Post-Stroke Fatigue and Functional Outcome in Young Adults
Post-Stroke Fatigue and Functional Outcome in Young Adults
We found that fatigue was a very common symptom in young patients with stroke. Even after almost a decade of follow-up, fatigue was associated with a poor functional outcome.
Strengths of our study include the large sample size and long follow-up, with inclusion of both ischaemic stroke and patients with TIA and the comparison with a stroke-free control group. Furthermore, the single centre design allowed us to collect information systematically, limiting information bias. Being a tertiary referral centre, our cohort is a representative sample of a Dutch young stroke population, since those young patients with stroke who survive usually visit a university medical centre during the course of their disease.
However, some limitations need to be addressed. Selection bias might have occurred due to patients who refused to participate, patients lost to follow-up, and patients who were deceased. It might be that higher levels of fatigue were an important reason for patients to refuse to participate. This would result in an underestimation of the prevalence of fatigue in our cohort. However, including these patients would then only have magnified the differences we found between patients and controls, and not alter our conclusions.
Patients who were lost to follow-up and deceased patients more often had an ischaemic stroke than a TIA and higher NIHSS-scores. These variables were however not associated with a higher prevalence of fatigue. Furthermore, we have no reason to assume that the nature of the relation between fatigue and functional outcome differs in participants from that in any of the three groups of non-participants (eg, refusers, lost to follow-up and deceased).
Confounding is another bias to be considered. The most important confounders in the relation between fatigue and functional outcome are the presence of depressive symptoms and anxiety. We tried to overcome this bias by making adjustments for the presence of depressive symptoms and anxiety in our analysis.
Statistical power was limited in patients with TIA, so we had to include the variables 'depressive symptoms' and 'anxiety' as continuous scores on the HADS in the model investigating their association with fatigue. They were found to be independently related to fatigue, with modestly elevated ORs per point on the HADS-score. If there were more patients in the TIA-group, and we could have included depressive symptoms and anxiety as dichotomous variables, we expect that this would have resulted in larger ORs, showing even more clearly that these symptoms are inter-related. Therefore, this probably would not have altered conclusions.
One previous study in young patients with stroke reported a prevalence of fatigue of 51.3% versus 31.6% in controls. These figures were higher than we found (41% vs 18.4%). This difference may be explained by the different methods that were used to assess fatigue. It might be that the threshold of detecting fatigue with our questionnaire is higher, but we are not aware of studies that directly compare the two different methods that assess fatigue (Fatigue Severity Scale (FSS) and CIS).
A surprising finding of our study was that fatigue was equally common in young patients with TIA as in young patients with ischaemic stroke. This may suggest that the size of the structural lesion caused by the initial TIA or stroke does not play a role in the presence of fatigue after long-term follow-up. In addition, we did not find any difference in prevalence of fatigue in different lesion locations. Our findings are in contrast with previous short-term follow-up studies. One study found that fatigue was more common after a minor stroke than after a TIA. Other short-term follow-up studies found that the lesion location, namely infratentorial or in the basal ganglia, was associated with the presence of fatigue. Our findings indicate that on the long term, initial stroke characteristics (eg, volume of brain damage or lesion location) are not an important determinant of the presence of fatigue, in contrast to previous studies that found stroke characteristics to be associated with fatigue shortly after the stroke.
Whereas the initial stroke may not have a large role in long-term post-stroke fatigue anymore, accumulating structural brain damage over the years may have. This is indeed suggested by our finding that recurrent cerebrovascular events were associated with the presence of fatigue (in patients with an ischaemic stroke). However, one would then also expect that the prevalence of fatigue would rise with longer follow-up, during which recurrent cerebrovascular events occur. We did however not find differences in prevalence during different tertiles of follow-up.
Our findings that stroke subtype (TIA or ischaemic stroke), lesion location and recurrent cerebrovascular events were not unequivocally related to the presence of fatigue, suggest that other factors may play an important role, for example psychosocial factors. In our study, depressive symptoms and anxiety were associated with the presence of fatigue. These symptoms often co-occur. A study in patients with a mean age of 68 years found that fatigue was associated with poor physical health 1.5 years after a stroke, independent of depressive symptoms. However, younger patients might have to cope with higher demands than older stroke survivors, due to active work participation and a young family. Therefore, results from studies in older stroke survivors cannot necessarily be extrapolated to young patients with stroke. Our study shows that fatigue was related physical health in young stroke survivors, measured with a motor performance scale (mRS), independently from depressive symptoms or anxiety. Moreover, we found that fatigue was also associated with impairment of speed of information processing and showed a trend towards a significant association with impaired working memory. One previous study in older patients with stroke found no longitudinal relationship between fatigue and mental health. This differs from the findings from our study, but this is probably explained by the fact that we assessed each cognitive domain separately, instead of one global score for cognitive performance. Although we found that fatigue was associated or showed a trend towards an association with only two cognitive domains (speed of information processing and working memory), impairment in these domains may very much interfere with regaining prestroke activities, for example returning to work.
One other study also found a relation between fatigue and poor functional physical outcome in young patients with ischaemic stroke. Since our study shows that this association also holds true for cognitive impairment, our study underlines the fact that the less visible (psychological) post-stroke consequences are at least as important to address in rehabilitation programmes as are the more visible motor symptoms.
Our findings can be used to properly inform young stroke survivors and caregivers, which may be a first step in optimising rehabilitation programmes. Although different kinds of treatments for post-stroke fatigue are in development for the general stroke population, no treatments have been proven unequivocally successful in clinical trials in these patients, let alone in specific young stroke cohorts.
In conclusion, even after almost a decade of follow-up after a stroke at young age, fatigue was present in approximately 40% of patients. This fatigue was associated with a poor functional outcome. Therefore, reducing fatigue can be an important tool in increasing functional independence in young stroke survivors.
