Sudden infant death syndrome in Australian aboriginal and non-aboriginal infants: an analytical comparison.

Abstract

Our previous research has shown that the sudden infant death syndrome (SIDS) rate for Aboriginal infants in Western Australia (WA) is markedly higher than that for non-Aboriginal infants. The aim of this study was to identify factors that may be important in explaining this disparity. A case-control study was conducted based on routinely collected data for the population of WA singleton births from 1980 to 1990 inclusive. Cases were infants born and classified as dying from SIDS in WA (Aboriginal n = 88, non-Aboriginal n = 409). Controls were infants born in WA and not classified as dying from SIDS; 2% samples of both Aboriginal and non-Aboriginal infants were included. The risk of dying from SIDS in Aboriginal infants was 3.86 times [95% confidence interval (CI) = 2.98 to 5.02] that in non-Aboriginal infants. Statistically significant univariable risk factors for SIDS in Aboriginal infants were preterm birth, low birthweight and small-for-gestational-age; for non-Aboriginal infants they included these factors as well as single marital status, young maternal age, parity of one or greater and male sex. Comparing Aboriginal with non-Aboriginal controls, most of the risk factors were more common in the Aboriginal population. Multiple logistic regression analysis indicated that Aboriginal infants were 1.43 times [95% CI = 1.04 to 1.95] more likely to die from SIDS than non-Aboriginal infants. Differences in the risk factor profile for Aboriginal and non-Aboriginal infants were sought using interaction terms. The only important differences were that the risk of SIDS in Aboriginal infants, unlike that in non-Aboriginal infants, appeared not to be strongly related to male sex or to single marital status. Thus, the results show that the disparity between the incidence of SIDS in the Aboriginal and non-Aboriginal populations can be explained largely, although not totally, by the high prevalence of routinely recorded risk factors in the Aboriginal population. A limitation of this study is that data on the postnatal risk factors of prone sleeping, maternal smoking and non-breastfeeding were unavailable. The residual excess risk for Aboriginal infants may be a result of these recognised postnatal risk factors and/or other infant care practices that are not routinely recorded in our data base, or to underlying social and economic conditions. Further study of all these potential risk factors is warranted.