Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Underweight status, household food security and associated characteristics among women ≥18y in Bình Phước… Brown, Matthew Ryan 2012

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

Item Metadata

Download

Media
24-ubc_2013_spring_brown_matthew.pdf [ 1.16MB ]
Metadata
JSON: 24-1.0073470.json
JSON-LD: 24-1.0073470-ld.json
RDF/XML (Pretty): 24-1.0073470-rdf.xml
RDF/JSON: 24-1.0073470-rdf.json
Turtle: 24-1.0073470-turtle.txt
N-Triples: 24-1.0073470-rdf-ntriples.txt
Original Record: 24-1.0073470-source.json
Full Text
24-1.0073470-fulltext.txt
Citation
24-1.0073470.ris

Full Text

Underweight Status, Household Food Security and Associated Characteristics among Women >18y in Bình Phước Province, Vietnam by  Matthew Ryan Brown  B.Sc, The University of British Columbia, 2007  SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENT FOR THE DEGREE OF  MASTER OF SCIENCE  in  The Faculty of Graduate Studies (Human Nutrition)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver) December, 2012  © Matthew Ryan Brown, 2012  Abstract Background: Chronic malnutrition among women is a significant issue in rice exporting Vietnam. In 2000, underweight prevalence (BMI<18.5 kg/m2) among women (20-49y) was 26.3% and was associated with maternal and infant mortality and intrauterine-growth-restrition. Known risk factors of underweight are poverty, rural-living and ethnicity but are insufficient to explain the causes of these high rates. Food insecurity, nutritional knowledge, dietary diversity, and cultural dietary behaviours may contribute to underweight prevalence in Vietnam. The province of Bình Phước was selected to explore determinants of high underweight status in a productive agrarian region. Objective: The primary objective was to assess underweight status and household food insecurity among non-pregnant women in Bình Phước. Secondary regression analysis of both underweight status and household food insecurity determined associated risk factors including psychosocial dietary behaviours, household demographics, dietary diversity and nutrition knowledge. Methods: A cross-sectional survey was conducted among non-pregnant women (n=397), >18y, in 10 villages. The Household-Food-Insecurity-Access-Scale (HFIAS) classified food insecurity and anthropometric data classified underweight. Further survey questions assessed dietary diversity, demographics, nutrition knowledge and dietary behaviours. Results: In 2006, 24.4% (97/397) of women were underweight (95% CI; [20.2 to 28.7]). Food insecurity (HFIAS) scores were 6.7 (5.7 SD) out of 27 classifying 51% (204/397) as having severe, 15% (61/397) as moderate and 19% (77/397) as mild food insecurity. ii  Logistic regression predicting underweight (vs. non-underweight) found that participants with mild food insecurity had a one-in-three odds of being underweight compared with those severely food insecure (OR 0.35, 95% CI; [0.17 to 0.75], p<0.001). Participants with no children (<5y) had lower odds of underweight compared to participants with >2 children (<5y) (OR 0.30, 95% CI; [0.14 to 0.65], P<0.01). The odds of higher food insecurity was lower among participants with moderate nutrition knowledge (OR 0.52, [95%CI, 0.29 to 0.92], P<0.05), higher education (grade 7-9) (OR 0.59, [95%CI, 0.36 to 0.97], P<0.05) and higher dietary diversity (OR 0.32, [(95% CI; 0.18 to 0.58], P<0.001). Conclusion: High levels of underweight were associated with HFIAS food insecurity and with having higher numbers of children (<5y). Significant predictors of food insecurity were education, nutrition knowledge and dietary diversity.  iii  Preface This thesis is based off data collected in a 2006 survey carried out in Binh Phuoc. The questionnaire was created by Dr. Judy McLean of the University of British Columbia (UBC). Dr. McLean headed the collaborative effort between UBC and the National Institute of Nutrition and facilitated the translation and adaptation of this survey in Binh Phuoc. I, Matthew Brown, was involved in the proposal and facilitation process with health professionals in Vietnam as well as organizing translators and training sessions. I assisted in the data collection with my translators as well as with Dr Xuan and a trained nurse. I also completed the data entry of the survey for an undergraduate directed studies project. This data was unpublished and I was later permitted to perform analysis using this data for my Master of Science thesis. This project was a concerted effort with UBC, National Institute of Nutrition [Vietnam], the hospital of Dong Xoai (Binh Phuoc) and regional health centres. University of British Columbia behavioral ethics committee (H06-80632) and the National Institute of Nutrition of Vietnam and ultimately the Ministry of Health Vietnam approved the survey.  iv  Table of Contents Abstract .......................................................................................................................................................... ii Preface ............................................................................................................................................................iv Table of Contents ........................................................................................................................................... v List of Tables ..................................................................................................................................................vi List of Figures ................................................................................................................................................ix List of Acronyms ............................................................................................................................................. x Acknowledgements ........................................................................................................................................xi Dedication ................................................................................................................................................... xii I INTRODUCTION..................................................................................................................................... 1 1.1 Chronic Under-nutrition among Women and Children .............................................. 1 1.2 Study Area and Population ......................................................................................... 3 II REVIEW OF LITERATURE ................................................................................................................. 5 2.1 Effects of Chronic Under-nutrition among Women .................................................... 5 2.2 Known Determinants of Underweight in Vietnam ..................................................... 7 2.2.1Poverty and Under-nutrition ........................................................................ 7 2.2.2 Rural Setting and Under-nutrition............................................................... 8 2.2.3 Ethnic Minorities and Under-nutrition......................................................... 9 2.3 Identifying Additional Risk Factors of Underweight among Women ....................... 10 2.3.1 Household Food Insecurity ....................................................................... 11 2.3.2 Low Dietary Diversity ............................................................................... 14 2.3.3 Demographics Factors ............................................................................... 16 2.3.4 Nutrition Knowledge of Women ............................................................... 18 2.3.5 Identifying Cultural Psychosocial Dietary Behavior ................................. 19 2.4 Study Objectives and Rationale ................................................................................. 22 III METHODS AND MATERIALS ......................................................................................................... 23 3.1 Participants ................................................................................................................ 23 3.2 Survey Tools .............................................................................................................. 24 3.2.1 Household Food Insecurity ........................................................................ 25 3.2.2 Dietary Assessment .................................................................................. 25 3.2.3 Anthropometric Assessment ..................................................................... 27 3.2.4 Nutrition Knowledge of Women............................................................... 27 3.2.5 Psychosocial Dietary Behaviours and Perceptions ................................... 28 3.3 Data Analysis ............................................................................................................. 29 3.3.1 Sample Size .............................................................................................. 30 3.3.2 Analysis of Underweight Status ................................................................ 30 3.3.3 Analysis of Household Food Insecurity .................................................... 31 IV RESULTS ............................................................................................................................................ 32 4.1 Underweight among Women ..................................................................................... 32 4.2 Household Food Insecurity ........................................................................................ 34 4.3 Adjusted FCS and Dietary Diversity ........................................................................ 35 4.4 Nutritional Knowledge .............................................................................................. 36 4.5 Dietary Behaviours .................................................................................................... 41 4.6 Predictor of Underweight........................................................................................... 45 4.7 Predictor of Household Food Insecurity .................................................................... 46 V DISCUSSION AND CONCLUSIONS ................................................................................................. 51 5.1 Underweight Prevalence of Previous Studies ............................................................ 51 5.2 Risk Factors for Underweight ................................................................................... 51 5.3 Risk Factors: Household Food Insecurity ................................................................. 52 5.4 Nutrition Knowledge ................................................................................................. 53 5.5 Dietary Behaviours and Perceptions .......................................................................... 55 5.6 Dietary Diversity ....................................................................................................... 57 5.7 Targeting Possible Interventions ............................................................................... 59 5.8 Study Limitations....................................................................................................... 60 5.8.1 Validity of HFIAS Categories................................................................... 60 5.8.2 Nutrition Knowledge Classifications: ‘Green Leafy Vegetables’ ............. 62 5.8.3 Future Questionnaire Amendments .......................................................... 64 5.9 Conclusion .................................................................................................................. 66 VI BIBLIOGRAPHY ................................................................................................................................ 68 VII APPENDIX A: Descriptive, Regression, Knowledge and Behaviour Tables ..................................... 75 APPENDIX B: Survey Questionnaire ................................................................................................. 93  v  List of Tables Table 2.3.1 Defining Household Food Insecurity Access Scale (HFIAS) by category of food insecurity .... 12 Table 2.3.2 Household Hunger Scale (HHS) scoring and categorization of cumulative cutoffs of household hunger ............................................................................................................................................................ 13 Table 2.3.3 Food Consumption Score weighting coefficients of food groups of the Food Frequency Questionnaire (FFQ) used in Bình Phước, Vietnam ...................................................................................... 15 Table 4.1 Characteristics of non-pregnant women >18y (n= 397) in Bình Phước, Vietnam 2006 ............... 33 Table 4.2.1 Household food insecurity indicators categorized by the Household Food Insecurity Access Scale (HFIAS) and Household Hunger Scale (HHS) among non-pregnant women (n= 397) surveyed in Bình Phước, Vietnam Nov-Dec 2006 .......................................................................................................... 34 Table 4.2.2 Household Food Insecurity Access Scale (HFIAS) 30-day recall with grey scale indicating mild, moderate and severe categorization of food security among non-pregnant women (n=397) surveyed in Bình Phước, Vietnam Nov-Dec 2006 ...................................................................................................... 35 Table 4.3.1 Adjusted Food Consumption Score (FCS) among non-pregnant women (n = 397) surveyed in Bình Phước, Vietnam Nov-Dec 2006 .......................................................................................................... 36 Table 4.4.1 Nutrition knowledge scores of women survey participants (n=397) in Bình Phước, Vietnam Nov-Dec 2006 ............................................................................................................................................... 37 Table 4.4.2 Food items commonly* considered as nutritional for the treatment of goiter [Iodine deficiency] assessed by free response questions** among non-pregnant women familiar with goiter (n=376) in Bình Phước, Vietnam ............................................................................................................................................. 38 Table 4.4.3 Food items commonly* considered by participants as good sources of vitamin A assessed by free response questions** among women (n=212) in Bình Phước, Vietnam ................................................ 39 Table 4.4.4 Food items commonly* considered by participants as good sources of iron assessed by three free response questions among women (n=325) in Bình Phước, Vietnam ................................................... 40 Table 4.4.5 Food items commonly* considered by participants as good sources of protein assessed by free response questions** among non-pregnant women familiar with protein (n=254) in Bình Phước, Vietnam ....................................................................................................................................................................... 41 Table 4.5.1 Food items commonly* considered important to consume during pregnancy (n respondents=266) and one month postpartum (nrespondents=357) assessed by three free response questionnaire among nonpregnant women with children (n=366) in Bình Phước, Vietnam ............................................................... 42 Table 4.5.2 Food items commonly* considered important by non-pregnant women with children for weaning infants during first year of life assessed by free response questions** (n=366) in Bình Phước, Vietnam ......................................................................................................................................................... 44 Table 4.6.1 Logistic regression for underweight status (BMI <18.5) among associated factors of Household Food Insecurity Scale (HFIAS) categories and number of dependents (<5y) among non-pregnant women (n = 397) surveyed in Bình Phước, Vietnam ..................................................................................................... 45 Table 4.7.1 Univariate Linear Regression of HFIAS scores on determinant factors among non-pregnant women indicating dietary preference during pregnancy (n=395) surveyed in Bình Phước, Vietnam ........... 46  vi  Table 4.7.2 Spearman's rank test correlation coefficient of BMI with three food security indicators; Household Hunger Scale (HHS), Household Food Insecurity Access Scale (HFIAS) and adjusted Food Consumption Score (FCS) among women (n=470) in Bình Phước, Vietnam .............................................. 49 Table 4.7.3 Logistic regression of Household Food Insecurity Access Scale from zero to mild food insecurity versus moderate to severe food insecurity with associated factors of education level, and nutrition knowledge and adjusted food consumption scores of non-pregnant women (n = 397) surveyed in Bình Phước, Vietnam .................................................................................................................................... 50 Table A1.1 Characteristics of women survey participants (n=397) by village in Bình Phước, Vietnam NovDec 2006 ....................................................................................................................................................... 75 Table A1.2 Logistic Regression between underweight (BMI<18.5) and non-underweight women (n=251) and associated characteristics of non-pregnant women (n=397) in Bình Phước, Vietnam Nov-Dec 2006 ... 78 Table A1.3 Univariate Linear Regression of HFIAS scores on determinant factors among non-pregnant women indicating dietary preference during pregnancy (n=360) surveyed in Bình Phước, Vietnam .......... 81 Table A1.4 Multi-nomial regression coefficient of mild Household Food Insecurity Access Scale (HFIAS) categories versus severe on determinant factors of age, household size, number of children, education level, vocation, adjusted Food Consumption Score and BMI of women (n = 397) surveyed in Bình Phước, Vietnam ......................................................................................................................................................... 83 Table A1.5 Table Household Food Insecurity Access Scale (HFIAS) score by characteristics of women participants (n = 397) in Bình Phước, Vietnam Nov-Dec 2006 ................................................................... 84 Table A1.6 Linear regression for BMI among associated factors of age and number of dependents (<5y) and among non-pregnant women (n = 397) surveyed in Bình Phước, Vietnam ........................................... 86  Table A2.1 Nutrition knowledge: Food items commonly* considered by participants as nutritional treatment of night blindness (XN) assessed by free response questions among women (n=276) in Bình Phước, Vietnam ............................................................................................................................................. 86 Table A2.2 Nutrition knowledge: non nutritional treatments for iron deficiency anemia assessed by free response questions among women (n=336) in Bình Phước, Vietnam ........................................................... 87  Table A3.1 Dietary behaviour: food items commonly* fed to ill children, three responses (nresponses=567) as surveyed among (nparticipants=264) in Bình Phước, Vietnam ........................................................................... 88 Table A3.2 Dietary behavior: common rational for foods items given to ill children (n participants=264) in Bình Phước, Vietnam ............................................................................................................................................. 88 Table A3.3 Dietary behaviour: food items commonly* avoided during pregnancy by respondents (n=266) in Bình Phước, Vietnam ............................................................................................................................... 89 Table A3.4 Dietary behaviour: food item avoidance during pregnancy according to trimester as reported by women (n = 217) in Bình Phước, Vietnam .................................................................................................... 89 Table A3.5 Dietary behaviour: food pattern increases during pregnancy according to trimester as reported by previously pregnant, non-pregnant women cognitively eating more specific food items (n= 366) in Bình Phước, Vietnam ............................................................................................................................................. 89  vii  Table A3.6 Dietary behaviour: common reasons given by women (n=217) for food item avoidance during pregnancy in Bình Phước, Vietnam .............................................................................................................. 90 Table A3.7 Dietary behaviour: Food Frequency Questionnaire with colour corresponding with response frequency of surveyed women (n=397) in Bình Phước, Vietnam ................................................................ 91  viii  List of Figures Figure 2.1 Geo-Political Vietnam ………………………………………………….5 Figure 3.1 Bình Phước survey sites…………………………………………….