ABSTRACT
Aim
International anticholinergic burden scales may incompletely capture country‐specific prescribing patterns and drug availability. Although the Japanese Anticholinergic Risk Scale (JARS) reflects Japanese clinical practice, its prognostic relevance in poststroke patients remains unclear. We examined whether JARS was associated with functional outcomes across activities of daily living (ADL), cognition, swallowing, nutrition, and muscle strength in older patients after stroke.
Methods
This retrospective cohort study enrolled consecutive patients aged ≥ 65 years undergoing poststroke rehabilitation. Anticholinergic burden was assessed using JARS at admission. The primary outcome was Functional Independence Measure (FIM)‐motor at discharge. Secondary outcomes were FIM‐cognitive, Food Intake Level Scale, Geriatric Nutritional Risk Index (GNRI), and handgrip strength at discharge. Multivariable linear regression was performed for total JARS and, as an exploratory analysis, psychotropic drug‐derived JARS, with adjustment for relevant covariates. p‐values were adjusted across the five outcomes using the Benjamini–Hochberg false discovery rate (FDR) procedure.
Results
Among 747 patients (median age, 80.0 years; 51% men), 484 (64.8%) had a JARS score of ≥ 1. Total JARS was not associated with discharge FIM‐motor score (B = 0.032, 95% CI −0.799 to 0.862), FIM‐cognitive score, FILS, or handgrip strength. However, higher total JARS was negatively associated with GNRI at discharge in the primary analysis (B = −0.603; 95% CI, −1.061 to −0.146; FDR‐adjusted p = 0.049) although this association did not remain statistically significant in the modified‐JARS sensitivity analysis (B = −0.666; 95% CI, −1.179 to −0.154; FDR‐adjusted p = 0.055). Psychotropic drug‐derived JARS was associated with lower FIM‐cognitive score (B = −0.800; 95% CI, −1.318 to −0.282; FDR‐adjusted p = 0.013), but not with the other outcomes after FDR adjustment.
Conclusions
JARS was not associated with discharge ADL, swallowing, or muscle strength. A modest association with nutritional risk was observed in the primary analysis, although this finding should be interpreted cautiously because statistical significance was not retained in the modified‐JARS sensitivity analysis. Psychotropic drug‐derived JARS was associated with cognitive level. These findings support domain‐ and drug‐class‐specific interpretation of anticholinergic burden.
Keywords: activities of daily living, cholinergic antagonists, nutritional status, rehabilitation, stroke
In older poststroke rehabilitation patients, total JARS was not associated with discharge ADL but was modestly associated with poorer nutritional status. Psychotropic drug‐derived JARS was associated with lower cognitive function in exploratory analysis, supporting domain‐ and drug‐class‐specific interpretation of anticholinergic burden.
1. Introduction
Anticholinergic burden is an important medication‐related risk in older adults. Anticholinergic effects can cause peripheral and central adverse events, including dry mouth, constipation, urinary retention, and somnolence, and have been linked to cognitive decline [1], delirium [2], falls [3], physical functional decline [4], and mortality [5]. Many drugs with anticholinergic properties are regarded as potentially inappropriate medications in older adults [6]. These drugs are heterogeneous: some are prescribed for their intended anticholinergic effects, whereas others have anticholinergic activity as an unintended pharmacologic property. In older adults, multimorbidity and polypharmacy can result in cumulative exposure to multiple anticholinergic drugs [7, 8]. Assessing anticholinergic burden is therefore important for identifying medication‐related risks in geriatric care.
Medication‐related problems may hinder functional recovery after stroke. Patients with stroke are increasingly older and often receive multiple medications for secondary prevention and for comorbidities such as hypertension, diabetes mellitus, atrial fibrillation, and dyslipidemia [9, 10]. Poststroke rehabilitation aims to restore premorbid living as far as possible by improving activities of daily living (ADL), walking ability, swallowing function, cognitive function, and other functional domains [11, 12]. Medication‐related symptoms, including somnolence, dizziness, and anorexia, may reduce participation in rehabilitation and interfere with oral intake [13, 14, 15]. Previous studies in stroke rehabilitation patients have linked polypharmacy, potentially inappropriate prescribing, anticholinergic burden, and sedative burden to poorer recovery of ADL, cognitive function, and swallowing function [10, 16, 17, 18, 19, 20]. These medication‐related problems are therefore potentially modifiable factors in poststroke rehabilitation.
The relationship between anticholinergic burden assessed with the Japanese Anticholinergic Risk Scale (JARS) [21] and rehabilitation outcomes after stroke remains unclear. Previous studies in stroke rehabilitation patients have reported associations between anticholinergic burden assessed using the Anticholinergic Risk Scale (ARS) and cognitive recovery [22], between the Drug Burden Index (DBI) and ADL or balance ability [23], and between increases in ARS scores and swallowing outcomes [24]. However, studies have used different anticholinergic burden scales, including the ARS, Anticholinergic Cognitive Burden scale, and DBI, and the target drugs and scoring methods differ among scales [25]. Scales developed outside Japan may not fully reflect drugs and prescribing patterns used in Japanese clinical practice. JARS was developed to address this issue by covering drugs used in Japan [21], but evidence is limited regarding its association with rehabilitation outcomes across multiple domains in poststroke rehabilitation patients.
This study examined the association between anticholinergic burden assessed with JARS and functional outcomes, with a primary focus on ADL, in older poststroke rehabilitation patients. By assessing multiple outcomes, we aimed to clarify whether JARS serves as a general prognostic marker for ADL or more specifically reflects medication‐related risks in domains such as nutrition and cognition.
2. Methods
2.1. Participants and Setting
This retrospective cohort study included patients with stroke admitted between January 2020 and June 2025 to a single hospital with a convalescent rehabilitation ward in Japan. The hospital accepts patients for convalescent rehabilitation after their condition has stabilized following acute stroke treatment at acute care hospitals. After admission, patients received rehabilitation according to their functional status and clinical condition, including physical therapy, occupational therapy, and speech‐language‐hearing therapy. Rehabilitation programs included range‐of‐motion exercises, resistance training, bed mobility, transfer and gait training, ADL training, swallowing therapy, and higher brain function training. Registered dietitians also provided nutritional assessment and management.
