Glucocorticoid drugs are used to speed up the development of a foetus’s lungs.
New Scientist magazine reports that University of Toronto researchers gave the drugs to pregnant guinea pigs.
They found behavioural problems, such as hyperactivity, were more likely not only in the animals’ offspring, but in the next generation too.
Babies normally spend 40 weeks in the womb, but some can survive even if they are born 15 or 16 weeks early.
However, their lungs lack enough of a substance called a surfactant to breathe unassisted.
Since the 1970s, doctors have been injecting women at risk of having a very premature baby with synthetic glucocorticoid drugs, such as betamethasone, to speed up development, and increase the odds that the child will survive.
The drugs have proved very effective – a single dose cuts the death rate of such babies by up to 40%.
Doctors believe the drug works best when given within a week of birth – so a baby who is not as premature as expected may be exposed to repeated doses.
The Toronto team chose to experiment on guinea pigs because they have similar placentas to humans, and give birth to relatively mature live young.
The guinea pigs received the equivalent of three injections of betamethasone, and compared them to a group given either three injections of saline, or nothing at all.
The offspring of the drug group showed signs of abnormal behaviour compared with the other groups.
This tallies with previous research which has found that human babies whose mothers are given multiple doses of betamethasone also show signs of hyperactivity.
But the drug affected the next generation of guinea pigs too.
When affected female offspring were mated with normal males, their young also had physiological and behavioural abnormalities.
Their male pups showed little interest in exploring a new environment, while females were hyperactive and made strange noises.
The Toronto team believes the drug may alter the way a foetus’s genetic material is translated.
Such changes are known to be passed down the generations.
Benefits outweigh risks
Professor Henry Halliday, an expert on perinatal medicine at Belfast’s Royal Maternity Hospital, told the BBC News website the use of betamethasone had helped to reduce death rates among premature babies.
However, he said there was no clear evidence that giving repeated doses was any more effective than a single course – and some evidence from human studies to suggest that repeated courses of steroids might affect foetal growth and later behaviour.
“This research in guinea pigs seems to confirm that these behavioural problems may pass down the generations after repeat dosing.
“Steroids are powerful dugs with widespread effects throughout the body; they do not act at a single site and they also have genomic effects which may be life-lasting.
“They should not be given to humans without careful consideration of risks versus benefits.”
“But in the case of threatened preterm birth before 34 weeks’ gestation the benefits would appear to outweigh the risks.”
Dr Andrew Lyon, of the Simpson Centre for Reproductive Health, Royal Infirmary of Edinburgh, cautioned against drawing conclusions from work in animals that have a very short gestation compared to humans.
He also stressed that in most cases doctors would only use one course of betamethasone.
But he said: “This is a good example of where continuing research is needed looking at a treatment that appeared to be so good that many thought that we had found the answer to many of our problems.
“Medicine is full of examples where new, and apparently ‘miraculous’, treatments turn out to have important downsides as more research is done.”