Aspirin increases mortality in diabetic patients without cardiovascular disease: a Swedish record linkage study.
Journal article, 2009

PURPOSE: Aspirin is effective in secondary prevention of cardiovascular disease. The results are less convincing when aspirin is used for primary prevention even in high-risk patients (i.e., patients with diabetes). We therefore analyzed the effect of aspirin on mortality and serious bleeding in diabetic patients with and without cardiovascular disease. METHODS: We performed a record linkage study of the patient registry of the Västra Götaland region in south-western Sweden, the Swedish mortality register and the Swedish register of dispensed drugs. All diabetic patients (n = 58 465) from 1 July 2005 to 30 June 2006 were followed up with respect to bleeding until 31 October 2006, and mortality until 31 December 2006. RESULTS: When 19 confounding factors (diseases and interventions) were assessed, aspirin significantly increased mortality in diabetic patients without cardiovascular disease from 17% (95% confidence interval; 95%CI, 1-36) at age 50 years to 29% (16-43) at age 85 years. In contrast aspirin tended to decrease mortality among elderly diabetic patients with cardiovascular disease. Theoretical calculations indicated that aspirin caused 107 excess deaths among diabetic patients without cardiovascular disease and prevented 164 deaths among diabetic patients with cardiovascular disease. Aspirin also increased the risk of serious bleeding by 46% (95%CI, 22-75) in diabetic patients without cardiovascular disease but decreased the risk among those with cardiovascular disease. CONCLUSION: Aspirin use in diabetes patients without cardiovascular disease remains controversial and current guidelines should be revised until results from ongoing large randomized controlled trials become available.



Record linkage study



L. Welin

Lidköping Hospital

Lars Wilhelmsen

University of Gothenburg

A. Bjornberg

Pollex Retro Research

Anders Odén

Chalmers, Mathematical Sciences, Mathematical Statistics

University of Gothenburg

Pharmacoepidemiology and Drug Safety

1053-8569 (ISSN) 1099-1557 (eISSN)

Vol. 18 12 1143-9

Subject Categories






More information