Clinical predictors for walking recovery within six months post stroke: A retrospective cohort study in Thailand

Abstract

BACKGROUND: Walking difficulty is one of the important neurological consequences after stroke, early prediction of factors related to walking ability may benefit for rehabilitation team to set suitable goals and discharge planning.

PURPOSE: This study was performed to identify independent prognostic determinants associated with mobility recovery from ischemic stroke during six months after onset.

METHODS: Medical records from first-ever stroke patients who were admitted to the rehabilitation unit were retrospectively reviewed. Baseline characteristics including demographic data, stroke risk factors, motor assessment and stroke-related complications were collected. Outcome measure was ability to walk. To identify clinical predictors, multivariable risk regression analysis was used for analysis.

RESULTS: Of a total 146 stroke survivors, 81 patients (55.5%) could be independent in walking and 65 (45.5%) were non-functional mobility at six months follow-up. Stepwise, multivariable risk regression analysis with generalized linear model demonstrated that follow-up leg score was the strongest independent predictor of walking with risk ratio [RR] of 2.23 (P < 0.01). The association of some stroke-related factors, aphasia and unilateral neglect were found in univariable analysis but revealed no significance from the final model.

CONCLUSION: Identification of early predictors associate walking recovery provided meaningful information for stroke care team to consider the amount of care needed and to initiate optimal plan according to realistic goal.

Keywords

Stroke walking predictor

1 Introduction

Over the last two decades, stroke has become a major cause of death and morbidity worldwide, including Thailand and developed countries [1 –3]. American Heart Association reported the incidence rate in 2011, new or recurrent stroke in the United States were approximately 795,000 per year, of which ischemic stroke accounts for the most (87%), whereas 10% are intracerebral hemorrhage, and 3% are subarachnoid hemorrhage stroke [2]. However, when compared age-adjusted incidence rates per 100,000 person-years between high-income and low-income countries during the study period from 1970 to 2008, it showed a divergent and statistically significant trend in stroke incidence rates, with a 42% reduction among high-income countries, whereas, in low to middle income countries, incidence rates increased greater than 100% [1]. In Thailand, a report from Public Health Statistics showed that stroke incidence has been increasing and mortality rate from stroke were found to have increased from 11.8/100 000 in 1983, to 24.4 in 2006 and 31.4/100 000 in 2010 but as high as 154.3/100 000 among elderly population [4]. Physical disability is a significant consequence following stroke where inability to walk is accounted for a major problem and affects throughout their lifespan. Thus, regaining mobility is a primary goal of stroke rehabilitation, since it is a key factor in determining whether a patient can live independently. Unfortunately, in developing countries such as Thailand, as intensive rehabilitation settings are not be widely provided and the current trend to shorten the length of stay in hospital by discharging to home or community care, it implies that any prognostic method and knowledge in terms of basic activities and mobility is crucial to optimize stroke management. A number of studies [5 –13] have been established to identify the variables to predict the recovery of mobility, these systematic reviews and previous studies based on cohort study revealed that age, motor performance, initial poor functional status, speech problems, memory and perceptual deficit, complications of ischemic stroke such as cerebral edema or extraparenchymal bleeding were potential prognostic indicators. However, it is still inconclusive, due to differences in selecting measurements, duration of stroke and a variety of methodology and statistical analysis [5 , 30]. Therefore, further valid prognostic research on stroke mobility outcome is still required, especially in subacute stage after stroke to initiate optimal rehabilitation according to realistic goal [21, 24]. In this study, our aim was to identify the independent prognostic determinants associated with mobility recovery in stroke during six months after onset that could extend an understanding of the prognostic factors associated with improvement in mobility and functional outcome in stroke to design appropriate rehabilitation programs, set suitable goals and enable early discharge planning.

