Poststroke fatigue in subacute and chronic stroke rehabilitation prevalence, associated factors and impact on self-efficacy and functional ability

Authors: Yohannes Girma Legese, Foad Akmel, Yazachew Mekonnen, Dheeraj Lamba

Abstract

Background

Stroke is a leading cause of death and long-term disability worldwide, with low-income and middle-income countries accounting for about 87% of stroke-related deaths and disability-adjusted life-years. Among poststroke complications, poststroke fatigue is a common but often under-recognised condition characterised by emotional, cognitive and physical exhaustion unrelated to exertion and not relieved by rest. Poststroke fatigue can hinder functional recovery, yet it remains underassessed in the study area. So this study investigates the prevalence, associated factors and impact of poststroke fatigue on self-efficacy and activities of daily living.

Methods

A hospital-based cross-sectional study was conducted through a systematic random sampling technique on 370 study participants. The Fatigue Severity Scale, Nottingham Extended Activities of Daily Living and Stroke Self-efficacy Questionnaire were used to collect questionnaires related to fatigue and its impact. The data were analysed on SPSS using binary logistic regression to assess associated factors and ordinal logistic regression to assess impacts of fatigue on self-efficacy and activities of daily living.

Result

Fatigue was reported by 65.4% of participants (95% CI 60.30% to 70.20%) during the subacute and chronic rehabilitation phases. Older age, both overweight and underweight body mass index, National Institutes of Health Stroke Scale ≥12, lacked physiotherapy follow-up and experiencing depression were associated with poststroke fatigue with 95% CI and p<0.05. Moreover, individuals with poststroke fatigue were approximately 5.4 times less likely to report higher functional levels and 3.6 times less likely to demonstrate greater self-efficacy compared with those without fatigue.

Conclusions

Poststroke fatigue is highly prevalent among stroke survivors and negatively impacts both self-efficacy and functional ability. These findings highlight the need for early identification and targeted management of fatigue to improve rehabilitation outcomes and quality of life in stroke survivors.

INTRODUCTION

Stroke is the leading cause of death and long-term disability worldwide, and it creates a significant public health burden. The interruption of blood flow during stroke can lead to a quick loss of brain function, causing issues that range from mild neurological problems to severe disabilities or death.^1^ The majority of stroke deaths and disability-adjusted life-years worldwide, approximately 87.0% and 89.0%, respectively, occur in lower-income and lower-middle-income countries.^2^

Poststroke fatigue (PSF) is a multidimensional emotional, cognitive and motor-perceptual experience, and it is a feeling of intense physical or mental exhaustion that does not get better with rest and may persist for weeks, months or even years after a stroke. PSF prevalence in different studies ranges between 25% and 85%.^3^ The consequences of PSF extend towards reducing the ability of stroke survivors to perform instrumental activities of daily living (IADLs) such as transportation use, managing work environment and managing medical and rehabilitation programmes. Limitations in IADLs not only compromise independence but also increase the burden on caregivers and the healthcare system. Furthermore, fatigue may negatively affect psychological traits such as self-efficacy, which is the individual’s belief in their ability to manage and perform tasks effectively. Low self-efficacy can create a vicious cycle, where fatigue reduces activity engagement, which in turn limits recovery and reinforces feelings of incapacity.^4^ The causes of PSF are believed to be complex, involving a mix of health-related factors, psychological issues, behaviour and sociodemographic influences. The condition has been linked to low-grade inflammation, disruptions in dopaminergic signalling and altered sensorimotor network dynamics, which may contribute to both the onset and maintenance of fatigue. Given its substantial impact and the growing burden of stroke worldwide, understanding the prevalence and broader consequences of PSF is essential for improving poststroke care and rehabilitation strategies.^5 6^

While different international studies have highlighted the prevalence and negative impact of PSF, limited data exist from low-income countries, particularly in sub-Saharan Africa. Moreover, in Ethiopia, and in the study area, routine stroke rehabilitation does not consistently address PSF among stroke survivors. Most stroke rehabilitation services in Ethiopia focus on physical recovery, with limited integration of fatigue assessment during the subacute and chronic phases of care. As a result, its burden and impact among stroke survivors are not clearly understood. So that, with an increasing number of stroke survivors in our country and in the study area, identifying PSF prevalence, factors and its impact on stroke survivors is critical for optimal care. This study aimed to investigate the frequency, associated factors and impact of PSF on functional ability and self-efficacy. Recognising its impact is essential to direct attention and resources towards the most affected areas of recovery. Also, by recognising risk factors and patterns of fatigue during the subacute and chronic rehabilitation phases, healthcare and rehabilitation providers can devise targeted therapies. These changes pave the way for better management and, ultimately, help stroke patients restore their quality of life, as it is the first in the study area and in Ethiopia to quantify the impact of fatigue among stroke survivors.

