Date of Award

January 2025

Document Type

Thesis

Degree Name

Medical Doctor (MD)

Department

Medicine

First Advisor

Kieran J. O'Donnell

Abstract

Children exposed to maternal symptoms of depression and anxiety are at increased risk of socioemotional and behavioral problems in childhood. However, few studies have examined how the developmental timing of exposure to maternal and paternal depression from pregnancy onwards influences objective measures of mental health in adulthood. This study sought to examine how the timing of parental depression from pregnancy to early adulthood affects risk for offspring mental illness.

In this prospective cohort study using data from the Avon Longitudinal Study of Parents and Children (ALSPAC), a British birth cohort, we utilized distributed lag models (DLM) to calculate the cumulative effect of parental depressive symptoms across time on multiple mental health outcomes, adjusted for confounders related to parental socioeconomic status and antenatal characteristics. Using DLM, we also generated a time-response curve to visualize changes in the weight of association between each longitudinal measure of parental depressive symptoms from pregnancy to young adulthood and offspring self-reported symptoms of mental illness in adulthood. We then applied the related distributed lag interaction model (DLIM) to test if offspring sex modifies the relationship between parental depression and offspring mental health outcomes. Parental depressive symptoms were assessed 10-12 times from pregnancy to 21 years using the Edinburgh Postnatal Depression Scale (EPDS). Outcomes included symptoms related to depression at 27, anxiety at 25, psychotic disorders at 24, and alcohol use disorder (AUD) at 22. Participants with mental health outcomes and parental depressive symptoms measured at least once over time were included. As a supplemental analysis, we analyzed the effects of parental anxiety symptoms, as assessed using the Crown-Crisp Experiential Index (CCEI) anxiety subscale from pregnancy to 6 years of age (representing all available measures), on the same four mental health outcomes.

Between 3,342 and 3,795 participants (2086-2506 [62.4%-66.0%] female) were included. Adults exposed to a one point increase in EPDS scores across development were more likely to experience clinically concerning symptoms of depression (maternal: adjusted odds ratio [aOR] , 2·598; 95% CI, 2·101-3·191; P < 0·001; paternal: aOR, 2·292; 95% CI, 1·722 -3·052; P < 0·001), anxiety (maternal: aOR, 2·821; 95% CI, 2·262-3·517; P < 0·001; paternal: aOR, 2·170; 95% CI, 1·627-2·894; P < 0·001), and psychosis (maternal: aOR, 2·040; 95% CI, 1·377-3·022; P < 0·001; paternal: aOR, 1·766; 95% CI, 1·034-3·018; P = 0·037). Parental depressive symptoms were not associated with problematic alcohol use. We found no evidence that offspring sex modifies the effects of parental depressive symptoms on any of the outcomes studied. Parental depressive symptoms were associated with adult symptoms of depression and anxiety for most time points. However, maternal effects emerged during pregnancy, whereas paternal effects were first detected in early childhood. Only maternal depressive symptoms during pregnancy were significantly associated with psychotic experiences.

In this study, we found that both timing and chronicity of parental depression influenced offspring adult mental health in a large, longitudinal birth cohort. The presence of maternal effects only in pregnancy is consistent with the ‘fetal origins’ hypothesis. Interventions that reduce cumulative exposure to parental depression may improve the mental health of the next generation.

Comments

This thesis is restricted to Yale network users only. This thesis is permanently embargoed from public release.

Download

Share

COinS