If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Corresponding author Marte Stine Einstad, MD, NTNU-Norwegian University of Science and Technology, Faculty of Medicine and Health Sciences, N-7491 Trondheim, Norway.
Department of Neuromedicine and Movement Science, Faculty of Medicine and Health Sciences, NTNU-Norwegian University of Science and Technology, Trondheim
Department of Neuromedicine and Movement Science, Faculty of Medicine and Health Sciences, NTNU-Norwegian University of Science and Technology, Trondheim
Department of Neuromedicine and Movement Science, Faculty of Medicine and Health Sciences, NTNU-Norwegian University of Science and Technology, Trondheim
Department of Neuromedicine and Movement Science, Faculty of Medicine and Health Sciences, NTNU-Norwegian University of Science and Technology, TrondheimDepartment of Geriatric Medicine, St. Olavs Hospital, Trondheim University Hospital, Trondheim, Norway
Department of Neuromedicine and Movement Science, Faculty of Medicine and Health Sciences, NTNU-Norwegian University of Science and Technology, Trondheim
To investigate whether cognition and physical performance, both separately and combined, 3 months post stroke predict change in instrumental activities of daily living (IADL) up to 18 months and whether different paths of IADL could be identified by different scenarios, defined by combinations of high and low scores on physical performance and cognition.
Design
The study is part of the Norwegian Cognitive Impairment After Stroke study, a prospective multicenter cohort study including patients with acute stroke.
Setting
Stroke outpatient clinics at 3 university hospitals and 2 local hospitals.
Participants
Adult survivors of stroke (N=544) were followed up at 3 and 18 months after stroke. Participants’ mean ± SD age was 72.6±11.8 years, and 235 (43.2 %) were female.
Interventions
Not applicable.
Main Outcome Measures
The primary outcome was IADL as measured by Nottingham Extended Activities of Daily Living. At 3 months, Short Physical Performance Battery (SPPB) and Montreal Cognitive Assessment (MoCA) were used to assess physical performance and cognition, respectively.
Results
Mixed-effects linear regression analyses showed that the regression coefficient (95% CI) for the interaction with time was significant for MoCA, 0.238 (CI, 0.030-0.445; P=.025) but not for SPPB. The model combining SPPB and MoCA was significantly better than separate models (likelihood ratio P<.001). Overall, there was no improvement in IADL over time. A combination of SPPB and MoCA score in the upper quartile at 3 months was associated with improved IADL of 1.396 (CI, 0.252-2.540; P=.017) over time.
Conclusions
Combining measures of cognition and physical performance gave the best prediction of change in IADL. Function at 3 months seems to be predictive for long-term IADL status, which highlights the importance of targeted rehabilitation in the early and subacute phases after stroke.
Furthermore, a significant proportion of survivors of stroke without functional dependency have been found to have problems with cognition and social participation up to 5 years post stroke.
Maintaining independence in IADL is important for quality of life, reduces risk of depression and anxiety, and is associated with better health outcomes.
Factors associated with basic and instrumental activities of daily living in elderly participants of a population-based survey: the Nord-Trondelag Health Study, Norway.
Age trajectories of disability in instrumental activities of daily living and disability-free life expectancy among middle-aged and older adults in Taiwan: an 11-year longitudinal study.
In the population with stroke, factors such as age, severity of stroke, and prestroke function are well-known predictors of poststroke functional decline.
Verstraeten et al reported that patients with stroke had reduced physical, cognitive, and IADL function 3 months after the event compared with controls without stroke.
Individuals with impairments in both physical performance and cognition have been described as having a phenotype with higher risk of morbidity and progression to dementia,
Poststroke cognitive impairment is increasingly recognized as a frequent condition, with reported prevalence ranging from 37%-67% for mild cognitive impairment and 21%-42% for dementia, depending on different populations, study design, and diagnostic criteria.
Cognitive dysfunction, especially within the domains of executive function, memory, and attention, have been reported to be associated with poorer IADL function after stroke.
Long-term performance of instrumental activities of daily living in young and middle-aged stroke survivors-impact of cognitive dysfunction, emotional problems and fatigue.
Regarding physical performance as a predictor of poststroke IADL function, evidence is diverging. Some studies have reported physical performance as a predictor of IADL function,
However, studies have reported that recovery and transition between dependency and independency in activities of daily living (ADL) happen up to 12 months post stroke,
indicating substantial heterogeneity among patients. Thus, it would also be of great value to identify the recovery paths beyond 3 months for subgroups of survivors of stroke based on measures of cognition and physical performance.
