They have found that premature infants given H2 blocker treatment were 70 per cent more likely to develop intestinal inflammation leading to necrotising enterocolitis than infants not treated with the drugs.
In necrotising enterocolitis, the tissue lining the wall of the intestines dies and the digestive tract becomes unable to function. In some cases, the damage to the intestines may require surgery and in others the damage is so severe that the infant dies.
The researchers analysed data from a national register of 11,072 infants who weighed under 1.5kg at birth and survived for at least 12 hours.
The data were collected between September 1998 and December 2001. They compared rates of necrotising enterocolitis in infants treated with the H2 blockers cimetidine (Tagamet), famotidine (Pepcid), ranitidine (Zantac) and nizatidine (Axid) with rates in untreated infants.
A total of 7.1 per cent of the infants developed necrotising enterocolitis, with the highest rates in those weighing less than 750g. Treatment with H2 blockers was found to be associated with a 71 per cent increased risk of developing the disease.
Lead researcher Dr Ronnie Guillet, associate professor of paediatrics at the University of Rochester Medical Center in New York, advised caution in using H2 blockers for treating premature infants.
She suggested that H2 blockers might increase their risk of infection by lowering the acidity of the digestive tract, allowing gram-negative bacteria to flourish.
Although these bacteria are normally harmless, at high levels they might lead to necrotising enterocolitis. She added that the results were worrying because a large number of premature infants received H2 blockers.
The drugs are prescribed to premature infants to treat acid reflux, to prevent apnoea or to prevent excess stomach acidity in infants who are tube-fed.
Dr Elias Zerhouni, director of the US National Institutes of Health which funded the study, said: 'This study suggests that the common practice of prescribing H2 blockers to prevent or treat acid reflux in premature infants be discontinued, until more evidence on the safety of this practice is available.'
Pediatrics 2006; 117: 137-42