Environment-Wide Association Study on Childhood Obesity




Uche, Uloma I.


0000-0003-1887-3419 (Uche, Uloma I.)

Journal Title

Journal ISSN

Volume Title



Background: Obesity is both a global and national public health issue with increasing prevalence over the last decades. It is associated with adverse health effects as well as social and economic costs. Both children and adults are affected; however, much more impact and prevalence are seen in children because of their growing bodies. While the etiology and prevention of childhood obesity are not fully understood, studies have linked it to lack of built environment, diet, lack of physical activities, and genetic susceptibility, with growing evidence that it could also result from other environmental factors. Studies linking it to other environmental factors are quite limited, unsystematic, incomprehensive and inconclusive. Thus, using the concept of an environment-wide association study (EWAS) and while accounting for already known risk factors (lifestyle factors) associated with childhood obesity, the aims of this dissertation were 1) to comprehensively and systematically investigate all the environmental factors available in the National Health and Nutrition Examination Survey (NHANES) to determine factors associated with childhood obesity and 2) to validate my findings from aim1 on a different cohort of children and adults to see if factors persist. Methods: I utilized NHANES datasets 1999-2016, retrieving data files for children/adolescents (6-17yrs) and adults (>18yrs). Obesity was measured using BMI measures and waist to height ratio. A multinomial and binary logistic regression was conducted while adjusting for age, sex, race/ethnicity, creatinine, calorie intake, physical activity, screen time (TV hours & computer/video games hours), limitation to physical activities, and socioeconomic status. As in EWAS, multiple hypothesis testing was controlled, and validation analyses were done. Results: I found that metals such as beryllium (OR: 3.305 CI: 1.460-7.479) and platinum (OR: 1.346 CI: 1.107-1.636); vitamins such as gamma- tocopherol (OR: 8.297 CI: 5.683-12.114) and delta- tocopherol (OR: 1.841 CI:1.476-2.297); heterocyclic aromatic amines such as 2-Amino-9H-pyrido (2,3-b) indole (A-a-C) [OR: 1.323 CI: 1.083-1.617] and 2-Amino-3-methyl-9H-pyriodo[2,3-b]indole (MeA-a-C) [OR: 2.799 CI: 1.442-5.433]; polycyclic aromatic amines such as 9- fluorene (OR: 1.509 CI: 1.230-1.851), 4- phenanthrene (OR: 2.828 CI: 1.632-4.899) and caffeine metabolites such as 1,3,7-trimethyluric acid (OR: 1.22 CI: 1.029-1.414) and 1,3,7-trimethylxanthine(caffeine) (OR: 1.258 CI: 1.075-1.473) were positively and significantly associated with childhood obesity. More so, I found that factors such as gamma- and delta-tocopherols, as well as manganese, copper, caffeine, 2-napthol and 2-phenanthrene were associated with both childhood and adulthood obesity. Finally, I found that vitamin B6, B12 and C as well as carotenoids, enterolactone, harmane and iron are protective factors of both childhood and adulthood obesity. Discussion: These novel findings are of public health significance since these factors are potentially modifiable risk factors of childhood obesity and they are valuable for prevention and reducing the risk of obesity among U.S. children and adolescents. Exposures to some of these factors are mainly from vehicle exhaust, tobacco combustion, tea, and contaminated air and water. They may have the capability of eliciting stress, inhibiting enzymes needed for metabolic processes or disrupting lipid homeostasis which subsequently increases the risk of obesity. Conclusion: Despite the difficulty of ascertaining causality, this dissertation found novel pathways to the etiology of childhood obesity as well as adulthood obesity that needs further investigation.