS 117 5.2. FEEDING PRACTICES 120 5.2.1. Duration of breast-feeding and prevalence of iron-deficiency anaemia and low iron status 120 5.2.2. Feeding of low iron "milk" and iron status 123 5.2.3. Introduction of solid foods and prevalence of iron-deficiency anaemia and low iron status 124 5.3. SOCIAL, ECONOMIC, AND ETHNIC BACKGROUND 125 5.3.1. Socioeconomic status and prevalence of iron-deficiency anaemia and low iron status 125 5.3.2. Parental ethnic association and its relation to the infant's iron status 126 5.4. IRON STATUS AND COGNITIVE FUNCTION 130 5.5. CONCLUSION 132 5.6. RECOMMENDATIONS FOR FUTURE RESEARCH 133 viii TABLE OF CONTENTS BIBLIOGRAPHY 135 APPENDIX A 144 SAMPLE SIZE CALCULATION 144 APPENDIX B 145 1) LETTER FROM THE VANCOUVER HEALTH DEPARTMENT 145 2) LETTER DESCRIBING THE STUDY 146 3) CLINIC SCHEDULE SET 1; OCTOBER-NOVEMBER, 1993 147 4) CLINIC SCHEDULE SET 2; MARCH-APRIL, 1994 148 APPENDIX C 149 INFORMED CONSENT FORM 149 APPENDIX D 150 CONFIDENTIAL DEMOGRAPHIC QUESTIONNAIRE 1 50 APPENDIX E 153 DIET QUESTIONNAIRE 153 APPENDIX F 162 ANTHROPOMETRIC FORM 162 APPENDIX G 163 CONTACT CARD FOR BLOOD RESULTS 163 APPENDIX H 1 64 HEMATOLOGY 164 APPENDIX I 165 AGE OF INTRODUCTION OF VARIOUS FOODS WITHIN EACH IRON STATUS GROUP (FIGURES 4.7-4.16) 1 65 APPENDIX J 175 AVERAGE AGE OF INTRODUCTION OF VARIOUS FOODS WITHIN EACH IRON STATUS GROUP 175 APPENDIX K 176 COWS' MILK FEEDING IN THE STUDY POPULATION 1 76 APPENDIX L 177 BREAST-FEEDING AMONG CAUCASIAN VERSUS NON-CAUCASIANS MOTHERS 177 ix LIST OF TABLES LIST OF TABLES Table 2.1. Summary of changes in laboratory parameters of iron status during iron-deficiency anaemia 24 Table 2.2. Common laboratory tests and cutoff values used in the diagnosis of iron-deficiency in young children 26 Table 2.3. Iron content of infant formulas on the market in the Greater Vancouver Area 39 Table 3.1. Summary of cutoff values used to classify infants iron status 68 Table 4.1. Sampling of study population 73 Table 4.2. Summary of the number of infants who participated in the study . . . . . 74 Table 4.3. Gender distribution of study population 75 Table 4.4. Marital status of the parents 79 Table 4.5. Parental ethnic association 87 Table 4.6. Number of mothers born or not bom in Canada 88 Table 4.7. Number of years immigrant mothers had lived in Canada 89 Table 4.8. Summary of Fagan Tests 90 Table 4.9. Reasons for low confidence in scores on the Fagan Test. 91 Table 4.10. Duration of breast-feeding 93 Table 4.11. Age of introduction of various foods 95 Table 4.12. Age of introduction of low iron "milk" for infants fed a low iron milk for 1^ month 97 Table 4.13. Prevalence of iron-deficiency anaemia and low iron status among nine month old infants in Vancouver 99 Table 4.14. Prevalence of iron-deficiency anaemia and low iron status in relation to duration of breast-feeding 101 Table 4.15. Number of infants with a history of feeding low iron "milk" by iron status 102 x LIST OF TABLES Table 4.16. Iron status by number of infants for whom various foods have or have not yet been introduced by 9 months of age 104 Table 4.18. Infants' iron status in relation to the mothers' reported level of education 106 Table 4.19. Prevalence of iron-deficiency anaemia and low iron status in relation to annual family income 108 Table 4.20. Prevalence of iron-deficiency anaemia and low iron status in relation to annual family income from two adult families 109 Table 4.21. Prevalence of iron-deficiency anaemia and low iron status in relation to annual family income from one adult families 110 Table 4.22. Prevalence of iron-deficiency anaemia and low iron status in relation to whether of not the mother was born in Canada 112 Table 4.23. Prevalence of iron-deficiency anaemia and low iron status in relation to the number of years an immigrant mother has been resident in Canada 113 Table 4.24. Prevalence of iron-deficiency anaemia and low iron status in relation to mothers' ethnic 115 Table 4.25. Fagan test results among infants in the different iron status groups . . . 116 Table 4.17. Average age of introduction of various foods within each iron status group for those infants who were actually fed the food before 9 months of age 175 xi LIST OF FIGURES LIST OF FIGURES Figure 2.1. Stages of development of iron-deficiency 18 Figure 2.2. A description of the overlapping periods of the major sources of dietary iron during infancy 43 Figure 4.1. Distribution of mothers' ages 77 Figure 4.2. Distribution of fathers' ages 78 Figure 4.3. Annual income of two adult families 81 Figure 4.4. Annual income of one adult families 82 Figure 4.5. Highest level of education attained by mothers 84 Figure 4.6. Highest level of education attained by fathers 85 xii ACKNOWLEDGEMENT ACKNOWLEDGEMENT Mukama ayinza bwona yebazibwe. I appreciate and sincerely thank my supervisor Dr. Sheila Innis for giving me the opportunity to do this research project and for her excellent guidance and advise throughout the completion of my thesis. Many thanks to my other committee members Dr. Gwen Champman and Dr. Louis Wadsworth for their assistance and direction. I also wish to thank Corrine Eisler and Dr. Joseph Leichter for serving as members of the examining committee. I gratefully acknowledge the Vancouver Public Health Department for funding the project. I would also like to thank all the parents of infants in the study whose co-operative participation made this research project possible. I wish to extend a very big thank you to the Vancouver Health Department Nutritionists, Corrine Eisler, Helen Yeung, Vicki Boere and Barbara Crocker for the time they dedicated to the clinics. Thank you to Dianne Jacbson the research project nurse whose expert organizational skills ensured the smooth running of the clinics. I would also like to especially thank Paula Walsen, France Rioux and Carolanne Nelson for the time and energy they dedicated to data collection during the clinics. Very special thanks to Laurie Nicol and Roger Dier for their invaluable help, patience, support and guidance in the research lab. To my fellow graduated students, Lynn, Charitini, Trish, Fabio, Sabina, Eileen, Carolanne, Patty and Elly, thank you for your unfailing support, help, encouragement and friendship during these challenging years. Finally, extra special thanks to my parents, Veronica and Tucker and my brothers Christopher, Shemu and Yosiya and my darling sisters Sheba and Emma, for their continued support, understanding and love in so many ways throughout this challenging experience - "Si non c'est fini". And to Micheal, thank you for your love and support that encouraged me to persevere through the completion of this degree. Mwena mwebale nyo. xiii INTRODUCTION Chapter I 1. INTRODUCTION 1.1. BACKGROUND Iron-deficiency anaemia is the most common nutritional deficiency in both developing and industrialized countries, with a high incidence among infants, young children and women (Lozoff, 1988). In Canada, nutritional iron-deficiency is still a problem among infants and young children despite advances in knowledge of infant nutrient needs and notable changes in feeding practices since the mid 1970's (Green-Finestone et al., 1989; Tanka et al., 1989). In Montreal, for example, the prevalence of iron-deficiency anaemia among Chinese children aged 6 to 36 months was reported to be 12.1% (Chan-Yip & Gary-Donald, 1987). The prevalence of iron-deficiency anaemia in infants in other parts of Canada, including British Columbia, is unknown. In infants, iron-deficiency is most prevalent between 6-24 months of age, which coincides with the later part of the brain growth spurt (Lozoff, 1988). This period of the brain growth spurt lasts from about mid-gestation until the second year of life (Dobbing, 1990). In the past decade, the relationship between iron-deficiency and the cognitive development of infants and young children has been the focus of a large body of research. Several studies have provided consistent and convincing evidence that iron-deficiency sufficient to cause anaemia is associated with lower developmental scores as measured by the Bayley Scales of Infant Development (Grindulis et al., 1986; Lozoff et al., 1982; Lozoff et al., 1987; Walter et al., 1983; Walter et al., 1989). Of even greater concern, iron-deficiency anaemia in infancy 1 INTRODUCTION may have irreversible effects on developmental function that may result later in reduced school performance (Lozoff et al., 1991; Walter et al., 1990). Of note, a recent study found that early identification and treatment of infants 11-12 months of age with iron-deficiency anaemia resulted in correction of both iron-deficiency anaemia and low developmental test scores (Idjradinata & Pollitt, 1993). The magnitude of iron-deficiency anaemia and its effects on infant development could have considerable implications for the health care, education and social systems. The high prevalence of iron-deficiency anaemia in infants and young children may be explained by the rapid growth that occurs during this stage in development. This leads to high iron requirements which may exceed the supply available from dietary intake and endogenous iron stores (Dallman et al., 1980). In addition, feeding practices such as prolonged exclusive breast-feeding, use of low iron infant formula, and use of cows' milk with delayed introduction of iron-containing foods, such as iron-fortified infant cereals and other rich sources of iron such as meat, are most likely to place an infant at risk for developing iron-deficiency anaemia. Consequently, the Nutrition Committee of the Canadian Paediatric Society (1991) and the American Academy of Paediatrics (1992) have written guidelines to address the issue of iron-deficiency anaemia in infants. The recommendations focus on the feeding practices that are most likely to place an infant at risk for the development of iron-deficiency anaemia. 2 INTRODUCTION 1.2. PURPOSE OF THE STUDY Inappropriate feeding practices are the primary cause of iron-deficiency anaemia among infants. The prevalence of iron-deficiency anaemia, feeding low iron-formula, weaning to cows' milk prior to 9 months of age, prolonged breast-feeding without introduction to iron-fortified infant cereals or supplemental iron in British Columbia is unknown. The prevalence of iron-deficiency anaemia seems to be highest among children 6 months to 24 months of age. It is also during the first 24 months after birth that rapid brain growth, differentiation and myelination occur. Numerous studies have provided consistent evidence that iron-deficiency anaemia can result in delayed or impaired mental and motor development. Therefore, if there is an association between iron-deficiency anaemia and poor developmental function in infants, it would be highly desirable to develop strategies to prevent iron-deficiency anaemia in infants. This requires information on the prevalence of iron-deficiency anaemia and the feeding practices during infancy that are associated with iron-deficiency anaemia. The main aim of the study, therefore, was to determine the prevalence of iron-deficiency anaemia and low iron status in infants at 9 months of age in Vancouver, and to determine which infants are at risk for iron-deficiency anaemia and low iron status based on their feeding history, and the economic and ethnic background of the family. 3 INTRODUCTION 1.3. STUDY OBJECTIVES The objectives of this study were to: 1. determine the prevalence of iron-deficiency anaemia and low iron status in infants 9 months of age. 2. determine the feeding practices, retrospectively from 9 months to birth of age which , are associated with low iron status and iron-deficiency anaemia, with respect to duration of breast-feeding, age of introduction of infant formula, cow's milk, juice and other sources of dietary iron such as infant cereals, meat, poultry, fish or legumes. 3. compare iron status among infants of different economic and ethnic backgrounds. 4. determine the relationship between iron status and scores on a visual recognition memory test. 4 INTRODUCTION 1.4. HYPOTHESES For the purpose of this research the null hypotheses were; 1. Iron-deficiency anaemia and low iron status will not be present among 9 month old infants in Vancouver. 2. There is no association between infants' iron status at 9 months of age and duration of breast-feeding, weaning to cow's milk, use of low iron infant formula, age of introduction of an iron-fortified infant cereal, or other dietary sources of iron. 3. There is no association between family income, parental ethnic association, and the prevalence of iron-deficiency anaemia or low iron status. 4. There is no difference in scores on the visual recognition memory test between infants with iron-deficiency anaemia and infants classified as iron-sufficient (i.e. infants having normal iron status). 5 REVIEW OF LITERATURE Chapter II 2. REVIEW OF LITERATURE 2.1. IRON METABOLISM Iron is an essential element in the body which plays a major role in hemoglobin synthesis, electron transport for cellular respiration, DNA synthesis and many enzymatic reactions (Massey, 1992). It is found mainly in its oxidative states as either ferric (Fe3+) or ferrous (Fe2+) iron, with higher oxidative states being short-lived as intermediates in certain redox reactions (Wells & Awad, 1993). Iron can also bind to and influence the structure and function of a number of macromolecules, leading to harmful effects on the organism. As a result, many of the iron-binding proteins function to store and transport iron (Wells & Awad, 1993). Body iron may be categorized into 1) iron essential for normal physiological functions, and 2) storage iron necessary for the regulation of iron homeostasis and maintenance of reserves. The reserves are important to ensure an adequate supply of iron for the production of essential iron compounds in times of need (Dallman, 1989; Finch & Huebers, 1982). 2.1.1. Heme-containing proteins Heme-containing proteins include hemoglobin, myoglobin, mitochondrial cytochromes and heme containing enzymes (e.g. tryptophan pyrrolase). Hemoglobin is the most abundant of the heme proteins and functions to transport oxygen via the circulatory system from the lungs to tissues. Myoglobin gives muscle its red colour and functions to store oxygen 6 REVIEW OF LITERATURE necessary for muscle contraction. The mitochondrial cytochromes are enzymes in the electron transport system and are essential for the production of cellular energy in the form of adenosine triphosphate (ATP) (Dallman et al., 1993). As a whole, these proteins all play a role in oxidative metabolism and make up the essential iron compounds necessary for normal physiological functions (Dallman et al., 1993; Wells & Awad, 1993). Iron binds to the protein by either being incorporated into a protoporphyrin IX ring or by interacting with other protein ligands. The ferrous protoporphyrin IX complex is known as hemoglobin and consists of one ferrous iron, a tetrapyrrole ring and a protoporphyrin IX molecule (Wells & Awad, 1993). 2.1.2. Non-heme proteins Non-heme proteins include ferritin and hemosiderin, and transferrin and lactoferrin which are involved in the storage and transport of iron, respectively. These proteins are involved in the regulation of iron homeostasis (Dallman et al., 1993; Wells & Awad, 1993). Other non-heme proteins include a number of redox enzymes that have iron at the active site, such as, iron-sulphur proteins and metalloflavo proteins. In addition, there are a number of compounds which do not contain iron, but need iron or heme co-factors. This group includes enzymes that are important in the metabolism of biogenic amine neurotransmitters such as epinephrine, norepinephrine, serotonin and dopamine (Wells & Awad, 1993). 2.1.2.1. Storage proteins: Ferritin and hemosiderin Excess iron not needed for immediate functional activity is stored in tissues in two related forms: ferritin and hemosiderin. Both forms are primarily located in the 7 REVIEW OF LITERATURE reticuloendothelial cells of the liver, bone marrow and spleen (Bothwell et al., 1979; Cook et al., 1992), and make up approximately 25% of the total body iron. Ferritin is the primary storage form for iron, and its concentration in serum reflects total iron body stores. The small quantities of ferritin present in serum are not involved in iron transport (Cook et al., 1992). Hemosiderin makes up the other half of storage iron in the liver, and is a product of various stages of degradation of ferritin (Cook et al., 1992). Stored iron can be mobilized and distributed via plasma transferrin to wherever it is required in the body, with some direct transfer occurring to a limited extent. Generally, the total amount of storage iron can vary over a wide range and may be almost entirely depleted before iron-deficiency anaemia develops (Bothwell et al., 1979; Dallman, 1989). 2.1.2.2. Transport protein: Transferrin Transferrin is the major iron transport protein and accounts for approximately 0.1 % of total body iron; iron is transported in plasma bound to this protein (Cook et al., 1974). Transferrin is a B-1 globulin with two binding sites for ferric iron. It can, therefore, be present as a di-ferric or mono-ferric molecule, or as apotransferrin with no iron (Finch & Huebers, 1982; Massey, 1992). The transferrin molecule has a high affinity for ferric iron, but does not bind ferrous iron. At any one time, approximately 11 % of all transferrin proteins are saturated with ferric iron at both sites in a normal physiological state (Wells & Award, 1993). Transferrin transports a large proportion of the iron to immature red blood cells where it binds to specific surface receptors; iron taken up can then be used for heme synthesis (Conrad & Barton, 1981; Dallman et al., 1993; Wells & Awad, 1993). The mechanism by which iron is transported from the cell surface to heme synthesizing mitochondria is still unknown. 8 REVIEW OF LITERATURE 2.1.2.3. Transport protein: Lactoferrin Lactoferrin is an iron transport protein present in high concentration in human milk. It is similar to transferrin in having 2 binding sites for iron. The iron content of lactoferrin may vary, but the protein is never saturated, and it is believed that it may be important in protecting the newborn from microbial gastrointestinal infections (Wells & Awad, 1993). Microbial organisms need iron to function and replicate, but since lactoferrin is never saturated, there is rapid binding of free iron. Consequently, microbial growth is inhibited as the iron, needed by the micro-organism to function, is limited (Wells & Awad, 1993). It has also been suggested that lactoferrin may play a role in the transportation of iron to the intestinal receptor sites in the infant, therefore, increasing absorption of iron from human milk (Rosa & Turgo, 1994). However, results from a recent study by Davidson et al. (1994) do not concur. These researchers used a randomized cross-over design to measure iron absorption in 8 term breast-fed infants (2-10 months of age) fed breast-milk (with its native content of lactoferrin) or the same milk with lactoferrin removed by treating with heparin-sepharose. The breast-milk was labelled with 5 8Fe. Iron absorption was measured by estimating the amount of isotope incorporated into the red blood cells 14 days after feeding the milk. The researchers found no enhancing effect of lactoferrin on the absorption of iron. The mean fractional iron absorption from breast-milk with lactoferrin was 11.8%, and this was significantly lower (p<0.05) than the absorption of 19.8% iron from lactoferrin-free breast-milk. The findings indicate an inhibitory effect of lactoferrin on iron absorption. Earlier reports in the literature on the effect of lactoferrin on the absorption of iron from breast-milk have ranged from 49-81% (Garry et al., 1981; Saarinen & Siimes, 1979; Saarienen et al., 1977). Differences in reported values of iron absorption have been attributed to inconsistency in the methods used 9 REVIEW OF LITERATURE by different researchers. Hence the role of lactoferrin in human milk is still unclear. 2.2. ABSORPTION AND REGULATION OF BODY IRON The metabolism of iron operates largely like a closed system, in which the iron stores are reused by the body, iron losses are minimal (< 1 mg/day) and absorption is minimized (Chaney, 1993). The absorption of iron is a complex process that starts with the intake of food and ends with the entrance of iron into the plasma (Charlton & Bothwell, 1983). The amount of iron absorbed from the diet will depend on; 1) the amount and form of iron present in the food, and the interaction of the iron from each food with other components of the meal in the intestinal environment, and 2) the homeostatic regulation of iron absorption by the intestinal mucosa (Massey, 1992). There are two forms of dietary iron to consider; heme and non-heme iron (Charlton & Bothwell, 1983; Dallman, 1989; Massey, 1992). The active site of absorption for both heme and non-heme iron is the duodenum and the upper jejunum, with absorption in the distal part of the small intestine being of less significance. However, the two forms differ in their mechanism of absorption and bioavailability (Charlton & Bothwell, 1983). 2.2.1. Non-heme iron Ninety percent of the iron in food occurs primarily as non-heme iron in the ferric form bound to organic acids (Chaney, 1993; Massey, 1992). Non-heme iron is present in legumes, eggs and wheat, with very small amounts of iron present in most fruits and vegetables (Massey, 1992). The absorption of non-heme iron varies widely and depends on how soluble the iron becomes in the intestine. This in turn is influenced by the composition of other foods 10 REVIEW OF LITERATURE eaten with the meal (Hallberg, 1981; Charlton & Both well, 1983). Foods such as coffee, tea and phosphates inhibit the absorption of non-heme iron (Charlton & Both well, 1983). The low pH of the stomach, on the other hand, enhances the absorption of non-heme iron by reducing ferric iron to the more readily absorbed ferrous iron. This process is further enhanced by the presence of ascorbic acid which results in the dissociation of iron from the ligands making it more available for absorption (Wells & Awad, 1993). In addition, meat, fish, poultry, some amino acids such as cysteine, and other components of the meal such as fructose and citrate also enhance the absorption of non-heme iron by forming readily absorbable complexes with it (Charlton & Both well, 1983; Dallman et al., 1993; Hurrell, 1984; Oski, 1993). Non-heme iron enters the mucosal cell as the free metal where it is then transferred to plasma by the transport protein transferrin (Massey, 1992). The exact mechanisms that allow the iron to enter the mucosal cell and regulate its transfer into the plasma are still poorly understood (Charlton & Both well, 1983; Massey, 1992). 2.2.2. Heme iron Heme iron, found in hemoglobin and myoglobin, is supplied from the diet mainly by meat and usually accounts for no more than 10% to 15% of dietary iron (Dallman, 1989). In comparison to non-heme iron, heme iron is well absorbed with less influence by many of the substances which promote (e.g. hydrochloric acid and ascorbic acid), or inhibit (e.g. tea, coffee, and phosphates) absorption of non-heme iron (Charlton & Both well, 1983). The main reason for this is that heme iron is absorbed into the intestinal mucosal cells as the intact metalloporphyrin (i.e. hemoglobin), whereas non-heme iron must first be solubilized before absorption (Charlton & Both well, 1983; Massey, 1992). In the mucosal cell, the iron from 11 REVIEW OF LITERATURE hemoglobin (i.e. heme) is split off the porphyrin ring and joins the soluble pool of absorbed non-heme iron (Charlton & Bothwell, 1983; Wells & Awad, 1993). The transport of iron derived from heme iron is similar to that of non-heme iron. 2.2.3. Intestinal regulation of iron absorption The amount of iron entering the mucosal cell and passing into the body is regulated so as to maintain normal body iron content; total body iron stores and erythropoietic activity are involved in the regulation of iron absorption (Massey, 1992). The entry of iron into the body is regulated at the mucosal cells of the small intestine, but the mechanism for this regulation is still unknown (Peters et al. 1988; Wells & Award, 1993). Nevertheless, one mechanism that has been proposed to regulate the transfer of iron is the synthesis of apoferritin by the mucosal cell. Increased amounts of apoferritin are synthesized in iron overload, trapping iron inside the mucosal cell and preventing its transfer to the capillary bed. The contents of the cells are extruded to the intestinal lumen, as the cells turn over, without absorption occurring (Peters et al., 1988). During iron-deficiency anaemia, very little apoferritin is synthesized to avoid competition with the transfer of iron to the deficient host. Because of this, the amount of iron transferred across the mucosal cell is inversely related to the individual's iron stores and erythropoietic activity of the body (Massey, 1992). If iron stores are low, the intestinal mucosa readily takes up iron and the proportion of iron absorbed from the diet is increased. Conversely, high iron stores are associated with a reduction in the amount of iron absorbed (Peters et al., 1988). 12 REVIEW OF LITERATURE 2.3. IRON STATUS IN INFANCY The main factors determining iron requirement in infants are the iron endowment at birth, the requirements for growth and the need to replace normal losses. Dallman et al. (1989) estimated that dietary iron must provide 30% of the needs for hemoglobin iron synthesis in a 1 year old 10 kg infant compared to only 5% in the adult male. As a result, this imposes a disproportionately higher requirement for iron in the infant, leading to a greater dependency on dietary sources of iron for the maintenance of erythropoiesis. 2.3.1. Maternal-fetal relationships Before birth the fetus receives its iron from the maternal circulation; iron is transported against a concentration gradient, rapidly and unidirectionally across the placenta from mother to fetus (Oski, 1989). The transport protein transferrin transports maternal iron to the placenta where it is released to transferrin receptors present on the placental microvillus membrane. The maternal transferrin does not cross the placenta, but returns to the maternal circulation (Gitlin et al., 1964). The iron is then transported through the placenta where it becomes associated with fetal transferrin in the fetal circulation (Oski, 1989). There is little evidence to show that the presence of maternal iron-deficiency, unless unusually severe, compromises the iron endowment of the fetus (Singla et al., 1978). Singla et al. (1978) found that the hemoglobin concentration in the cord blood of infants born to iron deficient anaemic mothers did not differ from that of infants born to iron-sufficient mothers until the maternal hemoglobin fell to below 60 g/L. Studies have also found little or no difference between serum ferritin concentrations at birth of infants born to mothers with high and low serum ferritin values (Rios et al., 1975a). These studies suggest that, except in the 13 REVIEW OF LITERATURE most unusual circumstances, maternal iron-deficiency by itself does not result in iron-deficiency in the infant at birth. Instead, factors such as birth weight, perinatal blood loss, an increase or decrease in the hemoglobin mass at birth (a result of early or late cord clamping), and the occurrence of fetal to maternal hemorrhage are more important determinants of the infant's iron endowment (Oski, 1989). 2.3.2. Iron status of the infant after birth The term infant is born with a body iron content of approximately 75 mg/kg body weight. Seventy-five percent of this iron is in the form of hemoglobin in circulating red blood cells; iron stores contribute the other 25% (Dallman et al., 1980). After birth, there are a number of changes that occur in iron metabolism and the rate of erythropoiesis (Dallman et al., 1989). During the first 2 months after birth, there is a linear decrease in hemoglobin that is termed the "physiological anaemia" of infancy. This decline is attributed to an abrupt decrease in erythropoiesis as a result of increased postnatal delivery of oxygen to tissues (Dallman et al., 1980). As the hemoglobin values fall from about 165 g/L to 110 g/L, the iron from the catabolized hemoglobin accumulates in the reticuloendothelial stores, which is reflected by a sharp increase in the level of iron stores (serum ferritin) in the first month after birth. During this time, the absorption of dietary iron is low. This may be attributable to the inhibitory effect of large iron stores on absorption. The iron from the catabolized hemoglobin, together with the iron stores is sufficient to meet the infant's iron needs (Fairweather-Tait, 1992; Dallman et al., 1980). Red cell synthesis resumes after the second month of life and the movement of iron into the iron stores (i.e. ferritin) is reversed. The storage iron now becomes utilized to support the increasing red cell mass of the rapidly growing infant, which 14 REVIEW OF LITERATURE is reflected by a decrease in the serum ferritin, and a rise in the hemoglobin level to about 125 g/L (Dallman et al., 1980). At about six months of age, the transplacental endowment of iron acquired during gestation becomes exhausted and the infant becomes dependent on dietary iron for maintenance of normal erythropoiesis. Consequently, any insufficiency in iron absorption from the diet during the second six months of life will result in iron depletion marked by a concomitant decrease in body iron stores (serum ferritin). If inadequate dietary intake continues, iron-deficiency anaemia will eventually develop. 