Factors Determining Not Returning to Full-Time Work 12 Months After Mild Ischemic Stroke

Highlights • A total of 40% of the patients did not return to full-time work 12 months after a first-ever mild ischemic stroke.• Low functional level expressed as modified Rankin scale score >1 at 12-month follow-up was significantly associated with not returning to full-time work.• Diabetes mellitus was a borderline significant factor of not returning to full-time work.

Unemployment is associated with poor physical health 1 and reduced quality of life. 2 For some patients with stroke, work can be a cause of stress and therefore potentially a risk, whereas for others it is a way of demonstrating recovery. 3 It has previously been found that 53% of patients returned to work 1 year after stroke, 4 with the proportion being lower in women than in men. 5 Another study 6 indicated that 47% of patients working fulltime prestroke were still working fulltime 1 year later, while 27% were no longer in work, and 24% were working fewer hours a week. Predictors of returning to work were less severe neurologic deficits, better cognitive ability, greater independency in daily life activities, age younger than 51 years, and male sex. 4,7 Psychiatric morbidity, depressive symptoms, and poststroke fatigue might hinder younger patients from returning to professional activity for as long as 2 years after first-ever mild stroke. 6,[8][9][10] Functional outcome after stroke has been improved during the last years because of both intravenous thrombolysis and mechanical thrombectomy. [11][12][13] Thrombolysis for firstever ischemic stroke was independently associated with a reduced rate of dementia in one study. 14 The relationship between thrombolytic treatment and return to paid work has not been much investigated, 15 but a recent article 16 reported a high proportion of patients treated with thrombolysis returning to work (64%) 1 year after stroke. Age 41-60 years, high stroke severity expressed as National Institutes of Health Stroke Scale (NIHSS) 17 score ≥5, and female sex were associated with lower odds of workforce attachment. 16 The effect of endovascular stroke treatment on work ability is largely unknown.
One incentive to initiate this study was our own experience from the stroke outpatient clinic, where we observed that several young working patients with minimal or no neurologic sequelae after a mild stroke reported difficulties involving cognitive functions. In addition, some patients struggled with fatigue, indifference or apathy, anxiety, depression, or mood swings. It was previously shown that hidden outcomes can hinder the patients from returning to normal daily activities and affect both family and work life. 18 Unemployment is often associated with poorer physical 1 and mental health 19 and reduced quality of life. 2 In a Dutch study, one-third of the participants who had not gone back to work 1 year after a mild-to-moderate stroke were unsatisfied with their occupational situation. 6 In addition, the work satisfaction of working patients increased with the daily number of hours spent in a paid work.
Most studies about returning to work after stroke have included patients with mild-to-moderate ischemic and hemorrhagic strokes defined as NIHSS score ≤15, 20 while our study is one of the few studies that has included patients with only mild ischemic strokes defined as NIHSS score ≤3. Other authors 21 have found that stroke size, localization, and etiology, i.e. large anterior strokes, and stroke caused by large artery atherosclerosis and cardioembolism, are associated with not returning to work 1 year after ischemic stroke. We previously 22 found that younger age at stroke onset and finding of multiple cerebral infarctions were associated with cognitive impairment 1 year after stroke.
Overall, the aim of this prospective cohort study was to identify the prevalence of not returning to full-time work and to study whether factors such as demographic characteristics, stroke etiology, localization, and reperfusion therapy were associated with not returning to full-time work 12 months after first-ever mild stroke in patients 70 years or younger.

Methods Population
The included patients were admitted with an acute stroke to the stroke units at Oslo University Hospital, Norway and Baerum Hospital, Norway in the period from December 2014 until December 2016. Details regarding the study design have previously been published. 23 In this substudy, we included patients employed fulltime, prescribed by law in Norway as working at least 37.5 hours a week, at stroke onset. Inclusion criteria were age 18-70 years and that the participants had a mild first-ever ischemic stroke defined as a NIHSS 17 score ≤3 points at discharge. 24 Patients who did not speak Norwegian, patients with diagnosed dementia or cognitive impairment defined as a score > 3.2 on the short form of the Informant Questionnaire on Cognitive Decline in the Elderly, 25 and patients with known psychiatric disease were not eligible. Patients who had a new stroke during the first year after stroke were not invited to follow-up.

