Effects of scapulohumeral rehabilitation protocol on trunk control recovery in patients with subacute stroke: A pilot randomized controlled trial

Abstract

BACKGROUND: Despite upper limb rehabilitation is widely investigated in patients with stroke, the effects of scapulohumeral rehabilitation on trunk stabillization are mainly unknown.

OBJECTIVE: To test the effects of scapulohumeral rehabilitation protocol on trunk control recovery in patients with subacute stroke.

METHODS: A pilot randomized controlled trial with two groups of 14 patients each one performing 20 minutes per day, 5 days a week, for 6 weeks in add on to standard therapy. Experimental group performed a specific scapulohumeral rehabilitation protocol aiming to improve trunk competencies whereas control group performed conventional arm rehabilitation. Clinical scale tests and accelerometric evaluations were performed pre- and post-treatment.

RESULTS: Experimental groups showed better scores at discharge at Trunk impairment Scale (p < 0.001), Barthel Index (p = 0.024), Trunk Control Test (p = 0.002), Sitting Balance Scale (p = 0.002), but neither at Fugl-Meyer Scale (p = 0.194) nor Modified Ashworth Scale (p = 0.114). Accelerometric analysis showed higher stability of trunk for experimental group especially during static and dynamic items.

CONCLUSIONS: The recovery of scapulohumeral functions also acts on trunk stabilization post-stroke.

Keywords

Scapulohumeral rehabilitation trunk stability stroke rehabilitation accelerometry

1 Introduction

Stroke is the leading cause of major adult disability in Western Societies (Duncan et al., 1992). Up to 30% of patients who survived to a stroke experience have a very limited motor recovery (Langhorne & Coupar 2009). After stroke, a proper walking recovery typically occurs in nearly 55% of rehabilitated patients (Paolucci et al., 2008). Recovery of upper limb motor functions is even poorer, with less than one third of patients recovering full dexterity after 1 year from the acute event (Kong et al., 2011), and only one half 4 years after stroke (Broeks et al., 1999). In Eighties it was observed that this recovery is just partially influenced by therapy (Heller et al., 1987; Wade, 1983), it leaded to claim for the need of more appropriate and adequate therapeutic interventions (Carr & Sheperd, 1998) still lacking (Pollock et al., 2014) and for more objective quantitative assessment of motor outcomes (Iosa et al., 2016). Population-based studies performed across the last thirty years indicated that about 80% of first-time strokes result in an acute hemiparesis of the upper limbs and/or lower limbs (Foulkes et al., 1988; Duncan et al., 1994; Langhorne et al., 2009). Lower limb impairments can severely limit mobility whereas upper limb impairments may increase disability reducing the independency in activities of daily living (Ostwald et al., 1989; Olsen, 1990; Langhorne et al., 2009) and the quality of life (Franceschini et al., 2010). A recent study conducted on 435 patients with stroke showed that the highest improvement after rehabilitation performed in subacute phase was related to bowel and bladder functions, transfer and mobility, whereas the lowest improvement was related to bathing, grooming, dressing, and stair climbing (Morone et al., 2015). Most of these activities of daily living require not only the recovery of limb functions, but a preliminary recovery of trunk control.

Many approaches have been proposed for training patients in trunk control recovery, from Bobath-based exercises (Kilinç et al., 2016) up to video-game based therapy (Morone et al., 2014) and mirroring technology (Shin & Song, 2016). The recovery of trunk control was sometimes exploited by patients for actuating compensatory strategies aiming at favouring reaching movements, but this approach has been seen as a limit for a proper upper limb recovery by clinicians. In fact, the long-term use of trunk compensatory strategies may contribute to secondary impairments, such as learned non-use, joint contractures and pain (Pain et al., 2015). Conversely, trunk restraint (similarly to the constraint movement induced therapy for the contralateral upper limb) enables functional reach practice, while limiting compensatory strategies in the moderately to severely impaired stroke population. Many studies investigated the effects of trunk restraint training on the rehabilitation of shoulder and elbow proper patterns in stroke (Wee et al., 2014; Jeon et al., 2015; Tielman et al., 2008).

An opposite approach could be acting on shoulder training with the double aim of upper limb recovery and for favouring trunk control rehabilitation. In fact, a more functional shoulder does not need compensatory strategies performed by trunk, including misalignment in the direction of the affected limb for enlarging the reachable space. But the effect of specific shoulder training on trunk control was poorly investigated. Some results showed that a shoulder girdle strengthening significantly contributes in improving the postural alignment of the trunk (Awad et al., 2015).