Patients admitted to the convalescent rehabilitation ward with a diagnosis of stroke were eligible. We excluded patients who died during hospitalization, were transferred to another acute care hospital or acute care ward during rehabilitation, or were younger than 65 years.
2.2. Data Collection
Patient information was retrospectively collected from electronic medical records. Demographic and clinical variables included age, sex, stroke type, date of stroke onset, premorbid modified Rankin Scale (mRS) [26], and Charlson Comorbidity Index (CCI) [27]. As indicators of stroke severity and functional status, Brunnstrom Recovery Stage (BRS) [28] and Functional Independence Measure (FIM) [29] scores at admission were assessed by physical and occupational therapists. Food Intake Level Scale (FILS) [30] at admission was assessed by speech‐language‐hearing therapists.
Nutritional risk was assessed using the Geriatric Nutritional Risk Index (GNRI) [31], calculated from height, body weight, body mass index (BMI), and serum albumin level at admission. The GNRI was calculated as follows: GNRI = 14.89 × serum albumin (g/dL) + 41.7 × actual body weight/ideal body weight. When actual body weight exceeded ideal body weight, the ratio of current body weight to ideal body weight was set to 1. Ideal body weight was calculated as height (m)2 × 22 kg/m2.
Handgrip strength was measured three times on the nonparetic side by occupational therapists using a Smedley‐type hand dynamometer, and the maximum value was used. When measurement was not feasible because of cognitive impairment or other reasons, handgrip strength was defined as 0 kg. As a sensitivity analysis, the analysis of discharge handgrip strength was repeated among patients whose handgrip strength was directly measurable at both admission and discharge.
Length of stay and rehabilitation units per day were also collected; one rehabilitation unit was defined as 20 min.
2.3. Assessment of Anticholinergic Burden
Medications prescribed at admission were collected from electronic medical records and medication reconciliation records completed by pharmacists. The number of medications was calculated from regularly prescribed oral medications at admission. As‐needed medications, topical agents, ophthalmic preparations, inhaled medications, and injectable medications were excluded from the medication count.
Anticholinergic burden was assessed using JARS [21]. In JARS, each drug is scored from 1 to 3 according to the strength of its anticholinergic activity. The total JARS score for each patient was calculated by summing the scores of all JARS‐listed medications prescribed at admission. Higher scores indicate greater anticholinergic burden. To assess changes in anticholinergic burden during hospitalization, JARS scores were also calculated at discharge.
Because JARS includes drugs from various therapeutic classes, and psychotropic medications may affect physical activity, cognitive function, and nutritional status [6, 32], psychotropic drug‐derived JARS was calculated separately. Psychotropic medications included antipsychotics, antidepressants, and benzodiazepine anxiolytics and hypnotics.
We also calculated a modified JARS score that excluded frequently prescribed low‐scoring medications used primarily to treat comorbid conditions. Specifically, the modified score excluded 1‐point medications that were prescribed to at least 5% of the study population, were used primarily for nonneuropsychiatric comorbidities, and had uncertain relevance to central anticholinergic exposure: lansoprazole, nifedipine, furosemide, metformin, and warfarin (shown in Table S1). A post hoc sensitivity analysis using the modified JARS score was conducted to assess whether the findings were driven by these frequently prescribed medications.
2.4. Outcome Measures
The primary outcome was FIM‐motor score at discharge. The FIM evaluates independence in ADL and consists of 13 motor items and 5 cognitive items [29]. The FIM‐motor score ranges from 13 to 91, with higher scores indicating greater independence.
Secondary outcomes were FIM‐cognitive score, FILS, GNRI, and handgrip strength at discharge. The FIM‐cognitive score ranges from 5 to 35, with higher scores indicating greater independence in cognitive domains. The FILS is a 10‐level scale for swallowing and oral intake status, with higher scores indicating better oral intake ability [30]. The GNRI was used as an indicator of nutritional risk, with lower values indicating higher risk [31]. Discharge outcomes were assessed as part of routine multidisciplinary care before discharge.
2.5. Statistical Analysis
Continuous and ordinal variables are presented as medians and interquartile ranges, and categorical variables as numbers and percentages. Baseline characteristics were compared between the JARS‐negative group (JARS score of 0) and the JARS‐positive group (JARS score of 1 or higher). Discharge outcomes were compared between the same groups. Between‐group comparisons were performed using the Mann–Whitney U test for continuous and ordinal variables and the chi‐square test or Fisher's exact test for categorical variables, as appropriate.
Multivariable linear regression analyses were performed to examine the association between JARS at admission and rehabilitation outcomes at discharge. In the primary analysis, JARS at admission was entered as a continuous variable, and FIM‐motor score, FIM‐cognitive score, FILS, GNRI, and handgrip strength at discharge were entered separately as dependent variables. Covariates were age, sex, days from stroke onset to admission, premorbid mRS, stroke type, CCI, number of medications at admission, and baseline values of the outcomes, namely FIM‐motor score, FIM‐cognitive score, FILS, GNRI, and handgrip strength at admission. These variables were selected based on previous studies and clinical relevance as factors potentially associated with rehabilitation outcomes after stroke. The same covariate set was used for all outcomes. As an exploratory analysis, we examined the association between psychotropic drug‐derived JARS and rehabilitation outcomes at discharge. This model included the same covariates as the primary analysis, with additional adjustment for nonpsychotropic drug‐derived JARS. Multicollinearity was assessed using variance inflation factors (VIFs), with values of ≥ 10 considered indicative of substantial multicollinearity. Results are presented as unstandardized regression coefficients (B), 95% confidence intervals (CIs), and p‐values.
