Proprioceptive based training or modified constraint-induced movement therapy on upper extremity motor functions in chronic stroke patients: A randomized controlled study

1Introduction

According to the World Health Organization, stroke is the leading cause of disability and mortality among neurological diseases (World Health Organization, 2020). Approximately 13.7 million people experience a stroke anually worldwide, of which 5.5 million people die. Five million people experience permanent activity limitations during their lifetime after a stroke (Johnson et al., 2019). In Turkey, 22.4% of total deaths in 2018 were reportedly caused by cerebrovascular events, and one-third of the stroke survivors in Turkey become dependent in their activities of daily living (İnme, 2018).

Studies have reported that 67% stroke survivors continue to experience disability in the upper extremity and cannot use the affected upper extremity (Broeks, Lankhorst, Rumping, & Prevo, 1999; Siebers et al., 2010). Despite the importance of the condition of upper extremities, rehabilitation often focuses on improving balance, gait, and general mobilization because of the inadequacies of rehabilitation resources, time constraints, and insufficient early motor recovery in upper extremities (Ward & Cohen, 2004). Futhermore, it is unclear which dose (frequency, duration, or intensity) of upper limb rehabilitation and which rehabilitation interventions are most effective; furthermore, the patient characteristics that would be benefitted the most remain unclear (Cho & Li, 2021; Kwakkel et al., 2004). Therefore, it is important to compare the techniques and determine which of them improve functional outcomes and limitations and the effect on function. In addition, Whitall claimed that exploring the relative effects of different research supported by training programs is important to design and develop efficient and effective rehabilitation programs for patients with stroke (Whitall et al., 2011).

Proprioception is the capacity of the central nervous system to determine where all body parts will be placed statically or dynamically. Formation or preservation of somatosensory integrity, and proprioceptive integrity in particular, is considered crucial for motor control and learning ((Bornstein, Konstantin, Alessandro, Tresch, & Zelzer, 2021; Vidoni, 2008). Proprioceptive based training (PT) is based on motor learning principles, such as the use of concurrent feedback and repetition of tasks; it is one of the bilateral arm training methods. PT is reported to be beneficial as it performs simultaneous movements with both the unaffected and the affected arm to promote motor recovery through some reciprocal connections of the interhemispheric and transcallosal pathway. PT in the upper hemiplegic extremity is supported by a physiotherapist with the aim to retrain movement and spatial location/position/movement awareness (Bolognini, Russo, & Edwards, 2016; Vidoni, 2008, Kiper, Baba, Agostini, & Turolla, 2015). Recognition of the position of both limbs, therefore, is always requested. The advantage of the proposed approach relies on the possibility of application since the early acute phase after stroke, when several therapeutic modalities— such as Constraint-Induced Movement Therapy— cannot be provided because of the absence of residual voluntary muscular activation (Kiper, Baba, Agostini, & Turolla, 2015).

Proprioceptive deficits have reportedly been negatively associated with initial motor function, independence during hospitalization, and upper extremity functional recovery, (Ocal, Alaca, & Canbora, 2020; Meyer, Karttunen, Thijs, Feys, & Verheyden, 2014). Although, in recent years, the importance of proprioceptive therapy targeting the somatic sensation has been increasing (Rand, 2018; Kiper et al., 2015; Smania, Montagnana, Faccioli, Fiaschi, & Aglioti, 2003; Ocal et al., 2020), studies examining the role of post-stroke proprioceptive rehabilitation are rather limited (Rand, 2018; Kiper, Baba, Agostini, & Turolla, 2015; Ocal, Alaca, & Canbora, 2020; Meyer, Karttunen, Thijs, Feys, & Verheyden, 2014; Smania, Montagnana, Faccioli, Fiaschi, & Aglioti, 2003; Yozbatiran, Donmez, Kayak, & Bozan, 2006; Findlater & Dukelow, 2017). Therefore, the field of proprioceptive rehabilitation is relatively new and well-designed studies are needed in this area particularly those pertaining to upper extremity rehabilitation (Findlater & Dukelow, 2017).