…24  ix  List of Acronyms APR: Absolute Poverty Rate ASF: Animal Source Foods BMI: Body Mass Index CAD: Canadian Dollars CED Chronic Energy Deficiency EM: Expectation Maximization FANTA: Food And Nutrition Technical Assistance FAO: Food and Agriculture Organization FCS: Food Consumption Score FFQ: Food Frequency Questionnaire GDP: Gross Domestic Product GSO: General Statistics Office of Vietnam HAZ: Height for Age Z-score HDI: Human Development Index HFIAS: Household Food Insecurity Access Scale HHS: Household Hunger Scale IDA: Iron deficiency Anemia IMR: Infant Mortality Rate IUGR: Intra-Uterine Growth Restriction MCAR: Missing Completely At Random MOH: Ministry Of Health of Vietnam NCHS: National Center for Health Statistics NIN: National Institute of Nutrition of Vietnam OR: Odds Ratio PEMCP: Protein Energy Malnutrition Control Program PPP: Purchasing Power Parity RAE: Retinol Activity Equivalent SD: Standard deviation UNICEF: United Nations International Children’s Emergency Fund USD: United States of America Dollars VAD: Vitamin A Deficiency VLSS: Vietnam Living Standards Survey WAZ: Weight-for-Age Z-score WFP: World Food Program WHO: World Health Organization WTO: World Trade Organization  x  Acknowledgements A sincere special thanks to my supervising professors. This master project and being my supervisor was more work and headache than anyone could rightly imagine. Thank you Dr. Timothy Green helping push me through, putting up with me and all your grace! Thank you for not giving up on me even when I gave up on this project myself! Thank you for meeting up with me and encouraging me while I was in Vietnam. Thank you Dr. Judy McLean for inspiring the students of UBC that they can be a part of the solution to one of the great travesties in the 21st century, malnutrition in a world of food sufficiency. Thank you for believing in me and giving me this opportunity to be involved in this pursuit. Thank you to all involved at the National Institute of Nutrition Vietnam. Bui Da, Dr. Hop, Dr.Nhien, Dr. Khan, Dr. Thuc, Dr. Hoa, Dr. Hoa, Dr. Dung, Thao and Thuy. Thank you Dr. Nga and Huong for translating and reviewing the protocols and surveys. Thank you for applying for ethics approval and facilitating approval from the regional health districts. Thank you Bui Da for your extra work on my behalf and your friendship. Thank you to the nurses and doctors at the research sites; Dr. Xuan, Dr. Nga, Dr. Khoa, Dr. Thuc, Dr. Hoa, Dr. Hop, Dr. Minh, Thao. Thank you Dr. Xuan for help with nurses training for your needed involvement. To my faithful translators; Ngo Bella, Ngo Mai, Nguyen Tu, Nguyen Hoa, Nguyen Nga, and Vu Hai thank you for all your personal sacrifice for this project. To my wonderful Vietnamese family who adopted me, and treated me like a son. Dr Frank Wieringa for inspiring me as he bridged the difficult worlds of academia and development work, effecting large scale health benefits to millions of people who do not even realize it. Dr Wieringa walks the tension of being an academic involved in research and development projects and also having a wonderful family in the context of foreign country. Thank you for all you taught me, and your example.  I owe a debt of gratitude to my parents, Rodger and Leslie Brown.  xi  Dedication  To benevolent God, who is not inhibited by our situations and can turn passing individual suffering into the healing of many. Though him, his grace, his son and new life we are empowered to make this surrounding world a little more like heaven.  xii  I INTRODUCTION 1.1 Effects of Chronic Under-Nutrition among Women Worldwide, over one third of deaths in childhood are attributed to malnutrition (Black 2007) and over 65% of the world's undernourished children live in Asia. Vietnam has high rates of chronic malnutrition at ~40% (Nguyen 2005; Black 2008) with higher rates in rural provinces. Low maternal Body Mass Index is a risk factor for infant mortality (Han Z. 2011). In Vietnam 26% of women are underweight (GNS 2000) which is a serious public health concern (20-39%) (WHO 2010). Rice is the main staple of the Vietnamese diet and it is uncertain what risk factors are leading to these high levels of underweight among rural living women and children where sufficiency of rice is available. Indeed, in 2007 Vietnam was the third largest rice exporter in the world (Prakash 2011). The rural province of Bình Phước in the South East region of Vietnam is a highly productive agricultural region, which has high rates of maternal underweight status and provides an ideal microcosm to explore household food insecurity and other possible risk factors for the high rates of undernutrition. Socio-economic factors, such as poverty and rural living, commonly are associated with underweight status among women and children (GSO 2004), but do not adequately explain the current situation of high malnutrition rates in Vietnam (Thang and Popkin 2003). Recent economic development has increased per capita income, but has resulted in only minimal change to the malnutrition rate among children (Thang and Popkin 2003). Furthermore, recent economic development has benefitted people living in the cities more than those living in rural areas. In rural provinces, such a Bình Phước,  1  childhood underweight prevalence is over 2.5 times higher in rural areas as compared with urban households (UNICEF 2010). A factor that has not been adequately studied in Vietnam is household food security, as it may be a possible determinant of the high rate of underweight status among women. Household food security is defined as: “people having at all times both physical and economic access to sufficient food to meet their dietary needs for a productive and healthy life” (FAO 1996). A significant literature gap is that, as of 2006, no surveys have used primary food security measures, such as the Household Food Insecurity Access Scale (HFIAS) score, for quantifying household food insecurity. It is unknown if HFIAS scores are related to underweight status of women in Vietnam. Tools for primary measures of food security are the Food And Nutrition Technical Assistance (FANTA) HFIAS questionnaire (Coates 2007) or secondary proxy indicators such as dietary diversity scales (Hoddinott J. 2002), Absolute Poverty Rate (APR), Gini index, Gross Domestic Product (GDP) per capita and national food production. One group of researchers used food insecurity from secondary, extrapolated income figures in Vietnam (FAO 2004) to identify groups vulnerable to food insecurity. The FAO survey identified small farmers in both the southern and northernmost parts of Vietnam, fisherman in the central area of Vietnam and urban labourers without stable income as the most food insecure groups (FAO 2004). But unfortunately, there is little data on primary indicators of household food security and its determinants in Vietnam.  2  A further gap in the Vietnam specific literature is the identification of determinants of household food insecurity. These determinants might include nutrition knowledge, dietary diversity, household demographic data and traditional psychosocial dietary behaviors such as beliefs around ‘healthy’ foods.  1.2 Study Area and Population A focus on undernutrition among rural women is essential, as women’s chronic energy deficiency is correlated with Intra-Uterine Growth Restriction (Han, Mulla et al. 2011), growth faltering (<2 y) (Shrimpton 2001) and maternal and infant mortality (Black, 2007: WHO, 2004). The rural province of Bình Phước, in the South East region of Vietnam, was selected to explore determinants of underweight status and household food security among women in Vietnam. Bình Phước has a high prevalence of underweight, between 30-35% (Nguyen 2005), among both women and children (<5 y) which is considered high public health concern by the World Health Organization (WHO) (Black, Allen et al. 2008). Also, there is a duality of sufficient food production and under-nutrition. There are extensive plantations of rubber, rice, cashews, coffee and pepper. In SE Asia, variability in underweight status, Infant Mortality Rate (IMR) and vitamin A deficiency (VAD) rates between countries is very high, which rules out strictly environmental or geographical factors and demands further research (World Bank 2006). Bình Phước borders the eastern edge of Cambodia with some of the highest malarial rates in Vietnam (Abe, Honda et al. 2009). In the survey site of Bình Phước is a mixture of high poverty and some relative affluence with the provincial average income of 695,000Dong/month (VLSS 2006) or conservatively 1.45$ USD/day (16,000Dong/1USD) keeping in mind the Absolute 3  Poverty Rate (APR) was previously defined as income < 1.08$USD/day Purchasing Power Parity (PPP) (Chen and Ravallion 2007). Rural rates of childhood underweight are 259% higher than in urban centers (Nguyen 2000) which is surprising due to Bình Phước’s relative proximity to the largest urban center in Vietnam, Ho Chi Minh City. Ethnically, the population is predominantly Kinh, the main ethnic group in Vietnam, and 19% are ethnic minority groups (i.e. X’tieng, Mnông and Nùng) (Abe, Honda et al. 2009; VNA 2002). There are many culturally related dietary practices especially concerning women (Lundberg and Trieu 2011; Wadd 1983) especially during pregnancy and post-partum where certain foods are perceived as unhealthy and to be avoided. Bình Phước enables study into culturally propagated perception and attitudes toward what is considered ‘healthy’ food. This diverse province with differing ethnic groups, a mixture of urban and rural living, and unknown levels of household food security is an important region to identify possible determinants of the relatively high rates of under-nutrition close to Vietnam’s largest city with determinants that remain largely unexplored.  4  II REVIEW OF LITERATURE 2.1 Chronic Under-nutrition among Women and Children The South East Asian country of Vietnam has been making great strides in economic development and decreasing under-nutrition. The economy of Vietnam has historically been agrarian based and focused on rice production with 80% of the population living rurally (Thang and Popkin 2003). However, after Vietnam’s ascent into the World Trade Organization (WTO) in 2006 the economy and exports have been greatly increased and have opened the way for large multinational corporations such as Intel to start development of large industrial factories. Vietnam is a mix of towering new 60 storied development projects overlooking rice patties and thatched houses. From 1999 to 2009 alone, the GDP/capita increased by 80% (World-Bank 2012). However, maternal and childhood chronic under-nutrition in Vietnam are still high and both are over 20% (Nguyen 2000) which is a public health concern (Black, Allen et al. 2008). Maternal undernutrition and poor infant feeding practices in Vietnam are Figure 2.1 Geo-political Vietnam  © 2012 Google Map data © 2012 Google  5  associated and infant mortality (UNICEF 2010). Maternal underweight raises significant concern as it is a risk factor for both preterm and low birth weight babies (Han, Mulla et al. 2011). A relationship between maternal underweight and child (<2 y) growth faltering has also been established (Shrimpton 2001; Han 2011). Underweight status defined as a Body Mass Index (BMI) <18.5 kg/m2 is an indicator for Chronic Energy Deficiency (CED). The rates among women (20-49y) in Vietnam are very high at 26% nationwide (Nguyen 2000) but have improved from 1990 rates of 33%. Maternal underweight (BMI<18.5kg/m2), is further associated with small for gestational age neonates (Harita, Kariya et al. 2012), antenatal anemia (Sebire, Jolly et al. 2001), Intra-Uterine Growth Restriction (IUGR) , preterm babies, low birth weight babies(Han 2011), and increased risk of infant morbidity (Villar and Belizan 1982; Ronnenberg, Wang et al. 2003). Currently, Vietnam has a high rate of low birth weight babies (<2.5kg at birth), ranging between 7% in 2000 to 5% in 2005 (UNICEF 2010). Furthermore, in Vietnam underweight women have 40% increased risk of giving birth to a small for gestational age infants (Ota, Haruna et al. 2011) compared to women with BMI >18.5. Maternal underweight is associated with childhood underweight status in Vietnam and high levels of underweight and malnutrition are found among children (-2σ Weightfor-Age Z-score (WAZ)) when compared with countries with similar levels of quality of life, education, income per capita and life expectancy. These factors are included in the Human Development Index (HDI) and Vietnam has a higher prevalence of childhood underweight (25% WAZ <5y) than countries with a similar HDI (0.53) (WHO 2003; UNDP 2011). For example, Morocco, with an HDI of 0.52 has a childhood underweight prevalence of 9.6%. Also, Vietnam has higher rates of underweight (<5y) than Laos or  6  Cambodia when adjusted for HDI (WHO 2003; UNDP 2011). In 2006, Vietnam’s HDI and quality of life lagged behind China’s by 20 years. When adjusted for this, Vietnam still has 2.5 times greater underweight prevalence (WAZ <5y) than China (WHO 2003; UNDP 2011). Collectively this suggests that Vietnam’s high rate of chronic energy deficiency is not fully described by HDI, income, healthcare and education suggesting that other factors may be involved.  2.2 Known Determinants of Underweight in Vietnam The National Institute of Nutrition (NIN) of Vietnam as well as the General Statistics Office (GSO) of Vietnam has worked extensively at surveying and mapping undernutrition, using anthropometric data. However, national surveys have mainly examined four risk factors as they pertain to chronic energy deficiency among women; demographics (age), poverty, rural setting, and belonging to an ethnic minority group to give an overview of risk factors in Vietnam (GSO 1998: GNS 2000). Poverty, rural living, and ethnicity are not independent factors as the majority of ethnic minorities live in rural and poor areas. However, the highest prevalence of chronic energy deficiency exists among poor and rural living ethnic minorities (GSO 1998; GNS 2000). What follows is a more in-depth discussion of the major factors previously studied concerning chronic under-nutrition rates specifically: poverty, rural setting, and belonging to an ethnic minority group.  2.2.1 Poverty and Under-nutrition Household poverty is a risk factor for chronic energy deficiency among children in Vietnam (Thang and Popkin 2003). Vietnam has been making great advances in its economic development as a country and as of 2007 Vietnam was the world’s third largest 7  rice exporter (Prakash 2011). The implemented policy of Đ i Mới (“reformation”) in 1986 allowed for more privatization of the economy and has stimulated economic growth measured by GDP to over 8% in 2006 (World Bank 2012). Average income (GDP/capita) is increasing, as population growth in Vietnam is only 1.35%, and most health indicators are also moving along this positive trend. Infant mortality rates decreased from to 37 per 1000 live births in 2000 (GSO 2002) to 27 per 1000 in 2006 (Source Data: World Bank 2006). The mean life expectancy is 67 years (GSO 2002). Both of these changes are quite impressive given the current GDP/capita is less than that of Sudan (Source Data: World Bank 2006). However, Vietnam’s development has lagged behind especially concerning the issue of maternal and child under-nutrition. In 1999 the prevalence of underweight children under the age of five years in East Asia was 19%, but Vietnam’s percentage was a shocking 46% (Thang and Popkin 2003). The early years of Đ i Mới saw an increase economic expansion and household income, but minimal change in malnutrition rates among children (Thang and Popkin 2003).  2.2.2 Rural Setting and Under-nutrition Though poverty is a factor, it does not fully explain the high rate of under-nutrition in Vietnam, a prominent rice exporting country. Recent economic development has seemed to benefit mainly urban dwelling households and thus rural households are 17.6% more likely to be malnourished than urban households (Thang and Popkin 2003). Most malnutrition is found in rural areas (Nguyen 2000). The General Nutrition Survey (Nguyen 2000) found underweight prevalence of women (20 to 49y) was 28.3% in rural areas versus 20.5% in urban areas, or about 40% higher in rural areas. Underweight rates  8  among women are lowest in the urban center of Ho Chi Minh City (Nguyen 2000). However, 80% of the population are rural living farmers (Thang and Popkin 2003). Maternal and infant mortality rates around the periphery of these urban centers are also lower than in extremely rural areas such as the northern mountainous region or the central highlands (UNICEF 2010).  Women are involved in farming and especially wet  rice cultivation, which without machinery requires a high degree of physical labour which, in turn, increases caloric need. Furthermore, rural villages are more vulnerable to natural disasters affecting household income such as typhoons and flooding that can devastate crop growth. Also vulnerability to disease and the decreased access to health centers is an issue. This lack of accessible health care widens the disparity between urban and rural centres where poverty, higher levels of illiteracy, higher fertility rates and lower education are socioeconomic factors are involved.  2.2.3 Ethnic Minorities and Under-nutrition One prominent risk factor of chronic underweight among women is belonging to an ethnic minority group. There are over 54 minority ethnic groups in Vietnam and there is a disproportionate presentation of malnutrition among these minority groups (Thang and Popkin 2003). Being part of an ethnic minority increases the probability of being malnourished by 14.1% (Thang and Popkin 2003). Ethnicity is not an entirely independent risk factor (Malqvist, Hoa et al. 2012) for malnutrition. Ethnic minority groups tend to live in predominantly rural areas and have difficulty obtaining access to health care and give birth without trained supervision (UNICEF 2010). Ethnic minorities tend to have larger families and also a higher prevalence and severity of malnutrition. However, even after controlling for predictors 9  of rural living and poverty, ethnic minority status is a significant risk factor (UNICEF 2010). Ethnic minorities are strongly predisposed to high household malnutrition that is still poorly understood in the literature (UNICEF 2010). This may be due to poor nutrition knowledge or dietary practices. Currently high rates of childhood malnutrition among ethnic minorities have been attributed to “irrational child raising methods” (UNICEF 2010). This is an unacceptable justification. It is true, however, that cultural values, perceptions and dietary behaviors are thought to play a large part in malnutrition in Asia. In unpublished work from the National Institute of Nutrition (2010) a study showed that 27.2% of ethnic minority mothers discard colostrum (Le 2011). Furthermore, 27.3% ethnic minority people in the central highlands gave juice or foods before breastfeeding after birth. A strong link to these practices and maternal and child malnutrition has yet to be well described in Vietnam and will be discussed further.  2.3 Identifying Additional Risk Factors of Underweight among Women These four previously identified risk factors of demographics (age), poverty, rural setting, and ethnic minority groupings are not sufficient to fully explain the high levels of underweight among women in Vietnam. There are still poorly studied risk factors predisposing women to underweight. This is further indicated in the South East Region, both with a major urban center (Ho Chi Minh City) and rural Bình Phước province. This is a region with high rice production, a relatively high standard of living in the country, and counter intuitively, high rates of underweight prevalence among women. It has been suggested that household level food insecurity and/or sociocultural factors may play a  10  large part in the underlying chronic under-nutrition in Vietnam (FAO 2004; UNICEF 2010). Previous surveys have not explored combined factors of household food insecurity, household demographics (specifically pertaining to risk factors of IMR such as child spacing and number of children <5 y), vocation, and cultural dietary behaviours, nutritional education, cultural perceptions on the importance of specific food items as risk factors in underweight status among women. Identifying the determinants and risk factors for underweight among women may give insight to what possible interventions would be efficient and efficacious in decreasing chronic energy deficiency and/or food insecurity and what target groups to focus on. The information gathered may also help suggest characteristic target groups to implement interventions and improve the nutritional status of women in Bình Phước.  2.3.1 Household Food Insecurity Currently, it is unknown if household food insecurity is a factor in underweight among women in Bình Phước due to the region’s food sufficiency. Food security indicators are sensitive to immediate risk of under-nutrition and provide a quantifiable scale measuring stable access to food, availability of food and the biological utilization of food (FAO 2005). Additionally, food security has been found to predict underweight status among women in other studies. For food security to be attained, food needs to be available, accessible and people must make appropriate use of the food. Hygienic and sound nutritional choices must be made and sources of food must be stable. Food security is dependent on many large-scale factors such as fluctuations in the job market, conflict, natural disasters or even  11  household level factors such as sickness or death in the family. These can all have a large effect on household food security. Surveying the access and availability of nutritious food can be quantified using the Household Food Insecurity Access Scale (HFIAS) which is the “gold standard” for measuring food insecurity (Coates 2007) and is found in Appendix B.5. The HFIAS questionnaire is a collection of nine questions where the respondent recalls the last thirty days. The questions take into account three food security domains; 1) anxiety and uncertainty, 2) insufficient quality and 3) insufficient food intake and its physical consequences” (Coates 2007). Respondents are then asked to give the frequency of occurrence for each question which are tabulated into HFIAS scores. The HFIAS scores are most sensitive to wasting followed by underweight status and least sensitive to stunting (FAO 2005). The HFIAS scale has a specific algorithm for calculating the HFIAS categories of mild, moderate and severe food insecurity among the household. These categories do not necessarily correlate with the overall HFIAS score. For example if a respondent consumes a less diverse diet than desired once or twice in the last month, that respondent is automatically categorized as moderately food insecure.  