Multigenerational Cohort Studies on Child Neurodevelopmental Outcomes

Date of Award

Spring 1-1-2024

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Public Health

First Advisor

Liew, Zeyan

Abstract

Worldwide, neurodevelopmental disorders (NDDs) are rising in incidence, with evidence supporting the role of exposures and experiences during pregnancy, perinatal, and early postnatal periods in contributing to this increase. Prior etiological research of NDDs has focused on the effects of these factors on the immediate offspring. However, accumulating evidence suggests that grandmaternal factors affecting early life development of the parent generation can subsequently influence NDD risk in the grandchildren, possibly through inheritable epigenetic alterations of the germ cells and/or disparities in other important determinants of health. As the field of multigenerational research on neurodevelopment is just emerging, this dissertation aimed to contribute additional knowledge that is currently needed. It is important first to understand the current landscape and extent of multigenerational research on neurodevelopment across generations. The limited availability of prospectively collected data spanning several generations has left many factors related to grandmothers during pregnancy and delivery unexplored. This dissertation examined two such factors and compared them with those of the parent generation. Finally, this dissertation addressed one source of selection bias, a potential methodological challenge facing this field, given its unique three-generation data structure. In the first chapter of my dissertation, Chapter I, I reviewed the current studies and methods of multigenerational research on neurodevelopment across generations by conducting a scoping review. This review set the foundation for the analyses presented in the subsequent chapters. From reviewing the existing literature, I identified two important factors—grandmaternal Cesarean section (C-section) and breastfeeding—that could influence NDD risk across generations but lacked thorough epidemiological investigation. I investigated these factors in Chapters II and III. Additionally, I found that grandmaternal effects were not always compared to maternal effects, which were addressed in these two chapters. I also identified an important methodological challenge: selection bias, which had not yet received much research attention in this field. One particular source of selection bias, due to preterm-born second-generation individuals not having children, was assessed using a case study on autism spectrum disorder (ASD) in the last chapter of this dissertation. In Chapter II, I investigated the impact of one grandmaternal factor, Cesarean section (C-section) delivery, using data from grandmaternal-maternal-child pairs enrolled in the Nurses' Health Study II (NHS II). Grandmaternal C-section is particularly important as global C-section incidence has far exceeded recommended levels. Recent research suggests a plausible biological pathway through the gut microbiota, heavily influenced by the mode of delivery, that could potentially pass risks across multiple generations. This chapter examined the associations between grandmaternal C-sections and two neurodevelopmental outcomes, ASD and attention-deficit/hyperactivity disorder (ADHD), in their grandchildren. I also compared the independent and joint effects of grandmaternal and maternal C-sections. Given the mixed results and potential controversies surrounding confounding by indication in the effects of C-sections, I separately assessed and compared the effects of maternal C-sections to results from previous studies, which formed the initial section of Chapter II. In this chapter, I found that within this cohort of nurses, their children, and their mothers, maternal C-section was not associated with offspring ASD and ADHD. However, grandmaternal C-section was associated with more than a two-fold higher risk for ASD and a 40% elevated risk for ADHD in grandchildren, after controlling for confounding factors. There was also some suggestive evidence that these associations were stronger when only the grandmother had C-sections, but further replication is needed. In Chapter III, I explored another important factor that has not yet been assessed in the multigenerational context: grandmaternal breastfeeding, one of the most important postnatal exposure factors influencing growth and development. Recent studies suggesting its potential role in epigenome remodeling through factors including nutrition, enhanced infant attachment, as well as gut microbiota. I assessed the association between grandmothers' breastfeeding practices during their daughter’s infancy and the risk of ASD and ADHD in their grandchildren, utilizing data from the NHS II. I also compared these with maternal effects to differentiate the contributions of grandmaternal factors from those of maternal ones, while controlling for potential socioeconomic confounders. No evidence of an association between grandmaternal breastfeeding during their nurse daughters’ infancy and the risk of ASD or ADHD in their grandchildren was found, regardless of the nurse mothers' breastfeeding status. In the final chapter, Chapter IV, I discussed one methodological challenge facing this field: selection bias, which, through my scoping review, I found had not received much research attention. Using data from the Danish Medical Birth Registry, I assessed and adjusted for selection bias in a case study involving three generations to investigate its influence in multigenerational studies of offspring ASD. Specifically, I examined whether mothers (second generation) who were delivered preterm are less likely to have children (third generation) and whether selection bias would arise from excluding those who have not reproduced during the study period due to their preterm birth status. I found that preterm-born women were less likely to reproduce, reducing their likelihood of being included in a population-based multigenerational cohort study with a follow-up through early and mid-adulthood. After accounting for the selection predicted by the women’s preterm birth and other sociodemographic factors, the association between maternal preterm birth and offspring ASD was strengthened. Effect modification by some of these sociodemographic factors was also observed, warranting additional investigations. These projects collectively contribute to the literature on multigenerational associations in neurodevelopment. Chapter I provided a thorough discussion of existing studies, knowledge gaps, and methodological issues in multigenerational neurodevelopment research. Chapters II and III presented new findings concerning two grandmaternal factors not previously evaluated in relation to ASD and ADHD in grandchildren and compared these associations to maternal factors. Chapter IV presented a quantitative assessment of a type of selection bias that could affect all multigenerational studies. These projects highlight topics for future research in this field and could have important public health implications regarding maternal and child health.

This document is currently not available here.

Share

COinS