Statins are commonly prescribed medications, effective in primary and secondary cardiovascular prevention and mortality. Controversy exists over whether statins promote incident diabetes. Past studies lack rigorous diabetes ascertainment, long follow-up or inclusion of diabetes-promoting covariates. This study investigates statin-use and incident glucose disorders over 8y in extensively-phenotyped participants in the Sydney Memory and Ageing Study (MAS).
Methods
MAS is a longitudinal cohort-study of community-dwelling participants (n=1037,70-90y) recruited from Eastern Sydney with biennial assessments.
Diabetes ascertainment used medical and medication history and fasting glucose ≥7.0mmol/L. Prevalent diabetes was excluded (n=161). Statin-use was ascertained by medication history. Incident diabetes was defined as newly meeting diabetes criteria. Impaired fasting glucose was defined as 5.6-6.9 mmol/L in the absence of diabetes history. An incident glucose disorder was defined as incident diabetes or incident impaired fasting glucose. Analyses used contingency tables, then multivariate regression (age, sex, BMI, and metabolic markers).
Results:
At baseline, 47% were statin users. Statin users had a higher prevalence of vascular disease and stroke and similar body mass index, waist, blood pressure, and glucose levels (Table 1). As expected, LDL and HDL cholesterol were lower in statin-users.
Incident diabetes rates ( n=28) were 3.7% for statin users (15/410) and 2.8% for non-users (13/466). The relative risk for incident diabetes with statin use was not significantly increased (RR 1.31, 95% CI 0.63-2.72, p = 0.467).
There was no significant difference in the risk of incident change in glucose status between statin users and non-users (relative risk 1.10, 95% CI; 0.81-1.50, p = 0.538).
Multivariate regression showed that only age and low HDL were predictors of incident diabetes in a model that contained age, sex, glucose and metabolic syndrome clinical phenotypes.
Conclusions:
This study of a well-characterised community-dwelling cohort found no evidence of increased risk of incident glucose disorders in older participants.