Estresse materno e duração da amamentação: estudo de coorte-BRISA

Abstract

Some research shows that coping with stressful situations during pregnancy decreases the breastfeeding time, however this association has been little studied in the world. This study had two objectives: to analyze the validity of the Perceived Stress Scale in a sample of 1447 pregnant women in Sao Luis, MA and 1392 in Ribeirão Preto-SP; and analyze the effect of stress during pregnancy on the duration of breastfeeding in 1134 mother pairs / son of St. Louis. In the first article we evaluated the PSS versions with 10 and 14 items through Confirmatory Factor Analysis using the mplus 7.0, and a new version was created (PSS12) from the deletion of two items that had low factor loadings (<0.40). The models were estimated by the method of least squares (WLSMV) adjusted for mean and variance. Models with two factors showed good fit (RMSEA <0.08; CFI / TLI> 0.90) and high internal consistency (α> 0.80) in the three tested versions. In the second article we used the perceived stress scale (PSS14) and Reporting of stressful life events (EVE) to assess stress. PSS14 evaluates perceived stress during the last month. The EVE assessed stressful life events in the last twelve months. The theoretical model was designed in Acyclic Directed Graph (DAG). Survival curves were estimated using the Kaplan-Meier method. We used model of Cox proportional hazards, estimating the hazard ratio (HR) and 95%. The perceived stress during pregnancy was associated with a shorter duration of total breastfeeding (AMT) (HR=1,54; IC95%:1,23-1,93), but had no effect on exclusive breastfeeding time (AME). It was found that the PSS was adequate to assess stress in pregnant women in two cities in different regions. It suggests the use of PSS10 in the evaluation of perceived stress in pregnant women, considering their practical use, because it is a short range and its excellent fit indices. It was also found that the PEACE proved to be a tool to better identify stress than EVE.