This lack is probably due to the focus given on the recovery of reaching abilities, but trunk stability is another key feature of neurorehabilitation being at the basis of upright posture, functional reaching movements and walking, as stated above (Hsieh et al., 2002; Iosa et al., 2012).

The aim of this study was to investigate if a specific training for shoulder rehabilitation after stroke performed in subacute phase may also favour trunk control recovery in terms of upright stability.

2 Material and methods

This study is a randomized controlled trial conducted at IRCCS Santa Lucia foundation, Rome from March 2015 to December 2015. The study was approved by the Local Ethical Committee (CE/PROG.546) and all participants signed the informed consent. Inclusion criteria were: patients with diagnosis of stroke, time from acute event <6 months (subacute phase), age between 20 and 80 years. Exclusion criteria were: not collaborative patients for their neurological, cognitive or psychiatric conditions, severe neglect, severe visual deficits not corrected with glasses, severe aphasia with comprehension deficits, severe shoulder spasticity (score >4 at modified Ashworth scale) (Bohannon & Smith, 1987), severe force deficit at shoulder level (Manual Muscle test score <3) (Daniels & Worthingham, 1986) inability of maintaining sitting position (Sitting Balance score <2) (Sandin & Smith B; 1990) enrolment in other research protocols, refuse to participate. Using a sequence generated by a proper computer software, patients were randomly assigned to Experimental Group (EG) or Control Group (CG). All patients performed 2 daily sessions of neurorehabilitation of 40 minutes each, 5 days a week, for 6 weeks. Two different add-on therapies (20 minutes per day, 5 days a week, for 6 weeks) were performed by the two groups: the experimental group performed a specific training for functional re-education of shoulder, whereas the add-on of the control group was based on non-specific arm/trunk rehabilitation.

The experimental training consisted of passive mobilization of scapula, task-oriented exercises of the shoulder in open kinetic chain (reaching activities) and closed kinetic chain (supporting activities), scapular stability and dynamic exercises, besides of proprioceptive stimulation (shoulder joint traction and approximation) aimed to the sensori-motor recovery of shoulder functions. In particular, as shown in Fig. 1, the following exercises were performed:

Patient lying in supine position: passive mobilization of paretic shoulder (in particular of the scapula) in every direction: up, down, forward, backward (Fig. 1A); patient with extended elbows and interlocked hands moves both upper limbs in every direction following variable targets as requested by physical therapist (PT, Fig. 1B); patient keeps extended elbow and open hand of the paretic limb against PT’s hand and pushes repeatedly along different directions (Fig. 1C)

Patient lying on unaffected side: passive mobilization of paretic shoulder (in particular of the scapula) in every direction (Fig. 1D); patient keeps his extended elbow and open hand of the paretic limb against PT’s hand and pushes repeatedly along different directions (Fig. 1E)

Patient sitting on the edge of the treatment bed without back support and feet on the floor: patient keeps his extended elbow and open hand of the paretic limb against PT’s hand and pushes and then holds the limb towards targets progressively away from central line (Fig. 1F); patient keeps both elbows extended and open hands on PT’s shoulders and pushes against his resistance along different directions (Fig. 1G); patient with arms supported on bed, PT pushes slowly on unaffected shoulder or on sternum toward the affected hand to stimulate a support reaction (Fig. 1H); patient leaning on the affected elbow and open hand on the table: PT’s hand pushes on the unaffected shoulder to stimulate the support on the affected side (fixation of scapula on ribe cage, Fig. 1I);

Patient sitting on a chair in front of a table: patient with both elbows and open hands leaning on the table: PT pushes on the back of trunk to stimulate both elbows support or on the unaffected shoulder to stimulate a support reaction of the opposite elbow and fixation of scapula (Fig. 1J).

Fig.1

Photos of the exercises forming the experimental treatment described into the material and method section.

Control training was composed of exercises for upper hemiparethic limb. Supine position: passive, active assisted and active mobilizations of hemiparethic arm, cervical mobilization, arm coordination exercises (cross patterns with and without tools such as a stick or a ball). Sitting position at the table with trunk rested on chair back and elbows on the table: coarse grasping exercises, bimanual coordination exercises (for example passing a ball from one hand to the other one, alternate movements of pronation and supination). Postural changes training: from supine to side-lying, to sitting, to standing position.