12  Table 2.3.1 Defining Household Food Insecurity Access Scale (HFIAS) by category of food insecurity Frequency Question  Rarely  Food insecurity domain: anxiety and uncertainty Worried about food Food secure Food insecurity domain: insufficient food quality Ate less preferred foods than desired Mild Ate less variety than desired Mild Ate undesirable foods Mild Food insecurity domain: insufficient food quantity Ate less food than felt needed Moderate Ate fewer meals than desired Moderate No food in the house* Severe Went to sleep hungry* Severe Lacked food for over 24hrs* Severe  Sometimes  Often  Mild  Mild  Mild Moderate  Mild Moderate  Moderate  Moderate  Moderate Moderate Severe Severe Severe  Severe Severe Severe Severe Severe  * Used also in HHS scale There were some issues of external validity in comparisons of the HFIAS groupings with studies done in different countries therefore the Household Hunger Scale (HHS) is the Food and Nutrition Technical Assistance’s (FANTA’s) response to this issue of poor external validity. The HHS is a distillation of three questions of the HFIAS questionnaire and has strong external validity at the national level regardless of the cultural situation (Deitchler 2010). The HHS cutoffs for categorizing levels of household hunger in situations of severe food insecurity are straight forward.  13  Table 2.3.2 Household Hunger Scale (HHS) scoring and categorization of cumulative cutoffs of household hunger Frequency Question No food in the house Went to sleep hungry Lacked food for over 24hrs  Rarely Sometimes Often +1 +1 +1  +1 +1 +1  +2 +2 +2  Mild or no hunger 0-1 Moderate household hunger 2-3 Severe household hunger 4-6  Literature on food insecurity in Vietnam is extremely sparse and primary indicators of household food security have never been studied in Vietnam before, let alone in Bình Phước. One FAO project (FAO 2004) profiled food insecurity but used income as a proxy indicator. This focus on income, given Vietnam’s poorly explained discrepancy between higher malnutrition rates and relatively higher levels of income (Thang and Popkin 2003) may therefore not be a valid in this context.  2.3.2 Low Dietary Diversity Secondary household food security indicators can also be derived from coping strategies such as low dietary diversity. Dietary diversity is a proxy indicator for household food insecurity as it is a common household coping mechanisms to decrease dietary diversity by consuming cheaper monotonous foods (WFP 2008). Focusing on coping mechanisms is an important identifying trait of food insecure groups. Coping mechanisms such as selling productive capital (e.g. plowing oxen) to pay medical bills, or eating seeds intended for planting for immediate needs or migration are strongly correlated with food insecurity and can trap families in food insecurity for decades. The World Food Program’s (WFP) Food Consumption Score (FCS) is a quantifiable measure 14  of dietary diversity. The FCS is based on a 7-day recall, un-weighed, Food Frequency Questionnaire (FFQ). An un-weighed FFQ only records the frequency of the foods consumed, and not the portion size nor amount eaten. The FCS then divides the food frequency data into 8 main food groupings. These groups are then scored based on relative nutrient density by the FCS weighting system which places heavy importance on Animal Source Foods (ASF) such as meats and dairy. This methodology is simpler and more reproducible than other methods of analysis of FFQ data to observe significant dietary diversity patterns and trends. Table 2.3.3 Food Consumption Score weighting coefficients of food groups of the Food Frequency Questionnaire (FFQ) used in Bình Phước, Vietnam Food groups Items FCS weight Rice (noodles, porridge) Cassavas, potatoes  x2  Beans Nuts  x3  green leafy vegetables Carrots, yellow sweet potatoes  x1  Mango, papaya Fruits (pineapple, bananas, etc.)  x1  Pork, beef, poultry, other Eggs Organs  x4  Milk  Milk (fresh, condensed)  x4  Sugar  Sugar (sugary foods, cake)  x 0.5  Oil  Cooking oil, fat  x 0.5  Staples Pulses Vegetables Fruit Meat  Condiments Fish sauce, soy sauce *adapted from WFP FCS (WFP 2008)  x0  The FCS cut-offs for poor, borderline and adequate dietary diversity are represented by FCS values of 0-28, 28-42 and >42 respectively. These are the values used in countries where oil and sugar are consumed in high frequencies, such as Vietnam. 15  One weakness of the FCS is it does not consider portion size. Thus one might have high dietary diversity but consume only small amounts of nutrient dense foods. For example, a diet that consists primarily of rice but includes meat, albeit in exceedingly small portions, would still score high on the FCS. However, this diet may be relatively low in actual total nutrient intake. In Vietnam the common form of milk consumed is sweetened condensed milk, but this is consumed in small amounts compared with western intakes of milk. Also sweetened, condensed milk has about a third of the calcium, protein and twice the refined sugar when comparing isocalorically to 1% milk (U.S. Department of Agriculture 2011). However, the FCS does not distinguish these differences. As an additional limitation, the FCS relies on an assumed non-variability of portion sizes of commonly consumed foods between different cultures.  2.3.3 Demographics Risk Factors of IMR Maternal underweight status is implicated in higher IMR (Black, Allen et al. 2008) and poor maternal nutrition during pregnancy directly correlates with neonatal (<1y) mortality (World Health Organization 2005). Maternal underweight in Vietnam contributes significantly to infant mortality (Swenson, Nguyen et al. 1993). Maternal occupation and education level has been found to be related to underweight among children in Vietnam (Hien and Kam 2008). It is unknown if risk factors of IMR in Vietnam such as fertility rates (Talwalkar 1981), number of children (>5y), age of mother at first birth (Friede, Baldwin et al. 1987) and birth spacing (Osrin, Costello et al. 2005) have predictive value of underweight status among women in Vietnam. The specific effects of the education of women on their underweight prevalence have been poorly expressed in the literature. In 2000, overall figures including men and 16  women suggest a non-linear trend with those having secondary school education with slightly higher rates of underweight than those with less education and even among illiterate participants (Ha do, Feskens et al. 2011). Participants with post-secondary education had the lowest risk of underweight among Vietnamese adults (Ha do, Feskens et al. 2011). The association between maternal education and underweight has been better studied with mother-child pairs. However, contradictory evidence still exists in Vietnam. In other developing countries, higher maternal education is associated with decreased childhood chronic under-nutrition in (De Onis et al 2003; Black, Allen et al. 2008; Semba, de Pee et al. 2008). Further studies found that maternal education is negatively correlated with maternal mortality among women giving birth at health care facilities (Karlsen, Say et al. 2011). In Vietnam, in a recent small study in rural Nghe An province (n=383) a significant correlation with maternal education and childhood malnutrition was found (Hien and Kam 2008). However; this is disputed in Vietnam, as the large national surveys (Vietnam Living Standards Survey) have shown little effect of maternal education on the stunting rates among children (Thang and Popkin 2003). In 2006, Vietnam had the highest level of childhood underweight (<5y WAZ) compared to any country with comparable mean years of schooling (6.5y for women and 7.9 for the men (>25y)) (WHO 2003; GSO 2004). Nevertheless, it is still hypothesized that maternal education affects child growth, and underweight status and health in Vietnam (Nguyen, Eriksson et al. 2012). Fertility rates, defined as the total number of children per woman, have long been observed to have a direct relationship with IMR (Talwalkar 1981). In Vietnam, the age  17  of the mother at the birth of her first child is also significant in terms of IMR. There seems to be a U-shaped distribution for risk of infant mortality and maternal age. The IMR for children born to a mother <20y was almost twice that of those born to mothers aged 20-29 y (GSO 2002). The IMR then increases for mothers aged 30-39. Infant birth order is positively correlated with IMR according to the VLSS (GSO 2002). The IMR for first born infants (20 per 1000) was a third lower than the IMR for the fourth to sixth child (GSO 2002). This also is not strictly an independent factor as higher birth orders correspond with larger families, and larger families are common in poor, rural areas which these surveys did not account for. Increasing birth spacing is correlated with decreasing IMR. IMR is twice as high for infants born less than two years apart, than for those born 2-3years apart (GSO 2002). The IMR is than halved again for children born over 4 years apart. Again this not strictly an independent factor as planned parenthood, contraceptive use, education and household income are all associated with birthing intervals. Thus far, these demographic predictors of IMR have not been explored as a predictor of underweight among women which is also associated to IMR.  2.3.4 Nutrition Knowledge of Women A large research gap exists in Vietnam concerning nutrition knowledge, attitudes and perceptions and their effects on underweight among women. Some work has found associations between nutrition knowledge and dietary intake among women (Ogle 2007). Nutrition education is an important intervention for decreasing under-nutrition among children in situations of food sufficiency (Bhutta 2008 ). It is unknown if poor nutrition knowledge is related to underweight status in Bình Phước. Assessing participant  18  familiarity with common macronutrients and common micronutrient deficiencies (Vitamin A Deficiency (VAD), iron deficiency and iodine deficiency) will give us insights into whether low nutrition knowledge as a possible risk factor to underweight status among women. Further, it is unknown if being able to identify good food sources of; protein, vitamin A/β-carotene and iron are associated with significant increases in consumption frequency of those foods among women in Bình Phước. Knowledge of common micronutrient deficiencies were assessed for iron deficiency (Laillou, Pham et al. 2012), VAD and borderline vitamin A status (Laillou, Pham et al. 2012), which rates are 14%, <2% and 14% respectively. Iodine deficiency is also a concern due to a reoccurrence since the government’s de-regulating of salt iodization in 2005 (Fisher 2011). Exploring the associations between nutrition knowledge with both underweight status and household food security may be important to identify low nutrition knowledge as a potential risk factor.  2.3.5 Identifying Cultural Psychosocial Dietary Behaviours Cultural dietary behaviours in South East Asia have a history of giving rise to nutritional deficiencies. Historically, 89% of the population chew Betel nuts (Hoang 2000) which contain thiaminases, and therefore, contribute to vitamin B1 deficiency. Currently, there are many dietary and cultural practices in Vietnam that pertain to antenatal and postpartum states. For at least one month postpartum the Vietnamese have a great number of cultural behaviours and practices that could affect women’s health. Great importance is placed on the practices of Vietnamese women as to what they are to do: sit by a fire, strictly rest, practice strict sexual abstinence, avoid bathing in cold water and specific dietary behaviours. The women must eat more of or avoid eating certain 19  types of foods depending on the cultural perception of “beneficial” or “harmful after pregnancy” (Lundberg and Trieu 2011; Wadd 1983). Specifically, identifying these dietary beliefs and behaviours pertaining to times of nutritional significance such as perinatal, postpartum and during childhood illness is of particular importance for women’s health. Furthermore, there are poor infant feeding practices such as: non-exclusive breast feeding, discarding of colostrum milk, feeding juice or rice water to the neonate before breastfeeding. Maternal food restriction in the first trimester of pregnancy among Vietnamese women is quite high (Le 2011). In Vietnam, exclusive breastfeeding rates for the first four months was about 16% for urban and 21% for rural areas (Nguyen 2005). In comparison, the WHO prescribes exclusive breastfeeding for six months (WHO 2003). Even studies with Vietnamese immigrants in Australia, found exclusive breast feeding practices to be poor (McLachlan and Forster 2006). The Lancet nutrition series found that child mortality could be reduced by 9.1% through following six-month exclusive breast feeding recommendations. Further reductions of 1.5% are projected with post six month complementary feeding programs (Black 2007). Though these are blanket figures, they are quite significant and it remains to be seen if these factors are at play in Vietnam. In rural areas exclusive breastfeeding is higher than major cities where rates of exclusive breastfeeding (six months) was 17% (GSO 2008). The WHO/UNICEF/UNAIDS compiled seven studies and found a median of 30% exclusively breastfeed their children in Vietnam (Stina 2004). It was often perceived by women that giving rice water to their infant was healthier than breast milk. One small study in Melbourne found among Vietnamese immigrants (n=100), 40% gave their infant baby formula directly after giving birth (McLachlan and Forster 2006).  20  Vietnamese often discard colostrum milk, which is high in antibodies and nutrients for newborns (Morse, Jehle et al. 1990). Additional, cultural practices that could be involved propagating high malnutrition rates in Vietnam are maternal food restriction during the first trimester of pregnancy, poor infant feeding practices and poor knowledge of weaning foods (Nguyen 2008). Also, giving low nutrient density rice porridge, which is perceived as healthy for ill children, may be a factor. Culturally related food beliefs could lead to possibly detrimental nutritional practices and identifying these practices could help in correcting them through education and other health initiatives. Looking at cultural dietary behavior necessitates a look at the importance placed on certain types of foods according to a traditional psychosocial branch of Chinese medicine called ‘food theory’ which is reflected in the Vietnamese culture. Within ‘food theory’ there are some specific important food items to eat to “increase blood” or “help with lactation” in and around pregnancy. There is a belief that foods can have “heat” or have “cold” properties, not pertaining to temperature or spice; these properties are considered important during different stages of pregnancy (Chen 2007). Certain types of foods are prescribed for illness and also according to the seasons. These are culturally accepted but not universal in Chinese medicine (Chen 2007) and still need to be further explored for identifying positive, socially accepted, dietary behaviours and commending the beneficial practices (Ladinsky, Volk et al. 1987). Women following detrimental cultural dietary behaviours may be a risk factor for underweight or food insecurity among women in Vietnam. Identifying common culturally propagated ideas about nutrition during pregnancy, postpartum and during weaning is important as these may have measurable impact on risk of underweight status  21  and food insecurity and be useful for suggesting viable interventions. This study sought to identify cultural perceptions and dietary behaviors which predisposed women to different dietary behaviors leading to malnutrition.  2.4 Study Objectives and Rationale The primary objective of this survey is to describe both the prevalence of underweight status (BMI < 18.5 kg/m2) and household food insecurity among women (>18y) in the rural province of Bình Phước. The secondary objective is to identify risk factors associated with both underweight status of women and household food insecurity in Bình Phước. Risk factors explored in this study included; 1) Food insecurity, 2) demographic predictors of IMR, 3) dietary diversity, 4) traditional psychosocial dietary behaviors concerning pregnancy, postpartum and feeding of ill children and 5) nutrition knowledge. This analysis could give insight as to what modifiable risk factors are associated with underweight and household food security among non-pregnant women (>18y) in Bình Phước.  22  III METHODS AND MATERIALS 3.1 Participants Data collection for this cross sectional survey was carried out between November and December 2006 in the rural province of Bình Phước, 90kms North East of Ho Chi Minh City. This province was selected because of its high prevalence of unexplained underweight despite being close to an urban center with high relative standards of living and food sufficiency. The National Institute of Nutrition (NIN) thought that poor dietary behaviours might explain, in part, the high rates of underweight. The random sampling of village study sites were done by an independent scientific committee (NIN). Six villages within Dong Phu commune and four within Dong Xoai commune were selected. In the commune of Dong Phu the villages of Tan Lap, Dong Tien, Dong Tam, Tan Hoa, Tan Hung, Thuan Loi were surveyed. In the commune of Dong Xoai the villages of Tien Thanh, Tan Thanh, Tien Hung and Tan Xuan were selected. The goal was to recruit 50 women per village for a total planned sample size of 500. Recruitment started at the demographic center of the village as described by health officials. From there, women (> 18 y) from surrounding households were invited to participate in the study. Women were selected within each village and a ‘random walk’ method was used to approach other subsequent households. All women who met the selection criteria (see below) were invited to participate until ~50 women were recruited in each village. The researcher and trained nurses and doctors conducted the interviews. Participants were interviewed either at home or at the local health centre. Participant inclusion criteria were as follows: female, over 18 years, having no physical or mental disabilities preventing informed consent, and residing within the study area at  23  the time of the study. A small stipend ($0.60 CAD) was given to each participant, regardless of whether or not the interview was completed. The study was approved by the NIN and the University of British Columbia behavioral ethics committee (H06-80632) and all participants gave informed consent. Figure 3.1 Bình Phước survey sites  © 2012 Google Map data © 2012 Google  3.2 Survey Tools All survey instruments were adapted and translated for use in Vietnam by the NIN in Hanoi. Household demographic information pertaining to vocation, demographic risk factors of IMR (fertility rate, child-spacing, number of children (<5y)) and education level were obtained by the questionnaire. This questionnaire’s intent was to observe possible interactions of fertility and indicators of socio-economic status with underweight and food insecurity indicators. 24  3.2.1 Household Food Insecurity Household food insecurity was measured using the Household Food Insecurity Access Scale (HFIAS) questionnaire, which quantifies the access, quality and availability of nutritious food. The HFIAS survey consists of nine validated questions, which assess three broad domains: 1) anxiety and uncertainty, 2) insufficient food quality; and insufficient food quantity and 3) its physical consequences (Coates 2007). The participant responses indicate a frequency of occurrence of; never, rarely (1to 2 times), sometimes (3 to 10 times), and often (>10 times) for each of the questions, over the previous 30 days. This is then used to calculate HFIAS scores (Coates 2007). HFIAS scores range from 0 to 27 with a higher score indicating greater food insecurity. The categorical classification of the HFIAS is given in Table 2.3.1. The categories are not simply based on numerical cutoffs as certain questions have differing weight according to the question and frequency of the experience. The last three questions of the HFIAS were used to calculate the Household Hunger Scale (HHS). The three questions inquired about whether participants ‘had no food in the house’, ‘went to sleep hungry’ or ‘lacked food for 24hrs’. The HHS score codes the frequency where scores of never = 0, rarely to sometimes = 1 and often =2 (Deitchler 2010). This results in a score from 0-6 score. HHS food security is then ranked from mild food insecurity (0-1), moderate (2-3) to severe (4-6) and has high external validity (Dietchler 2010).  3.2.2 Dietary Assessment Dietary assessment was completed using a modified version of a Food Frequency Questionnaire (FAO/Nutrition and Consumer Protection Division, version May, 2007). The Food Frequency Questionnaire (FFQ) did not include questions on portion size. The 25  unmodified FFQ focused on 16 food item groups, namely: Cereals, β-carotene rich tubers, white tubers, dark green leafy vegetables, other vegetables, β-carotene rich fruits, other vegetables, iron/vitamin A/B12 rich organs as well as meats, eggs, fish, legumes, milk, oils/fats, sweets and condiments. The FFQ was modified to a total of 20 groups and consumption of soft-drinks, juice, soups/broths and beer/alcohol was assessed. Consumption of tannin rich foods such as tea and beer was also questioned as they inhibit iron absorption. This was done to highlight possible micronutrient deficiency risk in the population as well as provide a proxy indicator of food security as a coping strategy. Frequency of consumption responses were given a per week range of 1-3days/4weeks, 1 day per week, 2-4 days per week or 5-6 days per week, every-day and finally 1-3 days per month (~4 weeks). For frequencies of 2-4 and 5-6 days per week the median of 3 and 5.5 days per week were used, respectively. Frequencies less than once a week were disregarded as they may overestimate results. The FFQ was scored using the World Food Program’s (WFP) Food Consumption Score (FCS). The FCS is a nutrient density weighting scale which heavily favors nutrient dense Animal Source Foods (ASF). However, due to the median of weekly ranged responses being used to extrapolate a 7day recall FFQ data used by the FCS, the score used in this study was inappropriate for both external comparison, and comparison with the WFP thresholds of poor (< 28), borderline (28.5 to 42) and adequate dietary diversity (<42) (WFP 2008). However, these scores can be useful and differentiate between tertiles of dietary diversity scores among participants. It can also observe ASF consumption between high moderate and low dietary diversity in respect to underweight status, and food insecurity indicators.  26  It is also important to note the timing of the survey in regard to the harvest, as there is a significant seasonality of food insecurity in rural areas due to the juxtaposition of harvest and lean times. However, the study area, typically has three rice harvests instead of the two normally found in Northern Vietnam and therefore access to a more consistent food source is available.  