1Institute of Population Health Sciences, National Health Research Institutes, Miaoli County, Taiwan
2Department of Psychiatry, National Taiwan University Hospital, Taipei, Taiwan
3Department of Psychiatry, College of Medicine, National Taiwan University, Taipei, Taiwan
4Department of Biomedical Sciences and Engineering, National Central University, Taoyuan City, Taiwan
5Department of Physical Medicine and Rehabilitation, College of Medicine, National Taiwan University, Taipei, Taiwan
6Department of Physical Medicine and Rehabilitation, National Taiwan University Hospital Hsin-Chu Branch, Hsinchu, Taiwan
7Department of Physical Medicine and Rehabilitation, National Taiwan University Hospital, Taipei, Taiwan
8Department of Medicine, College of Medicine, National Taiwan University, Taipei, Taiwan
Correspondence: Huey-Wen Liang Department of Physical Medicine and Rehabilitation, College of Medicine, National Taiwan University, No. 1, Sec. 1, Ren-Ai Rd., Zhongzheng Dist., Taipei 100233, Taiwan. Tel: +886-2-2312-3456 E-mail: lianghw@ntu.edu.tw
• Received: November 27, 2025 • Revised: May 28, 2026 • Accepted: June 26, 2026
This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (https://creativecommons.org/licenses/by-nc/4.0/) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.
To determine whether the stability of daily physical activity, quantified by interdaily stability (IS), is associated with activities of daily living (ADL) recovery during inpatient neurorehabilitation, independent of overall activity level, we examined the association between physical activity stability and ADL recovery.
Methods
This prospective study recruited 161 patients with stroke or spinal cord injury (mean age: 62.4±14.8 years; 42% female) from two hospitals. Physical activity was assessed using actigraphy for seven days after admission. IS represented activity regularity, and M10 the total activity volume during the most active 10-hour period. ADL was evaluated using the Barthel Index (BI) (baseline BI=32.5±22.9; discharge BI=61.1±28.6). Multiple linear regression analyses examined IS, M10, and ADL recovery, adjusting for baseline BI, age, sex, hospital stay, days from onset to rehabilitation admission, and diagnosis group.
Results
Higher IS during the first week after admission was significantly associated with better ADL recovery at discharge after controlling for all covariates (B=30.86, p=0.005). In contrast, total physical activity (M10) was not significantly associated with ADL recovery (B=-0.11, p=0.313). Younger age, lower baseline BI, longer length of stay, and shorter onset-to-admission time were also associated with greater ADL recovery.
Conclusion
Greater activity stability was associated with greater ADL improvement in patients with stroke, but not in those with spinal cord injury. IS may be a clinically relevant behavioral marker during inpatient rehabilitation, although its causal role remains to be established.
Neurological disorders are major global contributors to disability and mortality, frequently resulting in persistent impairments and functional limitations, including weakness, aphasia, dysphagia, cognitive decline, depression, and reduced independence in activities of daily living (ADL) [1,2]. Rehabilitation plays a critical role in the functional recovery of individuals with neurological disorders. For example, numerous studies have demonstrated that higher levels of physical activity are associated with lower mortality, better motor function, cardiovascular health, and greater ADL independence in community-dwelling individuals after stroke [3,4]. Nonetheless, the importance of physical activity during inpatient rehabilitation has been explored in only a few studies with small sample sizes [5]. Repetition, intensity, and progression during neurorehabilitation are essential components for promoting neuroplasticity, muscle strength, and mobility [6-8], as reflected by improved ADL independence and increased activity levels. Beyond total activity levels, emerging evidence suggests that the regularity and stability of daily activity patterns may also play a critical role in recovery outcomes [9,10].
Growing attention has been given to circadian rest-activity rhythms (RARs), which regulate physiological processes such as blood pressure, metabolism, and the sleep–wake cycle. Disruptions in these rhythms have been linked to cardiovascular and metabolic diseases [11]. Actigraphy, a noninvasive method for tracking movement, enables quantification of RARs by capturing fluctuations in activity and rest cycles over time [12]. Previous studies have associated disrupted RARs with increased stroke risk and adverse long-term outcomes such as cognitive decline and functional disability [9,13,14]. Additionally, patients with acquired brain injuries show a higher prevalence of disrupted RARs [15,16]. However, less is known about how these rhythms—particularly the day-to-day stability of physical activity—influence recovery during inpatient rehabilitation for stroke and other neurological conditions.
Among various RAR metrics, interdaily stability (IS) captures the regularity of activity patterns across days and reflects how well a person’s behavior is synchronized with environmental cues such as the light–dark cycle. In population-based cohorts, lower IS has been associated with poorer cognitive performance and adverse cardiometabolic and mental health profiles [13,17-19]. However, its role in inpatient neurorehabilitation remains largely unexplored, despite the potential for externally imposed schedules to influence behavioral rhythms. The importance of repetition for neuroplasticity is well recognized. Current stroke rehabilitation guidelines recommend delivering needs-based, multidisciplinary therapy for at least 3 hours per day, five days per week [20,21]. In Taiwan, standard inpatient rehabilitation follows a similar structure, with therapist-led sessions provided on weekdays and weekends typically lacking formal therapy [17]. Meanwhile, activity outside scheduled therapy is relatively low, particularly on non-therapy days [22-24]. Beneficial effects of additional therapy aimed at increasing time spent in rehabilitation have been documented, particularly with respect to ADL and motor impairments, primarily in studies involving stroke patients [25]. Accordingly, maximizing time spent in rehabilitation beyond therapist-delivered activities may play a key role in facilitating functional outcomes in patients undergoing neurorehabilitation. The stability of physical activity across days—defined as the consistency of movement patterns over multiple days [12]—may serve as a proxy for the stability of daily activities, and it is therefore worth investigating its impact on the prediction of functional outcomes in this setting.
