Examining the effects of nicotine versus non-nicotine constituents of tobacco smoke on depression using multivariable MR

Background: 

Smoking and depression commonly co-occur. There is growing evidence across multiple study designs that tobacco smoking may cause depression. However, questions remain around what constituents of tobacco smoke (e.g., nicotine, carbon monoxide) may confer negative effects for mental health. Considering the rise in prevalence of use of alternative nicotine delivery systems (e.g., e-cigarettes), it is essential to understand and separate the effects of nicotine use from the impact of tobacco smoke exposure. 

Methods:

We conducted univariable and multivariable Mendelian randomisation (MR) analyses, using published, summary-level GWAS data to extract genetic instruments for the exposures: smoking heaviness (measured via cigarettes per day [CPD]) and nicotine metabolite ratio (NMR). For the outcome, we conducted a GWAS of major depressive disorder (MDD) using individual-level data from UK Biobank stratified by smoking status (i.e., ever, never, current, former). To aid interpretation, effects were examined: (i) stratified by smoking status, and (ii) considered across multiple methods (inverse variance weighted [IVW], MR Egger, weighted median, and weighted mode).

Results:

Univariable MR indicated a causal effect of genetic liability for CPD on MDD (ORIVW = 1.13, 95%CI = 1.04 – 1.23, P = 0.003) but no clear evidence for a causal effect of genetic liability for NMR on MDD (ORIVW = 0.98, 95% CI = 0.97 – 1.00, P = 0.134), with results consistent across MR approaches. Multivariable MR indicated an effect of genetic liability for CPD on MDD when accounting for NMR (ORMVMR-IVW = 1.19, 95% CI = 1.03 – 1.37, P = 0.017; ORMVMR-EGGER = 1.13, 95% CI = 0.89 – 1.43, P = 0.300), with only weak evidence of a small, potentially negligible, effect of NMR on MDD, when accounting for CPD (ORMVMR-IVW = 0.98, 95% CI = 0.96 – 1.00, P = 0.056; ORMVMR-EGGER = 0.98, 95% CI = 0.96 – 1.00, P = 0.038).

Conclusions:

Our study provides supporting evidence of a causal link between tobacco smoking and depression, as reported in previous MR studies using different exposure instruments (e.g., lifetime smoking, smoking initiation). Results suggest that the link between tobacco smoking and depression may be largely independent of nicotine exposure and implies the role of alternative causal pathways (i.e., driven by other constituents of tobacco).