Method and material

Study area, design and period

An institutional-based cross-sectional study was conducted at Jimma University Medical Center (JUMC), physiotherapy ward. Jimma is one of the main towns in the country, 346 km southwest of Ethiopia’s capital, Addis Ababa. The settlement is geographically located in the latitude and longitude of 7°40′N and 36°50′E. The study was conducted in JUMC from January to July 2025.

Eligibility criteria

Stroke survivors aged ≥18 years, who were beyond the acute phase, able to communicate and able to complete the interview, and who attended JUMC during the study period, were included.

Sample size determination and sampling technique

The sample size was calculated based on the findings of a previous study conducted in Amhara regional state that included 399 participants, in which the overall prevalence of fatigue among chronic stroke patients was reported to be 62.2%.^7^

n=Zα/22P(1-P)d2=363.5.^8^ This is approximately 364 participants. To consider 10% non-response, we use n=n1-nonresponserate= 404.

The sample size was also calculated for second objectives based on the study conducted in Amhara regional state among stroke survivors, in which older age was recognised as one of the more important factors.^7^ By assuming a 5% margin of error (d), 95% confidence level (alpha, α=0.05, two-tailed) and 80% power (1−β) to detect the assumed difference. This is calculated using Epi Info V.7, Z=CI interval at 95%=1.96 the sample will be below 404, so the final sample size for this study conducted in JUMC, physiotherapy ward was 404.

A systematic random sampling method was conducted to select study participants. The sampling interval, Kth, was calculated by dividing the total number of study participants in JUMC during the study period by the study sample (N/n), which is 2. The first patient was selected using the lottery method between one and ‘k’=1–2, and it was 2, and other patients were selected within every second value until the required sample size was obtained in each hospital.

Data collection procedures and data quality

Data were collected through face-to-face interviews and patient medical chart reviews. A semistructured questionnaire was formed using a literature review.4912 The questionnaire had six Part I: sociodemographic variables sections, Part II: health-related factors questionnaire sections (comorbidity, type of stroke, location of stroke, National Institutes of Health Stroke Scale (NIHSS) during admission, rehabilitation follow-up), were recorded from patient’s medical charts, which is the most recent Part III: lifestyle and psychological factors questionnaire sections (sleep quality, alcohol consumption, smoking, anxiety and depression), were recorded using self-reported assessment tools, Part IV: Fatigue Severity Scale (FSS), Part V: Nottingham Extended Activities of Daily Living and Part VI: Stroke Self-Efficacy Questionnaire.

Data quality was monitored starting from the questionnaire. Two translators translated the questionnaire into Amharic, the local language Afan-Oromo, and then translated it back into English to verify consistency and data quality. To ensure data quality, the principal investigator trained all data collectors for 1 day on how to approach and gather data on study participants, use the questionnaire, obtain informed consent from study participants and address significant issues related to data collection. Prior to data collection, the instrument was pretested using 5% of the overall sample size to ensure response accuracy, language clarity, consistency and appropriateness.

Operational definition

Fatigue: According to the FSS, a score of ≥36 is indicative of fatigue.^9^

Functional by Nottingham Extended Activities of Daily Living and classified as high function (44–46), moderate function (22–43) and low functional ability (≤21) on IADL.^13^

Self-efficacy: based on Stroke Self-efficacy Questionnaire 0–13 as low, 14–26 moderate and 27–39 high self-efficacy.^14^

Physiotherapy follow-up:planned inpatient physiotherapy; follow-up within 7 days after discharge if outpatient; 2–3 sessions/week with reassessment every 2–4 weeks in the subacute phase; and, depending on progress, follow-up in the chronic phase.^15^

Anxiety and poststroke anxiety and depression with a score ≥8 on the Hospital Anxiety Disorder Scale (HAD-A) and Hospital Depression Scale (HAD-D), respectively.^16^

Poor sleep indicated by a Pittsburgh Sleep Quality Index score of more than 5.^17^

Smoker: who has smoked more than 100 cigarettes in their lifetime and currently smokes every day.^1^

Alcohol based on the Alcohol Use Disorders Identification Test-Concise, a screening tool to assess alcohol consumption and drinking behaviours, a score ≥5 from a total score of 12 is considered as alcohol consumers or alcohol users.^8^

Analysis Methods

Descriptive statistics were done for all the variables in the study using statistical measurements and displayed in frequency tables and charts.