3.2.3 Anthropometric Assessment Anthropometric measurements were carried out using an electronic scale (Taylor, CP7324.10.05, China) and height and measurement boards supplied by the health centres in Bình Phước. Underweight prevalence ( BMI<18.5 kg/m2) was the main anthropometric indicator and was further categorized by using WHO standards for mild (BMI 18.5-17 kg/m2), moderate (BMI 16.99-16.00 kg/m2), and severely (BMI<16 kg/m2) underweight (James, Ferro-Luzzi et al. 1988).  3.2.4 Nutrition Knowledge Nutrition knowledge was surveyed using a thirteen item questionnaire and included open-ended, free response questions. This questionnaire was designed to assess familiarity with macro and micronutrients as well as the ability to identify food sources of protein, iron, vitamin A/β-carotene, iodine and identifying specific nutrient deficiencies (goiter, night blindness, and iron deficiency anemia). This questionnaire is found in Appendix B.4. Foods identified as ‘good sources’ of protein, vitamin A/β-carotene and iron were defined by the Nutritive Composition Table of Vietnamese Foods (Giay 2000). This was used to explore if knowing good sources of these micronutrients increase the consumption frequency those food items. Identification of good food sources of nutrients was done in a free response form and participants responses were truncated at  27  three responses per question. Good sources of iron were defined as >3.80mg/100g for meat sources and >5.00mg/100g plant sources (Giay 2000). Sources of Retinol Activity Equivalent (RAE) were set for plant sources at β-carotene >1615mcg, and animal sources >180mcg (Giay 2000). Sources of iodine considered correct were: iodized salt, salt [general], seafood and avoidance of cabbage. Due to a maximum of three responses collected for each question one incorrect answer negated one correct answer. Thus the participant with three responses had to identify two good sources to be considered correct. For example, if participant responses to ‘good sources of protein’, defined as >12% protein g/100g (Giay 2000), were “meats, soybeans and fruit” the net score would be + 1 as they would have identified two correct good sources of protein and one incorrect source. Responses that were broad and indistinct (i.e. ‘fruits’ are a good source of vitamin A) were excluded due to lack of specificity. Micronutrient tablets were not considered as proper response for identifying food items. These total responses were entered and coded with a score ranging from 0 to 10, and segregated into high, moderate and low nutrition knowledge tertiles. This un-validated approach is within the scope of the assessing the nutrition knowledge women in Bình Phước to identify associations between nutrition knowledge, food security and underweight status.  3.2.5 Psychosocial Dietary Behaviours and Perceptions Cultural perceptions and traditional dietary behaviors were assessed by a twelve item questionnaire that included free response questions. Questions probed what food items were perceived as ‘healthy’ or to be avoided during pregnancy and are found in Appendix B.2. Questions also pertained to behaviours concerning foods consumed one month postpartum, during weaning or during childhood illness as well as taboos and importance  28  placed on specific trimesters of pregnancy. This was to attempt to identify specific cultural, traditional thoughts and dietary behaviors that may be predictor variable for underweight prevalence among women.  3.3 Data Analysis Descriptive statistics on indicators of BMI as well as food security HFIAS, HHS, dietary diversity as outcome measures were carried out using Statistical Package for the Social Sciences v.20 (SPSS, Inc., 2012, Chicago, IL, www.spss.com). The primary indicators of food insecurity were HFIAS score, discrete HFIAS categories and discrete HHS categories. A significance level of P<0.05 was used throughout the analysis. Data was assessed for normality using the Shapiro-Wilk test and then plotted graphically. Kurtosis and skewness was also taken into consideration. All outcomes were normally distributed with the exception of HFIAS score that could not be normalized with logarithmic transformations and the nonparametric Krurskal-Wallis test was used to explore differences between groups. Missing data in regard to household food insecurity and underweight were tested for randomness using Little’s Chi-squared Missing Completely At Random (MCAR) test, then MCAR data was subject to list-wise deletion. Missing at random entries were also tested; however, imputation by Expectation Maximization (EM) was not used as HFIAS uses ordinal groupings and EM imputation uses the mean rather than a whole number, and does not decrease the Standard Error (SE) as it imputes probability distribution around the mean and therefore does not decrease standard error. Inconsistent entry data was checked for out-of-range values in Microsoft Excel (2010) and were excluded.  29  3.3.1 Sample Size The sample size was determined using the following formula: D= Where Z = 1.96 (two tailed α=0.05), n = sample size, Pestimated= estimated prevalence rate and D is the difference between the estimated versus the actual prevalence with 95% confidence (Daniel 1999). The range between the estimated versus the actual prevalence (D) with 95% confidence was set to be <5%. Based on an estimated prevalence of underweight in Bình Phước of 26% (GNS 2000) and a sample size of 500 it would be possible to detect an underweight prevalence to within 3.9% with 95% confidence. Also, because primary indicators of Household Food Insecurity Access Scale (HFIAS) food security prevalence has not been previously studied in Vietnam, the prevalence of severe food insecure groups was conservatively estimated at 50% (Naing 2006). Therefore, based on a sample size of 500, the estimate of the prevalence of food security will be within 4.4 % with 95% confidence.  3.3.2 Analysis of Underweight Status The outcome measures of underweight status and food security were subsequently analyzed to identify predictor variables. BMI was used both as a dichotomous outcome measure between underweight and non-underweight prevalence and as a continuous outcome measure. Logistic regression explored predictor variables between underweight (BMI<18.5) and non-underweight (BMI>18.5) women.  30  3.3.3 Analysis of Household Food Insecurity HFIAS scores were categorized into two groups, one ranging from food secure to mildly food insecure versus the other ranging from moderately to severely insecure, for logistic regression analysis. Associations between HFIAS scores and dietary diversity were explored along with other possible predictor variables including age, family size, education, nutrition knowledge, dietary behaviours. Linear regression analysis of HFIAS categories was not opted for as relationships were not described well linearly. Use of multinomial regression was also rejected due to the relatively small group sizes of secure, mild and moderately food insecure groups with a mean sample size of 65 (10.8SD) being compared with the severely insecure group with 235 samples. Trends in dietary behaviour concerning pregnancy, postpartum and weaning were explored using free response questions. Each participant had a maximum of three responses and a χ2 test or Fisher’s exact test where appropriate to identify significant (P<0.05) differences in responses for dietary perceptions and behaviours that were later explored in regression analysis for associations with underweight or household food security. Dietary diversity scores were analyzed by multivariate regression analysis to possible determinants of dietary diversity.  31  IV RESULTS 4.1 Underweight among Women A total of 471 female participants gave informed consent and participated in the survey. Subsequently, sixteen subjects were list-wise deleted due to missing data pertaining to HFIAS and/or anthropometric measures. Pregnant women (n=58) were further excluded from post-hoc regression analysis of anthropometric data due to confounding problems associated with BMI. The recruitment target of 500 participants was not reached due to time constraints. The final village (Tan Xuan) was allotted one day, and only 27 participants could be recruited at this time. Response rate was 99.9 % and no participants withdrew from the study. Usable data for the HFIAS questionnaire and anthropometric assessment was 94% of women and 99.8% for the nutrition knowledge questionnaire and FFQ.  Descriptive characteristics of the study participants  are given in Table 4.1. The mean age of the women was 34.6 y with a standard deviation of 10.6 y. Underweight prevalence (BMI <18.5) was 24.4% (95% CI; 20.2 to 28.7) among non-pregnant participants (n=397), with frequencies of mild, moderate and severe underweight of 15.4%, 5.5% and 3.5% respectively. Almost half of the women were laborers (189) followed by housewives and small business owners at 20.0% (79) and 18.2% (72) respectively.  32  Table 4.1 Characteristics of non-pregnant women >18y (n= 397) in Bình Phước, Vietnam 2006 Characteristic  Mean ± SD % Age 397 34.6 ± 10.6 Household size 390 4.5 ± 1.5 Married 389 94.6 Number of children 366 2.2 ± 1.3 0-1 children 33.2 2-3 children 53.7 >4 children 12.8 Maternal age at first child 366 22.0 ± 4.0* Mean child spacing 263 3.4 ± 2.3 <2y 22.2 2-3y 13.9 >3y 30.2 Number of dependents (<5yrs) 397 0.5 ± 0.7 Number of dependents (<18yrs) 397 1.6 ± 1.1 6.7 ± 3.1 Education level 397 46.3 Completed Grade 0-6 36.8 Completed Grade 7-9 16.9 Completed Grade 10-12+ Currently working 396 80.8 Housewife 20.0 Labourer 47.8 Agriculturalist 7.6 Small Business 18.2 Professional 6.3 Nutritional supplement use 397 11.4 Third Trimester food item avoidance 366 9.8 Betel nut consumption 397 0.8 Nutrition knowledge score (0-10) 397 5.5 ± 2.4 397 98.8 ± 33.4 Adjusted Food Consumption Score Weight (kg) 397 47.1 ± 7.2 Height (cm) 397 151.8 ± 5.0 BMI (Kg/M2) 397 20.4 ± 2.7 Overweight (>25) 5.5 Normal (25-18.5) 70.0 Underweight (<18.5) 24.4 Mild (18.5-17) 15.4 Moderate (16.9-16) 5.5 Severe (<16) 3.5 * Median and inter-quartile range  33  There were significant differences between villages with respect to HFIAS food security scores (P<0.001) and dietary diversity (P<0.001) given by a Kruskal-Wallis test and ANOVA respectively and these variations are further explored in Table A1.1.  4.2 Household Food Insecurity The mean HFIAS score was 6.7 out of a possible 27 (n=397) with a standard deviation of 5.7. However, as described in Table 4.2.1, due to the HFIAS categorization rubric, 51.4% (204) of participants were characterized as severely food insecure, which indicates a high prevalence of food insecurity. Only 13.9% (55) were classified as food secure. This was juxtaposed by the HHS indicator classifying just 2.0% (8) of participants as having severe food insecurity with hunger and 19.1% (76) of participants classified as severely food insecure with moderate hunger.  Table 4.2.1 Household food insecurity indicators categorized by the Household Food Insecurity Access Scale (HFIAS) and Household Hunger Scale (HHS) among nonpregnant women (n= 397) surveyed in Bình Phước, Vietnam Nov-Dec 2006 Indicators HFIAS scores: (0-27) Food secure Mildly insecure  n Mean ± SD % 397 6.7 ± 5.7 55 13.9 77 19.4  Moderately insecure Severely insecure  61 204 Household Hunger Scale (0-6) 397 Mild (0-1) 313 Moderate (2-3) 76 Severe (4-6) 8  15.4 51.4 0.8 ± 1.1 78.8 19.1 2.0  Of the three food insecurity domains, indicated in Table 4.2.2, ‘insufficient food quality” had the highest prevalence with 70% (276) experiencing some form of food  34  insecurity due to quality. The domain for participant’s experiencing perception of ‘anxiety and uncertainty’ and ‘insufficient food quality’ had prevalence rates of 52% (208) and 50% (200).  Table 4.2.2 Household Food Insecurity Access Scale (HFIAS) 30-day recall with grey scale indicating mild, moderate and severe categorization of food security among non-pregnant women (n=397) surveyed in Bình Phước, Vietnam Nov-Dec 2006 Never % (n)  1-2 times % (n)  3-10 times % (n)  >10 times % (n)  Domain: Anxiety and uncertainty Worried about food  47.6 (189) 40.8 (162)  14.6 (58)  21.7 (86)  22.9 (91)  Domain: Insufficient food quality  30.4 (121) 22.4 (89) 28.7 (114) 21.2 (84) 22.9 (91) 25.4 (101) 33.8 (134)  8.3 (33) 2.5 (10) 5.3 (21)  11.6 (46) 11.6 (46) 17.4 (69) 10.6 (42)  5.8 (23) 5.0 (20) 4.8 (19) 1.5 (6) 1.3 (5)  Participant responses  Ate less preferred foods than desired Ate less variety than desired Ate undesirable foods  40.6 (161) 53.4 (212) 35.5 (141)  Domain: Insufficient food quantity  50.3 (200)  Ate less food than felt needed Ate fewer meals than desired No food in the house Went to sleep hungry Lacked food for over 24hrs.  58.7 (233) 68.8 (273) 53.9 (214) 82.4 (327) 87.7 (348)  7.8 (31)  23.9 (95) 14.6 (58) 23.9 (95) 5.5 (22) 3.3 (13)  The high prevalence of severe food insecurity under HFIAS was due to the high number of participants (46%; n=183) who indicated that they had experienced having ‘no food in their house’. According to FANTA’s characterization rubric this automatically classified this sizable number of participants as severely food insecure.  4.3 Adjusted Food Consumption Score (FCS) and Dietary Diversity A further proxy indicator of household food security is the FCS favoring nutrient dense ASF to observe relative dietary diversity. As shown in Table 4.3.1, the dietary 35  diversity measured by the adjusted FCS found 96% of participants had adequate dietary diversity, which usually corresponds to low food insecurity. The adjusted FCS score that was used also suggests lower levels of food security as <1% was classified as having poor dietary diversity.  Table 4.3.1 Adjusted Food Consumption Score (FCS) among non-pregnant women (n = 397) surveyed in Bình Phước, Vietnam Nov-Dec 2006 Indicators  N Mean ± SD % Adjusted Food Consumption Score 397 98.8 ± 33.4 Adequate (>34) 381 96 Borderline (28.5-42) 14 3.5 Poor (<28) 2 0.5 Dietary diversity score was positively associated among women with a self-reported ‘vitamin’ supplementation of their children in multivariate regression models (P<0.05). There was no significant association between dietary diversity and vocation, number of children, maternal age at first child, specifically education level, nutrition knowledge or participants taking supplements. There was also no association seen for women who intentionally restricted nutritional intake during the first trimester of pregnancy or other dietary behaviours.  4.4 Nutrition Knowledge The ten-item questionnaire exploring baseline nutrition knowledge, including familiarity and ability to identify good sources of protein, iron, iodine and vitamin A/βcarotene and related deficiencies had the mean score of 5.5 with a standard deviation of 2.3 as in Table 4.4.1. Concerning micronutrient deficiencies, participants indicated they were most were familiar with goiter 95% (376), followed by anemia 82% (325) then  36  night blindness 70% (276). Also for women who correctly identified sources of protein (n=228) the consumption frequency of protein rich meat, eggs, fish, organs and beans were significantly higher in χ2 tests (P<0.001). This was not so with consumption of vitamin A and β-carotene rich foods P=0.36) and there was too low a frequency of correct answers for iron sources of iron to observe significant associations. Table 4.4.1 Nutrition knowledge scores of women survey participants (n=397) in Bình Phước, Vietnam Nov-Dec 2006 Nutrition questionnaire % (n) 1. Familiar with protein (n = 397) 2. Identified sources of protein (n = 254)* Correct [<12% protein g/100g] Incorrect 3. Familiar with night blindness (n = 397) 4. Identified nutritional treatment of night blindness (n=276)* Correct [β-carotene >1615mcg, Retinol >180mcg] Incorrect 5. Familiar with vitamin A (n = 397) 6. Identified sources of vitamin A (n=212)* Correct [β-carotene >1615mcg, Retinol >180mcg] Incorrect 7. Familiar with anemia or iron deficiency (n = 397) 8. Identified nutritional treatment of anemia (n = 325)* Correct [meats>3.80mg/100g, plants>5.00mg/100g] Incorrect 9. Familiar with goiter (n =397) 10. Identified nutritional treatment of goiter (n = 376)*  64.5 (254) 89.8 (228) 10.2 (26) 69.5 (276) 16.3 (45) 83.7 (331) 53.4 (212) 37.7 (80) 62.3(132) 81.9 (325) 11.7(38) 88.3 (287) 94.7 (376)  Correct [Salt/Iodized Salt/Avoid Cabbage] 73.9 (278) Incorrect 26.1 (98) *Max 3 responses per participant and one poor source response negates one good source response  37  Among participants familiar with goiter, 74% (278) indicated the nutritional treatment of goiter was iodized salt or salt consumption related treatment as shown in Table 4.4.2. Participants mentioning seafood and fish had a combined frequency of 4% (16). Furthermore, 1.3% (5) of responses correctly indicated ‘avoiding cabbage’, rich in glucosinolates. Table 4.4.2 Food items commonly* considered as nutritional for the treatment of goiter [Iodine deficiency] assessed by free response questions** among nonpregnant women familiar with goiter (n=376) in Bình Phước, Vietnam Food Item** % (n) 40.1 (154) Nutritional Treatment of Goiter Iodized salt (nresponses=316) Salt 32.7 (123) Fish (saltwater) 3.2 (12) Medicine 16 (6) Avoid cabbage 1.3 (5) Fish 1.1 (4) Seafood 1.1 (4) *Food items <1% excluded  **Max 3 responses per participant Participants had greater difficulty identifying vitamin A rich foods and only 53% (212) of participants were familiar with vitamin A as described in Table 4.4.3. For sources of vitamin A/β-carotene, general categories of ‘fruit’ and ‘dark green leafy vegetables’ had the highest response rates of 41% (87) and 35% (74) respectively. High sources of carotenoids (such as β-carotene) were papaya, carrot and mango, which had response rates of 19 % (41), 16% (33) and 4% (9) respectively. The highest vitamin A (retinol) source, liver, had a low response rate of 2% (4). One respondent noted the Vietnamese vegetable ‘gac’ is exceptionally high in β-carotene which is correct.  38  Table 4.4.3 Food items commonly* considered by participants as good sources of vitamin A assessed by free response questions** among women (n=212) in Bình Phước, Vietnam Food Item (IU/100g) IU/100g % (n) Fruit 41.0 (87) Sources of vitamin A (nresponses=398**) Dark green leafy vegetables 34.9 (74) Papaya 950 IU 19.3 (41) Carrot 16,706 IU 15.6 (33) Fish 30 IU 12.7 (27) Meat 9.0 (19) Oranges Tomatoes Mango Prawn Egg Banana Liver [pork] Grapes Beets  225 IU 489 IU  7.1 (15) 5.2 (11)  1,048 IU 180 IU 700 IU 64 IU 17,996 IU 100 IU 33 IU  4.2 (9) 3.8 (8) 2.8 (6) 2.4 (5) 2.4 (5) 1.9 (4) 1.4 (3)  *Food items <1% excluded  **Max 3 responses per participant  Only 12% (38) of participants could identify good sources of iron according to the Nutritive Composition Table of Vietnamese Foods (Giay 2000). Liver and pork blood were the best sources at 4% (13) and 1.5% (5) found in Table 4.4.4. Participants identified specifically beef as a good source of iron. This is interesting due to the lower consumption of red meat suggesting participants knew beef is higher in iron than pork or fish. Response rates for both beef and meats were 10.8% (35). There were no responses for increased simultaneous eating with vitamin C rich foods or avoiding tannin rich teas.  39  Table 4.4.4 Food items commonly* considered by participants as good sources of iron assessed by three free response questions among women (n=325) in Bình Phước, Vietnam Food Item** % (n) 10.8 (35) Sources of Iron Beef 10.8 (35) (nresponses=303) Meat Fish 8.9 (29) Fe Tablets 8.0 (26) Beet 5.8 (19) Prawn 5.5 (18) Carrot Dark green leafy vegetables  5.2 (17) 4.6 (15)  Fruit Crab Liver Tomatoes Egg Potatoes Blood Bean  4.6 (15) 4.0 (13) 4.0 (13) 3.4 (11) 2.2 (7) 1.8 (6) 1.5 (5) 1.5 (5)  Bone  1.2 (4)  *Food items <1% excluded **Max 3 responses per participant  Further, contrasting the high degree of knowledge concerning iodine, only 57% (228) of participants could identify one good source of protein in Table 4.4.5. However, participants who could identify good sources of protein often gave more than one example, thus nresponse=303.  40  Table 4.4.5 Food items commonly* considered by participants as good sources of protein assessed by free response questions** among non-pregnant women familiar with protein (n=254) in Bình Phước, Vietnam Food Item % (n) Meats (general) 65.0 (165) Fish 63.0 (160) Egg 20.1 (51) Prawn 16.9 (43) Crab 10.6(27) Beef 9.8 (25) Dark green leafy vegetables Rice  6.7 (17) 6.3 (16)  Bean Fruit Soybean Pork Water morning glory Dog Milk  3.1 (8) 3.1 (8) 3.1 (8) 2.0 (5) 1.6 (4) 1.6 (4) 1.2 (3)  *Food items <1% excluded from graph **Max 3 responses per participant The response frequencies for nutritional treatment of night blindness (XN) was low and can be found in the appendix Table A2.1. Nutrition knowledge was strongly significantly correlated with maternal education (r=0.45, [95%CI 0.36 to 0.53], P<0.001)  4.5 Dietary Behaviours Among women with a history of pregnancy, significant behaviours and perceptions were observed on what types of foods should be eaten or avoided during pregnancy, one month postpartum, weaning and childhood illness. Firstly, Table 4.5.1 describes perceptions and participant response frequency between foods considered important between pregnancy versus one month postpartum was significantly different (P<0.001) according to χ2 and fisher’s exact tests. Response rates 41  of foods considered important one month post-partum were significantly higher for assorted meat (fowl, poultry and beef). Less importance was placed on fish and seafood, postpartum versus during pregnancy. Sticky rice and pork feet were not mentioned during pregnancy but were mentioned postpartum with a response rate of 5% (16) and 3% (9) respectively. Vitamin supplements were not mentioned. Table 4.5.1 Food items commonly* considered important to consume during pregnancy (nrespondents=266) and one month postpartum (nrespondents=357) assessed by three free response questionnaire among non-pregnant women with children (n=366) in Bình Phước, Vietnam Food items  Food groups Staples  Rice and rice products Sticky rice (fermented) Cassavas, potatoes  Pulses  Beans, legumes, nuts  Vegetables  Green leafy vegetables Carrots, yellow sweet potatoes  Fruit  Mango, papaya Fruits (pineapple, lychee, bananas) Meat (general) Pork Pig's feet Fowl, poultry, beef  Meat  Eggs Fish Seafood, crab, prawns  Pregnancy % (n) 54.1 (144) 0 (0) 0 (0)  Postpartum % (n)  0 (0)  1.4 (5)  17.7 (47) 1.9 (5) 1.5 (4)  22.1 (79) 1.1 (4) 0 (0)  24.8 (66)  12.6 (45)  59.0 (157) 0 (0) 0 (0) 0.8 (2) 6.4 (17) 57.5 (153) 12.4 (33)  60.2 (215) 5.3 (19) 2.5 (9) 14.3 (51) 5.6 (20) 36.7 (131)  55.4 (198) 4.5 (16) 0.6(2)  Organs, blood  2.3 (6)  3.1 (11) 0 (0)  Milk  Milk (fresh, condensed)  6.8 (18)  13.7 (49)  Sugar  Sugar (sugary foods, cakes)  4.5 (12)  1.1 (4)  Oil  Cooking oil, fat Bones  0.4 (1)  0 (0)  1.9 (5)  5.1 (19) 1.4 (5)  Condiments  Other  0 (0)  *Food items <0.5% excluded  42  During pregnancy, respondents perceived the most important food items were; meat (poultry, fowl and beef), fish and rice at 59% (157), 58 % (153) and 54% (154) respectively as illustrated in table 4.5.1. Previously pregnant participants increased consumption of specific food items (n=266) and indicated a significant (P<0.01) perceived importance of these foods in trimesters one by 90% (238) and trimester two by 94% (249) over the final trimester representing 35% (93) as presented in Table A3.5. Of women, with a history of pregnancy (n=366), 26.3% (100) did not intentionally eat more food or more of certain food types during pregnancy and 54.7% (217) avoided certain foods during pregnancy. Of women who avoided types of food during pregnancy, there was a significant (P<0.01) avoidance of food in the first trimester by 92.6% (201) but this was mainly due to ‘morning sickness’ 74% (160), followed by a simply ‘not being hungry’ 13% (29) as indicated in Table A3.5 and Table A3:6. Fish was avoided by 63.9% (170) followed by general meat consumption (poultry, fowl and beef) at 38.7% (103) shown in Table A3.3.  