Browse/search for people

Publication - Professor Marcus Munafo

    The Association of Cigarette Smoking with Depression and Anxiety

    A Systematic Review


    Fluharty, ME, Taylor, AE, Grabski, M & Munafò, MR, 2017, ‘The Association of Cigarette Smoking with Depression and Anxiety: A Systematic Review’. Nicotine and Tobacco Research, vol 19., pp. 3-13


    BACKGROUND: Many studies report a positive association between smoking and mental illness. However, the literature remains mixed regarding the direction of this association. We therefore conducted a systematic review evaluating the association of smoking and depression and/or anxiety in longitudinal studies.

    METHODS: Studies were identified by searching PubMed, Scopus, and Web of Science and were included if they: 1) used human participants, 2) were longitudinal, 3) reported primary data, 4) had smoking as an exposure and depression and/or anxiety as an outcome, or 5) had depression and/or anxiety as the exposure and smoking as an outcome.

    RESULTS: Outcomes from 148 studies were categorised into: smoking onset, smoking status, smoking heaviness, tobacco dependence and smoking trajectory. The results for each category varied substantially, with evidence for positive associations in both directions (smoking to later mental health and mental health to later smoking) as well as null findings. Overall, nearly half the studies reported that baseline depression/anxiety was associated with some type of later smoking behaviour, while over a third found evidence that a smoking exposure was associated with later depression/anxiety. However, there were few studies directly supporting a bidirectional model of smoking and anxiety, and very few studies reporting null results.

    CONCLUSIONS: The literature on the prospective association between smoking and depression and anxiety is inconsistent in terms of the direction of association most strongly supported. This suggests the need for future studies that employ different methodologies, such as Mendelian randomisation, that will allow us to draw stronger causal inferences.

    Full details in the University publications repository