Evaluation of Healthy User Effects With Metformin and Other Oral Antihyperglycemia Medication Users in Adult Patients With Type 2 Diabetes

Abstract

Healthy user bias, whereby health-seeking patients are more likely to initiate preventive therapies and engage in healthy lifestyle behaviours, is well known in observational studies, particularly with statins. However, its influence in studies of oral antihyperglycemic therapies is unknown. We sought to explore the healthy user effects in metformin users vs. nonusers on various health outcomes that should not be associated with metformin use. We conducted a retrospective cohort study using data from Alberta, between 2008 and 2015, to examine the association between metformin use and various health outcomes. We identified 135,301 new users of oral antihyperglycemic agents. The mean age was 55years, 75,949 (56%) were men and 130,725 (97%) had had at least 1 metformin prescription during a mean follow-up period of 3.4years. Metformin users were less likely to be involved in accident-related events (adjusted hazard ratio [aHR] 0.90; 95% CI 0.85 to 0.96), were more likely to have preventive screening services (aHR 1.16; 95% CI 1.11 to 1.21), were less likely to experience other clinical events, such as asthma and gout attacks (aHR 0.90; 95% CI 0.84 to 0.97), and had lower risks for all-cause mortality (aHR 0.57; 95% CI 0.51 to 0.63) compared to nonusers. Our results suggest that metformin users are more likely to initiate preventive therapies and engage in other healthy behaviours. Failure to account for these behaviours may introduce healthy user bias into studies evaluating the effects of oral antihyperglycemic therapies.