Studies on possible sociodemographic inequities in the survival of preterm infants are scarce. Individual and neighbourhood sociodemographic factors are related to preterm birth and to infant mortality in full-term infants. The aim here was to examine whether infant mortality in Swedish preterm infants is related to individual and neighbourhood sociodemographic factors, and to study whether the hypothesised association between neighbourhood deprivation and infant mortality persists after accounting for individual sociodemographic factors. The study included 46,470 infants with a gestational length of <37 weeks, born in Sweden between 1992 and 2006. Neighbourhood deprivation was assessed by an index (education, income, unemployment, welfare assistance) in small geographical units, and categorised into low, moderate and high deprivation. Adjusted odds ratios for infant mortality were examined in relation to individual and neighbourhood sociodemographic factors. After adjusting for maternal age, infant mortality was associated with the following sociodemographic variables: maternal non-married/non-cohabiting status, low family income, low maternal education and rural status. After full adjustment, the odds ratio [95% confidence interval] was 2.98 [2.42, 3.67] for low family income compared with high family income. An increase in infant mortality was also associated with high neighbourhood deprivation; however, this increased risk no longer remained statistically significant after adjusting for individual sociodemographic factors. In conclusion, this study showed an increased infant mortality in preterm infants born to women with a less favourable sociodemographic profile.
© 2011 Blackwell Publishing Ltd.