The standard physiotherapy, shared by both groups, was focused on the facilitation of movements on the paretic side, exercise to decrease spasticity and and improving balance, standing, sitting, and transferring and walking.

Assessment was performed at baseline (T0, after randomization) and at the end of therapy (T1), by an assessor blind to the type of rehabilitation administered to patients. The primary outcome measure was the Trunk Impairment Scale (Verheyden et al., 2004). It is a measure of motor impairment of the trunk after a stroke through the evaluation of static and dynamic sitting balance as well as co-ordination of trunk movement with a score ranging from 0 (no trunk control) up to 23 (perfect trunk control). It includes the assessment of static sitting balance, dynamic sitting balance, and coordination. Secondary outcome measures were: Trunk Control Test, Sitting Balance Scale (Sandin & Smith B, 1990), Fugl Meyer, Modified Ashworth Scale, Barthel Index, Canadian Neurological Scale.

The clinical assessment was coupled with a biomechanical quantitative assessment performed using a triaxial accelerometer (FreeSense, Sensorize, sampling frequency 100 Hz). The items of Trunk Impairment Scale was in fact executed by patient wearing an elastic belt at the chest level containing the accelerometer located on the back as shown in Fig. 2.

Fig.2

Photo of the assessment performed using triaxial accelerometer.

This triaxial accelerometer was lightweight (93 g) and allowed to measure accelerations along the three body axes (antero-posterior, AP; latero-lateral, LL; and cranio-caudal, CC). These signals were analyzed after the subtraction of their mean values and low-pass filtered at 20 Hz. Then, the Root Mean Square (RMS) of the signal was computed. RMS is a measure of acceleration dispersion, which coincides with the standard deviation because of signal mean subtraction (Iosa et al., 2012). For the items in which subject is asked to maintain his/her posture, the RMS was computed along all the three axes (static items 1–3: maintaining balance during sitting position with aligned legs, with unaffected leg crossed over the hemiplegic leg, with hemiplegic leg crossed over unaffected leg; dynamic items 7–10: maintaining balance during pelvis lifting at the affected and unaffected side) and used as an indicator of instability. For the items in which subject is asked to perform a movement in a specific body plane (formed by two body axes), the RMS was computed along the body axes perpendicular to that plane and used as an indicator of an improper pattern (the investigated axis was AP for dynamic items 1–3, 4–6; and CC for coordination items). RMS were then averaged for similar tasks such as dynamic items 1–3, 4–6, 7–10 and coordination items 1-2 and 3-4. For all the items, a reduction of RMS after treatment should be considered as an improvement.

Data where summarized in mean and standard deviation. Between group comparisons were performed at T0 and T1 on the clinical scale scores using the Mann-Whitney u-test being ordinal measures, whereas RMS, being a continuous measure, were compared using analysis of variance (ANOVA) with group as between subject factor, rehabilitation (T0 vs. T1) and body axis (AP, LL, CC) as within subject factor. Due to the multiple comparisons, the alpha-level was adjusted in accordance with Bonferroni and set at 0.007 for clinical scores (0.007 is about 0.05/7, being 7 the used clinical scales), 0.0125 for ANOVA performed on RMS (0.0125 = 0.05/4 being 4 analyses of variance performed, one for each domain), and 0.003 for post-hoc analyses following ANOVA (0.003 = 0.05/16, being 16 the number of RMS computed on two assessment time).

3 Results

Twenty eight patients have been enrolled in this study and divided into the two groups, demographical and clinical features were reported in Table 1.

Table 1
Demographical and clinical features of the two groups of patients

Feature Experimental Group Control Group p-value

Number of subjects 14 14 –

Age 54.21 ± 18.98 50.43 ± 17.49 0.588

Sex 8 male, 6 female 7 male, 7 female 0.705

Side of hemiparesis 6 left, 8 right 7 left, 7 right 0.705

CNS-score 6.8 ± 0.68 6.9 ± 1.14 0.734

Feature	Experimental Group	Control Group	p-value
Number of subjects	14	14	–
Age	54.21 ± 18.98	50.43 ± 17.49	0.588
Sex	8 male, 6 female	7 male, 7 female	0.705
Side of hemiparesis	6 left, 8 right	7 left, 7 right	0.705
CNS-score	6.8 ± 0.68	6.9 ± 1.14	0.734

The Clinical Scale scores assessed at T0 and T1 are reported in Table 2 for the two groups. All the scores at T0 were not statistically different between groups, whereas significant differences were noted in terms of Trunk Impairment Scale (the primary outcome measure), Trunk control Test, Sitting Balance Test, but neither for Fugl-Meyer Scale nor Modified Ashworth Scale. Taking into account the risk of alpha level inflation due to multiple comparisons we could not consider as statistically significant the difference observed on Barthel Index scores at T1.

