Preconception environments linked to pregnancy-related brain remodeling and better maternal mental health in a Singaporean cohort

This observational cohort study included 194 participants in the preconception phase and 61 participants in the postpartum phase within the Singaporean S-PRESTO preconception cohort. The population was followed from preconception to three months postpartum, with offspring assessed at age four. The study examined preconception environments, including lifestyle, sociodemographic, psychological, and social-support variables, comparing them against normative age trends for gray matter volume trajectories.

The main results indicated that pregnancy-related gray matter volume reductions occurred in the default mode, frontoparietal control, salience, and limbic networks, as well as the hippocampus and caudate. A higher preconception brain gray matter volume was associated with more advantaged socioeconomic status, healthier diet, better sleep, non-smoking, stronger social support, and higher conscientiousness or lower neuroticism. This environment-associated brain pattern related to better maternal executive function, lower depression, anxiety, stress, and more favorable metabolic health.

Additionally, the same brain resilience pattern predicted lower postpartum anxiety and fewer offspring internalizing symptoms at age four. Safety data, adverse events, and discontinuations were not reported. Tolerability was not reported. Key limitations include uncertainty regarding whether modifiable preconception environments shape individual variability in pregnancy-related brain remodeling and whether these environment-shaped brain features relate to subsequent maternal and offspring mental health.

The practice relevance implicates modifiable preconception environments in shaping brain structural resilience to pregnancy-related remodeling. This highlights preconception as a potential window to promote maternal brain health and positive intergenerational mental health outcomes. The study is observational, so association versus causation is not explicitly distinguished beyond the language used in the abstract.