PHWR Public Health Weekly Report

Stroke is a leading cause of death and disability worldwide; it significantly impacts the long-term functional status and quality of life of survivors [1]. Recent advances in treatment for stroke in the acute phase and in rehabilitation therapy have increased survival rate and life expectancy, but the long-term management of stroke in the chronic phase has become medically and socially challenging [2]. However, there are relatively few studies investigating changes in the functioning of patients in the chronic phase compared to studies investigating the improvement in functioning in the acute and subacute phases [3].

Functional changes in patients with stroke in the chronic phase are affected by various complex factors. In general, functional ability in the chronic phase may become stable or deteriorate gradually [4]. In previous studies conducted by this study team, the ability of performing routine activities decreased statistically significantly between three years and four years after the stroke [5,6]. Remarkably decreased long-term functional ability can increase the patient’s dependency and present a sizeable societal burden, resulting in decreased long-term survival rate [7]. Accordingly, predicting and preventing a decrease in functional ability in these phases is critical. Therefore, it is necessary to identify factors within patient characteristics three years after the onset that can predict decreased functional ability four years after a stroke. Identifying predictive factors can help detect high-risk patients early and enable customized interventions [8].

This study determines factors that can predict a decline in functional ability four years after the onset from the patient characteristics investigated three years after a stroke. This will provide basic data to establish efficient therapeutic strategies that can prevent a decline in the functional ability of patients with stroke in the chronic phase and ultimately contribute to improved quality of life and reduced societal burden.

1. Study Population

This study analyzed data from first-ever stroke patients of the Korean Stroke Cohort for Functioning and Rehabilitation (KOSCO), designed as a consignment project and long-term follow-up of stroke patients conducted by the Korean Disease Control and Prevention Agency. The first-ever stroke cohort included 7,858 patients who consented to a long-term follow-up, among 10,636 first-ever acute stroke patients (8,210 with cerebral infarction and 2,426 with cerebral hemorrhage) who were hospitalized seven days before the onset of stroke in nine hospitals nationwide between August 2012 and May 2015. The detailed protocol of the KOSCO is provided in the previous study [9]. Among the 7,858 participants who consented to the long-term follow-up of KOSCO, the functioning of 4,731 participants was evaluated face-to-face three years post onset. Of the 4,731 participants, this study analyzed data from stroke patients with the modified Rankin scale (mRS) [10] 0 to 4 (from no symptoms to moderate to severe symptoms). From those with mRS, five were excluded from this study because the level of disability could not be evaluated. Therefore, 3,266 participants with mRS 0 or 1, who were capable of independently performing routine activities, were defined as independent patients. The 1,165 participants with mRS 2 to 4, who were incapable of independently performing routine activities, were defined as dependent patients. Among them, 4,023 participants, including 3,028 independent and 955 dependent patients, who were evaluated face-to-face for functioning four years post onset, were analyzed (Figure 1). Furthermore, an increase in the mRS score due to decreased functioning four years compared to three years post onset was defined as functional deterioration. A reduction or maintenance of mRS was defined as no functional deterioration in both groups.

Figure 1. Flow chart of participants
KOSCO=Korean Stroke Cohort for Functioning and Rehabilitation.

Sociodemographic characteristics, such as age, sex, education level, family background, comorbidities, and pre-stroke functional level, were collected from medical records. The severity of comorbidity was evaluated with combined condition-and age-related score [11] of the Charlson comorbidity index. Pre-stroke functional level was evaluated using mRS. The features and severity of a stroke seven days post-onset were evaluated. Multifaceted assessment of functional outcomes three years post-onset was conducted by using the Korean Mini-Mental State Examination for cognitive function [12], Fugl-Meyer Assessment [13] for motor functioning, Functional Ambulatory Category for ambulation ability [14], American Speech-Language-Hearing Association National Outcome Measurement System Swallowing Scale [15] for swallowing ability, and the Korean Version of the Frenchay Aphasia Screening Test [16] for language. The evaluation was a one-on-one, face-to-face process. A questionnaire survey was conducted with the Geriatric Depression Scale-Short Form [17] to evaluate emotional level.

2. Analytical Methods

Descriptive statistics was used to describe the participants’ demographic and clinical characteristics. To compare those with functional deterioration to those with no functional deterioration in every independent and dependent patient, an independent t-test and Chi-square test were used for continuous variables and categorical variables, respectively. The results of continuous variables were presented as mean (standard deviation), and the results of categorical variables were presented as frequency (percentage). For variables that showed statistical differences in the patients with functional deterioration and those with no functional deterioration, the dependent variables were set as the patients with functional deterioration. Independent variables that can predict functional decline were tested via binary logistic regression analysis. A p-value less than 0.05 was considered to be statistically significant. All data were analyzed using SPSS version 24.0 (IBM Co.).