All p‐values were two‐sided, and p < 0.05 was considered statistically significant. To account for multiple testing across the five outcome measures, p‐values for the associations of total JARS, psychotropic drug‐derived JARS and modified JARS with each outcome were adjusted separately using the Benjamini–Hochberg false discovery rate (FDR) procedure. Both unadjusted and FDR‐adjusted p‐values were reported, and an FDR‐adjusted p‐value of < 0.05 was considered statistically significant. As additional sensitivity analyses, the two associations that remained significant after FDR adjustment were re‐examined to account for changes in JARS during hospitalization. The analyses were repeated using JARS scores at discharge and, separately, among patients whose corresponding JARS scores remained unchanged between admission and discharge, using the same covariates as in the primary analyses. Analyses were performed using R version 4.5.2 (R Foundation for Statistical Computing, Vienna, Austria) in RStudio version 2026.01.1 + 403 (Posit Software, PBC, Boston, MA, USA).
3. Results
During the study period, 1038 patients with stroke were admitted to the convalescent rehabilitation ward. After excluding 8 patients who died during hospitalization and 72 who were transferred to another acute care hospital or acute care ward during rehabilitation, 958 patients were eligible. Of these, 211 patients younger than 65 years were excluded, leaving 747 patients aged 65 years or older for the final analysis (Figure 1).
FIGURE 1.
Flowchart of participant screening and inclusion criteria.
Baseline characteristics are shown in Table 1. The median age was 80.0 [73.0, 86.0] years, and 383 patients (51%) were men. Ischemic stroke was the most common stroke type (535 patients, 72%). Among the 747 patients, 263 had a JARS score of 0 and 484 had a score of 1 or higher; the prevalence of JARS positivity was 64.8%. JARS scores were concentrated between 0 and 2, with progressively fewer patients at higher scores (Figure 2). At discharge, the median total JARS score remained 1 [0–2]; however, individual scores changed in 393 patients (52.6%), with increases occurring more frequently than decreases (34.0% vs. 18.6%). The prevalence of total JARS positivity increased from 64.8% to 70.8%, and psychotropic drug‐derived JARS positivity increased from 14.3% to 19.3% (Table S2).
TABLE 1.
Baseline characteristics stratified by Japanese Anticholinergic Risk Scale positivity among poststroke patients.
| Variable | Overall (N = 747) | JARS 0 (N = 263) | JARS ≥ 1 (N = 484) | p |
|---|---|---|---|---|
| Age, years | 80.0 [73.0, 86.0] | 80.0 [73.0, 86.0] | 81.0 [73.0, 86.0] | 0.817 |
| Sex (men) | 383 (51%) | 130 (49%) | 253 (52%) | 0.458 |
| Stroke type | 0.032 | |||
| Ischemic stroke | 535 (72%) | 173 (66%) | 362 (75%) | |
| Intracerebral hemorrhage | 175 (23%) | 75 (29%) | 100 (21%) | |
| Subarachnoid hemorrhage | 37 (5.0%) | 15 (5.7%) | 22 (4.5%) | |
| Days from onset to admission | 16 [12, 23] | 15 [12, 22] | 17 [12, 25] | 0.014 |
| CCI | 2 [1, 3] | 2 [1, 3] | 3 [1, 3] | 0.015 |
| Premorbid mRS | 1 [0, 2] | 0 [0, 2] | 1 [0, 3] | < 0.001 |
| BRS upper limb | 5 [3, 6] | 5 [2, 6] | 5 [3, 6] | 0.688 |
| BRS fingers | 5 [2, 6] | 5 [2, 6] | 5 [3, 6] | 0.669 |
| BRS lower limb | 5 [3, 6] | 5 [3, 6] | 5 [3, 6] | 0.138 |
| FIM‐total | 60 [29, 87] | 60 [27, 90] | 60 [31, 84] | 0.992 |
| FIM‐motor | 40 [18, 63] | 39 [16, 65] | 41 [19, 62] | 0.856 |
| FIM‐cognitive | 19 [10, 25] | 19 [10, 26] | 18 [10, 25] | 0.656 |
| FILS | 8 [7, 10] | 8 [7, 10] | 8 [7, 10] | 0.580 |
| GNRI | 92.3 [85.3, 98.8] | 92.3 [84.6, 98.3] | 92.3 [86.3, 99.8] | 0.435 |
| BMI, kg/m2 | 21.8 [19.7, 24.1] | 21.5 [19.4, 24.1] | 21.9 [19.7, 24.3] | 0.319 |
| Handgrip strength in men, kg | 23.1 [14.4, 30.0] | 23.3 [16.1, 31.0] | 23.0 [13.6, 29.3] | 0.431 |
| Handgrip strength in women, kg | 11.7 [5.3, 16.3] | 12.0 [5.1, 17.1] | 11.5 [6.1, 16.3] | 0.980 |
| Number of medications | 6 [4, 8] | 5 [3, 6] | 7 [5, 9] | < 0.001 |
| JARS score | 1 [0, 2] | 0 [0, 0] | 1 [1, 2] | < 0.001 |
| Length of hospital stay, days | 87 [55, 129] | 88 [54, 126] | 86 [55, 130] | 0.744 |
| Rehabilitation, units/day | 8.1 [6.5, 8.6] | 8.2 [6.6, 8.6] | 8.1 [6.5, 8.6] | 0.332 |
Abbreviations: BRS, Brunnstrom Recovery Stage; CCI, Charlson Comorbidity Index; FILS, Food Intake Level Scale; FIM, Functional Independence Measure; GNRI, Geriatric Nutritional Risk Index; JARS, Japanese Anticholinergic Risk Scale; mRS, modified Rankin Scale.
FIGURE 2.
Distribution of JARS scores at admission. The vertical axis indicates the percentage of patients. JARS, Japanese Anticholinergic Risk Scale.
Age and sex did not differ significantly between the JARS‐positive and JARS‐negative groups. Stroke type differed between the groups with a higher proportion of ischemic stroke and a lower proportion of intracerebral hemorrhage in the JARS‐positive group. The JARS‐positive group had higher CCI and premorbid mRS, suggesting greater comorbidity burden and poorer premorbid functional status. In contrast, BRS, FIM‐total score, FIM‐motor score, FIM‐cognitive score, FILS, GNRI, BMI, sex‐specific handgrip strength, length of stay, and rehabilitation units per day did not differ significantly between groups. The median number of medications at admission was 6 [4, 8] in the overall cohort. The JARS‐positive group used more medications than the JARS‐negative group (7 [5, 9] vs. 5 [3, 6]; p < 0.001).