2 Materials and methods

Data were retrospectively reviewed from first-ever ischemic stroke patients who were admitted or transferred to the rehabilitation unit, Maharaj Nakorn Chiangmai Hospital, Faculty of Medicine, Chiang Mai University from January 2010 to March 2015. All eligible subjects had to able to walk independently prior to stroke, had no disability from other causes that could limit mobility before stroke onset. Baseline characteristics; demographic data, stroke risk factors (i.e. prior-TIA, history of atrial fibrillation, hypertension, dyslipidemia etc.), co-morbidity, type of stroke were recorded on admission. Other predictors included motor assessment, stroke-related complications; speech problems, facial palsy, dysphagia, perceptual deficit (unilateral neglect or inattention), cognitive impairment (MMSE score less than 23), apraxia, visual impairment and shoulder subluxation and complications from other causes that might affect recovery were recorded during hospitalization. Reassessment of these variables were revised at follow up time within six months post onset. Brunnstrom motor assessment was used to evaluate voluntary movement [14]. They were classified into 6 subscore from 1 to 6 which score 1 was flaccid tone and no voluntary movement, score 2 – 3 represented some elicited movement but in synergy pattern and score 4 was movement occurring with deviation from synergy and score 5 – 6 were more independent from synergy and freely to perform near normal coordination with minimal spasticity. Outcome in this study included an ability to walk which was categorized as “independent walking”; able to walk on level surface, allowed using gait aid or orthotic device) and “non-functional mobility group”; wheelchair used or walk with assistance from another person, and it was recorded at discharge and follow up. All variables and outcomes were assessed by certified physicians. This study was approved by the Ethics Committee of Faculty of Medicine, Ching Mai University, Chiang Mai, Thailand, Research ID: NONE 2558-03123, on July 24, 2015.

2.1 Statistical analysis

Data analysis was performed with STATA version 12. Descriptive statistics (frequencies, percentages, means and standard deviations) were used to describe baseline characteristics and stroke risk factors. Comparison between groups were done by Fisher’s Exact test or Student’s t-test where appropriate with significance level of P < 0.05. Independent predictive variables derived from backward stepwise multivariable risk regression analyses and are shown with the risk ratio (RR) and 95% confident interval (CI). Any of demographic variables, prestroke risk factors and post-stroke complications with P < 0.1 from univariable analysis or significant clinical variables were included in the multivariable analysis model, with P < 0.05 indicating statistical significance.

3 Results

From 157 ischemic stroke patients admitted to the rehabilitation unit from January 2009 to March 2015, 11 were excluded because of recurrent stroke [3], bilateral hemispheric lesion [2], unable to walk independently prior to stroke [2] and loss to follow-up [4]. One hundred and forty-six were eligible for this study, 81 patients (55.5%) could recover to be independent walking and 65 (45.5%) were in non-functional mobility level at six months follow-up. Baseline characteristics; demographic data, premorbid health status and stroke-related characteristics were illustrated and compared between groups in Table 1. Most pre-stroke characteristics seemed to be similar with no significant differences between groups except for gender, age, and other comorbidities. Independent walking group were likely to be male, younger and ones with no other comorbidity compared to non-functional mobility group. Motor assessment during admission and within six months post onset and post stroke complications were shown in Table 2.

Table 1
Demographic data, premorbid health status and stroke-related characteristics according to walking recovery score at baseline (n = 146)

Characteristics Independent walking Non-functional P - value

(n = 81) mobility (n = 65)