The Norwegian Cognitive Impairment After Stroke (Nor-COAST) study previously reported that 103 (19%) of the participating survivors of stroke (n=553) had concomitant impairments in mobility (Short Physical Performance Battery [SPPB] <10 points) and global cognition (Montreal Cognitive Assessment [MoCA] <24 points) 3 months after stroke,
indicating that combining these measures might give a more precise prediction than applying 1 at a time.
The primary aim of this study was to investigate if measures of cognition and physical performance, separately or in combination, at 3 months after stroke can predict change in IADL function 15 months later. A secondary aim was to explore if different paths of IADL could be identified by different scenarios, defined by combinations of high and low scores on physical performance and cognition.
Methods
Study design and participants
The present study is part of Nor-COAST, a large multicenter prospective cohort study including participants from 5 different hospitals in Norway between May 2015 and March 2017.
Patients were screened for inclusion during the index stay, with follow-up assessments at 3, 18, and 36 months. Inclusion criteria were diagnosed with stroke according to the World Health Organization criteria
The World Health Organization MONICA Project (monitoring trends and determinants in cardiovascular disease): a major international collaboration. WHO MONICA Project principal investigators.
or with findings on magnetic resonance imaging compatible with intracerebral hemorrhage or infarction, symptom onset within 1 week before admission, being older than 18 years, fluency in a Scandinavian language, and living in the catchment area of the participating hospitals. Patients with less than 3 months expected survival were excluded from the study. Further details were published in the protocol article for the Nor-COAST study.
As indicated by the objective, the present study included data from the 3- and 18-month follow-up only. Hence, participants who completed Nottingham Extended Activities of Daily Living (NEADL), SPPB, and MoCA at 3-month follow-up were included (fig 1).
score at admission was used to measure stroke severity (range, 0-42), where higher score indicates more severe stroke. Functional dependency prior to the stroke was measured with modified Rankin scale (mRS) (range, 0-6),
The scale consists of 22 questions covering mobility, kitchen, domestic, and leisure activities.
Predictors
Global cognition was assessed with MoCA, a 10-item test covering 8 cognitive domains, with possible scores ranging from 0-30, including 1 additional point for education <12 years and higher scores indicating better cognition.
The SPPB is a measure of physical performance and consists of 3 tasks: 4-m preferred gait speed, balance, and sit-to-stand from chair 5 times, with 4-point scales for each task and a summary score ranging from 0-12, with higher scores indicating better physical function.
A short physical performance battery assessing lower extremity function: association with self-reported disability and prediction of mortality and nursing home admission.
Baseline characteristics were retrieved from participants, proxies, and medical records during hospital stay. At 3- and 18-month follow-up participants were assessed at a hospital outpatient clinic. NEADL score was collected by an interview of the participant or proxy. Clinical assessments were performed by trained health care personnel according to a standardized manual.
Ethical considerations
The present study was carried out according to the Declaration of Helsinki. Participation was voluntary and based on written informed consent from the participant or, in cases where participants were not able to give consent, by their proxy. The study has been approved by the Regional Committee for Medical and Health Research Ethics (REC Central 194265) and registered in ClinicalTrials.gov (NCT02650531).
Statistical analyses
Demographic and clinical characteristics were summarized using means and SDs or frequencies and percentages. Single mean imputation was carried out in cases with single items missing on NEADL at 3 months (n=13) and 18 months (n=7), and MoCA (n=4). Cases with more than 50% of the items missing were excluded from the analyses. Missing scores on the SPPB were not imputed because of too few variables being part of the total score.
We used mixed-effects linear regression models with NEADL score as the dependent variable in 3 different models. All 3 regression models included the known risk factors age, sex, stroke severity (NIHSS), and prestroke functional dependency (prestroke mRS) as covariates and patient as random effect. Further, we included time point (ie, 18 months vs 3 months) and either MoCA (model A), SPPB (model B), or both in combination (model C) and their interactions with time. The interaction between time and MoCA and/or SPPB was examined to investigate the effect of the clinical measures on change over time. Normality of residuals was confirmed by visual inspection of quantile-quantile plots. Prestroke mRS was treated as a categorical variable in the mixed-models analyses because of its nonlinear association with NEADL score. Because of few participants with an mRS score of 4 or 5, participants with a score of 3 (n=27), 4 (n=1), or 5 (n=1) were collapsed into score 3-5 in the regression analyses. The model fit was compared between the models by likelihood ratio tests, with model C defined as the reference model.