Given these gaps, we hypothesized that greater cross-day stability of physical activity would be associated with greater improvement in ADL among patients undergoing inpatient neurorehabilitation. To minimize heterogeneity while still representing the major patient population, we included the two most common diagnoses in neurorehabilitation: stroke and spinal cord injury (SCI). This study aimed to clarify the relative associations of activity stability and overall activity levels with functional improvement during inpatient neurorehabilitation. Specifically, we investigated whether IS was associated with changes in the Barthel Index (BI) from admission to discharge, and whether total physical activity levels (most active 10-hour period, M10) were associated with BI score change after accounting for IS and other clinical factors. We also explored whether these associations differed between patients with stroke and those with SCI. Our findings may help inform strategies to enhance functional recovery in subacute rehabilitation settings.
METHODS
Study design and participants
This prospective study was conducted in the rehabilitation wards of two university-affiliated hospitals in Northern Taiwan between February 2023 and June 2024. Participants were recruited within the first week of admission to the rehabilitation units and provided written informed consent prior to enrollment. Eligible patients were consecutively approached for participation during the study period.
Inclusion criteria were as follows: age ≥20 years; diagnosis of stroke or SCI; admission for intensive rehabilitation; and a stable medical and neurological condition; and a Glasgow Coma Scale score of 15, indicating sufficient consciousness and ability to follow commands to participate in rehabilitation. Patients with stroke included both ischemic and hemorrhagic stroke subtypes (Table 1). All participants had to be independent in daily living prior to the current admission. Exclusion criteria included pregnancy or intent to conceive, unstable vital signs due to an acute condition, terminal illness with a life expectancy of less than three months, and pre-existing brain disorders (e.g., dementia, previous stroke, Parkinson’s disease, brain tumors, or traumatic brain injury). Eligible patients who declined participation were not enrolled in the study. After enrollment, participants were required to complete at least seven full 24-hour periods of actigraphy recording for inclusion in the final analysis. Participants with insufficient actigraphy data for the predefined analysis period were excluded from the final analysis.
The required sample size was determined using G*Power software (version 3.1.9.7; Heinrich Heine University Düsseldorf). Based on multiple linear regression, a sample size of 110 participants was estimated to achieve 95% statistical power, with an alpha level of 0.05, an effect size (f2) of 0.15, and 10 predictors. This calculated sample size represented the minimum number of participants required for the planned analysis. Considering the possibility of non-participation and incomplete actigraphy recording, we continued recruitment throughout the prespecified study period to ensure that the final analytic sample would remain adequate. Consequently, the final analytic sample exceeded the minimum required sample size.
Ethical considerations
This study was reviewed and approved by the ethics committee of National Taiwan University Hospital (approval number: 202212047RINA, dated January 11, 2023) and conducted in accordance with the principles of the Declaration of Helsinki. Written informed consent was obtained from all participants before enrollment, and participants were informed that they could withdraw from the study at any time without any negative consequences. All actigraphy and clinical data used in this study were de‑identified by a trained research assistant—who was not involved in the subsequent statistical analysis—prior to data processing to ensure privacy and confidentiality. No monetary or material compensation was provided to participants for their participation in this study. A CONSORT chart to illustrate the process of participant screening, enrollment, and inclusion in the final analysis is summarized in Fig. 1.
In-hospital rehabilitation
Patients were admitted from acute care for intensive rehabilitation, most of whom had moderate to severe functional disabilities. Standard rehabilitation consisted of five sessions per week, including physical therapy, occupational therapy, and speech therapy if indicated, each lasting one hour. Bedside rehabilitation was initiated in the acute ward before transfer to the rehabilitation unit to ensure early mobilization and functional improvement. Therapy programs were structured and individualized, incorporating one-on-one therapy and supervised exercises.
The daily routine in the rehabilitation wards included basic nursing care three times a day (9 AM, 1 PM, and 8 PM), and meal distribution at 7 AM, 12 PM, and 6 PM. All patients participated in rehabilitation programs—physical therapy, occupational therapy, and speech therapy if indicated—scheduled between 8 AM to 12 PM or 1 PM to 5 PM, with each session lasting one hour, five days a week. Outside therapy hours, patients were encouraged to participate in standing exercises, ambulation, or ADL, with assistance from caregivers when necessary.
Data collection and outcome measures
Demographic and clinical data were extracted from medical records, including comorbidities and prior functional status. The primary outcome was ADL function, assessed using the 100-point version of BI at both admission and discharge. The BI is a validated measure of functional independence across 10 essential self-care tasks, with total scores ranging from 0 to 100, where higher scores indicate greater independence [26,27]. BI scores were routinely assessed, jointly reviewed by the attending and resident physiatrists responsible for each patient, and recorded in the admission and discharge notes at each facility. These assessments were conducted as part of standard clinical care, and the assessors were blinded to the actigraphy-derived metrics (IS and M10).
Actigraphy measurement and activity pattern analysis
Two non-parametric circadian rhythm metrics were used to characterize rest–activity patterns:
1. M10: The 10 consecutive hours of the day with the highest sustained activity, serving as an indicator of total physical activity [28,29]. M10 was determined using a 10-hour moving average of the processed activity counts to identify the period with the highest average acceleration.
2. IS: Reflects the regularity of activity patterns across days, indicating the degree of synchronization with environmental cues. IS values range from 0 to 1, with higher values denoting greater stability and more predictable activity–rest cycles [12], as follow:
IS=∑h=1pxh-x¯2N∑i=1Nxi-x¯2p
where N represents the number of hourly averaged acti-counts, indicating the total data points across all days considered; p is the number of data points per day (e.g., 24 for hourly bins); xh denotes the hourly averages of the 24-hour profile (the template values); and xi refers to the individual hourly observations; and x¯ is the grand mean of all observations (xi).
