Men who receive blood transfusions from women who have ever been pregnant are at risk of an early death, research suggests.
Experts suspect antibodies produced by a woman's body in pregnancy might provoke a severe immune response if injected into a man.
But scientists last night stressed that the study of 31,000 people, by experts at Leiden University in the Netherlands, does not definitively prove men are at risk from female blood donations and said NHS procedures should not be changed.
However, others said the study raises worrying concerns and warrants additional research.
Dr Ritchard Cable, of the American Red Cross Blood Services, said: These results are provocative and may if true have significant clinical implications.'
The researchers, whose work is published in the JAMA medical journal, found that men who received red blood cells donated by a woman who had ever been pregnant were 13 per cent more likely to die within a year than if given blood from another man. For every 1,000 men who received a blood transfusion from a woman who had been pregnant, 101 died per year.
But for every 1,000 men injected with blood from another man, there were 80 deaths 21 fewer.
Patients' mortality was unchanged if the donation came from a woman who had never been pregnant. And for female recipients, their chances of death remained the same regardless of donor gender.
About 400,000 people receive a transfusion of red blood cells in Britain each year, and are not told the gender of the donor.
The researchers said lung problems were the most likely cause of death. They are not sure exactly why men seemed to react so strongly to the blood of women who had been pregnant, but said an immunologic' reaction could be to blame.
They wrote: Male recipients who received a transfusion from an ever-pregnant female donor had a statistically significant increase in mortality compared with those who received a transfusion from a male donor or from a female donor without a history of pregnancy.'
But they advised caution, adding: Further research is needed to replicate these findings, determine their clinical significance, and identify the underlying mechanism.'
Professor Kevin McConway, from The Open University, said that the mortality increase in real terms was not large, adding: I don't think there is yet any real cause for men to be particularly concerned about this issue if they need a blood transplant. At least, I'm a man, and I wouldn't be concerned.'
A spokesman for NHS Blood and Transplant said: Blood donations from all our donors are life-saving, and we continue to encourage donations from women who have previously been pregnant.'