Heterogeneity in the association between a dietary pattern high in fat, sugar, and sodium and adverse pregnancy outcomes by maternal characteristics: a US pregnancy cohort study.

Background: "Precision nutrition" aims to recognize variation in response to dietary patterns to inform tailored advice based on behavioral, social, environmental, genetic, and metabolic factors.

Objective: We sought to identify characteristics of pregnant individuals that modify the associations between a high fat, sugar, and sodium diet and poor perinatal outcomes.

Design: We used data from 8054 participants in the Nulliparous Pregnancy Outcomes Study: monitoring mothers-to-be (8 US medical centers, 2010‒2013), a prospective cohort study. Usual periconceptional dietary intake was assessed at 6‒13 weeks gestation using a food frequency questionnaire. The exposure was a high fat, sugar, and sodium dietary pattern versus all other diet patterns. The outcome was a composite of one or more perinatal outcomes: preeclampsia, gestational diabetes, preterm birth, or small-for-gestational-age birth. We used the doubly-robust learner, which enables the use of machine learning to identify maternal characteristics that modify the effect of the dietary pattern on the composite outcome.

Results: Approximately 29% had a dietary pattern that was high in fat, sugar, and sodium. One quarter had any adverse pregnancy outcome. The confounder-adjusted association between a high fat, sugar, and sodium dietary pattern and the risk of the adverse composite pregnancy outcome was stronger among certain subgroups of the cohort than others, including individuals with a higher BMI, lower socioeconomic status, and non-Hispanic Black race/ethnicity. For instance, compared with other diet patterns, intake of a diet high in fat, sugar, and sodium was associated with 5.9 excess cases per 100 pregnancies (adjusted risk difference 0.059 (95% confidence interval 0.012, 0.11) among individuals living in a high-poverty neighborhood, but 2.3 excess cases per 100 pregnancies (0.023; 95% CI -0.011, 0.057) among those residing in a low-poverty neighborhood.

Conclusions: This work may provide clues that contribute to a deeper understanding of the heterogeneity in dietary responses in pregnancy.