Physical activity and the rest–activity metrics were assessed using a research-grade wrist actigraphy device (MiCor A100; MiTAC Inc.), worn on the less-affected (better functional) upper limb in stroke patients, or on the upper limb with better muscle strength in patients with SCI, for at least seven consecutive days during inpatient rehabilitation. Participants were instructed to wear the device continuously throughout the monitoring period, except when temporary removal was necessary, such as during bathing or medical procedures. The device recorded triaxial acceleration data, which were processed using the Euclidean distance method with a bandpass filter (0.5–3 Hz). Processed acceleration signals were converted into activity counts and averaged into 1-minute epochs [30,31], and M10 values were expressed in counts per minute (CPM). In the present sample, the 1-minute epoch activity counts were approximately distributed between 0 and 500 CPM (Supplementary Fig. S1). Periods with insufficient or invalid actigraphy recording were treated as missing data rather than as inactivity. To ensure adequate representation of daily activity patterns, only participants with at least seven full 24-hour periods of valid actigraphy recording were included in the final analysis.
In addition to IS and M10, intradaily variability (IV) was also examined to assess rhythm fragmentation. IV was included in the additional analysis to explore its potential association with functional recovery alongside other rest–activity rhythm parameters.
Statistical analysis
Descriptive statistics were used to summarize participant characteristics. Continuous variables were reported as means and standard deviations, and categorical variables as frequencies and percentages. Differences in baseline characteristics between participants with stroke and SCI were evaluated using independent-samples t-tests for continuous variables and chi-square tests for categorical variables.
To examine the association between physical activity patterns and ADL recovery, a multiple linear regression analysis was conducted. The dependent variable was change in BI from admission to discharge (BI change=discharge BI-admission BI). The primary variables of interest were IS and M10. The model adjusted for baseline BI at admission, age, sex, diagnostic group (stroke vs. SCI), days from onset to admission to the rehabilitation ward, and length of stay in the rehabilitation ward. Subgroup analyses were further conducted separately in the stroke and SCI groups to explore whether the observed associations differed by diagnostic category. All statistical analyses were performed using SPSS for Windows, version 15.0 (SPSS Inc.), with statistical significance set at p<0.05.
RESULTS
A total of 161 participants were included in the analysis. The mean age was 62.4±14.8 years, and 42% were female (Table 1). The most commonly reported comorbidities were hypertension (65.2%), dyslipidemia (46.0%), and diabetes mellitus (26.7%). Stroke accounted for 71.4% of the cases. Among the stroke participants, 62.6% had ischemic stroke and 37.4% had hemorrhagic stroke. The mean time from onset to admission to the rehabilitation ward was 20.5±17.7 days, and the mean length of stay in the rehabilitation ward was 35.8±17.0 days. Upon admission, most participants were severely dependent, with a mean BI score of 32.5±22.9. At discharge, BI scores improved to a mean of 61.1±28.6, reflecting overall functional gains achieved through rehabilitation.
Compared with participants with SCI, those with stroke were older (64.6±14.3 vs. 57.0±14.8 years, p=0.005), more often male (65.2% vs. 41.3%, p=0.005), and more likely to be married (70.4% vs. 21.7%, p<0.001). Participants with stroke also had higher prevalences of hypertension (78.3% vs. 32.6%, p<0.001) and dyslipidemia (55.7% vs. 21.7%, p<0.001). At admission, the stroke group had lower BI scores than the SCI group (30.1±24.0 vs. 38.3±18.8, p=0.024). The stroke group also had a longer stay in the rehabilitation ward (35.8±17.0 days vs. 30.2±12.5 days, p=0.043). IS was lower in the stroke group than in the SCI group (0.455±0.170 vs. 0.535±0.152, p=0.006), whereas M10 did not significantly differ between groups (Table 1).
The overall model explained 28.2% of the variance in BI change (R2=0.282; adjusted R2=0.244). After adjustment for baseline BI at admission, age, sex, stroke diagnosis, length of stay in the rehabilitation ward, and days from onset to rehabilitation admission, higher IS remained significantly associated with greater BI improvement (B=30.86, 95% confidence interval [CI]=9.64 to 52.09, standardized β=0.253, p=0.005). In contrast, M10 was not significantly associated with BI change (B=-0.11, 95% CI=-0.31 to 0.10, standardized β=-0.091, p=0.313). Greater BI improvement was also significantly associated with younger age (B=-0.38, 95% CI=-0.59 to -0.17, standardized β=-0.274, p<0.001), lower baseline BI at admission (B=-0.23, 95% CI=-0.38 to -0.08, standardized β=-0.255, p=0.004), shorter time from onset to rehabilitation admission (B=-0.35, 95% CI=-0.52 to -0.17, standardized β=-0.299, p<0.001), and longer stay in the rehabilitation ward (B=0.31, 95% CI=0.11 to 0.50, standardized β=0.238, p=0.002). Stroke diagnosis and sex were not significantly associated with BI change in the adjusted model (Table 2).
In the stroke subgroup, the regression model was significant and explained 41.4% of the variance in BI change (R2=0.414; adjusted R2=0.375; p<0.001). Higher IS remained significantly associated with greater BI improvement in patients with stroke (B=36.42, 95% CI=13.27 to 59.56, standardized β=0.288, p=0.002). Younger age, lower baseline BI, shorter time from onset to rehabilitation admission, and longer stay in the rehabilitation ward were also significantly associated with greater BI improvement in the stroke subgroup. In contrast, in the SCI subgroup, the model was not significant (R2=0.051; adjusted R2=-0.124; p=0.954), and neither IS nor M10 was significantly associated with BI change (Table 3).