The Constraint-Induced Movement Therapy (CIMT) method aims to increase the functional use of the affected extremity by restricting the less affected side extremity (Rocha et al. 2021; Wolf, 2007). Unsuccessful use of the affected upper extremity in stroke patients may cause “learned nonuse phenomenon,” in which patients habitually rely on their less affected upper extremity to accomplish daily activities. CIMT has been effective for overcoming learned nonuse. The three major components of CIMT include the following: [1] restraint of the less affected arm; [2] repetitive, structured, practice intensive therapy in the more affected arm; and [3] application of a package of behavioral techniques that transfers gains from the clinical setting to the real world (i.e. making it functional, shaping practice, and task specific training). Studies with CIMT have shown that patients with chronic stroke improve functional use of the upper limbs by 20% – 25% (Yadav, Sharma, Borah, & Kothari, 2016; Page, Sisto, Levine, & McGrath, 2004; Pauli et al., 2020; Fleet, Page, MacKay-Lyons, & Boe, 2014; Shi, Tian, Yang, & Zhao, 2011; Cho & Li, 2021). The CIMT approach, with its high evidence value, is recognized as an important therapeutic method in the treatment of patients with stroke (Pauli et al., 2020). The original CIMT approach required the use of restrictive gloves for 90% of the wake time and approximately 6 hours of goal-directed training per day in a 2-week period (Rocha et al. 2021; Page, Sisto, Levine, & McGrath, 2004). However, the difficulty of applying this method has been reported by patients, families, and physiotherapists. Many patients with stroke have reported that they would not want to participate in CIMT, but would prefer a longer treatment protocol with shorter activity sessions and/or wearing restrictive devices for fewer hours (Page, Levine, Sisto, Bond, & Johnston, 2002; Pauli et al. 2020). More than 60% of responding therapists also noted that patients were extremely unlikely to comply with CIMT (Pauli et al., 2020). Therefore, a modified version of CIMT has been created. The modified CIMT (mCIMT) application varies between 30 minutes and 6 hours of goal-directed training per day and 2–7 sessions per week (total treatment duration 2–12 weeks). Specifically, in comparison to the original CIMT, meta-analytic evidence suggests that mCIMT is similar functional recovery of an affected limb post-stroke (Yadav, Sharma, Borah, & Kothari, 2016; Page, Sisto, Levine, & McGrath, 2004; Pauli et al., 2020; Fleet, Page, MacKay-Lyons, & Boe, 2014; Shi, Tian, Yang, & Zhao, 2011; Cho & Li, 2021).

In recent years, the question of whether unilateral or bilateral training is better for improving upper extremity function in chronic stroke patients has received more attention in the literature. PT includes bilateral training methods as well as influencing proprioceptive receptors (Sethy et al., 2016). The mCIMT method is a unilateral training strategy that avoids the less affected side of the upper extremity and activates the affected side of the upper extremity (Sethy, Bajpai, Kujur, Mohakud, & Sahoo, 2016; Cho & Li, 2021). To the best of our knowledge, there is no study in the literature regarding their superiority over each other. Therefore, we hypothesized that PT— because it restores proprioception deficits and sensory integrity, as well as being a bilateral education— will show similar or more beneficial functional improvements as mCIMT in patients with chronic stroke and PT could be used for patients who could not adapt to CMIT or mCMIT. In the present study, we attempted to investigate the effectiveness of conventional therapy and PT or mCIMT therapy in patients with chronic stroke and to compare which of the two interventions showed more or smilar effectiveness in terms of recovery of upper extremity function in patients with chronic stroke.

2Methods

2.1Study design and patients

This study used a prospective three-arms randomized controlled trial. The study was approved by Acıbadem University and Acıbadem Healthcare Institutions Clinical Research Ethics Committee (reference no. 2020-24/05) and registered on ClinicalTrials.gov from the U.S. National Library of Medicine (registration no. NCT04873908), and performed in accordance with the 2010CONSORT statement. The participants for the study were patients who applied for private home care (May 2021–September 2021), were aged 40–80 years, and had a stroke at least 6 months ago. Verbal and written consent was obtained from the patients for the study in accordance with the Declaration of Helsinki. The inclusion criteria were (Wolf, 2007; Page, Sisto, Levine, & McGrath, 2004):

• • Patients who were diagnosed with stroke according to the World Health Organization criteria;

• • One-sided hemiparesis;

• • According to Brunnstrom the upper extremity value at a minimum of 3;

• • At least 10° active extension in the wrist, at least 10° active abduction in the thumb, any 2 of the metacarpophalangeal joints have at least 10° of active extension;

• • According to the Standardized Mini Mental State Examination, scoring a minimum of 24 points;

• • Upper extremity spasticity score of < 3 according to Modified Ashworth Scale were included from the study.

