Polypharmacy cutoff and outcomes: five or more medicines were used to identify community-dwelling older men at risk of different adverse outcomes

J Clin Epidemiol. 2012 Sep;65(9):989-95. doi: 10.1016/j.jclinepi.2012.02.018. Epub 2012 Jun 27.


Objective: This study aimed to determine an optimal discriminating number of concomitant medications associated with geriatric syndromes, functional outcomes, and mortality in community-dwelling older men.

Study design and setting: Older men aged ≥ 70 years (n=1,705), enrolled in the Concord Health and Aging in Men Project were studied. Receiver operating characteristic curve analysis using the Youden Index and the area under the curve was performed to determine discriminating number of medications in relation to each outcome.

Results: The highest value of the Youden Index for frailty was obtained for a cutoff point of 6.5 medications compared with a cutoff of 5.5 for disability and 3.5 for cognitive impairment. For mortality and incident falls, the highest value of Youden Index was obtained for a cutoff of 4.5 medications. For every one increase in number of medications, the adjusted odds ratios were 1.13 (95% confidence interval [CI]=1.06-1.21) for frailty, 1.08 (95% CI=1.00-1.15) for disability, 1.09 (95% CI=1.04-1.15) for mortality, and 1.07 (95% CI=1.03-1.12) for incident falls. There was no association between increasing number of medications and cognitive impairment.

Conclusion: The study supports the use of five or more medications in the current definition of polypharmacy to estimate the medication-related adverse effects for frailty, disability, mortality, and falls.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Accidental Falls / statistics & numerical data
  • Aged
  • Aged, 80 and over
  • Cognition Disorders / diagnosis
  • Disability Evaluation
  • Frail Elderly
  • Geriatric Assessment / methods*
  • Humans
  • Male
  • Odds Ratio
  • Outcome Assessment, Health Care / methods*
  • Polypharmacy*
  • Prescription Drugs / adverse effects*
  • ROC Curve
  • Survival Analysis


  • Prescription Drugs