Abstract
Background Smoking prevalence is higher amongst individuals with schizophrenia and depression compared to the general population. Mendelian randomisation (MR) can examine whether this association is causal using genetic variants identified in genome-wide association studies (GWAS).
Methods We conducted a GWAS of lifetime smoking behaviour (capturing smoking duration, heaviness and cessation) in a sample of 462,690 individuals from the UK Biobank, and validated the findings via two-sample MR analyses of positive control outcomes (e.g., lung cancer). Having established the validity of our instrument, we used bi-directional two-sample Mendelian randomisation to explore its effects on schizophrenia and depression.
Outcomes There was strong evidence to suggest smoking is a causal risk factor for both schizophrenia (OR = 2.27, 95% CI = 1.67 - 3.08, P < 0.001) and depression (OR = 1.99, 95% CI = 1.71 - 2.32, P < 0.001). We also found some evidence that genetic risk for both schizophrenia and depression cause increased lifetime smoking (β = 0.022, 95% CI = 0.005 - 0.038, P = 0.009; β= 0.091, 95% CI = 0.027 - 0.155, P = 0.005).
Interpretation These findings suggest that the association between smoking, schizophrenia and depression is due, at least in part, to a causal effect of smoking, providing further evidence for the detrimental consequences of smoking for mental health.
Funding This work was supported by the Medical Research Council Integrative Epidemiology Unit, the NIHR Biomedical Research Centre, University Hospitals Bristol NHS Foundation Trust and the University of Bristol.
Evidence before this study The association between smoking and mental health (especially schizophrenia and depression) is often assumed to be the result of self-medication (for example, to alleviate symptoms). However, more recent evidence has suggested that smoking might also be a risk factor for schizophrenia and depression. This alternative direction of effect is supported by meta-analyses and previous prospective observational evidence using related individuals to control for genetic and environmental confounding. However, observational evidence cannot completely account for confounding or the possibility of reverse causation. One way to get around these problems is Mendelian randomisation (MR). Previous MR studies of smoking and mental health have not shown an effect of smoking on depression and are inconclusive for the effects of smoking on schizophrenia. However, these studies have only looked at individual aspects of smoking behaviour and some studies required stratifying participants into smokers and non-smokers, reducing power.
Added value of this study We have developed a novel genetic instrument for lifetime smoking exposure which can be used within a two-sample MR framework, using publicly-available GWAS summary statistics. We were therefore able to test the bi-directional association between smoking with schizophrenia and depression to see if the effects are causal. We found strong evidence to suggest that smoking is a causal risk factor for both schizophrenia and depression. There was some evidence to suggest that risk of schizophrenia and depression increases lifetime smoking (consistent with the self-medication hypothesis) but the effects were stronger for depression than schizophrenia.
Implications of all the available evidence This study was the first to demonstrate evidence for an effect of lifetime smoking exposure on risk of schizophrenia and depression within a causal inference framework. This emphasises the detrimental public health consequences of smoking, not just for physical health, but also to mental illness.
Footnotes
The latest version of this manuscript has been revised to include additional sensitivity analyses and a slightly revised genetic instrument for lifetime smoking.
