Phthalate Exposure and Child Development -- The Infant Development and Environment Study

Phthalates are a class of chemicals widely used in manufacturing to increase the transparency, durability, and flexibility of plastics. They are commonly found in pharmaceutical coatings, food wrapping, personal care products, and many household items. Since phthalates are not covalently bound to their product matrix, they can easily leach into the surrounding environment and are thought to enter the human body through a number of routes: oral, dermal, inhalational, intravenous. Phthalate exposure in the modern environment has become nearly universal as their use in consumer products has become commonplace. Some studies have suggested that particular types of phthalates may function as endocrine disruptors and interfere with infant development.

To help fund my PhD thesis research, I worked part-time for The Infant Development and Environment Study (TIDES), under the direction of Dr. Shanna Swan. In this capacity, I created the code to organize subject data and conducted some preliminary analyses. During this time I also conducted an analysis to explore the socioeconomic correlates of phthalate urinary metabolite levels using the National Health and Nutrition Examination Survey.

I just finished examining some fairly interesting associations of prenatal phthalate exposure and behavior in 6-10 year old children, using Shanna’s Study for Future Families. We found associations between certain phthalate esters and inattention and problem behavior in boys. Our paper was accepted by Environmental Health Perspectives and will be publicly available soon.
Here is a clip of Shanna discussing phthalate exposure on CBS National News.

Limits of Detection

Environmental health studies often rely upon biomarkers to estimate exposures. For example, in a study of whether phthalate exposure during pregnancy was associated with timing of labor, exposure to phthalates was estimated through quantifying metabolites in maternal urine samples. Unfortunately, the laboratory techniques that quantify biomarkers often produce data that are subject to limits of detection (LOD). Chemical levels below the LOD are difficult to quanitfy. The presence of these observations complicates the statistical analysis of environmental health analysis.

While several methods for addressing LODs have been described, they have rarely been systematically compared. We are conducting a simulation study to compare the bias associated with each of these methods. Once we have results we indeed to create a simple and clear decision tree to help environmental health researchers in their work.

The Seychelles Child Development Study

The Seychelles Child Development Study is a 20+ year prospective study of Seychellois children and their exposure to methylmercury. The findings of this study have been published in The Lancet, Journal of the American Medical Association, and many other prestigious journals.

I spent three years working with highly interdisciplinary research group. I helped determined the direction of future research and conducted many different analyses relating to socioeconomic status. One such study was a life course analysis of maternal socioeconomic position in relation to young adult cognitive function in multiple domains, which was recently published. In addition, I’ve contributed to many of their analyses of child neurodevelopment.

PhD Thesis Study

Allostatic load is a developing epidemiologic concept that has been used to quantify the physiologic costs of adaptation and cumulative life stress, whether psychological or physical.

Although various forms of life stress have been linked with late-life depressive symptoms, the association between allostatic load and depressive symptoms has rarely been assessed. My thesis project entailed a cross-sectional study of two different study populations: older adults from the 2005-2008 National Health and Nutrition Examination Survey and 125 community-dwelling older adults of Rochester, NY. The objective was to examine the complex relationship of life stress, allostatic load, social factors, and psychosocial factors with the severity of late-life depressive symptoms.

Specific aims:
(1) To examine the relationship between multisystem physiologic dysfunction (renal, immune, hepatic, metabolic, and cardiovascular function) and the severity of global, cognitive/affective, and somatic depressive symptoms among a nationally-representative sample of older adults.

(2) To examine the relationship between allostatic load (neuroendocrine, immune, metabolic, and cardiovascular function) and the severity of global, cognitive/affective, and somatic depressive symptoms among a sample of Rochester older adults.

(3) To conduct path analyses to examine the complex causal web of biological, psychological, and social factors underlying late-life depressive symptoms. To situate allostatic load within this causal framework.

This study provided evidence in support of the allostatic load model of late-life depressive symptoms. In both a large, nationally representative sample and a smaller sample of community-dwelling Rochester, NY older adults, we observed associations between composite measures of multisystem bodily dysfunction with late-life depressive symptoms. In addition to demonstrating that these associations do exist, the effect sizes we observe indicate a clinically meaningful difference in depressive symptoms; this difference was comparable to the average effect size observed in clinical trials of antidepressant medication use, although occurring in the opposite direction. The full public health significance of this association remains to be elucidated, as additional research is needed to confirm these findings. However, given the prevalence of clinically significant late-life depressive symptoms observed in the samples represented here and elsewhere, as well as the debilitating nature of depressive symptoms, the relationships explored in this project appear to be of broad public health significance.


rubber ducks