2.3.3. Iron requirements of infancy Iron requirements during infancy have 3 components. First, is the amount of iron that must be absorbed from the diet to make up the relatively small but finite iron losses. These iron losses average about 20/yg/kg/day in a normal infant (Oski, 1993). Extrusion of cells from the intestinal mucosa accounts for approximately two thirds of the iron lost; the other one third is in cells desquamated from the skin and urinary tract. The second component of iron requirements in infancy is imposed by growth; the infant's growth is at its maximum during the first year of life, with a tripling in birth weight (Stekel, 1984). Thirdly, iron is needed to assure maintenance of iron stores. The Committee on Nutrition of the American Academy of Paediatrics (1976) have recommended dietary iron intakes of 1 mg/kg/day for term infants, beginning no later than 4 months of age and continuing until 3 years of age. This requirement indicates a need to absorb and incorporate about 0.7 mg of iron per day. 15 REVIEW OF LITERATURE 2.4. ASSESSMENT OF IRON STATUS There are three iron compartments that reflect the iron status of an individual: 1) storage iron, 2) transport iron, and 3) erythroid iron (Cook et al., 1992). Storage iron makes up about 25% of body iron. It is consists of two compounds, ferritin and hemosiderin, which are primarily found in the reticuloendothelial cells of the liver and bone marrow (Cook et al., 1992; Gibson, 1990). Serum ferritin values correlate significantly with total body iron stores (Cook et al., 1974). The transport protein transferrin accounts for about 0.1 % of total body iron. The erythroid iron compartment includes the iron in hemoglobin (70% of total iron), iron in myoglobin (4%) and the iron present in enzymes such as catalases, peroxidases and cytochromes (< 1 %). 2.4.1. Stages of development of iron-deficiency The development of iron-deficiency can be divided into three stages (Gibson, 1990). The first stage is characterized by a decrease in iron stores and is reflected by a fall in serum ferritin levels. Although the depletion of storage iron increases the risk of developing iron-deficiency, it by itself is not associated with any known functional or physiological abnormalities (Cook et al., 1992). At this stage, transport iron and hemoglobin levels are normal. In the second stage, termed "iron deficient erythropoiesis", the iron stores are depleted. As a result, there is a loss of transport iron which is indicated by a decrease in serum iron levels. This leads to a reduced supply of iron to the erythropoietic ceils and a decrease in the saturation of the iron transport protein, transferrin. The hemoglobin levels are, however, still within normal range. The third and last stage iron-deficiency occurs when the supply of transport iron decreases enough to restrict the synthesis of hemoglobin. 16 REVIEW OF LITERATURE Consequently, it is not until this final stage of iron-deficiency that hemoglobin levels begin to decrease. This stage is characterized by gradual development of detectable anaemia (exhausted iron stores and decreased levels of circulating iron) and red cell microcytosis (low mean cell volume) and hyperchromia (low mean cell hemoglobin): hence the name microcytic hypochromic anaemia. In summary, the development of iron-deficiency proceeds in a well-defined sequence that allows for its diagnosis. Depletion of iron stores is followed by a decrease in serum transferrin saturation which ultimately results in the decreased production of hemoglobin, clinically recognized as anaemia (Figure 2.1). 17 REVIEW OF LITERATURE First stage Loss of storage iron Iron Depletion 4 serum ferritin Second stage Loss of circulating iron Iron Deficient Erythropoiesis Third stage Decreased Hemoglobin production Iron-deficiency Anaemia I transport iron t total iron binding capacity > 4 hemoglobin 1 mean cell volume 4 mean cell hemoglobin t zinc erythrocyte protoporphyrin Figure 2.1. Stages of development of iron-deficiency 18 REVIEW OF LITERATURE 2.4.2. Laboratory measures of iron status among humans In order to provide the best assessment of iron status, a number of indices are often evaluated simultaneously. This is because many of the tests lack specificity and thus,no single laboratory measure can adequately characterize an individual's iron status. Serum ferritin concentration, transferrin saturation, erythrocyte protoporphyrin, and mean cell volume are the tests most widely used, in addition to hemoglobin and hematocrit. The laboratory tests used in the diagnosis of iron-deficiency may be grouped by body iron compartment, reflecting different aspects of body iron metabolism. Serum ferritin is most commonly used to evaluate iron stores; Serum iron, total iron binding capacity and transferrin saturation evaluate the plasma iron compartment and hemoglobin, hematocrit, red cell indices and erythrocyte protoporphyrin evaluate the red cell compartment (Dallman et al., 1993). Figure 2.1 includes a summary of the sequence of changes in some of the laboratory parameters used in the assessment of an individual's iron status. 2.4.2.1. Biochemical tests 2.4.2.1.1. Serum ferritin (iron stores compartment) Serum ferritin provides an index of the body iron stores (Dallman, 1990). It is the only measure of iron status that can show a deficient, excess or normal iron status (Gibson, 1990). During iron-deficiency, the serum ferritin levels decrease before any deficits occur in the iron transport and erythroid compartments. This characterizes the first stage in the development of iron-deficiency (Figure 2.1). A serum ferritin value of less than 10 //g/L is considered indicative of depleted iron stores in infants and children at all ages (Dallman et al., 1993). Below this value, the measurement of serum ferritin provides no indication of the severity of 19 REVIEW OF LITERATURE iron-deficiency (Cook et al., 1992). Serum ferritin values, however, must be interpreted with caution as values may be in the normal range despite the presence of iron-deficiency, especially in association with infection or inflammatory disease (Cook et al., 1992; Dallman et al., 1993). 2.4.2.1.2. Serum iron; total iron binding capacity and percent transferrin saturation (plasma compartment) Once iron stores are depleted, a further decline in body iron is characterized by a decrease in serum iron (Gibson, 1990) (Figure 2). As a result, the iron supply to the erythropoietic cell is gradually reduced, with a concomitant decrease in the saturation of the transport protein transferrin. The serum iron is a measure of the number of iron atoms bound to the iron transport protein, transferrin. The total number of available iron binding sites on the transport protein are determined as total iron binding capacity (TIBC), and this is increased in iron-deficiency. Transferrin saturation (TS) provides a measure of the iron supply to the erythroid bone marrow and is calculated from the ratio of the serum iron to the TIBC expressed as a percent (Cook et al., 1992; Gibson, 1990). The specificity of a decreased TS is limited since iron deficient erythropoiesis is present in disorders other than iron-deficiency, such as inflammation or malignant diseases. However, TIBC increases in iron-deficiency and remains low with inflammation. This allows for a distinction between the low transferrin saturation of chronic disease and that of true iron-deficiency (Gibson, 1990). The measurement of transferrin saturation also allows for differentiation of thalassaemia, which also produces a microcytic hypochromic anaemia like iron-deficiency. TS, however, is increased or normal in thalassaemia and decreased in iron-deficiency (Gibson, 1990). 20 REVIEW OF LITERATURE 2.4.2.2. Hematological tests 2.4.2.2.1. Hemoglobin (Hgb) Hemoglobin is the oxygen-carrying pigment of red blood cells in which iron is an essential component. Low hemoglobin levels become apparent only in the last and final stage of development of iron-deficiency. Thus, low measures of hemoglobin are associated with overt iron-deficiency anaemia. However, the specificity of hemoglobin as a measure of iron-deficiency is limited because its concentration is also decreased in chronic infections, inflammation, in vitamin B-12 and folate deficiencies and in thalassaemia (Gibson, 1990). 2.4.2.2.2. Red cell indices The red cell indices provide useful information for the diagnosis of different types of anaemia, and their reliability has been enhanced in recent years by the development of electronic cell counters (Cook et al, 1992., Gibson, 1990). The electronic methodology has greater advantage over the manual techniques previously employed in that it is more precise, easily reproducible and has the ability to complete a large number of measurements quickly. The red cell indices include the following: i) Mean cell volume (MCV) The MCV provides a measure of the size of the red blood cells, and is a reliable index of decreased hemoglobin synthesis (Gibson, 1990). Its major limitation is the length of time necessary for the levels to become abnormal after the onset of iron-deficiency anaemia. This is because the life span of circulating red blood cells is close to 4 months, so several weeks must go by before enough microcytic cells are released to alter the MCV measurement (Cook et al. 1992). The MCV is decreased in iron-deficiency. 21 REVIEW OF LITERATURE ii) Mean cell hemoglobin (MCH) The MCH gives a reflection of the hemoglobin content of each individual red cell. It is a ratio of hemoglobin (g/L) to red cell count (1012/L) (Gibson, 1990). In iron-deficiency anaemia the MCH is low, but it is high in the anaemia of vitamin B-12 and folate deficiencies. ii) Red cell distribution width (RDW) The RDW provides an index of the variation in volume of red cells and can be used to detect the presence of erythrocytes showing excessive variations in size (anisocytosis) (Oski, 1993). The RDW is determined from red cell histograms obtained from electronic tests of red cell morphology. In iron-deficiency, the red cells are microcytic and have greater variability in the extent of the size reduction. Therefore, there is an increased red cell distribution width, which is easily measured electronically. The RDW is also useful in distinguishing iron-deficiency from thalassaemia because an increased RDW does not occur in the latter (Oski, 1993). iv) Mean cell hemoglobin concentration (MCHC) The MCHC is used to determine the concentration of red blood cells in blood, and it is derived from the hemoglobin and hematocrit (volume fraction of packed red cells) (Gibson, 1990). The MCHC is low in cases of iron-deficiency anaemia but it is normal in anaemia related to vitamin B-12 and folic acid deficiencies, and the anaemia of chronic disease. The MCHC is the last of the red cell indices to become abnormal during the development of iron-deficiency (Dallman et al., 1993). 22 REVIEW OF LITERATURE 2.4.2.2.3. Zinc erythrocyte protoporphyrin (ZPP) Protoporphyrin is a heme precursor that normally occurs in small amounts in erythrocytes (Gibson, 1990). However, in the second stage of iron-deficiency, depleted iron stores result in the accumulation of protoporphyrin IX, a complex which would otherwise combine with iron to form hemoglobin, within the developing red cells. Lamola and Yamane (1974) found that the excess protoporphyrin does not exist free, but instead combines with zinc in the last step of heme synthesis to form zinc protoporphyrin (ZPP). Therefore, ZPP is increased in erythrocytes that develop under conditions of inadequate iron supply. Lamola and Yamane (1974) also showed that ZPP is fluorescent, such that cells with increased ZPP fluoresce when exposed to ultra-violet light of the correct wave length. ZPP is measured using a hematofluorometer (Blumberg et al, 1977), a specialized instrument designed to measure the reflected fluorescence from ZPP in only a small drop of whole blood. An advantage of the instrument is that reliable measures are obtainable with minimum laboratory training and it can be used in the field. A disadvantage is that elevated blood lead levels also result in increased ZPP levels (i.e. lead levels higher than 35//g/dl). This should be taken into account when interpreting elevated ZPP results. Elevated levels of ZPP are also seen in anaemia of chronic disease, such as chronic infection or malignancy, in which blockage of iron release from the reticuloendothelial cells occurs (Trundle, 1984). In spite of this, ZPP measurements provide a means of diagnosing subclinical changes in the biosynthetic pathway of heme due to iron-deficiency before anaemia develops. Table 2.1 provides a summary of changes in laboratory parameters of iron status during iron-deficiency anaemia. 23 REVIEW OF LITERATURE Table 2.1. Summary of changes in laboratory parameters of iron status during iron-deficiency anaemia Test Result with iron-deficiency anaemia Serum ferritin decreased (4) % TS decreased (4) ZPP increased (t) TIBC increased (t) Hemoglobin decreased (*) MCV decreased (4) MCH decreased (4) MCHC decreased (4) RDW increased (t) Many of these tests lack specificity, therefore, several tests are used in combination to determine an individual's iron status. 24 REVIEW OF LITERATURE 2.4.2.3. Diagnosis of iron-deficiency Table 2.2 shows the cut off values for common laboratory tests for the diagnosis of iron-deficiency anaemia in children. Generally, where the hemoglobin is less than 100 g/L, the diagnosis of severe iron-deficiency anaemia is considered straightforward (Dallman et al., 1993). If iron-deficiency is suspected on the basis of anaemia, the use of one or more additional tests (i.e. decreased serum ferritin and MCV, an increased TIBC or ZPP) can help substantiate the diagnosis. The schematic diagram showing the development of iron-deficiency anaemia (Figure 2.1) is over-simplified. In reality, as discussed, a number of other conditions, other than iron-deficiency can influence the results of each test producing false negative and false positive results. The results may, therefore, not agree with one another (Yip, 1990). For example, serum ferritin should be decreased in all cases of iron responsive anaemia, but this is not always the case (Dallman et al., 1993). Because of this, several of the tests are used in combination to improve the accuracy of diagnosing iron-deficiency. However, two important points emphasized in the literature are that 1) the absence of anaemia does not exclude the presence of iron-deficiency, and 2) the presence of anaemia indicates iron-deficiency is relatively severe (Dallman et al., 1993; Yip, 1990) . 25 REVIEW OF LITERATURE Table 2.2. Common laboratory tests and cutoff values used in the diagnosis of iron-deficiency in young children (Oski, 1993). Test Age (Years) Cutoff value Biochemical Serum iron Total iron binding capacity Transferrin Saturation Erythrocyte protoporphyrin Serum ferritin 1-2 1-3 1-2 1-5 1-5 < 5.4 //mol/L > 86 //mol/L <8% > 90 //mol/mol of heme 8 to 12 //g/L Hematological Hemoglobin MCV MCH MCHC R D W 1 1-2 1-2 1-2 1-2 1-5 NHANES II <107 g/L <67 fl <22 pg <320 g/L >14.5% AAP < 110 g/L <70 fl * NHANES denotes the Second National Health and Nutrition Examination Survey, and AAP the Committee on Nutrition of the American Academy of Pediatrics. 1 Not included in the NHANES II or AAP 26 REVIEW OF LITERATURE 2.5. IRON AND DEVELOPMENT The prevalence of iron-deficiency anaemia seems to be highest in infants between 6 months and 24 months of age (Lozoff, 1988). This is a time which coincides with the later part of the postnatal brain growth spurt when many mental and motor processes are still maturing. This period of rapid brain growth continues through to 24 months of age. Indeed, the uptake of iron by the brain is at its highest during this time (Beard et al., 1993). Hence, an inadequate supply of iron at this time may have deleterious effects on the brain's overall development. The biochemical explanation for this has been suggested to be related to the role of iron in synaptogenesis, that is, the biochemical and functional maturation of the brain and not the role of iron in hemoglobin (Ben-Shachar et al., 1986; Yehuda & Youdim, 1989; Youdim et al., 1980; Youdim et al., 1989). From their studies on rats, Yodium et al. (as reviewed by Yehuda and Youdim, 1988) suggest that many of the behavioral and cognitive deficits seen in iron-deficiency may be due to altered dopamine metabolism. Dopaminergic neurotransmission is potentially linked to iron metabolism, as the key enzyme in dopamine synthesis, tyrosine-3-monooxygenase, requires iron. 2.5.1. Location of brain iron The basal ganglia contains the highest level of iron in the brain, but iron is found throughout the brain with the white matter also having a high concentration of iron. The mechanism by which iron is transported within the brain is not well known. Iron is transferred across the blood brain barrier bound to the transport protein transferrin. The transferrin concentrations are higher in the white matter than the grey matter of the cerebral cortex in humans (Beard et al., 1993). 27 REVIEW OF LITERATURE 2.5.2. Function of brain iron Iron plays a role in the function and synthesis of dopamine, serotonin, catecholamines, gamma-aminobutyric acid (GABA), and myelin formation (Beard et al., 1993). Intraneuronal iron is incorporated into enzymes that are involved in 1) redox or electron transport, 2) synthesis and packaging of neurotransmitters, and 3) uptake and degradation of the neurotransmitters (Beard et al., 1993). Storage iron (ferritin) present in oligodendrocytes and microglia cells is involved in the synthesis and maintenance of myelin (Beard et al., 1993). 2.4.3. Functional abnormalities associated with iron-deficiency Some of the major consequences of iron-deficiency include defects in cognitive function, a reduction in work capacity, reduced tolerance to exercise, growth retardation, compromised immunity, and impaired thermoregulation and premature delivery. Among the most important deficits associated with iron-deficiency are impairment in psychomotor development and cognitive function (Fairweather-Tait, 1992). 2.5.4. Association between iron status and cognitive development 2.5.4.1. Evidence from animal studies Animal studies have indicated that iron-deficiency during early development has more lasting effects than a similar deficiency during adulthood (Ben-Shachar et al., 1986; Yehuda et al., 1989; Youdim et al., 1980; Youdim et al., 1989). The biochemical mechanisms linking iron-deficiency anaemia to developmental function comes from studies done by Youdim et al. (1980, 1989) and Yehuda et al. (1986) in experimental animals (i.e. rats). Neurotransmitter changes, decreased brain iron content and behavioral alterations due to iron-deficiency 28 REVIEW OF LITERATURE anaemia seemed to persist despite treatment when the deficiency occurred in the infant rat but not in the adult rat (Ben-Shachar et al., 1986). In addition, the number of dopamine D-2 receptors in the brain of rats subjected to a transient period of iron-deficiency during infancy have been shown to decrease (Ben-Shachar et al., 1986). These changes were not restored with an iron supplement regimen. 2.5.4.2. Evidence from human studies The establishment of a link between iron-deficiency anaemia in infancy and behavioral alterations has entailed the comparison of iron deficient and normal infants. In addition, because the lack of iron can range from depleted iron stores to iron-deficiency anaemia, the degree of iron-deficiency has been an important dimension in the studies (Lozoff, 1988). Since the original report by Oski and Hoing (1978), which showed that iron-deficiency anaemia in infants was associated with subtle behavioral abnormalities, research interests have focused on either infants between 9 and 24 months of age or school children between 9 and 12 years of age. To date, several studies have compared the scores on developmental tests between infants With iron-deficiency severe enough to cause anaemia and non-anaemic controls. Most of these studies have used the Bayley Scales of Infant Development, a standardized test of infant development with three components: a Mental Scale, a Motor Scale, and an Infant Behaviour Record (Lozoff, 1988). A study in Guatemala of 6-24 month old infants found that the mean mental developmental test score for the 28 infants with iron-deficiency anaemia (hemoglobin (Hgb) < 105 g/L) was 87 compared to a mean score of 100 for 40 non-anaemic infants (Hgb > 120 g/L) (Lozoff et al., 1982). The researchers found that the mental test score 29 REVIEW OF LITERATURE deficits were especially marked in older anaemic infants (19-24 months), and that the pattern in item failure suggested particular difficulty in verbal test items. A similar study in Chile found that the average mental test score for 10 iron deficient anaemic 15 month old infants was 98, 12 points less than the mean score of the 27 infants with Hgb levels >110 g/L (Walter et al., 1983). No differences in motor test scores were found. Similar results were obtained in a study in England of 145 Asian infants aged 21-23 months (Grindulis et ai., 1986). Iron-deficient infants (Hgb < 110 g/L) had significantly lower scores on a developmental screening test (The Sheidan Developmental Sequence) than infants with higher hemoglobin levels. In a second study in Chile, infants were followed prospectively from birth as part of a field trial of iron-fortified foods with random assignment to either an iron-fortified group or control group (Walter et al., 1989). The researchers found that the 39 infants with iron-deficiency anaemia (Hgb < 110 g/L) had significantly lower mental and motor test scores than either the non-anaemic iron-deficient infants (n = 127) or the iron-replete controls (n = 30). Similar results were found in a double-blind, randomized community study of 191 Costa Rican infants aged 12-23 months (Lozoff et al., 1987). In this study, the data from the anaemic infants were further analyzed with respect to actual hemoglobin levels, since lower hemoglobin levels indicate more severe iron-deficiency once anaemia is present. The mental test scores were evaluated in increments of 5 g/L for hemoglobin levels. The scores of infants with mild iron-deficiency anaemia (Hgb 101 -105 g/L) was similar to those of infants with Hgb 2=105 g/L. However, as a group, the infants with mild iron-deficiency anaemia had lower motor test scores but not lower mental test scores. Only those infants with moderate iron-deficiency anaemia (Hgb < 100 g/L) were found to have significantly lower mental and motor 30 REVIEW OF LITERATURE test scores. On the whole, infants with hemoglobin ^ 105 g/L had significantly lower motor test scores than those with hemoglobin of >105 g/L. In the studies of both Lozoff et al. (1987) and Walter et al. (1989) the infants with lesser degrees of iron-deficiency anaemia did not have impairment in developmental test performance. These findings support the hypothesis that the more severe the anaemia the greater the effect on developmental delays in infancy. Lozoff et al. (1991) and Walter et al. (1990) both found that iron-deficiency anaemia in early childhood may result in long-lasting changes in behavioral development that may not be reversed with iron therapy sufficient to correct the anaemia. A more recent study, however, found that supplementation of infants who had Hgb <105 g/L (iron-deficiency anaemia) at 12 months to 18 months of age, with ferrous sulphate (3 mg/kg/day) for 4 months was effective in reversing anaemia and the low developmental test scores (Idjriadinata & Pollitt, 1993). Before intervention, the iron-deficient anaemic infants had significantly lower mental and motor test scores than non-anaemic iron-deficient infants (Hgb a: 120 g/L, serum ferritin ^12 //g/L) and iron sufficient infants (Hgb a 120 g/L, serum ferritin >12 //g/L). Intervention in this study was done much sooner than in most of the other studies. This suggests that if identified early and if appropriate intervention is given, the long term consequences involving reduced school performance can be avoided. In summary, the results of research published to date document an association between iron-deficiency anaemia and impaired performance on developmental tests in infancy. Although the number of studies is limited, similar results have been obtained by different researchers working in different countries with different populations. This provides strong circumstantial evidence that iron-deficiency anaemia affects mental and motor development. In addition, although the detrimental effects of iron-deficiency anaemia on motor and mental 31 REVIEW OF LITERATURE development have been shown in infants, the same has not been shown for infants with only iron depletion (i.e. low serum ferritin values of < 1 0 //g/L). 32 REVIEW OF LITERATURE 2.5.5. Cognitive assessment 2.5.5.1. The visual recognition memory test (Fagan Test of Infant Intelligence) The Fagan Test of Infant Intelligence (FTII) is a standardized test of visual recognition memory that has been developed for use with infants between the ages of 6 and 12 months. The Bayley Scales of Infant Development (BSID) was the choice of measure for testing cognitive and psychomotor development in infancy in most of the studies reviewed above. However, this test, when given to infants under 24 months of age has been shown to have low reliability for predicting intelligence in later life. This may be due to the fact that the test relies on sensorimotor functions and early language skills which have poor predictive value for later intelligence (Fagan et al., 1983). The BSID is made up of items that test the development of simple sensory and motor skill functions which have been argued as not relevant to later intelligence (Fagan & McGrath, 1981). Fagan and McGrath (1981) argue that tests to predict later intelligence must involve early behaviours that are similar to those known to be related to later intelligence. For example, on intelligence tests, older children may be asked to "discriminate among stimuli, to identify similarities, to retain new information and to define words" (Fagan & McGrath, 1981). Thus, even though it is not plausible to give an infant a vocabulary test, it is possible to ask an infant to exhibit tasks of discrimination, identification and retention (Fagan & McGrath, 1981). Researchers involved in the field of infant perception have discovered that measures of infant recognition memory and novelty preference are predictive of later intelligence (Thompson et al., 1991). They argue that perceptual tasks require processes involving memory, discrimination and attention, abilities that are intrinsic to successful performance on later IQ tests. Fagan (1984) notes that infants, after being shown a visual target, will shift their fixation more toward a new target. He suggests that this 33 REVIEW OF LITERATURE attention to the new target is indicative of perceptual tasks involving memory, discrimination, identification and attention - the same abilities that are intrinsic to the successful performance on later intelligence tests (Fagan, 1984; Fagan & McGrath, 1981; Thompson et al., 1991). Furthermore, early visual recognition scores do not seem to be influenced by factors such as race, birth order and parental education (Fagan, 1988). Thompson et al. (1991) administered the FTII to 113 full term infants at 5 and 7 months of age. The infants were then followed longitudinally. The following sequence of tests were administered: the BSID were given at 12 and 24 months; the Sequence Inventory of Communication Development, which provides scores for verbal activity, spatial ability and perceptual speed for memory was administered at 24 and 36 months; and the Stanford-Binet and Colorado Specific Cognitive abilities test was given at 36 months. Correlational analyses showed that novelty preference predicts language and memory, independent of intelligence. Their results suggest that novelty preference may be reflective of specific cognitive processes during the first year of life, in addition to predicting later intelligence. Fagan and Montie (1988) in a study of 128 high risk infants (i.e. high risk for mental retardation), found the correlation between early novelty preferences at 3 to 7 months and later IQ at 3 years of age was high and significant. The FTII was highly sensitive (80% to 88%) and specific (83% to 89%) in diagnosing mental retardation in this high risk group. When the FTII was compared with the BSID, the predictive validity values yielded by the FTII for retardation and for normality were high at 77% and 93%, respectively. In contrast, the use of the BSID resulted in the misclassification of individuals more often than in the correct classification of individuals; the predictive validity for retardation or normality for the BSID were only 33% and 50% respectively. It is important to note, however, when interpreting the results from Fagan 34 REVIEW OF LITERATURE and Montie's study, that the test was administered to a group of high risk infants who, due to their medical history, fell into one or more risk groups established for later mental retardation. 