Background: As well as increasing the risk of new-onset type 2 diabetes (T2DM), there is evidence from intervention trials that statin therapy, especially the more potent drugs in this class, can worsen glycaemic control.
Aims: To determine whether clinically significant HbA1c increases occur with statin therapy in T2DM patients in a community setting.
Methods: We analyzed data from participants in the longitudinal observational Fremantle Diabetes Study Phase I who commenced a statin between 1994 and 2000, and who had an annual assessment either side of this therapeutic change.
Results: At baseline (1993-6), 86 (6.6%) of the 1,296 T2DM participants were taking a statin (94.2% simvastatin, 5.8% pravastatin). During 4.0±2.8 years of follow-up from entry to their last assessment, 317 of the remaining 1,204 started statin treatment, of whom 315 had HbA1c measured either side of this. Over half (52.7%) were prescribed simvastatin, 26.0% pravastatin, 18.7% atorvastatin, and 2.5% cerivastatin or fluvastatin. At baseline, these 315 participants had a mean age of 62.5±8.6 years, median [inter-quartile range] diabetes duration 3.0 [0.5-7.0] years, median HbA1c 7.2% [6.2%-8.5%] and 49.2% were male. Before statin initiation, the median HbA1c was 7.1% [6.3%-8.2%] vs 7.2% [6.4%-8.0%] immediately afterwards (P=0.38). Excluding 34 (10.8%) who had their diabetes treatment intensified concurrently (diet to oral therapy and/or insulin, or oral therapy to insulin), median HbA1c increased from 7.0% [6.3%-8.1%] to 7.1% [6.4%-8.0%], P=0.055 (mean change (95% CI): 0.12% (-0.02% to 0.26%)). Stratifying by statin type (151 simvastatin/74 pravastatin/49 atorvastatin) the largest mean HbA1c increase was for simvastatin (0.14% (-0.06% to 0.34%)) followed by atorvastatin (0.12% (-0.21% to 0.45%)) and pravastatin (0.05% (-0.18% to 0.28%); P=0.88 by ANOVA).
Conclusions: Usual-care statin initiation, especially those with highest potency, confers a similar increase in HbA1c to that in intervention trials in type 2 diabetes but this increase does not appear clinically significant.