Effective pain relief relies on widely used agents such as paracetamol, NSAIDs and opioids, yet their long-term safety and efficacy, especially for chronic non-cancer pain, remain uncertain. Randomised trials are scarce and post-marketing studies often suffer from confounding, limiting causal interpretation of adverse outcomes (e.g. cardiovascular events, cognitive decline). In parallel, patients differ markedly in analgesic response and toxicity because of genetic, epigenetic, physiological and environmental factors. Variants affecting drug metabolism (e.g. CYP2D6 for codeine/tramadol) and pharmacodynamic targets (OPRM1, COMT, PTGS2/COX-2, TRPV1, NMDA subunits) alter efficacy, tolerance and risk of harm. Sex, age, inflammation, co-medications and socio-environmental context further modify these effects, and drugs developed for one pain type may behave differently in others.
This project will leverage UK Biobank’s linked genetic, prescription, health-record and lifestyle data to clarify the benefit-risk profile of common analgesics across pain conditions. Using modern causal-inference techniques, we will estimate whether major adverse outcomes are attributable to analgesic exposure rather than underlying morbidity. Pharmacogenomic and phenotypic modifiers of response will be explored to identify subgroups with favourable benefit-risk balance. By integrating pharmacoepidemiology, causal methods and pharmacogenomics at population scale, the study aims to inform precision prescribing, optimising analgesic choice and dose while reducing serious adverse events.