DISCUSSION
Our findings underscore the potential relevance of cross-daily stability of physical activity in neurorehabilitation, identifying IS as a behavioral marker associated with functional recovery. In the adjusted analysis, higher IS obtained in the first week after admission to the rehabilitation ward was significantly associated with greater improvement in ADL, operationalized as change in BI from admission to discharge. Notably, this association was evident in patients with stroke but not in those with SCI, as shown in the subgroup analyses. Younger age, lower baseline BI at admission, shorter time from onset to rehabilitation admission, and longer length of stay were also significantly associated with greater BI improvement, well in accordance with prior studies in stroke or SCI patients [14,32]. Besides these variables, IS may be of particular clinical interest because, unlike age or baseline functional status, day-to-day behavioral regularity may be modifiable during inpatient rehabilitation.
There are several possible explanations for the association between IS and functional improvements. First, IS and other RAR parameters have been linked to a higher burden of arteriolosclerosis in community-dwelling older adults and to delirium in stroke patients [15,33], both of which are associated with poorer recovery. Among the factors contributing to IS in a hospital setting, stroke patients exhibited additional brain damage and a higher prevalence of cognitive impairment [15], which were reflected in lower IS levels and likely reduced engagement in rehabilitation activities. In contrast, a higher IS in SCI group may also explain its limited impact on functional improvement. In our study setting, inpatient stroke rehabilitation follows a structured 5-day therapy schedule [18], and variability in activity outside scheduled sessions may contribute to differences in activity regularity. More consistent activity patterns were associated with greater functional improvement. Similar observations have been reported for step counts among stroke and traumatic brain injury patients in subacute stages [34]. Importantly, these findings should not be interpreted as indicating that total activity volume is unimportant for functional status. Rather, within a model that simultaneously considered M10, baseline BI, and other clinical covariates, IS showed an independent association with BI improvement whereas M10 did not. This suggests that cross-daily regularity may provide prognostic information not fully captured by activity volume alone. Meanwhile, caution is warranted in interpretation, as such activity patterns may reflect caregiver- or staff-assisted mobility rather than solely patient-initiated behavior. Additionally, this association was not observed in the SCI group, which may be related to its smaller sample size and warrants further investigation.
Hypothetically, daily activity routines related to IS may be modifiable, and interventions promoting daily activity consistency (e.g., structured exercises beyond therapy hours, caregiver support, or digital health tools) could help reinforce stable activity rhythms; however, whether such interventions improve rehabilitation outcomes remains to be determined. Notably, evidence from inpatient stroke rehabilitation suggests that providing structured weekend therapy can reduce length of stay and may improve functional independence, supporting the clinical rationale for targeting activity opportunities on non-therapy days [35,36]. Since neurorehabilitation aims not only to restore motor function but also to support long-term reintegration into everyday life, fostering stable physical activity patterns—even on non-therapy days—may represent a potentially important target for future rehabilitation strategies.
This study leveraged continuous, objective 7-day actigraphy to quantify both activity volume (M10) and day-to-day regularity (IS), capturing patient movement patterns beyond therapist-led sessions and documented their predictive value in a prospective design [12,27,30]. The inclusion of patients with both stroke and SCI, together with adjustment for key clinical covariates, allowed us to examine whether the observed association between IS and functional improvement differed across diagnostic groups. In addition to the overall model, subgroup analyses further showed that the association between IS and BI improvement was evident in the stroke subgroup but not in the SCI subgroup, providing a more nuanced view of diagnostic heterogeneity in inpatient neurorehabilitation.
Several limitations must be acknowledged. First, as an observational study, we cannot establish causality. Whether the observed association reflects a true biological effect of rhythmic stability on neuroplasticity or simply identifies patients with higher recovery potential requires confirmation through randomized controlled trials. Second, the median IS in our sample was 0.5, whereas prior community-based studies have considered IS>0.6 to indicate high stability [12]. Because the present analyses focused on IS as a continuous variable rather than on arbitrarily defined high- and low-IS groups, the optimal IS threshold for neurorehabilitation remains unclear and requires further investigation. Third, we could not determine whether inconsistencies in activity were primarily due to structured therapy schedules or other factors such as motivation or medical complications. Access to detailed daily therapy attendance records may provide deeper insight. Fourth, although IS was significantly associated with BI change, M10 was not. This may reflect characteristics of the inpatient rehabilitation setting. Previous studies have shown that stroke and SCI patients exhibit low step counts and upper-limb repetitions during rehabilitation [37,38]. The structured rehabilitation environment may have reduced inter-individual variation in total activity levels, thereby weakening the association between M10 and functional improvement. Fifth, the SCI subgroup was relatively small, which limited statistical power for subgroup inference and may have contributed to the nonsignificant findings in that group. Finally, functional outcomes at discharge are influenced by multiple factors beyond physical recovery, including rehabilitation goals, social support, and discharge planning. Although we used BI change rather than discharge BI alone to better capture rehabilitation-related improvement, this measure may still not fully reflect patients’ longer-term recovery potential. Future studies should incorporate longer follow-up periods to evaluate how cross-daily activity stability affects recovery beyond inpatient care.
While previous research has primarily emphasized the role of overall physical activity in post-stroke recovery [3,4], our findings suggest that activity stability may provide additional prognostic information during inpatient rehabilitation. Observational studies in rehabilitation settings have reported generally low levels of movement practice during formal therapy sessions [37,38] and relatively low step-counts during the subacute inpatient phase [34], highlighting the potential importance of activity performed across days.
