Abstract: Despite significant reductions in child stunting over recent decades, 36% of children remain stunted in Nepal (2016). Poor linear growth can begin in utero and continue beyond the age of two years, making the first 1000 days of life a critical period for stunting prevention. Recent epidemiological studies suggest that exposure to aflatoxins could contribute to low weight at birth and pos... read moretnatal stunting. However, study findings showing linkages between in utero aflatoxin exposure and adverse birth outcomes remain inconclusive, and factors contributing to widespread exposure to aflatoxin during pregnancy are inadequately understood. This dissertation contributes to the evidence-base to inform the design of aflatoxin reduction interventions and a better understanding of the potential influence of aflatoxin exposure on adverse birth outcomes. All three studies used data from 1675 pregnant women and newborns participating in the ongoing USAID-funded Mycotoxin (AflaCohort) Birth Cohort Study in Banke, Nepal. In Study 1, we estimated pregnant women's frequency of consumption of aflatoxin-prone foods (i.e. maize and groundnuts) and calculated dietary diversity scores. Ordinary Least Squares (OLS) and Quantile Regression models were used to compare the strength of associations between frequencies of consumption of maize and groundnuts and dietary diversity, and serum aflatoxin levels (n=1648). After adjusting for wealth and other covariates, women who had consumed maize and/or groundnuts more frequently showed higher levels of aflatoxin albumin adducts. Findings indicated that dietary diversity was not predictive of aflatoxin exposure. Seasonality was a strong predictor of prenatal aflatoxin exposure, with the highest levels seen in the winter months following maize and groundnut harvest seasons. The second study examined the correlations between food handling procedures and good agricultural practices (GAPs) in maize, groundnut and chili farming households, and aflatoxin exposure as measured by aflatoxin albumin adducts during pregnancy. Multivariate OLS regression modeling revealed no evidence that the GAPs used in a minority of maize farming households (n=392) were associated with reduced exposure of pregnant women to aflatoxin in this sample. The infrequent use of recommended GAPs may have limited our ability to detect such an association. Moreover, off-farm food acquisition was common. Levels of aflatoxin exposure observed in this study likely reflect consumption of various foods susceptible to aflatoxin from multiple sources. Study 3 used linear and logistic regression models to explore the relationship between prenatal aflatoxin levels and selected adverse birth outcomes in a sub-sample of 1621 mother-newborn pairs. Twenty percent of infants were low birth weight, 52% small-for-gestational-age, 16% stunted, and 13% were born prematurely. None of the birth outcomes studied were associated with maternal aflatoxin levels, which were considerably lower than those observed in Africa and the Middle East where a relationship with low birth weight has been previously documented. Together, the results presented in this dissertation underscore the importance of viewing aflatoxin contamination as a component of food safety within complex food systems. Our study, together with the mixed results from previous studies, reiterates how incomplete the evidence of the relationship between aflatoxin and birth outcomes remains at this point. It also suggests that additional research is necessary to elucidate the aflatoxin-fetal growth relationship, including determination of threshold values.
Thesis (Ph.D.)--Tufts University, 2018.
Submitted to the Dept. of Food Policy & Applied Nutrition.
Advisor: Patrick Webb.
Committee: Shibani Ghosh, Beatrice Rogers, and Gerald Shively.
Keywords: Nutrition, Public health, and Agriculture.read less