Space time clustering of births in SIDS: do perinatal infections play a role?

The aetiology of sudden infant death syndrome (SIDS) remains uncertain; many causal pathways have been proposed. In this paper we have examined firstly the variation in the risk of SIDS with age, month of death and month of birth; and secondly the space time clustering of SIDS deaths, and, separately, space time clustering of their births. Data were obtained from the Office of Populations, Censuses and Surveys on all certified SIDS deaths in the period; children were assigned grid references for the address of birth and of death. Data on number of births were abstracted from published material. A log-linear modelling technique was used to investigate the separate effects of age, month of death and month of birth on the risk of SIDS. The Knox method was used to investigate space time clustering of deaths and of births of children who died of SIDS. Separate, statistically significant effects were found for age, month of death and month of birth. There was minor space time clustering of SIDS births and deaths at large time and space intervals, and a marked space time clustering of births in short space time intervals in the first quarter of the year. The finding of an effect of month of birth on the risk of SIDS, and of space time clustering of births suggest that a perinatal hazard--possibly of infectious origin--may play a role in the aetiology of SIDS.