To answer the secondary aim of the study we presented the estimated level of NEADL score over time for 4 scenarios based on the model with combined assessment, namely for SPPB scores of 8 and 12 as well as for MoCA scores of 22 and 28. These correspond to the lower and upper quartiles of SPPB and MoCA in our data set. These results were obtained by using the variables SPPB score minus 8, SPPB score minus 12, MoCA score minus 22, and MoCA score minus 28 in the analyses. The 4 scenarios were then defined by combining MoCA upper quartile with SPPB lower and upper quartile and MoCA lower quartile with SPPB upper and lower quartile in the linear mixed-effects model.
Statistical significance was defined as 2-sided P<.05. Data were analyzed using SPSS version 25.0a and STATA 16.b
Results
Participant characteristics
Of 815 participants included in the Nor-COAST study, 700 were assessed at 3 months. Because of missing data on MoCA, SPPB, NIHSS, and NEADL, 544 participants were included in the present study (see fig 1). Those lost to follow-up (n=271) were significantly older (mean ± SD age 77.9±10.3 years vs 71.4±11.8 years, P<.001) and had more severe strokes (NIHSS score, 6.8±7.8 vs 3.7±4.7, P<.001).
Of the participants, 235 (43.2%) were female, and 29 (5.3%) had a prestroke mRS score >2 (table 1). Three months post stroke mean ± SD SPPB score was 9.4±3.1 and mean MoCA score was 24.4±4.8 (table 2). Mean NEADL score at 3 months was 51.4±14.1, and at 18 months it was 52.3±14.8.
In model A the interaction between MoCA and time was significant, with a coefficient of 0.268 (95% CI, 0.086-0.449; P=.004), indicating that MoCA had an effect on change in NEADL score from 3-18 months. In model B there was a significant interaction between SPPB and time (coefficient, 0.331; 95% CI, 0.042-0.619; P=.025). In model C the interaction between MoCA and time remained stable (coefficient, 0.238; 95% CI, 0.030-0.445, P=.025), whereas the interaction between SPPB and time was no longer significant (P=.431). Each model extension, that is, model C vs model A and model C vs model B, was a statistically significant improvement of the model (likelihood ratio test, all P<.001). All the results from the mixed-effects regression analyses are displayed in table 3.
Table 3Results from linear mixed-effects models with Nottingham EADL as dependent variable (N=544)
Model
Nottingham EADL
Coefficient
95% CI
P Value
Model A: Cognition
MoCA
1.000
0.772−1.226
<.001
Time
−6.225
−10.772 to −1.678
.007
Time*MoCA
0.268
0.086 to 0.449
.004
Age
−0.206
−0.289 to −0.122
<.001
Female sex
−0.494
−2.196 to 1.208
.569
NIHSS
−0.334
−0.514 to 0.154
<.001
mRS prestroke, scores
1
−3.482
5.504 to −1.460
.001
2
−5.712
−8.339 to −3.085
<.001
3-5
−16.375
−20.591 to −12.158
<.001
Model B: Physical performance
SPPB
2.571
2.271 to 2.871
<.001
Time
−2.571
−5.798 to −0.005
.050
Time*SPPB
0.331
0.042 to 0.619
.025
Age
−0.164
−0.232 to −0.096
<.001
Female sex
−2.386
−3.834 to −0.938
.001
NIHSS
−0.209
−0.361 to −0.057
.007
mRS prestroke, scores
1
−2.362
4.072 to −0.652
.007
2
−3.095
−5.337 to −0.854
.007
3-5
−11.042
−14.657 to −7.426
<.001
Model C: Combined assessment
MoCA
0.570
0.369 to 0.770
<.001
SPPB
2.364
2.058 to 2.671
<.001
Time
−6.840
−11.394 to −2.285
.003
Time*MoCA
0.238
0.030 to 0.445
.025
Time*SPPB
0.132
−0.197 to 0.461
.431
Age
−0.085
−0.154 to −0.017
.015
Female sex
−2.220
−3.602 to −0.839
.002
NIHSS
−0.162
−0.307 to −0.016
.029
mRS prestroke, scores
1
−2.230
−3.861 to −0.599
.007
2
−2.755
−4.894 to −0.615
.012
3-5
−7.972
−11.515 to −4.429
<.001
Abbreviation: EADL, Extended Activities of Daily Living
Change in IADL based on 3-month cognition and physical performance
Figure 2 illustrates the change in NEADL score over time for the 4 scenarios. Only the combination of MoCA and SPPB score in the upper quartile predicted a statistically significant improvement in NEADL score (coefficient, 1.396; 95% CI, 0.252-2.540; P=.017) (table 4).