35 REVIEW OF LITERATURE 2 . 6 . IRON CONTENT IN DIETARY SOURCES OF IRON FOR THE INFANT Human milk, iron-fortified infant formulas and iron-fortified infant cereals represent the major postnatal source of dietary iron for much of the first year of life (Cook & Both well, 1984), in addition to meat, fish and poultry as the infant gets older. 2 . 6 . 1 . Human milk During the first month postpartum breast-milk contains about 0.5 -1 mg iron/L. This decreases to about 0.3 mg iron/L at 4-6 months (Fairweather-Tait, 1992; Oski, 1993). The iron content of breast-milk is believed to be virtually independent of the maternal iron status (Lonnerdal, 1986). Although the iron content of breast-milk is low, it has a high bioavailability. Studies show that 50% - 70% of the iron in breast-milk is absorbed by the infant (Saarinen & Siimes, 1979). Therefore, using the transplacental iron endowment and the iron absorbed from breast-milk, the breast-fed infant is able to maintain iron body stores and erythropoietic activity for at least the first six months of life (Lonnerdal, 1986; Duncan et al., 1985). The composition of the rest of the infant's diet, however, has a strong influence on the percentage of iron that is absorbed. For example, feeding solid foods near the time of a breast-feeding can greatly inhibit the absorption of the iron from the breast-milk (Oski & Landaw, 1980). 2 . 6 . 2 . Cows' milk and low iron infant formulas Cows' milk has a similar iron content to breast-milk, but the bioavailability of the iron is low. Low iron formulas in Canada contain about 1.5 mg iron/L - 3 mg iron/L (Table 1). The absorption of iron from low iron formulas and cows' milk has been reported to be about 10% (Saarinen & Siimes, 1979). An infant's diet which consists of whole cows' milk or low iron 36 REVIEW OF LITERATURE infant formula as the primary source of milk, particularly after six months of age, may not meet the dietary requirements for iron (RNI 7 mg iron/day for infants 5-12 months) unless other sources of iron are provided. The basis for the high bioavailability of iron in breast-milk in comparison to cows' milk is not known. Plausible explanations point to the relatively low calcium and protein and high lactoferrin content of human milk, when compared to cows' milk or infant formulas, in facilitating the absorption of iron (Fairweather-Tait, 1992). However, as noted earlier, recent studies by Davidson et al. (1994) indicate that lactoferrin does not play a role in the enhancement of iron absorption from human milk in infants. The mean absorption of iron from human milk was as low as 11.8% in the breast-fed infants compared to 19.8% absorption of iron from breast-milk in which lactoferrin had been removed through treatment with heparin-sepharose. It should be noted however, that the absorption of iron from human milk in Davidson et al.'s study is much lower than that of 50% - 70% iron absorption from breast-milk shown by Saarinen and Siimes (1979). The differences in percent iron absorption from breast-milk has been attributed (Lonnerdal & Iyer, 1995) to the inconsistency in the methodologies used by the different researchers. Another plausible explanation for the high bioavailability of iron in breast-milk compared to cows' milk is that the high casein present in cows' milk binds iron tightly thus, limiting its absorption. Furthermore, it is possible that the high calcium and phosphate, and low ascorbic acid in cows' milk also compromise the absorption of iron (Barton et al., 1983). 37 REVIEW OF LITERATURE 2.6.3. Fortified infant formulas Many of the infant formulas on the market are now fortified with additional iron in an attempt to improve iron nutrition for infants. Cow milk protein-based formulas, soy protein-based formulas and specialized formulas (such as protein hydrolysates (e.g. Nutramigen) for infants allergic to cow milk or soy protein) can be fortified to contain 7-13 mg of iron per litre (Table 2.3). The iron used for fortification is ferrous sulphate. Despite the low bioavailability of iron from these formulas, the absolute amount of iron absorbed is high. The fractional absorption of iron from the high iron-fortified infant formulas is inversely proportional to the iron concentration (McMillan et al., 1977). That is, the percentage of iron absorbed decreases as the concentration of iron increases. Because of this, infant formulas supplemented with 6 mg iron/L are as effective a source of iron as infant formulas with 12 mg iron/L. McMillan et al. (1977) found that about 6% of iron was absorbed from formula containing 6 mg iron/L and about 4% iron was absorbed from formula containing 12 mg iron/L. Although the percentage of iron absorbed from high iron formulas (7-13 mg iron/L) was much lower than that from low iron formulas or whole cows' milk, the absolute amount of iron absorbed was more than six times greater. The bioavailability of iron from soy-protein based formula, however, is lower than from cow milk protein-based formulas (Brennan et al., 1989). The optimal amount of iron needed in infant formulas is still not well understood. 38 REVIEW OF LITERATURE Table 2.3. Iron content of infant formulas on the market in the Greater Vancouver Area Brand name Iron content • (mg/L) Enfalac 3 Enfalac with iron 7 Similac 1.5 Similac with iron 12 SMA (ready to use & concentrate) 1.5 SMA with iron 12 Carnation good start 10 Carnation follow-up 13 Milupa Aptamil (birth) 2 Milupa Milumil (3 mth) < 14 Bonamil with iron 12 Nutramigen 13 Alimentum 12 Isomil 12 Prosobee 12 Nursoy 12 Soyalac 13 Alactamil 12 'Iron content (mg/L) from product labels of infant formulas on the market in Vancouver 39 REVIEW OF LITERATURE 2.6.4. Solid foods Generally, infants initially receive most of their dietary iron in the form of non-heme iron. They are then gradually introduced to an increasing number of heme sources of iron throughout the later part of infancy and early childhood. 2.6.4.1. Infant cereals Iron-fortified infant cereals are usually one of the first solid foods introduced into an infant's diet, and often become a large part of the solid food intake by 6 months of age (Cook & Bothwell, 1984). Most of the commercial infant cereals, at least in the USA, were fortified with iron in the form of iron pyrophosphate prior to the 1970's. This form of iron was found to have an extremely low bioavailability and, therefore, did not contribute much to the infant's iron nutrition (Rios et al., 1975b). Infant cereals marketed in Canada and the USA. are now fortified with an electrolytic iron powder to a level of 0.45 mg iron per gram of dry cereal, providing approximately 1 mg of iron per tablespoon of prepared cereal, of which about 4% will be absorbed (Fomon, 1987; Rios et al., 1975b). 2.6.4.2. Other solids Over the second six months of life, solid foods (e.g. fruits, vegetables meat, fish, poultry, legumes, egg yolk, tofu etc) are gradually introduced into the infant's diet. In the past, attention was directed to the iron content of different foods. Today, it is known that the percentage of iron absorbed from different foods varies over a wide range. Some iron-rich foods from which iron is well absorbed include meat, poultry and fish (heme iron). Foods that contain a good source of ascorbic acid, such as fruit juices (the most probable source of 40 REVIEW OF LITERATURE vitamin C in an infant's diet), will further enhance the absorption of iron from foods such as cereals and legumes (non-heme) from which the iron is less well absorbed (Cook & Bothwell, 1984). Of note, some iron-rich foods such as egg yolk and spinach that were at one time recommended as good sources of dietary iron have been found to contain iron in poorly available forms (Bothwell et al., 1979). 2.7. INFANT FEEDING PRACTICES AND IRON-DEFICIENCY ANAEMIA Feeding practices play an important role in determining the risk of developing iron-deficiency anaemia in infants and children. Infant feeding practices in Canada have changed considerably since the mid 1970's. In the early 1970's about 88% of infants in Canada were given solid food by 1 month - 2 months of age, and about 38% were given cows' milk by 6 months of age (Green-Finestone et al., 1989; Tanka et al., 1989). More mothers now breast-feed and do so for at least six months, and fewer introduce solid foods before four months, or use unmodified cows' milk before six months of age (Green-Finestone et al., 1989; Tanka et al., 1989). Nevertheless, iron-deficiency anaemia still seems to be a significant nutritional problem among Canadian infants. The age at which iron-deficiency anaemia has been found to be most prevalent in infants in Canada is between 9 and 14 months of age (Gary-Donald et al., 1990; Chan-Yip & Gary-Donald, 1987). Cook and Bothwell (1984) have given a useful description of the overlapping periods of the major sources of dietary iron in relation to age in infants in the United States (Figure 2.2). During the first 3 months after birth the infant's diet consists primarily of milk, and at about 3-4 months of age cereals are usually introduced into the infant's diet. A mixture of solid foods such as meat, poultry, fish, eggs and legumes are gradually introduced into the 41 REVIEW OF LITERATURE infant's diet after 6 months of age so that by the age of 1 year, the infant's diet begins to mimic that of other members of the household (Cook & Bothwell, 1984). 42 REVIEW OF LITERATURE S o u r c e s o f D i e t a r y Iron D u r i n g i n f a n c y Solid Foods Infant Cereals • Milk & Infant Formulas L 1 1 1 I L o 4 6 - 8 Age In months 10 .12 From Cook & Bothwell, pg 119 In Sketel A (ed) Iron Nutrition and Childhood. Raven Press, New York, 1984. Figure 2.2. A description of the overlapping periods of the major sources of dietary iron during infancy1. 1 The figure indicates major dietary sources of iron in infancy and not initiation or cessation of food. 43 REVIEW OF LITERATURE 2 . 7 . 1 . Breast-feeding Although breast-feeding is considered to be protective against iron-deficiency because of the high bioavailability of iron in breast-milk, prolonged breast-feeding without the introduction of iron containing cereals or other foods rich in iron is associated with an increased risk of iron-deficiency anaemia (Calvo et al., 1992; Pizarro et al., 1991; Siimes et al., 1984). In a study conducted in Santiago, Chile, Pizarro et al. (1991) found prevalences of iron-deficiency anaemia in 9 month old infants of 14.7% in breast-fed infants not given supplemental iron, 0.6% in infants given iron-fortified formula and 20.2% in infants fed cows' milk without supplemental iron. Similar results were found by Calvo et al. (1992) in a study conducted in Argentina. In this study, the prevalence of iron-deficiency anaemia of 9 month old infants who were exclusively breast-fed to 6 months of age was 27.8% and 7.1 % among infants who were fed an iron-fortified formula from birth. In addition, in the same group of infants exclusively breast-fed to 6 months of age, 27.8% had depleted iron stores compared to none of the formula-fed infants. Both groups of infants had received foods rich in iron and vitamin C, but no iron-fortified products such as iron-fortified infant cereal had been fed from 6 months of age. Results from a study by Siimes et al. (1984) are consistent with the findings of Calvo et al. and Pizarro et al. Siimes et al. (1984) found that 25% of exclusively breast-fed 6 month old infants had 3 out of 6 laboratory criteria for iron-deficiency. The findings from these studies suggest that infants who are breast-fed be given supplemental iron from about 6 months of life. This is consistent with the time at which the iron stores present at birth are likely to become depleted. 44 REVIEW OF LITERATURE 2.7.2. Cows' milk Although cows' milk contains approximately the same amount of iron as human milk (0.5 mg of iron/L), this iron is not readily absorbed (Saarinen & Siimes, 1979). The use of cows' milk in infants under 6 months of age has been associated with increased gastrointestinal bleeding (Fomon et al., 1981; Fuchs et al., 1993a; Ziegler et al., 1990). Data published to date (Fuchs et ai., 1993b; Penrod et al., 1990; Tunnessen & Oski, 1987) have shown that a significant number of infants who are fed cows' milk in the second 6 months of life are at risk of developing depleted iron stores, even with intakes of iron and vitamin C which meet or exceed the recommended daily allowance (Fuchs et al., 1993b). Three similar studies (Fuchs et al., 1993b; Penrod et al., 1990; Tunnessen & Oski, 1987) have compared the iron status of infants fed iron-fortified formula in the first 6 months of life and then fed whole cows' milk or iron-fortified formula in the second 6 months of life. Tunnessen and Oski (1987) found an incidence of iron-deficiency of 22% at 12 months of age among infants fed cow's milk (n= 69) as the primary source of milk from 6 months of age. Fuchs et al. (1993b) and Pendrod et al. (1990) had similar findings. In a cross-sectional, non-randomized study using a single 3 day food diary and a 24 hour record Penrod et al. (1990) found more than three times as many infants with low serum ferritin (< 12 //g/L) were infants who had been fed cows' milk than fed iron-fortified formula. The differences were statistically significant (p<0.05). Fuchs et al. (1993b) have reported data to show that feeding foods containing supplemental iron (e.g. iron-fortified infant cereal) does not prevent the development of low iron status in infants fed whole cows' milk as the primary source of milk. The researchers concluded that 6 to 12 month old infants fed cows' milk and iron-containing foods (i.e. iron-45 REVIEW OF LITERATURE fortified infant cereals) are at risk of developing depleted iron stores, although not iron-deficiency anaemia. These studies provide consistent evidence that the iron status of infants who are fed cows' milk is significantly lower than in infants fed iron-fortified formula even when iron-containing foods, such as iron-fortified infant cereals are included in the diet. 2.7.3. Iron-fortified infant cereals The bioavailability of the electrolytic iron used to fortify infant cereals is not well known and is controversial (Fomon,1987). Some studies, however, have shown iron-fortified cereals are efficacious in reducing the incidence of iron-deficiency anaemia (Walter et al., 1993). By the second 6 months after birth, a diet consisting only of milk is no longer nutritionally adequate and it is recommended that solid foods be introduced at 6 months of age to help maintain adequate iron stores. A recent study in Chile looked at the effectiveness of iron-fortified infant cereal in preventing iron-deficiency anaemia (Walter et al., 1993). The study used a rice cereal fortified with 55 mg of electrolytic iron/100 g of dry cereal, and an infant formula with 12 mg of ferrous sulphate/100 g of dry powder, similar to the infant cereals and formula currently marketed in North America. The infants in the study were all healthy term infants and were divided into five groups; fortified cereal, unfortified formula; unfortified cereal, unfortified formula; unfortified cereal, fortified formula; breast-fed (>4 months), fortified cereal; and breast-fed (>4 months), unfortified cereal. At 8 months of age, only 6% of infants fed fortified cereal with unfortified formula, and only 4% fed unfortified cereal with fortified formula, had iron-deficiency anaemia compared to 17% of the infants fed unfortified cereal and unfortified formula. Among the breast-fed groups, the prevalence of iron-deficiency 46 REVIEW OF LITERATURE anaemia was 15% in infants who had been fed the unfortified cereal compared to a prevalence of 3% among the breast-fed infants who had been fed the fortified cereal. In the study protocol, infants with a hemoglobin <105 g/L were removed from the study and treated with iron supplements. By 12 months of age, less infants from the groups fed fortified cereal or fortified formula had been removed from the study because of a hemoglobin of <105 g/L, than from the group fed unfortified formula and unfortified cereal. Among the breast-fed infants, fewer were withdrawn from the study from the group who had been fed fortified cereal than from the group who had been fed the unfortified cereal. The findings of this study show that the use of an iron-fortified infant cereal can play a major role in reducing the incidence of iron-deficiency in young infants. This study did not look at the effect of providing cows' milk with iron-fortified infant cereal. However, as discussed, feeding with iron-fortified infant cereals does not seem to prevent the development of low iron stores in infants over six months fed cows' milk (Fuchs et al., 1993b). 4 7 REVIEW OF LITERATURE 2.8. GUIDELINES FOR INFANT FEEDING The following recommendations for infant feeding have been made by The Canadian Paediatric Society Nutrition Committee, The American Academy of Paediatrics, and the City of Vancouver Public Health Department: 2.8.1. Canadian Paediatric Society Nutrition Committee, (1991). • Term infants who are exclusively breast-fed do not need supplemental iron until they are 6 months of age. If solid foods are introduced earlier they should contain an adequate amount of iron. After 6 months of age, breast-fed infants should receive extra iron in the form of iron-fortified infant cereals and other iron rich foods. These infants should be offered an iron-fortified infant formula after they have been weaned from breast milk. • Term infants who are not breast-fed should be given an iron-fortified formula from birth. Studies are still under way to determine the optimal iron content of these formulas. Until results are known, the use of currently available iron-fortified formulas seems appropriate. After 4 to 6 months of age, iron-fortified infant cereals provide a good additional source of iron. • Cows' milk should not be introduced until an adequate amount of solid food containing iron and vitamin C is included in the diet, preferably at 9 to 12 months of age. 48 REVIEW OF LITERATURE 2.8.2. American Academy of Paediatrics Committee on Nutrition, (1992). • Infants should be breast-fed for the first 6 months to 12 months after birth. The only acceptable alternative to breast-milk is an iron-fortified infant formula. • Consumption of breast-milk or iron-fortified formula, along with age appropriate solid foods and juices, are recommended during the first 12 months of life to allow for more balanced nutrition. • Whole cows' milk and low iron formula should not be used to feed infants under one year of age. • Appropriate solid foods should be introduced between 4 and 6 months of age. 2.8.3. The Vancouver Public Health Department; Position Statement on infant feeding, (1993). • Breast-feeding is recommended for the first year of life or longer. Healthy full-term breast-fed infants are not at risk for iron-deficiency during the first six months of life, as evidenced by current literature. The iron in breast milk is highly bioavailable to the infant and the breast-fed infant does not require an iron-fortified formula when given the occasional breast milk substitute. • Whole cows' milk or evaporated cows' milk formula are not appropriate feeding choices as a replacement for breast milk before nine to twelve months of age. 49 REVIEW OF LITERATURE • Healthy term infants who are not breast-fed should be given formula for nine to twelve months though they do not physiologically require an iron-fortified formula for the first three months of life. • Solids may be offered when the infant can sit with support and maintain good head control. Iron-fortified infant cereals followed by a variety of vegetables and fruits should start to be offered between 4-6 months. At approximately 6 to 9 months, iron rich foods such as beef, chicken, pork, lamb, fish and beans, peas and lentils should be offered. The non breast-fed infant who does not receive a variety of the above solids should be given an iron-fortified formula until twelve months of age. 2.9. FAMILY INCOME, ETHNICITY AND IRON STATUS 2.9.1. Family Income In contrast to Canada, the prevalence of iron-deficiency anaemia among infants and preschool children has declined steadily in the United States over the past 20 years (Yip et al., 1987a; Yip et al, 1987b). This decline has occurred in infants and young children from both middle and low income families. For example, the prevalence of iron-deficiency anaemia in middle and low income families in the United States, as published in 1987, was 2.8% and 2.9%, respectively. The Special Supplement Food Program for Women, Infants and Children (WIC) in the United States has been credited for this decline (Miller et al., 1985; Vasquez-Seoane et al., 1985). Three quarters of infants in the United States who are fed formula are fed an iron-fortified formula, which is also the formula provided by the WIC program (Miller et al., 1985; Vasquez-Seoane et al., 1985). Canada has no program equivalent to the WIC 50 REVIEW OF LITERATURE program. Studies done in Canada have found an association between low income and the risk for iron-deficiency anaemia. A study in Montreal reported that the prevalence of iron-deficiency anaemia among infants 10 to 14 months of age from low-income families was 24.3% (Gary-Donald et al., 1990). Another study in Ottawa similarly found an inverse relationship between socioeconomic status and iron-deficiency anaemia (Feldman et al., 1985). In the latter study, the prevalence of iron-deficiency anaemia among 6 to 18 months old infants (n = 320) was 8.8% for those in the low socioeconomic group compared to 1.7% and 2.7% for those in the high and middle socioeconomic groups, respectively. These two studies suggest that infants from low-income families are at higher risk for developing iron-deficiency anaemia than infants from middle or high income families. 2.9.2. Ethnic background and prevalence of iron-deficiency anaemia Very few studies in Canada have looked at the prevalence of iron-deficiency anaemia in relation to ethnic background. A study in Montreal found a prevalence of 12.1% of iron-deficiency anaemia among Chinese infants of 6 to 36 months of age (Chan-Yip et al., 1987). Studies done in England found that iron-deficiency anaemia was more common among infants and children of some ethnic minorities, particularly among infants of Asian parentage than White infants (Aukett et al., 1986; Ehrhardt, 1986; Duggan et al., 1991). Studies in the United States have reported a higher prevalence of iron-deficiency anaemia among Black than White infants (Yip et al., 1987 a&b). 51 DESIGN AND METHODS Chapter III 3. DESIGN AND METHODS The study protocol and procedures were approved by the University of British Columbia Screening Committee for Research Involving Human Subjects. 3.1. STUDY DESIGN This was a cross-sectional study of iron status in infants at 39 ± 1 week (9 months) of age in relation to infant feeding history and family economic and ethnic background. 3.2. SUBJECTS 3.2.1. Subject identification and selection criteria A sample of infants was systematically identified for participation in the study using birth and death lists provided by the City of Vancouver Public Health Department. Selection criteria were that the infant was full-term (gestational age ^37-<42 weeks), with a birth-weight 2500-4500 gms, born between January 1st, 1993 and March 2nd, 1993 or between June 4th, 1993 and August 7th, 1993, to parents resident in Vancouver, with an address to enable contact. 3.2.2. Subject recruitment All eligible full-term infants were assigned a 4 digit identification number. A letter (Appendix B, No.1) was mailed to the parents/guardians of all eligible infants, using the address given on the birth list, inviting them to participate in this study by attending a clinic 52 DESIGN AND METHODS offering assessment of their infant's iron status. An enclosed sheet (Appendix B, No.2) also informed the parents that they would be asked to provide information on their infant's diet, given the opportunity for their infant to participate in a preferential looking test (Fagan Test), as well as have their infant's growth measured. The letters were mailed to the parents about 3 weeks in advance of clinics which coincided with the time when the infant would be 39 ± 1 week old. The letter was followed by a phone call about 1 week later, by a trained research nurse, to find out if the parent/guardian had received the letter and if they were interested in having their infant participate in the study. If the parent/guardian agreed to have their infant participate, an appointment was made to attend a clinic at a time suitable for the parent/guardian. Ail the parents/guardians who booked an appointment were telephoned again 1 -2 days prior to the scheduled appointment as a reminder. At this time, it was asked if the infant was free of infection, taking medication (e.g. antibiotics), or if there was any reason they could not attend. In theses cases, and where possible, appointments were rescheduled. Up to 3-6 attempts were made to reach parents/guardians by telephone to make an appointment. 3.3. CLINIC SCHEDULING The clinics were scheduled in 2 blocks, each covering about 2 months. The clinics were held at Public Health Units, Community Centres and Neighbourhood Houses in the North, South, East and West sections of the City of Vancouver. The appointments were scheduled in half-hour blocks with about 4 infants booked each half hour. Fifteen clinics were offered from October to November 1993 (Appendix B, No.3), and were scheduled to run in the morning, afternoon or the evening on weekdays, or in the morning on Saturdays. Each clinic 53 DESIGN AND METHODS was about 3-4 hours in duration. Another 9 clinics were offered from March to April 1994 (Appendix B, No.4) and were scheduled to run once every week. These clinics had a longer duration of about 6 hours. All the clinics provided an opportunity for interested parents to obtain nutrition counselling after completion of the questionnaires. Refreshments were available for the study participants, and toys and video entertainment were usually available to help occupy older siblings. 3.4. INFORMED CONSENT The purpose of the study and what was going to be done was explained to each parent/guardian by the research supervisor or research nurse. The parent/guardian was then asked to read the consent form (Appendix C) and sign it if they agreed to participate in the study. If the parent/guardian agreed to participate, they were then asked the name of the infant's doctor and told that the infant's blood results would be forwarded to the doctor if iron-deficiency anaemia or any test results requiring follow-up were found. The parents/guardians were also given a card with a telephone number and date to call (Appendix G) if they wanted to know their infant's blood results. 3.5. ANTHROPOMETRIC MEASURES Growth measurements were obtained for all participating infants by one of two trained personnel. The measurements made were body weight, length, head circumference and skinfold thickness (as an index of adiposity). The parent/guardian was present as all the anthropometric measures were being taken. Anthropometric data were collected to allow for consideration of potential confounding 54 DESIGN AND METHODS effects of general undernutrition in any infant subsequently found to have iron-deficiency anaemia. Infants below the 5th percentile for weight or weight for length, with reference to the Canadian reference growth charts (Nutrition Canada, 1980) would be considered at potential risk for delayed growth. i) Body weight The parent/guardian was asked to undress the infant but leave a dry diaper on (dry diapers were available to the parent/guardian when needed). The weight of the infant was then measured using an electronic balance accurate to 5 gms (Digital baby scale model 727, Lux & Zwingenberger LTD, Lakeshore, Toronto). The weight was recorded immediately on to a form with the infant's identification number (Appendix F). The scale was not adjusted for the weight of the diaper. ii) Length Crown to heel length was measured in the recumbent position using a paediatric length board accurate to the nearest millimetre (Ellard instrumentation LTD, Seattle, WA). The measure was recorded immediately on to the same form with the weight-measure. iii) Head circumference The head circumference was measured to the nearest millimetre using a disposable paper tape (Mead Johnson, Evansville, IN) placed over the part of the occiput which gives the maximum circumference (Gibson, 1990). The measure was recorded immediately on to the same form with the weight and length measures. 