Binary logistic regression analysis was used to analyse the relationship between PSF and independent variables. In the bivariate logistic regression analysis, variables with a p<0.2 were evaluated as prospective candidates in the final multivariable logistic regression analysis. The results were considered statistically significant in the multivariable logistic regression with a p<0.05. An adjusted OR (AOR) with 95% CI at a p<0.05 was analysed and reported.

To assess the impact of PSF on functional ability and self-efficacy, an ordinal logistic regression was performed. Model fit was evaluated using the Pearson and Deviance goodness-of-fit tests, and the proportional odds assumption was tested using the test of parallel lines.

Patient and public involvement

The design, conduct, reporting and dissemination strategies for this study do not include patients or the public.

RESULTS

Descriptions of study participants

Sociodemographic factors of study participants

From those 404 invited participants, 92% of them (370 poststroke patients) in the study participated. The remaining participants were unwilling to complete the interview due to personal reasons and time constraints. The participants ranged in age from 32 to 73 years, with an average age of 53.45 (±10) years, with 247 (66.8%) being male and 214 (57.8%) being above 55 years (table 1).

Health-related factors

Of the individuals in the study, 233 (63%) had proven medical comorbidities, in which most conditions are related to cardiovascular issues like hypertension, and metabolic disease such as diabetes mellitus. Most of the study participants, 236 (63.8%), had a disease duration less than 6 months, and only 133 (35.9%) had NIHSS <12. Furthermore, more than half (56.5%) of the study participants did not receive ongoing physiotherapy follow-up (table 2).

Lifestyle and Psychological factors

From participants, 169 (45.7%) smoked and 289 (78.1%) were not alcohol users. In terms of psychological traits, 246 (66.5%) of the study participants, or more than half, experienced depressed symptoms. Anxiety affected 175 (47.3%) of the study participants (table 3).

PSF prevalence among stroke survivors

Fatigue was seen in 65.4% of study subjects (95% CI 60.3% to 70.2%) (figure 1*).* This emphasises the necessity of focused interventions to help stroke survivors manage their fatigue.

Factors associated with fatigue among stroke survivors

The model’s fitness was assessed using Hosmer and Lemeshow’s goodness of fit test, which results 0.637 indicating a good fit, and the multicollinearity diagnostic was performed by the variance inflation factor (<10). Age group, body mass index (BMI), sleep quality, type of stroke, stroke duration, comorbidity, NIHSS, physiotherapy follow-up, anxiety and depression were among the factors that demonstrated significant associations at p<0.2 during bivariate logistic analysis. These variables were then taken into consideration jointly in multivariable analysis. Age group above 50 (AOR=2.29; 95% CI 1.06 to 4.91); BMI overweight (AOR=3.75; 95% CI 1.71 to 8.24) and underweight (AOR=2.44; 95% CI 1.26 to 4.72); NIHSS≥12 (AOR=4.66; 95% CI 2.52 to 8.63); lack of physiotherapy follow-up (AOR=5.07; 95% CI 2.76 to 9.31) and depression (AOR=3.53; 95% CI 1.84 to 6.79) were among the factors that demonstrated significant associations in multivariable analysis (table 4).

Impact of fatigue on functional ability (IADL)

An ordinal logistic regression was conducted to examine the impact of PSF on functional ability levels by controlling the effect of other variables, categorised as low, moderate and high functional ability based on IADL. The analysis model fit the data well, with the final model showing a significant improvement over the intercept-only model (p<0.001). Goodness-of-fit tests indicated an adequate model fit (Pearson, p=0.558; Deviance, p=0.152) and no violations of the proportional odds assumption were observed, supporting the validity of the ordinal regression model. Importantly, fatigue was found to be a significant predictor of functional ability (p<0.001). Specifically, individuals with PSF had significantly lower odds of reporting higher functional ability than those without fatigue (AOR=0.186, 95% CI 0.109 to 0.317), indicating they were approximately 5.4 times less likely to report higher levels of function (table 5).

Impact of Fatigue on Self-Efficacy

An ordinal logistic regression was used to investigate the effect of PSF on self-efficacy levels, which were classified as low, moderate or high. The analytical model accurately suited the data, and the final model outperformed the intercept-only model (p<0.001). Goodness-of-fit tests revealed a satisfactory model fit (Pearson, p=0.898; Deviance, p=0.395), and no violations of the proportional odds assumption were found (p=0.132), confirming the validity of the ordinal regression model. Fatigue was shown to significantly predict self-efficacy (p<0.001) with stroke survivors experiencing PSF being about 3.6 times less likely to report higher levels of self-efficacy compared with those without fatigue (table 5).