Fig 2Predictive margins plot with 95% CI of model C with MoCA and SPPB centered on quartiles. SPPB low=8p; SPPB high=12p; MoCA low=22p; MoCA high=28p.
In this longitudinal observational study of survivors of stroke accessible for clinical assessments 3 months after stroke incident, we found that combining measures of physical performance and cognition was superior to only assessing 1 domain to predict change in IADL from 3-18 months post stroke. However, cognition appeared to be a stronger predictor than physical performance when applied in the same model (model C: Combined assessment). Overall, there was a stable IADL function from 3-18 months for all 4 predefined scenarios. High scores on both cognition and physical performance were associated with a statistically significant improvement in IADL over time. The results indicate that physical performance and cognitive status at 3 months post stroke is predictive of IADL function in the chronic phase after stroke.
Our main finding was that measures of physical performance and cognition in combination added value to the prediction of poststroke IADL function, even when adjusting for known risk factors of functional decline. This result highlights the need for clinical assessments of both cognition and global physical function in the subacute phase beyond application of traditional stroke severity scales and functional screening tools.
MoCA was the strongest predictor of change in IADL function, a finding supported by previously published results, because even in a subgroup scoring 0 points on NIHSS, cognitive dysfunction correlated with a decline in poststroke IADL function.
Long-term performance of instrumental activities of daily living in young and middle-aged stroke survivors-impact of cognitive dysfunction, emotional problems and fatigue.
Executive function, memory, and attention, which are all incorporated domains in MoCA, are reported to be the most important for maintenance of IADL function in survivors of stroke.
MoCA is a sensitive screening tool for cognitive impairment in patients with stroke, in addition to being more feasible in a clinical setting than extensive neuropsychological testing.
This is in line with our results, which showed that physical performance may, when applied alone, be an important predictor of change in IADL function. However, the prediction was better in combination with measures of cognition. This is interesting because NEADL mainly includes tasks in which physical performance seems to be more important than cognition.
Many of the motor tasks in NEADL might in fact challenge the dual task capacity and the motor-cognitive interphase in addition to physical performance, which is thought to be closely linked to cognition and cognitive reserves.
Further, our results are in line with a recently published Delphi study regarding factors important for IADL function, where cognition was listed as the most important feature, while physical performance was also reported as an essential factor.
We identified 4 distinct scenarios based on the model that included measures of both physical performance and cognition (model C: Combined assessment) where adjustments were made for other known risk factors. Functional dependency at 3 months post stroke, mainly reported as mRS, has been reported as a predictor of long-term function, in terms of ADL dependency, comorbidity, and death,
and the present study shows that applying clinical tests at 3 months adds value to the prediction in addition to known risk factors for functional decline.
Three of the 4 scenarios had stable, nonsignificant changes in IADL function from 3-18 months, indicating that the vast majority of the participants did not change in IADL function. This contrasts with the findings reported by Rejnö et al, who found that transition from ADL independency to dependency mainly happened during the first year after stroke.
Further, Buvarp et al found that patients with moderate stroke declined in functional mobility from 3-12 months post stroke, which is an argument for prolonged rehabilitation in a selected group of patients with stroke.
The observed plateau in IADL function after 3 months might be explained by the fact that most functional recovery as well as rehabilitation takes place prior to this.
The statistically significant increase in NEADL score found in the best preforming scenario did not reach clinical significance on group level but may be an expression of continuous recovery in this subgroup.
Responsiveness, minimal detectable change, and minimal clinically important difference of the Nottingham Extended Activities of Daily Living Scale in patients with improved performance after stroke rehabilitation.
Time was a statistically significant predictor of change in the scenario representing survivors of stroke with the best scores on both physical performance and cognition. This scenario had an increase in NEADL score of 1.4, which is lower than the 6 points regarded as a clinically significant change.
Responsiveness, minimal detectable change, and minimal clinically important difference of the Nottingham Extended Activities of Daily Living Scale in patients with improved performance after stroke rehabilitation.
