Mental health symptom clusters and network changes over the course of pregnancy: Links to biomarkers of allostatic load

Pregnancy is marked by dynamic shifts in psychological and physiological states, yet standard diagnostic categories often overlook individual variability in symptom patterns and their biological underpinnings. The current study applied a multidimensional approach to examine mental health symptom networks, identify symptom clusters over the course of pregnancy, and link these profiles to multisystem biomarkers among women in Pakistan exempt from pre-existing physical and mental health conditions. At early to mid- (T1) and mid- to late (T2) gestation, 50 symptoms of anxiety, depression, and stress were used to estimate symptom networks. Exploratory factor analyses and hierarchical clustering analyses were conducted to uncover subgroups of mental health patterns. Cluster memberships were then mapped onto peripheral biomarkers (cortisol, C-reactive protein (CRP), glycosylated hemoglobin, total cholesterol, diastolic blood pressure and systolic blood pressure), as well as a composite allostatic load score. Somatic anxiety symptoms were the most central node across both timepoints, while positive affect was protective within the network. The network demonstrated overall strengthening at T2. Five clusters emerged at each timepoint, with substantial mental health profile shifts across gestation. Biomarker mapping revealed elevated CRP in the Flattened Affect cluster compared to Healthy and Somatic Anxiety at T1. At T2, the Depression cluster exhibited the highest CRP and allostatic load, and the Lack of Control cluster displayed the lowest. This study highlights significant heterogeneity and comorbidity of prenatal maternal mental health, and the shifts that occur across pregnancy. These data provide a personalized framework for tracking how symptom subtypes and physiological dysregulation evolves across pregnancy.