The exclusion criteria were (Yadav, Sharma, Borah, & Kothari, 2016; Page, Sisto, Levine, & McGrath, 2004):

• • Severe heart disease with upper extremity pain [score ≥4 according to the visual pain scale (0–10 cm)];

• • Minimal nonuse upper extremity (Motor Activity Diary score > 2.5);

• • Major neurological or rheumatological disorders affecting the musculoskeletal system other than stroke;

• • Impaired cognitive functions;

• • Significant vision or hearing impairment;

• • For safety reasons patients who had balance problems in the less affected side upper extremity restrictions were excluded from the study.

After the end of the study, the most effective rehabilitation program (according to results of our study) was applied to the control group in accordance with the Declaration of Helsinki.

The study power was calculated using the G^*Power V.3.1.7 (Kiel University, Kiel, Germany) program. A total of 45 patients were needed to achieve the power ratio of the study as 80% (effect size = 0.45 according to the Fugl-Meyer Upper Extremity Motor Rating Scale [FM]) with an error margin of 0.05 at a confidence level of 95% (Page, Sisto, Levine, & McGrath, 2004; Ocal, Alaca, & Canbora, 2020). A random allocation software program was used to randomly allocate the participants into three groups using the block randomization method, and the allocation was concealed. The recruitment process is reported as per the CONSORT (Fig. 1) and whose results could be analyzed after the completion of their treatment in groups as follows (n = 40):

Fig. 1

CONSORT flowchart of the study process.

• i. Conventional therapy training group (PTR, n = 14);

• ii. Conventional therapy and proprioceptive based training group (PTR-PT, n = 13);

• iii. Conventional therapy and Modified Constraint-Induced Movement Therapy group (PTR-mCIMT, n = 13).

3Results

The sociodemographic characteristics of the groups and the initial scores of the evaluations that were performed are shown in Table 1. The intergroup comparisons demonstrated that the baseline scores of the groups were not statistically significant (P = 0.217–0.965, Table 1).

Spasticity and TLT evaluations were conducted to analyze a Chi-square. In spasticity evaluation before and after the treatment; no change was observed in the PTR (P = 1.000, Table 2) and PTR-PT group (P = 0.165, Table 2). In the PTR-mCIMT group, there was a statistically significant decrease in spasticity in the in-group evaluation (P = 0.031, Table 2). In the comparison of the groups with each other, the PTR-mCMIT group decreased spasticity more significantly than both the PTR (P < 0.001, Table 2) and the PTR-PT group (P = 0.006, Table 2).

In determining the intra-group differences in TLT evaluations, compared with the pre-treatment there was no difference in the PTR group (P = 1.000, Table 2), while there was a significant decrease in the PTR-PT group (P = 0.001, Table 2) and the PTR-mCIMT group (P = 0.002, Table 2). Among the groups, the PTR-PT group showed a significant decrease compared to the PTR group (P = 0.020, Table 2), while the PTR-mCIMT group compared to both the PTR (P < 0.001, Table 2) and the PTR-PT group (P = 0.008) decreased more significantly (Table 2).

In the in-group evaluation of FM, ARAT, and MAL-28-QOM and MAL-28-AOU tests; while there was no significant increase in the PTR group (P = 0.088, P = 0.082, P = 0.061, P = 0.058, respectively, Table 3), with the PTR-PT (P = 0.002, P < 0.001, P = 0.006, P = 0.005 respectively, Table 3) and PTR-mCMIT (P = 0.002, P < 0.001, P < 0.001, P < 0.001, respectively, Table 3) groups there was a statistically significant increase compared to the pre-treatment.

