Causal impacts of smoking on pain conditions and the mediating pathways: a mendelian randomization study

Abstract Smoking is a risk factor for multiple diseases. We performed mendelian randomization (MR) analyses to investigate the causal association of smoking initiation on pain conditions and the potential mediating pathways. Genetic associated with smoking initiation at the genome-wide significance level were selected as instrumental variables. Genetic associations with 10 pain conditions were derived from the FinnGen and UK Biobank study. Multivariable MR analysis was conducted to explore the mediation effects of depression, insomnia and sedentary behavior. A series of sensitivity analyses were conducted to assess the stability of our research findings. Genetic liability to smoking initiation was associated with an increased risk of angina pectoris, dorsalgia, low back pain, pain in limb, pain in joint, pain in thoracic spine and sciatica in both FinnGen and UK Biobank study. These causal associations were largely mediated by major depression (2.9- 39.5%), sedentary behavior (13.0- 31.2%), insomnia (10.3- 33.1%) and combination of all three mediators (30.2- 65.3%). The effects of smoking on outcomes were partly attenuated after adjusting for depression, sedentary behavior and insomnia respectively, and the direct effect of smoking initiation on pain was diminished toward null after adjusting for combined three mediators. These results were robust to sensitivity analyses. Our findings illustrated the causal effect of smoking and a broad range of pain conditions, and major depression, sedentary behavior and insomnia mediate many of these associations.