Significant perceptions and dietary behaviours concerning weaning and complimentary foods were shown and women with children predominantly selected rice pabulum and other rice products followed by milk (beast and dairy milk) as considered most important during the infant’s first year of life. Relatively low response rates were given for any type of protein rich complimentary foods and fruits were the second highest scoring complimentary food was fruit (general) at 11% (41) as displayed in Table 4.5.2.  43  Table 4.5.2 Food items commonly* considered important by non-pregnant women with children for weaning infants during first year of life assessed by free response questions** (n=366) in Bình Phước, Vietnam Food Items % (n) Staples  Rice (noodles, porridge) Cassavas, potatoes  Pulses  Beans, legumes, nuts  1.9 (7)  Vegetables  Green leafy vegetables Mango, papaya  2.5 (9) 0 (0)  Fruit  Fruits (pineapple, bananas)  11.2 (41)  Meat  Meat (general) Pork Eggs Fish  9.0 (33) 0.5 (2)  Milk  >100 (375) 0.5 (2)  2.2 (8) 6.0 (22)  Milk (breast, dairy, condensed) >100 (373)  Cooking oil, fat Miscellaneous Bones Oil  0.5 (2) 3.3 (12)  *Food items <0.5% excluded  **Max 3 responses per participant Responses for foods given to ill children were predominantly rice porridge and rice noodles 69% (182) followed by fruits and vegetables 30% (79) and beef and fowl 31% (81) meat. Among participants with sick children (n=264), 40% (106) participants responded saying this was simply due to choosing the food that is easier to eat and only 27% (71) of respondents thought rice porridge was a nutritious and ‘therapeutic’ food for illness. Of respondents, 19.9% (50) stated their child does not want to eat anything else (shown in Table A3.2) However; of all the dietary behaviours and perceptions, no significant association between underweight among women or HFIAS food insecurity categories were observed.  44  4.6 Predictor Variables of Underweight The secondary, analytical objectives were to assess hypothesized predictor variables of underweight and household food insecurity. Logistic regression explored predictor variables between underweight (BMI <18.5) and non-underweight women in Bình Phước as shown in table 4.6.1. Severe food insecurity was a significant predictor of underweight and analysis found the likelihood of underweight status among the mildly food insecure was about a third of that the severely food insecure (OR 0.35, [95% CI; 0.17 to 0.75], P<0.001). Participants with no children dependents, under five years old, had a significantly lower odds of underweight compared with those who had two or more dependents under five years old (OR 0.30, [95% CI; 0.14 to 0.65], p<0.01). Table 4.6.1 Logistic regression for underweight status (BMI <18.5) among associated factors of Household Food Insecurity Scale (HFIAS) categories and number of dependents (<5y) among non-pregnant women (n = 397) surveyed in Bình Phước, Vietnam 95% Confidence P Predictor variable n OR Interval value Dependents <5y 0 Dependents 222 0.30 0.14 to 0.65 <0.005 1 Dependent 140 0.53 0.24 to 1.17 0.12 2 Dependent 35 Ref* HFIAS Categories Secure 0.34 to 1.48 55 0.71 0.36 Mild <0.001 77 0.35 0.17 to 0.75 Moderate 0.70 to 2.46 0.40 61 1.31 Severe 204 Ref* Ref* = reference category The significant predictor variables of BMI from unadjusted multiple linear regression models were participant’s age and number of dependents (<5y) (P<0.01) and are given in appendix Table A1.6.  45  4.7 Predictors of Household Food Security The HFIAS scale scores ranged from 0-27 and univariate analysis (n=397) showed significant positive association in between age, the village study sites, particularly with Dong Tien and Tan Hoa being positively associated with increased HFIAS scores. Table 4.7.1 Univariate Linear Regression of HFIAS scores on determinant factors among non-pregnant women (n=395) surveyed in Bình Phước, Vietnam Predictor variable n Demographic Characteristics  Regression coefficient  95% CI  P value  Age (y) BMI (kg/m2) Vocation Housewife Labourer Agriculturalist Professional Small business Education level  395 395  0.05  0.01 to 0.09  <0.01  -0.05  -0.16 to 0.06  0.40  79 189 30 25 72  0.30 -0.01 0.50 -1.06 *Ref  -0.66 to 1.25 -0.87 to 1.45 -0.80 to 1.79 -2.45 to 0.32  0.54 0.27 0.45 0.13  Completed Grade 0-6 Completed Grade 7-9 Completed Grade 10-12+ Number of children 0-1 children 2-3 children >4 children  182 146 67  0.65 0.25 *Ref  -0.29 to 1.59 -0.65 to 1.15  0.18 0.58  132 212  0.29  -0.87 to 1.45 -0.41 to 1.45  0.63 0.27  Dependents (<5y) 0 1 >2  51  0.52 *Ref  221  -0.44  -1.67 to 0.78  0.78  139 35  -0.36 *Ref  -1.53 to 0.81  0.55  46  Predictor variable n Demographic Characteristics  Regression coefficient  95% CI  P value  35 43 45 40 44 42  1.67 2.47 1.50 2.99 1.14 1.35  0.07 to 3.27 0.90 to 4.05 -0.05 to 3.04 1.43 to 4.56 -0.39 to 2.68 -0.18 to 2.89  0.04 <0.01 0.06 <0.01 0.15 0.08  43 Tien Thanh 35 Tan Thanh 43 Tien Hung 25 Tan Xuan Food Security Characteristics  -0.05 0.24  -1.53 to 1.43 -1.30 to 1.78  0.95 0.76  1.07 *Ref  -0.42 to 2.56  0.16  -6.76 -4.79 -2.19 *Ref  -7.77 to -5.75 -5.69 to -3.88 -3.12 to -1.25  <0.01 <0.01 <0.01  -14.35 to -9.77 -9.06 to -4.50  <0.01 <0.01  -0.03 to 0.02  0.48  -2.11 to 1.39 -1.43 to 0.84  0.69 0.61  -0.88 to 0.64 -0.40 to 1.09  0.76 0.36  Village Tan Lap Dong Tien Dong Tam Tan Hoa Tan Hung Thuan Loi  HFIAS categories Secure Mild Moderate  55 77 61 202  Severe Household Hunger Scale 312 Low -12.06 76 -6.78 Moderate 7 *Ref High 395 -0.01 Adjusted FCS score Adjusted Relative Food Consumption Score 134 -0.36 High dietary diversity -0.29 Moderate dietary diversity 131 130 *Ref Low dietary diversity Nutrition Knowledge Characteristics Nutrition knowledge score Low Moderate High  142 107 146  -0.12 0.34 *Ref  47  Predictor variable  n  Regression coefficient  95% CI  P value  -0.07 to 1.86  0.07  Dietary Behaviour Characteristics Nutrition Supplements Yes No  47 348  0.89 *Ref  *Ref= Reference group  Similar results were found among women with children (n=360) with the exception of some dietary behaviour characteristics observed were mainly concerning pregnancy and postpartum and therefore only concerned women who already had children. This diminished the sample size and is shown in Appendix table A1.3. Factors of maternal age at first child, child-birth spacing, meat avoidance and restriction of protein rich (>12g/100g) foods and rice considered the primary weaning food were all insignificant. Food avoidance was also broken down by trimester, but was also non-significant. The only significant behavioral dietary factor observed was that 95 of 360 women who did not consciously try to eat more food or more types of foods during pregnancy had slightly lower levels of food insecurity (regression coefficient 0.97, 95%CI [0.24 to 1.70], P<0.01).  HFIAS scores were not normally distributed and furthermore, the data was not normalized by transformation. Therefore, a further non-parametric Kruskal-Wallis test explored the median HFIAS scores between participant characteristics shown in appendix table A1.5.  Correlations exploring the convergent validity of HHS, HFIAS categories and adjusted FCS dietary diversity as indicators for household food insecurity were explored  48  using bivariate analysis in Table 4.2.2. BMI was negatively correlated with HHS Score (r= -0.12, P<0.05). HFIAS score also had a negative association with BMI (r= -0.11, P<0.05) suggesting BMI was lower with higher food insecurity scores. There was a high correlation coefficient with HHS categories and HFIAS score (r=0.78, P<0.005). Dietary diversity was negatively associated with both food insecurity indicators HFIAS and HHS. However; no significant relationship was observed between dietary diversity scores and BMI.  Table 4.7.2 Spearman's rank test correlation coefficient of BMI with three food security indicators; Household Hunger Scale (HHS), Household Food Insecurity Access Scale (HFIAS) and adjusted Food Consumption Score (FCS) among women (n=470) in Bình Phước, Vietnam HHS HHS HFIAS Score Adjusted FCS BMI  HFIAS Score  Adjusted FCS  1 0.78**  1  -0.29*  -0.31**  1  -0.12*  -0.11*  -0.02ns  * P value (< 0.05) ns Non-significant Due to lack of strong linearity or ordinal relationship between different categories of foods insecurity, logistic regression was selected in lieu of ordinal or multinomial regression. Logistic regression analysis explored HFIAS food insecurity among nonpregnant women, grouped dichotomously between lower levels of food insecurity (secure to mildly) versus higher levels (moderate to severe). Three significant predictor variables were identified and are shown in Table 4.7.1. Women who had completed between grade six to nine had lower odds of being moderately or severely food insecure compared with those who had not completed at least grade six (OR 0.59, [95% CI; 0.36 to 0.97], 49  P<0.05). Relatively low dietary diversity defined by tertile was also a significant predictor to higher food insecurity (P<0.01). Also women with the lowest nutrition knowledge score had twice the odds of food insecurity (moderate to severe) than those with a moderate nutrition knowledge (OR 0.52, 95% CI [0.29 to 0.92], P<0.05). Furthermore, low relative dietary diversity scores from the adjusted FCS score found three times higher odds of high levels of food insecurity than among participants with high dietary diversity (OR 0.32, 95% CI [0.18 to 0.58], P<0.001). Associations with underweight was strongly significant (P<0.01), but in analysis was considered an outcome rather than a predictor of food insecurity.  Table 4.7.3 Logistic regression of Household Food Insecurity Access Scale from zero to mild food insecurity versus moderate to severe food insecurity with associated factors of education level, and nutrition knowledge and adjusted food consumption scores of non-pregnant women (n = 397) surveyed in Bình Phước, Vietnam Predictor variable Education level Completed Grade 0-6 Completed Grade 7-9 Completed Grade 10-12+  n  OR  95% CI  184 Ref* 146 0.59 0.36 to 0.97 67 0.47 0.25 to 0.89  <0.05 0.02  Nutrition knowledge Low Moderate High  143 Ref* 108 0.52 0.29 to 0.92 146 0.73 0.42 to 1.28  <0.05 0.28  Adjusted Food Consumption Score Low dietary diversity 134 Ref* Moderate dietary diversity 131 0.25 0.14 to 0.45 132 0.32 0.18 to 0.58 High dietary diversity Ref* = reference  P value  <0.001 <0.001  No significant associations were found with other factors of household demographics and socio-economic factors. 50  V DISCUSSION 5.1 Underweight Prevalence of Previous Studies This study found a high rate of underweight women with rates very similar to the previous national surveys. In this study, among non-pregnant women (25y-64y) (n=357), the rate of underweight was 24.9 % (95% CI: 20.4% to 29.1%), which is comparable to the 2005 national rates (n=8,483) of 22% (Ha do, Feskens et al. 2011). In 2000, data specific to Bình Phước and the South East region found 25.9% of women (20-49y) were underweight which was comparable the results of this survey (n=357), 24.9% (95% CI [20.4% to 29.4%]). This is down from the rate of 28% (n=7408) found in another 2000 survey (Khan and Khoi 2008). In terms of the dichotomy between rural Bình Phước and urban centers, discrepancies still exist and Bình Phước had higher underweight rates among women (2060y) than Ho Chi Minh city (19%, n=771) (Cuong, Dibley et al. 2007) or rates among urban women in Hanoi (Walls, Peeters et al. 2009).  5.2 Risk factors for Underweight Logistic regression analysis found only two main risk factors associated with underweight rates among women. The two factors were categorized HFIAS food insecurity and increasing number of dependents (<5 y). Both were positively correlated with underweight among women. This is noteworthy because the total fertility rate (total number of children per women) and other predictors of IMR (maternal age, birth spacing) showed no association with underweight status. No significant associations of socio-economic variables such as vocation with underweight status was found in regression models. Agricultural workers had the lowest  51  adjusted mean BMI of the vocations though differences were not significant. This finding was similar to a study in Vietnam looking at the rural province of Nghe An. This study reported no association between the vocation of women and their underweight status (Hien and Kam 2008) . This study was not able to observe a significant association between maternal education level and underweight status. Previous studies in Vietnam have shown no consistent relationship between underweight and education (Hien and Kam 2008; Khan and Khoi 2008). Surprisingly, in some studies with Vietnamese adults, participants who had secondary or high school education had a higher odds ratio for underweight versus those with illiterate and those with little or no schooling (Ha do, Feskens et al. 2011). However; this lack of significant effect in the current study may be due to a small sample size with insufficient power to detect such findings.  5.3 Risk Factors: Household Food Insecurity Household Food Insecurity was very high and 51% (204) of participants were characterized as severely food insecure. This was primarily due to participants experiencing a lack of food in the home within the last thirty days.  Household food  insecurity was a significant factor associated with underweight status. Women with moderate to severe food insecurity had twice the risk of being underweight. This was a significant finding as this study was the first to look at primary indicators of household food insecurity using the HFIAS questionnaire in Vietnam. In Bình Phước, the three factors with strongest effect size for household food insecurity were relative dietary diversity (adjusted FCS score tertile), and both the education of women and nutrition knowledge score. This is noteworthy because factors  52  such as maternal vocation, fertility rate and young primipara maternal age were insignificant. Household food insecurity studies in rural Guatemala found rural living, low socio economic quintiles, more children <5y were the major determinants of food insecurity (Chaparro 2012). In Bình Phước, it was interesting to note that the number of children (<5 y) was associated with maternal underweight but not with food insecurity. Unsurprisingly, the relative dietary diversity given by the adjusted FCS was strongly associated with household food insecurity, as dietary diversity is a proxy indicator of food insecurity. However, it was interesting to note that relative dietary diversity (high/moderate/low) and not the absolute, externally valid cut-offs, showed associations with household food insecurity. Nutrition knowledge was a predictor variable of food insecurity; however, no direct correlation was seen with underweight status. There have been conflicting studies on the effect of maternal education on nutritional status in Vietnam but these results found nutritional knowledge and education level of women both have a significant predictive value on food insecurity, which has not been demonstrated before in Vietnam. A further major issue was the lack of convergent validity of this indicator and the HHS, which is derived from the same questionnaire. This will be discussed further, in the limitations of the study.  5.4 Nutrition Knowledge Nutrition knowledge, while not associated with underweight among women was associated with the education level and household food insecurity. This association was significant and not previously examined among women in Vietnam. Household food insecurity consist of both access and utilization components. This association with  53  nutrition knowledge may indicate the utilization component of food insecurity, and not strictly the access component, may be involved in the high levels of severe insecurity found. This also is evidenced by socioeconomic factors (such as vocation) playing an insignificant role in predicting food insecurity in this study. Nutrition knowledge was generally poor concerning vitamin A and iron deficiency. Furthermore, among women with a history of pregnancy, responses on weaning foods strongly preferred rice over any other type of non-milk weaning food. This suggests that more complementary feeding education is needed. Thirty percent of women fed their ill children rice porridge because of perceptions that it is therapeutic and nutritious even though relatively low nutrient density and low protein intake and could be involved in the high rates of growth faltering among children (<5 y) but is beyond the scope of this study. However, these findings do stress the important need for nutrition education. Notably there was such little effect of self-reported vitamin supplementation between underweight status and food insecurity. This may be due to a lack of knowledge and familiarity of what nutritional supplements actually are. Many responses incorrectly indicated a topical, lipid based ointment used in Chinese medicine as a ‘vitamin’ supplement or tonic. This study found nutrition knowledge concerning iodine was the highest. Currently the governmental nutritional education programs (2001-2010 National Nutrition Strategy) are focused on public iodine education program. This high degree of familiarity with iodine deficiency may suggest the possible effectiveness of the government’s public education program in Bình Phước. Some participants even knew avoiding cabbage was  54  important for iodine status. This is due to thiocyanate which competitively inhibits iodine absorption (Han and Kwon 2009). Nutrition knowledge was also significantly associated with eating patterns. Participants who identified good sources of protein had higher intake frequencies (table A3.7) of protein rich foods (>12% protein/100g) (P<0.05). However, knowing good food sources of vitamin A or iron which commonly are associated with micronutrient deficiencies in Vietnam, and was not observed to be associated with consumption frequency of those foods or dietary diversity. This broad un-validated approach looking at ‘nutrition knowledge’ and perception, though not an externally valid or international recognized tool, gave this study insight into possible effective interventions as women with higher basic nutrition knowledge on vitamin A, protein, iodine and iron were less food insecure.  5.5 Dietary Behaviours and Perceptions Significant dietary behaviours and perceptions were reported around pregnancy, the postpartum period and for child rearing among women with a history of pregnancy (n=360). However, only one behaviour showed an association with increasing food insecurity (HFIAS scores). Among previously pregnant participants, 26% (95/360) did not consciously increase food consumption or food item consumption during pregnancy and had a slight but significantly lower level of food insecurity. This result seems conceptually anomalous though statistically significant. However, there is a 45% probability that at least one of the nine dietary behaviours observed would result in a Type I error due to the α value of 0.05. The possibility of a Type I error was also  55  suggested as other behaviours studied such as a cognitive decrease of foods or ASF during pregnancy showed no significance.  Other significant behaviours were observed concerning food consumption around pregnancy, perceived importance of nutrition during each trimester, weaning foods and the feeding practice of ill children. However; the significance of these behaviours were muted and did not correspond with significant effects as predictor variables for food insecurity or underweight. It is interesting to note that dietary behaviours that could be seen as detrimental such as cognitive decrease of protein rich meats during pregnancy and particularly during the third trimester showed no significant effect.  Previous studies have observed a lack of exclusive breastfeeding, discarding of colostrum and feeding of rice water to neonates in Vietnam (UNICEF 2010; Nakamori, Nguyen et al. 2010). This study found poor knowledge about weaning foods as rice was perceived as the most important weaning food. Studies in the 1980’s found the use of other poor infant feeding practices such as feeding exclusively dairy milk in Vietnam (Mathews and Manderson 1980). However, the current study also endeavored to identify what weaning foods were considered important among Vietnamese women. Practices involved with possible detrimental dietary practices were much lower than anticipated. It was initially suggested that traditionally lower birth weight babies were desired as it was thought smaller infants made the birthing process easier (Le 2011). However, in this survey only 2 of 217 women said the reason for caloric restriction or food avoidance during pregnancy was to have a smaller infant. Also, rates of betel nut chewing (implicated in B1 deficiency) was less than <1%.  