Comparison of risk factors for preterm births in two different social contexts in Northeast and South regions, Brazil

INTRODUCTION:  plausible and significant associations with social disparities have been identified as part of multifactor chain of preterm (PT) birth and there are differences for each population.

METHODS: To compare socioeconomic effects on preterm births in two different social contexts in Brazil, case-control population-based studies conducted in Londrina (LD), South region (2006-2007) and Campina Grande (CG), Northeast (2008-2009). It was used a conceptual framework via structural equations modeling. The latent variables involved and the model generated were built and validated with the weighted least square estimators in the software MPlus.

The latent variables generated represent socioeconomic vulnerability (SEV), family structure pattern (FSP) and non-acceptance of pregnancy (NAP). Prenatal care was classified into: adequate; intermediate and inadequate. Gestational age (GA) was used as a continuous variable.

RESULTS: Latent variable SEV comprises the number of residents/room (std.coef. LD:0.66, CG:0.70; p<0.001), income per capita (LD:-0.87;<0.001), maternal education (LD:0.84; CG:-0.74;<0.001), education of the household head (LD:0.74, CG:-0.81;<0.001) mother's (CG:0.55;<0.001) and household head occupation (CG:0.62;<0.001). FSP was formed by family type (LD:0.96, CG:0.96;<0.001), mother living with a partner <2 years (LD:0.69, CG:-0.60;<0.001), head household (LD:0.80, CG:0.95;<0.001). NAP considers the reactions of the mother to pregnancy (LD:0.77, CG:0.95;<0.001), father (LD:0.81, CG:0.72; <0.001) and family (LD:0.82, CG:0.76;<0.001). The effects on GA included a direct effect of inadequate prenatal care (LD:-0.21, CG:-0.27;<0.001). It was found a direct effect of NAP (0.19;<0.01) and FSP (-0.14;0.04) on PT only in Londrina. The effect of social vulnerability (SEV) was mediated through prenatal care in both cities (LD: 0.18, CG: 0.24;0,001).

CONCLUSIONS:  Despite of the lower income in CG than in Londrina, the effect of SEV on PT was similar in both cities, contributing to understand the meaning of SEV in the prematurity pathway. In CG the SEV was more relevant for PT than psychosocial factors.