55 DESIGN AND METHODS iv) Skinfold thickness Skinfold thicknesses were determined using a Lange skinfold thickness caliper (Cambridge Scientific Inc, Cambridge, MD). The skinfold measures were recorded to the nearest 0.5 mm. These precision calipers are designed to exert a defined and constant pressure of 10 gm/mm2 throughout the range of measured skinfolds, and to have a standard contact surface or "pinch" area of 20-40 mm2 (Gibson, 1990). The skinfold measures were recorded immediately, on to the same form as the other measures mentioned above. Two readings were recorded for each site and if any two readings differed by more than 0.6 mm a third measure was taken. The average of the two measures with a difference of <0.6 mm was recorded into the data base. The infant was held by the parent/guardian as the measures of skinfold thickness were taken. i) subscapular skinfold was measured by grasping a pinch of skin between the thumb and forefinger just below and laterally to the angle of the left shoulder blade (Gibson, 1990). While maintaining the pinch of skin, the caliper was applied and the measure read after 2 to 3 seconds. The grasp was maintained throughout the procedure and the subsequent measures. ii) tricep skinfold was measured by grasping a vertical pinch of skin and subcutaneous fat between the thumb and forefinger at the midpoint of the upper left arm over the triceps muscle located at the back of the arm. The calipers were then applied about 1 cm below the pinch and the measurement read after about 2 to 3 seconds. The results were recorded immediately. The grasp was maintained throughout the procedure and the subsequent measures. 56 DESIGN AND METHODS iii) abdominal skinfold was measured by grasping a pinch of skin between the thumb and forefinger about 1 cm inferolateral to the umbilicus. While maintaining the grasp, a reading was taken 2 to 3 seconds after the calipers were applied. Once again the grasp was maintained throughout the procedure and the subsequent measures. 3.6. STUDY QUESTIONNAIRES The parents/guardians were given two questionnaires to complete; one on information regarding their family background and the other on their infant's nutritional history. 3.6.1. Developing the questionnaires Questionnaires were used to collect information on the family background, and on the infant's nutritional history. The questionnaires were developed by the research supervisor and myself, with input from nutritionists with the Vancouver Public Health Department. The diet history questionnaire was designed to collect current and retrospective information which would allow for identification of feeding practices associated with risk for iron-deficiency anaemia or low iron status. The history included questions on: the duration of breast-feeding, age of introduction of formula, the type of formula used (low iron or iron-fortified), age of introduction and types of cows' milk (whole, 2%, 1 % or skim) used, age of introduction of solid foods including cereals, vegetables, fruits, dairy and other animal products (e.g. meat, poultry or fish) and fruit juices and the use of vitamin/mineral supplements (Appendix E). Information on the family background included questions on the parent's/guardian's ethnic background, how long they had lived in Canada, age, family income, marital status and the highest level of education attained (Appendix D). 57 DESIGN AND METHODS The questionnaires were translated into Cantonese and were available for use at one clinic that was scheduled in an area of Vancouver with a high density of Chinese, as well as for other clinics as needed. These questionnaires were translated back into English by an independent, second individual to assure the accuracy of the translation. 3.6.2. Validating the questionnaires In developing the questionnaires, several steps were taken to ensure a high credibility of the instrument. Consultation was made with Ruth Milner, Head Research Support Group at B.C. Research Institute for Child and Family Health and Vancouver Health Department Public Health Nutritionists to address the content validity of the questionnaires. The questionnaires were pilot tested to address their face validity, as well as their content validity, with a group of 8 mothers from average income homes who had completed a high school education. The mothers were all residents of New Westminster, B.C. The mothers were asked to complete the questionnaires, then give constructive criticism on the wording of the questions, comment on the content, how comfortable they felt answering the questions, and if there was anything they thought should be added. Revisions to the questionnaires were then made. These included, for example, reasons for stopping breast-feeding and the different types of grain and cereal products used. The revised questionnaires were then taken for comments from the City of Vancouver Public Health Nutritionists and Ruth Milner, and final revisions were then made. 58 DESIGN AND METHODS 3.6.3. Completing the questionnaires The parents/guardians were instructed in the correct method of completing the questionnaire by the research nurse or nutritionists working on the project, who were familiar with the questionnaires and their objectives. These personnel were readily available to help or answer any queries if needed. The information on family background was collected on a separate questionnaire that was marked confidential. All the forms were identified by the infant's study number, but no names or addresses were on the nutritional history or family background forms. All the questions, except those on the confidential questionnaire (relating to the family's income, marital status, education and age) were checked by a nutritionist to ensure completeness. Any problems or inconsistencies in the information recorded were then discussed with the parent/guardian and corrections made if necessary. In a few incidences a relative brought the infant to the clinic. In these cases the relative was asked to take the questionnaires to be filled in at home by the parents and then to mail them back. 3.7. THE VISUAL RECOGNITION MEMORY TEST A standardized test of visual recognition memory, the Fagan Test of Infant Intelligence (Infantest Corporation, Cleaveland Heights, OH), was used to assess cognitive development. The test was administered at each clinic by this student or one other student according to test procedures described in the test manual (Infantest Corporation, Cleaveland Heights, OH). The tests were given in a quiet room or space as free from distraction as possible. Specific tests are available for infants of different ages, for example, 67, 69, 79 and 92 weeks postconception. A term infant with gestational age of 40 weeks, for example, would be tested at 27, 29, 39, and 52 weeks after birth, respectively with these particular tests (Fagan, 59 DESIGN AND METHODS 1991). For this study, the 39 ±1 week old infants were tested with the 79 week postconception test. Before the test was administered, the parent/guardian was given a synopsis of what was going to be done and what the test was going to measure. The parent/guardian was informed that the test would last for about 15-30 minutes, but could be stopped at any point if the infant became fussy or disturbed. The parent/guardian was also told that the baby could eat (but not breast-feed) as the test was being administered as long as the baby's view of the screen was not obstructed. The parent/guardian was then asked to hold the baby on their lap facing the front of the test stage. The parent/guardian was instructed not to point or comment on the pictures as this could influence the baby's looking behaviour. The test was then started. The test consists of 10 novelty problems made up of achromatic and coloured photographs of the faces of men, women and babies. The test involves first familiarizing the infant to a pair of identical faces for a specified amount of time. One of the pictures is then removed and replaced with a non-identical (novel) picture. The tester, unseen by the infant, observes the baby's looking behaviour by watching through a small peep hole in the center of the stage. The infant's looking direction (familiar versus novel) is recorded directly into a portable computer by pressing keys on the keyboard designated for a left or right look. The time (duration) of the look is recorded as the time during which the computer key is depressed. The time allowed for familiarization, and for comparing the novel and the familiar face is standardized, with the tester being prompted to remove the face cards by a beep from the computer as each interval is completed. The positioning of the novel and familiar faces (i.e. left or right) and the sequence in which they are placed on the stage is consistent for 60 DESIGN AND METHODS every test. At the end of the testing period, the computer calculates the infant's mean novelty preference score over the 10 novelty problems by dividing the amount of time looking at the new picture (novel face) over the total amount of looking time for both pictures. The testers made notes of their level of confidence with each infant's overall test performance at the time of the test and before seeing the score. This was recorded as okay (confident) or not okay (not confident). The test was recorded as not okay, if it was felt that the infant for example, due to inattentiveness, fussiness or tiredness did not attend to the test in a "reliable" way. The test results were then printed out from a portable printer. The parents/guardians were not usually shown the print out, but the tester explained how well their infant had performed. 3.8. HEMATOLOGY 3.8.1. Blood collection The blood draw was always done after all other measures were made on the infant and after the parent/guardian had completed the questionnaires. Three tubes of capillary blood were collected by a trained phlebotomist from a finger prick using a sterilized lancet and after warming the infant's finger. Blood was collected as follows; • 1 tube of 250 //L of blood collected into an ethylene diamine tetra-acedic acid (EDTA) coated microtainer tube. • 1 tube of 120 pL of blood collected into a microtainer tube coated with lithium heparin. • 1 tube of 800//L of blood collected into a second microtainer tube coated with lithium heparin. 61 DESIGN AND METHODS The tubes were labelled with the infant's 4 digit study ID number only. The two lithium heparin tubes and not the EDTA tube were placed on ice until the completion of the clinic. At the end of each clinic, the tubes were transported to the laboratory at the Children's Variety Research Centre. 3.8.2. Blood analyses 3.8.2.1. Analysis of hemoglobin The 250 pL EDTA tube of blood was taken to the hematopathology laboratory at B.C. Children's Hospital as soon after each clinic as possible. This was used for analysis of hemoglobin. The analyses were done on whole blood using a TOA Sysmex NE series 23, parameter blood counter (TOA Sysmex, Los Alamitos, CA, USA) according to the manufacturer's instructions. This instrument has been in routine use in the hospital laboratory for the past 3 years. The counter also generates values for the red cell indices and white blood cell and platelet counts. 3.8.2.2. Analysis of zinc erythrocyte protoporphyrin fZPP) The lithium heparin tubes with 120 pL of whole blood were stored at 4°C in the nutrition laboratory at the Children's Variety Research Centre until analysis of ZPP the following day. ZPP was measured using a hematofluorometer (Helena Laboratories, Beaumont, TX, USA), an instrument designed to measure the zinc protoporphyrin/heme ratio in whole blood, as described by Blumberg et al. (1977). Before the analysis, the hematofluorometer was allowed to warm up for 30 minutes according to the manufacturer's instructions. The protofluor calibrators, one high and one low (Helena Laboratories, Beaumont, TX, USA) and 62 DESIGN AND METHODS the stored blood samples were allowed to equilibrate to room temperature. The hematofluorometer was then calibrated using the protofluor calibrators according to the manufacturer's instructions. A drop of high protofluor calibrator was placed directly on a sample cover slip and the hematofluorometer was calibrated by pressing to scale, up or down, until the calibrator value was obtained (the reading should be ± 10 of insert on initial reading). The calibrator value for the high protofluor calibrator given by the manufacturer was 168 //mol. A reading of another drop of high protofluor calibrator was taken to verify calibration. Using the same procedure as above, the hematofluorometer was then calibrated using the low protofluor calibrator. The value of the low calibrator given by the manufacturer was 74//mol and a reading of ±3 of the assigned value was acceptable. Once the instrument was calibrated, 100//L of protofluor reagent (Helena Laboratories, Beaumont, TX, USA) was added to 50 fjL of whole blood sample in a test tube (12X75 mm), vortexed, and a reading taken within 5 minutes by inserting a coverslip (10x125)(Helena Laboratories, Beaumont, TX, USA) with the mixed solution on it into the hematofluorometer. All measurements were done in duplicate and the average calculated. 3.8.2.3. Analyses of ferritin, total serum iron and unsaturated iron binding capacity The lithium heparin tube with 800 //L of blood was used for the analysis of ferritin, total serum iron and unsaturated iron binding capacity (UIBC). The tubes were centrifuged (DPR-6000 Centrifuge, Damon/I EC Division, Needham Hts, MASS) at 3000 rpm (2000 g) at 4°C for 5 minutes. The serum was transferred into labelled 600//L eppendorf tubes and then stored at 4°C for analysis the following day. 63 DESIGN AND METHODS i) Ferritin analysis Ferritin was determined by a two site immunoradiometric assay as modified by Miles et al (1974) using a custom made "fer-iron" radioimmunoassay kit (Ramco Laboratories Inc, Houston, TX, USA). Before analysis, the reagents, the controls and patient sera were allowed to equilibrate to room temperature. Ten/sL of serum or calibrator (prediluted ferritin calibrator solutions of 6, 20, 60, 200, 600 or 2000 pg/mL of human spleen ferritin; i.e. the standards) were pipetted into ferritin antibody coated plastic microtiter tubes to which the ferritin in the patient sera or calibrator binds. Ten JJL of an anaemic control and 3 immunoassay control serum levels, 1, 2 and 3 (Lyphocheck, Biorad, Anaheim, CA) were each then pipetted into separate ferritin antibody microtiter tubes and were run with each assay for quality control. Two hundred microliters of radiolabeled [125l] antiferritin was then pipetted into each of the tubes. A "sandwich" is formed as the radiolabeled antiferritin binds to the ferritin in the solid phase. The tubes were then incubated for 2 hours at room temperature in a Dubnoff shaking incubator bath (Precision Scientific Inc. Chicago, IL, USA) at 150 cycles/minute. After the 2 hour incubation period, the unbound labelled antiferritin was removed by washing 3 times with distilled water into a sink with running water. The radioactivity in the washed microtiter tubes was then counted using a Clinigamma counter (Model 1272, LKB Wallac, Fishers Scientific, Ottawa, Canada). The ferritin concentration was then computed by entry of calculations provided by the manufacturer of the kit into a spread sheet (VP-Planner Plus, Stephenson Software Inc., Berkeley, CA). All ferritin analyses were done in duplicate, and the average of the counts was used in the calculation of the ferritin concentration. Used radioactive solutions and solids were disposed of in accordance with approved regulations as stipulated by the University of British Columbia Department of Occupational Health and Safety. 64 DESIGN AND METHODS ii) Total serum iron and UIBC analysis The serum remaining after the analysis of ferritin was used for analysis of total serum iron using a diagnostic kit manufactured by Hoffman-La Roche Ltd (Basel, Switzerland) and unsaturated iron-binding capacity (UIBC) using a diagnostic kit from Diagnostic Chemicals LTD (Charlottetown, PEI). The analyses were done according to the manufacturer's instructions provided in the kits. i) . Iron Assay: Five hundred //L of buffer (200 mmol/L sodium acetate pH 4.5, 120 mmol/L thiourea, 300 mmol/L hydroxylamine and 4.5 mol/L guanidine hydrochloride 38%) was added to 100 //L of sample serum, iron standard reagent (70 //mol/L iron) provided in the Dignostic Chemicals LTD kit (see below for method of preparation of the iron standard reagent) and water in a glass test tube (13 x 100 mm). The solutions were then mixed and a reading of the sample blanks (A1) (each sample acts as its own blank) was taken at 562 nm using a SP8-400 UV/VIS spectrophotometer (Pye Unicam LTD, Phillips, Ontario, Canada). At timed intervals (30 seconds), 25 //L of chromogen (iron colour reagent; 200 mmol/L sodium acetate and 39 mmol/L ferrozine) was added to each sample and the solution immediately vortexed. After exactly 20 minutes, the absorbency of each of the samples (A2) was read at 562 nm. All samples were run in duplicate. ii) . UIBC Assay: The iron standard reagent for this assay was prepared by mixing 1 package (200 mg) of ascorbic acid with 10 mL of the iron standard (70 //mol/L iron) provided in the diagnostic kit. Two hundred microliters of this iron standard reagent was added to each of 200//L of sample serum or 200//L distilled water for the standard. Another test tube with 400 //L of distilled water was used for the blank and had no iron standard reagent added to it. The samples were vortexed and then 200 //L of binding reagent was added to all of the 65 DESIGN AND METHODS samples. The samples were then left to stand for 20 minutes. After the 20 minutes, the absorbency of the sample blanks (A3) was taken at 562 nm (each sample acts as its own blank therefore, A3 is the absorbency of each of the samples before addition of the colour reagent). Twenty//L colour reagent was then added, the samples vortexed and left to stand for 5 minutes. After 5 minutes the final absorbency of the samples (A4) was read at 562 nm. All samples were run in duplicate. 3.8.2.4. Calculation of total serum iron. UIBC and TIBC. The equations below were entered into a spread-sheet (VP-Planner Plus, Stephenson Software Inc, Berkley, CA) for computation of the total serum iron, UIBC and TIBC. Total serum iron total iron //mol/L = (A2 ; A1) sample X 70 //mol/L (concentration of iron standard) (A2 - A1) standard A1 sample = absorbency of the sample blank A2 sample = final absorbency of the sample A1 standard = absorbency of the standard blank A2 standard = final absorbency of the standard ii) jJJBC. Excess iron added, //mol/L (A4 - A31 sample X 70 //mol/L (A2 - A1) standard A1 standard = absorbency of the standard blank from the iron assay A2 standard = final absorbency of the standard from the iron assay A3 sample = absorbency of the sample blank (i.e. sample serum, distilled water for the standard and distilled water for the blank (no iron standard reagent added) A4 sample = final absorbency of the samples 66 DESIGN AND METHODS • UIBC = total iron added (i.e. 70 //Imol/L) - excess iron added iii) Total iron binding capacity Total iron binding capacity (TIBC) was calculated as the sum of the total iron and UIBC. • TIBC = total iron + UIBC 3.9. CLASSIFICATION OF IRON STATUS The cut off values for the biochemical tests used to classify iron status were as follows; 1. Hemoglobin :S 110 g/L 2. Serum ferritin ^10 //g/L 3. Total iron binding s= 60//mol/L capacity 4. Zinc protoporphyrin s= 70 //molZPP/mol heme These cut off values are based on values commonly used in the literature to diagnose iron-deficiency anaemia or low iron status. Infants were classified as follows based on the results of the above tests: Group 1: Iron deficient anaemic All infants with a hemoglobin of <101 g/L, or a hemoglobin of ^110 g/L with 2 or 3 abnormal biochemical test results (i.e. serum ferritin ^10 //g/L, TIBC s= 60 //mol/L and ZPP >70 //mol ZPP/mol heme). Group 2: Low iron status (non-anaemic iron depleted) All infants without iron-deficiency anaemia with a serum ferritin of ^ 10 //g/L. 67 DESIGN AND METHODS Group 3: Non-anaemic iron-sufficient This group included all infants with hemoglobin > 110 g/L (non-anaemic) and a serum ferritin > 10 //g/L (iron-sufficient). Group 4: Uncertain This group included all infants with a hemoglobin > 101 g/L - s 110 g/L, serum ferritin > 10 //g/L and 0 or 1 abnormal biochemical test results. These infants were classified as such because their ferritin values were all greater than 10 //g/L, i.e. consistent with normal iron stores. However, there is the possibility that infection may have resulted in serum ferritin values within the normal range. Thus, iron-deficiency may or may not have been the reason for the low hemoglobin. Table 3.1. Summary of cutoff values used to classify infants iron status Group Classification Group 1 Iron deficient anaemic Hgb :£101 or Hgb :£ 110 g/L & 2 or 3 abnormal biochemical test results Group 2 Low iron status (non-anaemic iron depleted) serum ferritin < 10 //g/L without iron-deficiency anaemia Group 3 Non-anaemic iron sufficient Hgb > 110 g/L and serum ferritin > 10 //g/L Group 4 Uncertain Hgb > 101 - ^110 g/L with 0 or 1 abnormal biochemical test result and a serum ferritin > 10 //g/L 68 DESIGN AND METHODS 3.10. DATA ANALYSIS A Chi-square test was used to determine if there was any relationship between iron status and the variables of interest. These included data from the nutritional history (duration of breast-feeding, age of introduction of various foods and age of introduction of low iron formula) and demographic data (family income, mother's education, and ethnic background). The Goodman and Kruskal Tau test was used to determine the strength of the relationship between each of the independent variables (nutrition history and demographic data) and the dependent variable, which was always iron status. A One-Way Analysis of Variance was used to examine if there were any differences in scores on the visual recognition memory test among infants in the different iron status groups. The Tukey and Scheffe tests were done simultaneously with the One-Way Analysis of Variance to determine which groups, if any, were significantly different from each other. Descriptive statistics were calculated for all blood parameters used to assess iron status; hemoglobin, serum ferritin, total iron binding capacity, zinc protoporphyrin, and for the test scores on the visual recognition memory test. All statistics were done using the Statistical Package for the Social Sciences (SPSS) (Release 4.0, SPSS Inc. Chicago, Illinois). 69 RESULTS Chapter IV 4. RESULTS 4.1. CLINIC ATTENDANCE 4.1.1. Clinics set 1 A birth list was obtained from the City of Vancouver Public Health Department for infants born between January 1st, 1993 and March 3rd, 1993. The list was used to identify those infants who would be 39 ±1 week of age during the first set of clinics that were scheduled in October and November, 1993. This list had 2007 births. From this list, 870 (43.3%) infants met the eligibility criteria of a gestational age ^37-^42 weeks, birth weight of 2500-4500 gms with an address in the city of Vancouver. Infants on the birth list who had either no address or an incomplete address could not be selected for potential participation in the study. Of the 870 letters mailed out, 91 (10.5%) were returned due to a wrong address. However, a phone call was attempted to all the 870 parents/guardians. The phone call clarified if the parents/guardians had received the letter of information about the study. A total of 213 parents/guardians could not be reached by phone (wrong number, no-one at home). Telephone contact was made with 657 of the 870 parents/guardians. However, 103 of these parents/guardians could not speak English sufficiently to allow an adequate description of the study protocol. From this group, a subgroup of 25 Cantonese speaking families with an infant who would be 39 ± 1 week old coincident with a clinic scheduled at Strathcona Community Centre, located in an area in Vancouver with a high density of Chinese, were selected. This 70 RESULTS was done specifically to ensure good representation of infants from families of Chinese background in the study. These 25 Chinese families were then telephoned by an assistant fluent in Cantonese. Fourteen families agreed to participate in the study.2 Thus, a total of 579 (66.6%) (including the 25 Chinese families) of the 870 parents/guardians of eligible infants, were successfully contacted by phone. Of these 579,302 (52.2%) parents/guardians agreed to participate in the study, and booked an appointment to attend a clinic coinciding with when their infant would be 39 ± 1 week of age. Two hundred and twenty-five (225) of the 302 (74.5%) parents/guardians with an appointment attended the clinics. Another 2 parents without an appointment brought their infant to a clinic. These infants were born outside Vancouver, but the parents had heard about the study from another mother. The infants met the eligibility criteria for gestational age and birth weight, and age of the infant (39 ± 1 week) at the clinic they attended. A total of 227 infants were seen in the first set of clinics. 4.1.2. Clinics set 2 Another birth list of 1715 babies was obtained from the City of Vancouver Public Health Department for the second set of clinics scheduled between March and April, 1994 (Table 4.1). Nine hundred and forty-three (54.9%) infants born between June 4th, 1993 and July 8th, 1993, inclusive, met the eligibility criteria (gestational age >37-<42 weeks, birth weight of 2500-4500 gms and had an address in the City of Vancouver. Fifty-four of the 943 (5.7%) letters mailed were returned due to a wrong address. Again, an attempt was made to telephone all parents/guardians identified from the birth list. The consent form and the questionnaires were translated into Cantonese. 71 RESULTS One hundred and sixty-four of the 943 (17.4%) parents/guardians could not be reached by phone. A total of 779 parents/guardians were contacted by phone. However, communication was not possible with 80 of these parents/guardians as their English language skills were not sufficient for discussion of the study protocol. No language other than English was used during phone calls in preparation for this set of clinics. However, posters in Cantonese were posted in Public Health Units around the city of Vancouver inviting parents with 9 month old infants to participate in the study. Successful telephone communication was made with 699 (74.1 %) parents/guardians of the 943 eligible infants. Of these 669, 290 (41.5%) booked an appointment to participate in one of the study clinics. Two hundred and two (202) of the 290 (69.7%) parents/guardians attended a clinic. An additional 7 infants were brought to the clinics by parents/guardians who had heard about the study through the Public Health Units, or from another mother. Of these infants, one did not meet the eligibility criteria of 39 ± 1 week of age when brought to the clinic. A total of 208 eligible infants were seen in the second set of clinics. In all, the study surveyed 25.1 % (225/870) and 21.4% (202/943) of all infants who met the eligibility criteria for participation in the first and second set of clinics, respectively. 72 RESULTS Table 4.1. Sampling of study population Study population Oct-Nov 1993 Mar-Apr 1994 (clinics set 1) (clinics set 2) No. infants on birth list 2007 1715 No. infants who met eligibility criteria1 (mail out) 870 (43.3%) 943 (54.9%) No. contacted but unable to communicate with on the phone2 103J (11.8%) 80 (8.5%) No. wrong phone number, or not at home 213 (24.5%) 164 (17.4%) No. contacted with effective communication 579* (66.6%) 699 (74.1%) No. contacted successfully who agreed to 302 290 participate in study (booked appointments)5 (52.2%) (41.5%) No. with a booked appointment who attended 225 202 aclinic6 (74.5%) (69.7%) No. infants who attended with no booked 2 6 appointment Total clinic attendance by eligible infants 227 208 No, number 1 Eligibility criteria; gestational age 2:37-^42 weeks, birth weight of 2500-4500 gms and an address in Vancouver for contact. Letters describing the study protocol were mailed to all eligible infants, 91 and 54 were returned marked incorrect address for clinic set 1 and 2, respectively. 2 Unable to communicate with because parents/guardians had limited English language skills 3 Twenty-five (25) Chinese families with whom communication was difficult were selected for contact by a Cantonese speaking assistant. 4 The 25 Chinese families telephoned by a Cantonese speaking assistant are included. 5 percent contacted successfully by phone (i.e. 302/579 and 290/699) in brackets 6 percent with booked appointment (i.e. 225/302 and 202/290) in brackets. 