Gender 0.02^*

Male 47 58.02 25 38.46

Female 34 41.98 40 61.54

Hemisphere 0.51

Left 35 43.21 27 49.23

Right 46 56.79 33 50.77

Subtype 0.27

Large artery 54 67.50 51 76.69

atherosclerosis

Cardioembolism 3 3.75 1 1.56

Small-vessel occlusion 23 28.75 12 18.75

Age, mean(±sd) 57.5 (15.05) 63.2 (11.93) 0.02^*

BP

Admission SBP, mean (±sd) 1150.3 (25.99) 153.9 (33.61) 0.48

Admission DBP 9 88.7 (21.10) 89.1 (19.96) 0.92

GCS 13.8 (2.14) 13.0 (2.40) 0.06

Prior TIA 7 43.75 9 56.25 0.43

Myocardial infarction 2 33.33 4 66.67 0.40

Atrial fibrillation 19 55.88 15 44.12 1.00

Smoking 16 51.61 15 48.39 0.69

Alcohol 17 62.96 10 37.04 0.52

Co-morbidity

Hypertension 59 52.21 54 47.79 0.17

Dyslipidemia 46 56.10 36 43.90 1.00

Diabetes mellitus 22 47.83 24 52.17 0.21

Other co-morbidity 0.04^*

No others 39 48.15 19 29.23

1 comorbidity 25 30.86 22 33.85

2 or more comorbidity 17 20.99 24 36.92

Characteristics	Independent walking		Non-functional		P - value
Gender					0.02^*
Male	47	58.02	25	38.46
Female	34	41.98	40	61.54
Hemisphere					0.51
Left	35	43.21	27	49.23
Right	46	56.79	33	50.77
Subtype					0.27
Large artery	54	67.50	51	76.69
atherosclerosis
Cardioembolism	3	3.75	1	1.56
Small-vessel occlusion	23	28.75	12	18.75
Age, mean(±sd)	57.5 (15.05)		63.2 (11.93)		0.02^*
BP
Admission SBP, mean (±sd)	1150.3 (25.99)		153.9 (33.61)		0.48
Admission DBP	9 88.7 (21.10)		89.1 (19.96)		0.92
GCS	13.8 (2.14)		13.0 (2.40)		0.06
Prior TIA	7	43.75	9	56.25	0.43
Myocardial infarction	2	33.33	4	66.67	0.40
Atrial fibrillation	19	55.88	15	44.12	1.00
Smoking	16	51.61	15	48.39	0.69
Alcohol	17	62.96	10	37.04	0.52
Co-morbidity
Hypertension	59	52.21	54	47.79	0.17
Dyslipidemia	46	56.10	36	43.90	1.00
Diabetes mellitus	22	47.83	24	52.17	0.21
Other co-morbidity	0.04^*
No others	39	48.15	19	29.23
1 comorbidity	25	30.86	22	33.85
2 or more comorbidity	17	20.99	24	36.92

Abbreviations: rt-PA, intravenous tissue plasminogen activator, TIA, transient ischemic attack. ^*p < 0.05 between group.

Table 2

Comparison of motor assessment and poststroke complications between independent walking group and non-functional mobility group

Characteristics	Independent Walking		Non-functional mobility
	N	%	N	%
Initial motor assessment
Leg^**
Score 1-2	30	39.47	49	79.03
Score 3-4	32	42.11	11	17.74
Score 5-6	14	18.42	2	3.23
Arm^**
Score 1-2	41	53.95	52	83.87
Score 3-4	20	26.32	4	6.45
Score 5-6	15	19.74	6	9.68
Follow up motor assessment
Leg^**
Score 1-2	8	11.94	31	64.58
Score 3-4	39	58.21	14	29.17
Score 5-6	20	29.85	3	6.25
Arm^**
Score 1-2	15	22.06	35	72.92
Score 3-4	31	45.59	8	16.67
Score 5-6	22	32.35	5	10.42
Stroke-related complications
Aphasia^**
Dysarthria	2	2.47	9	13.85
Motor/Sensory aphasia	8	9.88	5	7.69
Global aphasia	11	13.58	26	40.00
Dysphagia	14	17.28	20	30.77
Inattention^*	13	16.05	22	33.85
Apraxia	3	3.70	3	4.62
Depression	9	11.11	7	10.77
Cognitive impaired	27	34.18	17	26.56
Shoulder subluxation	18	22.22	15	23.08
Other complications
1 complication	21	25.93	18	27.69
2 or more complications	8	9.88	10	15.38

^*p < 0.05. ^**p < 0.001.