Unadjusted comparisons of discharge outcomes are shown in Table 2. None of the discharge outcomes differed significantly between the JARS‐positive and JARS‐negative groups.
TABLE 2.
Discharge rehabilitation outcomes stratified by Japanese Anticholinergic Risk Scale positivity among poststroke patients.
| Variable | JARS 0 (N = 263) | JARS ≥ 1 (N = 484) | p |
|---|---|---|---|
| FIM‐motor | 75 [34, 87] | 73 [37, 86] | 0.887 |
| FIM‐cognitive | 26 [16, 32] | 24 [15, 31] | 0.203 |
| FILS | 10 [8, 10] | 10 [8, 10] | 0.943 |
| GNRI | 92.7 [86.7, 98.3] | 93.3 [87.6, 98.3] | 0.875 |
| Handgrip strength in men, kg | 26.3 [18.0, 32.2] | 24.5 [16.5, 30.4] | 0.278 |
| Handgrip strength in women, kg | 13.4 [8.0, 18.0] | 13.8 [8.0, 17.5] | 0.937 |
Abbreviations: FILS, Food Intake Level Scale; FIM, Functional Independence Measure; GNRI, Geriatric Nutritional Risk Index; JARS, Japanese Anticholinergic Risk Scale.
Table 3a shows the multivariable linear regression analyses examining the association between JARS at admission and rehabilitation outcomes at discharge. All VIFs were below 10 in all multivariable models, indicating no evidence of substantial multicollinearity. JARS at admission was not significantly associated with the primary outcome, FIM‐motor score at discharge (B = 0.032; 95% CI, −0.799 to 0.862; FDR‐adjusted p = 0.941). Among secondary outcomes, JARS at admission was negatively associated with GNRI at discharge (B = −0.603; 95% CI, −1.061 to −0.146; p = 0.010; FDR‐adjusted p = 0.049), but not with FIM‐cognitive score, FILS, or handgrip strength at discharge. In post hoc sensitivity analyses, excluding five frequently prescribed 1‐point medications yielded a similar estimate for GNRI, although the association did not remain significant after FDR adjustment (B = −0.666; 95% CI, −1.179 to −0.154; FDR‐adjusted p = 0.055; Table S3). In analyses restricted to patients with measurable handgrip strength at both admission and discharge, neither total JARS at admission (n = 581; B = −0.076; 95% CI, −0.309 to 0.157; p = 0.521) nor psychotropic drug‐derived JARS at admission (n = 581; B = 0.015; 95% CI, −0.400 to 0.430; p = 0.944) was associated with handgrip strength at discharge.
TABLE 3.
Multivariable analysis of the association between Japanese Anticholinergic Risk Scale at admission and rehabilitation outcomes.
| Outcomes | B | 95% CI | p | FDR‐adjusted p‐value |
|---|---|---|---|---|
| (a) Total Japanese Anticholinergic Risk Scale score | ||||
| FIM‐motor at discharge | 0.032 | −0.799 to 0.862 | 0.941 | 0.941 |
| FIM‐cognitive at discharge | −0.243 | −0.531 to 0.045 | 0.098 | 0.163 |
| FILS at discharge | −0.013 | −0.103 to 0.077 | 0.778 | 0.941 |
| GNRI at discharge | −0.603 | −1.061 to −0.146 | 0.010 | 0.049 |
| Handgrip strength at discharge | −0.311 | −0.653 to 0.032 | 0.075 | 0.163 |
| (b) Psychotropic drug‐derived Japanese Anticholinergic Risk Scale score | ||||
| FIM‐motor at discharge | 0.267 | −1.233 to 1.767 | 0.727 | 0.727 |
| FIM‐cognitive at discharge | −0.800 | −1.318 to −0.282 | 0.003 | 0.013 |
| FILS at discharge | −0.034 | −0.197 to 0.129 | 0.680 | 0.727 |
| GNRI at discharge | −0.934 | −1.758 to −0.110 | 0.026 | 0.066 |
| Handgrip strength at discharge | −0.163 | −0.779 to 0.452 | 0.602 | 0.727 |
Note: Models were adjusted for age, sex, days from stroke onset to admission, premorbid modified Rankin Scale, stroke type, Charlson Comorbidity Index, admission FIM‐motor, admission FIM‐cognitive, admission FILS, admission GNRI, admission handgrip strength, and number of medications, and JARS derived from nonpsychotropic drugs. FDR‐adjusted p‐values were calculated using the Benjamini–Hochberg procedure across the five outcome measures.
Abbreviations: FDR, false discovery rate; FILS, Food Intake Level Scale; FIM, Functional Independence Measure; GNRI, geriatric nutritional risk index.
The exploratory analysis of psychotropic drug‐derived JARS is shown in Table 3b. After adjustment for the same covariates as in the primary analysis and for nonpsychotropic drug‐derived JARS, psychotropic drug‐derived JARS was negatively associated with FIM‐cognitive score at discharge (B = −0.800; 95% CI, −1.318 to −0.282; p = 0.003; FDR‐adjusted p = 0.013). Although psychotropic drug‐derived JARS was associated with lower discharge GNRI before FDR adjustment (B = −0.934; 95% CI, −1.758 to −0.110; p = 0.026), this association did not remain statistically significant after FDR adjustment (FDR‐adjusted p = 0.066). No significant associations were observed with FIM‐motor score, FILS, or handgrip strength at discharge.
