Effect of Trunk Targeted Interventions on Functional Outcomes in Children with Cerebral Palsy- A Systematic Review

ABSTRACT Objective of this review was to collate information on the effectiveness of trunk targeted intervention on children with cerebral palsy (CP) on three functional outcomes that are gross motor function, trunk control and balance. A comprehensive search was conducted on online databases from inception to August 2021, using relevant keywords. A total of 15 randomized controlled trials which enrolled children with cerebral palsy under 18 years met the inclusion criteria. A significant improvement was seen in the trunk targeted training groups on applying trunk targeted interventions. Trunk targeted interventions improve gross motor function, trunk control as well as balance, hence should be incorporated in the conventional physical therapy program delivered to children with CP and would help in greater functional recovery.


Introduction
As the brain matures over time, the clinical presentation of children with cerebral palsy (CP) varies significantly, despite being a non-progressive disorder. 1 These children are present with delayed motor development, accompanied by limitations in motor performance due to reduced strength and abnormal muscle tone. 2,3 Poor trunk control remains one of the chief problems of children with CP, which directly affects activities of daily living. 4 The trunk, being the central key point of the body, influences the functional movements of the limbs. [5][6][7] Children with CP find it challenging to perform selective trunk movements, 8 thus leading to impaired trunk control which affects activities like sitting, reaching, and walking. [9][10][11] Trunk control, which is impaired in children with CP, is one of the determinants of motor function, mainly gross motor function and mobility, 12 and both static and dynamic trunk control functions are affected in children with CP. Static trunk control requires anticipatory as well as compensatory postural adjustments to maintain the stable trunk position while performing upper and lower limb movements. 8,13,14 In contrast, dynamic reaching is the ability to actively move the trunk forwards, laterally and across midline beyond the limits of base of support. 8 Amongst the different subcategories of cerebral palsy, quadriplegics show a greater involvement of trunk control when compared to diplegics and hemiplegics. 8 Deficit in trunk control varies amongst the levels in Gross Motor Function Classification Scale; there is a more significant impairment in the trunk control as the level of disability increases. 8 The above findings make the incorporation of trunk targeted interventions necessary in the rehabilitation program of children with CP.
Physical therapy aims to improve the functional ability in children with developmental conditions like CP and plan interventions that will ensure the child's well-being. 15 Among these therapies, interventions that have been effective in improving trunk stability and posture have been compared to conventional therapies, and have gained recognition in the last few years. 16 These interventions include therapies pertaining to either general trunk exercises, 17 using neurodevelopmental treatment strategies, [18][19][20] proprioceptive neuromuscular facilitation, 21 loaded weights, 22 using additional intervention strategies for the trunk such as taping, [23][24][25][26][27] electrical stimulation, [27][28][29] using external trunk stabilizers like trunk belt 30 or a trunk training equipment system. 31 Individual studies showed a more remarkable improvement in the functional outcome on incorporating trunk targeted training along with the conventional rehabilitation program of children with CP. However, to date, no systematic review has been performed to assess the effectiveness of such trunk targeted interventions on different body functions, in children with CP. This review, therefore, aims to examine the effect of trunk targeted interventions on the gross motor function, trunk control and balance in children with CP.
Register of Systematic Reviews (PROSPERO). Registration number-CRD42021237023

Search Strategy
A thorough data search was performed on databases such as PubMed, Scopus, Web of Science, PEDro, Cochrane, ProQuest, and EMBASE from inception to August 2021. The search term used for trunk was torso (MeSH). For intervention, the search terms were Exercise interventions, therapy, treatment, exercise, therapeutics (MeSH), exercise therapy (MeSH), exercise (MeSH), physical therapy modalities (MeSH). Trunk control was the search term used for trunk control. The search terms used for Gross motor function were Motor skill, gross motor skills, motor skills, Motor skills (MeSH), motor skill (MeSH), and motor activity (MeSH). For balance, Balance, postural balance, and postural balance (MeSH) were the search terms that were used. Lastly, the search terms used for cerebral palsy were cerebral palsy and cerebral palsy (MeSH). Boolean operators "AND" or "OR" were used to combine the search terms wherever relevant.