56  The semi-qualitative free response nature of the questionnaire for dietary behaviours and feeding trends during pre/post-pregnancy used an un-validated, open-ended responses and thus is difficult to compare with other studies, but similarities were found with other studies in the cultural importance of postpartum dietary practices (Wadd 1983; Lundberg and Trieu 2011). For example, postpartum sticky rice and pork consumption among women was culturally important and reported as being beneficial in small semiqualitative studies in rural Vietnam (Le 2002). For pork consumption, and particularly pork feet were specifically noted to ‘increase lactation’ and was due to a traditionally held version of ‘Chinese food theory’ in Vietnam as suggested elsewhere (Le 2002). Specific food item avoidance such as culturally related decreased post-partum fruit consumption (Le 2002) trend was seen was not significant.  5.6 Dietary Diversity The adjusted FCS tertiles showing relative dietary diversity was found strongly significant in predicting lower levels of food insecurity (mild to secure) versus higher levels (moderate to severe) among participants. However, when absolute cut-off categories of FCS were used no significant association between food insecurity or underweight were observed. Furthermore, higher levels of dietary diversity suggested much lower levels of food insecurity than the HFIAS food insecurity classifications indicated.  It was interesting in the current study that no significant association was observed between dietary diversity and socioeconomic and demographic factors such as vocation of women, education level and number of child. Small business owners had the highest  57  level of dietary diversity, albeit non-significant. Conversely, professionals having the highest income are expected to have the highest dietary diversity but in this study had the lowest, albeit insignificant, dietary diversity. There was also no significant effect seen in self-reported ‘vitamin’ supplementation on dietary diversity which is interesting because usually in the Western context, individuals taking nutritional supplements often have confounding correlations with participants practicing other ‘healthier’ behaviors (Hoggatt 2003). One of the main predictors of dietary diversity was for households with children on supplements (P<0.01). Mothers with children taking nutritional supplements had much more diverse diet than children not on supplements, possibly showing better knowledge, or more care given to childhood growth. The diet of women in Bình Phước was found to be sufficiently diverse and animal source foods consumption frequency was 5.5 days/week for pork, beef, and poultry (see Appendix Table A3.7). The highest median consumption frequency was for rice, green leafy vegetables, fish/soy sauce and soup broth which is arguably not a particularly nutrient dense diet for underweight women. Plant and animal sources of β-carotene and iron were also low among participants. However, meat portion sizes differ greatly between North America and Vietnam and may overestimate the nutrients these women are consuming and a weighed FFQ or 24hr recall would give clearer indication of the true consumption of ASF, but was beyond the scope of this study.  58  5.7 Targeting Possible Interventions There was a significant association between nutrition knowledge and household food insecurity suggesting the possible need to increase nutritional education/training. Although this study focused on determinants of underweight and food insecurity among non-pregnant women, a need for training concerning complementary weaning foods was identified in Bình Phước. Nutritional training could be included on complementary feeding as it is effective at reducing childhood (<5 y) chronic energy deficiency (Black 2007). In this study the effectiveness of public nutrition education initiatives in Vietnam, was in some post-hoc way demonstrated with high overall participant’s familiarity with iodine.  In terms of a target population, pregnant women and particularly women pregnant within 5 years of a previous child’s birth are an important group due to associations with underweight status and food insecurity. Education interventions could suggest supplementation for women during pregnancy due to nationwide concerns of micronutrient deficiency. Nutrition education/training during pregnancy and pertaining to complimentary weaning foods and exclusive 6 month breastfeeding may be beneficial. Furthermore, zinc is commonly the most deficient nutrient in complementary food mixtures fed to infants during weaning (Winichagoon 2002) and zinc deficiency is extremely prevalent among women in Vietnam (Laillou, Pham et al. 2012) and could be an important training component to include. Training could be done at the health centers when women come for check-ups during pregnancy as 89% of women have some contact with a skilled health provider during pregnancy (GSO 2002). Nutrition training could  59  also utilize the volunteer health network of the Protein Energy Malnutrition Control Program (PEMCP).  Possible applications of this study suggested a possible need of nutritional training on weaning and complimentary foods could prove beneficial specifically for pregnant women with more than one child (<5y).  5.8 Study Limitations There were a number of limitations of this 2006 survey in Bình Phước province. The randomization of the village study sites were done by an independent scientific committee (NIN) in Vietnam and not by the principal investigator. However; no significant indications of selection bias were found as mean age between villages did not differ significantly also the underweight prevalence was similar to previous studies. Primarily one woman per household was selected but selection by random walk method allowed for multiple participants in each house. Further limitations that will be explored below are; a lack of convergent validity for the food insecurity scale, the FFQ’s frequencies were inadequate for external FCS comparison, the lack of non-validated questionnaire for observing dietary behaviors and feeding practices in the context of Vietnam and a lack of consensus and conflictions among influential bodies (FAO, WFP, FANTA) in defining broad categories (dark-green-leafy-vegetables) as good sources of either vitamin A or iron for quantifying nutrition knowledge scores.  5.8.1 Validity of HFIAS Categories There was great discrepancy between the high food insecurity suggested by HFIAS categories, and low food insecurity suggested by other food insecurity indicators (HFIAS  60  scores and dietary diversity). This classification of high levels of severely food insecure was mainly due of one question of experiencing ‘no food in the house’ which may not be a valid in the context of Bình Phước. These issues prompted major concerns with convergent validity between the high food insecurity described by the HFIAS categories and low food security suggested by low overall HFIAS scores, HHS categories or level of dietary diversity. The HHS, HFIAS score and higher levels of dietary diversity all indicated a low level of household food insecurity and household hunger. These indicators were correlated suggesting convergent validity in this regards. The degree of dietary diversity also suggested congruency with the HHS, describing lower household hunger than the severly food insecure catagories suggests. The HHS has robust external validity and further catagorizes severe food insecurity into including the experience of hunger. However, most of the severe food insecurity described in this study had low hunger. After this survey was carried out in 2006, FANTA issued a paper highlighting the HFIAS’s external validity and previous cut-offs are under review concerning this very issue, poor external validity. This suggests poor generalizability in different countries settings (Deitchler 2010) which is extremely significant to the discussion and findings of our study. FANTA’s confounding issue with HFIAS score, HFIAS catagories and HHS indicators influences the interpretation of this study, even though all are based off the same questionaire. Furthermore, FANTA give more unessicary complications such as using different classifications (Cook 2000), or in recent Guatamalian study used a further derivation using five questions of the HFIAS questionnaire (Chaparro 2012). All these major issues drecreases confidence in this measure, use of this questionnaire and the original scoring rubric set forth by FANTA.  61  A further strike against the external validity of the HFIAS questionnaire is that in developed countries, food insecure groups seems to be experiencing higher obesity rates (Wilde and Peterman 2006). Food insecurity is sometimes an important factor in underweight in developing countries but not in all situations. Environmental conditions and other determinant variables are also contributing factors (Chaparro 2012) which confound the validity of this construct and suggest possible validity issues of the HFIAS categories in the context of Vietnam.  5.8.2 Nutrition Knowledge Classifications: Green Leafy Vegetables In quantifying nutrition knowledge and identifying good sources of vitamin A and iron there is a major concern due to lack of international consensus for definitions (FAO 2002) for classifying dark-green-leafy-vegetables for vitamin A and iron content. Participants were asked to identify what foods were rich in vitamin A and iron, many responded dark-green-leafy-vegetables. The issue is that the nutrient component of this broad grouping of vegetables differ greatly depending local diet and the specific species of leafy greens common to the study region. Regardless; important bodies such as FANTA and the WFP and FAO still classify dark-green-leafy-vegetables differently in terms of RAE and iron status and are at odds with each other. For example, the WFP FCS describes dark-green-leafy-vegetables as a good source of β-carotene but overall as not a good source of RAE (FAO/WHO 2002). However; FANTA defines all dark-greenleafy-vegetables being considered rich in RAE rich (Swindale and Bilinsky 2006). Using FANTA’s approach to classifying all dark-green-leafy-vegetables as a good source of RAE would not have been valid in this study. This is primarily due to the concurrent high daily rates of green leafy vegetable consumption found in this study and 62  prevalent national level borderline VAD among women in Vietnam (Laillou, Pham et al. 2012). Thus women, by FANTA’s definition, are consuming “RAE rich” dark-greenleafy-vegetables every day. In terms of RAE sources of β-carotene to retinol conversions are 12 to one, other carotenoids 24 to one retinol but in a randomized control trial in Vietnam found may be upwards of 28 to one RAE (Khan 2007), possibly due to intestinal parasites. Β-carotene in yellow fleshed fruit has twice the bio-availability of green leafy vegetables (Khan 2007), further indicating, at least in this study region, that green leafy vegetables should be treated as a low RAE source.  Another approach is to categorize representative values and weighted means for food values from representative food items identified from FFQ data used in previous studies. Breaking down broad categories such as dark-green-leafy-vegetable was done in the VLSS 1998, 2002, 2004 data that showed consumption of similarly classed foods. For dark-green-leafy-vegetables category in Vietnam, Water morning glory (Ipomoea aquatic, or Rau muong) has the highest consumption in frequency (95% of households) and quantity (17.4 kg/capita) (GSO 1998). The consumption/capita was for water morning glory by far the single most consumed for any dark-green-leafy-vegetable. This Vietnamese FFQ data gave validation for choosing water morning glory as a representative food items for the nutrient composition of that group. However; data given by the VLSS had combined groups of other vegetables into a non-distinct ‘other vegetables’ group therefore adjusting the weighed means intake was not possible. Iron content was not given in the Nutritive Composition Table of Vietnam however, a separate study found water morning glory was high in iron (210.30mg/100g) (K.J. Umar 2007) but absorption from plant based diets are about 5%, three times less bioavailable than 63  heme-iron sources (15%) (WHO/FAO 2002). Additional increases in iron absorption above 50% can be observed, among anemic patients, however; surveying IDA was beyond the scope of this survey. Due to this confusion, it was considered that the threshold definitions of ‘good sources’ of vitamin A/iron/protein set forth by the Nutritive Composition Table of Vietnam was valid for assessing nutrient knowledge. Questionnaire amendments should prohibit the use of broad categories in the free response questions and only allow for specific food item examples.  5.8.3 Future Questionnaire Amendments Incorporating questions on breastfeeding practices, length of exclusive breastfeeding, occurrence of mothers discarding colostrum milk or feeding of rice water to neonates would bring a fuller picture to the dietary behaviour pertaining to pregnancy portion of the survey. Information on exclusive breastfeeding rates would give more insight and enable further external comparisons with other studies and possible correlate poor breastfeeding practices with levels of underweight and food security among women in Vietnam. Also for exploring nutritional knowledge, questions concerning zinc deficiency and sources of zinc should be added. Zinc deficiency is the most prevalent micronutrient deficiency among women in Vietnam with rates of 67.2 ±2.6% (n=1522) (Laillou, Pham et al. 2012). Future studies could include helminth data as infection rates in some regions of Vietnam are high (Lundberg and Trieu 2011) (Pasricha, Caruana et al. 2008). This nonnutritional factor maybe a large determinant of IDA, VAD and possibly underweight  64  prevalence among women and has been a major confounder of other papers in Vietnam (Khan, West et al. 2007). This may be especially significant among women who work directly with agricultural production and with those without proper access to latrines which can lead to re-infections (Do, Molbak et al. 2007). The FFQ would also benefit by separating the grouping of fresh fish from dried and shellfish as shellfish is high in zinc and river fish/rice patty fish are much lower in zinc content (U.S. Department of Agriculture 2011). Also concerning iron bioavailability, it would be beneficial to look at the intake of vitamin C rich, citrus fruits, as an enhancer of iron absorption. Also, using either a 24 hour recall or a FFQ with a full range of 7 day recall frequencies would enable further dietary diversity externally valid comparisons with other studies. Possible inclusion of ethnic minority status and clearer segregations of rural and urban characteristics would allow better comparisons with the national level surveys completed by the GSO, but ethnicity were outside of the ethical scope of our study.  65  5.9 Conclusion High rates of both underweight status and household food insecurity were found among women in Bình Phước province. This study found underweight status was predicted by higher levels of food insecurity and higher numbers of children (<5 y) in a family. Customary dietary behaviours such as caloric restriction during pregnancy and culturally related consumption/avoidance of certain food particularly around pregnancy and one month postpartum were observed but had no significant association with the underweight status of women in Bình Phước. Significantly higher levels of food insecurity, were correlated with low maternal education and lower nutrition related knowledge, thus suggesting knowledge, not vocation, had a higher predictive value on food insecurity in Bình Phước. Therefore, food insecurity was not purely dependent on household socio-economic status as described in national level studies. Also, it was noteworthy that though cultural dietary behaviours were observed, these behaviours had no significant association with food insecurity or the underweight status of women in Bình Phước. The strengths of the study were that it was the first study to use primary household food insecurity indicators and use logistic regression analysis to explore possible risk factors of both food insecurity and underweight among rural women in Vietnam. These primary indicators were observed for convergent validity with proxy measures of food insecurity such as dietary diversity. This study also had probed possible factors of cultural dietary behaviors and possible associations with basic nutrition knowledge. Some of the main limitations of the study were the lack of strong convergent validity between the HFIAS categories with other indicators of food security. Furthermore, the 66  questionnaires for nutrition knowledge and cultural dietary behaviours allowed for free responses answers preventing biases found in multiple choice questionnaires. This allowed for identification of dietary behaviours and perceptions, but the limitation of this was that is difficult to validate externally and therefore is inappropriate to compare with other studies in different settings. Nutrition knowledge on protein, vitamin A and iron was low and associated with household food insecurity. The need for nutritional knowledge on complimentary weaning foods was highlighted and nutrition training could be targeted to reproductive aged women and those who recently had a child (within 5 y) and are currently pregnant. This education could include complementary feeding education, education on good sources of protein. Women who could identify good sources of protein had a correspondingly higher consumption of protein rich food intake, and may prove beneficial to the health of women of Bình Phước. Future directions of research could include more nutrition knowledge questions pertaining to zinc deficiency. Also helminth data would be beneficial in analysis and may be a risk factor in underweight among women in Bình Phước. Longer term study could explore the relationship between maternal anthropometric data of mother/child parings and observe if pre-pregnancy underweight is a risk factor for longer term growth failure and childhood stunting (<5y) as the literature has noted predictors in IGUR, preterm children, and low birth weight babies, but the relationship with longer-term growth faltering (<5y) could be further defined in the context of Vietnam.  67  BIBLIOGRAPHY Abe, T., S. Honda, et al. (2009). Risk factors for malaria infection among ethnic minorities in Bình Phước, Vietnam. Southeast Asian J Trop Med Public Health 40(1): 18-29. Allen, L. and R. Shrimpton (2005). The International Research on Infant Supplementation Study: Implications for Programs and Further Research. The Journal of Nutrition 135(3): 666S-669S. Beard, J. L. (2001). Iron Biology in Immune Function, Muscle Metabolism and Neuronal Functioning. The Journal of Nutrition 131(2): 568S-580S. Bhutta Z., B. S., Black R., Brown K., Gardner J., Hidayat A., Khatun F., Martorell R, Ninh N., Penny M., Rosado J. (2000). Therapeutic effects of Oral Zinc in Acute and Persistent Diarrhea in Children in Developing Countries: Pooled Analysis of Randomized Controlled Trials. American Journal Clinical Nutrition (72): 1516–22. Bhutta ZA, A. T., Black RE, Cousens S, Dewey K, Giugliani E, Haider BA, Kirkwood B, Morris SS, Sachdev HP, Shekar M; (2008 ). What works? Interventions for maternal and child undernutrition and survival. Lancet 371(9610): 417-40 Black R., B. Z., Bryce J., Morris S, Victoria C. (2007). Executive summary: Maternal and Child Undernutrition. Lancet Series Maternal and Child Undernutrition. Black, R. E., L. H. Allen, et al. (2008). Maternal and child undernutrition: global and regional exposures and health consequences. The Lancet 371(9608): 243260. Brown KH, P. J., Allen LH, Rivera J. (2002). Effect of Supplemental Zinc on the Growth and Serum Zinc Concentrations of Pre-pubertal Children: a Metaanalysis of Randomized, Controlled Trials. Am J Clinical Nutrition (75): 1062–71. Caulfield, L. E., S. A. Richard, et al. (2004). Undernutrition as an Underlying Cause of Malaria Morbidity and Mortality in Children Less Than Five Years Old. 71: 55-63. Coates J., S. A., Bilinsky P. (2007). Household Food Insecurity Access Scale (HFIAS) for Measurement of Food Access: Indicator Guide. Washington, DC, Food and Nutrition Technical Assistance Project (FANTA) Academy for Educational Development. Cook, G. B. M. N. C. P. W. H. J. (2000). Guide to Measuring Household Food Security. Office of Analysis, Nutrition, and Evaluation Food and Nutrition Service. Alexandria, USDA. Cuong, T. Q., M. J. Dibley, et al. (2007). Obesity in adults: an emerging problem in urban areas of Ho Chi Minh City, Vietnam. Eur J Clin Nutr 61(5): 673-81. Chaparro, C. (2012). Household Food Insecurity and Nutritional Status of Women of Reproductive Age and Children under 5 Years of Age in Five Departments of the Western Highlands of Guatemala: An Analysis of Data from the  68  National Maternal-Infant Health Survey 2008–09 of Guatemala, FANTA-2 Bridge. Chen, S. and M. Ravallion (2007). Absolute poverty measures for the developing world, 1981-2004. Proc Natl Acad Sci U S A 104(43): 16757-62. Chen, X. D. (2007). Brief research on the theory of cold and heat of the foods in Zhou li (The rites of Zhou). Zhonghua Yi Shi Za Zhi 37(4): 248-50. de Onis, M. and M. Blössner (2003). The World Health Organization Global Database on Child Growth and Malnutrition: methodology and applications. International Journal of Epidemiology 32(4): 518-526. Deitchler M, B. T., Swindale A, Coates J (2010) Validation of a Measure of Household Hunger for Cross-Cultural Use. Food and Nutrition Technical Assistance II Project Do, T. T., K. Molbak, et al. (2007). Helminth infections among people using wastewater and human excreta in peri-urban agriculture and aquaculture in Hanoi, Vietnam. Trop Med Int Health 12 Suppl 2: 82-90. FAO (1996). Report of the World Food Summit. World Food Summit, Rome, FAO. FAO (2004). Food Insecurity and Vulnerability in Viet Nam: Profiles of Four Vulnerable Groups. ESA Working Paper No. 04-11, FAO. FAO/WHO (2002). Human Vitamin and Mineral Requirements. Bangkok, WHO/FAO. Friede, A., W. Baldwin, et al. (1987). Young maternal age and infant mortality: the role of low birth weight. Public Health Rep 102(2): 192-9. arc a-Casal, M. a. N., M. Layrisse, et al. (1998). Vitamin A and β-Carotene Can Improve Nonheme Iron Absorption from Rice, Wheat and Corn by Humans. The Journal of Nutrition 128(3): 646-650. Grantham-McGregor, S., Y. B. Cheung, et al. (2007). Developmental potential in the first 5 years for children in developing countries. Lancet 369(9555): 60-70. GSO (1998). Viet Nam Living Standard Survey (VLSS). S. P. House. Hanoi. GSO (2002). Viet Nam Living Standard Survey (VLSS). G. S. Office. Hanoi, Statistical Publishing house. GSO (2002). Vietnam Demographic Health Survey (VDHS). GSO, Hanoi, Committee for Population, Family and Children, GSO (2004). Viet Nam Living Standard Survey (VLSS). G. S. Office, Statistical Publishing house. GSO (2008). Viet Nam Multiple Indicator Cluster Survey 2006, UNICEF. Giay T., B. D., Ed. (2000). Nutritive Composition Table of Vietnamese Foods Hanoi, Medical Publishing House. Ha do, T. P., E. J. Feskens, et al. (2011). Nationwide shifts in the double burden of overweight and underweight in Vietnamese adults in 2000 and 2005: two national nutrition surveys. BMC Public Health 11: 62. Ha Huy Khoi, N. C. K., Tu Giay, Nguyen Trong Nhan, Hoang Viet Thang, and N. C. D. Nguyen Xuan Ninh, Nguyen Thu Nhan, Dao Ngoc Dien, Hoang Thi Luy (1999). Progress of Vitamin A Deficiency Control Program in Vietnam. Hanoi Scientific Publishing House.  