The Diabetogenic Effect of Statin Use May Interact With Polygenic Risk Scores for Type 2 Diabetes: Evidence From the UK Biobank

Abstract

Objectives: Statins are essential in the prevention of cardiovascular disease; however, their association with type 2 diabetes mellitus (T2DM) risk is concerning. We examined whether genetic susceptibility to T2DM modifies the association between regular statin use and T2DM risk.

Methods: This study included 447 176 individuals from the UK Biobank without baseline diabetes or major cardiovascular disease. Statin use was recorded at baseline, and T2DM incidence was determined using clinical records. Polygenic risk scores (PRS) for T2DM risk were provided by the UK Biobank. Using propensity scores adjusted for age, sex, body mass index, and comorbidities, 14 831 statin users were matched with 37 060 non-users. Cox proportional hazards models were used to estimate the interaction effect of statin use and PRS on T2DM incidence, adjusting for key confounders.

Results: In the propensity-matched cohort, 3675 of 51 891 participants developed T2DM over a mean follow-up period of 13.7 years. Within the top 5% of the PRS distribution, per 1000 person-years, the incidence of T2DM was 15.42 for statin users versus 12.18 for non-users. Among the lowest 5%, the incidence was 1.90 for statin users and 1.65 for non-users. Based on the Cox proportional hazards model, regular statin use was associated with a 1.24-fold increased T2DM risk (95% confidence interval [CI], 1.15 to 1.33). Furthermore, PRS exhibited a significant multiplicative interaction with regular statin use (odds ratio, 1.10; 95% CI, 1.02 to 1.19).

Conclusions: PRS may help identify individuals particularly susceptible to the diabetogenic effects of statins, providing a potential path for personalized cardiovascular disease management.

Keywords: Diabetes mellitus type 2; Genetic risk score; Hydroxymethylglutaryl-CoA reductase inhibitors.

Figure 1

Eligible population selection process. Participants diagnosed with diabetes of any type (ICD-10 codes E10–E14) or major atherosclerotic cardiovascular disease at baseline were excluded. As a result, 333 840 complete cases were extracted from a total population of 502 236. Using these 333 840 complete cases, propensity scores were calculated, and up to five controls were matched for each treated individual with a caliper of 0.2 pooled standard deviations. HbA1c, hemoglobin A1c; BMI, body mass index; T2DM, type 2 diabetes mellitus; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; ICD-10, International Classification of Diseases, 10th revision. ¹The numbers of missing data categories include duplicates.

Figure 2

(A) Incidence rate of T2DM by PRS groups and regular statin use status in the matched population, shown in units of 1000 person-years. (B) Cumulative incidence plot for T2DM risk stratified by regular statin use status. Error bars and shaded areas represent the 95% confidence intervals, highlighting statistically significant differences between groups. T2DM, type 2 diabetes mellitus; PRS, polygenic risk score.