Discussion
We found that fatigue was a very common symptom in young patients with stroke. Even after almost a decade of follow-up, fatigue was associated with a poor functional outcome.
Strengths of our study include the large sample size and long follow-up, with inclusion of both ischaemic stroke and patients with TIA and the comparison with a stroke-free control group. Furthermore, the single centre design allowed us to collect information systematically, limiting information bias. Being a tertiary referral centre, our cohort is a representative sample of a Dutch young stroke population, since those young patients with stroke who survive usually visit a university medical centre during the course of their disease.
However, some limitations need to be addressed. Selection bias might have occurred due to patients who refused to participate, patients lost to follow-up, and patients who were deceased. It might be that higher levels of fatigue were an important reason for patients to refuse to participate. This would result in an underestimation of the prevalence of fatigue in our cohort. However, including these patients would then only have magnified the differences we found between patients and controls, and not alter our conclusions.
Patients who were lost to follow-up and deceased patients more often had an ischaemic stroke than a TIA and higher NIHSS-scores. These variables were however not associated with a higher prevalence of fatigue. Furthermore, we have no reason to assume that the nature of the relation between fatigue and functional outcome differs in participants from that in any of the three groups of non-participants (eg, refusers, lost to follow-up and deceased).
Confounding is another bias to be considered. The most important confounders in the relation between fatigue and functional outcome are the presence of depressive symptoms and anxiety. We tried to overcome this bias by making adjustments for the presence of depressive symptoms and anxiety in our analysis.
Statistical power was limited in patients with TIA, so we had to include the variables 'depressive symptoms' and 'anxiety' as continuous scores on the HADS in the model investigating their association with fatigue. They were found to be independently related to fatigue, with modestly elevated ORs per point on the HADS-score. If there were more patients in the TIA-group, and we could have included depressive symptoms and anxiety as dichotomous variables, we expect that this would have resulted in larger ORs, showing even more clearly that these symptoms are inter-related. Therefore, this probably would not have altered conclusions.
One previous study in young patients with stroke reported a prevalence of fatigue of 51.3% versus 31.6% in controls. These figures were higher than we found (41% vs 18.4%). This difference may be explained by the different methods that were used to assess fatigue. It might be that the threshold of detecting fatigue with our questionnaire is higher, but we are not aware of studies that directly compare the two different methods that assess fatigue (Fatigue Severity Scale (FSS) and CIS).
A surprising finding of our study was that fatigue was equally common in young patients with TIA as in young patients with ischaemic stroke. This may suggest that the size of the structural lesion caused by the initial TIA or stroke does not play a role in the presence of fatigue after long-term follow-up. In addition, we did not find any difference in prevalence of fatigue in different lesion locations. Our findings are in contrast with previous short-term follow-up studies. One study found that fatigue was more common after a minor stroke than after a TIA. Other short-term follow-up studies found that the lesion location, namely infratentorial or in the basal ganglia, was associated with the presence of fatigue. Our findings indicate that on the long term, initial stroke characteristics (eg, volume of brain damage or lesion location) are not an important determinant of the presence of fatigue, in contrast to previous studies that found stroke characteristics to be associated with fatigue shortly after the stroke.
Whereas the initial stroke may not have a large role in long-term post-stroke fatigue anymore, accumulating structural brain damage over the years may have. This is indeed suggested by our finding that recurrent cerebrovascular events were associated with the presence of fatigue (in patients with an ischaemic stroke). However, one would then also expect that the prevalence of fatigue would rise with longer follow-up, during which recurrent cerebrovascular events occur. We did however not find differences in prevalence during different tertiles of follow-up.
Our findings that stroke subtype (TIA or ischaemic stroke), lesion location and recurrent cerebrovascular events were not unequivocally related to the presence of fatigue, suggest that other factors may play an important role, for example psychosocial factors. In our study, depressive symptoms and anxiety were associated with the presence of fatigue. These symptoms often co-occur. A study in patients with a mean age of 68 years found that fatigue was associated with poor physical health 1.5 years after a stroke, independent of depressive symptoms. However, younger patients might have to cope with higher demands than older stroke survivors, due to active work participation and a young family. Therefore, results from studies in older stroke survivors cannot necessarily be extrapolated to young patients with stroke. Our study shows that fatigue was related physical health in young stroke survivors, measured with a motor performance scale (mRS), independently from depressive symptoms or anxiety. Moreover, we found that fatigue was also associated with impairment of speed of information processing and showed a trend towards a significant association with impaired working memory. One previous study in older patients with stroke found no longitudinal relationship between fatigue and mental health. This differs from the findings from our study, but this is probably explained by the fact that we assessed each cognitive domain separately, instead of one global score for cognitive performance. Although we found that fatigue was associated or showed a trend towards an association with only two cognitive domains (speed of information processing and working memory), impairment in these domains may very much interfere with regaining prestroke activities, for example returning to work.
One other study also found a relation between fatigue and poor functional physical outcome in young patients with ischaemic stroke. Since our study shows that this association also holds true for cognitive impairment, our study underlines the fact that the less visible (psychological) post-stroke consequences are at least as important to address in rehabilitation programmes as are the more visible motor symptoms.
Our findings can be used to properly inform young stroke survivors and caregivers, which may be a first step in optimising rehabilitation programmes. Although different kinds of treatments for post-stroke fatigue are in development for the general stroke population, no treatments have been proven unequivocally successful in clinical trials in these patients, let alone in specific young stroke cohorts.
In conclusion, even after almost a decade of follow-up after a stroke at young age, fatigue was present in approximately 40% of patients. This fatigue was associated with a poor functional outcome. Therefore, reducing fatigue can be an important tool in increasing functional independence in young stroke survivors.
Source...