Changes in Maternal Diet Quality Across Pregnancy and Postpartum and Associations with Sociodemographic Characteristics

Abstract

Objectives: This study described diet quality across the perinatal continuum and investigated associations of sociodemographic characteristics with diet quality change from pregnancy to postpartum. Methods: Pregnant women (n = 365) in a prospective cohort study completed a demographic questionnaire at recruitment and six 24-hour dietary recalls (one in each pregnancy trimester and 2, 6, and 12 months postpartum). Sample-level diet quality (Healthy Eating Index-2015, HEI total and component scores) at each of the six time points was quantified using the population ratio method and compared using bootstrapping and permutation tests. Individual-level HEI scores during the pregnancy (pooling 3 pregnancy recalls) and postpartum (pooling 3 postpartum recalls) periods were quantified using the by-person method and compared using paired t-tests. Multiple linear regression models examined independent associations of sociodemographic characteristics with change in HEI (postpartum subtracted from pregnancy score). Results: Sample-level HEI total and component scores were generally consistent among time points within pregnancy and within post-partum but showed varied changes between periods. Within-person analyses indicated both positive (total protein, fatty acids, refined grains, saturated fat) and negative (total fruit, whole fruit, dairy) changes in HEI from pregnancy to postpartum. Greater improvements in HEI total score were observed in white non-Hispanic versus black non-Hispanic participants (β ± SEM: 6.7 ± 3.2), participants with at least a college degree versus less than a college degree (6.8 ± 2.7), and participants who were not working by choice versus working full time (7.0 ± 2.6). In addition to associations with change in HEI total score, race/ethnicity was associated with change in greens and beans, seafood and plant proteins, and added sugars component scores, education was associated with change in total fruit and refined grains scores, and employment was associated with change in refined grains score. Age, smoking status, and body mass index were also associated with changes in HEI component scores. Conclusions: Perinatal decreases in fruit and dairy intake and sociodemographic disparities in dietary changes may represent critical intervention targets.