Article Text

Download PDFPDF

A call to include people with mental illness and substance use disorders alongside ‘regular’ smokers in smoking cessation research
  1. Anna Lembke1,
  2. Keith Humphreys1,2
  1. 1Department of Psychiatry and Behavioral Sciences, Stanford University, School of Medicine, Stanford, California, USA
  2. 2Center for Innovation to Implementation, VA Palo Alto Health Care System
  1. Correspondence to Professor Keith Humphreys, Department of Psychiatry and Behavioral Sciences, Stanford University School of Medicine, 401 N Quarry Road (MC: 5717), Stanford, CA 94305, USA; knh{at}


This commentary points out that smoking is increasingly concentrated among people with psychiatric problems and other substance use disorders (eg, alcohol use disorder), and argues that for clinical, ethical and efficiency reasons, such individuals should be routinely enrolled in smoking cessation research.

  • Co-substance use
  • Priority/special populations
  • Disparities

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Over the past half century, the smoking population of the developed world has evolved to include an increasing proportion of individuals who have comorbid mental health disorders (eg, alcohol use disorder, depression, schizophrenia).1 Yet most smoking cessation trials continue to exclude these growing populations of vulnerable smokers from research participation, thereby reducing the clinical utility and public health impact of the field's science. This commentary calls for a more inclusive approach to smoking cessation research. Rather than excluding smokers with mental illness and alcohol/drug problems, or creating a separate area of research to study smoking cessation in these subpopulations, smoking cessation researchers should as a matter of course allow these smokers to enrol in future treatment research alongside ‘regular’ smokers.

Certain smokers (some with comorbid mental health problems, some without) must be excluded from clinical trials because of legitimate ethical concerns, for example, patients who lack capacity to provide informed consent or who are taking a medication that interacts with a pharmacotherapy understudy. However, this reality cannot account for the large proportion of smokers who are denied research participation. Based on typical smoking cessation study exclusion criteria, two-thirds of smoking adults and three-fourths of smoking adolescents would be barred from participating in most smoking cessation clinical trials.2 ,3 The widespread practice of not allowing smokers with mental health and substance use disorders (other than tobacco) to enrol in smoking cessation trials is the largest contributor to this high rate of exclusion. One review of the smoking cessation trial literature found that 40% excluded depressed smokers, 55% excluded smokers with alcohol use disorders and 59% excluded those taking psychiatric medications.2 One critic described the practice of excluding smokers with mental health issues as ‘scandal’,4 which is reasonable given the stunning 62% rate of smoking among people with schizophrenia, the 42.6% rate of binge drinking among all smokers, and the enormous tobacco-related health damage in the seriously mentally ill population.5

No reasonable observer could doubt that smoking cessation researchers care about the well-being of smokers who have comorbid mental health problems. What then accounts for the tradition of excluding such individuals from smoking cessation trials?

In some cases, exclusion may reflect a commendable concern that treatments that help ‘regular’ smokers are ineffective for or even harmful to smokers who have alcohol, illicit drug or psychiatric disorders. In some particular clinical circumstances, for example, following discharge from psychiatric inpatient care, typical interventions may indeed not always be sufficiently potent to induce smoking cessation.6 However, for more prevalent clinical care situations, common interventions in the field (eg, varenicline or nicotine replacement therapy coupled with brief behavioural counselling) perform similarly for smokers with and without co-occurring mental health problems.7 Even when the question is restricted to the most severely troubled subset of smokers with comorbid mental health problems—those with serious mental illness—meta-analytic work indicates that the interventions that are effective among the general population of smokers are usually comparably beneficial.8 Further, contrary to fears that smoking cessation may perturb the course of mental illness, recent trials indicate that psychiatric symptoms are not worsened and indeed may even diminish because of smoking cessation.9 ,10

A second concern of some smoking cessation researchers is that if more impaired smokers are included, the costs and challenges of conducting research will increase. For some subpopulations, for example, individuals with schizophrenia who are homeless, and illicit drug-dependent individuals who go in and out of incarceration, tracking and assessing patients at follow-up will be more challenging. Grant review committees should be sympathetic to this reality when assessing proposed budgets, and researchers should take advantage of the empirically grounded methods available for conducting research with mobile, socially disadvantaged populations.11 That said, recall that most smokers with mental health problems are not in such markedly unstable situations: the modal case is a smoker who is simply depressed.7 Further, relaxing exclusion criteria can in fact reduce some study costs by shortening the time needed for sample recruitment.

What might be gained if smoking cessation research moved from typically excluding to typically including smokers with comorbid substance use and psychiatric disorders? Three major benefits are within reach.

First, smoking cessation research would be more credible and useful to the many clinical practitioners who treat large numbers of smokers who have comorbid mental health problems (eg, clinicians in the UK National Health Service and US Veterans Health Administration). Such an expansion of the impact on clinical practice is obviously in keeping with the central goals for medical research.

Second, ending the segregation of those smokers with comorbid mental health problems into a separate clinical research domain would be a more efficient use of scarce research dollars.12 Given that treatments that benefit ‘regular’ smokers also typically benefit smokers who have comorbid alcohol, illicit drug or psychiatric disorders, conducting a second set of trials on these smokers at a later date represents money poorly spent.

Third, while it is admirable indeed that individual researchers have in recent years conducted some of the first rigorous studies of smokers with comorbid psychiatric disorders,9 ,10 ,13 one can still question the research traditions that create the need for such studies in the first place. Excluding a population from most clinical trials in a field poses ethical problems if these patients are required to wait longer for determination of safety and efficacy, that is, until a second or third trial. Arguably, given the disproportionate damage of tobacco to people with mental illness and alcohol/drug use disorders, such populations should be at the front of the line to learn what treatments can help them quit smoking. However, under current research traditions, their interests are addressed last, if they are addressed at all. Their welfare would be better secured if they and those clinicians who care for them benefited from new research as promptly as the rest of the smoking population. This can only occur if these patients are included in such research as a matter of course. That substance use and mental health-focused exclusion criteria tend to disproportionately deny research participation to low-income people and racial minorities only heightens the ethical case for change.14 ,15

The status quo thus deserves to be questioned: inclusion should become the norm rather than the exception in the smoking cessation field. With smoking becoming more concentrated in less healthy and less privileged segments of society, the science of tobacco cessation must alter its ways to bring its considerable blessings to the current population of smokers.



  • Funding US Department of Veterans Affairs RCS-04-141, The Greenwall Foundation, Ethics of Exclusion in Medical Research/GF 14-213.

  • Competing interests None declared.

  • Provenance and peer review Not commissioned; externally peer reviewed.