Selection of Studies
After eliminating the duplicates, two reviewers performed the title and abstract screening independently (AT and PJ). Any disagreements after screening were sorted by discussion with AN. Finally, a pre-set criterion was decided for screening the eligible articles which included only randomized controlled trials specific to interventions performed on the trunk in children having cerebral palsy. The full-text screening was done for the selected eligible studies, independently by AT and PJ. The selected studies met the following inclusion criteria-1. Articles evaluating participants under the age group of 18, diagnosed with cerebral palsy, 2. Articles performing trunk targeted interventions on children with cerebral palsy, 3. Articles assessing at least one of the outcomes (gross motor function, trunk control, or balance), and 4. Articles published in the English language. Preferred Reporting items for Systematic Reviews and Meta-analyses (PRISMA) flow diagram was used for displaying the results of the search ( Figure 1).

Data Extraction and Management
The studies are summarized in a data extraction table (Table 1), which includes details of the study such as the author and the year in which the study was published, details of the subjects, description and duration of the intervention provided, the outcome measures used, the key findings and the conclusion. The data extraction procedure was carried out by AT and PJ and discrepancies were settled by discussions with AN and SK.

Assessment of Risk of Bias
The Cochrane risk of bias tool 32 was used to assess the risk of bias for the included studies, which was performed independently by AT and PJ. A judgment of either high, low, or unclear risk was marked under domains like random sequence generation, concealment of allocation, blinding of participants and personnel, and outcome assessment, incomplete outcome data, selective reporting, and any other bias that might have occurred while performing the study. Any disagreements between AT and PJ were resolved by discussions with AN.

Characteristics of Studies
In total, 678 articles were identified from the various databases (PubMed-30, Scopus-116, Web of Science-202, Cochrane-7, ProQuest-167, EMBASE-147, and PEDro-0) and through other sources (9). On removing the duplicates, the titles and abstracts for a total of 533 articles were screened. This was followed by full text screening for the potentially eligible 87 articles, out of which 15 articles met the inclusion. The reasons for exclusion of the articles were studies other than randomized controlled trials, different outcome measures and the most common reason being interventions not specific to the trunk.

Population of Included Studies
Out of the 15 RCT studies, ten studies included children with spastic diplegia, [19][20][21][22][23]25,[27][28][29][30] one with spastic hemiplegia, 17 two included all types of spastic cerebral palsy, 18,24 and lastly, two studies included all types of cerebral palsy. 26,31 Seven studies classified children based on their functional ability using GMFCS. Two articles used children in the GMFCS levels I-III, 20,21 two included children within levels I-II, 17,22 one article had children within levels II-III, 19 and two articles included children falling between levels III-V. 26,31

Intervention of Included Studies
The articles included in this systematic review include interventions that are targeted directly on the trunk for the physical rehabilitation of children with CP. The trunk targeted interventions delivered to the study group included general trunkspecific exercises which included either general exercise focused on the trunk(n = 1), 17 exercises based on NDT/ Bobath principles(n = 3), 18-20 PNF principles(n = 1), 21 resistance training by using loaded weights on the trunk to perform sit to stand(n = 1), 22 use of external trunk stabilizers such as a trunk belt made of leather adjustable with Velcro, a head and trunk control equipment system(n = 2), 30,31 electrical stimulation(n = 3), [27][28][29] and lastly, giving proprioceptive input to the trunk in the form of taping(n = 5). [23][24][25][26][27] In the included studies, the control group received conventional therapy based on the NDT principles. The conventional therapy comprised different types of exercises which included positioning, facilitating head control, general range of motion and strengthening exercises for upper limb such as grabbingreleasing, pegboard exercises, and catching-throwing the ball, exercises for lower limb such as passive stretching to reduce spasticity, maintaining standing, squatting, stepping, ball kicking and gait training, and exercises for trunk such as activities on a balance board, reach out activities, weight shifts, lifting upper trunk in supine or prone using a mat or a wedge and facilitating equilibrium reactions. All these interventions were given for a duration of 4 weeks to 6 months. A total of 535 children were included in the trial out of which 501 children completed the study. No adverse events were reported related to the interventions in either of the studies.

