Relationships between psychosocial factors during pregnancy and preterm birth in Puerto Rico

Abstract

Psychosocial stress during pregnancy has been associated with adverse pregnancy outcomes including preterm birth (PTB). This has not been studied in Puerto Rico, an area with high PTB rates. Our objective was to develop a conceptual model describing the interrelationships between measures of psychosocial stress and depression, a result of stress, among pregnant women in Puerto Rico and to examine their associations with PTB. We used data from the Puerto Rico Testsite for Exploring Contamination Threats pregnancy cohort (PROTECT, N = 1,047) to examine associations among depression and different continuous measures of psychosocial stress using path analysis. Psychosocial stress during pregnancy was assessed using validated measures of perceived stress, negative life experiences, neighborhood perceptions and social support. Logistic regression was used to examine associations between psychosocial stress measures in tertiles and PTB. Perceived stress, negative life experiences, and neighborhood perceptions influenced depression through multiple pathways. Our model indicated that perceived stress had the strongest direct effect on depression, where one standard deviation (SD) increase in perceived stress was associated with a 57% SD increase in depression. Negative life experiences were directly but also indirectly, through perceived stress, associated with depression. Finally, neighborhood perceptions directly influenced negative life experiences and perceived stress and consequently had an indirect effect on depression. Psychosocial stress was not associated with PTB across any of the measures examined. Our study examined interrelationships between multiple measures of psychosocial stress and depression among a pregnant Puerto Rican population and identified negative neighborhood perceptions as important upstream factors leading to depression. Our findings highlight the complex relationship between psychosocial stress measures and indicate that psychosocial stress and depression, assessed using 5 different scales, were not associated with PTB. Future research should investigate other environmental and behavioral risk factors contributing to higher rates of PTB in this population.

Fig 1. Path diagram indicating the relationship between psychosocial stress measures and depression in PROTECT study population.

Maternal age, marital status, and education are included as covariates in model (N = 1,047). Note: All paths are significant at p<0.05; missing data handled using full information maximum likelihood. Model fit statistics: X² = 0.71, p value = 0.40, CFI = 1.00, TLI, 1.02, RMSEA = 0.00, SRMR = 0.00. Abbreviations: X²/df, chi-square to degree of freedom index; RMSEA, Root Mean Square Error of Approximation; CFI, Comparative Fit Index; TLI, Tucker-Lewis Index; SRMR, Standardized Root Mean Square Residual.