One in Five Older Americans Still Face Dangerous Drug Combinations

Polypharmacy — the concurrent use of multiple medications and supplements — is a growing concern for older Americans, and a new nationally representative study published in JAMA puts hard numbers on the problem. Analyzing data from the National Social Life, Health, and Aging Project, researchers from the University of Southern California compared medication patterns in adults aged 62–85 between 2015–2016 and 2021–2023.

Overall prescription and supplement use held steady across the two periods, with roughly 84% of older adults using prescriptions and about 66% using dietary supplements regularly. However, polypharmacy — defined as five or more concurrent medications or supplements — rose meaningfully: prescription polypharmacy climbed from 31% to nearly 36%, and supplement polypharmacy jumped from 12.6% to 16.7%. These shifts matter because more drugs in combination exponentially increase the risk of harmful interactions.

The headline finding cuts both ways. The proportion of older adults with potentially major drug-drug interactions fell from 25.7% to 22.3% — a modest but statistically significant improvement. This decline was largely driven by reduced opioid and benzodiazepine prescribing. Yet antidepressants remained the most frequently implicated drug class in dangerous interactions, raising concerns about serotonin syndrome and cardiac arrhythmias.

A particularly alarming trend was the near-doubling of interacting regimens involving muscle relaxants, rising from 1.35% to 2.49%. Researchers link this to increasing rates of unintentional overdose deaths among older adults and are urging clinical guidelines to specifically address this class.

For health-conscious older adults and their clinicians, the practical message is clear: regularly auditing medication and supplement combinations — especially antidepressants and muscle relaxants — is essential. Deprescribing initiatives should be prioritized, and supplement use must be included in any interaction review, as it is often overlooked. Caveats include the cross-sectional design, which limits causal inference.