Advancing understanding of perinatal binge eating attitudes and behaviors: A longitudinal investigation of the role of ovarian hormone and executive function changes across pregnancy and postpartum

Binge eating attitudes/behaviors (BEs) are common, clinically significant, and change across the perinatal period, with marked elevation during pregnancy and relative attenuation postpartum. There is extensive non-perinatal research identifying biopsychosocial factors that contribute to BEs, but comparatively little is known about how perinatal-related changes in these factors relate to the course of BEs. Reliance on cross-sectional data and narrow, dichotomous measures of BEs has limited the field’s ability to determine why susceptibility to BEs seems to change across the perinatal period and obscured understanding of the multifaceted and dimensional nature of BEs within the context of the perinatal period. Longitudinal research that comprehensively and continuously assesses BEs, and that models change in ovarian hormones and executive functioning (EF) as potential key biopsychological risk processes across the perinatal period, is a crucial next step toward clarifying etiology and informing prevention and intervention efforts tailored to this vulnerable population. Accordingly, the proposed study seeks to examine associations between changes in ovarian hormones and BEs, as well as between changes in EF and BEs, across the perinatal period. The proposed study also seeks to address methodological limitations of existing perinatal BEs research, including by integrating two longitudinal studies with pre-collected data from approximately 300 participants at four-to-five timepoints across the entire perinatal period, using a more comprehensive measure of BEs, employing structural equation models (SEM) that can examine complex patterns of change in multiple variables over time and how these changes influence each other, and covarying for depressive symptoms in a sensitivity analysis. Findings will enhance understanding of biopsychological factors associated with risk for perinatal BEs, informing future research that aims to address this highly prevalent and deleterious phenomenon. Specifically, the proposed study aims to: 1) Test associations between changes in ovarian hormones and changes in perinatal BEs. Using three bivariate dual latent change score models (LCSMs), associations will be examined between the (a) effect of estradiol x progesterone interaction, (b) main effect of estradiol, and (c) main effect of progesterone on changes in BEs across the perinatal period. Consistent with prior research in non-perinatal samples (Edler et al., 2007; Klump et al., 2008), it is hypothesized that (H1a) increases in progesterone relative to estradiol during pregnancy will be significantly associated with increases in BEs, whereas increases in estradiol relative to progesterone postpartum will be significantly associated with decreases in BEs. It is also hypothesized, given that the nature of hormonal changes across the perinatal period may amplify associations between ovarian hormones and BEs, that (H1b) increases in estradiol will be significantly associated with decreases in perinatal BEs and that (H1c) increases in progesterone will be significantly associated with increases in perinatal BEs. 2) Test associations between changes in EF and changes in perinatal BEs. Using one bivariate dual LCSM, associations will be examined between changes in EF (latent variable comprised of behavioral task and/or self-report measures) and changes in BEs across the perinatal period. Based on prior research in non-perinatal (Azevedo et al., 2004; Manasse et al., 2015) and perinatal (Davies et al., 2018; Henry & Sherwin, 2012) samples, it is hypothesized that (H2) decreases in EF will be significantly associated with increases in perinatal BEs.