69  Ha K., H. V., Le M. (1997). The Situation and Risk Factors of IDA in Vietnam: Nutritional Situation and Strategies for Action in Vietnam. Hanoi Medical Publishing House: 71-77. Han, H. and H. Kwon (2009). Estimated dietary intake of thiocyanate from Brassicaceae family in Korean diet. J Toxicol Environ Health A 72(21-22): 1380-7. Han Z., Mulla S., Beyene J., Liao G., McDonald S., (2011) Maternal underweight and risk of preterm birth and low birth weight: a systematic review and meta analyses. International Journal of Epidemiology. 40(1):65-101. Harita, N., M. Kariya, et al. (2012). Gestational bodyweight gain among underweight Japanese women related to small-for-gestational-age birth. J Obstet Gynaecol Res 38(9): 1137-44. Heijmans, B. T., E. W. Tobi, et al. (2008). Persistent epigenetic differences associated with prenatal exposure to famine in humans. Proceedings of the National Academy of Sciences 105(44): 17046-17049. Hien, N. N. and S. Kam (2008). Nutritional status and the characteristics related to malnutrition in children under five years of age in Nghean, Vietnam. J Prev Med Public Health 41(4): 232-40. Hoang, N. (2000). Vietnam: Image of the Community of 54 Ethnic Groups, The Ethnic Cultures Publishing House. Hoddinott J., Y. Y. (2002). Dietary Diversity as a Household Food Security Indicator. Food and Nutrition Technical Assistance Project, Academy for Educational Development,. Washington, D.C. Hoggatt, K. J. (2003). Commentary: Vitamin supplement use and confounding by lifestyle. International Journal of Epidemiology, 32: 553-555. James, W. P., A. Ferro-Luzzi, et al. (1988). Definition of chronic energy deficiency in adults. Report of a working party of the International Dietary Energy Consultative Group. Eur J Clin Nutr 42(12): 969-81. Jane Fisher, T. T., Beverley Biggs, Tuan Tran, Terry Dwyer, Gerard Casey, Dang Hai Tho & Basil Hetzel (2011). Iodine Status in Late Pregnancy and psychosocial determinants of iodized salt use in rural northern Vietnam Bulletin of the World Health Organization. K.J. Umar, L. G. H., S.M. Dangoggo and M.J. Ladan (2007). Nutritional Composition of Water Spinach (Ipomoea aquatica Forsk.) Leaves. Journal of Applied Sciences 7(6): 803-809. Karlsen, S., L. Say, et al. (2011). The relationship between maternal education and mortality among women giving birth in health care institutions: Analysis of the cross sectional WHO Global Survey on Maternal and Perinatal Health. BMC Public Health 11(1): 606. Kuczmarski, R. J., C. L. Ogden, et al. (2000). CDC growth charts: United States. Adv Data(314): 1-27. Khan NC, K. H., Giay T, Nhan NT, Nhan NT, Dung NC, Thang HV, Dien DN, Luy HT (2002). Control of vitamin A deficiency in Vietnam: achievements and future orientation. Food Nutr Bull 23(3): 133-42. Khan, N. C. and H. H. Khoi (2008). Double burden of malnutrition: the Vietnamese perspective. Asia Pac J Clin Nutr 17 Suppl 1: 116-8. 70  Khan, N. C., C. E. West, et al. (2007). The contribution of plant foods to the vitamin A supply of lactating women in Vietnam: a randomized controlled trial. Am J Clin Nutr 85(4): 1112-20. Ladinsky, J. L., N. D. Volk, et al. (1987). The influence of traditional medicine in shaping medical care practices in Vietnam today. Social Science &amp; Medicine 25(10): 1105-1110. Laillou, A., T. V. Pham, et al. (2012). Micronutrient Deficits Are Still Public Health Issues among Women and Young Children in Vietnam. PLoS ONE 7(4): e34906. Le, H. (2011). Nutritional Practices of Ethnic Minorities and Child Malnutrition in Mountainous Areas of Central Vietnam. Hanoi, Hanoi Medical University: 16. Le Minh Thi, W. P., Oratai Rauyajin (2002). Traditional Postpartum Practices among Vietnamese Mothers in An thi District, Hung Yen Province. Hanoi School of Public Health. Bangkok, Mahidol University, Thailand. Le Minh Thi, W. P., Oratai Rauyajin (2002). Traditional postpartum practices among Vietnamese mothers in Anthi distric, HungYen Provence. Public Health. Hanoi, Hanoi School of Public Health. . Leenstra, T., L. P. Acosta, et al. (2006). Schistosomiasis japonica, anemia, and iron status in children, adolescents, and young adults in Leyte, Philippines. The American Journal of Clinical Nutrition 83(2): 371-379. Lundberg, P. C. and T. N. Trieu (2011). Vietnamese women's cultural beliefs and practices related to the postpartum period. Midwifery 27(5): 731-6. Mahalanabis, D., K. N. Jalan, et al. (1976). Vitamin A absorption in ascariasis. Am J Clin Nutr 29(12): 1372-5. Malqvist, M., D. Hoa, et al. (2012). Causes and determinants of inequity in maternal and child health in Vietnam. 12: 641. Mathews, M. and L. Manderson (1980). Infant feeding practices and lactation diets amongst Vietnamese immigrants. Aust Paediatr J 16(4): 263-6. McLachlan, H. and D. Forster (2006). Initial breastfeeding attitudes and practices of women born in Turkey, Vietnam and Australia after giving birth in Australia. International Breastfeeding Journal 1(1): 7. Morse, J. M., C. Jehle, et al. (1990). Initiating breastfeeding: a world survey of the timing of postpartum breastfeeding. International journal of nursing studies 27(3): 303-13. My, L. ( 1997). The Situation of Iodine Deficient Disorders in Vietnam Scientific Symposium. National Institute of Nutrition, Medical Publishing House Hanoi. Nakamori, M., X. N. Nguyen, et al. (2010). Nutritional status, feeding practice and incidence of infectious diseases among children aged 6 to 18 months in northern mountainous Vietnam. J Med Invest 57(1-2): 45-53. Nam, M. a. H. P. G. i. V. (2008). Joint Annual Health Review 2007. D. L. Nguyen X., Nguyen L., Pham T., Bales S., Duong T. Nguyen, H., B. Eriksson, et al. (2012). Physical growth during the first year of life. A longitudinal study in rural and urban areas of Hanoi, Vietnam. BMC Pediatrics 12(1): 26.  71  Nguyen Khan, L. T., Phan Duong, Truong Son, Pam Hoan, Nguyen Tien, Nguyen Chung, Nguyen Phuong, Do Kiem, Nugyen Nga (2005). Nutritional status of children and their Mothers in 2005. N. I. o. Nutrition, Medical Statistics Publishing House. Nguyen, N. (2008). Caloric Restriction during the first trimester. Nguyen, S. (2000). General Nutrition Survey. K. Ha. Hanoi, Ministry of Health, National Institute of Nutrition. Nguyen, S. (2005). Protein Energy Malnutrition Control Program. Hanoi, Ministry of Health, National Institute of Nutrition. Nguyen Xuan Ninh, N. C. K., Ha Huy Khoi, Chuyen NV, Kawakami M. Lien DTK (1998). Actual nutrition problems in Vietnam and Japan. Medical Publisher Hanoi: 187-200. Nguyen Xuan Ninh, N. C. K., Ha Huy Khoi, Chuyen NV, Kawakami M. Lien DTK (1998). Micronutrient Deficiencies and Strategies for Controlling in Vietnam. Hanoi Scientific Publishing: 187-200. Nhien N, K. N., Ninh N, Huan P, Hop le T, Lam NT, Ota F, Yabutani T, Hoa V, Motonaka J, Nishikawa T, Nakaya Y (2008). Micronutrient deficiencies and anemia among preschool children in rural Vietnam. Asia Pac J Clin Nutr 17(1): 48-55. Nyberg W., (1958 ). Absorption and excretion of vitamin B12 in subjects infected with Diphyllobothrium latum and in non-infected subjects following oral administration of radioactive B12. Acta Haematol. 19(2): 90-8. Osrin, D., A. Costello, et al. (2005). Birth spacing and neonatal mortality. The Lancet 365(9457): 383-384. Ogle B., Hop T., Son T. (2007) Improvement of nutrition knowledge and practices among reproductive-aged women through participatory communication in two communes of yenthe district in bacgiang province, Vietnam. Malays J Nutr. 13(2):121-9 Ota, E., M. Haruna, et al. (2011). Maternal body mass index and gestational weight gain and their association with perinatal outcomes in Viet Nam. Bull World Health Organ 89(2): 127-36. Pasricha, S. R., S. R. Caruana, et al. (2008). Anemia, iron deficiency, meat consumption, and hookworm infection in women of reproductive age in northwest Vietnam. Am J Trop Med Hyg 78(3): 375-81. Prakash (2011). Safeguarding Food Security in Volatile Global Markets. P. A. Ronnenberg, A. G., X. Wang, et al. (2003). Low Preconception Body Mass Index Is Associated with Birth Outcome in a Prospective Cohort of Chinese Women. The Journal of Nutrition 133(11): 3449-3455. Saloojee H., P. J. (2001). Iron deficiency and Impaired Child Development. . BMJ (323): 1377-1378 Sebire, N. J., M. Jolly, et al. (2001). Is maternal underweight really a risk factor for adverse pregnancy outcome? A population-based study in London. Bjog 108(1): 61-6.  72  Semba, R. D., S. de Pee, et al. (2008). Effect of parental formal education on risk of child stunting in Indonesia and Bangladesh: a cross-sectional study. The Lancet 371(9609): 322-328. Shankar, A. H., B. Genton, et al. (2000). The influence of zinc supplementation on morbidity due to Plasmodium falciparum: a randomized trial in preschool children in Papua New Guinea. The American Journal of Tropical Medicine and Hygiene 62(6): 663-9. Shrimpton R., Victora CG, de Onis M., Lima RC., Blossner M, Clugston G. (2001) Worldwide timing of growth faltering: implications for nutritional interventions. Pediatrics. 107 (5):E75 Shrimpton R., G. R., Darnton-Hill I., Young M. (2005). Zinc Deficiency: What Are The Most Appropriate Interventions? BMJ (330): 347-349. Stina, A. (2004) Study To Guide Adaptation of Global WHO/UNICEF/UNAIDS Guidelines on HIV and Infant Feeding for Viet Nam. Volume, DOI: Swenson, I. E., M. T. Nguyen, et al. (1993). Factors influencing infant mortality in Vietnam. J Biosoc Sci 25(3): 285-302. Swindale, A. and P. Bilinsky (2006). Household Dietary Diversity Score (HDDS) for Measurement of Household Food Access: Indicator Guide. Food and Nutrition Technical Assistance Project (FANTA). Talwalkar, M. A. (1981). Association of infant mortality and high fertility: an empirical investigation. IIPS Newsl 22(1): 2-11. Tomáš Scholz, H. H. ., Roman Kuchta, Barbara Wicht (2009). Update on the Human Broad Tapeworm (Genus Diphyllobothrium), Including Clinical Relevance. Clin Microbiol Rev 22(1): 146-160. Thang, N. M. and B. Popkin (2003). Child malnutrition in Vietnam and its transition in an era of economic growth. J Hum Nutr Diet 16(4): 233-44. U.S. Department of Agriculture, A. R. S. (2011). SDA National Nutrient Database for Standard Reference, Release 24, Nutrient Data Laboratory Home Page. Umeta M, W. C., Haidar J., Deurenberg P., Hautvas J (2003). Zinc Supplementation and Stunted Infants in Ethiopia: A Randomized Controlled Trial Lancet 355(9220): 2021 - 2026. UNDP (2011). 2011 Human Development Report, UNDP. UNICEF (2010). An Analysis of the Situation of Children in Viet Nam. V. H. Kohn R., Nguyen G. UNICEF (2010). An Analysis of the Situation of Children in Viet Nam. Vietnam, UNICEF. Van Nhien N, K. N., Ninh NX, Van Huan P, Hop le T, Lam NT, Ota F, Yabutani T, Hoa VQ, Motonaka J, Nishikawa T, Nakaya Y. (2008). Micronutrient deficiencies and anemia among preschool children in rural Vietnam. Asia Pac J Clin Nutr. 17(1): 48-55. Villar, J. and J. M. Belizan (1982). The relative contribution of prematurity and fetal growth retardation to low birth weight in developing and developed societies. Am J Obstet Gynecol 143(7): 793-8. Vu Thi Thu Hien, N. T. L., Nguyen Cong Khan, Nguyen Tri Dung, C Murray Skeaff, Bernard J Venn, Trevor Walmsley, Peter M George, Judy McLean, 73  Matthew R Brown and Timothy J Green (2009 ). Folate and vitamin B12 status of women of reproductive age living in Hanoi City and Hai Duong Province of Vietnam. Public Health Nutrition 12 (07): 941-946. Wadd, L. (1983). Vietnamese postpartum practices. Implications for nursing in the hospital setting. JOGN Nurs 12(4): 252-8. Walls, H. L., A. Peeters, et al. (2009). Prevalence of underweight, overweight and obesity in urban Hanoi, Vietnam. Asia Pac J Clin Nutr 18(2): 234-9. Watanabe, K., R. Flores, et al. (2005). Early childhood development interventions and cognitive development of young children in rural Vietnam. J Nutr 135(8): 1918-25. Watanabe H. Inoue K, Doi M., Matsumoto M., Ogasawara K., Fukuoka H., Nagai Y., (2010). Risk factors for term small for Gestational age infants in women with low prepregnancy body mass index. J Obstet Gynacol Res. 36(3):50612. West, K. P., Jr. (2002). Extent of vitamin A deficiency among preschool children and women of reproductive age. J Nutr 132(9 Suppl): 2857S-2866S. West, K. P., Jr., R. P. Pokhrel, et al. (1991). Efficacy of vitamin A in reducing preschool child mortality in Nepal. Lancet 338(8759): 67-71. WFP (2008). Food Consumption Analysis: Calculation and use of the food consumption score in food security analysis. Rome, World Food Program. WHO (2003). Global Database on Child Growth and Malnutrition, The World Bank. WHO (2003). Global Strategy for Infant and Young Child Feeding. Fifty-fifth World Health Assembly, WHO Library Cataloguing-in-Publication Data. WHO (2004). Comparative Quantification of Health Risks. L. A. Ezzati M, Rodgers A, Murray C. Geneva, World Health Organization. Volume 1. WHO/FAO (2002). Human Vitamin and Mineral Requirements. Report of a joint FAO/WHO expert consultation Bangkok, Thailand. Rome. WHO (2010). Nutrition Landscape Information Systems (NLIS) country profile indicators; interpretation guide. WHO library Cataloguing-in-Publication Data. Geneva Wilde, P. E. and J. N. Peterman (2006). Individual weight change is associated with household food security status. J Nutr 136(5): 1395-400. Winichagoon P. (2002). Prevention and Control of Anemia: Thailand Experience. Journal of Nutrition(132): 862S-866S. World-Bank (2012). GDP growth (annual %). World Development Indicators. http://api.worldbank.org/datafiles/NY.GDP.MKTP.KD.ZG_Indicator_Meta Data_en_EXCEL.xls. World Health Organization, R. O. f. t. W. P. (2005). Maternal Mortality in Vietnam 2000-2001: 128. Yamada, R. T. and C. a. R. Leone (2008). Intrauterine Growth Restriction and Zinc Concentrations in Term Infants during the First Month of Life. 27: 485-491. Zeba, A. N., H. Sorgho, et al. (2008). Major reduction of malaria morbidity with combined vitamin A and zinc supplementation in young children in Burkina Faso: a randomized double blind trial. Nutr J 7: 7.  74  APPENDIX A Appendix Table A1.1 Characteristics of women survey participants (n=397) by village in Bình Phước, Vietnam Nov-Dec 2006 Village Characteristics Number of Subjects Age (y) Household Size Marital Status (% yes) Children Maternal Age at First Child (y) Child Spacing (y) Dependents (<5y) Dependents (<18y) Education level Grade 0-6 Grade 7- 9 Grade 10-12+ * Statistically Sidak correction  Tan Lap  Dong Tien  Dong Tam  Tan Hoa  Tan Hung  Thuan Loi  Tien Thanh  Tan Thanh  Tien Hung  Tan Xuan  35  43  45  41  44  42  43  36  43  25  40.6 (37.0 to 44.2) 4.6 (4.2 to 5.1)  38 (34.33 to 42.0) 5.0 (4.6 to 5.4)  38.1 (33.2 to 43.0) 4.3 (3.8 to 4.7)  36.3 (33.6 to 38.9) 5.0 (4.6 to 5.4)  42.8 (38.3 to 47.3) 5.5 (4.7 to 6.3)  39.7 (35.5 to 43.9) 5.0 (4.4 to 5.5)  37.1 (33.9 to 40.3) 4.8 (4.4 to 5.2)  34.2 (31.2 to 37.3) 4.9 (4.4 to 5.4)  42.1 (38.1 to 46.2) 4.8 (4.4 to 5.2)  34.4 (31.6 to 37.1) 5.0 (4.7 to 5.3)  94%  95%  96%  88%  95%  93%  100%  94%  77%  96%  2.9± 1.1  2.9 ± 1.1  2.6 ± 0.7  2.8 ± 0.6  3.0 ± 1.0  3.0 ± 0.9  2.8 ± 1.3  2.5 ± 0.7  3.2 ± 1.2  2.9 ± 0.9  24.2 ± 3.0  22.6 ± 4.8  22.2 ± 4.3  23.9 ± 3.5  22.9 ± 3.9  22.9 ± 2.6  23.2 ± 3.1  23.4 ± 3.0  2.4.0 ±3.6  21.9 ± 3.0  3.1 ± 2.0  3.1 ± 1.8  3.2 ± 2.0  3.0 ± 2.4  3.0 ±1.7  3.7 ± 3.5  4.5 ± 3.0  3.5 ± 2.0  2.8 ± 1.5  3.9 ±2.2  0.37 ( 0.15 to 0.59) 1.5 (1.1 to 1.8) 7.0 (6.0 to 8.0) 12  0.51 (0.31 to 0.72) 1.6 (1.3 to 2.0) 4.3 (3.3 to 5.4) 28  0.38 (0.22 to 0.54) 1.3 (1.0 to 1.7) 6.1 (4.7 to 7.4) 19  0.66 (0.41 to 0.91) 2.0 (1.6 to 2.3) 5.7 (4.7 to 6.6) 23  0.32 (0.15 to 0.49) 1.5 (1.1 to 1.8) 7.0 (5.8 to 8.2) 18  0.50 (0.30 to 0.70) 1.4 (1.0 to 1.8) 5.3 (3.8 to 6.8) 21  0.74 (0.53 to 0.96) 1.9 (1.5 to 2.2) 7.1 (6.2 to 7.9) 19  0.78 (0.55 to 1.01) 1.8 (1.5 to 2.0) 7.3 (6.0 to 8.7) 11  0.44 (0.27 to 0.61) 1.3 (1.0 to 1.6) 6.8 (5.7 to 7.8) 14  0.68 (0.42 to 0.94) 2.0 (1.5 to 2.5) 5.9 (5.0 to 6.7) 19  18  12  14  11  18  11  19  18  21  4  5  3  12  7  8  10  5  7  8  2  P<0.05  ** P<0.01  P value  0.056 0.054  0.006 0.015 0.003  Appendix Table A1.1 Characteristics of women survey participants (n=397) by village in Bình Phước, Vietnam Nov-Dec 2006 Characteristics  Number of Subjects  Village Tan Lap  Dong Tien  Dong Tam  Tan Hoa  Tan Hung  Thuan Loi  Tien Thanh  35  43  45  41  44  42  43  6  10  5  14  4  4  17  29  26  19  18  7  0  1  3  5  3  9  0  1  11  P value  Tien Hung  Tan Xuan  36  43  25  8  7  12  9  17  23  20  13  7  2  2  3  3  6  3  2  15  11  7  4  11  5  4  2  5  8  2  1  1  1  0  6  1  4  7  10  4  2  2  10  0  6  1  4  7  10  4  2  2  45.3 (43.0 to 47.6) 151.9 (150.0 to 153.8)  46.2 (43.7 to 48.6) 151.3 (149.7 to 153.0)  47.0 (44.7 to 49.3) 150.8 (149.5 to 152.2)  45.9 (43.7 to 48.0) 150.2 (148.5 to 151.9)  48.1 (46.3 to 49.9) 151.6 (150.2 to 152.9)  47.3 (44.5 to 50.1) 152.0 (150.5 to 153.5)  48.2 (45.9 to 50.4) 152.6 (151.1 to 154.2)  48.2 (46.0 to 50.5) 152.9 (151.2 to 154.6)  48.1 (45.9 to 50.3) 153.2 (151.7 to 154.6)  46.3 (44.0 to 48.6) 151.3 (149.6 to 153.1)  Tan Thanh  Occupation Housewife Laborer Agriculturalist Small business Professional Supplement use Child supplement use Weight (kg) Height (cm)  76  Characteristics  Village Tien Hung  Tan Xuan  36  43  25  20.9 (19.9 to 22.0) 6  20.6 (19.5 to 21.6) 2  20.5 (19.2 to 21.7) 2  19.9 (19.0 to 20.8) 0  25  22  28  30  22  7  13  15  6  11  3  4  5  10  12  4  7  3  1  0  1  2  2  1  3  0  2  2  3  1  1  1  1  1  0  6.8 (4.5 to 9.1)  11.9 (10 to 13.9)  8.5 (6.1 to 10.9)  9.5 (7.6 to 11.4)  7.6 (5.1 to 10.1)  8.7 (6.0 to 11.3)  3.3 (1.8 to 4.8)  3.7 (1.8 to 5.6)  6.3 (4.1 to 8.6)  2.1 (0.9 to 3.3)  Secure  7  2  1  2  4  3  12  9  7  8  Mild  7  1  6  3  8  7  12  12  9  12  7  1  6  5  8  9  8  3  10  4  14  39  32  31  24  23  11  12  17  1  101.3 (87.4 to 115.2)  110.1 (96.3 to 123.8)  89.6 (78.2 to 101.0)  118.3 (107.7 to 128.9)  107.9 (95.0 to 120.8)  104.8 (92.6 to 117.0)  102.5 (88.5 to 116.4)  Number of Subjects BMI (kg/m2) Overweight Normal 25-18.5 Underweight <18.5 Mild18.5-17 Moderate 16.9-16 Severe <16 HFIAS scores: (0-27)  Tan Lap  Dong Tien  Dong Tam  Tan Hoa  Tan Hung  Thuan Loi  Tien Thanh  35  43  45  41  44  42  43  20.0 (18.9 to 21.2) 1  20.4 (19.3 to 21.6) 3  20.9 (19.9 to 22.0) 2  20.3 (19.3 to 21.3) 0  21.4 (20.4 to 22.3) 2  21.2 (19.6 to 22.9) 4  20  27  35  34  35  14  13  8  7  7  4  5  5  7  2  Tan Thanh  0.001 **  Food Insecurity  Moderate Severe  Food Consumption 83.4 (74.3 to 72.8 (61.2 to 9.1.9 (80.8 to Score 92.5) 84.4) 103.1) * Statistically significant (P <0.05), ** (P<0.01), Include Sidak correction  77  0.001 ***  Appendix A1.2 Logistic Regression between underweight (BMI<18.5kg/m2) and non-underweight women (n=251) and associated characteristics of non-pregnant women (n=397) in Bình Phước, Vietnam Nov-Dec 2006 Characteristic  N  OR  95% CI  P value  Village  251  0.91  0.76 to 1.10  0.333  Age (y) Household size  251 251  1.01 1.04  0.90 to 1.13 0.69 to 1.55  0.92 0.86  Currently working  251  0.00  0.00  1.00  Vocation Small business  251 42  *Ref  Agricultural worker  27  1.28  0.22 to 7.64  0.79  Housewife  38  0.00  0.00  1.00  Labourer  134  2.79  0.79 to 9.85  0.11  Professional  13  1.18  0.12 to 11.88  0.89  Education level 0-6grade  251 127  7-9grade  98  0.73  0.29 to 1.87  0.52  >High school  1.51  0.36 to 6.30  0.57  Children  29 251  0.80  0.32 to 2.03  0.64  Number of dependents (<5y)  251  Demographic characteristics  0.51  0.53  0  147  0.08  1  73  0.18  0.041 to 0.82  0.03  2 Number of dependents (<18y)  34  0.49  0.13 to 1.84  0.29  251  0.85  0.43 to 1.67  0.63  Mean child spacing >2y  251 82  0.35  2 to 3y  54  2.47  0.73 to 8.35  0.15  >3y  1.71  0.433 to 6.75  0.44  Maternal age at first Child (y)  118 251  1.08  0.93 to 1.25  0.31  Food Security characteristics HFIAS score  251  1.08  0.92 to 1.26  0.34  Household Food Insecurity Access Scale  251 2.34  0.30 to 18.28  0.42  Mildly insecure  34 51  0.48  0.06 to 3.66  0.48  Moderately insecure  39  3.81  0.60 to 24.09  0.15  Severely insecure  130  Food secure  0.04  Characteristic  N  OR  95% CI  P value  251  0.92  0.21 to 4.05  0.92  Low  197  0.96  0.83 to 555.88  0.96  Moderate  51 6  0.83  0.59 to 0.01  67.80  251  1.00  0.97 to 1.037  0.93  251  1.00  0.97 to 1.04  0.93  90  0.74  0.06 to 9.23  0.82  Moderate diversity  79  2.21  0.43 to 11.41  0.34  High diversity  85  Food Security characteristics Household Hunger Scale  High Food Consumption Score (FCS) FCS (adjusted) Low diversity  0.84  0.13  FCS categories(adjusted) Adequate  147  0.53  0.00 to 77.20  0.80  Borderline  10  0.79  0.01 to 109.31  0.93  Poor  2  0.92  Nutrition Knowledge characteristics 251 99  0.07  0.01 to 0.84  Moderate  67  0.53  0.13 to 2.07  High  88  Nutrition knowledge score low  Dietary Behaviour characteristics Supplement use  0.04 0.36 0.06  251  0.68  0.19 to 2.41  0.68  Yes  134  1.31  0.15 to 11.31  0.81  No  117 2.15  0.15 to 30.21  0.57  Restriction of protein rich food items during pregnancy  Third trimester food avoidance Yes  23  No  228  Avoided fish during pregnancy Yes  123  No  128  0.28  0.02 to 3.83  0.34  No  100  4.07  