Additional sensitivity analyses were performed to account for changes in JARS during hospitalization. When discharge JARS scores were used as exposures, total JARS was not significantly associated with discharge GNRI (B = −0.312; 95% CI, −0.658 to 0.035; p = 0.078), whereas psychotropic‐derived JARS remained associated with discharge cognitive FIM score (B = −1.118; 95% CI, −1.550 to −0.685; p < 0.001). Among patients whose corresponding JARS scores remained unchanged between admission and discharge, admission total JARS remained associated with discharge GNRI (B = −0.961; 95% CI, −1.752 to −0.170; p = 0.018), and psychotropic‐derived JARS remained associated with discharge cognitive FIM score (B = −0.997; 95% CI, −1.864 to −0.129; p = 0.024) (Table S4).
4. Discussion
This study examined the association between anticholinergic burden at admission, assessed with JARS, and functional outcomes in poststroke rehabilitation patients. Three findings were obtained. First, JARS was not associated with ADL at discharge. Second, total JARS was associated with nutritional risk at discharge. Third, in an exploratory analysis, psychotropic drug‐derived JARS was associated with cognitive level at discharge.
JARS was not associated with ADL at discharge. This finding differs from previous reports linking anticholinergic burden to poorer ADL recovery after stroke [23]. The drug profile of the present cohort may partly explain this discrepancy. Most JARS‐positive patients in the present cohort had relatively low scores, and frequently prescribed JARS‐listed medications included lansoprazole, nifedipine, furosemide, and metformin (Table S1). JARS was developed by integrating scores from existing anticholinergic risk scales and using expert consensus for medications whose scores could not be determined by the predefined algorithms [21]. Notably, lansoprazole is not a conventional muscarinic receptor antagonist; however, experimental studies have shown that proton pump inhibitors, including lansoprazole, can inhibit choline acetyltransferase, the enzyme responsible for acetylcholine synthesis, suggesting a potential anticholinergic‐like mechanism [33, 34]. Nevertheless, the clinical relevance of this mechanism remains uncertain. Thus, total JARS may have reflected comorbidity‐related prescribing and overall treatment burden as well as exposure to centrally acting anticholinergic medications. The null association with ADL persisted in the post hoc sensitivity analysis using modified JARS. Because ADL recovery is strongly influenced by neurologic severity, premorbid and admission ADL, cognitive function, and nutritional status [35, 36, 37], the independent contribution of JARS may have been small in this cohort. These findings should therefore be interpreted in the context of the relatively low JARS scores and the heterogeneous drug composition of the scale.
JARS was associated with nutritional risk at discharge but not with cognitive function, oral intake status, or handgrip strength. This association with GNRI supports previous reports linking anticholinergic burden to malnutrition risk [38, 39]. Dry mouth, reduced gastrointestinal motility, and central nervous system depression can affect appetite, dietary intake, nutritional status, and body weight. However, some frequently prescribed JARS‐listed medications may also affect GNRI through nonanticholinergic mechanisms, such as gastrointestinal symptoms or changes in body weight and fluid status. Thus, the association between JARS and GNRI may reflect both anticholinergic effects and other medication‐ or disease‐related factors. Psychotropic drug‐derived JARS also showed an inverse association with GNRI before FDR adjustment, suggesting that anticholinergic activity may have contributed, although this association did not remain statistically significant after correction for multiple testing. In addition, GNRI includes serum albumin and is influenced by inflammation and disease burden [40]. The established GNRI risk categories are > 98 for no risk, 92 to ≤ 98 for low risk, 82 to < 92 for moderate risk, and < 82 for major risk [31]. The observed 0.6‐point difference was small relative to the widths of these risk categories. Given the cohort median GNRI of 92.3 and the absence of an established minimal clinically important difference for GNRI in poststroke rehabilitation, the clinical relevance of this association remains uncertain. The estimate remained similar after excluding five frequently prescribed 1‐point medications but did not remain significant after FDR adjustment. In additional sensitivity analyses, the association was observed among patients with unchanged total JARS scores but was not significant when total JARS at discharge was used, suggesting that the association with GNRI was less robust to the timing of exposure assessment. The present findings should therefore not be interpreted as evidence of a direct causal effect of anticholinergic burden on nutritional deterioration.
Total JARS was not significantly associated with cognitive level, swallowing level, or handgrip strength. These outcomes are influenced by diverse factors, including stroke lesion characteristics, paresis, aphasia, stroke‐related dysphagia, muscle mass, and physical activity [41, 42, 43], and total JARS alone may not capture these pathways. In contrast, psychotropic drug‐derived JARS was associated with cognitive function at discharge. Psychotropic medications may affect arousal, attention, memory, and executive function [44], making their anticholinergic burden more relevant to cognitive functional status than total JARS. This association remained consistent when psychotropic drug‐derived JARS at discharge was used and when the analysis was restricted to patients with unchanged psychotropic drug‐derived JARS scores during hospitalization. However, this exploratory association may have been influenced by confounding by indication or reverse causation. Previous observational studies have attempted to reduce this bias through lagged exposure analyses [45], drug‐class‐specific analyses [46], and adjustment for baseline cognition and related clinical conditions [47, 48]; however, residual confounding by indication remains difficult to eliminate. Psychotropic medications may have been prescribed for delirium, behavioral symptoms, sleep disturbance, or preexisting cognitive impairment, which may themselves predict poorer cognitive outcomes. Although we adjusted for cognitive FIM score at admission, detailed information on the indications for psychotropic medications and symptom severity was unavailable. Psychotropic drug‐derived JARS may therefore reflect both pharmacological exposure and the conditions prompting treatment.
Taken together, these findings suggest potential domain‐specific associations of JARS, particularly with nutrition and cognition, rather than a general association with ADL. A strength of this study is that it evaluated JARS in relation to several rehabilitation outcomes in a real‐world cohort of poststroke rehabilitation patients. This broad coverage is useful for real‐world assessment, but it also means that the total score aggregates drugs from heterogeneous therapeutic classes with different clinical implications. In stroke rehabilitation, ADL is strongly shaped by disease‐ and treatment‐related factors, including lesion characteristics, paresis, premorbid function, ADL at admission, cognitive function, nutritional status, and rehabilitation intensity [11, 35, 36, 37, 49], and it can change substantially during the rehabilitation course. Anticholinergic burden alone is therefore unlikely to explain overall ADL recovery. JARS may therefore be more informative when interpreted by outcome domain and drug class rather than as a comprehensive prognostic indicator of ADL recovery. Future studies should examine JARS by drug class and pharmacologic profile to clarify which types of anticholinergic burden are most relevant to specific rehabilitation outcomes.