73 RESULTS 4.2. STUDY SAMPLE SIZE A summary of the study sample size is shown in Table 4.2. A total of 24 clinics were held between October - November, 1993 and March - April, 1994. Four hundred and thirty five (435) eligible infants attended the clinics. Data for 1 infant was later excluded from the study because of a history of prenatal substance abuse. The final study sample size consisted of 434 infants, of which all lived with their parents. Table 4.2. Summary of the number of infants who participated in the study Number of Number of Number of infants clinics offered infants seen included in the study n n n Clinics set 1 October-November, 1993 15 227 2261 Clinics set 2 March-April, 1994 9 208 208 Total 24 435 434 One infant excluded due to a history of prenatal substance abuse. 74 RESULTS 4.3. DESCRIPTION OF THE STUDY POPULATION Overall, the study population comprised a broad spectrum of racial and ethnic backgrounds. However, the infants were predominately from higher-education, higher-income two-adult families. No incentives were given to parents to encourage participation in this study. Thus, the parents who did participate were likely to be relatively highly motivated with respect to their infant's well being. 4.3.1. Gender distribution There were more male infants (54.4%) than female infants (45.6%) in the group of infants in the study. A similar gender distribution was present in both sets of clinics (Table 4.3). Table 4.3. Gender distribution of study population Total Gender Female Male Clinics set 1 226 n 105 121 % 46.5 53.5 Clinics set 2 208 n 93 115 % 44.7 55.3 Both clinics 434 n 198 236 % 45.6 54.4 %, percent female and male infants within each clinic set and overall. 75 RESULTS 4.3.2. Parental age distribution Figures 4.1 & 4.2 show the distribution of age of the parents who participated in the study. Most of the mothers (Figure 4.1) and fathers (Figure 4.2) were in the older age groups. Only 0.7% (n = 3) of the mothers, and none of the fathers were under 20 years of age, and only 7.8% (n = 34) of the mothers and 2.5% (n = 11) of the fathers were under the 25 years of age. Twenty four percent (n = 105) of the mothers and 14.7% (n = 64) of the fathers were 25 to 29 years of age, and 64.7% and 74.6% of the mothers and fathers, respectively, were over 30 years of age. Two point five percent of the mothers and 8.1 % of the fathers, did not report their age. 76 R E S U L T S 180 -i 165 J o. 150 -I I 135 H S 120 H o o 4> 2 SZ *-> o £ 105 -90 -75 -60 -45 30 15 H 0 (0.7X) (2.5X) <20 20-24 25-29 30-34 Age (years) 35+ unknown Figure 4.1. Distribution of mothers' ages n = 434; No, number; percent of mothers within each age group in brackets. 77 RESULTS 195 i <20 20-24 25-29 30-34 354- unknown Age (years) Figure 4.2. Distribution of fathers' ages n = 434; No, number; percent of fathers within each age group in brackets. 7 8 RESULTS 4.3.3. Marital status The marital status of the parents is shown in Table 4.4. Single parents (including those who were separated or divorced) represented 7.8% of the parents and 90.3% of the parents were married or lived common-law. Marital status was not reported by 1.8% of the infants' parents. No parents reported that they were a widow or widower. Table 4.4. Marital status of the parents Marital Status Clinic set 1 (n = 226) Clinic set 2 (n = 208) Both clinics (n = 434) Married/common-law n 208 184 392 % 92.0 88.5 90.3 Single n 13 14 27 % 5.8 6.7 6.2 Separated/divorced n 2 5 7 . % 0.9 2.4 1.6 Widow/widower n 0 0 0 % Unknown n 3 5 8 % 1.3 2.4 1.8 n, number; percent of parents within each marital status 79 RESULTS 4.3.4. Family income The data on family income was separated into one adult and two adult families; the former included single and separated/divorced parents (n = 34). Most of the two-adult families (n = 392) were in the higher income brackets; 36.9% had an annual family income over $50,000, 27.0% had a family income between $30,000 and $50,000, and 15.8% reported a family income between $20,000 and $29,999 per year. Only 11.2% of the two adult families reported an annual family income between $10,000 and $19,999, and only 5% reported an income of less than $10,000 per year (Figure 4.3). Annual family income was not reported by 4.1 % of the two adult families. Of the 34 single parent families, 14.7% reported a family income of less than $10,000 per year, and 52.9% reported an annual family income between $10,000 and $19,999. About 9% of the single adult families had an annual family income of $20,000 to $29,999 and another 9% had an income of $30,000 to $50,000. Only 2.9% of the single parents reported an annual family income over $50,000. Annual income was not reported by 11.8% of the single adult families (Figure 4.4). 80 RESULTS 1 5 0 -1 1 3 5 • Q. Z3 1 2 0 • O V 1 0 5 -<U E o 9 0 • o c. 7 5 -SZ o o 6 0 • <u c 4 5 • d 3 0 -1 5 -0 • (36.9*) (27.0%) (15.8*) (11.2*) (4.8%) Legend nTTTl <$10,000/yr $10.000-19,999/yr [==3 $20.000-29,999/yr fggg $30,000-50,000/yr ^ >$50,000/yr \ 1 Not known (4.1%) Family Income Figure 4.3. Annual income of two adult families n = 392; No, number; yr, year; percent of two adult families within each income grouping in brackets 81 RESULTS 2 0 -I 18 -CL Z3 16 -2 14 -£ O 12 -O c 10 -sz o a 8 -cu c 6 -d 4 -2 -0 -Legend PTffl <$10.000/yr $10,000- 19,999/yr $20,000-29,999/yr $30,000-50,000/yr >$50.000/yr Not known (14.7* (8.8%) (8.8%) (11.8%) (2.9%) Family Income Figure 4.4. Annual income of one adult families n = 3 4 (includes those separated or divorced); No, number; yr, year; percent one adult families within each income grouping in brackets 82 RESULTS 4.3.5. Parental education The highest level of education reported by the mothers and fathers of the infants studied is shown in Figures 4.5 and 4.6, respectively. Most of the mothers and fathers had some form of a higher level of education. Only 7.6% of the mothers and 8.1 % of the fathers reported that they had not completed high school. Eighteen percent of the mothers and 19.8% of the fathers had completed high school but had no further educational or vocational training, 32 .7% of mothers and 26.7% of fathers had some college or vocational education, and about 4 0 % of each of the mothers and the fathers had completed university. The highest level of education attained was not reported by 1.6% of the mothers and 5.8% of the fathers. 83 R E S U L T S 0> 1 8 0 -1 1 6 5 • ~B 1 5 0 • o 1 3 5 • o o D 1 2 0 • x» CD 1 0 5 • _c o 9 0 • D <D 7 5 • C * 6 0 -cn a> 4 5 -+^  o 3 0 -E 1 5 • 6 0 • (39.9X) (32.7X) (mm) Legend Incomplete high school f=\ High school only College/vocational training rTTTTl University I I Not known Level of education Figure 4.5. Highest level of education attained by mothers n = 434; No, number; percent of mothers within each level of education in brackets 84 RESULTS 180 -j > 165 -o 150 -c o 135 -o o 120 • TJ CU 105 • .c o 90 • o <D 75 -C 60 • 01 <u 45 • -4-< o 30 -Q 15 -z 0 -(39.6%) (26.7X) 09.8%) (8.155) Legend KSSSI Incomplete high school High school only College/vocational training rrnrn Universi ty J" Not known <5.8X) Level of education Figure 4 . 6 . Highest level of education attained by fathers n = 4 3 4 ; No, number; percent of fathers within each level of education in brackets 85 RESULTS 4.3.6. Parental ethnic association Ethnic association was determined by self-report. Questions were also asked on the main language(s) spoken at home, whether or not the parents had been born in Canada, and the number of years they had been resident in Canada. Table 4.5 shows a summary of the ethnic backgrounds as reported by the mother. For presentation of the results, the information on ethnic background obtained from the questionnaire (Appendix D) is grouped as follows: 1) Canadian/American including all who reported they were Canadian or American only, 2) British/North West (NW) European, including English, Scottish or Irish (British), German, Dutch, French or Scandinavian (Swedish, Norwegian or Danish), 3) East European, including Ukrainian, Polish, Croatian, Latvian, Yugoslavian, Russian, Czech, Romanian or Hungarian, 4) Mediterranean, including Italian, Greek, Spanish or Portuguese, 5) Chinese, 6) Filipino, 7) East Indian, 8) other Asian, including Japanese, Korean or Vietnamese, 9) African, and 10) other, including Jewish, El Salvadorian, North American Indian, Arabic, Iranian, Palestinian or Egyptian and all those who checked "other" ethnic group on the questionnaire but did not specify any ethnic background. The last group is heterogenous, but because each group was small, meaningful information could not be expected by considering them separately. 86 RESULTS Table 4.5. Parental ethnic association Ethnic group Total number of infants (n = 434) Percent of total (%) Canadian/American 120 27.6 British/NW European 66 15.2 East European 27 6.2 Mediterranean 32 7.4 Chinese 81 18.7 Filipino 18 4.1 East Indian 43 9.9 Other Asian 23 5.3 African 10 2.3 Other 13 3.0 Unknown 1 0.2 British/NW European: English, Scottish, Irish, German, Dutch, French, Norwegian, Swedish, or Danish; East European: Ukrainian, Polish, Croatian, Latvian, Yugoslavian, Russian, Czech, Romanian or Hungarian; Mediterranean: Kalian, Greek, Spanish or Portuguese; Other Asian: Japanese, Korean or Vietnamese; Other: Jewish, Arabic, Iranian, Palestinian, Egyptian, North American Indian, El Salvadorian or "other" ethnic not specified. 87 RESULTS 4.3.7. Immigrant and non-immigrant mothers Not all parents born in Canada necessarily identified their ethnic background as Canadian. Therefore, to determine whether an infant's mother was or was not born in Canada, a comparison was made between the number of years the mother had lived in Canada and her age. Forty-seven percent (47.0%) of the mothers were born in Canada, and 41.9% were born outside Canada (Table 4.6). Information on whether the mother was born in Canada could not be reliably deduced for 11.1% of the mothers. Data for immigrant mothers was further analyzed to determine how long they had been residing in Canada. Of the 182 immigrant mothers, 44.0% had lived in Canada for 5 years or less, and 56.0% had lived in Canada for more than 6 years (Table 4.7). Table 4.6. Number of mothers born or not born in Canada Born in Canada1 Yes No Uncertain n 204 182 48 % 47.0 41.9 11.1 n= 434 1 extrapolated by comparison between the number of years the mother had lived in Canada and her age. 88 RESULTS Table 4 .7. Number of years immigrant mothers had lived in Canada1 Years immigrant mothers Total Percent total had lived in Canada (n = 182) (%) <2 28 15.4 3-5 52 28.6 6- 10 39 21.4 >10 63 34.6 1 Based on the response to the question " how many years have you lived in Canada?" 89 RESULTS 4.4. VISUAL RECOGNITION MEMORY TEST (FAGAN TEST OF INFANT INTELLIGENCE) Table 4.8 summarizes the number of the Fagan Tests that were administered and the tester's confidence in the infant's test performance. A total of 401 Fagan Tests were administered (92.4%), of which 351 (87.5%) were recorded as completed with high confidence in the infant's test performance. Low test confidence was recorded for 12.5% (50/401) of the infants. Two infants did not complete the test due to instrument failure, and 31 infants (7.1 %) were not tested because their parent did not have time for the test. For the 50 tests recorded with low test confidence, 30 of the infants were inattentive, 13 were fussy/crying, 5 infants were scared, and 2 infants fell asleep before the test was completed (Table 4.9). In some instances, for example, where the infant was crying, the test was stopped before completion. Test scores for these infants were not computed. Table 4.8. Summary of Fagan Tests Clinic set Total number infants Number tested Number not tested1 Test confidence2 high low Tests not completed Clinic set 1 n 226 203 21 176 27 2 % 89.8 9.3 (86.7) (13.3) 1.0 Clinic set 2 n 208 198 10 175 23 0 % 95.2 4.8 (88.4) (11.6) Both clinics n 434 401 31 351 50 2 % 92.4 7.1 (87.5) (12.5) 0.5 1 Infants not tested because the parent did not have sufficient time to have the test administered. 2 In brackets are the percents of all tests done with high or low confidence. Low test confidence, infants for whom the tester felt that the infant did not attend to the test in a reliable way due to fussiness, inattentiveness or tiredness. 90 RESULTS Table 4.9. Reasons for low confidence in scores on the Fagan Test Reason for Clinic set 1 Clinic set 2 Both clinics rejecting test n - 2 7 n = 23 n = 50 Baby inattentive 15 15 30 Baby fussy/crying 8 5 13 Baby scared 3 2 5 Baby fell asleep 1 1 2 Testers made notes on their level of confidence with the infant's overall test performance at the end of the testing period and before seeing the infant's computer generated test score. A recording of low confidence in the test score was made if it was felt by the tester that the infant did not attend to the test in a "reliable" way due to inattentiveness, fussiness or tiredness. 91 RESULTS 4.5. FEEDING HISTORY The feeding history data for one infant was lost, hence the analysis of feeding history is based on 433 infants, rather than 434. 4.5.1. Duration of breast-feeding For the purposes of this study, exclusive breast-feeding was defined as breast-feeding with bottle-feeding with less than 12oz of formula or cows' milk per week. If the infant was fed formula with no breast-feeding by 7 days of age, then the infant was classified as never breast-fed. Infants who could not be classified as breast-fed or formula-fed by 1 month of age were considered to have received mixed feeds from birth. The mixed feed group also includes infants for whom there was more than 1 month overlap in breast-feeding and bottle feeding with > 12oz/week of formula or cows' milk, and infants who were bottle-fed formula/cow's milk at 39 ± 1 week and for whom the age at which breast-feeding was stopped is uncertain. Table 4.10 shows the duration of exclusive breast-feeding among infants who participated in this study. Of the 433 infants, 68 (15.7%) were never breast-fed. These infants were fed a variety of commercially available infant formulas. Breast-feeding was stopped between 7 days and 1 month after birth by 10.6% of the mothers. About 7% of the infants were classified as given mixed feeds, and 21.2% of the infants were still receiving breast-milk as their main source of milk (i.e. exclusive breast-feeding) at the time the infant attended the clinic. 92 RESULTS Table 4.10. Duration of breast-feeding Duration of breast-feeding1 Mixed3 (months) feed group never2 7d-<1 >1-<3 >3-=s6 >6-^8 >8 Number 68 46 51 95 49 92 32 (n = 433) Percent 15.7 10.6 11.8 21.9 11.3 21.2 7.4 total Refers to duration of exclusive breast-feeding, defined as breast-feeding with bottle feeding with < 12oz of formula or cows' milk per week. Infants fed formula with no breast-feeding by 7 days of age Infants who could not be classified as either breast-fed or formula fed by 1 month of age, infants for whom the age breast-feeding stopped was uncertain, and infants for whom there was an overlap of ^1 month in breast and bottle feeding. 93 RESULTS 4.5.2. Introduction of solid foods Table 4.11 shows the age of introduction of various foods to the infants. Some infants were given solid foods before 4 months of age. The solid foods given were predominately commercial iron-fortified infant cereal (7.6% of infants), fruits (2.3% of infants) and/or vegetables (1.6% of infants). No infants were fed egg yolk, tofu or meat before 4 months of age, and less than 1 % of the infants were fed chicken, fish or legumes before 4 months of age. Most of the infants were introduced to solid foods between 4 and 6 months of age, consistent with recommendations by the Canadian Paediatric Society Nutrition Committee (1991). By 6 months of age, 87.7% of the infants had been introduced to iron-fortified cereals, 72.5% to vegetables, 72.5% to fruits, and 54.8% to fruit juice. Foods that provide a source of iron in the infant's diet such as meat, chicken, fish, egg yolk, and legumes were predominately introduced at 7 to 9 months of age, consistent with current recommendations on infant feeding (Canadian Paediatric Society, 1991; Vancouver Health Department; Position Statement on infant feeding, 1993). Nevertheless, at the time of the study, when the infants were 39 ± 1 week of age, a large proportion of infants had not been introduced to these foods. That is, 33.2%, 24.2%, 62.4%, 57.6%, and 25.6% of the infants had not been introduced to meat, chicken, fish, egg yolk, and legumes, respectively. 94 RESULTS Table 4 .11. Age of introduction of various foods Food Age of introduction Total Not Not (months) introduced yet known <z3 4-6 7-9 Fruit ju ice* % 6.5 48.3 23.6 78.3 18.7 3.0 n 28 209 102 339 81 13 Infant cereal* % 7.6 80.1 5.8 93.5 4.0 2.5 n 33 347 25 405 17 / / Meat* % 0.0 26.1 36.3 62.4 33.2 4.4 n 0 113 157 270 144 / S Chicken* % 0.2 28.4 43.9 72.5 24.2 3.2 n / 123 123 314 105 / 4 F ish* % 0.7 11.8 19.2 31.6 62.4 5.7 n 3 51 83 137 271 25 Egg yolk* % 0.0 10.4 25.6 36.0 57.6 6.2 n 0 45 111 156 250 27 Legumes* % 0.7 44.1 22.6 67.4 25.6 6.9 n 3 191 98 292 111 Vegetables % 1.6 70.9 19.9 92.4 3.7 3.9 n 7 307 86 400 / S 17 Fruit* % 2.3 70.4 21.5 94.2 2.8 3.0 n 10 305 93 408 / 2 17 Tofu % 0.0 5.8 10.6 16.4 75.1 8.5 n 0 25 46 71 325 37 n = 433, %, percent of total Age of introduction refers to the period in which the food was first introduced Not yet, food had not yet been introduced at the time of the study Unknown, age of introduction of food was not reported by the parent Infant cereals refers to commercial iron-fortified infant cereals * Foods that provide a source of iron in the infant's diet * Sources of vitamin C; vitamin C enhances the absorption of iron 95 RESULTS 4.5.3. Use of low iron milk Table 4.12 shows the age of introduction of low iron "milk" to infants who had been breast-fed less than 7 months and who had been bottle-fed with a low iron "milk" for 2s 1 month. Breast-feeding as used here refers to any breast-feeding (i.e not exclusive breast-feeding). The term low iron milk includes; low iron infant formula, cows' milk, goats' milk. No infants were fed home-made formula. About 11 % of the infants were fed a low iron formula from birth, 14.8% were fed a low iron formula from between 1 and 3 months of age, 13.7% between 4 and 6 months of age, and 1.4% and 2.3% were fed a low iron formula from 7 or 8 months of age, respectively. 96 RESULTS Table 4.12. Age of introduction of low iron "milk" for infants fed a low iron milk for ^1 month. Age of introduction of low iron milk Number or infants introduced to low iron milk Percent of infants fed low iron milk out of all infants (n = 433) From birth 49 11.3 1 month 29 6.7 2 months 16 3.7 3 months 19 4.4 4 months 18 4.2 5 months 18 4.2 6 months 23 5.3 7 months 6 1.4 8 months 10 2.3 Low iron "milk" given for at least 1 month, includes cows' milk, goats' milk, low iron infant formula. Not included in the table are infants who were breast-fed for > 7 months or who were not fed a low iron milk. f 97 RESULTS 4.6. PREVALENCE OF IRON-DEFICIENCY ANAEMIA AND LOW IRON STATUS The mean, median and standard deviation values for the iron indices used to classify the iron status of the study population are shown in Appendix G. The prevalence of iron-deficiency anaemia and low iron status among the 9 month old infants in this study was 6.9% and 24.4%, respectively (Table 4.13). Classification was difficult for 9.9% of the infants as they had a haemoglobin greater than 101 g/L and less than 110 g/L, a serum ferritin > 10 //g/L and 1 or no abnormal parameter of iron status (i.e. TIBC 2=60 //mol/L or ZPP 2:70 //mol ZPP/mol heme). These infants were classified as iron status uncertain. More than half (58.8%) the infants involved in the study were classified as non-anaemic, iron sufficient, that is, normal according to the criteria used for the classification of iron status. 98 RESULTS Table 4.13. Prevalence of iron-deficiency anaemia and low iron status among nine month old infants in Vancouver Iron Status Total Percent total (n=434) (%) Iron deficient anaemic1 30 6.9 Low iron status2 106 24.4 (non-anaemic iron depleted) Non-anaemic,iron sufficient3 255 58.8 Uncertain4 43 9.9 1 infants with a hemoglobin ^101 g/L or s 110 g/L with 2 or 3 abnormal parameters of iron status from serum ferritin ^10 //g/L, TIBC s60 //mol/L and ZPP a 70 //mol ZPP/mol heme. 2 infants without iron-deficiency anaemia, but with a serum ferritin of :S 10 //g/L. 3 infants with a hemoglobin > 110 g/L (non-anaemic) and a serum ferritin > 10 //g/L (iron-sufficient). 4 infants with a hemoglobin > 101 g/L - £ 110 g/L and a serum ferritin > 10 //g/L, but with 0 or 1 abnormal parameter or iron status. 99 RESULTS 4.7. IRON STATUS AND ITS RELATION TO FEEDING PRACTICES 4.7.1. Relationship between iron status and duration of breast-feeding Chi-square analysis of the data on the relationship between iron status and duration of breast-feeding show a strong, significant association (p< 0.0001) between iron status in the infants and the duration of breast-feeding, with iron status as the dependent variable (Table 4.14). As the duration of breast-feeding increased, the risk of iron-deficiency anaemia and low iron status also increased. The prevalence of iron-deficiency anaemia and low iron status among infants who were breast-fed for more than 8 months was 15.2% (14/92) and 30.4% (28/92) respectively, compared to a prevalence of 1.5% (1/68) and 10.3% (7/68), respectively among infants who were never breast-fed. 100 RESULTS Table 4.14. Prevalence of iron-deficiency anaemia and low iron status in relation to duration of breast-feeding. Iron status Duration of breast-feeding Mixed (months) feeds never 7d- >1- >3- >6- >8 £ 1 £ 3 ^6 £ 8 n 68 46 51 95 49 92 32 Iron deficient 30 n 1 3 0 10 2 14 0 anaemic % 3.3 10.0 0.0 33.3 6.7 46.7 0.0 Low iron 106 n 7 12 17 23 14 28 5 status % 6.6 11.3 16.0 21.7 13.2 26.4 4.7 Non-anaemic 254 n 53 28 26 54 24 42 27 iron sufficient % 20.9 11.0 10.2 21.3 9.4 16.5 10.6 Uncertain 43 n 7 3 8 8 9 8 0 % 16.3 7.0 18.6 18.6 21.0 18.6 0.0 Significant association between iron status and duration of breast-feeding (P<0.0001). d, day; %, percent of infants within each duration of breast-feeding and mixed feed. Never breast-fed, infants fed formula with no breast-feeding by 7 days of age. Mixed feed, includes all infants who could not be classified as either breast-fed or formula-fed by 1 month of age. Also includes infants for whom the age breast-feeding stopped was uncertain and for whom there was an overlap in feeding types of >1 month. 101 RESULTS 4.7.2. Iron status in relation to use of low iron "milk" Table 4.15 shows the number of infants breast-fed for less than 7 months who had also been fed a low iron milk for at least a month from birth, 1 -3 months or from 6-8 months, within each iron status group. A statistically significant association (p<0.05) was found between feeding with low iron milk and iron status, with iron status as the dependent variable. Table 4.15. Number of infants with a history of feeding low iron "milk" by iron status. Iron status n # infants introduced to low iron milk # infants breast-fed <7 mths % infants breast-fed <7 mths fed low iron milk during 3 mths preceding testing birth 1-3 mths 4-5 mths 6-8 mths iron-deficient anaemic 30 1 2 1 5 16 31.3 Low iron status 106 6 19 11 8 78 10.3 Non-anaemic iron sufficient 254 37 40 22 21 212 9.9 Uncertain 43 5 3 21 5 35 14.3 mths, months Significant association between iron status and feeding low iron formula (p<0.05). Low iron "milk" includes cow's milk, goat's milk, or low iron infant formula. 102 RESULTS 4.7.3. Iron status in relation to the age of introduction of various foods Table 4.16 shows the number of infants for whom various foods had or had not yet been introduced by 9 months of age within the different iron status groups. There was no significant association between iron status and introduction of any of the foods: fruit juice, iron-fortified infant cereal, vegetables, fruits, egg yolk, legumes, tofu, fish, chicken or meat, with iron status as the dependent variable. Of note, 90% and 86.7% of the infants classified as iron-deficient anaemic had been introduced to iron-fortified infant cereal and fruit juice, respectively, by 9 months of age. The age of introduction of the various foods to the infants within each iron status group for those infants who were given the food is shown in Figures 4.7 to 4.16 (Appendix I). Table 4.17 (Appendix J) shows the mean age of introduction of each food item and fruit juice within each iron status group for those infants who were actually fed the food by 9 months of age. It should be noted that these calculations take no account of the infants who had not been introduced to the food, or the infants for whom the age of introduction of the food was not reported. The averages thus do not reflect the average age of introduction of the food to either the group of infants studied, or the infant population in general. 103 RESULTS Table 4.16. Iron status by number of infants for whom various foods have or have not yet been introduced by 9 months of age. Food Fruit juice Infant cereal Veg1 Fruit Egg Yolk Leg-umes Tofu Fish Chic-ken Meat introduced n 339 405 400 408 156 292 71 137 314 270 H not yet introduced n 81 17 16 12 250 111 325 271 105 144 Not known2 n 13 11 17 13 27 30 37 25 14 19 # introduced by iron status Iron deficient anaemic 30 n % 26 86.7 27 90.0 30 100 29 97.7 12 40.0 22 73.3 7 23.3 16 53.3 22 73.3 17 56.7 Low iron status 106 n % 80 75.5 102 96.2 100 94.3 101 95.3 41 38.7 75 70.7 13 12.3 32 30.2 77 72.6 66 62.3 Non-anaemic iron sufficient 254 n % 197 77.6 235 92.5 232 91.3 239 94.1 87 34.3 166 65.4 44 17.3 76 29.9 183 72.0 164 64.6 Uncertain 43 n % 36 83.7 41 95.3 38 88.4 39 90.7 16 37.2 29 67.4 7 16.8 13 30.2 32 74.4 23 53.5 No significant association between iron status and introduction of any one food n=433 %, percent of infants introduced to food within each iron status group. 1 vegetables 2 unknown whether infant introduced to food. 104 RESULTS 4.8. IRON STATUS AND SOCIOECONOMIC FACTORS Data on occupation of the parents was not collected. Socioeconomic status, as used in this study, relates to the parents' education and income. 4.8.1. Relationship between iron status and maternal education For the purposes of this study, the relationship between parental education and iron status was only assessed in relation to maternal education (Table 4.18). There was no statistically significant association between the iron status of infants and the mothers' level of education (p = 0 . 1 2 ) , with iron status as the dependent variable. 1 0 5 RESULTS Table 4.18. Infants' iron status in relation to the mothers' reported level of education. Iron status Highest level of education attained by mother Incomplete High College/Vocational University high school school only training n 33 78 142 173 Iron deficient 30 n 2 2 8 17 anaemjc % 6.1 2.6 5.6 9.8 Low iron 106 n 9' 12 35 49 status % 27.3 15.4 24.6 28.3 Non-anaemic 254 n 19 54 84 92 iron sufficient % 57.6 69.2 59.2 53.2 Uncertain 43 n 3 10 15 15 % 9.1 12.8 10.6 8.7 %, percent within each level of education 106 RESULTS 4.8.2. Relationship between iron status and annual family income Information on what is considered low income in Vancouver was obtained by personal communication from Statistics Canada. According to Statistics Canada, low income in Vancouver is assessed based on family size. In 1994, low income for 1 individual was classified as an income less than $16,511 /year, and for 2 individuals as less than $20,00C7year. Based on this information, family income brackets as indicated on the questionnaire (Appendix D) were combined into two categories: a family income of < $20,000/year and st $20,000/year, the former representing low income families. Chi-square analysis showed no significant association between a family income of < $20,000 or > $20,000 per year and iron status (p = 0.76), with iron status as the dependent variable, for the whole group of infants. The data on iron status and family income were also analyzed separately for two adult and one adult families (Table 4.20 & 4.21, respectively). There was no significant association between the infants' iron status and family income for the 2 adult families. A statistically significant association, however, was found between the infants' iron status and family income for the one adult families (p<0.05), with iron status as the dependent variable. However, the results cannot be over-interpreted because the number of single adult families surveyed is small (n = 30), and only 7 single adult families with an annual income > $20,000/year were studied. This makes it difficult to determine the strength of the apparent association between iron status and income. 