0.16 to 104.44  0.40  1  109  1.55  0.29 to 8.135  0.61  2  2  0.00  0.00  1.00  Cognitive food, or food item, avoidance by trimester  3  7  0.00  0.00  1.00  2 to 3  1  0.00  0.00  1.00  1 to 3  15  0.98  79  Characteristic Dietary Behaviour characteristics Culturally related post-partum food item consumption (pork feet, sticky rice) Yes No  N  OR  95% CI  237 17  P value  0.29 0.80  0.17 to 3.75  0.78  0.63  0.12 to 3.43  0.59  Cognitive caloric, or food item, restraint during pregnancy Yes  157  No  94  Cognitive caloric, or food item, increase during pregnancy Yes  157  No  68  0.80  0.04 to 15.35  No  73  1.59  0.27 to 9.53  2  9  0.97  0.09 to 10.12  3  8  0.00  0.00  1 to 2  114  0.90  1 to 3  47  0.88  Cognitive caloric, or food item, increase consumption by trimester  0.26 to 3.07  0.61 0.98 1.00 0.87 0.98  Rice considered primary weaning food Yes  128  No  123  1.40  0.58 to 3.32  0.45  80  Table A1.3 Univariate Linear Regression of HFIAS scores on determinant factors among nonpregnant women indicating dietary preference during pregnancy (n=360) surveyed in Bình Phước, Vietnam Predictor variable  n  Regression coefficient  95% CI  P value  360  0.04  -0.01 to 0.08  0.10  360  -0.07  -0.18 to 0.05  0.25  Housewife  70  -0.03  -1.02 to 0.96  0.95  Labourer  176  -0.41  -1.26 to 0.44  0.35  Agriculturalist  28  -0.09  -1.42 to 1.24  0.89  Professional  19  -1.71  -3.22 to -0.20  0.03  Small business  67  *Ref  Completed Grade 0-6  171  1.15  0.12 to 2.17  0.03  Completed Grade 7-9  138  0.77  -0.22 to 1.75  0.13  Completed Grade 10-12+  51  *Ref  Number of children  100  0.34  -0.82 to 1.50  0.57  0-1 children  210  0.62  -0.30 to 1.55  0.18  2-3 children  50  *Ref  187  -0.19  -1.43 to 1.05  0.76  138  -0.36  -1.52 to 0.81  0.55  35  *Ref  34  0.94  -0.67 to 2.56  0.25  39  2.09  0.42 to 3.75  0.01  40  0.97  -0.61 to 2.56  0.23  35  2.25  0.61 to 3.90  0.01  41  1.01  -0.56 to 2.58  0.21  34  1.14  -0.45 to 2.73  0.16  43  -0.06  -1.53 to 1.42  0.94  33  0.36  -1.18 to 1.90  0.65  37  0.88  -0.66 to 2.41  0.26  24  *Ref  Demographic characteristics Age (y) BMI (kg/m2) Vocation  Education level  >4 children Dependents (<5y) 0 1 >2 Village Tan Lap Dong Tien Dong Tam Tan Hoa Tan Hung Thuan Loi Tien Thanh Tan Thanh Tien Hung Tan Xuan  81  Predictor variable  n  Regression coefficient  95% CI  P value  51  -6.92  -7.96 to -5.89  0.00  71  -4.93  -5.86 to -4.00  0.00  57  -2.28  -3.24 to -1.32  0.00  181  *Ref  6  6.35  3.90 to 8.80  0.00  71  *Ref  283  -5.16  -6.06 to -4.26  0.00  360  -0.02  -0.04 to 0.01  0.19  -1.02  -2.84 to 0.81  0.27  -0.87  -2.05 to 0.31  0.15  Food Security characteristics HFIAS categories Secure Mild Moderate Severe Household Hunger Scale High Moderate Low FCS score (adjusted)  Adjusted Relative Food Consumption Score 116 High dietary diversity 121 Moderate dietary diversity 123 Low dietary diversity Nutrition Knowledge characteristics  *Ref  Nutrition knowledge score low Moderate High  131  -0.25  -1.03 to 0.52  0.52  94  0.31  -0.46 to 1.08  0.43  135  *Ref  184  *Ref  176  -3.01  -8.72 to 2.70  0.30  -2.01 to 9.54  0.20  -1.44 to 0.84  0.60  Dietary Behaviour characteristics Avoided meat during pregnancy Yes No  Restricted protein rich foods (>12%/g) during pregnancy 185 *Ref Yes 175 3.76 No Cognitive food, or food item, restraint during pregnancy Yes No  215  *Ref  144  -0.30  82  Predictor variable  n  Regression coefficient  Dietary Behaviour characteristics Cognitive food, or food item, increase during pregnancy 265 *Ref Yes 95 0.97 No Rice considered primary weaning food Yes  188 172  No  *Ref 0.64  95% CI  P value  0.24 to 1.70  0.01  -0.015 to 1.30  0.06  * Maternal age at first child and mean child spacing non-significant  Table A1.4 Multi-nomial regression coefficient of mild Household Food Insecurity Access Scale (HFIAS) categories versus severe on determinant factors of age, household size, number of children, education level, vocation, adjusted Food Consumption Score and BMI of women (n = 397) surveyed in Bình Phước, Vietnam Predictor variable BMI (kg/m2)  n 397  OR 1.13  95% CI 1.02 to 1.25  P value 0.02  0.24  0.11 to 0.53  <0.01  1.32  0.71 to 2.48  0.38  0.35  0.17 to 0.70  <0.01  1.16  0.60 to 2.24  0.65  Nutrition knowledge score tertile Low score  134  Moderate score  131  High score  132  *Ref  Modified Food Consumption Score tertile Low dietary diversity  143  Moderate dietary diversity  108  High dietary diversity  146  *Ref  *Ref= reference  83  Appendix A1.5 Table Household Food Insecurity Access Scale (HFIAS) score by characteristics of women participants (n = 397) in Bình Phước, Vietnam Nov-Dec 2006 Characteristic  n  Median (P25, P75)  Interquartile Range  72  4  9  Agricultural Worker  30  4  8  Housewife  79  5  11  Labourer  189  7  9  Professional  25  3  8  184  7.5  10  3  8  3  7 10  P value  Demographic characteristics Vocation Small Business  0.10  Education Level 0-6grade 7-9 grade >High School  <0.01  146 67  Number of dependents (<5y)  <0.05  0  222  6.5  1  140  4  8  35  4  10  0  0  Mild Insecurity  55 77  2  2  Moderately Insecure  61  4  4  Severely Insecure  204  10  7  Low  313  3  7  Moderate  76 8  14  6  21  9  134  9  10  Moderate diversity  131  3  7  High diversity  132  4  8  Adequate  381  5  9  Borderline`  14  13.5  10  Poor  2  16  2  Food Security characteristics Household Food Insecurity Access Scale Food Secure  <0.01  Household Hunger Scale  High FCS (adjusted) Low diversity  <0.01  <0.01  FCS Categories(adjusted) <0.01  84  Characteristic  n  Median (P25, P75)  Interquartile Range  143  8  10  Moderate  108  4  9  High  146  4  8  6  10  P value  Food Security characteristics Nutrition Knowledge characteristics Nutrition knowledge score low  <0.01  Dietary Behaviour characteristics Supplement use Yes  47  <0.01  No  3 350 Restriction of protein rich food items during pregnancy  6 0.64  Yes  186  6  9  No  211  5  9  Yes  36  8  11  No  330  5  9  Yes  170  6  9  No  227  5  8  Third trimester food avoidance  <0.01  Avoided fish during pregnancy  0.49  Cognitive food avoidance by trimester  0.34  No  185  5  9  1  144  5  9  2  2  6  6  3  7  6  12  2 to 3  30  5  9  1 to 3  27  7  Culturally related post-partum food item consumption (pork feet, sticky rice) Yes  23  6  7  No  374  5  9  Cognitive food, or food item, restraint during pregnancy  0.82  0.61  Yes  217  5  9  No  146  5  10  Cognitive food, or food item, increase during pregnancy  <0.01  Yes  266  9  10  No  96  4  8  85  Characteristic  n  Median (P25, P75)  Interquartile Range  Dietary Behaviour characteristics Cognitive food increase consumption by trimester  <0.01  No  134  8  10  2  11  4  4  3  14  8  7  1 to 2  159  4  8  1 to 3  79  3  7  Rice considered a primary weaning food Yes  192  No  205  P value  <0.01 4 7  7 9  Table A1.6 Linear regression for BMI among associated factors of age and number of dependents (<5y) and among non-pregnant women (n = 397) surveyed in Bình Phước, Vietnam n  β  95% Confidence Interval  P value  Age Dependents <5y  397  0.06  0.03 to 0.09  <0.001  0 Dependents  222  1.54  0.57 to 2.51  <0.01  1 Dependent  140  1.29  0.32 to 2.26  <0.01  2 Dependent  35  Ref*  Predictor variable  Ref* = reference  Table A2.1 Nutrition knowledge: Food items commonly* considered by participants as nutritional treatment of night blindness (XN) assessed by free response questions among women (n=276) in Bình Phước, Vietnam Food Item Nutritional Treatment of Night-blindness (nresponses=74)  % (n)  Vitamin A  9.4 (26)  Fish oil  5.1 (14)  Carrot  1.8 (5)  Dark Green Leafy Vegetables  1.4 (4)  Fish  1.1 (3)  Papaya  1.1 (3)  Fish heads  0.7 (2)  Liver  0.7 (2)  Meat  0.7 (2)  Fruit  0.7 (2)  *Food items <0.5% excluded **Max 3 responses per participant  86  Table A2.2 Nutrition knowledge: non nutritional treatments for iron deficiency anemia assessed by free response questions among women (n=336) in Bình Phước, Vietnam  Non-nutritional treatment of goiter (nresponses=175)  Food Item  % (n)  Health services  28.6 (50)  TV  16.6 (29)  Iron Tablets  7.4 (13)  Radio  7.4 (13)  Books  5.7 (10)  Doctors  5.7 (10)  Newspapers  5.7 (10)  Neighbors  3.4 (6)  School  3.4 (6)  Fruit  2.3 (4)  Mother  1.7 (3)  Carrots  1.1 (2)  Eat enough  1.1 (2)  Tonic [topical]  1.1 (2)  *Responses <1% excluded **Max 3 responses per participant  87  Table A3.1 Dietary behaviour: food items commonly* fed to ill children, three responses (nresponses=567) as surveyed among (nparticipants=264) in Bình Phước, Vietnam Food Items  % (n) 68.9 (182)  Rice (noodles, porridge)  Staples  Cassavas, potatoes  3.4 (9)  Pulses  Beans, legumes, nuts  1.5 (4)  Vegetables  Green leafy vegetables  1.9 (5)  Onions, corn, tomatoes  0.4(1)  Carrots, yellow sweet potatoes  3.0(8)  Mango, papaya  0.4(1)  Fruit and vegetables  Meat  Fruits (pineapple, lychee, bananas)  29.9(79)  Fowl, poultry, beef, other Pork  30.7 (81)  Fish  14.4 (38)  Eggs  1.9 (5)  1.1 (3)  40.2 (106)  Milk (fresh, condensed)  Milk Sugar  1.1 (3)  Sugar (sugary foods, cake) Bones  19.8 (29)  Condiments *Food items <1% excluded  Table A3.2 Dietary behavior: common rational for foods items given to ill children (nparticipants=264) in Bình Phước, Vietnam Response  % (n)  1. Easy to eat and digest  40.2 (106)  2. This food is nutritious and healthy  26.9 (71)  3. Child does not want to eat  18.9 (50)  4. Child's preference  8.3 (22)  5. No response given  3.0 (8)  6. Other  2.7 (7)  88  Table A3.3 Dietary behaviour: food items commonly* avoided during pregnancy by respondents (n=266) in Bình Phước, Vietnam Food items  % (nrespondents)  Staples  Rice (noodles, porridge)  3.0 (8)  Vegetables  Green leafy vegetables  1.9 (5) 3.0 (8)  Fruit  Onions, corn, tomatoes Fruit (pineapple, lychee, bananas)  Meat  Meat (general)  38.7 (103)  Fish  63.9 (170)  Milk Oil  3.8 (10)  2.6 (7)  Milk (fresh, condensed) Cooking oil, fats, butter  4.1 (11)  Other  1.5 (4)  *Food items <1% excluded  Table A3.4 Dietary behaviour: food item avoidance during pregnancy according to trimester as reported by women (n = 217) in Bình Phước, Vietnam Variable % (n) Decreased intake  54.7 (217)  Trimester 1  92.6 (201)  Trimester 2  28.1(61)  Trimester 3  16.7 (36)  Table A3.5 Dietary behaviour: food pattern increases during pregnancy according to trimester as reported by previously pregnant, non-pregnant women cognitively eating more specific food items (n = 366) in Bình Phước, Vietnam Variable % (nresponses) Increased intake  72.7 (266)  Trimester 1  89.5 (238)  Trimester 2  93.6 (249)  Trimester 3  34.9 (93)  * <1.5% non-response  89  Table A3.6 Dietary behaviour: common reasons given by women (n=217) for food item avoidance during pregnancy in Bình Phước, Vietnam Primary reasons  % (n)  Morning sickness  73.7 (160)  Not hungry  13.3 (29)  Cannot eat  2.3 (5)  Prefer smaller birth weight baby  0.9 (2)  Other  9.6 (21)  90  Table A3.7 Dietary behaviour: Food Frequency Questionnaire with grey scale corresponding with response frequency of surveyed women (n=397) in Bình Phước, Vietnam Food items  Food groups Staples  Rice and rice products  n  Never  1-3 days/month  1 day/week  2 to 4 days/week  5 to 6 days/week  7 days/week  396  0  1  0  3  1  391  65  77  133  85  16  19  Cassavas, potatoes Pulses  Beans, legumes, nuts  395  Vegetables  Green leafy vegetables  395  5  3  7  27  7  346  carrots, yellow sweet potatoes*  397  36  71  122  117  16  35  Mango, papaya *  397  64  90  86  84  19  54  Fruits (pineapple, lychee, other)  394  18  21  56  63  23  213  Juice  396  154  54  55  77  10  46  Pork, beef†, poultry, other  393  4  20  42  117  46  164  Eggs†  397  23  47  80  134  23  90  Fish  397  11  32  44  137  33  140  Organs*‡  396  113  124  116  28  12  3  Milk  Milk (fresh, condensed)  396  270  15  20  36  10  45  Sugar  Sugar (sugary foods, cake)  396  37  15  48  36  17  243  Soft drinks  395  212  70  61  35  5  12  Oil  Cooking oil, fat  396  1  4  4  9  5  373  Miscellaneous  Fish sauce, soy sauce  397  5  4  3  4  3  378  Tea¹ or Coffee  395  271  10  12  20  5  77  Alcohol/beer  394  369  12  6  0  2  Soup broth  396  6  2  4  5  359  Fruit  Meat  5 20  *Rich beta-carotene/Vitamin A source  91  †Good iron source (10-19%DV)/100g ‡Rich iron source (<20%DV)/100g ¹Contain iron absorption inhibiting tannins 2  Source of B12  3  Source of Zinc  92  APPENDIX B DEMOGRAPHICS, FOOD CHOICES, NUTRITION KNOWLEDGE AND FOOD SECURITY OF RURAL WOMEN IN BÌNH PHƯỚC PROVINCE, VIETNAM Principal Investigator: Dr. Susan I. Barr Co-investigators: Dr. Judy A. McLean Matthew Brown (UBC student) Candice Rideout (Ph.D. candidate) Contact person: Dr. Judy A. McLean (604-228-0888 or judym@telus.net ) We are conducting a survey of rural Vietnamese women to gather information about the foods you eat, what you know about nutrition, and whether you have enough to eat. The purpose of this survey is to learn about things that could influence you and your children’s health and nutrition. This knowledge may be used to help health care workers and educators develop ways to improve the nutrition of the Hmong. You are being invited to take part in this survey as we believe you can provide us with very useful information on your family’s food habits and nutrition knowledge. The survey will take about 30 minutes to complete. At the end of the survey we would like to measure your height and weight. It is up to you whether you take part or not, and you may stop taking part at any time. Your answers are important to us but if we come to a question you would rather not answer, just let us know and we will go on to the next question. You will receive 10,000 VND (~$0.60 Canadian) for taking part in this survey whether or not you answer all the questions. All your responses are completely confidential and cannot be used to identify you in any way. Only the investigators named on this form will have direct access to your answers. If you do not understand any part of the survey, or have any questions about how this information will be used, please feel free to ask us at anytime. You will be given a copy of this form should you wish to contact the researchers or the Office of Research Services at UBC (phone number 604-822-8598). If you answer the questions it is assumed that you have agreed to take part in the study. You can decide now whether you want to answer the questions or not, or you can let us know tomorrow. Would you like to answer the questions?  93  B.1 Questions about you 1. How old are you? _____________ 2. Are you married?  ____________  3. How many people live in the same household as you? (by “household” we mean those of you who sleep under the same roof and take meals together at least 4 days a week) _____________ 4. How many years did you attend school? ____________ 5. What grade (level) did you reach in school? ___________ 6. Do you earn money for any work you do)? ______ No (go to # 7) ______ Yes (Describe your paid work: _______________________________________________) 7. Do you have any children? ______ No (go to # 8) ______ Yes (Tell us if they are girls or boys, their ages and whether they are going to school) Gender  Age  Attending School  _________  ________  _____________  _________  ________  _____________  _________  ________  _____________  (more  space if needed) 8. Are you pregnant now? ______ No (go to section C if no to #7 and #8) ______ Yes  94  B.2 Food habits and reproduction (I am going to ask about the food you ate during your most recent pregnancy.) 1a. Did you (or do you) try to eat more food, or more of some foods, during your pregnancy? ______ No (go to # 2) ______ Yes (list the foods: _______________________________________________________) 1b. When during your pregnancy did you eat more food (i.e. early, middle or latter part)? __________________________________________________________________ ____________ 1c. Why did you eat more food or more of these foods? ___________________________________ ________________________________________________________________ ______________ 2a. Did you (or do you) try to eat less food, or less of some foods, during your pregnancy? ______ No (go to # 3) ______ Yes (list the foods: ________________________________________________________) 2b. When during your pregnancy did you try to eat less food (i.e. early, middle or latter part)? ________________________________________________________________ _____________ 2c. Why did you eat less food or less of these foods? ______________________________________  95  3. What foods and/or beverages were important for you to try to eat in the first month after your baby was born? ________________________________________________________________ _____________ (now I have a few questions about feeding your children) 4. What foods and/or beverages were important to feed your baby during the first year of life? ____________________________________________________________________________ Do you give your children any supplements (vitamins and/or mineral pills)? ______ No ______ Yes (list the supplements: __________________________________________________) 5. Do you try to feed your children more food, less food or the same amount of food when they are sick? ____________________________ 7a. When your children are sick do you feed them any different foods than usual? ______ No (go to section C) ______ Yes (list the foods: ________________________________________________________) 7b. If yes, why do you feed them differently? ___________________________________________________________________ ____________  96  B.3 Food frequency We would like to know if you eat certain foods and beverages and if so, on how many days in the past month (~4 weeks) you have eaten these items. For each food we will ask if you ever ate the food in the past month. If you did it, we will ask how often: on 13 days, 1 day a week, 2-4 days a week, 5-6 days a week or every day). FOOD OR BEVERAGE  Neve r  1–3 days  1 d/wk  2-4 d/wk  5-6 d/wk  1. Milk (powdered, tinned or fresh animal milk) 2. Tea or coffee 3. Coke, Sprite, Fanta or other soft drink, number 1 4. Juice 5. Beer or other alcoholic beverages 6. Any other liquids (e.g. broth or soup) 7. Rice, noodles, maize, bread 8. Yellow sweet potatoes (yams) or carrots 9. White potatoes, cassava or white sweet potatoes (yams) 10. Any dark green leaves 11. Ripe (orange) mangoes or papayas 12. Any other fruits or vegetables (e.g. tomatoes, onions, bananas, pineapples, lychee) 13. Liver, kidney, heart, blood, intestine or other organs 14. Meat such as pork (pig), cow or water buffalo, chicken or duck (other fowl) or other animal flesh 17. Eggs 18. Fresh or dried fish or shellfish 19. Beans or foods made from beans (e.g. tofu, soy milk), other pulses or nuts (e.g. cashews or peanuts) 20. Any oil, fats, butter (including foods made with any of these) 21. Sugar or sugary foods such as sweets, candies, cakes, or biscuits 22. Any condiments (e.g. fish sauce, soy sauce, fish paste, chili sauce )  97  Every day  B.4. Health conditions, food knowledge and related habits 1a. Have you heard of anemia or iron deficiency? ________ No (go to # 2) ________ Yes 1b. If yes, do you know what foods can help to prevent or treat iron deficiency? ________ No (go to # 1c) ________ Yes (list the foods: _____________________________________________________) 1c. Describe any other ways you know of to prevent or treat anemia or iron deficiency. ________ I don’t know of any other ways (go to # 2) ________ These are the ways I know of: (describe) ______________________________________  2. Do you know of any foods that are good sources of the nutrient vitamin A? ________ No (go to # 3) ________ Yes (list the foods: ______________________________________________________)  3. Do you know of any foods that are good sources of protein? ________ No (go to # 4) ________ Yes (list the foods: ______________________________________________________)  4a. Have you ever heard of night blindness? ________ No (go to # 5) ________ Yes  98  4b. If yes, do you know what foods can help to prevent or treat night blindness? ________ No (go to # 5) ________ Yes (list the foods: _____________________________________________________)  5a. Have you heard of, or seen a goiter? ________ No (go to # 6) ________ Yes 5b. If yes, do you know what foods can help to prevent or treat a goiter? ________ No (go to # 6) ________ Yes (list the foods: __________________________________________________)  6a. Do you chew betel nuts? ________ No (go to # 7) ________ Yes 6b. If yes, how often? ________ infrequently (a few times/month) ________ occasionally (a few times/wk) ________ daily ________ several times/day  7a. Do you take any supplements (vitamins and/or mineral pills)? ________ No (go to section D.) ________ Yes 7b. If yes, what do you take and how often?  99  What  How often  _________________  ___________________  _________________  ___________________  _________________  ___________________  100  B.5 Household food security For each of the following questions, consider what has happened in the past 30 days. Consider if this happened: never (not even once), rarely (once or twice), sometimes (310 times) or often (more than 10 times)? 1. Did you worry that your household would not have enough food? Never Rarely Sometimes Often 2. Were you or any household member not able to eat the kinds of foods you preferred because of a lack of resources? Never Rarely Sometimes Often 3. Did you or any household member eat just a few kinds of food day after day due to a lack of resources? Never Rarely Sometimes Often 4. Did you or any household member eat food that you preferred not to eat because of a lack of resources to obtain other types of food? Never Rarely Sometimes Often 5. Did you or any household member eat a smaller meal than you felt you needed because there was not enough food? Never Rarely Sometimes Often 6. Did you or any household member eat fewer meals in a day because there was not enough food? Never Rarely Sometimes Often 7. Was there ever no food at all in your household because there were not resources to get more? Never Rarely Sometimes Often 8. Did you or any household member go to sleep at night hungry because there was not enough food? Never Rarely Sometimes Often 9. Did you or any household member go a whole day without eating anything because there was not enough food? Never Rarely Sometimes Often E. Measurements Weight: _________________ kg  Height: _________________ cm  Thank you very much for your participation in this survey!  101  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0073470/manifest

Comment

Related Items