Date of Award
1-1-2023
Language
English
Document Type
Dissertation
Degree Name
Doctor of Philosophy (PhD)
College/School/Department
Department of Epidemiology and Biostatistics
Content Description
1 online resource (xiv, 158 pages) : illustrations (some color)
Dissertation/Thesis Chair
Marilyn L Browne
Committee Members
Martha M Werler, Melissa Tracy
Keywords
birth defects, congenital malformations, pregnancy, self-report, sleep aid, zolpidem, Zolpidem, Hypnotics, Pregnant women, Abnormalities, Human
Subject Categories
Epidemiology
Abstract
This dissertation examines the validity of maternal self-reported medication exposures and then explores the associations between self-reported sleep aid use, including zolpidem specifically, and birth defects using data from three population-based case-control studies conducted within the United States. The first aim of this dissertation was to explore agreement between two sources of prescription medication use in early pregnancy: medical records and self-reported interview data from the Birth Defects Study To Evaluate Pregnancy exposureS (BD-STEPS). We observed that agreement of prescription data between self-report and medical records varies by whether the medication is used chronically or episodically. While self-reported medication information for medications used chronically were found to have substantial to near perfect agreement and high sensitivity (89%) and specificity (94%), agreement and sensitivity (51%) for medications used episodically were lower. We did not observe differences in agreement across any of the explored maternal or interview characteristics. Based on our findings, future analyses exploring self-reported medication use would benefit from estimates of its accuracy and validity. Analyses that seek to assess the impact of exposure misclassification on the observed estimates of the effects of early pregnancy medication use on the risk of birth defects or other pregnancy outcomes could rely on estimates of the specificity and sensitivity from this dissertation.Very few studies have examined the associations between sleep aids and birth defects and this dissertation helps address this gap in the literature. Using data from the National Birth Defects Prevention Study (NBDPS) and the Birth Defects Study (BDS), we calculated estimates of the association between early pregnancy zolpidem use and specific birth defects. Early pregnancy zolpidem use was rare, reported by 84 (0.2%) case and 46 (0.2%) control participants. The main analysis included 47 specific birth defects, and adjusted odds ratios (ORs) were <1.0 for cleft lip, pulmonary valve stenosis, and secundum atrial septal defect, and were greater than 1.8 for cleft palate, anorectal atresia, gastroschisis, and tetralogy of Fallot. All of the corresponding 95% confidence intervals (CIs) included the null. Three sub-analyses were conducted to explore the impact of potential biases, including additional confounding adjustment via propensity scores, comparison of early pregnancy zolpidem use to late pregnancy zolpidem use to control for factors prompting treatment with zolpidem, and an exposure misclassification bias analysis. In all of these analyses, results were largely similar in terms of direction and magnitude. The adjusted OR for the four defects with the most elevated results in the main analysis (cleft palate, anorectal atresia, gastroschisis, and tetralogy of Fallot) remained elevated and moved further from the null in the sub-analyses; however, the CIs were wide in all analyses. The third aim of this dissertation was to explore the use more generally of sleep aid medications in early pregnancy and calculate estimates of the association with birth defects using data from BD-STEPS. Given that BD-STEPS is ongoing, these results are preliminary and might change if results are updated with additional years of data prior to publication in a scientific journal. We found largely null results for the nine defects that had five or more exposed cases, which adjusted ORs ranged from 0.64 for gastroschisis to 1.46 for anotia/microtia and all of the corresponding 95% CI included the null. The results from two sub-analyses that explored confounding by indication and the impacts of exposure misclassification were of similar magnitude and in the same direction as results that adjusted only for confounding.
Recommended Citation
Howley, Meredith M., "Sleep aid use during early pregnancy and the risk of birth defects" (2023). Legacy Theses & Dissertations (2009 - 2024). 3153.
https://scholarsarchive.library.albany.edu/legacy-etd/3153
