Researchers compared 609 normal pregnancies with 717 that had one or more of four adverse outcomes: premature birth, small for gestational age, bleeding during pregnancy or pre-eclampsia.

The scientists tested each baby's blood within five days of birth for the DNA or RNA of eight different viruses, including five strains of herpes. Virus exposure was associated with all four negative outcomes, but especially highly with pre-eclampsia. Mothers of babies who tested positive for any virus were more than five times as likely to have had pre-eclampsia, and those whose babies tested positive for a herpes virus were almost six times as likely.

The study appears in the March issue of The British Journal of Obstetrics and Gynaecology.

The authors acknowledged that their cases and controls were not matched and that the small number of cases of pre-eclampsia in their sample (23) made it difficult to draw firm conclusions.

Still, Dr. Alastair MacLennan, a co-author of the paper and head of the obstetrics department at the University of Adelaide, said the finding is interesting, although it needs confirmation with studies that include more cases. "As yet,"he said, "we don't know its clinical significance."