Outcome Variables
According to the International classification of Functioning, Disability and Health (ICF), the outcome variables can be classified under body structure and function, activity, and participation. Therefore, the outcome variables that have been given attention in this review are gross motor function, trunk control, and balance, all of them which fall under the first component of ICF framework, i.e., body structure and function.

Gross Motor Function
Five studies used GMFM-88, 19 18,19,26,31 i.e., A-lying and rolling, B-sitting, C-crawling and kneeling, D-standing and E-walking, jumping, and running. Five studies used only dimension B, [25][26][27][28][29] one study used dimensions D and E22 and one study used dimensions B and D. 23,30 An improvement was noted in the gross motor function in nine studies. 18,19,22,23,25,[27][28][29][30] Sah et al. observed a significant between-group mean difference of 8.53 for the overall gross motor function (p < .001). 19 Similarly, Karabay et al. assessed only the sitting subset, 28 and Wahsh et al. assessed sitting as well as standing subsets of GMFM 30 and reported a significant improvement in the respective subsets (p < .001) between the two groups. 19,28,30 On application of taping on the trunk in the study conducted by Ibrahim et al., significant improvements were Figure 2. Cochrane risk of bias. Green-low risk, yellow-unclear risk, and red-high risk of bias seen in sitting (p = .005) and standing (p = .003) between the two groups. 23 The remaining five studies that incorporated exercises focused on the trunk, loaded weights and kinesiotaping along with conventional therapy 18,22,25,27,29 also followed a similar trend of significant improvement among the children in the study group (p < .01,27 p = .014,18 p = .02,22 p < .0525,29).
Karabay et al. compared gross motor function in three groups, out of which one group received kinesiotaping, second group received neuromuscular stimulation, and the third group received only conventional therapy, which was common for all the groups. A remarkable improvement was seen in all the groups, however the group that received neuromuscular electrical stimulation reported a greater mean difference in their gross motor function (11) when compared to kinesiotaping (6.84) and conventional therapy (4.47), along with a significant between-group p value of<0.01. 27 Two studies, out of which one used external head and trunk stabilizer and the other one used kinesiotaping on the trunk, 26 did not show any significant difference in the gross motor function between the groups, thus indicating that these interventions do not have any notable effect over the trunk, even after being incorporated with conventional therapy.

Trunk Control
Included studies used TCMS, 18,20,21 TIS 17,19 and SATCo 31 to assess trunk control. El Basatiny et al. reported significant inter group improvements in trunk control (p = .0001), 17 18 The interventions used in these studies were exercises based on NDT17, 18 and pelvic PNF 21 principles. The functional component within trunk control that showed significant improvement was dynamic sitting (p = .000117 and p = .00418). Sah et al. 19 also reported remarkable improvements in trunk control between the study and the control group (p = .014).
On the contrary, no statistically significant improvements in trunk control were observed by Curtis et al. 31 and Ari et al. 20 post intervention.

Balance
Balance was assessed using either PBBS [17][18][19][20] or FFR/FRT. 23,24 Sah et al. 19 reported a mean difference of 4.86 between the study and control group, thus observing a significant betweengroup improvement (p < .001). Similarly, Ibrahim et al. 23 also reported significant improvements between the two groups (p = .0001), on the application of taping.
Partoazar et al. used BBS as well as FFR to assess balance, which was assessed at four different intervals-before the application of kinesiotaping, immediately after application, after 48 hours with kinesiotaping in situ, and 48 hours after its removal. Inter-group comparisons reported significant improvements in dynamic as well as static balance, short time after application (p = .02, p = .06) and after removal of kinesiotaping (p = .01, p = .07). 24 On the application of trunk-specific exercises based on NDT principles, significant between-group improvements (p = .006) were observed in dynamic balance, by Akbas et al. 18 Remarkable improvements were seen in the study carried out by Ari et al. (p = .038), between the study and control group. The study evaluated the effects of NDT/Bobath based trunk protocol on balance in children with CP. 20

Discussion
The effect of trunk interventions in cerebral palsy has an impact on the different components of ICF. However, this review focuses on body functions, mainly three outcome variables, i.e., gross motor function, trunk control, and balance. Therefore, this review aimed to understand the importance of trunk targeted interventions in children with CP and to study its effect on gross motor function, trunk control, and balance.