For the total 4,023 participants, the mean age 3 years post-onset was 65.6±12.4 years, 60.7% were male participants, and 79.5% had experienced ischemic stroke. Among the total participants, the mean mRS at 3 years and 4 years post onset of stroke was 1.05±1.17 and 1.10±1.23, respectively, showing a statistically significant reduction (p<0.05). A functional decline 4 years post onset was reported in 392 (12.9%) patients with independence and 120 (12.1%) patients with dependence; no statistically significant difference in a decline in functioning was observed between the groups (Figure 2).

Figure 2. Rate of functional deterioration from 3 years to 4 years after stroke onset according to independence at 3 years after stroke
There was no significant difference in the rate of functional deterioration between the two groups.

1. Analysis of Independent Patients Three Years Post Onset of Stroke

For the 3,028 participants with mRS 0 or 1, who did not present disability 3 years post onset, Table 1 shows the demographic and clinical characteristics of 392 participants with functional deterioration and 2,636 participants with no functional deterioration. Compared to the latter, the former group was older with a higher rate of not having a spouse (p<0.05). The prevalence of hypertension, diabetes mellitus, and atrial fibrillation was statistically higher (p<0.05), which caused a statistically significantly higher rate of comorbidity (p<0.05). A statistically higher rate of ischemic cerebral infarction was observed in the participants with functional deterioration (p<0.05), but no significant difference in pre-stroke functional level and the severity of stroke seven days post onset was observed between the two groups. Participants with functional deterioration four years post onset showed a statistically significant decline in cognitive function, motor functioning, ambulation ability, swallowing ability, language skill, and emotional level three years post onset of stroke compared to those with no functional deterioration (p<0.05, Table 1).

Table 1 General and functional characteristics of stroke patients with independence at 3 years after stroke

Variable	Functional deterioration (n=392)	No functional deterioration (n=2,636)	p-value
General and clinical characteristic
Age (yr)	67.2±12.3*	63.6±11.9	<0.001
Sex, male:female	64.0:36.0	63.4:36.6	0.822
Education			0.071
Elementary school degree or less	27.9	22.8
Middle school degree	49.2	50.9
Bachelor’s degree or more	22.9	26.3
Spouse, yes	80.7*	85.3	0.026
Medical history
Hypertension	55.2*	49.6	0.043
Diabetes mellitus	24.8*	18.7	0.006
Coronary heart disease	7.0	4.9	0.084
Atrial fibrillation	8.1*	4.9	0.014
Hyperlipidemia	12.3	9.5	0.091
CCAS	2.6±1.7*	2.1±1.8	<0.001
Premorbid mRS (score)	0.6±1.2	0.6±1.2	0.418
Stroke type, ischemic:hemorrhagic	84.7:15.3*	80.2:19.8	0.033
NIHSS at 7 days after stroke onset	2.1±3.2	1.8±3.4	0.181
Functional characteristics at 3 years after onset
K-MMSE	27.1±3.6*	28.1±2.7	<0.001
Fugl-Meyer Assessment	97.9±7.0*	99.0±4.5	0.003
Functional Ambulatory Category	4.9±0.4*	5.0±0.2	0.001
AHSA-NOMS	6.8±0.5*	6.9±0.3	0.001
Short K-FAST	17.3±3.4*	18.1±2.7	<0.001
GDS-SF	3.6±3.6*	2.9±3.2	0.001

Data are presented as mean±standard deviation or %. CCAS=combined condition- and age-related score; mRS=modified Rankin scale; NIHSS=National Institutes of Health Stroke Scale; K-MMSE=Korean Mini-Mental State Examination; AHSA-NOMS=American Speech-Language-Hearing Association National Outcome Measurement System Swallowing Scale; Short K-FAST=Short Korean Version of the Frenchay Aphasia Screening Test; GDS-SF=Geriatric depression scale-short form. *p<0.05..

When factors with statistical differences between the participants with functional deterioration and those with no functional deterioration were verified using the binary logistic regression analysis, age, comorbidity severity, and ambulation ability three years post-onset were analyzed as independent predictive factors (p<0.05). Presence of a spouse, cognitive function, swallowing ability, and emotional level three years post onset were analyzed as predictive factors with a tendency (Table 2).