This study has several limitations. First, the retrospective observational design precludes causal inference. In particular, the association between psychotropic drug‐derived JARS and FIM‐cognitive may have been influenced by confounding by indication, as detailed information on the indications for psychotropic medications and the severity of the underlying symptoms was unavailable. Second, the primary exposure was JARS assessed at admission, which may not have represented anticholinergic burden throughout hospitalization. Although the median total JARS score was unchanged at discharge, individual scores changed in 52.6% of patients. In addition, medications used before stroke onset could not be distinguished from those newly initiated during acute care, and dose, duration of use, adherence, and pharmacological tolerance were not fully captured. These limitations may have resulted in exposure misclassification and influenced the observed associations. Third, JARS was first published in 2024, and evidence regarding its reliability, construct validity, and external validity remains limited compared with more established anticholinergic burden scales. Therefore, the present findings should be interpreted as preliminary outcome‐based evidence, and further validation in independent and multicenter populations is required. Finally, this was a single‐center study conducted in a Japanese convalescent rehabilitation ward. Facility‐specific prescribing practices, medication review processes, patient characteristics, and the relatively high rehabilitation intensity in this cohort may have influenced the observed associations. Therefore, the findings may not be directly generalizable to other convalescent rehabilitation wards in Japan with different prescribing patterns, medication management systems, or rehabilitation resources, and multicenter studies are needed to confirm their external validity.
5. Conclusion
In poststroke patients, anticholinergic burden assessed with JARS was not associated with discharge ADL. Total JARS was modestly associated with nutritional risk in the primary analysis; however, statistical significance was not retained in the modified‐JARS sensitivity analysis, warranting cautious interpretation. Psychotropic drug‐derived burden was associated with poorer cognitive status. The clinical implications of anticholinergic burden may therefore differ by outcome domain and score composition.
Author Contributions
Ayaka Matsumoto: conceptualization, data curation, formal analysis, investigation, methodology, resources, supervision, validation, visualization, writing – original draft. Yoshihiro Yoshimura: methodology, project administration, resources, supervision, writing – review and editing.
Funding
The authors have nothing to report.
Ethics Statement
This study was conducted in accordance with the Declaration of Helsinki and the Ethical Guidelines for Medical and Biological Research Involving Human Subjects in Japan. The study protocol was approved by the ethics committee of the participating institution (approval ID: 2025‐32).
Consent
Because this was a retrospective observational study using existing medical records, written informed consent was not obtained from individual participants. Information about the study was disclosed through an opt‐out approach, and participants were given the opportunity to refuse participation.
Conflicts of Interest
The authors declare no conflicts of interest.
Supporting information
Table S1: JARS‐listed medications prescribed in patients at admission.
Table S2: JARS scores at admission and discharge and changes during hospitalization.
Table S3: Sensitivity analysis of the associations between modified JARS score and discharge outcomes.
Table S4: Sensitivity analyses addressing changes in JARS during hospitalization.
Acknowledgments
The authors would like to thank the staff of the Department of Rehabilitation and the Department of Nursing for their clinical assessments and patient care. We also thank the staff of the Department of Pharmacy for their cooperation in providing medication‐related information and all staff members who assisted with data collection. During the preparation of this work, the authors used ChatGPT to improve the readability and language of the manuscript. The authors reviewed and edited the output as needed and take full responsibility for the content of the published article.
Data Availability Statement
The datasets generated and/or analyzed during the current study are not publicly available because they contain patient clinical information. De‐identified data may be available from the corresponding author upon reasonable request and with permission from the relevant institutional authority.
References
- 1. Mancini V., Latreche C., Fanshawe J. B., et al., “Anticholinergic Burden and Cognitive Function in Psychosis: A Systematic Review and Meta‐Analysis,” American Journal of Psychiatry 182 (2025): 349–359. [DOI] [PubMed] [Google Scholar]
- 2. Egberts A., Moreno‐Gonzalez R., Alan H., Ziere G., and Mattace‐Raso F. U. S., “Anticholinergic Drug Burden and Delirium: A Systematic Review,” Journal of the American Medical Directors Association 22 (2021): 65–73.e4. [DOI] [PubMed] [Google Scholar]
- 3. Stewart C., Taylor‐Rowan M., Soiza R. L., Quinn T. J., Loke Y. K., and Myint P. K., “Anticholinergic Burden Measures and Older People's Falls Risk: A Systematic Prognostic Review,” Therapeutic Advances in Drug Safety 12 (2021): 20420986211016645. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Stewart C., Yrjana K., Kishor M., et al., “Anticholinergic Burden Measures Predict Older People's Physical Function and Quality of Life: A Systematic Review,” Journal of the American Medical Directors Association 22 (2021): 56–64. [DOI] [PubMed] [Google Scholar]
