Abstract
OBJECTIVE: We sought to investigate social differences in Swedish infant mortality by cause of death. METHODS: All live single births in Sweden between 1983 and 1986 to mothers 15 to 44 years old with Nordic citizenship were studied. The causes of death were classified into six major groups. Mother's education was used as a social indicator. Logistic regression analysis was used with identical models for all groups of causes of death. RESULTS: There were 355,601 births and 2012 infant deaths. Only for sudden infant death syndrome were significant social differences found, with crude odds ratios of 2.6 for mothers with less than 10 years of education and of 1.9 for mothers with 10 to 11 years, compared with 1.0 for mothers with 15 years or more. After adjusting for age, parity, and smoking habits, these ratios were no longer significant. CONCLUSIONS: The social differences obtained could be explained by the fact that mothers with less education smoke more, are younger, and have higher parity than those with more education.