Article does not prove that smoking cessation has an “effect” on mental healthBMJ 2014; 348 doi: https://doi.org/10.1136/bmj.g2018 (Published 10 March 2014) Cite this as: BMJ 2014;348:g2018
- Saskia C Sanderson, adjunct assistant professor1,
- Amy Taylor, research associate2,
- Marcus Munafò, professor of biological psychology2
- 1Icahn School of Medicine at Mount Sinai, New York, NY 10029, USA
- 2School of Experimental Psychology, University of Bristol, Bristol, UK
Taylor and colleagues show a clear association between smoking cessation and improved mental health.1 This may help reassure patients that quitting smoking need not have a lasting adverse effect on mental health, and that it might even improve mental health. Nevertheless, readers of the press release and media coverage could be forgiven for believing that the study provided evidence of a causal effect of smoking cessation on improved mental health, which it did not.
The authors present three possible explanations for their findings: smoking cessation improves mental health; improved mental health causes cessation; or a common factor explains both. We see little in their findings to support the apparent endorsement of the first over the second and third of these explanations in the media coverage.
Making causal inferences from observational data is problematic because of confounding and reverse causality. Changes in lifestyle factors other than smoking cessation may also influence mental health. In addition, although these data may not suggest that mental health affects the likelihood of attempting to quit, reverse causality could still be an explanation if change in mental health influences the ability to remain abstinent.
The comparison of the magnitude of the association between smoking cessation and mental health with effect sizes obtained from randomised controlled trials of antidepressants should also be treated with caution. Observational studies are generally more prone to bias, so effect sizes may be inflated.2 3 Thus, even if the association is causal, the true effect of smoking cessation will not necessarily be this big. We agree that additional methods, such as mendelian randomisation,4 will shed light on whether smoking cessation does causally affect mental health.
Cite this as: BMJ 2014;348:g2018
Competing interests: None declared.
Full response at: www.bmj.com/content/348/bmj.g1151/rr/687868.