- 5. Phutietsile G. O., Fotaki N., Jamieson H. A., and Nishtala P. S., “The Association Between Anticholinergic Burden and Mobility: A Systematic Review and Meta‐Analyses,” BMC Geriatrics 23 (2023): 161. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. By the 2023 American Geriatrics Society Beers Criteria Update Expert Panel , “American Geriatrics Society 2023 Updated AGS Beers Criteria for Potentially Inappropriate Medication Use in Older Adults,” Journal of the American Geriatrics Society 71 (2023): 2052–2081. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Al Rihani S. B., Deodhar M., Darakjian L. I., et al., “Quantifying Anticholinergic Burden and Sedative Load in Older Adults With Polypharmacy: A Systematic Review of Risk Scales and Models,” Drugs & Aging 38 (2021): 977–994. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Ruxton K., Woodman R. J., and Mangoni A. A., “Drugs With Anticholinergic Effects and Cognitive Impairment, Falls and All‐Cause Mortality in Older Adults: A Systematic Review and Meta‐Analysis,” British Journal of Clinical Pharmacology 80 (2015): 209–220. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Miyamoto S., Ogasawara K., Kuroda S., et al., “Japan Stroke Society Guideline 2021 for the Treatment of Stroke,” International Journal of Stroke: Official Journal of the International Stroke Society 17 (2022): 1039–1049. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Matsumoto A., Yoshimura Y., Nagano F., et al., “Polypharmacy and Potentially Inappropriate Medications in Stroke Rehabilitation: Prevalence and Association With Outcomes,” International Journal of Clinical Pharmacy 44 (2022): 749–761. [DOI] [PubMed] [Google Scholar]
- 11. Gittler M. and Davis A. M., “Guidelines for Adult Stroke Rehabilitation and Recovery,” JAMA 319 (2018): 820–821. [DOI] [PubMed] [Google Scholar]
- 12. Yoshimura Y., Yasumoto Y., Matsumoto T., et al., “Clinical Practice Guideline for Rehabilitation Nutrition 2026: Executive Summary of the Japanese Association of Rehabilitation Nutrition Guideline 2026—A Secondary Publication,” JMA Journal 9 (2026): 1360–1378. [Google Scholar]
- 13. Yoshimura Y. and Matsumoto A., “Pharmacotherapy in Geriatric Rehabilitation: A Paradigm Shift From Disease Management to Functional Enhancement,” Progress in Rehabilitation Medicine 10 (2025): 20250022. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Clarke C. L. and Witham M. D., “The Effects of Medication on Activity and Rehabilitation of Older People – Opportunities and Risks,” Rehabilitation Process and Outcome 6 (2017): 1179572717711433. [Google Scholar]
- 15. Geller A. I., Nopkhun W., Dows‐Martinez M. N., and Strasser D. C., “Polypharmacy and the Role of Physical Medicine and Rehabilitation,” PM & R: The Journal of Injury, Function, and Rehabilitation 4 (2012): 198–219. [DOI] [PubMed] [Google Scholar]
- 16. Matsumoto A., Yoshimura Y., Nagano F., et al., “Polypharmacy and Its Association With Dysphagia and Malnutrition Among Stroke Patients With Sarcopenia,” Nutrients 14 (2022): 4251. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Matsumoto A., Yoshimura Y., Nagano F., et al., “Potentially Inappropriate Medications Are Negatively Associated With Functional Recovery in Patients With Sarcopenia After Stroke,” Aging Clinical and Experimental Research 34 (2022): 2845–2855. [DOI] [PubMed] [Google Scholar]
- 18. Matsumoto A., Yoshimura Y., Shimazu S., et al., “Association of Polypharmacy at Hospital Discharge With Nutritional Intake, Muscle Strength, and Activities of Daily Living Among Older Patients Undergoing Convalescent Rehabilitation After Stroke,” Japanese Journal of Comprehensive Rehabilitation Science 13 (2022): 41–48. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Kose E., Maruyama R., Okazoe S., and Hayashi H., “Impact of Polypharmacy on the Rehabilitation Outcome of Japanese Stroke Patients in the Convalescent Rehabilitation Ward,” Journal of Aging Research 2016 (2016): 7957825. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Ogawa Y., Sakoh M., Nibe F., and Mihara K., “Polypharmacy and Functional Decline in Activities of Daily Living in Older Rehabilitation Inpatients: A Retrospective Observational Study,” BMC Geriatrics 25 (2025): 1028. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Mizokami F., Mizuno T., Taguchi R., et al., “Development of the Japanese Anticholinergic Risk Scale: English Translation of the Japanese Article,” Geriatrics & Gerontology International 25 (2024): 5–13, 10.1111/ggi.15001. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Kose E., Hirai T., and Seki T., “Psychotropic Drug Use and Cognitive Rehabilitation Practice for Elderly Patients,” International Journal of Clinical Pharmacy 40 (2018): 1292–1299. [DOI] [PubMed] [Google Scholar]
- 23. Ogawa Y., Nibe F., Ogawa R., and Sakoh M., “Anticholinergic and Sedative Drug Burden and Functional Recovery After Cerebrovascular Accident: A Retrospective Descriptive Study,” Progress in Rehabilitation Medicine 5 (2020): n/a. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Kose E., Hirai T., Seki T., Okudaira M., and Yasuno N., “Anticholinergic Load Is Associated With Swallowing Dysfunction in Convalescent Older Patients After a Stroke,” Nutrients 14 (2022): 2121. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Matsumoto A., Yoshimura Y., Nagano F., et al., “Exploring the Impact of Anticholinergic Burden on Urinary Independence: Insights From a Post‐Stroke Cohort of Older Adults,” International Journal of Clinical Pharmacy 46 (2024): 910–917. [DOI] [PubMed] [Google Scholar]
- 26. Banks J. L. and Marotta C. A., “Outcomes Validity and Reliability of the Modified Rankin Scale: Implications for Stroke Clinical Trials: A Literature Review and Synthesis,” Stroke 38 (2007): 1091–1096. [DOI] [PubMed] [Google Scholar]
- 27. Charlson M. E., Pompei P., Ales K. L., and MacKenzie C. R., “A New Method of Classifying Prognostic Comorbidity in Longitudinal Studies: Development and Validation,” Journal of Chronic Diseases 40 (1987): 373–383. [DOI] [PubMed] [Google Scholar]
- 28. Brunnstrom S., “Motor Testing Procedures in Hemiplegia: Based on Sequential Recovery Stages,” Physical Therapy 46 (1966): 357–375. [DOI] [PubMed] [Google Scholar]
