New-Onset Diabetes After Statin Exposure in Elderly Women: The Australian Longitudinal Study on Women's Health.

INTRODUCTION: Extensive clinical research has consistently shown statins lower the risk of cardiovascular events and mortality. Some studies also suggest statins increase the risk of new-onset diabetes. Research to date has rarely included elderly women, hence little is known about the risk of diabetes after statin exposure in this population.

OBJECTIVES: Our objectives were to evaluate and estimate the risk of new-onset diabetes associated with statin exposure in a cohort of elderly Australian women.

METHODS: We performed an analysis of a population-based longitudinal cohort study with data linkage to the national death index and to national databases of non-hospital episodes of medical care and prescription medications dispensing. Participants included 8372 Australian women born between 1921 and 1926, alive at 1 January 2003, free of diabetes, and eligible for data linkage. Statin exposure was ascertained based on prescriptions dispensed between 1 July 2002 and 31 August 2013.

RESULTS: Over 10 years of follow up, 49% of the cohort had filled a prescription for statins and 5% had initiated treatment for new-onset diabetes. Multivariable Cox regression showed statin exposure was associated with a higher risk of treatment for new-onset diabetes (hazard ratio 1.33; 95% confidence interval [CI] 1.04-1.70; p = 0.024). This equates to a number needed to harm (NNH) of 131 (95% CI 62-1079) for 5 years of exposure to statins. Risk increased with increasing dose of statin from the hazard ratio of 1.17 (95% CI 0.84-1.65) for the lowest dose to 1.51 (95% CI 1.14-1.99) for the highest dose.

CONCLUSION: The dose-response for statins on new onset of diabetes suggests elderly women should not be exposed to higher doses of statins. Elderly women currently taking statins should be carefully and regularly monitored for increased blood glucose to ensure early detection and appropriate management of this potential adverse effect, including consideration of de-prescribing.