Regular long-term use of aspirin may be associated with an increased risk of pancreatic cancer in women, research has suggested.
Aspirin has many beneficial effects
Harvard Medical School scientists found women who regularly used the drug for more than 20 years had a 58% increased risk compared to infrequent users.
Aspirin, and similar painkillers, have been shown to cut the risk of many types of cancer and heart disease.
However, the overall risk of cancer remains tiny, despite the increase.
Previous research has suggested that aspirin use may inhibit the development of pancreatic cancer.
However, these studies have been carried out on human cells, and laboratory animals, and have not examined the actual effect of taking aspirin, or other nonsteroidal anti-inflammatory drugs (NSAIDS).
The researchers' findings are based on an 18-year analysis of 88,378 women. Of these, 34% were current regular aspirin users, taking two or more 325mg aspirin tablets per week.
The researchers found there was no statistically significant difference in risk of pancreatic cancer between women who were currently taking aspirin and those who were not.
But when they considered longer-term use of the drug, the results were very different.
Not only did regular use of the drug over a 20-year period appear to increase risk, reasonably long periods of relatively heavy use also appeared to pose a problem.
The researchers quizzed the women about their aspirin use once every two years.
Women who reported that they regularly took 14 or more aspirin tablets per week in at least two out of three consecutive surveys were 86% more likely to develop pancreatic cancer than those who did not take the drug at all.
The study is published in the Journal of the US National Cancer Institute.
Lead researcher Dr Eva Schernhammer told BBC News Online: "Our study suggests that long-term use of aspirin could increase the risk of pancreatic cancer.
"However, the findings need to be confirmed by other large, well designed trials before doctors should consider changing recommendations for aspirin use."
Dr Schernhammer said it was possible that aspirin triggered inflammation of the pancreas.
In an accompanying editorial, Dr John Baron, of Dartmouth Medical School in New Hampshire, said: "There are no easy answers to the question of what aspirin and other NSAIDs do to pancreatic carcinogenesis.
"These findings are provocative and force us to think carefully about the actions of aspirin and other NSAIDs and the mechanisms underlying pancreatic cancer."
Dr Richard Sullivan, Head of Clinical Programmes at Cancer Research UK,
said: "Compared to the risk of developing pancreatic cancer due to smoking, excessive alcohol and weight gain the risk presented by aspirin is negligible.
"To say there is an 86% increased risk of pancreatic cancer is highly misleading as the absolute risk remains tiny.
"Interestingly, according to leading Cancer Registries in the US, the incidence of pancreatic cancer in American women is decreasing.
"However, this observational study indicates that aspirin may not necessarily be a panacea for all cancers."