Assessments
We recorded age, sex, educational level in years, and marital state at stroke onset in addition to vascular risk factors as treated hypertension, hyperlipidemia, coronary heart disease, atrial fibrillation, cigarette smoking, diabetes mellitus, body mass index, and presence of any apolipoprotein E-epsilon 4 alleles.
The diagnosis of ischemic stroke was based on the medical history, findings on the neurologic examination, and findings of an acute infarction on cerebral computed tomography (CT) or magnetic resonance imaging (MRI) scans. Patients presenting with neurologic deficits lasting more than 24 hours, but with no acute lesions on CT or MRI, were also considered to have an ischemic stroke. 26 The ischemic strokes were classified by etiology according to the Trial of ORG 10172 in Acute Stroke Treatment (TOAST) classification. 27 The Barthel Index score and the modified Rankin scale (mRS) score 28,29 was used to assess the general functional level at discharge and at 12-month follow-up.

Twelve-month follow-up
Study participants were invited to follow-up in the outpatient clinic. The cognitive and emotional assessments were performed by either a stroke physician, an occupational therapist, or a research nurse. These professionals had been trained by experienced physicians with high competence and interest in the field.

Cognitive and emotional assessments
Cognitive and emotional functions were evaluated using an extended battery of screening tests. 23 The Mini-Mental Status Examination Norwegian Revision 2 (MMSE-NR2) is considered to be a screening test of global cognitive functioning. 30 To examine particular cognitive domains, we used the Clock Drawing Test 31 (executive and visuospatial function), the Trail Making Test (TMT) A 32 (focused attention, psychomotor speed) and B (divided attention, executive function, psychomotor speed), the Controlled Oral World Association (COWA) Verbal Fluency Test 33 (attention, executive function, psychomotor speed, language ability), the Rey-Osterrieth Complex Figure Test (ROCF) 34 (executive and visuospatial function, memory), and the California Verbal Learning Test II (CVLT II) 35 (learning ability, memory). Because of the participants' high educational level and low mean age (52.4 years), we chose a higher score (<27/30 points) 36 than usual as a cutoff for the MMSE-NR2. 30 For the Clock Drawing Test, the standard cutoff of 4 of 5 points was used, 31 while the correct administrations within 60 seconds for the TMT A and 120 seconds for the TMT B were chosen as cutoffs. 32 Because of the sparse evidence for differences related to age and sex on CVLT II scores, to age on ROCF, and to sex and education level on COWA and Trail Making Tests, raw scores were converted into standardized scores (mean §SD, 50 §10) based on appropriate normative sample according to test manuals used by the Norwegian Registry of Persons Assessed for Cognitive Symptoms in Specialist Health Care services. In accordance with the criteria of the International Society for Vascular Behavioral and Cognitive Disorders, 37 mild cognitive impairment can be present when the performance on validated cognitive tests in 1 or more of the cognitive domains of attention, processing speed, executive function, learning and memory, language, visuoconstructional-perceptual ability, praxis-gnosis-body schema, and social cognition is 1-2 SD below appropriate norms. Accordingly, we defined a score of at least À1 SD or any score outside the reference range on at least 1 of the cognitive tools as indicative for impairment of that particular cognitive domain. 23 The presence of depression and anxiety symptoms, fatigue, and apathy was defined using well-established cutoffs of the Hospital Anxiety and Depression Scale (HADS) (>7 points for the depression subscale; >7 points for the anxiety subscale), 38 the Fatigue Severity Scale (FSS) (≥4 points), 39 and the Apathy Evaluation Scale−Selfreport (AES-S) (≥34 points). 40 Details about the testing of the cognitive and the emotional domains have previously been published. 23

