Risk Factors for Potentially Inappropriate Medication Use in Older Adults: A 10-Year Cohort Study with Generalized Estimating Equation Models

Abstract

BackgroundMuch of the knowledge on the use of potentially inappropriate medications (PIM) in older adults is derived from cross-sectional studies, with little known about the risk factors over time. AimLongitudinal analysis was applied to estimate the occurrence and risk factors of PIM use among older adults in a 10-year follow-up.Methods Longitudinal study with 418 older adult residents of a capital city of Central-West Brazil. The PIM were classified according to the Beers criteria 2019. The usage rate was calculated at baseline (2008) and at the 10-year follow-up moment (2018). Analysis of predictors (sociodemographic, self-rated health, hospitalization, number of comorbidities, polypharmacy, diabetes, hypertension, hypercholesterolemia and nutritional status) was performed using Generalized Estimating Equation (GEE) models.ResultsMean age at baseline was 70.6 years (SD 7.1) and 76% were women; 221 older adults took part in the follow up. The rate of PIM use was 50.4% at baseline and 57.5% at the 10-year follow-up. Multiple analysis showed that PIM use in the cohort was statistically higher in the older adults with a history of hospitalization (RRadj: 1.20; 95%CI: 1.01-1.40), with three or more diseases (RRadj: 1.41; 95%CI: 1.14-1.74), with polypharmacy (RRadj: 1.81; 95%CI: 1.47-2.24) and with diabetes mellitus (RRadj: 1.24; 95%CI: 1.05-1.47).ConclusionA high level of potentially inappropriate medication use was observed, reaching 50% of the older adults, with a 7% increase in the prevalence over the 10-year follow-up period. Hospitalization, multimorbidities, polypharmacy and diabetes mellitus were associated with the use of these medications. Interventions for surveillance of the deprescribing process need to be encouraged to avoid potential harm caused by the use of medications.