OBJECTIVE: Studies in Western populations have shown the association of nonsteroidal anti-inflammatory drugs (NSAIDs) and upper gastrointestinal bleeding (UGIB). The role of Helicobacter pylori infection in NSAIDs-related UGIB remains to be studied. We conducted a case-control study in Japan to investigate these related topics.METHODS:Cases of UGIB due to duodenal or gastric ulcer, or gastritis were identified in 14 study hospitals in various areas of Japan. For each case, two controls were identified from population registries in the same district. Information on drugs and other risk factors was obtained from 175 cases and 347 controls by telephone interviews. Anti-H. pylori antibody in the urine was measured in a single laboratory for all the cases and 225 controls.RESULTS:The odds ratio (OR) of UGIB was 5.5 for aspirin and 6.1 for other NSAIDs (NANSAIDs) (p<0.01). The OR for regular use was higher than for occasional use both for aspirin (7.7 vs 2.0) and NANSAIDs (7.3 vs 4.1). Loxoprofen (5.9), frequently used in Japan as a safe 'prodrug', was significantly associated with UGIB. The odds ratio for H. pylori infection was 4.9 and the relative excess risk due to the interaction between H. pylori and the use of NSAID was 1.2 (95% CI: -5.8-8.1).CONCLUSION:NSAIDs including loxoprofen increase the risk of UGIB in Japan as in Western countries, with a similar magnitude of association. There was no evidence of biological interaction between NSAIDs and H. pylori infection.