To develop a prediction model, we used the step-backward method of multivariable risk regression analysis with generalized linear model, variables with significant level less than 0.1 from univariable analysis and meaningful clinical variables were included in multivariable analysis. Initially, gender, age older than 72, other comorbidities prior to stroke, admission and follow-up leg and arm score below 3 and post stroke complications as aphasia, perceptual impairment (inattention) and dysphagia were selected. Final stepwise multivariable model demonstrated that the strongest independent predictors for walking recovery in this setting was having leg-score of motor assessment within six months after onset with risk ratio [RR] = 2.23, 95% confidence interval [CI] = 1.33–3.73 (P = 0.002) whereas aphasia and inattention (risk ratio of aphasia = 1.24, P = 0.072; inattention risk ratio = 1.67, P = 0.126) were promising but without statistical significance (Table 3).

Table 3

Independent predictors of walking recovery within six months post stroke using a multivariable analysis generalized linear model

Predictors	RR	95% CI	p-value
Follow-up leg score	2.23	1.33 – 3.74	0.002^**
Inattention	1.68	0.87 – 3.24	0.126
Aphasia	1.24	0.98 – 1.56	0.072

Abbreviations: RR, risk ratio; CI, confidence interval. ^**p < 0.01.

4 Discussion

This study was conducted retrospectively 5 consecutive years from all ischemic stroke patients admitted rehabilitation unit for intensive rehabilitation program, Maharaj Hospital, Faculty of Medicine, Chiang Mai University, Thailand. At six months post stroke, 55.5 percent were able to walk independently on level surface, but most still required any assistive devices (cane, ankle-foot-orthosis, etc.). Stepwise, multivariable generalized linear regression model showed that the ability to move leg within six months post onset was independent predictors of walking recovery with risk ratio 2.23 (P < 0.01). Motor performance at immediate after onset and in subacute period found to be a strong predictor in several studies [9 , 15–17], our findings are consistent with these studies. Recovery of motor function after stroke was dramatic recovery occurred over the first 30 days, regardless of initial severity [18], but still continued over the first year after stroke [19]. This finding emphasized significance of intensive rehabilitation program as early as medical condition stabilized to promote recovery for stroke survivors and muscle strengthening or facilitation of normal movement program should be addressed since during hospitalization and continued to homecare. Aphasia and unilateral neglect (inattention) were shown a tendency to be independent predictors of walking recovery in this cohort but not significant in the final model. Evidence from the literature [9 , 19–22] identified that these stroke-related factors influenced poor functional outcome. Global aphasia is the most serious communication problem which mostly correlated to the left superior temporal gyrus and more likely to lower rate of recovery itself [23]. There were 38 global aphasic in this study, 26 (68.4%) could not walk independently, may supported the evidence that severe type of aphasia were associated with decreased rate of functional recovery, health-related quality of life and also higher mortality rate in stroke survivors [20 , 23–25]. Unilateral neglect, a disorder of spatial attention after stroke was reported as independent predictor of walking ability [9 , 22], shown to associated with the outcome on univariable analysis. Patients with unilateral neglect may exhibit errors in spatial perception which caused negative effects to mobility and walking ability [22] and increased more than 5-fold evidence of fall [21].

Other meaningful but nonsignificant variables in the final model including the presence of more than two other pre-stroke comorbidities and dysphagia. Comorbidities in our cohort were ranging from degenerative musculoskeletal problems as gout, osteoarthritis of knee joints, lumbar spondylosis, shoulder impingement syndrome to systematic disease as thalassemia, chronic kidney disease, COPD and hypothyroid. Weakness of hemiplegic limb together with multi-diseases prior to stroke could account to impact to recovery of function after stroke [17]. Difficulty to swallow (dysphagia) was reported an association for pulmonary complications especially aspiration and pneumonia and found higher incidence in elderly stroke patients as aging causes changes in muscles strength of masticatory apparatus [26 –28]. A systematic study by Martino [27], concluded that stroke patients with dysphagia had increased more than 3-fold risk for pneumonia and 11-fold increase in risk for aspiration, therefore, dysphagia and its consequences may limit functional recovery in stroke survivors and may be the leading cause of mortality in elderly stroke patients [27 , 30].