An one-way ANOVA revealed that there was a statistically significant difference in FM (F (2, 36) =[14.485], P < 0.001), ARAT (F (2, 36) = [14.004], P < 0.001) and MAL (F (2, 36) = [6,690], P = 0.03) between three groups. Tukey’s HSD Test for multiple comparisons found that FM was significantly different between PTR (95% C.I. = [–1.09–0.07]) and PTR-mCMIT (95% C.I. = [–8.00–2.50], p < 0.001), PTR-PT (95% C.I.  = 15.00–6.00]) and PTR-mCMIT (P = 0.011, Table 3) groups. FM test, the PTR-PT group did not show a significant difference compared to the PTR group (P = 0.070, Table 3). ARAT score was significantly different between PTR (95% C.I. =  [–0.19–2.05]) and PTR-mCMIT (P < 0.001, 95% C.I. =  [4.97–11.35], Table 3), PTR-PT (95% C.I. =  [2.26–5.90]) and PTR-mCMIT (P = 0.015, Table 3) groups but the PTR-PT group did not show a significant difference compared to the PTR group (p = 0.062, Table 3). MAL-28-QOM was significantly different between PTR (95% C.I. =  [0.01–0.08]) and PTR-PT (P = 0.035, 95% C.I. =  [0.09–0.050], Table 3), PTR and PTR-mCMIT groups (P = 0.004, 95% C.I. = [0.20–0.58], Table 3). MAL-28- AOU was significantly different between PTR (95% C.I. = [0.01–0.09]) and PTR-PT (P = 0.041, 95% C.I. = [0.09–0.065], Table 3), PTR and PTR-mCMIT groups (P = 0.005, 95% C.I. = [0.20–0.60], Table 3). MAL-28-QOM and MAL-28-AOU scores, the PTR-PT group did not show a significant difference compared to the PTR-mCMIT group (P = 0.623, P = 0.688, respectively, Table 3).

In the effect value results, the PTR-PT group (ES = 0.87–1.43, Table 3) showed a higher effect value than the PTR group (ES = 0.51–0.57, Table 3) in the FM, ARAT, and MAL evaluations. The PTR-mCIMT group showed a higher effect value than the other groups (ES = 1.33–1.69, Table 3).

4Discussion

The present study aimed to compare the effect of PT therapy or mCMIT applied in addition to conventional therapy on upper extremity motor functions in patients with chronic stroke. It was determined that both PT and mCMIT helped improve posttreatment proprioception, motor function, and activities of daily life. The only group where these improvements were not observed was the conventional therapy group. In addition, the group in which mCMIT was applied, compared to the other groups, showed further improvements in spasticity, and motor function.

Various studies in both animals and humans show that the central nervous system is rearranged differently at molecular and synaptic connectivity levels depending on the type of behavioral interaction with the external environment, that is, it develops plasticity. Therefore, conventional therapy methods used to regulate upper extremity motor functions are insufficient (due to low intensity and less repetitions) and techniques with intense stimulation (repetitive and task-oriented movement) including motor learning strategies may be more beneficial (Alıpsatıcı, Alaca, & Canbora, 2020; Barreca, Wolf, Fasoli, & Bohannon, 2003; Krakauer, 2005; Pauli et al, 2020; Dimyan & Cohen, 2011). Similar to previous review articles (Barreca, Wolf, Fasoli, & Bohannon, 2003; Krakauer, 2005; Pauli et al, 2020; Dimyan & Cohen, 2011), probably due to the abovementioned reasons, the current study demonstrated that PT and mCMIT improved upper limb functions, while these improvements were not observed was the only conventional therapy group.