Statistics
The Statistical Package for Social Sciences versions 27.0 and 28.0 were used for all statistical analyses. a Descriptive statistics and table analyses were performed, and data are presented with means and SDs for continuous variables and with proportions and percentages for categorical variables. The rating scale scores are represented with median and range where the data distribution was skewed. To compare the baseline and the follow-up characteristics of the participants who returned to full-time work and those who did not, we performed chi-square (x 2 ) test for the categorical and independent samples t test for continuous variables.
Logistic regression analyses were performed separately to identify possible factors associated with the dependent variable of not returning to full-time work 12 months after a mild stroke. First, unadjusted regression analyses were performed using baseline patients' characteristics, stroke characteristics such as topography, TOAST classification, functional level expressed as NIHSS, Barthel Index score at discharge from hospital, treatment with intravenous thrombolysis and thrombectomy, mRS score at discharge and follow-up, presence of cognitive impairment, and the scores on the anxiety, depression, fatigue, and apathy scales at 12-month follow-up as independent variables in unadjusted regression analyses.
Age, sex, and variables associated with the outcome with a P value <.1 in unadjusted analyses were then entered into multivariate hierarchical logistic regression models. In model 1 potential explanatory variables were baseline characteristics, that is, age, sex, years of education, diabetes mellitus, and NIHSS score at admission and discharge. The mRS score at follow-up and the scores of the anxiety, depression, and fatigue scales at 12-month follow-up were entered in model 2. The factors that remained significant from the hierarchical logistic models were considered to be the independent predictors of the outcome. Regarding the use of the anxiety, depression, fatigue, and apathy scales, we decided not to dichotomize these variables but to use the raw scores in order not to lose information and statistical power. 41 Results are presented as odds ratio with 95% confidence interval. A P value <.05 was considered as the limit for statistical significance.

Ethics
The Regional Committee South East for Medical and Health research ethics (register no. 2014/1268) and the Oslo University Hospital's Data Protection Authority approved the study. Before inclusion, written informed consent was given by all patients.

Results
In the main study, we included 127 patients, with follow-up data on 117 of them. This substudy comprises 84 patients (17 female [20%]), with mean age 52.4 §10.8 years who worked full-time at stroke onset. A total of 22 patients (25%) were treated with intravenous thrombolysis. Four patients (5%) were treated with mechanical thrombectomy, all combined with intravenous thrombolysis. Demographic and clinical baseline characteristics are presented in Table 1. At 12-month follow-up, 78 patients (93%) attended. During the 1-year poststroke period, 1 patient died, 3 had a recurrent stroke, and 2 withdrew from the follow-up. A total of 63 patients (81%) had returned to work, 47 (60 %) of them to full-time work. Table 2 shows the distribution of the employment status as well as cognitive and emotional impairments at follow-up for all patients included in the study. Tables 3 and 4 show the baseline and follow-up characteristics of patients who returned to full-time work 1 year after stroke and those who did not.
Factors determining not returning to full-time work 12 months after a mild ischemic stroke In the adjusted regression models, a low functional level (mRS score >1) at follow-up was significantly associated with not returning to full-time work, and diabetes mellitus showed a borderline (P=.052) significant association. The relationship between not returning to full-time work and female sex, NIHSS score at discharge, and a higher score on the anxiety, depression, and fatigue scales were deflated after adjusting for relevant covariates (Table 5).

