Genetic and Clinical Determinants of Variation in Drug Response in Type 2 Diabetes: Insights From the Scottish and UK Biobank Cohorts

Partitioned polygenic risk scores by biological pathway (beta-cell function, obesity, liver lipids, bilirubin) associated with glycaemic response by drug class

Two population-based cohorts were analysed in parallel: GoDARTS, with 41 802 patients initiating one of six major glucose-lowering drug classes, of whom 11 615 had genotype data, and UK Biobank, with 9 371 individuals including 8 293 with genetic data. The primary outcome was change in HbA1c 12 months after treatment initiation, set against clinical factors and 14 partitioned polygenic risk scores representing type 2 diabetes biological pathways. Baseline HbA1c was the strongest determinant of response (p < 0.001), older age was consistently associated with greater HbA1c reduction, while BMI and total cholesterol showed drug-class-specific associations, higher BMI being linked to better response to thiazolidinediones. In meta-analysis across cohorts, higher overall genetic risk for type 2 diabetes was associated with greater HbA1c reduction on sulfonylureas (β = −0.46 mmol/mol, p = 0.013). Specific genetic profiles also emerged: beta-cell function clusters with sulfonylureas (β = −0.57, p = 0.002), obesity-related variants with GLP-1 receptor agonists (β = −1.49, p = 0.04), liver-lipid variants with SGLT2 inhibitors (β = −0.84, p = 0.05) and a bilirubin score with DPP-4 inhibitors (β = −0.69, p = 0.006).

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

The methodological contribution is partitioning the polygenic score by biological pathway rather than handling a single global score: the resulting associations become mechanistically interpretable, the beta-cell pathway with sulfonylureas being the most coherent example. The effect sizes must nonetheless be read for what they are, a few tenths of mmol/mol of HbA1c, far below what would shift a prescribing decision for an individual patient. The most practically useful message remains the dominance of clinical determinants, baseline HbA1c and age above all, over the genetic component; the polygenic score here is a tool for understanding pharmacological pathways rather than a criterion for drug choice.

Analysis by Dr Thibaut Benquey