In conclusion, within six months post stroke which recovery was greatest [31], identified of independent predictors associate walking recovery provided meaningful information for stroke care team to initiate optimal plan and decide an appropriate management according to realistic goal. However, our study had some limitations. We recruited stroke patients from only a single unit, all participants received rehabilitation program twice daily for 4-6 weeks that might gained potentially benefit to the outcome from an intensive training. A relative small numbers of patients reduced statistical power, most of dependent variables had to be categorized into dichotomous consideration with cut off point from distribution and clinical supported. Some meaningful data such as socioeconomic status, NIHSS, Barthel Index could not be completely reviewed from medical records. Therefore, results could not be generalized and should be interpreted with caution.

Conflict of interest

None to report.

Footnotes

Acknowledgments

The author thank Professor Jayanthorn Patumanond, MD, Ph.D and Assistant Professor Chidchanok Ruengorn, Ph.D. for their contribution in data analysis.

References

Feigin

, Lawes

, Bennett

, Barker-Collo

, Parag

. Worldwide stroke incidence and early case fatality reported in 56 population-based studies: A systematic review. Lancet Neurol 8(4):355–69.

Roger

, Go

, Lloyd-Jones

, Adams

, Berry

, Brown

, et al. Heart disease and stroke statistics— update a report from the American heart association. Circulation 2011;123(4):e18–209.

Viriyavejakul

, Senanarong

, Prayoonwiwat

, Praditsuwan

, Chaisevikul

, Poungvarin

. Epidemiology of stroke in the elderly in Thailand. J Med Assoc Thail Chotmaihet Thangphaet 1998;81(7):497–505.

Public Health Statictics 2011. [cited 2015 Jul 12]. Available from: http://bps.ops.moph.go.th/E-book/statistic/statistic54/statistic54.html

Craig

, Wu

, Bernhardt

, Langhorne

. Predictors of poststroke mobility: Systematic review. Int J Stroke Off J Int Stroke Soc 2011;6(4):321–7.

Forti

, Maioli

, Procaccianti

, Nativio

, Lega

M-V

, Coveri

, et al. Independent predictors of ischemic stroke in the elderly: Prospective data from a stroke unit. Neurology 2013;80(1):29–38.

Paquet

, Desrosiers

, Demers

, Robichaud

, BRAD Group. Predictors of daily mobility skills 6 months post-discharge from acute care or rehabilitation in older adults with stroke living at home. Disabil Rehabil 2009;31(15):1267–74.

Patterson

, Forrester

, Rodgers

, Ryan

, Ivey

, Sorkin

, et al. Determinants of walking function after stroke: Differences by deficit severity. Arch Phys Med Rehabil 2007;88(1):115–9.

Kollen

, Port

I van de

, Lindeman

, Twisk

, Kwakkel

. Predicting improvement in gait after stroke a longitudinal prospective study. Stroke 2005;36(12):2676–80.

10.

Masiero

, Avesani

, Armani

, Verena

, Ermani

. Predictive factors for ambulation in stroke patients in the rehabilitation setting: A multivariate analysis. Clin Neurol Neurosurg 2007;109(9):763–9.

11.

Reid

, Dai

, Christian

, Reidy

, Counsell

, Gubitz

, et al. Developing predictive models of excellent and devastating outcome after stroke. Age Ageing 2012;41(4):560–4.

12.

van de Port

IGL

, Kwakkel

, Schepers

VPM

, Lindeman

. Predicting mobility outcome one year after stroke: A prospective cohort study. J Rehabil Med 2006;38(4):218–23.

13.

Petrilli

, Durufle

, Nicolas

, Pinel

, Kerdoncuff

, Gallien

. Prognostic factors in the recovery of the ability to walk after stroke. J Stroke Cerebrovasc Dis Off J Natl Stroke Assoc 2002;11(6):330–5.

14.

