Drug treatment for diabetes in nursing home residents

Aims The aim of this study was to describe drug treatment for diabetes in a large sample of nursing home residents and to compare subjects' health outcomes according to the anti-diabetic agents used. Methods The cross-sectional data of 6275 residents [average age 86years (+/- 8.2); 73.7% women] from 175 nursing homes in France were analysed. Participants were divided into one of the following four groups: diabetes non-drug treatment, diabetes hypoglycaemic (e.g. insulins, sulphonylurea) treatment, diabetes non-hypoglycaemic (e.g. metformin) treatment and no diabetes. Group comparisons were made on functional ability (activities of daily living score) and on the prevalence of the following variables (yes vs. no): emergency department visits, falls and fractures. Results Of the participants, 1076 (17.1%) had diabetes: 222 participants in the non-drug treatment group, 722 in the hypoglycaemic group and 132 in the non-hypoglycaemic group. The remaining 5199 participants made up the group without diabetes. Insulin and metformin were used by 549 and 185 participants, respectively. Activities of daily living scores differed across the four groups, with those in the non-drug treatment group being the most disabled. Adjusted multivariate analyses showed that, compared with the group without diabetes, those in the hypoglycaemic group had a higher probability of emergency department visits (odds ratio1.26, 95%CI 1.03-1.54) and increased the incidence rate ratios (1.02, 95%CI 1.00-1.04) of disability (activities of daily living score), whereas the non-hypoglycaemic group was not significantly associated with these outcomes. Conclusions The use of hypoglycaemic drugs was associated with poor health outcomes in nursing home residents. Therefore, more attention must be paid to adapting anti-diabetic treatment in this complex population.