Actigraphy offers an objective approach for characterizing circadian RARs and daily activity patterns in clinical populations [12,30,31]. Prior work has described RAR disruption in acquired brain injury populations during in-hospital rehabilitation [15,16]. In addition, Nakagawa et al. [10] reported circadian activity rhythm characteristics in elderly inpatients with stroke or motor diseases and evaluated their relationship to overall activity levels. More recently, a prospective inpatient study in subacute stroke rehabilitation found that IS was an independent predictor of discharge BI after accounting for relevant covariates and other RAR indicators [39]. Our study extends these inpatient findings by including patients with SCI as well as stroke, by examining BI change rather than discharge BI alone, and by directly comparing the associations of activity regularity (IS) and activity volume (M10) within the same adjusted model. The subgroup analyses further suggest that this association may be more apparent in stroke than in SCI, although the SCI subgroup was smaller and more heterogeneous. This between-group comparison is potentially informative rather than merely a source of heterogeneity. In the same acute rehabilitation setting, prior work has shown that poststroke patients and non-brain disease controls, including patients with SCI, can exhibit different RAR profiles, with stroke patients showing lower relative amplitude and higher L5 [15]. Against this background, including both stroke and SCI in the present study provided an opportunity to examine whether the relationship between RAR regularity and functional improvement may also differ across neurological populations receiving broadly similar inpatient care. Although the SCI sample was smaller, the shared ward environment and comparable rehabilitation context still provide a useful basis for exploratory comparison.
Conclusion
Greater IS was associated with greater improvement in ADL, measured as change in BI from admission to discharge. In subgroup analyses, this association was evident in patients with stroke but not in those with SCI. Among RAR metrics, IS—which captures regularity across days—may be the most clinically relevant for neurorehabilitation, because rehabilitation requires sustained engagement over time, whereas indicators that capture within-day fluctuations may be less applicable to recovery over extended rehabilitation periods. IS may serve as a clinically relevant behavioral marker for functional improvement in ADL during rehabilitation, although its causal role and usefulness as an intervention target remain to be established. Future interventional studies should test whether strategies that support consistent daily activity can improve functional recovery.
CONFLICTS OF INTEREST
No potential conflict of interest relevant to this article was reported.
FUNDING INFORMATION
This work was supported by the National Taiwan University Hospital Hsin-Chu Branch and National Health Research Institutes [grant numbers 112-HR0008, 12A1-PHSP05-014].
AUTHOR CONTRIBUTION
Conceptualization: Lin YH, Liang HW. Data curation: Lin YH, Liang HW, Chen SY, Sung KL, Hsu WC, Lin C. Formal analysis: Lin YH, Liang HW, Lin C, Chang HC. Funding acquisition: Lin YH, Liang HW, Wu CH. Project administration: Lin YH, Liang HW, Wu CH. Visualization: Lin YH, Lin C, Chang HC. Writing – original draft: Lin YH, Liang HW. Writing – review and editing: Lin YH, Liang HW, Lin C, Hsu WC. Approval of final manuscript: all authors.
ACKNOWLEDGMENTS
The authors thank for the joint funding from the National Taiwan University Hospital Hsin-Chu Branch and National Health Research Institutes.
This figure illustrates the statistical findings from Table 2 by presenting seven-day actigraphy data from four representative participants. Panels (A) and (B) show two participants with poor activities of daily living (ADL) function at the start of rehabilitation, with baseline Barthel Index (BI) scores of 15 and 0, respectively. Despite having similar physical activity levels (M10=34.7 and 35.8), the participant in panel (A) had a higher IS of 0.72, while the participant in panel (B) had a lower IS of 0.24. At discharge, the participant in panel (A) demonstrated substantial ADL recovery (BI=95), whereas the participant in panel (B) remained functionally dependent (BI=0). Panels (C) and (D) depict two participants with better initial ADL function, with baseline BI scores of 55 and 50, respectively. Again, their M10 values were comparable (40.9 and 43.3), but IS values differed (0.54 vs. 0.28). At discharge, the participant in panel (C) achieved full ADL independence (BI=100), while the participant in panel (D) had a BI of 55. These examples support the regression findings, highlighting the positive association between the stability of physical activity and functional recovery, regardless of total activity levels.
Flow diagram of participant selection and inclusion in the final analysis. Of 701 patients assessed for eligibility, 292 were excluded after screening. Among the 409 eligible patients approached for participation, 227 declined participation or did not provide written informed consent. A total of 182 participants were enrolled, of whom 21 were excluded from the final analysis because of insufficient actigraphy data for the predefined analysis period, leaving 161 participants in the analytic sample.
Table 1.
Baseline characteristics, rehabilitation process, functional outcomes, and actigraphy-based physical activity measures in patients with stroke and spinal cord injury
Baseline characteristic
Stroke (n=115)
Spinal cord injury (n=46)
p-value
Age (yr)
64.6±14.3
57.0±14.8
0.005
Sex
0.005
Female
40 (34.8)
27 (58.7)
Male
75 (65.2)
19 (41.3)
Body mass index (kg/m2)
24.9±4.0
24.5±6.0
0.674
Married
81 (70.4)
10 (21.7)
<0.001
Nonsmoker
84 (73.0)
38 (82.6)
0.282
Stroke subtype
Ischemic stroke
72 (62.6)
-
-
Hemorrhagic stroke
43 (37.4)
-
-
Comorbidities
Hypertension
90 (78.3)
15 (32.6)
<0.001
Diabetes mellitus
35 (30.4)
8 (17.4)
0.091
Dyslipidemia
64 (55.7)
10 (21.7)
<0.001
Cancer
13 (11.3)
7 (15.2)
0.496
Coronary artery disease
16 (13.9)
4 (8.7)
0.435
Atrial fibrillation
20 (17.4)
4 (8.7)
0.222
Initial functional status
Nasogastric tube use
41 (35.7)
0 (0)
<0.001
Indwelling catheter use
27 (23.5)
4 (8.7)
0.045
Barthel Index at admission
30.1±24.0
38.3±18.8
0.024
Days from onset to admission to the rehabilitation ward
18.7±13.7
25.2±24.6
0.095
Rehabilitation process and functional outcomes
Length of stay in rehabilitation ward (day)
35.8±17.0
30.2±12.5
0.043
Barthel Index at discharge
63.0±25.0
60.3±30.0
0.585
Change in Barthel Index from admission to discharge
30.2±21.5
24.8±17.7
0.104
Actigraphy-based physical activity measures
Interdaily stability
0.455±0.170
0.535±0.152
0.006
Activity during the most active 10-hour period (M10)
30.9±17.1
31.7±19.5
0.792
Intradaily variability
1.0±0.2
1.4±0.3
<0.001
Values are presented as mean±standard deviation or number (%).