- 29. Ottenbacher K. J., Hsu Y., Granger C. V., and Fiedler R. C., “The Reliability of the Functional Independence Measure: A Quantitative Review,” Archives of Physical Medicine and Rehabilitation 77 (1996): 1226–1232. [DOI] [PubMed] [Google Scholar]
- 30. Kunieda K., Ohno T., Fujishima I., Hojo K., and Morita T., “Reliability and Validity of a Tool to Measure the Severity of Dysphagia: The Food Intake LEVEL Scale,” Journal of Pain and Symptom Management 46 (2013): 201–206. [DOI] [PubMed] [Google Scholar]
- 31. Bouillanne O., Morineau G., Dupont C., et al., “Geriatric Nutritional Risk Index: A New Index for Evaluating At‐Risk Elderly Medical Patients,” American Journal of Clinical Nutrition 82 (2005): 777–783. [DOI] [PubMed] [Google Scholar]
- 32. Yoshimura Y., Matsumoto A., and Momosaki R., “Pharmacotherapy and the Role of Pharmacists in Rehabilitation Medicine,” Progress in Rehabilitation Medicine 7 (2022): 20220025. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Kumar R., Kumar A., Nordberg A., Långström B., and Darreh‐Shori T., “Proton Pump Inhibitors Act With Unprecedented Potencies as Inhibitors of the Acetylcholine Biosynthesizing Enzyme‐A Plausible Missing Link for Their Association With Incidence of Dementia,” Alzheimer's & Dementia: The Journal of the Alzheimer's Association 16 (2020): 1031–1042. [DOI] [PubMed] [Google Scholar]
- 34. Baidya A. T. K., Das B., Devi B., et al., “Mechanistic Insight Into the Inhibition of Choline Acetyltransferase by Proton Pump Inhibitors,” ACS Chemical Neuroscience 14 (2023): 749–765. [DOI] [PubMed] [Google Scholar]
- 35. Brown A. W., Therneau T. M., Schultz B. A., Niewczyk P. M., and Granger C. V., “Measure of Functional Independence Dominates Discharge Outcome Prediction After Inpatient Rehabilitation for Stroke,” Stroke 46 (2015): 1038–1044. [DOI] [PubMed] [Google Scholar]
- 36. Kobayashi R. and Kobayashi N., “Performance of a Prediction Method for Activities of Daily Living Scores Using Influence Coefficients in Patients With Stroke,” Frontiers in Neurology 15 (2024): 1419405. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37. Lee Y.‐C. and Chiu E.‐C., “Nutritional Status as a Predictor of Comprehensive Activities of Daily Living Function and Quality of Life in Patients With Stroke,” NeuroRehabilitation 48 (2021): 337–343. [DOI] [PubMed] [Google Scholar]
- 38. Naharci M. I., Katipoglu B., and Tasci I., “Association of Anticholinergic Burden With Undernutrition in Older Adults: A Cross‐Sectional Study,” Nutrition in Clinical Practice: Official Publication of the American Society for Parenteral and Enteral Nutrition 37 (2022): 1215–1224. [DOI] [PubMed] [Google Scholar]
- 39. Ates Bulut E., Erken N., Kaya D., Dost F. S., and Isik A. T., “An Increased Anticholinergic Drug Burden Index Score Negatively Affect Nutritional Status in Older Patients Without Dementia,” Frontiers in Nutrition 9 (2022): 789986. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40. Evans D. C., Corkins M. R., Malone A., et al., “The Use of Visceral Proteins as Nutrition Markers: An ASPEN Position Paper,” Nutrition in Clinical Practice 36 (2021): 22–28. [DOI] [PubMed] [Google Scholar]
- 41. Lee B. and Pyun S.‐B., “Characteristics of Cognitive Impairment in Patients With Post‐Stroke Aphasia,” Annals of Rehabilitation Medicine 38 (2014): 759–765. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42. Jones C. A., Colletti C. M., and Ding M.‐C., “Post‐Stroke Dysphagia: Recent Insights and Unanswered Questions,” Current Neurology and Neuroscience Reports 20 (2020): 61. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43. Yoshimura Y., Wakabayashi H., Bise T., et al., “Sarcopenia Is Associated With Worse Recovery of Physical Function and Dysphagia and a Lower Rate of Home Discharge in Japanese Hospitalized Adults Undergoing Convalescent Rehabilitation,” Nutrition 61 (2019): 111–118. [DOI] [PubMed] [Google Scholar]
- 44. Chandramouleeshwaran S., Khan W. U., Inglis F., and Rajji T. K., “Impact of Psychotropic Medications on Cognition Among Older Adults: A Systematic Review,” International Psychogeriatrics 36 (2024): 1110–1127. [DOI] [PubMed] [Google Scholar]
- 45. Gray S. L., Anderson M. L., Dublin S., et al., “Cumulative Use of Strong Anticholinergics and Incident Dementia: A Prospective Cohort Study,” JAMA Internal Medicine 175 (2015): 401–407. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46. Richardson K., Fox C., Maidment I., et al., “Anticholinergic Drugs and Risk of Dementia: Case‐Control Study,” BMJ 361 (2018): k1315. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47. Lockery J. E., Broder J. C., Ryan J., et al., “A Cohort Study of Anticholinergic Medication Burden and Incident Dementia and Stroke in Older Adults,” Journal of General Internal Medicine 36 (2021): 1629–1637. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48. Taylor‐Rowan M., Kraia O., Kolliopoulou C., et al., “Anticholinergic Burden for Prediction of Cognitive Decline or Neuropsychiatric Symptoms in Older Adults With Mild Cognitive Impairment or Dementia,” Cochrane Database of Systematic Reviews 8 (2022): CD015196. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 49. Ren H., Liu Y., Zhao M., et al., “Stroke: Epidemiology, Risk Factors, Signaling Pathways, and Clinical Management,” MedComm 6 (2025): e70558. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Table S1: JARS‐listed medications prescribed in patients at admission.
Table S2: JARS scores at admission and discharge and changes during hospitalization.
Table S3: Sensitivity analysis of the associations between modified JARS score and discharge outcomes.
Table S4: Sensitivity analyses addressing changes in JARS during hospitalization.
Data Availability Statement
The datasets generated and/or analyzed during the current study are not publicly available because they contain patient clinical information. De‐identified data may be available from the corresponding author upon reasonable request and with permission from the relevant institutional authority.