Background The elderly population is often in continuous use of several medications and is more subject to the "iatrogenic triad" of polypharmacy, potentially inappropriate medication use and drug-drug interactions. However, few studies have investigated these three factors concomitantly. Purpose To assess the prevalence and inter-relationship of potentially inappropriate medication (PIM) use, polypharmacy and drug-drug interactions in older adults, together with their associated factors. Setting city of Juiz de Fora, Brazil. Methods a cross-sectional, observational and door-to-door epidemiologic study in community-dwelling older adults was conducted. Main outcome measure The primary outcomes were polypharmacy, inappropriate medication use (2012 Beers and 2015 STOPP criteria) and drug-drug interactions. Associated factors were also investigated using bivariate and multivariate analyses. Results a total of 368 (92%) older adults were in continuous use of at least one drug. There was a high prevalence of polypharmacy (44.6%), drug-drug interaction (72.3%) and PIMs by Beers (42.1%) and PIMs by STOPP (46.2%). Analysis of the inter-relationship of the criteria (polypharmacy, PIMs STOPP and drug-drug interactions) revealed that 108 (29.3%) of the older adults had all three criteria concomitantly and multivariate analysis showed that frailty and having a caregiver were associated with this "iatrogenic triad". Conclusion A high prevalence of iatrogenic effects from drugs was found in the older adults assessed. One in every three elderly participants of the study had all three iatrogenic criteria concomitantly, highlighting the major public health impact of this problem. The results of this study can serve to inform new preventive and educational strategies for health professionals.

The present study found moderate-to-high levels of concordance among the four PIM criteria assessed, pointing to a consensus in this field. However, each criterion showed particular characteristics: the EU(7)-PIM criterion had higher sensitivity, the Taiwan criterion higher specificity, and the Beers and STOPP a more balanced profile. These results highlight that each criterion has its own characteristics, and should be used according to health providers' objectives. Geriatr Gerontol Int 2017; 17: 1628-1635.