Table 2.
Multiple linear regression analysis of factors associated with change in BI from admission to discharge
B (95% CI)
Standardized β
p-value
Intercept
39.92 (18.03 to 61.82)
-
<0.001
Age
-0.38 (-0.59 to -0.17)
-0.274
<0.001
Sex
3.90 (-2.12 to 9.92)
0.094
0.203
Baseline BI (admission BI)
-0.23 (-0.38 to -0.08)
-0.255
0.004
IS
30.86 (9.64 to 52.09)
0.253
0.005
M10
-0.11 (-0.31 to 0.10)
-0.091
0.313
Stroke group
3.87 (-3.08 to 10.81)
0.085
0.273
Days from onset to rehabilitation admission
-0.35 (-0.52 to -0.17)
-0.299
<0.001
Length of stay
0.31 (0.11 to 0.50)
0.238
0.002
Model summary: R=0.531; R2=0.282; adjusted R2=0.244; overall model p<0.001.
BI, Barthel Index; CI, confidence interval; IS, interdaily stability; M10, mean activity during the most active 10 hours of the day.
Table 3.
Subgroup multiple linear regression analyses of factors associated with change in BI from admission to discharge in patients with stroke and spinal cord injury
Group/Predictor
B (95% CI)
Standardized β
p-value
Stroke group
Intercept
47.07 (22.22 to 71.92)
-
<0.001
Age
-0.44 (-0.67 to -0.21)
-0.291
<0.001
Sex
4.60 (-2.39 to 11.58)
0.102
0.195
Baseline BI (admission BI)
-0.33 (-0.50 to -0.16)
-0.368
<0.001
IS
36.42 (13.27 to 59.56)
0.288
0.002
M10
-0.03 (-0.28 to 0.21)
-0.027
0.777
Days from onset to rehabilitation admission
-0.56 (-0.82 to -0.30)
-0.358
<0.001
Length of stay
0.37 (0.17 to 0.57)
0.291
<0.001
Model summary
R=0.643; R2=0.414; adjusted R2=0.375
Overall model p<0.001
Spinal cord injury group
Intercept
47.10 (-6.28 to 100.49)
-
0.082
Age
-0.20 (-0.64 to 0.25)
-0.165
0.378
Sex
0.82 (-12.23 to 13.86)
0.023
0.900
Baseline BI (admission BI)
-0.12 (-0.52 to 0.27)
-0.129
0.536
IS
3.54 (-47.68 to 54.76)
0.030
0.890
M10
-0.13 (-0.54 to 0.28)
-0.144
0.519
Days from onset to rehabilitation admission
-0.15 (-0.40 to 0.11)
-0.206
0.245
Length of stay
-0.03 (-0.57 to 0.52)
-0.018
0.926
Model summary
R=0.225; R2=0.051; adjusted R2=-0.124
Overall model p=0.954
The dependent variable was change in BI (discharge BI-admission BI). Baseline BI (admission BI) denotes the BI recorded at admission and was entered as a covariate; it is distinct from the dependent variable (change in BI). Sex was entered as a binary variable. Stroke and spinal cord injury groups were analyzed separately in subgroup models.
BI, Barthel Index; CI, confidence interval; IS, interdaily stability; M10, mean activity during the most active 10 hours of the day.
REFERENCES
1. GBD 2021 Nervous System Disorders Collaborators. Global, regional, and national burden of disorders affecting the nervous system, 1990-2021: a systematic analysis for the Global Burden of Disease Study 2021. Lancet Neurol 2024;23:344-81.
2. Tsao CW, Aday AW, Almarzooq ZI, Alonso A, Beaton AZ, Bittencourt MS, et al. Heart disease and stroke statistics-2022 update: a report from the American Heart Association. Circulation 2022;145:e153-639.
3. Mediano MFF, Mok Y, Coresh J, Kucharska-Newton A, Palta P, Lakshminarayan K, et al. Prestroke physical activity and adverse health outcomes after stroke in the Atherosclerosis Risk in Communities Study. Stroke 2021;52:2086-95.
4. Buvarp D, Viktorisson A, Axelsson F, Lehto E, Lindgren L, Lundström E, et al. Physical activity trajectories and functional recovery after acute stroke among adults in Sweden. JAMA Netw Open 2023;6:e2310919.
5. Błaszcz M, Prucnal N, Wrześniewski K, Pasiut S, Mika P, Kucia M, et al. Physical activity, psychological and functional outcomes in non-ambulatory stroke patients during rehabilitation-a pilot study. J Clin Med 2022;11:7260.
6. Noé E, Gómez A, Bernabeu M, Quemada I, Rodríguez R, Pérez T, et al. Guidelines: basic principles of neurorehabilitation for patients with acquired brain injury. Recommendations of the Spanish Society of Neurorehabilitation. Neurologia (Engl Ed) 2024;39:261-81.
7. Maier M, Ballester BR, Verschure PFMJ. Principles of neurorehabilitation after stroke based on motor learning and brain plasticity mechanisms. Front Syst Neurosci 2019;13:74.
8. Winstein CJ, Stein J, Arena R, Bates B, Cherney LR, Cramer SC, et al. Guidelines for adult stroke rehabilitation and recovery: a guideline for healthcare professionals from the American Heart Association/American Stroke Association. Stroke 2016;47:e98-169.
10. Nakagawa Y, Noda K, Inoue Y. The circadian activity rhythms for elderly inpatients with stroke or motor diseases in a rehabilitation facility and its relationship to physical activity level. Sleep Biol Rhythms 2024;22:125-35.