Gross Motor Function
The use of individually structured trunk training in children with spastic CP facilitates upper limb, lower limb as well as trunk motor functions, 18 suggesting that exercise training for the trunk is an essential component in the rehabilitation of children with CP. Previous studies have shown an improvement in gross motor functions like rolling, laying, sitting, crawling, and standing after the application of NDT in children with spastic diplegia and quadriplegia. 33 NDT approaches, which help in achieving correct alignment of the trunk through trunk elongation and weight shifts to perform a controlled reach activity via facilitation through a hands-on approach, contribute toward achieving an optimal motor recruitment required to improve balance, trunk control, and gross motor functions. 19 Studies in the past that have assessed the impact of functional strength training on muscle strength, physical performance and psychological behavior showed an improvement in each function. [34][35][36] The loaded STS exercise program which used weights on the trunk in the study group showed similar effects by improving the GMFM score and muscle strength, along with decreasing energy consumption during walking. 22 Kara et al. performed taping on specific regions on upper limb such as wrist (buttonhole taping to correct wrist extension), scapula and forearm ("I" taping for stabilization), and lower limb such as hip ("I" band for hip abduction), knee (to correct hyperextension) and ankle (to correct dorsiflexion), in children with hemiplegic spastic CP. A remarkable improvement was observed in performance-related physical fitness, gross and fine motor function, and functional independence in the activities of daily living of the study group. 37 A systematic review by Ungar et al. to see the effect of kinesiotaping on children with cerebral palsy on gross motor function included five studies in total, out of which four studies are a part of this review, as taping was performed on the trunk. The results of this review suggest that kinesiotaping is beneficial and can be used as an adjunct to conventional therapy for children who fall under GMFCS levels I and II. 38 Electrical stimulation, which is well tolerated by pediatric population, when combined with task-oriented activities activates muscle contraction, thus improving the motor control. 39 Several studies have evaluated the effect of neuromuscular electrical stimulation on gross motor function in children with CP; the results of which have shown a significant improvement in the standing dimension of GMFM scale. [40][41][42] Karabay et al. 28 and Park et al., 29 however observed improvements in the sitting subset of GMFM on the application of electrical stimulation on the trunk in children with CP. The results of all these studies therefore suggest that the incorporation of electrical stimulation along with the conventional therapy can be considered as a safe and effective method of intervention, in children with CP.
Karabicak et al. compared the effect of electrical stimulation and taping on improving shoulder function in hemiplegic stroke patients; which showed that kinesiotaping was more effective than electrical stimulation. 43 However, on the contrary, neuromuscular electrical stimulation, when compared to kinesiotaping, given in adjunct to trunk based neurodevelopmental therapy, gave better results on the gross motor function in children with CP, and hence was considered to be superior than taping. 27 The reason behind this is that NMES stimulates an intact motor neuron, thus causing an active muscle contraction, whereas KT produces a tactile stimulation, unlike NMES. 44,45 Trunk stability is necessary for performing coordinated movements at the extremities, which is required for performing daily functional activities and motor tasks of higher levels. 30,46 The application of trunk belt showed significant increase in the sitting and standing scores of GMFM. 30 Improvement was attributed to the stability provided by the trunk belt that helps in performing rhythmical, smooth and precise movements, along with providing increased sensory awareness and proprioceptive input. 30 This suggests that the external trunk stabilizer should be incorporated along with the conventional physiotherapy program, to bring about an overall functional improvement.

