Psychiatric adverse events in randomized, double-blind, placebo-controlled clinical trials of varenicline: a pooled analysis

Drug Saf. 2010 Apr 1;33(4):289-301. doi: 10.2165/11319180-000000000-00000.


Background: Varenicline (Chantix), Champix) has shown efficacy and tolerability as an aid to smoking cessation. In postmarketing surveillance, neuropsychiatric symptoms have appeared; however, their incidence and causal relationship to varenicline is not known.

Objective: We assessed the incidence and relative risk (RR) of psychiatric disorders in ten randomized, double-blind, placebo-controlled trials of varenicline for smoking cessation.

Methods: All smoking cessation phase II, III and IV randomized controlled clinical trials of varenicline versus placebo completed as of 31 December 2008, on file with the manufacturer (Pfizer, Inc.), were included. All studies have been published. All 3091 participants who received at least one dose of varenicline and all 2005 participants who received placebo were included in this analysis. These were men and women smoking > or =10 cigarettes/day, aged 18-75 years and without current psychiatric disease who received varenicline or placebo for 6 (one study), 12 (eight studies) or 52 (one study) weeks. Adverse events were recorded at each study visit and classified according to standard Medical Dictionary for Regulatory Activities (MedDRA) terms (version 11.0).

Results: The incidence of psychiatric disorders other than solely sleep disorders and disturbances was 10.7% in subjects treated with varenicline and 9.7% in subjects treated with placebo, with an RR of 1.02 (95% CI 0.86, 1.22). The RRs (95% CI) versus placebo of psychiatric adverse events with an incidence > or =1% in the varenicline group were 0.86 (0.67, 1.12) for anxiety disorders and symptoms, 0.76 (0.42, 1.39) for changes in physical activity, 1.42 (0.96, 2.08) for depressed mood disorders and disturbances, 1.21 (0.79, 1.83) for mood disorders and disturbances not elsewhere classified and 1.70 (1.50, 1.92) for sleep disorders and disturbances. There were no cases of suicidal ideation or behaviour in varenicline-treated subjects in the ten placebo-controlled studies analysed. However, among three trials that were excluded from the analysis because of their open-label design, two cases of suicidal ideation and one completed suicide were reported in patients who had been treated with varenicline. With the exception of sleep disorders and disturbances, there was no evidence of dose-responsivity.

Conclusions: There was no significant increase in overall psychiatric disorders, other than sleep disorders and disturbances, in varenicline-treated subjects in this sample of smokers without current psychiatric disorders. Ongoing studies are testing the use of varenicline in psychiatric patients.

Publication types

  • Meta-Analysis
  • Research Support, Non-U.S. Gov't

MeSH terms

  • Adolescent
  • Adult
  • Aged
  • Benzazepines / adverse effects*
  • Benzazepines / therapeutic use*
  • Double-Blind Method
  • Female
  • Humans
  • Incidence
  • Male
  • Mental Disorders / chemically induced*
  • Mental Disorders / epidemiology
  • Middle Aged
  • Nicotinic Agonists / adverse effects*
  • Nicotinic Agonists / therapeutic use*
  • Placebos
  • Quinoxalines / adverse effects*
  • Quinoxalines / therapeutic use*
  • Randomized Controlled Trials as Topic* / statistics & numerical data
  • Risk
  • Sleep Wake Disorders / chemically induced
  • Smoking Cessation / methods
  • Varenicline


  • Benzazepines
  • Nicotinic Agonists
  • Placebos
  • Quinoxalines
  • Varenicline