Table 2 Binary logistic regression analysis results to identify factors at 3 years after stroke associated with the functional deterioration at 4 years after stroke

Predictor	Regression coefficient	SE	Wald χ2	p-value	Odds ratio (95% CI)
Stroke patients with independence at 3 years
Age	0.013*	0.006	4.801	0.028	1.013 (1.001–1.025)
Spouse	–0.288	0.153	3.545	0.060	0.750 (0.555–1.012)
CCAS	0.094*	0.034	7.561	0.006	1.099 (1.028–1.175)
K-MMSE	–0.041	0.022	3.674	0.055	0.960 (0.419–0.959)
Functional Ambulatory Category	–0.456*	0.211	4.661	0.031	0.634 (0.419–0.959)
AHSA-NOMS	–0.301	0.175	2.953	0.086	0.740 (0.525–1.043)
GDS-SF	0.032	0.017	3.470	0.062	1.033 (0.998–1.068)
Stroke patients with dependence at 3 years
Age	0.030*	0.010	9.392	0.002	1.031 (1.011–1.050)
K-MMSE	–0.026	0.014	3.463	0.063	0.974 (0.948–1.001)

SE=standard error; CI=confidence interval; CCAS=combined condition- and age-related score; K-MMSE=Korean Mini-Mental State Examination; AHSA-NOMS=American Speech-Language-Hearing Association National Outcome Measurement System Swallowing Scale; GDS-SF=Geriatric depression scale-short form. *p<0.05..

2. Analysis of Dependent Patients Three Years Post Onset of Stroke

Among the 995 participants with mRS 2 to 4, who presented disability, Table 3 shows the demographic and clinical characteristics of 120 participants with functional deterioration and 875 participants with no functional deterioration three years post onset. Compared to the participants with no functional deterioration, those with functional deterioration were older (p<0.05), but no statistical difference in education level, presence of spouse, and comorbidities was observed. No significant difference in the features of a stroke, pre-stroke functional level, and the severity of stroke seven days post onset was observed between the two groups. Participants with functional deterioration four years post onset showed a statistically significant decline in cognitive function, ambulation ability, swallowing ability, and language skill three years post onset of stroke compared to those with no functional deterioration (p<0.05, Table 3).

Table 3 General and functional characteristics of stroke patients with dependence at 3 years after stroke

Variable	Functional deterioration (n=120)	No functional deterioration (n=875)	p-value
General and clinical characteristics
Age (yr)	74.7±12.2*	69.6±12.2	<0.001
Sex, male:female	53.3:46.7	51.9:48.1	0.771
Education			0.379
Elementary school degree or less	43.9	44.9
Middle school degree	39.5	42.9
Bachelor’s degree or more	16.7	12.1
Spouse, yes	70.2	72.9	0.576
Medical history
Hypertension	61.7	57.8	0.480
Diabetes mellitus	27.4	26.3	0.824
Coronary heart disease	7.6	6.0	0.540
Atrial fibrillation	7.8	7.3	0.850
Hyperlipidemia	15.4	10.3	0.112
CCAS	2.6±2.2	2.6±1.9	0.764
Premorbid mRS (score)	0.5±1.1	0.7±1.4	0.150
Stroke type, ischemic:hemorrhagic	80.8:19.2*	75.0:25.0	0.174
NIHSS at 7 days after stroke onset	6.7±7.2	6.9±6.8	0.693
Functional characteristics at 3 years after onset
K-MMSE	19.8±7.7*	21.9±7.1	0.007
Fugl-Meyer Assessment	73.9±28.5	73.8±27.5	0.972
Functional Ambulatory Category	3.2±1.3*	3.5±1.4	0.017
AHSA-NOMS	6.4±1.0*	6.6±0.7	0.009
Short K-FAST	11.0±5.9*	12.6±5.6	0.006
GDS-SF	3.8±4.5	4.3±4.8	0.249

Data are presented as mean±standard deviation or %. CCAS=combined condition- and age-related score; mRS=modified Rankin scale; NIHSS=National Institutes of Health Stroke Scale; K-MMSE=Korean Mini-Mental State Examination; AHSA-NOMS=American Speech-Language-Hearing Association National Outcome Measurement System Swallowing Scale; Short K-FAST=Short Korean Version of the Frenchay Aphasia Screening Test; GDS-SF=Geriatric depression scale-short form. *p<0.05..

When factors with statistical differences between the participants with functional deterioration and those with no functional deterioration were verified using the binary logistic regression analysis, age and cognitive function three years post onset were analyzed as independent predictive factors (p<0.05, Table 2).

