Treatment with bupropion and varenicline for smoking cessation and the risk of acute cardiovascular events and injuries: A Swedish case-crossover study

Abstract

Introduction: Bupropion and varenicline are non-nicotine medications used for smoking cessation that mitigate craving and withdrawal symptoms. We aim to investigate whether these drugs increase the risk of selected acute adverse outcomes when used in medical practice. Methods: Population-based case-crossover design using data from Swedish health and administrative registers. Adult individuals diagnosed with acute myocardial infarction, stroke, suicide, suicide attempt, fall injury, or that suffered a road traffic crash from 01.10.2006 for bupropion, or from 01.03.2008 for varenicline, until 31.12.2013 were included. Different lengths of exposure periods were analyzed within the 12-week hazard period prior to the adverse outcome (1-14, 15-28, and 29-84 days). The control period was matched using the interval preceding the hazard period (85-168 days), and breaking it up into equivalent periods (85-98, 99-112, and 113-168 days). Conditional logistic regression with each case considered as one stratum was used to estimate adjusted odds ratios (OR) and confidence intervals (CI). Results: Neither medication was associated with consistent higher risks for any of the adverse outcomes. For bupropion and varenicline, respectively, in the 1-14 days hazard period, OR (95% CI) were: myocardial infarction 1.14 (0.55 to 2.34) and 1.06 (0.70 to 1.62); stroke 1.16 (0.39 to 3.47) and 1.26 (0.72 to 2.17), and traffic crashes 0.85 (0.39 to 1.85) and 1.48 (0.90 to 2.41). In the other periods, ORs were similar or even lower. For falls and suicidal events ORs were generally below one for both drugs. Conclusion: The available evidence suggests that if prescription guidelines are properly followed regarding potential contraindications both of these medications could be considered relatively safe. Implications: The reliable exposure and diagnosis assessment used in this nationwide registerbased study, along with the number of cases gathered makes this sample one of the largest of its type to assess potential side effects associated with the use of these drugs. Neither medication was associated with consistent higher risks for any of the adverse outcomes studied.