107 RESULTS Table 4.19. Prevalence of iron-deficiency anaemia and low iron status in relation to annual family income Iron status Family income ($/yr) n < 20,000 86 S: 20,000 323 Iron deficient anaemic 30 n 7 21 % 8.1 6.5 Low iron status 106 n 21 82 % 24.4 25.4 Non-anaemic iron 255 n 47 190 sufficient % 54.7 58.8 Uncertain 43 n 11 30 % 12.8 9.3 1 Family income was unknown for the parents of 2 infants with iron-deficiency anaemia, 3 infants with low iron status, 18 infants with normal iron status and 2 infants classified as uncertain (n = 25) %, percent within each income level; yr, year 108 RESULTS Table 4.20. Prevalence of iron-deficiency anaemia and low iron status in relation to annual family income from two adult families Iron status Two adult family income ($/yr) < 10,000 10,000- 20,000 - 30,000 - > 50,000 19,999 29,999 50,000 n 19 44 62 106 145 Iron deficient n 0 4 6 4 11 anaemic % 0.0 9.1 9.7 3.8 7.6 Low iron n 4 13 12 26 40 status % 21.0 29.5 19.4 24.5 27.6 Non-anaemic n 10 21 38 65 83 iron sufficient % 52.6 47.7 61.3 61.3 57.2 Uncertain n 5 6 6 11 11 % 26.3 13.6 9.7 10.4 7.6 n = 376, family income not reported by 16 of the two adult families. %, percent within each income bracket; yr, year 109 RESULTS Table 4.21. Prevalence of iron-deficiency anaemia and low iron status in relation to annual family income from one adult families Iron status One adult family income <$/yr) < 10,000 10,000- 20,000- 30,000- > 50,000 19,999 29,999 50,000 n 5 18 3 3 1 Iron deficient n 2 1 0 0 0 anaemic % 40.0 5.5 0.0 0.0 0.0 Low iron n 0 4 2 2 0 status % 0.0 22.2 66.6 66.6 0.0 Non-anaemic n 3 13 0 1 0 iron sufficient % 60.0 72.2 0.0 33.3 0.0 Uncertain n 0 0 1 0 1 % 0.0 0.0 33.3 0.0 100.0 Significant association between iron status and family income from one adult family (P<0.05) n = 30, family income not reported by 4 of the one adult families. %, percent within each income bracket; yr, year 110 RESULTS 4.9. IRON STATUS AND ETHNIC BACKGROUND 4.9.1. Infants iron status in relation to whether or not the mother was born in Canada Table 4.22 shows the prevalence of iron-deficiency anaemia and low iron status among infants of immigrant and non-immigrant mothers. There was a statistically significant association between iron status and whether or not the mothers had been born in Canada (p<0.05), with iron status as the dependent variable. These results suggest a higher risk for iron-deficiency anaemia and low iron status among infants' whose mother had been born in Canada. 4.9.2. Relationship between iron status and number of years immigrant mother has been resident in Canada As results suggested an increased risk for iron-deficiency anaemia among infants whose mothers had been born in Canada, the data were further analyzed to see if there was a relationship between iron status and the number of years an immigrant mother had been resident in Canada. There was no significant difference in the distribution of iron status among infants of immigrant mothers who had been resident in Canada for a: 5 years compared to infants of mothers who had been resident in Canada for < 5 years (Table 4.23). This suggests that the number of years a mother resided in Canada did not increase the risk of her infant having iron-deficiency anaemia or low iron status. 111 RESULTS Table 4.22. Prevalence of iron-deficiency anaemia and low iron status in relation to whether of not the mother was born in Canada Iron status Mother born in Canada1 Yes No Not known n 204 182 48 Iron deficient anaemic 30 n 20 8 2 % 9.8 4.4 4.2 Low iron status 106 n 58 40 8 % 28.4 21.9 16.7 Non-anaemic iron 254 n 106 115 34 sufficient % 52.0 63.2 70.8 Uncertain 43 n 20 19 4 % 9.8 10.4 8.3 Significant association between iron status and mother born or not born in Canada (P<0.05) 1 Mother born or not born in Canada was determined by comparing the number of years the mother reported having lived in Canada and her age. %, percent of infants within each group; mother born or not born in Canada and not known. 112 RESULTS Table 4.23. Prevalence of iron-deficiency anaemia and low iron status in relation to the number of years an immigrant mother has been resident in Canada Iron status n Number of years immigrant mother has been resident in Canada ^5 Years 80 > 5 Years 102 Iron deficient n 4 4 anaemic % 5.0 3.9 Low iron status n 11 8 % 13.8 7.8 Non-anaemic iron n 52 63 sufficient % 65.0 61.8 Uncertain n 13 27 % 16.3 26.5 No significant association between iron status and number of years an immigrant mother has been in Canada %, percent of infants within each group of years an immigrant mother has been resident in Canada- s 5 years or >5 years. ' 113 RESULTS 4.9.3. Relationship between mothers' ethnic association and the infants' iron status The data were also analyzed to see if there was any relationship between the infants' iron status and the parents' ethnic background. A statistically significant association (p < 0.05) was found between iron status and ethnic background, with iron status as the dependent variable (Table 4.24). Specifically, the prevalence of iron-deficiency anaemia and low iron status was highest among infants born to mothers with a European background (36.4%) (British/NW European, East European, Mediterranean; Caucasians mothers) than among infants of Chinese (3.7%) or East Indian (0.0%) background; Non-Caucasian mothers. Indeed, the highest prevalence of iron-deficiency anaemia and low iron status was among the infants of mothers with an East European background (14.8%). No infants with an East Indian mother had iron-deficiency anaemia. The next lowest prevalence of iron-deficiency anaemia was found among infants of Chinese mothers (3.7%). Appendix L shows the duration of breast-feeding among Caucasian (Canadian/American and European mothers) versus non-Caucasian mothers (Chinese, East Indian, Filipino, other Asian and other). A significantly higher number of Caucasian than non-Caucasian mothers breast-fed their infants, 91.6% versus 56.8%, respectively. Among Caucasian mothers, 29.3% breast-fed their infants for >8 months. In contrast, only 6.5% of the non-Caucasian mothers were still breast-feeding by 8 months. 114 RESULTS Table 4.24. Prevalence of iron-deficiency anaemia and low iron status in relation to mothers' ethnic association l r o n s t a t u s Ethnic group Can/ Am Brit/ NW Eur E. Eur Med Chin ese Fili-pino E. Indian Other Asian Afri-can Oth n 120 66 27 32 81 18 43 23 10 13 Iron deficient anaemic 30 n % 6 5.0 6 9.1 4 14.8 4 12.5 3 3.7 2 11.1 0 0.0 2 8.7 1 10.0 2 15.4 Low iron status 106 n % 35 29.2 18 27.3 10 37.0 8 25.0 10 12.3 2 11.1 11 25.6 4 17.4 4 40.0 4 30.8 Non anaemic iron 254 n % 70 58.3 34 51.5 10 37.0 16 50.0 61 75.3 12 66.7 27 62.8 15 65.2 4 40.0 5 38.5 sufficient Uncertain 43 n % 9 7.5 8 12.1 3 11.1 4 12.5 7 8.6 2 11.1 5 11.6 2 8.7 10 10.0 2 15.4 Significant association between iron status and mothers' ethnic background (p<0.05) n=433 %, percent infants within each ethnic grouping Can/Am; Canadian/American, Brit/NW Eur; British NW European, English, Scottish, Irish, German, Dutch, French, Norwegian, Swedish or Danish; E. Eur; East European, Ukrainian, Polish, Croatian, Latvian, Yugoslavian, Russian, Czech, Romanian or Hungarian; Med; Mediterranean, Greek, Italian, Spanish or Portuguese; Other Asian, Japanese, Korean or Vietnamese; Oth; Other, Jewish, Arabic, Iranian, Palestinian, Egyptian, North American Indian or El Salvadorian. 115 4.10. IRON STATUS AND PERFORMANCE ON THE VISUAL RECOGNITION MEMORY TEST There were no statistically significant differences between the scores on the visual recognition memory test (Fagan Test of Infant Intelligence) among infants in the different iron status groups, as shown by a one way analysis of variance and the Tukey and Scheffe tests (Table 4.25). Table 4.25. Fagan test results among infants in the different Iron status n Iron deficient anaemic 23 iron status groups Mean score Standard deviation 62.1 5.4 Low iron status 91 60.4 6.8 Non-anaemic iron sufficient 196 60.4 6.3 Uncertain 41 59.5 7.2 n = 351 z?z£zzzrs ,he averaoe ,Mt ra ,o'a"in,anB *°' , m b •*« -*h «*«»"—«** 116 DISCUSSION Chapter V 5. DISCUSSION Despite advances in the understanding of infant nutrient requirements, iron-deficiency anaemia is still a common nutritional deficiency among infants and young children world wide. The main aim of this study was to assess the iron status of infants 9 months of age in Vancouver, British Columbia. Another aim of the study was to determine which groups of infants are at risk for iron-deficiency anaemia and low iron status based on their feeding history, and the economic and ethnic background of the family. The following section will discuss the main findings pertaining to each of the hypotheses. 5.1. PREVALENCE OF IRON-DEFICIENCY ANAEMIA AND LOW IRON STATUS. There is no reported information on the prevalence of iron-deficiency anaemia and low iron status among infants in British Columbia. One of the objectives of this study, therefore, was to assess the prevalence of iron-deficiency anaemia and low iron status among a group of infants 9 months of age in Vancouver. The prevalence of iron-deficiency anaemia among the group of 434 otherwise healthy 9 month old infants studied was 6.9%, as defined by a hemoglobin (Hgb) of < 101 g/L or Hgb <; 110 g/L with at least 2 other abnormal indicators of iron status (serum ferritin < 10 //g/L, zinc protoporphyrin >70//mol ZPP/mol heme and total iron binding capacity > 60//mol/L). A further 24.4% of the infants had low iron status, defined as a serum ferritin ^10 //g/L without iron-deficiency anaemia. Another 9.9% of the infants had low Hgb levels (Hgb > 101 117 DISCUSSION -^110 g/L) without definite low iron status (i.e. serum ferritin £ 10 //g/L). Reasons for this are not certain. Some of these infants may have been iron-deficient with borderline serum ferritin levels (serum ferritin > 10//g/L - s 12//g/L), or they may have had low iron status with raised serum ferritin due to recent infection (Elin et al.,1977). It may also be that these infants have physiologically low normal Hgb values with no hematological abnormality. Whether infants with low iron stores are at increased risk for iron-deficiency anaemia at older ages, or whether low iron stores are a normal physiological course in infants that is corrected as the dietary variety and amounts of solid food consumed increases through early childhood is not known. The prevalence of iron-deficiency anaemia among the infants in the current study is more than double that of 3% reported for infants and children in the United States. The prevalence of iron-deficiency anaemia in infants and young children in the United States has declined from 7.8% in 1975 to 2.9% in 1985 in infants and children from 6-60 months of age (Yip et al., 1987a&b). This decline has been attributed in part to the special supplement food program for Women, Infants and Children (WIC program) which provides iron-fortified formulas and cereals for the infants. Children in the United States also undergo periodic screening for anaemia, making possible early diagnosis and treatment of iron-deficiency (Miller et al., 1985; Vasques-Seone et al., 1985). Direct comparison of the prevalence rate of iron-deficiency anaemia in the United States with the prevalence of iron-deficiency anaemia found in the current study, however, must be cautious. This is because the infants in this study were tested at 9 months of age, but the reports from the United States include infants and children from 6-60 months of age, without indication of the relative rates of iron-deficiency anaemia within each individual age group. It seems possible that older children, for example 118 DISCUSSION at 60 months of age whose diets are more varied may have a lower risk of iron-deficiency anaemia than infants of 9 months. Canada has no food supplement programs equivalent to WIC and there are currently no screening programs to provide early detection of iron-deficiency anaemia. However, the Canadian Task Force on the Periodic Health Examination does recommend screening of infants at high risk for iron-deficiency anaemia at 9 months of age. Infants considered at high risk are those with delayed introduction of iron containing foods, those fed low iron formula or cows' milk as the primary source of milk, and those with poor weight gain (Blatherwick, 1995, Canadian Paediatric Society Nutrition Committee, 1991). Nevertheless, despite the recommendation for screening, physicians do not routinely see or assess the iron status of 9 month old infants. In the present study, 6.9% of the otherwise healthy infants screened at 9 months of age had iron-deficiency anaemia. This does indicate enough concern to develop strategies for the early detection of iron-deficiency anaemia. Possibly, greater emphasis on screening infants identified as high risk for iron-deficiency anaemia, as recommended by the Task Force, could help with diagnosis and hence, treatment and prevention of longer standing iron-deficiency anaemia. However, the cost/benefit ratio of screening programs must be weighed, especially since many of the laboratory tests are expensive. Nonetheless, screening infants defined as high risk for iron-deficiency anaemia seems justifiable in light of the consistent evidence that some of the effects of iron-deficiency anaemia on the developing central nervous system may be irreversible (Walter et al., 1989; Lozoff et al., 1988). 119 DISCUSSION 5.2. FEEDING PRACTICES The feeding practices believed to place infants at risk for iron-deficiency anaemia include: 1) feeding of low iron formula for more than 4 months from birth without introduction of high iron foods, 2) breast-feeding for more than 6 months without the introduction of iron-fortified infant cereals or supplemental iron, and 3) feeding cows' milk as the primary milk source before 9-12 months of age (Canadian Paediatric Society Nutrition Committee, 1991). One objective of this study was to determine the feeding history associated with iron-deficiency anaemia and low iron status among infants at 9 months of age in Vancouver. The null hypotheses were that there would be no association between iron status and duration of breast-feeding, weaning to cows' milk, use of low iron formula, introduction of an iron-fortified cereal from 4-6 months of age or other dietary sources of iron, such as meat from 7-9 months of age. 5.2.1. Duration of breast-feeding and prevalence of iron-deficiency anaemia and low iron status. The null hypothesis with regards to duration of breast-feeding was rejected. Iron-deficiency anaemia and low iron status were both more prevalent among infants who had received breast-milk as the exclusive source of milk for more than 8 months (15.2% and 30.4%, respectively) than among infants who were never breast-fed (1.5% and 10.3%, respectively). Breast-milk is thought to protect infants under 6 months of age from iron-deficiency anaemia because of the high bioavailability of the iron in breast-milk. However, the amount of iron supplied by breast-milk is not adequate to support the iron requirements of the rapidly growing infant over 6 months of age (Dallman et al., 1989). Therefore, it is 120 DISCUSSION recommended that iron-fortified cereals or supplemental iron be provided to infants who are breast-fed for more than 6 months (Canadian Paediatric Society Nutrition Committee, 1991). Pizarro et al. (1991) found similar prevalence rates of iron-deficiency anaemia and low iron stores among infants to those found in this study. The researchers found a prevalence of iron-deficiency anaemia of 14.7% and a prevalence of low iron status of 26.5% among 9 month old infants fed human milk as the only milk. In contrast, those infants fed an iron-fortified formula had a lower prevalence of iron-deficiency anaemia (0.6%) and low iron stores (12.3%) (Pizarro et al., 1991). Calvoet al. (1992) also found a significantly higher prevalence of iron-deficiency anaemia (27.8%) among 9 month old infants who had received breast-milk as the main source of milk to 6 months of age than among infants who had been fed iron-fortified formula (7.1%). The infants in the studies of Pizarro et al. (1991) and Calvo et al. (1992) had been introduced to solid foods at appropriate ages (between 4-6 months of age), however, the solid foods were not iron-fortified. A high prevalence of iron-deficiency anaemia among breast-fed infants has been attributed to weaning practices which do not include iron-fortified complementary foods (Calvo et al., 1992; Garry et al., 1981). Siimes and Salmenpera (1989) reported improvement in markers of iron status, such as serum ferritin and total iron binding capacity (TIBC), although not in hemoglobin, in 7.5 or 9 month old breast-fed infants introduced to solid foods when compared to infants who continued to be exclusively breast-fed. This is consistent with the understanding that the iron content of human milk becomes limiting for infants over six months of age (Canadian Paediatric Society Nutrition Committee, 1991). It seems plausible that the high prevalence of iron-deficiency anaemia and low iron status among infants in the current study who were breast-fed for more than 8 months may 121 DISCUSSION have been due to delayed introduction of iron containing foods (e.g. iron-fortified infant cereals) or foods which enhance the absorption of iron (e.g. vitamin C from fruit juice). However, the results show that the high prevalence of iron-deficiency anaemia and low iron status among these infants can not be explained by delayed introduction of iron-fortified infant cereals, or vitamin C in the form of fruit juice. In a recent study, Walter et al. (1993) concluded that the use of iron-fortified infant cereal can play a major role in the prevention of iron-deficiency anaemia in infants and young children. The study used criteria similar to the current study to define iron-deficiency anaemia (Hgb < 110 g/L with 2 or 3 additional abnormal measures of iron status). The researchers found a prevalence of iron-deficiency anaemia of 3% at 8 months of age among breast-fed infants receiving a fortified infant cereal compared to a prevalence of 15% among breast-fed infants fed an unfortified cereal. These results clearly show that feeding with an iron-fortified cereal reduces the risk of iron-deficiency anaemia among young infants, although it does not prevent it completely. In the current study, despite age appropriate introduction of an infant fortified cereal, iron-deficiency anaemia occurred in 15% of the infants who were breast-fed for more than 8 months. Reasons for this are not certain. It is important to note that the bioavailability of the electrolytic iron used in the fortification of infant cereals has been questioned (Fomon, 1987). Other possible explanations may relate to the quantity of cereal consumed by the infants. Breast-fed infants in the study by Waiter et al. had a mean intake of 25 g of rice cereal (fortified with 55 mg of electrolytic iron/100 g dry cereal) per day. In the current study, data on the quantity of iron-fortified cereal received by individual infants was not collected. Rather, the objective was to identify the feeding practices associated with a high prevalence of iron-deficiency anaemia and low iron status among 9 month old infants 122 DISCUSSION living in Vancouver, British Columbia. 5.2.2. Feeding of low iron "milk" and iron status. The null hypothesis of no association between weaning to cows' milk, or a low iron infant formula was also rejected. A significant association was found between iron status and use of low iron "milk"3 for the group of infants who had been breast-fed for 7 months or less and fed a low iron milk for at least one month. Other studies have also reported a higher prevalence of iron-deficiency anaemia and low iron stores in 6-12 months old infants fed an unfortified cow milk protein based formula (low iron formula) or whole cows' milk than among infants who were fed an iron-fortified infant formula (Pizarro et al., 1991; Fuchs et al., 1993). There appears to be no published information on the absorption of iron from foods providing a source of iron in infants over 6 months fed cows' milk or low iron infant formula. It is plausible that certain characteristics of cows' milk, such as its high protein, calcium and phosphate, and low ascorbic acid content may compromise the absorption of iron from food (Barton et al., 1983). Reports in the literature show that even adequate supplemental feedings may not supply enough iron for infants fed whole cows' milk (Penrod et al., 1990; Tunnessen & Oski, 1987). Indeed, Fuchs et al. (1993a&b) demonstrated that despite the consumption of iron-fortified cereal, infants of 6-12 months of age fed whole cows' milk were at risk for low iron stores, but not anaemia. This was the case even though these infants were receiving adequate intakes of iron from iron-fortified infant cereal and vitamin C in the form of fruit juice. The selection of cows' milk for infant feeding may be based on factors such as its low 3 This includes cows' milk, low iron infant formula, or goats' milk. 123 DISCUSSION cost in comparison to the cost of infant formula, or possibly lack of knowledge regarding its low iron bioavailability. The CPS currently recommends that cows' milk not be introduced to infants until at least 9-12 months of age. The results of this study are consistent with this recommendation. 5.2.3. Introduction of solid foods and prevalence of iron-deficiency anaemia and low iron status. The CPS recommends that solids not be introduced to infants until 4 to 6 months after birth. Results of the current study indicate that most mothers do not introduce solids until 4 months of age consistent with current CPS recommendations on infant feeding practices. Nonetheless, it was noted that a large proportion of mothers had not introduced dietary sources of iron such as meat, fish, chicken and legumes, 33.2%, 62.4%, 24.4%, and 25.6%, respectively, to their infants by the recommended age of 7-9 months. No significant associations were found between iron status and the age of introduction of specific foods (e.g. iron-fortified cereal, meat, fish, chicken, vegetables, fruits or tofu) to the infants. There are, however, factors not addressed by this study that should be considered. First, the composition of the diet as a whole and how it will affect the absorption of iron should be taken into consideration. For example, meat and vitamin C enhance the absorption of non-heme iron whereas the polyphenols in many vegetables and phosvitin in egg yolk inhibit the absorption of non-heme iron (Cook & Monsen, 1976). Secondly, the quantity of iron and vitamin C ingested is also important to the infant's overall iron status. It should also be noted that mothers were asked to recall their infant's diet retrospectively from 9 months to birth. Studies have shown that the reliability of reported data on breast-feeding and 124 DISCUSSION age of introduction of solid food by mothers decreases over time (Persson and Carlgen, 1984; Rios et al., 1994). The possibility of recall bias in this study, thus may limit interpretation of cause and effect relationships between diet and iron status. Summary In summary, the results of this study show that the feeding practices associated with a high prevalence of iron-deficiency anaemia and low iron status among a group of 9 month old infants in Vancouver are the feeding of breast-milk as the main source of milk for more than 8 months, and the feeding of cows' milk or low iron infant formula. 5.3. SOCIAL, ECONOMIC, AND ETHNIC BACKGROUND 5.3.1. Socioeconomic status and prevalence of iron-deficiency anaemia and low iron status. The null hypothesis that there would be no association between family income and the prevalence of iron-deficiency anaemia or low iron status among a group of 9 month old infants in Vancouver was accepted. Other studies in Canada have found that infants and young children from low socioeconomic backgrounds are at higher risk for iron-deficiency anaemia than infants from high socioeconomic backgrounds (Feldman et al., 1985; Gary-Donald et al., 1990; Green-Finestone et al., 1991). For example, a study of 320 infants of 6 to 18 months of age from all social classes in Ottawa found an inverse relationship between socioeconomic status and iron-deficiency anaemia (Feldman et al., 1985). In the latter study, 8.8% of the infants in the low socioeconomic group had iron-deficiency anaemia compared to 1.7% and 2.7% in the high and middle socioeconomic groups, respectively. The lack of association between low family income (annual family income < $20,000) and iron status in 125 DISCUSSION the present study may be explained by the small number of infants from low income families surveyed. Alternatively, it may be that low income families now feed iron-fortified infant formula rather than low-iron formula or cows' milk. In contrast to the 2 adult families, the annual family income from one adult families was significantly associated with the infants' iron status. However, only a small number of infants from one adult families (n = 34) were surveyed. Of the 34 one adult families, 23 reported an annual income of < $20,000, and 7 had an annual income of > $20,000. Because of this small number of infants studied in this group, the apparent significant association between family income and iron status should be interpreted with caution. As participation in the study required that the parent have an address and telephone number to enable contact, a potential sampling bias against the one adult families who may more frequently change their addresses may have occurred. Thus, it is possible that the true prevalence of iron-deficiency anaemia among infants from single parent or low income families in Vancouver may not have been accurately estimated and may be higher than that suggested by this study. 5.3.2. Parental ethnic association and its relation to the infant's iron status. The null hypothesis that there is no association between parental ethnic association and iron status of the infant was rejected. Ethnic background, as reported by the mother, was significantly associated with the infants' iron status. Specifically, the prevalence of iron-deficiency anaemia and low iron status was higher among infants of European and Canadian parentage (Caucasians) than among infants of non-Caucasian mothers (Chinese and East Indian). Of note, no infants of East Indian parentage had iron-deficiency anaemia and the prevalence of iron-deficiency anaemia among infants of Chinese parentage was only 3.7% in 126 DISCUSSION contrast to the other ethnic groups. Contrary to our findings of a low prevalence of iron-deficiency anaemia among infants of Chinese parentage, Chan-Yip and Gray-Donald (1987) found a higher prevalence of iron-deficiency anaemia (12.1%), based on a hemoglobin rise of at least 10 g/L in response to treatment, among 346 Chinese infants 6-36 months old in Montreal. The prevalence of iron-deficiency anaemia in the group who were 6-12 months (n=166) was 11.4%; still much higher than the prevalence of 3.7% found among the 9 month old Chinese infants in the current study. The study in Montreal which was based on retrospective dietary recording, found that many of the infants were fed excessive amounts of cows' milk (the quantity consumed was not reported) and weaned to a traditional Chinese beikost (Congee; cooked rice porridge), both of which have low iron bioavailability. This may well explain the higher prevalence of iron-deficiency anaemia among the Chinese infants in Montreal compared to the Chinese infants in Vancouver in this study. In the current study, it is probable that the Chinese infants who were never breast-fed or who were weaned from breast-milk before 3 months of age were fed an iron-fortified infant formula. Studies have shown that the absolute amount of iron that is absorbed from iron-fortified formulas (iron concentration of 6-13 mg/L) is high enough to meet the infant's iron requirements (Dallman, 1990). Information on the prevalence of iron-deficiency anaemia among Blacks, Whites and Hispanics has been reported from studies in the United States. A higher prevalence of iron-deficiency anaemia has been reported among Black infants than among White infants in the United States (Yip et al., 1987a&b). Other ethnic groups, such as Asians have not been well studied. Interestingly, studies in England found iron-deficiency anaemia was a common problem in some ethnic minorities, especially in infants of Asian parentage; in particular, East 127 DISCUSSION Indian children (Aukett et al., 1986; Duggan et al., 1991; Erhardt, 1986; Grindulis et al., 1986; Mills, 1990). Erhardt (1986) reported a higher prevalence of both iron-deficiency anaemia (24.6%) and low iron stores (serum ferritin < 10//g/L) (32.8%) among children aged 6 months to 4 years from some Asian groups (Hindus, Sikhs, Moslems; East Indian) than from White children. This study, however, did not assess feeding practices. Another study in England also found iron-deficiency anaemia was more common among some non-Caucasian infant groups (Asians; East Indians and others; West African, South East Asian, Chinese, and mixed races) than among White children (Mills, 1990). Mills (1990) found that at all age ranges studied (8-11, 12-14, 15-17, 18-24 months), Asian children drank more cows' milk than White children or children from other groups. In the present study, iron-deficiency anaemia was not found in any of the 9 month old infants of East Indian parentage. It is probable, therefore, that these infants were fed an iron-fortified formula or were breast-fed for a shorter duration (less than 4 months) and then weaned to an iron-fortified formula. This may well explain the lack of iron-deficiency anaemia in the 9 month old East Indian infants in this group. It is important to note that it was not the intent of this study to identify which ethnic groups breast-feed and which feed formula to their infants. In the present study, the risk of iron-deficiency anaemia and low iron status was higher among infants of mothers born in Canada than among infants of mothers not born in Canada. A plausible explanation for the differences in the prevalence of iron-deficiency anaemia among infants of mothers born in Canada and immigrant mothers may be related to cultural differences in feeding practices. Several studies (Green-Finestone et al., 1989; Rassin et al., 1984) have shown that maternal ethnicity is a strong determinant of choice and duration of infant feeding practices, such as breast-feeding. Yeung et al. (1981) reported a higher rate 128 DISCUSSION of breast-feeding among Canadian Anglophones than Francophones, and among Europeans and Americans than Anglophone Canadians. Their study, however, did not report the rate of breast-feeding among other ethnic groups in Canada, such as Chinese or East Indians (Asians). Rassin et al. (1984) also found ethnic differences in the incidence of breast-feeding in the United States. They noted that the incidence of breast-feeding was higher among White women than Black or Hispanic women. Studies in the United States have indicated a dramatic decline in the rate of breast-feeding among Asian women after immigration to the United States (Ghaemei-Ahamadi, 1992; Eomero-Gwynn, 1989). Ghaemi-Ahamadi (1992) undertook interviews of 150 Persian and Southeast Asian immigrants who were WIC participants. He found that 95% of the women had exclusively breast-fed in their native country, with 85% doing so for 5 months or more. After immigrating to the United States, only 32% exclusively breast-fed and they did so for a shorter period. It is possible that the low prevalence of iron-deficiency anaemia among infants of Chinese and East Indian parentage in Vancouver is explained by a short duration of breast-feeding and use of an iron-fortified formula as the main source of milk for bottle-feeding. In summary, based on the results of this study, the 9 months old infants at highest risk for iron-deficiency anaemia in Vancouver are infants of Caucasian mothers (European and Canadian background), rather than infants of non-Caucasian mothers (Chinese or East Indian backgroud). This may well be explained by cultural differences in infant feeding practices, in particular the choice and duration of breast-feeding. Hence, as this study found that the duration of breast-feeding was significantly associated with a higher risk for iron-deficiency anaemia, the infants of Caucasian mothers may be at highest risk for iron-deficiency anaemia 129 DISCUSSION because of a higher incidence of longer breast-feeding than among infants of non-Caucasian mothers (i.e. East Indian and Chinese). Indeed, analysis of the data (Appendix L) shows that the duration of breast-feeding was significantly higher among Caucasian than non-Caucasian mothers. Among Caucasian mothers, 29.3% breast-fed their infants for >8 months of age. In contrast, however, only 6.5% of the non-Caucasian mothers were still breast-feeding by 8 months post-partum. 