Table 2

Mean ± standard deviation of clinical scale score and relevant p-values (obtained using Mann Whitney u-test) at T0 and T1

Scale	T0			T1
	EG	CG	p-value	EG	CG	p-value
Trunk impairment scale	8.0 ± 3.5	8.1 ± 3.4	0.946	20.1 ± 1.6	11.8 ± 3.15	<0.001
Barthel Index	35.9 ± 8.7	36.1 ± 7.6	0.910	85.6 ± 8.2	72.2 ± 14.6	0.024
Trunk Control Test	52.3 ± 16.4	52.5 ± 17.8	0.910	95.3 ± 8.2	74.1 ± 18.2	0.002
Sitting Balance Test	22.6 ± 5.8	22.7 ± 5.8	0.910	34.3 ± 1.9	26.8 ± 5.8	0.002
Fugl Meyer Scale	29.0 ± 11.0	32.2 ± 12.6	0.285	56.7 ± 10.9	46.8 ± 17.3	0.194
Modified Ashworth Scale	5.1 ± 1.7	3.8 ± 2.3	0.164	1.8 ± 1.1	3.1 ± 1.9	0.114

Bold values indicate a statistically significant difference at post-hoc analyses.

Accelerometric data showed a significant reduction of accelerations along the axis in which movements should be not performed for all of the items of Trunk impairment Scale in EG with respect to CG, as shown in Fig. 3. ANOVA highlighted that rehabilitation was in general effective in improving trunk balance during static items (F(1,26) = 14.045, p = 0.001), dynamic items 4–10 (F(1,26) = 32.256, p < 0.001), dynamic items 7–10 (F(1,26) = 67.516, p < 0.001), and coordination items (F(1,26) = 9.590, p = 0.005). However a significant interaction of rehabilitation per group was observed for static items (F(1,26) = 24.803, p < 0.001), dynamic items 4–10 (F(1,26) = 18.820, p < 0.001) dynamic items 7–10 (F(1,26) = 13.794, p = 0.001), and coordination items (F(1,26) = 13.261, p = 0.001). Post-hoc analyses revealed that instabilities were reduced more in EG than in CG along all the three body axes in static items. During dynamic items the improper movements were reduced along AP-axis for items 4–10, and along LL-axis for items 7–10, and along CC-axis for coordination items.

Fig.3

Mean±standard deviation of acceleration Root Mean Square for the items of Trunk Impairment Scale analysed also with accelerometer. Stars indicate a statistically significant difference at post-hoc analyses.

4 Discussion

The hypothesis of this study was that shoulder rehabilitation may improve trunk stability in patients with subacute stroke. Results confirmed this hypothesis, showing significant differences between experimental and control groups at the end of rehabilitation, especially in the domains related to trunk stability and balance functions. The independency in daily living activities approached the statistical significance, without reaching that for the conservatory choice of applying Bonferroni correction on the statistical analysis for taking into account the multiple comparisons. The absence of significant differences in terms of Fugl Meyer Scale and Modified Ashworth Scales could be explain by the fact that the domains covered by these two scales are already well taken into account by conventional therapy, and hence also the control group improved in these aspects. Clinical results were supported by instrumental assessment performed by means of the triaxial accelerometer. Interestingly, during dynamic items 7–10, the improvement occurred in latero-lateral axis, in accordance with a recovery of a more symmetric stability.

A major limit of our study was that assessments were performed only at the beginning and at the end of rehabilitation. In add a follow-up to verify that the permanence of benefits is lacking. Further, middle time assessments were not performed, but they could be helpful to understand if the trunk control recovery followed improvements in shoulder functions and also if then trunk control may favour further improvements for recovering upper limb functions. Further studies should investigate these aspects. The importance of trunk control was highlighted in a study reporting it as the most important clinical predictor of motor outcome in patients with stroke (Cote et al., 1989). Trunk control recovery was in turn predicted by age, Fugl Meyer score and Barthel Index score (Hsieh et al., 2002), three parameters also found as predictors of outcomes related to activities of daily living (Kwakkel et al., 1996). Trunk stability have many benefit related to motor abilities and independency in activities of daily living (Hsieh et al., 2002), reduction of risk of fall (Kao et al., 2014), and also on vital functions such as swallowing and breathing (Lanini et al., 2003; Reid & Samrai, 1995). Our results are in line with previous ones showing as strengthening of the scapular muscles is correlated with a reduction of spinal lateral deviation angle (Awad et al., 2015; Ko et al., 2016). However, previous studies were focused on the trunk muscle strengthening, whereas our intervention was focused on scapulo-humeral joint for stabilising the trunk, both in static and dynamic conditions.