Discussion
We included patients who worked full time at onset of a first-ever mild ischemic stroke. At 12-month follow-up, 4 in 5 patients were back to any kind of employment, either full time or part time, of whom 3 in 4 worked full time. Reduced functional status (mRS score >1) at follow-up and having diabetes mellitus were associated with not returning to fulltime work after 12 months. Female sex, NIHSS score at discharge, and a high score on anxiety, depression, and fatigue scales were also associated with not returning to full-time work, although these associations were only seen in the unadjusted regression models.
It has previously been reported that approximately 50% of patients return to work 1 year after stroke, and approximately 60% return 2 years after stroke. 4,19,42 A greater percentage of our patients returned to paid work. This may not fully be explained by the young age in our sample because other studies 4 also report on an average age between 37 and 55 years. Tables 1 and 3 show that the percentage of patients 45 years or younger (14/26 patients [54%]) returning to full-time work is nearly the same as the percentage of patients older than 45 years (33/58 [57%]). Generally, young working patients wish to return to their job and daily life as soon as possible after stroke. 19 Older employees may be unable to work because of age-related changes or other health issues or prefer not to work as much as they did before stroke onset. 43 On the other hand, it is possible that employers as well as the rehabilitation system focus more on younger than older working patients. Not returning to full-time work after stroke We found that low functional status at 12-month followup, operationalized as an mRS score >1, was an independent associative factor of not returning to full-time work. This may be because high mRS score expresses a higher dependence in daily life, which hinders patients from returning to work. An Indian study 44 showed that functional disability after stroke and the type of job can determine return to work rather than psychosocial factors such as depression and anxiety. In addition, functional level at discharge 19 , 1 month 45 , and 3 months 46 post stroke can affect work participation.
In patients with stroke, diabetes mellitus has been shown to be both a risk factor for cognitive impairment 47 and reduced ability to return to work. 6,8,19,43 We found that diabetes is borderline significantly associated with not returning to full-time work, but we did not find any significant relationship between cognitive impairment and unemployment 12 months after stroke. Anxiety, depressive symptoms, and fatigue were related to not returning to full-time work but only in the unadjusted analyses. Such invisible emotional impairments have previously been described as factors affecting both quality of life 48 , work, and social participation 10,42,46,49 after stroke. High education (especially at university level) is related to high socioeconomical status, private insurance, and high income, and these factors are associated with a higher probability of returning to work up to 4 years after stroke onset 19,45,50 . In our study, we did not show that educational level was associated with returning to full-time work. However, other socioeconomic factors, such as patients' income and profession were not recorded, and so the importance of these factors on returning to full-time work remains unclear. According to data from several studies 19,43,45,50 , men seem to return to work sooner than women. One explanation might be that women are less likely to have a favorable outcome 51 and a good health-related quality of life 48 after stroke than men. In our study there were fewer women (56%) with a favorable functional outcome (mRS score 0-1) than men (79%).
Some studies have identified a relationship between living alone and not returning to any kind of paid work. 44,52 Possible explanations for these findings could be that people who live alone after stroke may have more anxiety and less support and encouragement from a partner in their daily life 52 . We did not, however, find that living alone at stroke onset predicts not returning to full-time work.
The effect of reperfusion therapy on the outcome after stroke is widely investigated during the last years 12 , but there are only a few studies 15,16 that describe that treatment of acute ischemic stroke with intravenous thrombolysis may be a positive predictor of returning to full-time work after stroke. Approximately 50% of our patients treated with intravenous thrombolysis returned to full-time work 1 year later, but the association between thrombolysis and return to work was not statistically significant. This may be because of low power related to few patients or the study sample with patients who had a high functional level expressed as mRS score ≤1 at discharge (74%).

Study limitations and strengths
One limitation was that only 1 in 5 participants were female. The large number of male patients younger than 70 years, however, is in line with stroke epidemiology. 53 One of our inclusion criteria was that patients were previously cognitively intact. To reduce the possibility of underreported prestroke cognitive decline in some patients, the information about cognitive health was obtained through interviews with patients, their next of kin, and their family physicians and by reviewing patients' medical records. The dependents of all patients filled in the Informant Questionnaire on Cognitive Decline in the Elderly. 25 Strengths of the study are the clear diagnosis and classification of a first-ever mild stroke. We used validated and widely used cognitive and emotional instruments to evaluate several cognitive and emotional domains. The tests were performed by either a physician, occupational therapist, or nurse. We did not perform any intrarater reliability tests of the tools used in our study, and this may represent a limitation.
Almost all patients aged 18-70 years admitted to the 2 stroke units during the inclusion period and who had been evaluated as eligible for inclusion, wished to participate. Only 3 of the included patients withdrew their consent before follow-up. Consequently, attrition bias was low.

Conclusions
Younger patients can have difficulties with returning to paid work even after a mild stroke without visible  .003* NOTE. Cognitive impairment: a score outside the reference range on at least 1 of the used cognitive tests. 20 Abbreviations: AES-S, Apathy Evaluation Scale−Self-report; APOE-e4, apolipoprotein E-epsilon 4; BMI, body mass index (calculated as weight in kilograms divided by height in meters squared); CI, confidence interval; CT, computed tomography; FSS, Fatigue Severity Scale; HADS, Hospital Anxiety and Depression Scale; MRI, magnetic resonance imaging; OR, odds ratio; TOAST, Trial of Org 10172 in Acute Stroke Treatment. * Statistical differences (P<.05).