12. Witting W, Kwa IH, Eikelenboom P, Mirmiran M, Swaab DF. Alterations in the circadian rest-activity rhythm in aging and Alzheimer's disease. Biol Psychiatry 1990;27:563-72.
13. Sun X, Yu W, Wang M, Hu J, Li Y. Association between rest-activity rhythm and cognitive function in the elderly: the U.S. National Health and Nutrition Examination Survey, 2011-2014. Front Endocrinol (Lausanne) 2023;14:1135085.
15. Liang HW, Wu CH, Lin C, Chang HC, Lin YH, Chen SY, et al. Rest-activity rhythm differences in acute rehabilitation between poststroke patients and non-brain disease controls: comparative study. J Med Internet Res 2024;26:e49530.
16. Kjeldsen SS, Brunner I, Nielsen JF, Pedersen AR. Estimating day-to-day circadian rhythm in patients with severe acquired brain injury at the beginning of in-hospital rehabilitation. J Integr Neurosci 2022;21:58.
17. Sung KL, Lin YH, Lin C, Wu CH, Chang HC, Liang HW, et al. Impact of actigraphy-based circadian rest-activity rhythms on functional outcomes in post-stroke rehabilitation. J Formos Med Assoc 2026;125:971-6.
18. Yang L, Feng H, Chen J, Kwok Wing Y, Benedict C, Tan X, et al. Association of circadian rest-activity rhythms with cardiovascular disease and mortality in type 2 diabetes. Diabetes Res Clin Pract 2023;197:110262.
19. Chen IM, Lin C, She GJ, Chang HC, Chuang HH, Chen TY, et al. Rest-activity rhythm patterns and their associations with depression and obesity: a study using actigraphy and human-smartphone interactions. Depress Anxiety 2025;2025:2617282.
21. Tang E, Moran N, Cadman M, Hill S, Sloan C, Warburton E; guideline committee. Stroke rehabilitation in adults: summary of updated NICE guidance. BMJ 2024;384:q498.
22. Whiteneck G, Gassaway J, Dijkers M, Backus D, Charlifue S, Chen D, et al. The SCIRehab project: treatment time spent in SCI rehabilitation. Inpatient treatment time across disciplines in spinal cord injury rehabilitation. J Spinal Cord Med 2011;34:133-48.
23. Kenah K, Tavener M, Bernhardt J, Spratt NJ, Janssen H. "Wasting time": a qualitative study of stroke survivors' experiences of boredom in non-therapy time during inpatient rehabilitation. Disabil Rehabil 2024;46:2799-807.
24. Eng XW, Brauer SG, Kuys SS, Lord M, Hayward KS. Factors affecting the ability of the stroke survivor to drive their own recovery outside of therapy during inpatient stroke rehabilitation. Stroke Res Treat 2014;2014:626538.
25. Clark B, Whitall J, Kwakkel G, Mehrholz J, Ewings S, Burridge J. The effect of time spent in rehabilitation on activity limitation and impairment after stroke. Cochrane Database Syst Rev 2021;10:CD012612.
28. Gonçalves BS, Adamowicz T, Louzada FM, Moreno CR, Araujo JF. A fresh look at the use of nonparametric analysis in actimetry. Sleep Med Rev 2015;20:84-91.
29. Jones SE, van Hees VT, Mazzotti DR, Marques-Vidal P, Sabia S, van der Spek A, et al. Genetic studies of accelerometer-based sleep measures yield new insights into human sleep behaviour. Nat Commun 2019;10:1585.
31. Ancoli-Israel S, Cole R, Alessi C, Chambers M, Moorcroft W, Pollak CP. The role of actigraphy in the study of sleep and circadian rhythms×. Sleep 2003;26:342-92.
32. Hsu WC, Hsu SC, Cheng JC, Pan SL. Prognostic factors for functional disability in persons with spinal cord injury. J Formos Med Assoc 2025. doi: 10.1016/j.jfma.2025.08.001 [Epub ahead of print].
33. Lin C, Chen IM, Chuang HH, Wang ZW, Lin HH, Lin YH. Examining human-smartphone interaction as a proxy for circadian rhythm in patients with insomnia: cross-sectional study. J Med Internet Res 2023;25:e48044.
34. Mischler B, Hilfiker R, Hund-Georgiadis M, Maguire C. Physical activity based on daily step-count in inpatient setting in stroke and traumatic brain injury patients in subacute stage: a cross-sectional observational study. NeuroRehabilitation 2023;52:435-50.
35. English C, Shields N, Brusco NK, Taylor NF, Watts JJ, Peiris C, et al. Additional weekend therapy may reduce length of rehabilitation stay after stroke: a meta-analysis of individual patient data. J Physiother 2016;62:124-9.
36. Peiris CL, Shields N, Brusco NK, Watts JJ, Taylor NF. Additional Saturday rehabilitation improves functional independence and quality of life and reduces length of stay: a randomized controlled trial. BMC Med 2013;11:198.
37. Lang CE, Macdonald JR, Reisman DS, Boyd L, Jacobson Kimberley T, Schindler-Ivens SM, et al. Observation of amounts of movement practice provided during stroke rehabilitation. Arch Phys Med Rehabil 2009;90:1692-8.
38. Zbogar D, Eng JJ, Miller WC, Krassioukov AV, Verrier MC. Movement repetitions in physical and occupational therapy during spinal cord injury rehabilitation. Spinal Cord 2017;55:172-9.
39. Iddagoda MT, Inderjeeth CA, Chan K, Raymond WD. Post-stroke sleep disturbances and rehabilitation outcomes: a prospective cohort study. Intern Med J 2020;50:208-13.
Association of Daily Activity Regularity With Functional Recovery in Patients With Stroke and Spinal Cord Injury: A Prospective Actigraphy Study
Fig. 1. Flow diagram of participant selection and inclusion in the final analysis. Of 701 patients assessed for eligibility, 292 were excluded after screening. Among the 409 eligible patients approached for participation, 227 declined participation or did not provide written informed consent. A total of 182 participants were enrolled, of whom 21 were excluded from the final analysis because of insufficient actigraphy data for the predefined analysis period, leaving 161 participants in the analytic sample.