In conclusion, a specific shoulder rehabilitation can be effective on the double objectives of favouring the recovery of hemiplegic upper limb functioning and that of trunk stability.

Conflict of interest

None to report.

References

Awad

, Shaker

, Shendy

, & Fahmy

(2015). Effect of shoulder girdle strengthening on trunk alignment in patients with stroke. J Phys Ther Sci, 27(7), 2195–2200.

Bohannon

R. Smith, M.

(1987). Interrater reliability of a modified Ashworth scale of muscle spasticity. Physical Therapy, 67(2), 206.

Broeks

J. G.

, Lankhorst

G. J.

, Rumping

, & Prevo

A. J.

(1999). The long-term outcome of arm function after stroke: Results of a follow-up study. Disabil Rehabil, 21(8), 357–364.

Carr

, & Shepherd

(1998). Neurological rehabilitation: Optimizing motor performance, Butterworth-Heinemann.

Cote

, Battista

R. N.

, Wolfson

, Boucher

, Adam

, & Hachinski

V. C.

(1989). The canadian neurological scale: Validation and reliability assessment. Neurology, 39, 638–643.

Daniels

, & Worthingham

(1986). Muscle Testing Techniques of Manual Examination. 5 ed., Philadelphia: WB Saunders.

Duncan

P. W.

, Goldstein

L. B.

, Matchar

, Divine

G. W.

, & Feussner

(1992). Measurement of motor recovery after stroke. Outcome assessment and sample size requirements. Stroke, 23, 1084–1089.

Duncan

P. W.

, Goldstein

L. B.

, Horner

R. D.

, et al (1994). Similar motor recovery of upper and lower extremities after stroke. Stroke, 25, 1181–1188.

Foulkes

M. A.

, Wolf

P. A.

, Price

T. R.

, et al (1988). The stroke data bank: Design, methods, and baseline characteristics. Stroke 547–554.

10.

Franceschini

, La Porta

, Agosti

, & Massucci

(2010). ICR2 group. Is health-related-quality of life of stroke patients influenced by neurological impairments at one year after stroke? Eur J Phys Rehabil Med, 46(3), 389–399.

11.

Heller

, Wade

D. T.

, Wood

V. A.

, Sunderland

, Hewer

R. L.

, & Ward

(1987). Arm function after stroke: Measurement and recovery over the first three months. J Neurol Neurosurg Psychiatry, 50(6), 714–719.

12.

Hsieh

C. L.

, Sheu

C. F.

, Hsueh

I. P.

, & Wang

C. H.

(2002). Trunk control as an early predictor of comprehensive activities of daily living function in stroke patients. Stroke, 33(11), 2626–2630.

13.

Iosa

, Fusco

, Morone

, Pratesi

, Coiro

, Venturiero

, De Angelis

, Bragoni

, & Paolucci

(2012). Assessment of upper-body dynamic stability during walking in patients with subacute stroke. J Rehabil Res Dev, 49(3), 439–450.

14.

Iosa

, Picerno

, Paolucci

, & Morone

(2016). Wearable inertial sensors for human movement analysis. Expert Rev Med Devices, 13(7), 641–659.

15.

Jeon

S. H.

, Lee

S. M.

, & Kim

J. H.

(2015). Therapeutic effects of reaching with forward bending of trunk on postural stability, dynamic balance, and gait in individuals with chronic hemiparetic stroke. J Phys Ther Sci, 27(8), 2447–2451.

16.

Kao

P. C.

, Dingwell

J. B.

, Higginson

J. S.

, & Binder-Macleod

(2014). Dynamic instability during post-stroke hemiparetic walking. Gait Posture, 40(3), 457–463.

17.

Kilinç

, Avcu

, Onursal

, Ayvat

, Savcun Demirci

, & Aksu Yildirim

(2016). The effects of Bobath-based trunk exercises on trunk control. functional capacity. balance. and gait: A pilot randomized controlled trial. Top Stroke Rehabil, 23(1), 50–58.

18.