5.4. IRON STATUS AND COGNITIVE FUNCTION There seems to be no reported information on scores on the Fagan Test of Infant Intelligence (FTII) in infants in relation to iron status. Nonetheless, the test was chosen in preference to the widely used Bayley Scales of Infant development (BSID) because visual recognition scores, unlike scores on the BSID, are not influenced by factors such as, race, birth order or parental education (Fagan, 1988). In addition, the FTII is quick and simple to administer in comparison to the BSID. The null hypothesis that there would be no difference in test scores on the visual recognition memory test between iron-deficient anaemic infants and infants classified as iron-sufficient was accepted. This finding does not concur with published studies showing that iron-deficiency severe enough to cause anaemia also results in reduced performance on developmental tests in infancy (Grindulis et al., 1986; Lozoff et al., 1987; Lozoff et al., 1988; Walter et al., 1983; Walter et al., 1989). Several explanations for the lack of concurrence with published studies are possible. Unlike this study, the studies which have found lower developmental test scores in infants with iron-deficiency anaemia have used the Bayley Scales of Infant Development (BSID) or the Sheridan Developmental Sequence test. The lack of 130 DISCUSSION concurrence from this study, therefore, may be related to aspects of cognition that are affected by anaemia; the FTII may not measure them. Studies using the BSID have noted specific patterns of test failure in items that required language comprehension, but did not involve a visual demonstration among infants with iron-deficiency anaemia. Further, in the BSID psychomotor test, balance in the standing position and walking were accomplished by significantly fewer infants with iron-deficiency anaemia than by infants classified as iron-sufficient (i.e. normal iron status) (Lozoff et al., 1987; Walter et al., 1989). The FTII, in contrast to the BSID, measures attention/habituation, which involves basic intellectual processes such as encoding, abstraction, detection of invariant features and categorization (Fagan, 1982). The skills involved in the FTII thus differ from those measured in the BSID, which relies on the child's ability to perform certain motor as well as cognitive tasks (Fagan & McGrath, 1981; Fagan & Singer, 1983). Attention, one of the skills required for performance on the FTII has been argued to be necessary for successful performance on later intelligence tests (Fagan et al., 1983). In studies in preschool children, iron-deficiency anaemia has been associated with cognitive deficits, particularly in attention (Pollitt et al., 1986). Because the duration of the anaemia in the infants in the current study is unknown, it is possible that the infants may have been tested before the effects of iron-deficiency anaemia on cognitive functions such as, attention were functionally expressed, i.e. the duration of the anaemia was too short to have had any effect. 131 DISCUSSION 5.5. CONCLUSION The study set out to determine the prevalence of iron-deficiency anaemia and low iron status in group of 9 month old infants in Vancouver, British Columbia, and to determine which groups of infants are at highest risk for iron-deficiency anaemia based on their feeding history, economic and ethnic family background. The observed prevalence of iron-deficiency anaemia was 6.9% and the prevalence of low iron status was 24.4% among an otherwise healthy group of 9 month old infants who participated in this study. The findings of this research suggest the need to develop strategies for the early detection and hence, prevention of longer standing iron-deficiency anaemia. One strategy would be to place greater emphasis on the screening of infants at high risk for iron-deficiency anaemia at 9 months of age as recommended by the Canadian Task Force on the Periodic Health Examination. In practice, however, screening all infants at 9 months is not practical or cost effective. Screening for iron-deficiency anaemia should, therefore, target only those infants identified as being at highest risk for developing iron-deficiency anaemia. The iron status of these infants would be tested and those infants with a Hgb > 101 g/L and <110 g/L may be retested by measurement of the serum ferritin to distinguish infants with normal iron status from those with iron-deficiency anaemia. The results of this study also show that the infants at highest risk for iron-deficiency anaemia and low iron status in Vancouver, when defined by feeding history are infants who are breast-fed as the only source of milk for more than 8 months and infants who are bottle-fed with low iron milk (cows' milk, goats' milk or low iron infant formula). When defined by the ethnic background of the family, the infants at highest risk of iron-deficiency anaemia and low iron status are infants of Caucasian mothers compared to infants of non-Caucasian 132 DISCUSSION mothers (East Indian and Chinese). It is these infants, at least in Vancouver, for whom there is reason to suspect iron-deficiency. 5.6. RECOMMENDATIONS FOR FUTURE RESEARCH • The composition of the diet as a whole is important to the absorption of dietary iron (Charlton & Bothwell, 1983). Studies on the absorption of iron from the diet as a whole have been carried out predominately in adults. Future studies are needed to examine the effects of solid foods on the absorption of iron in relation to the main source of milk that infants are receiving. • Further investigations are still necessary to determine if breast-fed infants consume sufficient amounts of cereal on a continuous basis to meet the recommended iron intake of 7 mg/day (Health and Welfare Canada Nutrition Recommendations, 1990). These investigations are warranted as the results of this study do indicate an increased risk for iron-deficiency anaemia and low iron status among infants who are fed human-milk as the main source of milk for more than 8 months. • It is not known whether those infants classified as having low iron stores are at an increased risk of developing iron-deficiency anaemia at an older age. Further studies are needed to investigate this potential risk and to determine whether low iron status in 9 month old infants is corrected as the diet becomes more varied through early 133 DISCUSSION childhood. A longitudinal study following infants from 9 to 24 months of age could be used to investigate this. • It is unknown whether the increasing variety and amount of solid foods consumed by the infant in the second six months after birth is adequate to meet their requirements for maintenance of normal erythropoiesis. Consequently, quantitative data on intake of dietary iron in infancy and early childhood, in conjunction with hematologic assessment of iron status, should be obtained in future research. Any methods used to assess dietary intake would have to take account of the nutritionally beneficial, culturally specific foods that should be promoted in infant feeding, especially for those foods that may have a positive or negative impact on the infant's iron status. • Because inappropriate feeding practices and hence, poor nutrition are a the primary cause of iron-deficiency anaemia, another method of screening infants at risk for iron-deficiency anaemia could involve the use of dietary intake methodologies such as diet histories. A food frequency questionnaire, for example, can be used to help detect infants at risk for iron-deficiency anaemia due to inadequate intake of iron rich foods. Development of a valid dietary screening tool could provide a non-invasive, easy to administer and cost effective method of screening. 134 BIBLIOGRAPHY BIBLIOGRAPHY American Academy of Pediatrics (Committee on Nutrition). The use of whole cow's milk in infancy. Pediatr. 89:1105-1109, 1992. American Academy of Pediatrics (Committee on Nutrition). Iron supplementation for infants. Pediatr. 58:765-766, 1976. Aukett MA, Parks YA, Scott PH, and Wharton BA. Treatment with iron increases weight gain and psychomotor development. Arch. Dis. Child. 61:849-857, 1986. Barton JC, Conrad ME, and Parmley RT. Calcium inhibition of inorganic iron absorption in rats. Gasteroenterol. 84:90-101, 1983. Beard JL, Connor JD and Jones BC. Brain iron: Location and Function. Prog. Food and Nutr. Sci. 17:183-221, 1993. Ben-Shachar D, Youdim BMH, and Ashkenazi R. Long term consequences of early iron deficiency on dopaminergic neurotransmission in rats. Int. J. Dev. Neurosci. 4:81-88, 1986. Blumberg WE, Eisinger J, Lamola AA, and Zuckerman DM. The Hematofluorometer. Clin. Chem. 23:270-274, 1977. Blatherwick J. The need for iron. BCMA Newsletter. 17: 1995. Bothwell TH, Charlton RW, Cook JD, and Finch CA. Iron metabolism in Man.Blackwell Scientific Publications, Oxford, 1979 Brennan MM, Flyn A, and Morrissey PA. Absorption of iron and zinc from soya and cow's milk based infant formula in suckling rats. Proc. Nutr. Soc. 48:39A, 1989. Brown MK, and Picciano MF. Relative effectiveness of food sources of supplemental iron to the infant. Fed. Proc. 46:1160-1163, 1987. Calvo EB, Galindo AC, and Aspres NB. Iron status in exclusively breast-fed infants. Pediatr. 90:375-379, 1992. Canadian Paediatric Society Nutrition Committee. Meeting the iron needs of infants and young children. Can. Med. Assoc. 144:1451-1453, 1991. Canadian Paediatric Society Nutrition Committee. Infant feeding. Can J Public Health 70:376-385,1979. 135 BIBLIOGRAPHY Canadian Task Force on the Periodic Health Examination. The periodic health examination. Can. Mad. Assoc. J. 121:1193-1194, 1979. Chan-Yip A, and Gray-Donald K. Prevalence of iron-deficiency among Chinese children aged 6 to 36 months in Montreal. Can. Med. Assoc. J. 136:373-378, 1987. Chaney SG. Principles of nutrition II: Micronutrients. In Devlin TM (ed) Textbook of biochemistry, with clinical correlations (3rd edition). Wiley-Liss Inc, New York, pp 115-1145,1993. Charlton RW, and Bothwell TH. Iron absorption. Annu. Rev. Med. 34:55-68, 1983. Conrad ME, and Barton JC. Factors affecting iron balance. Am. J. Hematol. 10:199-225, 1981. Cook JD, Baynes RD, and Skine B$. Iron-deficiency and the measurement of iron status. Nutr. Res. Rev. 5:189-202, 1992. Cook JD, and Bothwell TH. Availability of iron from infant foods. In Stekel A (ed) Iron Nutrition in Infancy and Childhood. Raven Press, New York, pp 119-143, 1984. Cook JD, Lipschitz DA, Miles LEM, and Finch CA. Serum ferritin as a measure of iron stores in normal subjects. Am. J. Clin. Nutr. 27:681-687, 1974. Cook JD, Layrisse M, Martinez-Torres C, Walker R, Monsen E, and Finch CA. Food iron absorption measured by an extrinsic tag. J. Clin. Invest. 51:805-807, 1972. Dallmam PR, Yip R, and Oski FA. Iron-deficiency and related nutritional anaemias. In Nathan DG, and Oski FA (Eds.) Hematology of infancy and childhood (4th ed.). W.B. Saunders Company: Philadelphia, pp 413-449, 1993. Dallman PR. Progress in the prevention of iron-deficiency in infants. Acta. Peadiatr. Scand. Suppl. 365:28-34, 1990. Dallman PR. Review of iron metabolism. In Filer LR (ed) Dietary iron: Birth to two year. Raven Press Ltd, New York, pp 1-11, 1989. Dallman PR, Martti AS, and Stekel A. Iron-deficiency in infancy and childhood. Am. J. Clin. Nutr. 33:86-118, 1980. Dallman PR. Inhibition of iron absorption by certain foods. Am. J. Dis. Child. 134:453-454, 1980. Davidson L, Kastenmayer P, Yuen M, Ldnnerdal B, and Hurrel RF. Influence of lactoferrin on iron absorption from human milk in infants. PedSatr. Res. 35:117-124, 1994. 136 BIBLIOGRAPHY Dobbing J. Vulnerable periods in the developing brain. In Dobbing J (ed) In brain behaviour and iron in the infant. London: Springer Verlag, pp 1-17, 1990. Duncan B, Schifman RB, Corrigan JJ Jr, and Schaefer C. Iron and the exclusively breast-fed infant from birth to six months. J. Pediatr. Gasteroenterol. Nutr. 4:421-425, 1985. Ehrhardt P. Iron-deficiency in young Bradford children from different ethnic groups. Br. Med.J. 292:90-93, 1986. Elin RJ, Wolff SM, Finch CA. Effects of fever on serum iron and ferritin concentrations in man. £/o«/49:147-153, 1977. Ernst JA, Brady MS, and Richard KA. Food and nutrient intake of 6 to 12 month old infants fed formula or cow's milk: A summary of four nutritional surveys. J. Pediatr. 117:S86-S100, 1990. Fagan JF, and Montie JE. Behavioral assessment of cognitive well being in the infant. In Kavanagh JF (ed) Understanding Mental Retardation: Research accomplishments and new frontiers, Baltimore, pp 207-221, 1988. Fagan JF. Evidence for the relationship between responsiveness to visual novelty during infancy and later intelligence: A summary. Bum. Bull. Cognit. Psyc. 8(5) 469-475, 1988. Fagan JF, and Singer LT. Infant recognition memory as a measure of intelligence. In Lipsitt L.P (eds). Advances in Infancy Research (vol 2), Norwood, NJ: Ablex, pp 31-78, 1983. Fagan JF, and Mcgrath SK. Infant recognition memory and later intelligence. Intelligence 5:121-130, 1981. Fairweather-Tait SJ. Iron-deficiency in infancy: easy to prevent or is it ? Euro. J. Clin. Nutr. 46(suppl 4):S9-S14, 1992. Fairweather-Tait SJ. The concept of bioavailability as it relates to iron nutrition. Nutr. Res. 7:319-325, 1987. Feldman W, Greene-Finestone L, and Heick H. Do all infants need to be screened for anaemia? Clin. Invest. Med. 8(3):182A, 1985. Finch CA, and Huebers HA. Perspectives in iron metabolism. N. Eng. J. Med. 306:1520-1528, 1982. Fomon SJ. Bioavailability of supplemented iron in commercially prepared dry infant cereals. J. Pediatr. 110:660-661, 1987. 137 BIBLIOGRAPHY Fomon SJ, Ziegler EE, Nelson SE, and Edwards BB. Cow milk feeding in infancy: gastrointestinal blood loss and iron nutritional status. J. Pediatr. 98:540-545, 1981. Fuchs GJ, DeWier M, Hutchinson SW, Sundeen M, Schwartz S, and Suskind RM. Gastrointestinal blood loss in older infants: Impact of Cow's milk versus formula. J.Pediatr. Gasteroenterol. Nutr. 16:4-9,1993a. Fuchs GJ, Farris RP, Dewier M, Hutchinson SW, Warrier HD, Doucet H, and Suskind RM. Iron status and intake of older infants fed formula vs cow milk with cereal. Am. J. Clin. Nutr. 58:343-348, 1993b. Gabriel A, Gabriel KR, and Lawrence RA. Cultural values and biomedical knowledge: choices in infant feeding. Analysis of a survey. Soc. Sci. Med. 23:501-548, 1986. Gary-Donald K, Di-Tommaso S, and Leamann F. The prevalence of iron-deficiency anaemia in low income Montreal infants aged 10-14 months. J. Can. Diet. Assoc. 51:424A, 1990. Garry PJ, Owen GM, Hooper EM, Gilbert BA. Iron absorption from human milk and formula with or without iron supplementation. Pediatr Res. 15:822-828, 1981 Ghaemi-Ahmadi S. Attitudes toward breast-feeding and infant feeding among Iranian, Afghan, and Southeast Asian immigrant women in the United States: implications for health and nutrition education. J. Am. Diet. Assoc. 92:354-355, 1992. Gibson RS. Principles of Nutritional Assessment. Oxford University Press, Oxford pp 163-183, 349-372, 1990. Gitlin D, Kumate J, Urrust J, and Morales C. The selectivity of the human placenta in transfer of plasma proteins from mother to fetus. J. Clin. Invest. 43:138-140, 1964. Greene-Finestone L, Feldman W, Heick H and Luke B. Prevalence and risk factors of iron depletion and iron deficiency anaemia among infants in Ottawa-Carleton. J Can Dietetic Assoc. 52:20-23, 1991. Greene-Finestone L, Feldman W, Heick H, and Luke, B. Infant-feeding practices and socio-demographic factors in Ottawa-Carlton. Can. J. Public Health 80:173-176, 1989. Grindulis H, Scott PH, Belton NR, and Wharton BA. Combined deficiency of iron and vitamin D in Asian toddlers. Arch. Dis. Child. 61:843-848, 1986. Jacobson SW, Jacbson JL, and Frye KF. Incidence and correlates of breast-feeding in socioeconomically disadvantaged women. Pediatr. 88:728-736, 1991. Hallberg L. Bioavailability of dietary iron in man. Annu. Rev. Nutr. 1:123-147, 1981. 138 BIBLIOGRAPHY Hallberg L, and BjOrn-Rasmussen E. Determination of iron absorption from whole diet. A new two-pool model using two radioiron isotopes given as heme and non-heme iron. Scand. J. Hemeatol. 9:193, 1972. Health and Welfare Canada. Present patterns and trends in infant feeding in Canada. Ottawa: Ministry of Supply and Services Canada, 1990. Health and Welfare Canada. Nutrition Recommendations. The Report of the Scientific Review Committee. Ottawa:Supply and Services Canada, pp 204, 1990. Hertrampf E, Cayazzo M, Pizarro MT, and Stekel A. Bioavailability of iron in soy-based formula and its effect on iron nutriture in infancy. Pediatr. 78:640-645, 1986. Heubers HA, and Finch CA. Transferrin: Physiologic behaviour and clinical implications. Blood 64:763-765, 1984. Hill PD. Predictors of breast-feeding duration among WIC and non-WIC mothers. Public Health Nurs. 8:46-52, 1991. Hurell RF. Bioavailability of different iron compounds used to fortify formulas and cereals: Technological problems. In Stekel A (ed). Iron nutrition in Infancy and Childhood. Raven Press, New York, pp 147-175, 1984. Idjradinata P, and Pollitt E. Reversal of developmental delays in iron-deficient anaemic infants treated with iron. Lancet 341:1-4, 1993. Lamola AA, and Yamane T. Zinc protoporphyrin in erythrocytes of patients with lead intoxication and iron-deficiency anaemia. Science 186:936-938, 1974. Lonnerdal B and Iyer S. Lactoferrin: Molecular structure and biological function. Annu. Rev. Nutr. 15:93-110, 1995. Lonnerdal B. Iron and breast milk. In: Sketal A (ed) Iron Nutrition in infancy and childhood. Raven Press, New York, pp 94-114, 1984. Lozoff B, Jimenze E, and Wolf W. Long term developmental outcome of infants with iron deficiency. N. Engl. J. Med. 325:687-694, 1991. Lozoff B. Behavioral alterations in iron-deficiency. Adv. Pediatr. 35:331-360, 1988. Lozoff B, Brittenham GM, Wolf AW, McClish DK, Kuhnert PM, Jimenez E, Jimenez R, Mora LA, Gomez I, and Krauskoph D. Iron-deficiency anaemia and iron therapy effects on infant developmental test performance. Pediatr. 79:981-995, 1987. 139 BIBLIOGRAPHY Lozoff B, Brittenham GM, Viteri FE, Wolf AW, and Urrutia JJ. The effects of short-term oral iron therapy on developmental deficits in iron-deficient anaemic infants. J. Pediatr. 100:351-357, 1982. Massey AC. Microcytic anaemia: Differential diagnosis and management of iron-deficiency anaemia. Med. din. N. Am. 76(3): 549-567, 1992. McMillan JA, Oski FA, Lourie G, Tomarelli RM, and Landaw SA. Iron absorption from human milk, simulated human milk and proprietary formulas. Pediatr. 60:896-900, 1977. McNally E, Hendricks S, and Horowitz I. A look at breast-feeding trends in Canada (1963-1982). Can. J. Public Health 76:101-107, 1985. Miles LEM, Lishitz DW, Bieber CP, and Cook JD. Measurement of serum ferritin by a 2-site immunoradiometric assay. Anal. Biochem. 61:29-224, 1974. Miller V, Swaney S, and Deinard AS. Impact of the WIC program on the iron status of infants. Pediatr. 75:1619-1623, 1985. Nutrition Canada. Anthropometry Report-Height, Weight and Body Dimensions (Bureau of National Sciences, Health Protection Branch, Health and Welfare , Ottawa), 1980. Oski FA. Iron-deficiency in infancy and childhood. AY. Engl. J. Med. 329:190-193, 1993. Oski FA. The causes of iron-deficiency in infancy. In Filer LR (ed) Dietary iron: Birth to two years. Ravan Press Ltd, New York, pp 63-70, 1989. Oski FA, and Landaw SA. Inhibition of iron absorption from human milk by baby food. Am. J. Dis. Child. 134:459-460, 1980. Oski FA, and Hoing AS. The effects of therapy on the developmental scores of iron-deficient infants. J. Pediatr. 92:21-25, 1978. Penrod C, Anderson K, and Acosta PB. Impact on iron status of introducing cow's milk in the second six months of life. J. Pediatr. Gastroenterol. Nutr. 10:462-467, 1990. Peters TJ, Raja KB, Simpson RJ, and Snapes. Mechanisms and regulation of iron absorption. Ann. N.Y. Acad. Sci. 526:141-147, 1988. Pizarro F, Yip R, Dallman PR, Olivares M, Hertrampf E, and Walter T. Iron status with different infant feeding regimens: Relevance to screening and prevention of iron deficiency. J Pediatr. 118:687-692, 1991. Pollitt E, Saco-Pollitt C, Liebel R, and Viteri F. Iron-deficiency and behavioural development in infants and preschool children. Am J Clin Nutr. 43:555-565, 1986. 140 BIBLIOGRAPHY Rassin DK, Richardson CJ, Baranowski T, Nader PR, Guenther N, Bee DE, and Brown JP. Incidence of breast-feeding in a low socioeconomic group of mothers in the United States: ethnic patterns. Pediatr. 73:132-137, 1984. Rios E, Lipschitz DA, Cook JD, and Smith NJ. Relationship of maternal and infant iron stores as assessed by determination of plasma ferritin. Pediatr. 55:694-699, 1975a. Rios E, Hunter RE, Cook JD, Smith NJ, and Finch CA. The absorption of iron as supplements in infant cereal and infants formulas. Pediatr. 55:686-693, 1975b. Romero-Gwynn E. Breast-feeding pattern among Indochinese immigrants in Northern California. Am. J. Dis. Child. 143:804-808, 1989. Rosa G, and Trugo NMF. Iron uptake from lactoferrin by intestinal brush-border membrane vesicles of human neonates. Brazilian J. Med. Biol. Res. 27:1527-1531, 1994. Rossiter JC. Attitudes of Vietnamese women to baby feeding practices before and after immigration to Sydney, Australia. Midwifery 8:103-112, 1992. Saarinen UM, and Siimes MA. Iron absorption from breast milk, cow's milk and iron-supplemented formula: An opportunistic use of changes in total body iron determined by hemoglobin, ferritin and body weight in 132 infants. Pediatr. Res. 13:143-147, 1979. Saarinen UM, and Siimes MA. Iron absorption from infant formula and the optimal level of iron supplementation. Acta. Pediatr. Scand. 66:719-721, 1977. Saarinen UM, Siimes MA, and Dallman PR. Iron absorption in infants: high bioavailability of breast-milk iron as indicated by the extrinsic tag method of iron absorption and the concentration of serum ferritin. J. Pediatr. 91:36-39, 1977. Singla PN, Chand S, Khanna S, and Agarwal KN. Effect of maternal anaemia on the placenta and newborn infant. Acta. Paediatr. Scand. 67:645, 1978. Siimes MA, Salmenpera L, and Perheentupa J. Exclusive breast-feeding for 9 months: Riskb iron-deficiency. J. Pediatr. 104:196-199, 1984. Siimes MA and Salmenpera L. The weanling: Iron for all or one. Acta Paediatr Scand Suppl. 361:103-110,1989. Stekel A. Iron requirements in infancy and childhood. In Stekel A (ed) Iron Nutrition in Infancy and Childhood. Raven Press, New York, pp 1-7, 1984. Tanka PA, Yeung DL, and Anderson GH. Infant feeding practices 1984-85 versus 1977-78. Can. Med. Assoc. J. 136:940-944, 1989. 141 BIBLIOGRAPHY The Vancouver Health Department. Position on infant feeding for the first year of life. November, 1993. Thompson LA, Fagan JF, and Fuller DW. Longitudinal prediction of specific cognitive abilities from infant novelty preference. Child. Dev. 62:530-538, 1991. Trundle DS. Clin Chem News pp 8 Aug, 1984. Tunnessen WW. Jr, and Oski FA. Consequences of starting whole cow milk at 6 months of age. J. Pediatr. 111:813-816, 1987. Vazquez-Seoane P, Windom R, and Pearson HA. Disappearance of iron-deficiency anaemia in a high risk infant population given supplemental iron. N. Engl. J. Med. 313:1239-1240, 1985. Walter T, Dallman PR, Pizarro F, Velozo L, Pefta G, Bartholmey SJ, Hertrampf C, Olivares M, Letetier A, and Arrendondo M. Effectiveness of iron-fortified infant cereal in prevention of iron deficiency anaemia. Pediatr. 91:976-982, 1993. Walter T, deAndraca I, Castillo F, Rivera T, and Cobo C. Cognitive effect at 5 years of age in infants who were anaemic at 12 months: A longitudinal study. Pediatr. Res. 28:295A, 1990. Walter T, deAndraca I, Chadud P, and Perales CG. Iron-deficiency anaemia: Adverse effects on infant psychomotor development. Pediatr. 84:7-17, 1989. Walter T, Kovalskys J, and Stekel A. Effect of mild iron-deficiency on infant mental development scores. J. Pediatr. 102:519-522, 1983. Weller SC, and Dungy Cl. Personal preferences and ehtnic variations among Anglo and Hispanic breast and bottle feeders. Soc. Sci. Med. 23:539-588, 1986. Wells MS, and Award WM. Iron and heme metabolism. In Devlin TM (ed) Textbook of biochemistry, with clinical correlations (3rd edition). Wiley-Liss Inc, New York, pp 1001-1019, 1993. Williams PL, Innis SM, and Vogel AMP. Breast-feeding and weaning practices among infants in Vancouver. (In progress). Yates F. Sampling methods for census and surveys. NewYork, MacMillan, 1981. Yehuda S, and Youdim MBH. Brain iron: a lesson from animal models. Am. J. Clin. Nutr. 50:S618-629, 1989. 142 BIBLIOGRAPHY Yehuda S, and Youdim MBH. Brain iron-deficiency: Biochemistry and behaviour. In Youdim (Ed.) Brain iron: Neurochemical and Behavioral Aspects. Taylor and Francis, London, pp 89-114, 1988. Yehuda S, Youdim MBH, and Mostofsky Dl. Brain iron-deficiency causes reduced learning capacity in rats. Pharmacol. Biochem. Behav. 25:141-144, 1986. Yeung DL, Penned MD, Leung M and Hall J. Breast-feeding: Prevalence and influencing factors. Can. J. Public Health 72:323-330, 1981. Yip R. The epidemiology of childhood iron-deficiency: Evidence for improving iron nutrition among US children. In Dobbing J. (Ed.) Brain, Behaviour and Iron in the infant diet. London: Springer-Verlag, pp 27-29, 1990. Yip R, Walsh KM, Goldfarb MG, Binkin NJ. Declining prevalence of anaemia in childhood in a middle-class setting: A pediatric success story? Pediatr. 