Older patients recovering from stroke and other cerebrovascular events who took six or more medications did not end their rehabilitation with less total functional improvement than patients on fewer drugs, according to a large observational study drawing on Japanese real-world claims data. Yet the pace of that recovery told a different story: patients whose medication burden persisted, grew, or was resolved during hospitalization all improved more slowly on a standard measure of daily living ability than those who stayed below the polypharmacy threshold throughout their stay.
The data came from the Japan Medical Data Center claims database, which contains anonymized inpatient medical and pharmacy claims from multiple healthcare institutions across Japan covering the period from 1 April 2014 to 31 January 2025. Patients who died during hospitalization, stayed 14 days or less, received anticancer agents, or lacked Functional Independence Measure (FIM) data at admission or discharge were excluded.
Polypharmacy was defined under Japan’s national health insurance system as the use of six or more medications, a threshold that also triggers a reimbursement incentive encouraging deprescribing during hospitalization. When counting drugs, the researchers excluded as-needed medications and topical agents such as ointments, but counted topical medications intended for systemic effects and used daily, such as isosorbide dinitrate patches. Combination drugs were counted as a single medication. Medication use at admission was based on drugs patients brought with them or prescriptions filled within seven days of admission, while discharge medication use was based on prescriptions written on the discharge date itself.
By comparing medication counts at admission and discharge, the team sorted patients into four groups. Nearly half of the cohort, 48.9 percent, had persistent polypharmacy, meaning six or more drugs at both admission and discharge. Another 14.8 percent developed incident polypharmacy during their stay, 8.8 percent achieved resolution of polypharmacy, and 27.6 percent remained below the threshold throughout hospitalization. Roughly 20 percent of all patients had hyper-polypharmacy, defined as ten or more medications, at both admission and discharge. The mean age of the cohort was 79 years, 96.3 percent were treated in non-public hospitals, and 57.3 percent were admitted to hospitals with fewer than 200 beds.
The primary outcome was FIM gain, the difference between FIM scores at discharge and admission. The FIM is an 18-item scale covering motor and cognitive domains including self-care, sphincter control, transfers, locomotion, communication, and social cognition, with total scores ranging from 18 to 126 points. The secondary outcomes were FIM efficiency, calculated as FIM gain divided by length of hospital stay, and length of stay itself. After multivariable linear regression adjusting for age, sex, body mass index, admission FIM score, smoking status, primary diagnosis, total hospital bed capacity, comorbidities such as hypertension, diabetes, dyslipidemia, heart failure, dementia, chronic kidney disease, and Parkinson’s disease, and admission period, polypharmacy status showed no significant association with FIM gain in either crude or adjusted models.
FIM efficiency, by contrast, was significantly lower in all three polypharmacy groups compared with the persistent non-polypharmacy group. The strongest negative association appeared in the incident polypharmacy group, with an adjusted coefficient of B = −0.114 and p < 0.001 in the model that also adjusted for persistent anticholinergic drug use, followed by the persistent polypharmacy group at B = −0.069. These associations held in sensitivity analyses stratified by pre- and post-2020 admission periods, and when polypharmacy was redefined by counting fixed-dose combination products according to their number of active ingredients, which reclassified only 1.6 percent of patients. Persistent anticholinergic drug use, defined as use recorded at both admission and discharge, was independently associated with lower FIM efficiency but not with FIM gain.
Length of hospital stay emerged as a critical piece of the puzzle. All three polypharmacy groups stayed significantly longer than the persistent non-polypharmacy group, with the largest adjusted difference in the incident polypharmacy group at 16.478 days, followed by 9.539 days for the persistent polypharmacy group and 9.533 days for the polypharmacy-resolved group, all statistically significant. Because length of stay sits in the denominator of FIM efficiency, the authors caution that the slower recovery pace may partly reflect longer hospitalization rather than a genuinely smaller amount of functional improvement. Patients who recover slowly may simply remain hospitalized longer while still ultimately reaching comparable levels of independence in daily activities.
The composition of newly added medications also proved revealing. Supplementary analyses of prescribing patterns showed that in the incident polypharmacy group, prescriptions of antipsychotics rose by roughly 10 percentage points and hypnotics and sedatives, including benzodiazepine receptor agonists, by roughly 15 percentage points between admission and discharge. Such centrally acting drugs are known to cause sedation, cognitive impairment, impaired balance, and increased fall risk, which could interfere with rehabilitation. However, the authors note that these prescriptions may equally have responded to delirium, behavioral symptoms, insomnia, pain, or difficulty participating in therapy during the hospitalization itself. Meanwhile, increases in antithrombotic, antihypertensive, antidiabetic, and lipid-lowering medications may have represented appropriate treatment of newly identified conditions or secondary prevention rather than harmful burden.
A subgroup analysis restricted to patients who already had polypharmacy at admission found that resolving it was not significantly associated with either FIM gain or FIM efficiency after adjustment, with an estimate of 0.345 for FIM gain and 0.015 for FIM efficiency when comparing the resolved group with the persistent polypharmacy group as reference. This suggests that deprescribing itself was not clearly linked to poorer recovery, and that the lower efficiency seen in the resolved group in the main analysis may reflect baseline differences or unmeasured factors. The timing of medication reduction was not recorded, so resolution occurring shortly before discharge may simply have had too little time to influence outcomes. The authors also point out that patients with polypharmacy at discharge had a higher prevalence of heart failure, a condition whose guideline-directed treatment often requires multiple drug classes, though the database lacked information on disease stage and severity.
The researchers emphasize several limitations. Detailed information on rehabilitation intensity and pharmacist-led medication optimization was unavailable, residual confounding and confounding by indication cannot be excluded because stroke severity and heart failure stage were not recorded, and patients without medication data at admission or discharge were excluded in ways that may have affected both the estimated prevalence of polypharmacy and the observed associations. Most fundamentally, because medication changes, functional outcomes, and length of stay were all assessed during the same hospitalization, the temporal direction of the associations cannot be established, and reverse causation remains possible. The authors conclude that polypharmacy status was not significantly associated with FIM gain, the primary outcome, but that all three polypharmacy groups showed lower FIM efficiency and longer hospital stays, and that these observational findings do not establish a causal relationship between changes in medication burden and functional recovery. They suggest that polypharmacy itself may nonetheless serve as a clinically meaningful marker of delayed functional recovery in post-acute rehabilitation, warranting attention to which drugs are being added and why.
Subject of Research: The association between polypharmacy and functional recovery in older patients undergoing cerebrovascular rehabilitation in Japan.
Article Title: Polypharmacy and Activities of Daily Living in Older Patients Undergoing Cerebrovascular Rehabilitation: An Observational Study Using Japanese Real-World Data
Article References: Ogawa, Y., Horii, T., Sakoh, M., & Mihara, K. (2026). Polypharmacy and Activities of Daily Living in Older Patients Undergoing Cerebrovascular Rehabilitation: An Observational Study Using Japanese Real-World Data. Drugs – Real World Outcomes. https://doi.org/10.1007/s40801-026-00589-z
Image Credits: AI Generated
DOI: 10.1007/s40801-026-00589-z
Keywords: polypharmacy, cerebrovascular rehabilitation, activities of daily living, Functional Independence Measure, older adults, convalescent rehabilitation ward, FIM efficiency, length of hospital stay, anticholinergic drugs, deprescribing, Japanese claims database, observational study
News Source: Cassandra Pierce. (October 28, 2026). Many Medications Tied to Slower Daily Living Recovery After Stroke Rehabilitation. Scienmag.