E. J.

, Chun

M. H.

, Kim

D. Y.

, Yi

J. H.

, Kim

, & Hong

(2016). The Additive Effects of Core Muscle Strengthening and Trunk NMES on Trunk Balance in Stroke Patients. Ann Rehabil Med, 40(1), 142–151.

19.

Kong

K. H.

, Chua

K. S.

, & Leem

(2011). Recovery of upper limb dexterity in patients more than 1 year after stroke: Frequency, clinical correlates and predictors. NeuroRehabilitation, 28(2), 105–111.

20.

Kwakkel

, Wagenaar

R. C.

, Kollen

B. J.

, & Lankhorst

G. J.

(1996). Predicting disability in stroke: A critical review of the literature. Age Ageing, 25, 479–489.

21.

Lanini

, Bianchi

, Romagnoli

, et al (2003). Chest wall kinematics in patients with hemiplegia. Am J Respir Crit Care Med, 168, 109–113.

22.

Langhorne

, Coupar

, & Pollock

(2009). Motor recovery after stroke: A systematic review. Lancet Neurol, 8, 741–754.

23.

Morone

, Paolucci

, & Iosa

(2015). In what daily activities do patients achieve independence after stroke? J Stroke Cerebrovasc Dis, 24(8), 1931–1937.

24.

Morone

, Tramontano

, Iosa

, Shofany

, Iemma

, Musicco

, Paolucci

, & Caltagirone

(2014). The efficacy of balance training with video game-based therapy in subacute stroke patients: A randomized controlled trial. Biomed Res Int, 2014, 580861.

25.

Ostwald

S. K.

, Snowdon

D. A.

, Rysavy

D. M.

, et al (1989). Manual dexterity as a correlate of dependency in the elderly. J Am Geriatr Soc, 37, 963–969.

26.

Olsen

T. S.

(1990). Arm and leg paresis as outcome predictors in stroke rehabilitation. Stroke, 21, 247–251.

27.

Paolucci

, Bragoni

, Coiro

, De Angelis

, Fusco

F. R.

, Morelli

, et al (2008). Quantification of the probability of reaching mobility independence at discharge from a rehabilitation hospital in nonwalking early ischemic stroke patients: A multivariate study. Cerebrovasc Dis, 26, 16–22.

28.

Pain

L. M.

, Baker

, Richardson

, & Agur

A. M.

(2015). Effect of trunk-restraint training on function and compensatory trunk, shoulder and elbow patterns during post-stroke reach: A systematic review. Disabil Rehabil, 37(7), 553–562.

29.

Pollock

, Farmer

S. E.

, Brady

M. C.

, Langhorne

, Mead

G. E.

, Mehrholz

, & van Wijck

(2014). Interventions for improving upper limb function after stroke. Cochrane Database Syst Rev, 12(11), CD010820.

30.

Reid

W. D.

, & Samrai

(1995). Respiratory muscle training for patients with chronic obstructive pulmonary disease. Phys Ther, 75, 996–1005.

31.

Shin

D. C.

, & Song

C. H.

(2016). Smartphone-Based Visual Feedback Trunk Control Training Using a Gyroscope and Mirroring Technology for Stroke Patients: Single-blinded. Randomized Clinical Trial of Efficacy and Feasibility. Am J Phys Med Rehabil, 95(5), 319–329.

32.

Thielman

, Kaminski

, & Gentile

A. M.

(2008). Rehabilitation of reaching after stroke: Comparing 2 training protocols utilizing trunk restraint. Neurorehabil Neural Repair, 22(6), 697–705.

33.

Verheyden

, Nieuwboer

, Mertin

, Preger

, Kiekens

, & De Weerdt

(2004). The Trunk Impairment Scale: A new tool to measure motor impairment of the trunk after stroke. Clin Rehabil, 18(3), 326–334.

34.

Sandin

K. J.

, & Smith

(1990). The measure of balance in sitting in stroke rehabilitation prognosis. Stroke, 21, 82–86.

35.

Wade

D. T.

, Langton-Hewer

, Wood

V. A.

, Skilbeck

C. E.

, & Ismail

H. M.

(1983). The hemiplegic arm after stroke: Measurement and recovery. J Neurol Neurosurg Psychiatry, 46, 521–524.

36.

Wee

S. K.

, Hughes

A. M.

, Warner

, & Burridge

J. H.

(2014). Trunk restraint to promote upper extremity recovery in stroke patients: A systematic review and meta-analysis. Neurorehabil Neural Repair, 28(7), 660–677.