Graphical abstract
Fig. 1.
Graphical abstract
Association of Daily Activity Regularity With Functional Recovery in Patients With Stroke and Spinal Cord Injury: A Prospective Actigraphy Study
Baseline characteristic
Stroke (n=115)
Spinal cord injury (n=46)
p-value
Age (yr)
64.6±14.3
57.0±14.8
0.005
Sex
0.005
Female
40 (34.8)
27 (58.7)
Male
75 (65.2)
19 (41.3)
Body mass index (kg/m2)
24.9±4.0
24.5±6.0
0.674
Married
81 (70.4)
10 (21.7)
<0.001
Nonsmoker
84 (73.0)
38 (82.6)
0.282
Stroke subtype
Ischemic stroke
72 (62.6)
-
-
Hemorrhagic stroke
43 (37.4)
-
-
Comorbidities
Hypertension
90 (78.3)
15 (32.6)
<0.001
Diabetes mellitus
35 (30.4)
8 (17.4)
0.091
Dyslipidemia
64 (55.7)
10 (21.7)
<0.001
Cancer
13 (11.3)
7 (15.2)
0.496
Coronary artery disease
16 (13.9)
4 (8.7)
0.435
Atrial fibrillation
20 (17.4)
4 (8.7)
0.222
Initial functional status
Nasogastric tube use
41 (35.7)
0 (0)
<0.001
Indwelling catheter use
27 (23.5)
4 (8.7)
0.045
Barthel Index at admission
30.1±24.0
38.3±18.8
0.024
Days from onset to admission to the rehabilitation ward
18.7±13.7
25.2±24.6
0.095
Rehabilitation process and functional outcomes
Length of stay in rehabilitation ward (day)
35.8±17.0
30.2±12.5
0.043
Barthel Index at discharge
63.0±25.0
60.3±30.0
0.585
Change in Barthel Index from admission to discharge
30.2±21.5
24.8±17.7
0.104
Actigraphy-based physical activity measures
Interdaily stability
0.455±0.170
0.535±0.152
0.006
Activity during the most active 10-hour period (M10)
30.9±17.1
31.7±19.5
0.792
Intradaily variability
1.0±0.2
1.4±0.3
<0.001
B (95% CI)
Standardized β
p-value
Intercept
39.92 (18.03 to 61.82)
-
<0.001
Age
-0.38 (-0.59 to -0.17)
-0.274
<0.001
Sex
3.90 (-2.12 to 9.92)
0.094
0.203
Baseline BI (admission BI)
-0.23 (-0.38 to -0.08)
-0.255
0.004
IS
30.86 (9.64 to 52.09)
0.253
0.005
M10
-0.11 (-0.31 to 0.10)
-0.091
0.313
Stroke group
3.87 (-3.08 to 10.81)
0.085
0.273
Days from onset to rehabilitation admission
-0.35 (-0.52 to -0.17)
-0.299
<0.001
Length of stay
0.31 (0.11 to 0.50)
0.238
0.002
Group/Predictor
B (95% CI)
Standardized β
p-value
Stroke group
Intercept
47.07 (22.22 to 71.92)
-
<0.001
Age
-0.44 (-0.67 to -0.21)
-0.291
<0.001
Sex
4.60 (-2.39 to 11.58)
0.102
0.195
Baseline BI (admission BI)
-0.33 (-0.50 to -0.16)
-0.368
<0.001
IS
36.42 (13.27 to 59.56)
0.288
0.002
M10
-0.03 (-0.28 to 0.21)
-0.027
0.777
Days from onset to rehabilitation admission
-0.56 (-0.82 to -0.30)
-0.358
<0.001
Length of stay
0.37 (0.17 to 0.57)
0.291
<0.001
Model summary
R=0.643; R2=0.414; adjusted R2=0.375
Overall model p<0.001
Spinal cord injury group
Intercept
47.10 (-6.28 to 100.49)
-
0.082
Age
-0.20 (-0.64 to 0.25)
-0.165
0.378
Sex
0.82 (-12.23 to 13.86)
0.023
0.900
Baseline BI (admission BI)
-0.12 (-0.52 to 0.27)
-0.129
0.536
IS
3.54 (-47.68 to 54.76)
0.030
0.890
M10
-0.13 (-0.54 to 0.28)
-0.144
0.519
Days from onset to rehabilitation admission
-0.15 (-0.40 to 0.11)
-0.206
0.245
Length of stay
-0.03 (-0.57 to 0.52)
-0.018
0.926
Model summary
R=0.225; R2=0.051; adjusted R2=-0.124
Overall model p=0.954
Table 1. Baseline characteristics, rehabilitation process, functional outcomes, and actigraphy-based physical activity measures in patients with stroke and spinal cord injury
Values are presented as mean±standard deviation or number (%).
Table 2. Multiple linear regression analysis of factors associated with change in BI from admission to discharge
Model summary: R=0.531; R2=0.282; adjusted R2=0.244; overall model p<0.001.
BI, Barthel Index; CI, confidence interval; IS, interdaily stability; M10, mean activity during the most active 10 hours of the day.
Table 3. Subgroup multiple linear regression analyses of factors associated with change in BI from admission to discharge in patients with stroke and spinal cord injury
The dependent variable was change in BI (discharge BI-admission BI). Baseline BI (admission BI) denotes the BI recorded at admission and was entered as a covariate; it is distinct from the dependent variable (change in BI). Sex was entered as a binary variable. Stroke and spinal cord injury groups were analyzed separately in subgroup models.
BI, Barthel Index; CI, confidence interval; IS, interdaily stability; M10, mean activity during the most active 10 hours of the day.
, Chen Lin, PhD4
