Genotype-predicted drug response phenotypes and their co-occurrence with dispensed medicines among 738,531 participants in the UK Our Future Health study

Drug response phenotypes are predicted from imputed genotypes at 17 pharmacogenes with prescribing guidelines and then matched to medicines actually dispensed; the dominant pairings involve CYP2C19 (proton pump inhibitors, antidepressants) and SLCO1B1 (statins).

This cross-sectional analysis of Our Future Health, a new UK national biobank, applied PharmCAT v3.1.1 to imputed genotypes from 738,531 participants across 17 pharmacogenes with established prescribing guidelines. Every participant carried at least one actionable pharmacogenomic phenotype, with a mean of 6.1 (SD 1.3), a count similar across genetically inferred ancestry groups even though the contributing pharmacogenes differed. Using linked primary care dispensing records, 36.8% (95% CI 36.7-36.9) had been dispensed at least one medicine between April 2018 and June 2025 matched to a gene for which they carried an actionable phenotype, a proportion rising with age and reaching 43.7% to 58.9% across ancestry groups among those aged 70 years or older. Participants carried an actionable phenotype for a mean of 13.8 of the 33 medicines dispensed in English primary care with pharmacogenomic guidance (SD 6.5), of which a mean of 0.6 (SD 1.0) had actually been dispensed. Co-occurrence was concentrated in a few widely dispensed classes, mainly proton pump inhibitors and antidepressants acting through CYP2C19, and statins through SLCO1B1.

Synthesis written by Geno'X. For the full original abstract, please refer to the source publication.

This work provides the denominator missing from debates on pre-emptive testing: one third of adults, and up to nearly six in ten after age 70, have already been dispensed a medicine relevant to one of their actionable phenotypes. The concentration of the signal on CYP2C19 and SLCO1B1 supports starting with a targeted test covering a few high-yield genes rather than a broad panel, a convenient argument against cost objections. Two caveats matter: genotypes are imputed, which limits resolution for structurally complex pharmacogenes, and co-occurrence is not demonstrated clinical benefit — this preprint measures an opportunity, not an effect on adverse drug events.

Analysis by Dr Thibaut Benquey