Sudden infant death syndrome in Scandinavia. An epidemiological study

Abstract

In the light of the increasing incidence of sudden infant death syndrome (SIDS) in the 1970s and 1980s, the first Nordic case-control study of SIDS was initiated in 1990. The main question this study set out to answer was to what extent internationally shown risk factors, such as prone sleeping position and smoking, operate in the Scandinavian population. This was especially interesting in Sweden, where the incidence of SIDS was already very low in spite of prone being the recommended sleeping position. In this study, special focus is on the effects of smoking, alcohol and caffeine on SIDS. The parallel introduction of the advice to reduce the risk of SIDS by promoting the non-prone sleeping position, non-smoking and breastfeeding offered an opportunity to study the effects of this on the incidence and on known risk factors. Using the control material, a description of normal child care practices in Scandinavia in the early 1990s has been obtained.The thesis is mainly based on the material from the Nordic Epidemiological SIDS Study (NESS), consisting of 244 cases of sudden infant death and 869 controls, matched for age, sex and maternity ward, recruited between 1 September 1992 and 31 August 1995. As a measure of relative risk, odds ratios have been calculated with conditional logistic regression. To evaluate the changes in the epidemiology, a material from the Medical Birth Registry of Sweden with 1724 SIDS cases and 24000 controls from the years 1973 to 1996 has been analysed with unconditional logistic regression. This method has also been utilised to quantify the different child care practices in the Scandinavian countries. To assess seasonal variation, the fitting of a cosine curve has been applied.The study shows that the decrease in incidence of sudden infant death has paralleled a decrease of the prevalence of prone sleeping position, and that the odds ratios for prone sleeping, smoking and bottle feeding have increased during the study period (paper I). The material from the Medical Birth Registry of Sweden in addition shows that the odds ratios for low birth weight, low gestational age and heavy smoking in early pregnancy have increased in the period after the risk reducing campaign. There is a significant seasonal variation in periods of high incidence, but not in periods of low incidence (paper II). Maternal smoking, especially smoking during pregnancy, is an independent risk factor for SIDS (paper III). Alcohol during pregnancy was not shown to be a risk factor, but high intake postnatally increased the risk (paper IV). We could not demonstrate a significant effect of caffeine during pregnancy after adjustment for smoking habits (paper IV). The analysis of the con-trols has provided valuable data on normal child-care practices, and shown that living conditions of infants in the Scandinavian countries in the early 1990s are, with small exceptions, similar (paper V).To further decrease the incidence of SIDS in Scandinavia, continual information to parents about the risks of prone and side sleeping position and smoking, and the benefits of breast-feeding, is essential. The greatest gains are expected from smoking discontinuation. We see no evidence for advice against moderate intake of caffeinated beverages. The increasing odds ratios for low birth weight and low gestational age emphasises the importance of high quality maternal and neonatal care.