This study divided the total participants into two groups, namely independent and dependent patients. Among the characteristics investigated three years post onset of stroke in each group, the factors that can predict a decline in functioning four years post onset were identified. Of the 4,023 participants, 12.9% independent patients and 12.1% dependent patients showed a functional decline four years post onset. The predictive factors for functional decline four years post onset were age, comorbidity severity, and ambulation ability in the independent patients and age and cognitive function in the dependent patients.

In 2021, the annual incidence of stroke was reported to be 108,950 cases in the Republic of Korea [18]. Since this study shows that approximately 12.1–12.9% of patients reported functional decline 4 years post onset of stroke, an annual incidence of functional decline can be projected to be 13,618 patients (calculated as 12.5%) with chronic stroke. Predicting functional decline in those patients using the predictive factors identified in this study and providing customized rehabilitation therapy to prevent decline can significantly reduce societal burden and improve quality of life.

Predictive factors for functional decline were different between the two groups, which implies a need for management strategies that differ depending on the functional level of chronic phase stroke patients [19]. Among the independent patients three years post-onset of stroke, age, comorbidity severity, and ambulation ability were analyzed as predictive factors for a statistically significant decline in independent functioning. Among the factors, the ambulation ability is the sole modifiable factor that can fluctuate with additional rehabilitation therapies [7]. A decline in ambulation ability reduces activity level and causes deconditioning, which supposedly resulted in functional decline [20]. Accordingly, appropriate rehabilitation therapy to maintain or enhance ambulation ability for chronic phase stroke patients, who can perform daily routine activities independently, is expected to prevent functional decline [20]. The incidence of functional decline is higher especially among older adult stroke patients in the chronic phase since they are more likely to have comorbidities. They should be prioritized while administering appropriate rehabilitation therapy for chronic phase stroke to promote ambulation ability [21].

However, age and cognitive function three years post onset were analyzed as predictive factors for statistically significant decline in independent functioning among the dependent patients three years post onset of stroke. Owing to a relatively low ambulation ability, cognitive function is assumed to play a more important role in performing daily routine activities [22]. Accordingly, appropriate rehabilitation therapy to maintain or enhance cognitive function for stroke patients in the chronic phase, who cannot perform routine activities independently, is expected to prevent functional decline [23]. Since the incidence of functional decline is higher especially among older chronic stroke patients, they should be prioritized while administering appropriate rehabilitation therapy in the chronic phase to promote cognitive function [24].

The results emphasize the importance of customized rehabilitation strategies to maintain and improve functioning for chronic stroke patients. A program focused on improved ambulation ability for independent patients and an intervention that emphasizes improved cognitive function for dependent patients can be effective [25,26]. Furthermore, since irreversible factors, such as age and comorbidity, were also identified as critical predictive factors, there is a need for more aggressive and persistent management for high-risk patients [2,27].

In conclusion, this study provides a basis for establishing customized rehabilitation strategies by investigating the predictive factors for functional decline in stroke patients in the chronic phase. Further research that develops a customized rehabilitation program in consideration of the predictive factors and verifies its effectiveness is needed [27]. Additionally, the results of this study are expected to contribute to the establishment of policy for maintaining long-term functioning and improving the quality of life of chronic phase stroke patients [2].

Ethics Statement: This multicenter cohort study was approved by the Institutional Review Board of each participating hospital.

Funding Source: This study was supported by the Research Program funded by the Korea Disease Control and Prevention Agency (no. 3300-3334-300-260-00, 2013-E33017-00, 2013E-33017-01, 2013E-33017-02, 2016-E33003-00, 2016-E33003-01, 2016-E33003-02, 2019-E3202-00, 2019-E3202-01, 2019-E3202-02, 2022-11-006).

Acknowledgments: None.

Conflict of Interest: The authors have no conflicts of interest to declare.

Author Contributions: Conceptualization: YHL, YTK, YHK, and WHC. Data curation: DYK, JML, Min Kyun Sohn, YIS, YSL, Min-Keun Song, MCJ, SYL, DNC, GJO, JHA, JHH, YHK, and WHC. Formal analysis: DHK and WHC. Funding acquisition: YHK and WHC. Methodology: DHK, HSL, DYK, JML, Min Kyun Sohn, YIS, YSL, Min-Keun Song, MCJ, SYL, DNC, GJO, JHA, JHH, YHK, and WHC. Supervision: DYK, JML, Min Kyun Sohn, YIS, YSL, Min-Keun Song, MCJ, SYL, GJO, JHA, JHH, YHK, and WHC. Writing – original draft: DHK and HSL. Writing – review & editing: YHL, YTK, YHK, and WHC.