• myocardial infarction;
  • cohort study;
  • NSAIDs;
  • coxibs



Meta-analyses of observational studies show variability in the risk of acute myocardial infarction (AMI) among non-steroidal anti-inflammatory drugs (NSAIDs), with an increase in risk for rofecoxib and diclofenac, and no increase in risk for celecoxib, naproxen, or ibuprofen.

Methods and Results

We identified a cohort of 364 658 individuals aged 40–84 years who were enrolled in Saskatchewan Health, Canada, from 15 November 1999 to 31 December 2001. A nested case–control analysis compared 3252 incident cases of hospitalized AMI and out-of-hospital CHD deaths with 20 002 controls randomly sampled from the cohort. The incidence of AMI/CHD was 5.1 per 1000 person-years (95%CI: 5.0–5.3). The adjusted ORs (95%CI) of AMI/CHD in current users of individual NSAIDs compared with non-use were: celecoxib (1.11; 0.84–1.47), rofecoxib (1.32; 0.91–1.91), diclofenac (1.02; 0.75–1.38), naproxen (1.57; 0.98–2.52), ibuprofen (1.59; 0.88–2.89), and indomethacin (1.34; 0.81–2.19). Long-term use of rofecoxib was compatible with an increased risk (OR = 1.46; 0.97–2.22) while estimates of other individual NSAIDs were close to unity. Overall NSAID use was associated with a 30% increased risk of nonfatal AMI but was absent for fatal AMI/CHD.


This study showed a modest increased risk of AMI/CHD with various traditional NSAIDs and COX-2 inhibitors. Confidence intervals of estimated ORs included the null value for most comparisons. The study confirmed that the differentiation between traditional NSAIDs and COX-2 inhibitors is not a reliable tool for predicting cardiovascular risk associated with NSAIDs. Copyright © 2009 John Wiley & Sons, Ltd.