Multidimensional Maternal Perinatal Stress: Integrating Variable- and Person-Centered Approaches to Improve Risk Identification

This is the authors' abstract. We don't add key points for this paper.

Purpose Maternal perinatal stress is widely studied but remains conceptually fragmented and inconsistently operationalized, limiting its clinical utility for screening and early identification in perinatal care. This study aimed to integrate variable-centered and person-centered approaches to examine the structure and heterogeneity of maternal psychosocial stress across pregnancy and early postpartum. Methods A longitudinal study was conducted with 198 mothers recruited from public primary health care centers. Psychological assessments were performed during the second trimester, third trimester, and three months postpartum. Measures included anxiety, depressive symptoms, perceived stress, and pregnancy-related distress. Principal component analyses were used to derive composite stress indices, and latent profile analyses were conducted to identify subgroups of mothers with distinct stress patterns. Results Across all time points, psychological indicators converged into a single latent dimension explaining between 72% and 85% of the variance. Latent profile analyses identified three distinct maternal stress profiles characterized by low, moderate, and high levels of perinatal stress. Profiles differed significantly across all psychological indicators but showed no differences in sociodemographic variables. Conclusions Maternal perinatal stress can be conceptualized as a coherent multidimensional construct with meaningful interindividual variability. Integrating variable-centered and person-centered approaches may enhance the identification of women at risk and inform the development of more comprehensive screening strategies in perinatal mental health care.