Safaz

, Ylmaz

, Yaşar

, Alaca

. Brunnstrom recovery stage and motricity index for the evaluation of upper extremity in stroke: Analysis for correlation and responsiveness. Int J Rehabil Res 2009;32(3):228–31.

15.

Mercer

, Freburger

, Yin

, Preisser

. Recovery of paretic lower extremity loading ability and physical function in the first six months after stroke. Arch Phys Med Rehabil 2014;95(8):1547–1555.e4.

16.

Hendricks

, van Limbeek

, Geurts

, Zwarts

. Motor recovery after stroke: A systematic review of the literature. Arch Phys Med Rehabil 2002;83(11):1629–37.

17.

Bates

, Xie

, Kwong

, Kurichi

, Ripley

, Davenport

, et al. Development and validation of prognostic indices for recovery of physical functioning following stroke: Part 2. PM&R [Internet]. [cited 2015 Mar 19]; Available from: https://www-sciencedirect-com.web.bisu.edu.cn/science/article/pii/S1934148215000441

18.

Duncan

, Goldstein

, Matchar

, Divine

, Feussner

. Measurement of motor recovery after stroke. Outcome assessment and sample size requirements. Stroke J Cereb Circ 1992;23(8):1084–9.

19.

Ween

, Alexander

, D’Esposito

, Roberts

. Factors predictive of stroke outcome in a rehabilitation setting. Neurology 1996;47(2):388–92.

20.

Oliveira

FF de

, Damasceno

. Global aphasia as a predictor of mortality in the acute phase of a first stroke. Arq Neuropsiquiatr 2011;69(2B):277–82.

21.

Chen

, Hreha

, Kong

, Barrett

. Impact of spatial neglect on stroke rehabilitation: Evidence from the setting of an inpatient rehabilitation facility. Arch Phys Med Rehabil 2015;96(8):1458–66.

22.

Oh-Park

, Hung

, Chen

, Barrett

. Severity of spatial neglect during acute inpatient rehabilitation predicts community mobility after stroke. PM R 2014;6(8):716–22.

23.

Watila

, Balarabe

. Factors predicting post-stroke aphasia recovery. J Neurol Sci 2015;352(1-2):12–8.

24.

Plowman

, Hentz

, Ellis

. Post-stroke aphasia prognosis: A review of patient-related and stroke-related factors. J Eval Clin Pract 2012;18(3):689–94.

25.

Oliveira

, Damasceno

. Short-term prognosis for speech and language in first stroke patients. Arq Neuropsiquiatr 2009;67(3B):849–55.

26.

Lugger

. Dysphagia in the elderly stroke patient. J Neurosci Nurs J Am Assoc Neurosci Nurses 1994;26(2):78–84.

27.

Martino

, Foley

, Bhogal

, Diamant

, Speechley

, Teasell

. Dysphagia after stroke incidence, diagnosis, and pulmonary complications. Stroke 2005;36(12):2756–63.

28.

Oliveira

AR de S

, Costa

AG de S

, Morais

HCC

, Cavalcante

, Lopes

MV de O

, de Araujo

. Clinical factors predicting risk for aspiration and respiratory aspiration among patients with Stroke. Rev Lat Am Enfermagem 2015;23(2):216–24.

29.

Nguyen

VQC

, PrvuBettger

, Guerrier

, Hirsch

, Thomas

, Pugh

, et al. Factors associated with discharge to home versus discharge to institutional care after inpatient stroke rehabilitation. Arch Phys Med Rehabil 2015;96(7):1297–303.

30.

Kenmuir

, Hammer

, Jovin

, Reddy

, Wechsler

, Jadhav

. Predictors of outcome in patients presenting with acute ischemic stroke and mild stroke scale scores. J Stroke Cerebrovasc Dis Off J Natl Stroke Assoc 2015;24(7):1685–9.

31.

Hankey

, Spiesser

, Hakimi

, Carita

, Gabriel

. Time frame and predictors of recovery from disability following recurrent ischemic stroke. Neurology 2007;68(3):202–5.