DISCUSSION

The purpose of this study was to determine the magnitude of fatigue during the sub-acute and chronic rehabilitation phases and its associated factors among stroke survivors attending JUMC. In this study, the prevalence of fatigue during the sub-acute and chronic stroke rehabilitation phase while attending JUMC was found to be 65.4%. The age group of patients, BMI, NHISS, depression and rehabilitation follow-up were significantly associated with PSF among stroke survivors. The prevalence estimates of PSF reported in previous systematic reviews and meta-analyses using the FSS, the Multidimensional Fatigue Inventory-20 (MFI-20) and the Fatigue Assessment Scale were 47.4%, 51.7% and 36.1%, respectively.^18^ These figures are lower than the prevalence observed in our study. This discrepancy underscores a considerable gap in the assessment and management of PSF in our country, particularly within the study area.

The prevalence of fatigue during the subacute and chronic stroke rehabilitation phase in this study was comparable with a study done in our country, the Amhara regional state, 62.2%^7^ and Egypt, 62%.^19^ This resemblance might be attributed to the use of a comparable PSF indicator for detecting fatigue, as well as a similar operational definition.

On the other hand, the prevalence of fatigue in our study was higher than the findings of a previous study in Sweden, 51%.^20^ Another study done in Denmark also had a high prevalence of PSF: 46% in 6 months after stroke and 56% in 1 year after stroke.^21^ The possible reason for the difference between our study and other studies might be due to differences in sociodemographic characteristics, study design and sample size. The difference with the above studies was also due to different measurement tools used to assess PSF; the study conducted in Denmark used the MFI, and the study conducted in Sweden used the Mental Fatigue Scale to assess fatigue.

However, the prevalence of fatigue in our study was lower than the findings of two previous studies in the Netherlands, 76%^22^ and Belgium, 71%.^23^ The possible reason for the difference between our study and other studies might be the sociodemographic difference and sample size difference. The studies done in the Netherlands and Belgium (62 and 84 participants) had fewer participants than our study, in which 370 poststroke patients participated.

In this study, BMI was a significant predictor of PSF among stroke survivors, where overweight respondents were 3.7 times and underweight respondents were 2.4 times more likely to have PSF than those who had normal BMI. The result of this study was supported by a study done in China that provides an overview of the relationship between obesity and PSF.^24^ BMI can influence the risk and severity of PSF; overweight individuals may be more prone to PSF due to increased systemic inflammation, metabolic dysregulation and cardiovascular strain. Excess adipose tissue, especially visceral fat, is known to produce pro-inflammatory cytokines such as interleukin-6 (IL-6) and tumour necrosis factor-alpha (TNF-α), which have been linked to fatigue and poorer neurological recovery poststroke.^25^ On the other hand, underweight patients may experience higher PSF due to malnutrition, muscle wasting and reduced physiological reserves, which compromise the body’s ability to recover from neurological insults. Malnutrition is common in underweight individuals and is associated with deficiencies in essential nutrients such as B vitamins and iron, which are important for energy metabolism and cognitive function. Additionally, low muscle mass, often seen in underweight individuals, contributes to physical fatigue due to decreased strength and endurance.^26^

According to our study, old age was significantly associated with PSF. Our study implies that participants who were aged >50 were 2.5 times more likely to fatigue after stroke as compared with those aged below 35. A systematic review and meta-analysis also supported the finding in this study.^27^ As individuals age, they may naturally experience disrupting neurotransmitter systems, and with advancing age, reduced cardiovascular and pulmonary reserve, as well as muscle loss, may result. This lowers physiological resilience and increases energy required for everyday tasks, making older individuals more vulnerable to fatigue.^28^

In the current study, patients with depression symptoms were 3.5 times more likely to have PSF than those who did not experience depression. By a systematic review and meta-analysis done to identify the risk factors contributing to PSF, depression was associated with PSF.^29^ This relationship between depression and PSF was also supported by a cross-sectional study done in the Amhara regional state, Ethiopia, and a systematic review and meta-analysis conducted in China.^7 30^ This could be because depression is linked to higher levels of pro-inflammatory cytokines (eg, TNF-α, IL-1β and IL-6). Such inflammatory reactions can have a detrimental impact on mood, motivation and physical energy, which in turn can contribute to PSF.^31^

About the NIHSS, high NIHSS was positively associated with PSF, in which PSF was about 4.4 times more common among participants who had NIHSS≥12 during admission. The result was in line with a systematic review and meta-analysis done to identify the risk factors contributing to PSF.^29^ Because severe strokes often result in significant damage to brain regions responsible for motor control, cognition and emotional regulation.^32^

