Depression and smoking raise back pain risk, and back pain returns the favour

Chronic back pain sits alongside a long list of correlated risk factors, from smoking and drinking to depression and low educational attainment, but untangling cause from consequence is notoriously hard. People in pain may smoke or drink more because of the pain, and shared genetic influences can create apparent links that neither variable causes. Conventional observational studies cannot easily separate these possibilities, and randomising people to smoke or to be depressed is impossible. Mendelian randomisation offers a way in by using inherited genetic variants, fixed at conception, as proxies for lifelong exposure.

The authors ran a bidirectional Mendelian randomisation analysis, testing the causal effect of education, smoking, alcohol consumption, physical activity, sleep and depression on chronic back pain, and then the causal effect of chronic back pain on those same factors. Genetic instruments were drawn from the largest published genome-wide association studies of each trait in individuals of European ancestry. Three analytic approaches were used: inverse variance weighted meta-analysis, Causal Analysis Using Summary Effect, and sensitivity analyses. A relationship was interpreted as causal only if it remained statistically significant after accounting for multiple testing and if the direction and size of the estimates agreed across all methods.

Four exposures met the criteria for a causal effect on chronic back pain: greater education was protective, with an odds ratio of 0.54 per 4.2 years of schooling, while ever smoking (1.27), greater alcohol consumption (1.29 per category increase) and major depressive disorder (1.41) each raised risk. In the reverse direction, chronic back pain appeared to increase alcohol consumption (1.19) and smoking (1.21), suggesting a feedback loop for both behaviours. The remaining relationships tested, including physical activity and sleep, did not meet the pre-defined criteria for a causal association in either direction.

This adds genetic support to the long-standing observation that back pain is not purely a matter of spines and mechanics, and it puts numbers on the bidirectional nature of the smoking and alcohol links. Where earlier studies could only report associations, the design here narrows the plausible explanations. It also complicates the assumption that physical activity and sleep are causally protective, since neither cleared the study's threshold, though failing to detect an effect is not the same as showing none exists. The genetic instruments and the European ancestry samples both limit how far the findings travel.

This is a reason to think of persistent back pain as entangled with mood and habits rather than sitting apart from them. It suggests smoking and drinking may both contribute to back pain and be increased by it, which is a loop worth being aware of if you are managing long-term pain. Depression showed the strongest effect on risk of the factors tested, which supports treating mental health as part of the same conversation rather than an afterthought. The absence of a clear causal signal for physical activity and sleep here does not mean either is unimportant; it means this particular method did not detect one.