Are all placebo respondents non-smokers?

In smoking cessation trials, placebo response rates are reported to be rather low, and they are lowest when compared to the placebo response rates of treatment of other addictions. We hypothesized that high placebo response rates in trials outside of cessation treatment may predict low participation of smokers, and that non-smoking may be a behavioral marker of the placebo response.

We re-analyzed raw data from a randomized controlled drug trial in functional dyspepsia (n=315) for the number of smokers and non-smokers in both treatment arms (drug, placebo) and varied the responder definition in a sensitivity analysis.

An equal number of smokers and non-smokers were assigned to drug and placebo. With the pre-defined responder definition (40% symptom improvement of a patient-reported outcome scale), 3% of placebo responders were smokers, but around 20% of patients among the placebo non-responder, and the drug responders and non-responders. With a more restrictive response definition (50% improvement) none of the placebo responders (n=29) was a smoker while the percentage of smokers remained similar in all other groups (p<.001). Age and gender did not affect this distribution.

Non-smoking behavior may be association with higher placebo response rates and may indicate a biomarker for reward sensitivity of the nicotine-dopamine pathway. Common genotypes may underlie both behaviors.