80:330-334, 1987a. Yip R, Binkin NJ, Fleshood L, and Trowbridge FL. Declining prevalence of anaemia among low-income children in the United States. JAMA. 258:1619-1623, 1987b. Youdim MBH, Ben-Shachar D, and Yehuda S. Putative biological mechanisms of the effect of iron-deficiency on brain biochemistry and behaviour. Am. J. Clin. Nutr. 50:S607-615, 1989. Youdim MBH, Green AR, BloomField MR, Mitchell BD, Heal DJ, and Grahame-Smith DG. The effects of iron-deficiency on brain biogenic monamine biochemistry and function in rats. Neuro. Phar. 19:259-267, 1980. Ziegler EE, Fomon SJ, Nelson SE, Rebouche CJ, Edwards BB, Rogers RR, and Lehman LJ. Cow milk feeding in infancy: Further observations on blood loss from the gastrointestinal tract. J. Pediatr. 116:11 -18, 1990. Zlotkin SH. Iron-deficiency in Canadian children? Implications and prevention. Can. J. Pediatr. -2:18-23, 1992. 143 APPENDICES APPENDICES APPENDIX A SAMPLE SIZE CALCULATION A sample size calculation was performed in order to ensure that the point prevalence of iron-deficiency anaemia had a certain precision. That is that the confidence interval around that point estimate was not too wide (Yates, 1981). The table below summarizes the width corresponding to the different sample sizes for the 95% confidence intervals calculated for prevalence rates of 10%, 15%, 20% and 25%. From rates observed in the literature, we expect our estimated prevalence to lie somewhere between 10% and 25%. From the table it can be seen that for a prevalence rate of 15% for example, a sample size of 400 would give us a 95% confidence interval of (11.5%, 18.5%). Width of the 95% confidence interval around the estimated prevalence rate Sample size Prevalence 10% 15% 20% 25% 300 ± 3.4 ± 4.0 ± 4.5 ± 4.9 400 ± 2.9 ± 3.5 ± 3.9 ± 4.2 500 ± 2.6 ± 3.1 ± 3.5 ± 3.8 The above confidence interval values are based on the hypothesis that we have a random sample. What we actually have is a systematic sample from the lists provided by the Vancouver Public Health Department. Treating such a sample as a random sample for the purpose of sample size calculation gives us an overestimation of the sampling error so that the above confidence intervals are on the conservative side. 144 APPENDICES APPENDIX B MAIL OUT 1) LETTER FROM THE VANCOUVER HEALTH DEPARTMENT Vancouver Health Department Administrative Office Fax:734-7897 1060 West 8th Avenue. Vancouver. B.C. V6H 1C4 Telephone: 736-2033 August 9th, 1993 Dear Parent: The Vancouver Health Department and the Nutrition Research Team of the Department of Pediatrics. University of BC, are conducting a study to determine the iron status of infants in the City of Vancouver. We are inviting all parents with infants who will be nine months old in the near future to participate. While we encourage you to participate, you are under no obligation to do so. Attached is more detailed information related to the stud/. You will be contacted by a member of the project team shortly. Should you wish to enrol in the study, you may do so at that time. In the meantime, if you have any questions about the study, please call Dr. Sheila Innis or Diane Jacobsen, RN, at 875-2418. Thank you very much for your consideration. Sincerely, V - W 7 Dr. F. J. Blatherwick. MD, FRCP(C) Medical Health Officer Vancouver Health Department Dr. Sheila Innis. PhD Professor Dept of Pediatrics University of BC CITY OF VANCOUVER FffHTEO ON RECMXEO PAPER 145 APPENDICES 2) LETTER DESCRIBING THE STUDY T H E U N I V E R S I T Y O F B R I T I S H C O L U M B I A DEPARTMENT OF PAEDIATRICS NUTRITION RESEARCH TEAM The Research Centre Faculty of Medicine Department of Pediatrics 950 West 28th Avenue Vancouver, B.C. Canada V5Z 4H4 Tel: (604) 875-Fax:(604) 875-2496 Dear Iron deficiency is the most common and preventable nutrient deficiency in infants under 1 year of age. Although the Canadian Task Force on Periodic Health Examination has recommended all infants be screened for blood iron status (hemoglobin) at 9 months of age, this is not usually done. If detected, low iron status can be easily and effectively treated. Your infant will be 9 months old on . There will be clinics held at which you can obtain a blood iron status test. This screening is offered to you free of charge. Please review the list of marked clinics on the attached sheet and select a clinic that best suits you. You will be contacted by phone about one week prior to these clinic dates to confirm if you can attend one of these clinics. The test results will be sent to your family doctor or paediatrician, so that any necessary treatment can be started as soon as possible. You will be asked to provide some information on your infant's diet and offered the opportunity to participate in a test of visual preference looking. There are no risks to your infant, and any and all information which you will provide will be kept strictly confidential. The screening clinic will give you the opportunity to meet a Public Health Nurse or Nutritionist to discuss any concerns you may have regarding your infant. Light refreshments will be offered. Should you require any additional information you may call either Dianne Jacobsen R.N. or Dr. Sheila Innis at 875-2418. We look forward to seeing you at the clinic. Sincerely, (X-Sheila M. Innis, Ph.D. Professor Department of Paediatrics University of British Columbia 146 APPENDICES 3) CLINIC SCHEDULE SET 1; OCTOBER-NOVEMBER, 1993 C L I N I C S C H E D U L E CLINIC DATE CLINIC TIME CLINIC NAME & ADDRESS 1 TUESDAY OCT 5 , 1 9 9 3 2 : 0 0 - 5 : 0 0 PM MOUNT P L E A S A N T H E A L T H U N I T 3 2 3 E A S T BROADWAY, V A N C O U V E R 2 THURSDAY OCT 7 , 1 9 9 3 4 : 0 0 - 8 : 0 0 PM STRATHCONNA COMMUNITY C E N T R E : 6 0 1 K E E F E R S T , V A N C O U V E R 3 THURSDAY OCT 1 4 , 1 9 9 3 9 : 0 0 AM -1 2 : 0 0 NOON E A S T H E A L T H U N I T 2 6 1 0 V I C T O R I A D R , V A N C O U V E R -*' 4 SATURDAY OCT 1 6 , 1 9 9 3 1 1 : 0 0 AM -4 : 0 0 PM DUNBAR COMMUNITY C E N T R E 4 7 4 7 DUNBAR S T , V A N C O U V E R 5 TUESDAY OCT 1 9 , 1 9 9 3 5 : 0 0 - 8 : 0 0 PM R A Y - C A M C O - O P E R A T I V E C E N T R E 9 2 0 E A S T H A S T I N G S , V A N C O U V E R 6 SATURDAY OCT 2 3 , 1 9 9 3 1 1 : 0 0 A M -4 : 0 0 PM K I T S I L A N O COMMUNITY C E N T R E 2 6 9 0 L A R C H S T , V A N C O U V E R 7 TUESDAY OCT 2 6 , 1 9 9 3 9 : 0 0 AM -1 2 : 0 0 NOON WEST M A I N H E A L T H U N I T 2 1 1 0 WEST 4 3 R D , V A N C O U V E R 8 WEDNESDAY OCT 2 7 , 1 9 9 3 9 : 0 0 AM -1 2 : 0 0 NOON TROUT L A K E COMMUNITY C E N T R E 3 3 5 0 V I C T O R I A D R , V A N C O U V E R 9 TUESDAY NOV 2 , 1 9 9 3 9 : 0 0 AM -1 2 : 0 0 NOON SOUTH V A N C O U V E R F A M I L Y P L A C E -2 2 8 5 EAST* 6 1 S T , V A N C O U V E R 1 0 TUESDAY NOV 2 , 1 9 9 3 4 : 0 0 - 8 : 0 0 PM MOUNT P L E A S A N T H E A L T H U N I T 3 2 3 E A S T BROADWAY} V A N C O U V E R 1 1 SATURDAY NOV 6 , 1 9 9 3 1 1 : 0 0 AM -4 : 0 0 PM B R I T A N N I A COMMUNITY C E N T R E 1 6 6 1 N A P I E R S T , V A N C O U V E R 1 2 TUESDAY NOV 9 , 1 9 9 3 9 : 0 0 AM -1 2 : 0 0 NOON WEST M A I N H E A L T H U N I T 2 1 1 0 WEST 4 3 R D , V A N C O U V E R 13 TUESDAY NOV 1 6 , 1 9 9 3 4 : 0 0 - 8 : 0 0 PM C H A M P L A I N H T S . COMMUNITY C T R 3 3 5 0 M A Q U I N N A , V A N C O U V E R 14 WEDNESDAY NOV 1 7 , 1 9 9 3 9 : 0 0 A M -1 2 : 0 0 NOON NORTH H E A L T H U N I T 2 0 0 - 1 6 5 1 C O M M E R C I A L D R , V A N . 1 5 SATURDAY NOV 2 0 , 1 9 9 3 1 1 : 0 0 AM -4 : 0 0 PM K I T S I L A N O COMMUNITY C E N T R E 2 6 9 0 L A R C H S T , V A N C O U V E R 16 TUESDAY NOV 2 3 , 1 9 9 3 4 : 0 0 - 8 : 0 0 PM A C A D I A F A I R V I E W COMMONS B L K . 2 7 0 7 T E N N I S C R E S C . V A N C O U V E R 17 WEDNESDAY NOV 2 4 , 1 9 9 3 9 : 0 0 AM -1 2 : 0 0 NOON MARPOLE COMMUNITY C E N T R E 9 9 0 WEST 5 9 T H , V A N C O U V E R Clinics for which an infant was eligible to attend, that is those that coincided with when the infant would be 39 ± 1 week, were highlighted with a coloured fluorescent marker for clarification. 147 APPENDICES 4) CLINIC SCHEDULE SET 2; MARCH-APRIL, 1994 CLINIC SCHEDULE FOR THE IRON STUDY DATE TIME CLINIC LOCATION INFANTS BIRTH DATE 20 FRIDAY MARCH 11/94 10 AM to 4 PM SOUTH HEALTH UNIT 6405 KNIGHT STREET JUNE 4 to JUNE 18 21 WEDNESDAY MARCH 16/94 1 PM to 4 PM • CRABTREE CORNER 101 EAST CORDOVA STREET JUNE 9 to JUNE 23 22 MONDAY MARCH 21/94 10 AM to 4 PM NORTH HEALTH UNIT 200-1651 COMMERCIAL DRIVE JUNE 14 to JUNE 28 23 SATURDAY MARCH 26/94 10 AM to 4 PM K3WASSA NEIGHBOURHOOD HOUSE 2425 OXFORD STREET JUNE 19 to JULY 3 24 WEDNESDAY APRIL 6/94 10 AM to 4 PM EAST HEALTH UNIT 2610 VICTORIA DRIVE " JUNE 30 to JULY 14 25 MONDAY APRIL 11/94 10 AM to 4 PM MOUNT PLEASANT HEALTH UNIT 323 EAST BROADWAY JULY 5 to JULY 19 26 SATURDAY APRIL 16/94 10 AM to 4 PM KITSILANO COMMUNITY CENTRE 2690 LARCH STREET JULY 10 to JULY 24 j 27 FRIDAY APRIL 22/94 10 AM to 4 PM NORTH HEALTH UNIT 200-1651 COMMERCIAL DRIVE JULY 16 to JULY 30 28 SATURDAY APRIL 30/94 10 AM to 4 PM REACH CLINIC 1145 COMMERCIAL DRIVE JULY 24 to AUG 7 The clinics for which an infant was eligible to attend, that is those that coincided with when the infant would be 39 ± 1 week, were highlighted with a coloured fluorescent marker for clarification. 148 APPENDICES APPENDIX C INFORMED CONSENT FORM T H E U N I V E R S I T Y O F B R I T I S H C O L U M B I A The Research Centre Faculty of Medicine Department of Paediatrics 950 West 28th Avenue Vancouver. B.C. Canada V5Z 4H4 Tel: (604) 875-Fax: (604) 875-2496 Prevalence of low Iron Status and Feeding Practices of Nine Month Old Infants in Vancouver Summary of Project: The purpose of this study is to define the prevalence of tow iron status in 9 month old infants in Vancouver, identify infants at risk, and assess infant feeding practices and visual preferential looking behaviour. The study involves no risk. My infant and I win participate in the following ways: 1. A small blood sample (about 1/5th teaspoon) will be taken by a registered technologist or nurse from B.C. Children's Hospital from a heel or finger prick for testing iron status. The prick may cause very brief discomfort, but otherwise no pain or risk. 2. The results of the blood tests will be sent to Dr. Louis Wadsworth. B.C.'s Children's Hospital If further testing or treatment is required, my doctor and I will be notified of the results. 3. Measurements of my infant's weight, length, head circumference and fat (skinfold thickness) will be i taken. 4. My infant will have the opportunity to participate in a visual preference looking test which will take about 10-15 min. This involves showing your infant pictures of peoples' faces and recording the amount of time he/she takes to look at the pictures. 5. I will give information regarding my infant's diet and my family's social cultural background. All this information will be strictly confidential. The care and treatment of my infant win be the same as if he/she were not a participant in this study, t may decline or withdraw any participation at any time. If I have any questions about the study procedures and my infant's participation, I may contact Dianne Jacobsen R.N or Dr. Sheila Innis at 875-2418. CONSENT: The objectives and procedure of the study have been explained to me to my satisfaction. Refusal to participate will not jeopardize the future care of my infant My or my infant's name will be treated confidentially and will not be mentioned in any report of the study. I voluntarily give consent for myjnfant to participate in the study of iron status, and I acknowledge receipt of a copy of the consent form. SIGNED: (Parent or person legally authorized to give consent) RELATIONSHIP TO INFANT: W I T N E S S : D A T E : INFORMED CONSENT 149 A P P E N D I C E S APPENDIX D CONFIDENTIAL DEMOGRAPHIC QUESTIONNAIRE QUESTIONAIRE: Form 1 CONFIDENTIAL NUMBER: (4 digit 0001-9,999) • • • • CLINIC DATE: Day Month CLINIC SITE: (2 digits) BABY'S BIRTH DATE: BABY'S SEX: Day Month Male Female • • Year Year Please Check the Appropriate Response I am the baby's: 1) 2) 3) 4) 5) Mother Father Relative Nanny Other • • • • • (specify)_ 1 5 0 APPENDICES CONFIDENTIAL DEMOGRAPHIC QUESTIONNAIRE (cont'd...) Baby's mothers age is: 1) < 20 years 2) 20-24 years 3) 25-29 years 4) 30-34 years 5) 35+ years • Baby's fathers age is-• • • • • 3. What is your marital status? 1) Single • 2) Married/Common-law 3) Separated/Divorced 4) Widow/Widower 4. How many other children live in the household? Did you complete high school? Did you go to college or vocational training? Did you go to University? Mother Father Yes Yes • • • • • • 151 APPENDICES CONFIDENTIAL DEMOGRAPHIC QUESTIONNAIRE (cont'd...) Which of the following describes your family income per year? • • 1) Less than $10,000 2) $10,000 -$19,000 3) $20,000 - $29,000 4) $30,000 - $50,000 5) over $50,000 • THANK YOU , 152 APPENDIX E DIET QUESTIONNAIRE QUESTIONNAIRE: Form 2 (completed with nutritionist/nurse) NUMBER: (4 digit 0001-9,999) • • • • CLINIC DATE: Day Month Year • • • • . • • CLINIC SITE: (2 digits) BABY'S BIRTH DATE: Day Month Year BABY'S SEX: Male Female • • To which ethnic background(s) do you belong? 1) Canadian • 12) French 2 > American (USA) • 13) Italian 3) Chinese • 14) Dutch 4) Japanese • 15) Scottish 5) East Indian • 16) Ukrainian 6) North American Indian 0 17) Polish 7) English • 18) Irish 8) German • 19) Greek 9) Vietnamese • 20) Jewish 10) Korean • 21) Hispanic 11) African 22) other • • • • • • • • • • • (specify). 153 APPENDICES DIET QUESTIONNAIRE (cont'd...) 2 1 b). How many years have you lived in Canada? 1c). Main language spoken at home? a) English b) Other specify 2. ^ Your family diet includes which of the following? Yes No 1) Red meat (beef, pork) • • 2) Fish • • 3) Poultry • 4) Dairy products • • 5) Eggs • • 6) Legumes (beans, peas) • • 7) Nuts • • 8) Fruit and vegetables • • 9) Breads/Cereals • • 10) Pasta/Rice • • 11) Special/Medical/Allergy related diet (specify) 3. Was the baby breast-fed? Yes (complete Q 4 to 6) No (go to Q 7) 4. How long was the baby exclusively breast-fed? (can include fruit juices &/or up to one 8oz bottle of formula/cow's milk/wk) • Months Weeks or Weeks 154 APPENDICES DIET QUESTIONNAIRE (cont'd...) 4b). How much fruit juice/day? 5. Is the baby still breast-feeding? Yes (go to Q 6) No d 5b). At what age was breast feeding completely stopped? • Months Weeks or Weeks 6. Breast feeding was replaced/supplemented with formula/milk: 1) Infant formula (powdered/concentrate a) regular formula (low iron) b) formula with iron (fortified) c) soybased formula d) other 2) Cow's milk 3) Home made formula (evaporated milk) 4) other Comments: Check for used Age -started Changed/ stopped mths wks mths wks • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • Brand &/or label colour/or specify 155 APPENDICES DIET QUESTIONNAIRE (cont'd...) 4 6b). Which of the following influenced your choice to breast-feed? 1) Public health clinic or nurse's advice • 2) Pediatrician • 3) Family Doctor • 4) Family advice • 5) Friend's or other mother's advice • 6) Availability • 7) Cost • 8) Advertising • 9) Dietitian • 10) In hospital support • 11) Prenatal class • 12) Personal choice • 13) Books • 14) Previous experience • 15) Other • specify • SKIP TO Q X $ 7. What type(s) of formula do you feed your infant? NOTE: (baby's fed 1) Infant formula (powdered/concentrate a) regular formula (low iron) b) formula with iron (fortified) c) soybased formula formula from birth only) Check for used Age started Changed/ stopped mths wks mths wks • • • • • • • • • • • • • • • Brand &/or label colour/or specify 1 5 6 A P P E N D I C E S D I E T Q U E S T I O N N A I R E (cont'd...) d) other 2) Cow's milk 3) Home made formula (evaporated milk) 4) other Comments: • • • • • • • • • • • • • • • • • • • • 7b). Which of the following influenced your choice to feed formula/cow's milk? Formula Cow's milk 1) Public health clinic or nurse's advice • • 2) Pediatrician • • 3) Family Doctor • • 4) Family advice • • 5) Friend or other mother's advice • • 6) Availability • • 7) Cost • • 8) Advertising • • 9) Dietitian • • 10) Used in hospital • • 11) Prenatal class • • 12) Personal choice • • 13) Books • • 14) Personal experience • • 15) Other • (specify). 157 DIET QUESTIONNAIRE (cont'd...) (all infants receiving formula/cow's milk) Which of the following influenced your choice of formula or choice to use 1) Public health clinic or nurse's advice 2) Pediatrician 3) Family Doctor 4) Family advice 5) Friend or mother's advice 6) Availability 7) Infant tolerance/allergy 8) Cost 9) Advertising 10) Dietitian 11) Used in hospital 12) Prenatal class 13) Personal choice 14) Books 15) Previous experience 16) Other (specify) Formula Cow's milk • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 9. At what age did you start to introduce cereal foods? • Months Weeks or Weeks 158 APPENDICES DIET QUESTIONNAIRE (cont'd...) 10. What type of cereals did you first use? 1) Commercial infant cereals 2) Cooked rice 3) Bread 4) Crackers 5) Breakfast cerea!s(hot or cold) (specify if yes) -Yes No • • • • Yes No • • • • • • 6) Other • specify. 11. Which of the following influenced your choice of cereal to use? 1) Baby did not like it 2) Family advice 3) Pediatrician 4) Friend or other mother's advice 5) Availability 6) Cost 7) Public health clinic or nurse's advice 8) Books/magazines 9) Advertising 10) Infant tolerance (allergy) 11) Family Doctor 12) Previous experience 13) Personal choice 14) Dietitian 15) Other (specify) • • • • • • • • • • • • • • • • 1 5 9 APPENDICES DIET QUESTIONNAIRE (cont'd...) 12. Does the baby eat any of the following foods? (for all yes response indicate at what age these foods were introduced). Age Yes No 1) fruit juices • • 2) meat/beef • • 3) egg yolk • • 4) chicken • • 5) fish • • 6) vegetables • • 7) fruits • • 8) legumes (beans/peas/da!) • • 9) tofu • • 13. Does your baby currently drink cow's milk? Yes Q No go to Q14 13b). If yes, how much cow's milk/day? cup(s). 14. What type of cow's milk? 1) Whole milk • 2) 2% milk • 3) 1% milk • 4) Skim • Does the baby : Yes No 1) eat from a spoon ? • • 2) eat finger foods? • • 1 6 0 DIET QUESTIONNAIRE (cont'd...) 15. Do you give any vitamin or other nutritional supplements to the baby? yes (specify) . no • 16. How many of the following prepare the baby's food? (from 6 months) 1) Mother 5) Nanny/baby sitter • 2) Father 6) Daycare • 3) Grandmother , 7) Other • 4) Grandfather specify Note: (next question to be answered by all mothers who breast fed) 17. Your choice to start using formula/cow's milk was influenced by which of the following? 1) Returned to work 2) lack of support 3) Illness of mother 4) Illness of baby • 5) Concerned about baby's nutrition, adequate milk • 6) Uncomfortable with breast feeding • 7) Painful • 8) Baby biting the breast 9) Not enough time 10) Personal choice 11) Other C] specify Name of interviewer: 161 APPENDIX F ANTHROPOMETRIC FORM GROWTH: Form 3 (completed by nurse/nutritionist) NUMBER: (4 digit 0001-9,999) CLINIC DATE: CLINIC SITE: (2 digits) BABY'S BIRTH DATE: • ••• Day Month Day Month • • BABY'S SEX: Male • Female • ANTHROPOMETRIC MEASUREMENTS GROWTH MEASURES 1. Weight gm 2. Crown-Heel Length cm 3. Head circumference cm Year Year BODY COMPOSITION 1 2 3 1. Abdominal mm 2. Triceps mm 3. Sub-scapular mm Nurse/nutritionist 162 CONTACT CARD FOR BLOOD RESULTS University of British Columbia Nutrition Research Group F O R B L O O D R E S U L T S : Please call (between the hours 8:00 am and 10:00 am.) Please have available your baby's ID# Telephone: 875-2418 163 APPENDICES APPENDIX H HEMATOLOGY The mean, median and standard deviation values for the iron indices used to classify the iron status of the study population are shown in the table below. Haemoglobin and ferritin were determined for all 434 infants who participated in the study. The TIBC was not assessed for 10 infants because an inadequate amount of blood was collected. ZPP was also not assessed for 9 infants because the collected blood sample had clotted before the analysis could be done. Mean, standard deviation and median values of the iron indices used to classify iron status of the study population . Iron indices1 n Mean Standard deviation Haemoglobin 434 118.0 8.3 118.0 Ferritin 434 20.2 18.4 14.9 TIBC 424 55.5 9.2 55.4 ZPP 425 64.4 18.4 61.0 1 Parameters used to classify iron deficiency in the study population. TIBC, total iron binding capacity; ZPP, zinc protoporphyrin. 164 APPENDICES APPENDIX I AGE OF INTRODUCTION OF VARIOUS FOODS WITHIN EACH IRON STATUS GROUP (FIGURES 4.7-4.16) Iron deficient anaemic (n=30) v o 3 TJ O U 60 50 40 30 20 10 0 O o 5 £ 3 4-6 7-9 N M Age of Introduction (months) Non—anaemic iron sufficient (n=254) 60 H 50 40 H 30 20 10 -i 0 £ 3 4-6 7-9 N Age of introduction (months) Low iron status (n=106) <D O 3 •a o a •a • i H 60 50 40 -| 30 -I 20 10 H 0 < 3 4-6 7-9 N M Age of introduction (months) Uncertain (n=43) x) 60 <u o 0 is 4 0 - i d 1 20-] .a 0 10 J I M i 3 4-6 7-9 N Age of introduction (months) Figure 4 . 7 . Age of introduction of fruit juice within each iron status group N = not yet introduced; M = missing 165 APPENDICES o T3 CD O •g o (-, a Iron deficient anaemic (n=30) go 80 H 70 60 H 50 40 30 H 20 10 0 £ 3 4-6 7-9 N Age of introduction (months) 90 80 70 60 50 40 30 20 10 0 Non—anaemic iron sufficient (n=254) £ 3 4-6 7-9 N M Age of introduction (months) CD O 3 X) o SH X ) CD O o a Low iron status (n=106) 90 80 70 j 60 A 50 40 30 20 10 0 1 £ 3 4 - 6 7-9 N Age of introduction (months) Uncertain (n=43) 90 80 70 60 50 40 30 -| 20 10H 0 3 4-6 7-9 N M Age of introduct ion (months) Figure 4.8. Age of introduction of infant cereal wi thin each i r o n status group N = not yet; M = missing 166 APPENDICES V O 3 •O o -4-> .5 ro t-> Ci a) Iron deficient anaemic (n=30) 90 -i <L> 80 -O 70 -3 •d o 60 -i i ^> d 50 -03 40 -1 30 -20 -10 -0 - — i — n * 3 4-6 7-9 N M Age of introduction (months) 80 70 60 50 ^ 40 30 20 10H 0 Non—anaemic iron sufficient (n=254) £ 3 4-6 7-9 N M Age of introduction (months) Low iron status (n=106) 80 -i 70 -a> o •3 60 -o 50 -u "3 40 -w a 30 -infa 20 -infa 10 -0 -£ 3 4-6 7-9 N Age of introduction (months) Uncertain (n=43) 80 -i 7 0 -«J o 3 T3 60 -O 50 -u .a 40 -M •> d 30 -•a 20 -• 3 10 -0 - J Z L Age = 3 4-6 7-9 N M of introduction (months) Figure 4 . 9 . Age of introduction of vegetables within each iron status group N = not yet introduced; M = missing 167 APPENDICES xf <v o 3 X» O SH fc? x> co o 3-x> o u a Iron deficient anaemic (n=30) 80 70 60 50 40 3 0 20 10 0 M i 3 4 - 6 7 - 9 N Age of introduction (months) 8 0 70 H 60 50 40 30 20 H 10 0 Non—anaemic iron sufficient (n=254) J Z F L I M £ 3 4 - 6 7 - 9 N Age of introduction (months) xi « o 3 X) o SH XI CD O O SH .9 6? Low iron status (n=106) 8 0 - , 7 0 60 5 0 -4 0 -3 0 -2 0 -1 0 -0 M £ 3 4 - 6 7 - 9 N Age of introduction (months) Uncertain (n=43) 8 0 7 0 H 6 0 5 0 4 0 3 0 2 0 10 0 • a A g e < 3 4 - 6 7 - 9 N M of introduction (months) Figure 4.10. Age of introduction of fruit within each iron status group N = not yet introduced; M = missing 1 6 8 APPENDICES v o 3 TJ O f-< -»-» £ •*-> !=l <0 "a •H O o kl Iron deficient anaemic (n=30) 6 0 -5 0 -4 0 -3 0 2 0 -I 1 0 - | 0 s: 3 4 - 6 7 - 9 N M Age of Introduction (months) Non—anaemic iron sufficient (n=254) 6 0 5 0 4 0 3 0 2 0 1 0 0 - a M * 3 4 - 6 7 - 9 N Age of introduction (months) Low iron status (n=106) •0 o o 3 T3 0 3 a *a 6 0 - ^ 50-I 4 0 3 0 2 0 -I 1 0 -I 0 £ 3 4-6 7-9 N U Age of introduction (months) Uncertain (n=43) 6 0 H (0 o 5 5 0 - | TJ o vi 3 0 -§ 2 0 -•9 10 0 M * 3 4 - 6 7 - 9 N Age of introduction (months) Figure 4.11. Age of introduction of egg yolk within each iron status group N = not yet introduced; M =. missing 169 APPENDICES Iran deficient anaemic (n=30) 4> O 3 X> © +-> a 73 rt 60 50 40 -| 30 20 10 A 0 < 3 4-6 7-9 N M Age of introduction (months) Non—anaemic iron sufficient (n=254) 60 -Q> O 3 50 -Xf o £ 40 -fl CO 30 -§ 20 -s 10 -n - * 3 4-6 7-9 N M Age of introduction (months) Low iron status (n=106) xi CO o 3 •o o u H-> a n * J cl ca •a 60 50 4 0 -I 30 -| 20 10 0 £ 3 4-6 7-9 N M Age of introduction (months) Uncertain (n=43) xi 60 H CD o 3 XI o u 50 -\ 40 30 -\ d w § 20 •IH S 10 fc? 0 J Z F L £. 3 4-6 7-9 N M Age of introduction (months) Figure 4.12. Age of introduction of legumes within each iron status group N = not yet introduced; M = missing 170 APPENDICES xf v o 3 T3 o u a> o o kl Iron deficient anaemic (n=3Q) 80 70 60 -I 50 40 30 20 -I 10 -I 0 £ 3 4-6 7-9 N M Age of Introduction (months) 80 70 60 50 -\ 40 30 20 10 0 Non—anaemic iron sufficient (n=254) _0TL — i — ~ r s: 3 4-6 7-9 N Age of introduction (months) xi « o 3 o u X ) <D O o t-l .a tf Low iron status (n=10G) 80 70 -60 -50 -40 -30 -20 -10 0 JTJL H 3 4 - 6 7 - 9 N M Age of introduction (months) Uncertain (n=43) 80 70 60 -50 -40 -30 -20 -10 -0 < 3 4-6 7-9 N M Age of introduct ion (months) Figure 4.13. Age of introduction of tofu within each iron status group N = not yet introduced; M = missing 171 APPENDICES Iron deficient anaemic (n=30) x> 4> O 3 X> O d es 70 -, 60 50 40 30 H 20 10 0 £ 3 4-6 7-9 N M Age of introduction (months) X f CD O 3" o 5 fc? 70 -, 60 50 H 40 30 20 10 H 0 Non—anaemic iron sufficient (n=254) i M £ 3 4-6 7-9 N Age of introduction (months) Low iron status (n=106) o o 3 X ) o fc? 70-! 60 50 -40 -30 -20 10 0 £ 3 4 - 6 7-9 N Age of introduction (months) Uncertain (n=43) x) CD O 3 XI o &? 70 n 60 50 40 30 20 10 0 ^ 3 4-6 7-9 N M Age of introduction (months) Figure 4 . 1 4 . Age of introduction of fish within each iron status group N = not yet introduced; M = missing 172 APPENDICES Iron deficient (n=30) x> v o 3 TJ 0 u 1 tn .«-> <0 tf 60 50 -40 -30 -20 10 0 £ 3 4-6 7-9 N M Age of introduction (months) Non—anaemic iron sufficient (n=254) xJ a> o 3 •O o u a 60 50 40 30 20 10 0 s: 3 4-6 7-9 N M Age of introduction (months) Low iron status (n=106) TJ 60 -41 O 3 50 -T3 O SH -a 40-.3 m 30 -3 a 20 -a 10 -0 -£ 3 4-6 7-9 N Age of introduct ion (months) Uncertain (n=43) T J (D O 3 •0 o .a 60 H 50 40 -I 30 20 -| 10 0 M * 3 4-6 7-9 N Age of introduction (months) Figure 4.15. Age of introduction of chicken within each iron status group N = not yet introduced; M = missing 173 APPENDICES Iron deficient anaemic (n=30) 50 -, TJ 4> O 73 40 -T j O u 30 -a 20 -an 3 10 -0 -* 3 4 - 6 7 - 9 N M Age of introduction (months) TJ 4> O a TJ O f-, -*-» OQ a «J *? 50 40 30 20 10 0 Non—anaemic iron sufficient (n=254) < 3 4 - 6 7 - 9 N Age of introduction (months) Loir iron status (n=106) 50 - i TJ <0 o n 40 -TJ O u H-) 30 -d 20-3 3 10-0 - X Z L T M £ 3 4 - 6 7 - 9 N Age of introduction (months) Uncertain (n=43) •d a> o d TJ O u d 3 tf 50 40 ^ 30 20 10 0 * 3 4 - 6 7 - 9 N M Age of introduction (months) Figure 4 . 1 6 . Age of introduction of meat within each iron status group N = not yet introduced; M = missing 174 APPENDICES APPENDIX J AVERAGE AGE OF INTRODUCTION OF VARIOUS FOODS WITHIN EACH IRON STATUS GROUP 4.17. Average age of introduction of various foods within each iron status group for those infants who were actually fed the food before 9 months of age Food Average age if introduction of food (months) Iron deficient anaemic (n = 30) Low iron status in =106) Non-anaemic iron-sufficient (n = 254) Uncertain (n = 43) Fruit juice 6.6±1.2 5.5±1.6 5.6±1.6 5.4±1.6 (13.3%) (24.5%) (22.4%) (16.2%) Infant cereals 5.1 ±0.8 4.5±1.2 4.7±1.1 4.9 ±1.3 (10.0%) (3.8%) (7.4%) (4.6%) Vegetables 6.1 ±0.7 5.6±1.1 5.7±1.1 5.6±1.1 (0.0%) (5.7%) (8.6%) (11.6%) Fruits 6.0±0.8 5.5±1.3 5.7±1.2 5.7±1.2 (3.3%) (4.7%) (5.9%) (9.3%) Egg yolk 6.7±1.5 7.3±1.1 7.1 ±1.2 7.3±1.0 (60.0%) (61.3%) (65.7%) (62.8%) Legumes 6.7 ±1.0 5.8 ±1.4 5.9±1.2 5.7±1.2 (26.6%) (29.2%) (34.6%) (32.5%) Tofu 7.1 ±1.2 7.2±1.0 7.0±1.2 6.3 ±1.0 (76.7%) (87.7%) (82.7%) (83.7%) Fish 7.1 ±1.2 6.9 ±1.4 6.9±1.2 6.6±1.9 (46.6%) (69.8%) (70.0%) (69.8%) Chicken 7.0 ±1.0 7.0 ±1.3 6.8±1.1 6.9 ±1.4 (26.6%) (27.3%) (27.9%) (25.6%) Meat 7.0±1.0 6.9±1.3 6.8±1.2 6.6±1.1 (43.3%) (37.7%) (35.4%) (46.5%) Values are means ± standard deviation % in brackets indicates percent of infants within each category who had not received the food by 9 months. The mean and standard deviation values are calculated for infants who had received the food by 9 months of age. 175 APPENDICES APPENDIX K COWS' MILK FEEDING IN THE STUDY POPULATION Cows' milk feeding and age of introduction of cows's milk Total infants introduced Age of introduction of cows' milk1 Total infants not to cows' milk (months) introduced to cows' milk 8 75 1 2 3 ; 4 10 25 18 358 1 Age of introduction of cows' milk was unknown for 12 infants introduced to cows' milk No., Number 176 APPENDICES APPENDIX L BREAST-FEEDING AMONG CAUCASIAN VERSUS NON-CAUCASIANS MOTHERS Duration of breast-feeding was significantly higher among Caucasian than non-Caucasian mothers (p<0.00001) (Williams et al, 1995). Among Caucasian mothers, 29.3% breast-fed their infants for more than 8 months. In contrast, only 6.5% of the non-Caucasian mothers were still breast-feeding by 8 months. Duration of breast-feeding among Caucasian versus Non-Caucasian Mothers in the study population Duration of breast-feeding (months) Total breast-fed Never <3 >3-<6 >6-<8 >8 % Caucasian mothers breast-feeding (n = 249) 8.4 16.5 28.9 16.9 29.3 91.61 % Non-Caucasian mothers breast-feeding (n = 185) 43.2 30.3 10.8 9.2 6.5 56.81 1 p<0.0001 %, percent Infant formula and cows' milk < 12oz/week Caucasian, comprised of Canadian/American (n = 120) and European (n = 129) Non-Caucasian, comprised of Chinese (n = 81), East Indian (n=43), Filipino (n = 18), other Asian (n = 23) and Other (n = 20) 177 

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.831.1-0087071/manifest

Comment

Related Items