However, approximately 15 points on the NEADL differed between this scenario and the one with scores on both MoCA and SPPB in the lower quartiles, which suggests that the level on performance-based tests at 3 months might predict the level of IADL function at which patients with stroke will stabilize in the long-term. This finding highlights the importance of high-quality acute treatment and early rehabilitation to achieve best possible cognition and physical function the first months post stroke.
To our knowledge this is the first study to date addressing the role of physical performance and cognition in prediction of IADL function, which is an important outcome, after stroke. The large sample size and the longitudinal multicenter design should also be considered a significant strength. The repeated measures allow us to identify predictors of change in IADL function, and the mixed-model statistical analyses use data from all participants, including those with partially missing data, thus avoiding loss of statistical power. NEADL is shown to have few floor or ceiling effects,
Comparison of 2 extended activities of daily living scales with the Barthel Index and predictors of their outcomes: cohort study within the South London Stroke Register (SLSR).
making it applicable for a heterogeneous population with stroke, and MoCA and SPPB are widely used feasible clinical measures and therefore easily incorporated into clinical evaluations of patients with stroke.
The cohort design of the study inhibits us concluding on causality, and there is a need for external validation to draw any conclusion on this prediction model outside of the present study sample. The Nor-COAST population is representative of the majority of the general population with stroke who experience mild to moderate strokes.
However, the present study sample was slightly younger and healthier than the original Nor-COAST cohort, and conclusions for individuals with severe stroke cannot be drawn. There are some sex biases in the application of NEADL in which several of the questions favor women, which is why sex was adjusted for in the analyses. Furthermore, NEADL includes more mobility than cognitive questions compared with other IADL tools, which might have led to an underestimation of associations with cognition.
Conclusions
Being able to predict change in IADL function after stroke is of great value to patients and health care providers. This is the first study to examine how the combination of cognition and physical performance in an early stage after a stroke predicts change in daily life activities over time. Combining measures of physical performance and cognition provided the best prediction of change in IADL over time. Cognitive and physical function at 3 months post stroke were predictive of IADL function over the next 15 months, and the observed effects of cognition over physical performance on IADL outcomes are arguments for increased attention on cognitive impairment in the acute phase.
Further research should target how measures of physical and cognitive performance in the early stage might be used to identify specific risk profiles and personalize long-term rehabilitation.
Suppliers
a.
SPSS version 25; BMI.
b.
STATA version 16; StataCorp.
Acknowledgments
We thank all participants and their families for their contributions, the Nor-COAST research group, and the study staff at St. Olavs Hospital, Haukeland University Hospital, Ålesund Hospital, Vestre Viken Hospital Trust; Bærum Hospital and Oslo University Hospital; and Ullevål for their contribution in collecting and preparing data.
References
Feigin VL
Vos T
Nichols E
et al.
The global burden of neurological disorders: translating evidence into policy.
Factors associated with basic and instrumental activities of daily living in elderly participants of a population-based survey: the Nord-Trondelag Health Study, Norway.
Age trajectories of disability in instrumental activities of daily living and disability-free life expectancy among middle-aged and older adults in Taiwan: an 11-year longitudinal study.
Long-term performance of instrumental activities of daily living in young and middle-aged stroke survivors-impact of cognitive dysfunction, emotional problems and fatigue.
The World Health Organization MONICA Project (monitoring trends and determinants in cardiovascular disease): a major international collaboration. WHO MONICA Project principal investigators.
A short physical performance battery assessing lower extremity function: association with self-reported disability and prediction of mortality and nursing home admission.
Responsiveness, minimal detectable change, and minimal clinically important difference of the Nottingham Extended Activities of Daily Living Scale in patients with improved performance after stroke rehabilitation.
Comparison of 2 extended activities of daily living scales with the Barthel Index and predictors of their outcomes: cohort study within the South London Stroke Register (SLSR).
The European Stroke Organisation Conference is hosted by the European Stroke Organisation, and in 2021 the conference was organised as a virtual conference from September 1 to 3. The poster was presented at September 3.
The Nor-COAST study is funded by the Norwegian Health Association, and additional funding was provided by the Department of Neuromedicine and Movement Science, Faculty of Medicine and Health Science, NTNU-Norwegian University of Science and Technology.
Clinical Trial Registration No.: NCT02650531.
Disclosure: I.S. has been an investigator in the drug trial Boehringer-Ingelheim 1346.0023 and on the advisory board for Biogen. The remaining authors declare no conflicts interest.
00212-X&id=gr1.jpg){kind=link}
00212-X&id=gr2.jpg){kind=link}