Proprioception training is a bilateral training besides improving proprioception, whereas mCMIT is a unilateral training method. Studies have reported that bilateral training is superior to unilateral training (Kelly, Borstad, Kline, & Gauthier, 2018; Lin, Chen, Chen, Wu, & Chang, 2010; Van der Lee, 1999) while there were studies reporting no difference (Morris, van Wijck, Joice, Ogston, Cole, & MacWalter, 2008; Morris & Van Wijck, 2012; Whitall et al., 2011). Although the neural mechanisms remain unclear, it is speculated that bilateral administration may facilitate co-activation and interhemispheric activation and thus a possible change in the contralesional cortical network. Additionally, cortical motor area damage directly affects 15%–20% of uncrossed corticospinal fibers and provides ipsilateral control of unilateral movements. After stroke, the undamaged primary motor cortex receives reduced interhemispheric inhibition from the lesioned hemisphere. These conditions cause ipsilesional dysfunction with normative motor control. For these reasons, bilateral training suggests that training may cause beneficial changes not only in the ipsilesional brain area but also in the contralesional brain area (the entrainment effect). Another important aspect of bilateral arm training is repetition, a well-established motor learning principle, and recent animal studies have shown that forced use involving a motor task rather than restrictive use alone can best improve central nervous system flexibility (Sethy, Bajpai, Kujur, Mohakud, & Sahoo, 2016; Lin, Chen, Chen, Wu, & Chang, 2010; Van der Lee, 1999; Morris & Van Wijck, 2012; van Delden et al., 2013; Chen, Abel, Janecek, Chen, Zheng, & Cramer, 2019, Yu et al., 2017). On the other hand, CIMT and mCIMT; provide positive reinforcement with the use of the paretic extremity, overcoming learned nonuse with practice through cortical reorganization and strengthening of neural connections, greatly increasing the use of the paretic extremity, improving mobility, and supporting its use in activities of daily living. Neuroimaging studies have shown that stroke patients participating in CIMT and mCIMT undergo neuroplastic changes in brain function and structure. Several cortical areas, such as the primary motor cortex, dorsal premotor cortex, and supplementary motor area, show increased electrical and metabolic neuronal activity during CIMT and mCIMT (Szaflarski, Page, Kissela, Lee, Levine, & Strakowski, 2006; Yadav, Sharma, Borah, & Kothari, 2016; Lucas et al., 2013).

CIMT and mCMIT were more effective in improving motor functions than the Bobath technique, bilateral training, mirror therapy, and other treatment methods especially in evaluations of hand functions and daily life (Huseyinsinoglu, Ozdincler, & Krespi, 2012; Brunner, Skouen, & Strand, 2012; Sethy, Bajpai, Kujur, Mohakud, & Sahoo, 2016; Ju & Yoon, 2018). Although it is a bilateral training method, the number of studies on PT is limited (Rand, 2018; Kiper, Baba, Agostini, & Turolla, 2015, 2015; Ocal, Alaca, & Canbora, 2020; Samania et al., 2003; Meyer, Karttunen, Thijs, Feys, & Verheyden, 2014; Yozbatiran, Donmez, Kayak, & Bozan, 2006; Findlater & Dukelow, 2017) and there are very few studies on its divergence from other methods (Ocal, Alaca, & Canbora, 2020). In a pilot study conducted by Kiper, Baba, Agostini, & Turolla (2015) with six patients with subacute stroke, they stated that 3 weeks of proprioception-based training had a positive effect on spasticity but a difference was not observed in the FM test. Smania, Montagnana, Faccioli, Fiaschi, & Aglioti, (2003) conducted 30 treatment sessions, similar to the present study, they evaluated the effects of a training program targeting somatic sensory and sensory deficits in motor control in a chronic stroke patient. During the 6-month follow-up, they observed that this treatment provided significant improvements. In addition, the authors’ study published in 2020 (Ocal, Alaca, & Canbora, 2020) compared conventional therapy and the additional PT model in chronic stroke patients. It was observed that upper extremity PT yielded better results in patients with chronic hemiplegia developed after stroke than conventional therapy in increasing the frequency and quality of movement in upper extremity similar to our study.

In the present study, mCIMT in the spasticity, FM, and ARAT evaluations, showed better results than PT and our hypothesis that both treatments might show similar effects was not confirmed. As the reasons for these consequences; it was our understanding that one of the causes for inclusion — in accordance with the working principles of mCMIT— in our study is a moderate lack of proprioception, which occured owing to the need to have slightly improved hand functions. It is unclear whether a patient with severe proprioception deficit will benefit from these therapies. The assessors were not blinded to group allocation, which is a major limitation. Futhermore, the sample size was not large and the long-term follow-up assessment of our patients. We could not be evaluated and the limited external generalisability of the findings, which was a very stringent inclusion criteria. These situtations constitute limitations of our study. In the future, more studies with high sample numbers, long-term results, with suitable masking, patient with different propriocepsion deficits are needed to show the superiority of both treatments more clearly.

5Conclusion

This study demonstrated it was determined that PT and mCMIT applied in addition to conventional therapy in patients with chronic stroke had therapeutic effects on proprioception, motor function, and activities of daily living when compared to only the conventional treatment. In addition, mCMIT showed greater improvement in spasticity and motor functional outcomes than PT.