PARIS (AFP) - British researchers say they have found strains of germs that appear to play a stealthy role in the rare but mysterious disorder known as sudden infant death syndrome (SIDS). Also known as cot death, SIDS is defined as the sudden and unexpected death of an infant aged between a week and a year, but the causes are believed to be many and are fiercely debated. Writing in next Saturday's issue of The Lancet, paediatricians from Great Ormond Street Hospital for Children in London reviewed autopsies of 507 infants who had died of SIDS between 1996 and 2005. The team divided the cases into three groups. Of the 507 deaths, 379 were categorised as unexplained. Fifty-six were ascribed to a bacterial infection such as pneumonia or meningitis. And 72 deaths were attributed to non-infective causes, such congenital heart disease or an accident. The team then looked at post-mortem samples that tested the infants' blood for the presence of germs. These samples showed no difference in overall bacteria level between the three groups. What was intriguing, though, was that specific strains of germs were more present in the "unexplained" and "bacterial infection" groups than in the non-infective category. These germs, including two common microbes, Staphylococcus aureus and Escherichia coli, are part of a pathogenic group that can cause blood poisoning yet now show the focus of infection. The presence of these germs in the samples is not proof in itself that they caused the babies' deaths. But the investigators believe they could be linked at least to some SIDS cases that so far cannot be explained. At present, the risk factors for SIDS are commonly accepted to be a genetic susceptibility; a period of particular vulnerability, between two and five months; an infant's exposure to tobacco smoke; and if the baby sleeps on its tummy rather than its back. The British charity that funded the research, the Foundation for the Study of Infant Deaths (FSID), said the study raised key questions about how microbes could be indirectly responsible for an infant's death. One potential bacterial path is non-infective, "for example by colonising the pharynx, or upper airway, from where toxins produced by the bacteria may enter the bloodstream," said FSID's scientific advisor, George Haycock. Another possible avenue is that, among some infants, the immune system runs amok when exposed to certain bacteria, causing an overwhelming, fatal inflammatory response, he said.