Abstract 16008: Association Between Statin Exposure and Diabetes Incidence in a Nationwide Claims Database
Background: Statins reduce risk for cardiovascular events; however, concerns have been raised about their association with diabetes mellitus (DM). We used a national health insurance claims database, which captures the health care experiences of >50 million privately insured individuals, to determine if there was a significant association between statin exposure and incidence of DM when controlling for multiple covariates not accounted for in prior analyses.
Method: A cohort of non-diabetic patients who initiated treatment with a statin was matched with a control group of patients who did not receive a statin based on age, gender and geographic location. Patients were then followed for a minimum of 2 years to determine whether they developed DM (i.e. ≥2 ICD-9 code for DM and/or pharmacy claim for DM medications). In addition to Pearson chi-square tests for unadjusted comparisons, Cox proportional hazard models were used to examine the relationship between statin exposed and unexposed groups adjusting for HTN, CAD, PAD, CVA, CHF, hyperlipidemia and healthcare costs in the prior year. Predefined subgroup analyses were performed based on age, gender, obesity status and statin type.
Results: The study population included 231,478 matched pairs of exposed and unexposed subjects (age 51.3 ±7.5 years; follow up duration of 1140 ±212.77 days). Statin exposure was associated with an increased DM incidence [adjusted HR 2.236 (2.182, 2.291)]. However, the effect was significant only after the age of ≥50 years [for age groups 30-39, 1.0 (0.86, 1.163); 40-49, 1.053 (0.91, 1.22); 50-59, 1.19 (1.028, 1.374); 60-69, 1.276 (1.101, 1.478) with age group 20-29 as reference]. Compared to males, statin-exposed females had an increased incidence of DM [HR 1.08 (1.06, 1.10)]. In obese individuals DM incidence was 1.820 (1.610, 2.057) in statin exposed versus unexposed. All statin types were associated with increased DM incidence without any statistically significant differences between them.
Conclusion: In this large study, there was an association between statin exposure and diabetes incidence. These findings are consistent with other data reported in literature and strengthen the causal link between statin exposure and diabetes.
Author Disclosures: T. Ahmad: None. G. Liu: None. P. Alagona Jr: None. C.H. Chuang: None. A.J. Foy: None. S.M. Bokhari: None. D. Leslie: None.
- © 2015 by American Heart Association, Inc.