Women who have a baby who dies from cot death have an increased risk of suffering complications in other pregnancies, scientists have found.
Putting a baby to sleep on its back reduces the risk of cot death
The Cambridge University researchers say these complications - prematurity and low birth weight - themselves increase the risk of cot death.
The scientists looked at data on 258,000 women and their babies.
They say the findings, published in the Lancet, offer a partial explanation as to why recurrent cot deaths occur.
The assessment of the risk of more than one cot death in a family has caused controversy.
Professor Sir Roy Meadow said during the trial of solicitor Sally Clark, who was convicted and then cleared of killing two of her babies, that the probability of two natural unexplained cot deaths in the family was 73 million to one.
But subsequent research has suggested women who have one cot death have a five-times higher risk that a subsequent baby will die from the same condition, compared to the general population.
The Cambridge University team analysed data for the women, who had had two consecutive single-baby births between 1985 and 2001.
Women who had an infant die from cot death - also known as Sudden Infant Death Syndrome (SIDS) - were two to three times more likely to go on to have premature babies and to have a baby who had not developed properly.
The researchers, led by Professor Gordon Smith, say the reason for the link is that women whose baby dies of SIDS are more likely to smoke, to be unmarried, to be very young and to live in a deprived area - all factors which contribute to an increased risk of future pregnancy complications.
In addition, babies who die from SIDS are more likely to be premature and have poor growth.
The reason why poorly grown or premature babies have an increased risk of SIDS is not known, but there is a hypothesis that three factors interact to cause the condition.
One of them is that developing in a poor environment in the womb leads to babies being less healthy - leading them to be born early or developing less well.
The wrong sleep position and exposure to cigarette smoke also appear to play a part.
The incidence of cot deaths also appear to peak at around three months of age.
'Body of evidence'
Professor Smith said: "The primary significance of the study is that it provides at least a partial explanation for why women who have a baby who dies due to SIDS are at increased risk of SIDS deaths in the future."
But he said it was likely that other factors would be found to explain the occurrence of recurrent cot deaths, such as genetic disposition.
Dr Richard Wilson, a trustee for the Foundation for the Study of Infant Deaths (FSID), which funded Professor Smith's research, said: "This important research increases our understanding of risk factors for sudden infant death within families.
"However, parents should not be overly alarmed by these findings, since the risk of sudden infant death, even with such predisposing factors, is still very small."
Parents are advised to take a series of steps to reduce the risk of cot death, including avoiding smoking during pregnancy and around the baby - and putting the baby to sleep on its back are known to reduce the risk.