Regarding rehabilitation follow-up, patients who did not adhere to physiotherapy follow-up were found to be 5.2 times more likely to experience PSF. This finding is consistent with a study conducted in the USA.^33^ PSF was considerably more likely to occur in patients who did not complete their poststroke physical treatment because of several interconnected reasons. To improve cardiovascular fitness, restore physical function and increase general energy levels, physiotherapy is essential. This may be due to some of the central (disturbances in cerebral blood flow, cellular energy reserves and neuronal circuit activity) and peripheral (aerobic deconditioning and muscular atrophy) issues that may be linked to PSF and lessened by exercise.^6^

In addition to the variables that remained significantly associated with PSF in the multivariable analysis, several others including type of stroke, duration since stroke, comorbidity, anxiety and sleep quality showed associations in the bivariate analysis but did not retain statistical significance after adjusting for confounders. Notably, sleep quality showed a borderline association, suggesting its potential relevance in the development or persistence of fatigue among stroke survivors. Sleep disturbances are common after stroke and may contribute to fatigue by impairing physical recovery, cognitive function and emotional well-being.^34^ Anxiety and comorbidities may contribute to fatigue due to their physical and psychological burden.^35 36^ The type and duration of stroke may also influence fatigue through their effects on recovery patterns and neurological damage.^37^ Overall, these findings highlight the multifactorial nature of PSF and the need for holistic approaches to its assessment and management, with attention to potentially modifiable factors.

The present study demonstrates a significant impact of PSF on both functional ability and self-efficacy among stroke survivors. The result of this study indicated that individuals with PSF were substantially less likely to report higher functional ability levels, with an odds of 5.4, highlighting the profound negative effect fatigue can have on daily functioning. This finding underscores the importance of addressing fatigue in poststroke rehabilitation, as it appears to considerably hinder the capacity to perform IADL. PSF may impair functional ability through multiple, interlinked pathways. It can reduce the energy reserves and motivation necessary for engaging in rehabilitation or daily tasks, thereby slowing gains in strength, balance and mobility. Recent mechanistic insights suggest that PSF may stem from underlying neurobiological disruptions, including low-grade inflammation, dopaminergic system dysfunction and impaired brain network connectivity, all of which can negatively influence neuroplasticity and functional recovery.^6^

Additionally, fatigue was also a significant predictor of self-efficacy, with those experiencing fatigue being nearly 3.6 times less likely to report higher self-efficacy. This association can be understood through the way fatigue influences psychological resilience and perceived capability. Persistent fatigue can undermine a person’s belief in their ability to manage everyday tasks, particularly when physical or cognitive effort consistently results in exhaustion or failure. Over time, these experiences may lead to a reduced sense of control and confidence, both of which are critical components of self-efficacy. Fatigue also contributes to emotional distress, reduced motivation and a passive coping style, which further diminish one’s belief in their ability to influence outcomes in rehabilitation or daily life. Moreover, fatigue-related impairments in attention and executive function can hinder planning and decision-making, reinforcing a perception of helplessness. These psychological and behavioural patterns reflect a broader interaction between physical symptoms and self-perception, in which fatigue plays a central role in lowering self-efficacy among stroke survivors.^38^

This suggests that PSF not only limits functioning but may also diminish confidence in managing poststroke challenges, potentially affecting motivation and engagement in rehabilitation programmes. The impact of fatigue on functional level and self-efficacy in this study was supported by a systematic review and meta-analysis done in Australia.^39^ The relationship between fatigue, self-efficacy and activity of daily living was also supported by a study done in Ireland.^4^

Conclusion and Recommendation

This study revealed that PSF is a highly prevalent and significant concern among stroke survivors at JUMC, with a prevalence rate of 65.4%. Several factors were found to be significantly associated with PSF, including older age, abnormal BMI (both underweight and overweight), depression symptoms, higher stroke severity (NIHSS ≥12) and poor adherence to rehabilitation follow-up. The findings suggest that PSF has a substantial impact on both functional ability and self-efficacy, thereby affecting recovery and quality of life.

Given its strong association with modifiable factors such as depression and rehabilitation adherence, early identification and targeted intervention for PSF should be a priority in stroke rehabilitation programmes. Integrating psychological support, nutritional management and consistent physiotherapy follow-up may help reduce fatigue and enhance functional outcomes for stroke survivors. Future longitudinal and interventional studies are recommended to further explore causality mechanisms and evaluate effective strategies to manage PSF in this population.

Strengths and Limitations of the Study

This study addresses a significant health issue influencing stroke survivors by focusing on the main concerns that happen during stroke recovery. It also addressed the impact on functional activity and self-efficacy. For the sake of further studies, it is crucial to acknowledge a few limitations. The study may miss important trends or changes that occur on PSF over time.