The use of low-dose aspirin has been reported to be associated with an increased risk of upper gastrointestinal complications (UGIC). The coating of aspirin has been proposed as an approach to reduce such a risk. To test this hypothesis, we carried out a population based case-control study.

We identified incident cases of UGIC (bleeding or perforation) aged 40 to 79 years between April 1993 to October 1998 registered in the General Practice Research Database. Controls were selected randomly from the source population. Adjusted estimates of relative risk (RR) associated with current use of aspirin as compared to non use were computed using unconditional logistic regression.

We identified 2,105 cases of UGIC and selected 11,500 controls. Among them, 287 (13.6%) cases and 837 (7.3%) controls were exposed to aspirin, resulting in an adjusted RR of 2.0 (1.7-2.3). No clear dose-effect was found within the range of 75-300 mg. The RR associated with enteric-coated formulations (2.3, 1.6-3.2) was similar to the one of plain aspirin (1.9, 1.6-2.3), and no difference was observed depending on the site. The first two months of treatment was the period of greater risk (RR= 4.5, 2.9-7.1). The concomitant use of aspirin with high-dose NSAIDs greatly increased the risk of UGIC (13.3, 8.5-20.9) while no interaction was apparent with low-medium doses (2.2, 1.0-4.6).

Low-dose aspirin increases by twofold the risk of UGIC in the general population and its coating does not modify the effect. Concomitant use of low-dose aspirin and NSAIDs at high doses put patients at a specially high risk of UGIC.