Trunk Control
Trunk exercises mainly comprise selective trunk movements, which help in increasing trunk position awareness, strengthening the trunk muscles as well as making anticipatory postural adjustments. 17,18,20,47 Previous studies have showed a positive relationship between trunk position sense and muscle strength with trunk control. 48,49 Trunk targeted exercises included trunk elongation, active weight shifts and strengthening exercises such as pelvic bridging and trunk rotations, in order to achieve controlled mobility along with improving the strength of the back muscles. [17][18][19]50 These exercises bring about a significant improvement in static and dynamic sitting balance, dynamic reaching, as well as co-ordination. [17][18][19][20] However, the component that showed maximum improvement was dynamic sitting and dynamic reaching. 17,18 Pelvis and trunk are considered to be in a close relation concerning impairment and association. 51 Oommen et al. performed a cross-sectional study that examined the correlation between pelvic symmetry and trunk control, suggesting that pelvic training strategies should be incorporated in the rehabilitation program of children with CP. 51 Pelvic PNF techniques help in correcting the malalignment of the pelvis, thus improving trunk control, as pelvis is the framework that connects the trunk and lower extremity. [52][53][54][55] The improvement in trunk control suggests that pelvic PNF techniques can be included to improve trunk control in the exercise program of children with spastic diplegia.
Taping over the trunk has shown positive results in improving lower trunk range of motion. 56 Taping plays a vital role in supporting weak muscles, modulating tone as well as increasing proprioceptive perception. 23 Therefore, on considering all the above factors, taping over the trunk will allow children with CP to maintain a good and stable sitting posture, thus improving their trunk control.
There have been studies that have investigated the effect Swiss ball 57 and Both Sides Utilized Ball (BOSU) 58 on trunk. Both the studies observed efficient trunk muscle activation on the use of this equipment, the reason being that exercises performed on a smaller base activates more muscles when compared to a broad base. The results of these studies recommend the use of trunk stabilization exercises in the rehabilitation of children with CP. 57,58 Similarly, Curtis et al. suggested the incorporation of an external trunk stabilizer system along with conventional therapy, rather than using it as an independent approach. 31 Balance NDT principles state that the control over movements of the extremities occurs in a cephalo-caudal direction with the trunk, where the trunk has an important role in controlling the movements of the extremities, thus further contributing toward development of functional mobility as well as balance. 7,59,60 A positive relation between trunk control and balance has been determined in children with spastic CP, 7 thus implying that trunk targeted interventions contribute in improving balance, in children with CP. A non-randomized controlled trial performed by Ramya et al. reported improvements in static balance in the intervention group, on the application of NDT based trunk protocol, in children with CP. These results are like the studies performed by Akbas et al. 18 and Sah et al., 19 who observed positive results in dynamic balance, on using a trunk focused NDT program.
Choi et al. reported that strengthening the trunk muscles improved balance among children with CP, 61 the findings of which can be correlated to the study performed by Arı and Güne which reported an improvement in the trunk extensor strength on using Bobath-based trunk protocol. 20 Literature also provides evidence for the improvements seen in balance in children with spastic CP on the application of Swiss ball and BOSU.
Iosa et al. in 2015 in a case study stated that kinesiotaping can affect static and dynamic balance in children with cerebral palsy, as a result of which it can be used as a positive forward step toward the rehabilitation of these children. 62 Another study carried out on children with hypotonic CP where taping was performed over the trunk showed good sitting control in these children. 63 The results of all these studies are favorable to the incorporation of trunk taping along with conventional therapy, similar to the findings in the trials that used taping in this review. The study by Partoazar et al. in the year 2020 on taping showed an improvement in the dynamic balance in the selected population short time after application, and after removal, thus indicating that taping provides instantaneous and short-term results on dynamic balance. 24 On analyzing all the articles, this review suggests that a trunk targeted intervention for a minimum duration of 6 weeks, in the form of exercises based on Bobath principles such as pelvic bridging, upper and lower trunk rotation, and forward and lateral reaches, is required to bring about an improvement in PBS, thus significantly improving balance in children with spastic CP. 17,20,64

Strengths and Limitations
This is the first systematic review summarizing the effect of interventions specific to only trunk on children with CP, to the best of our knowledge. Most of the articles used in this review have performed their trials on children with spastic cerebral palsy, which gives us enough evidence for future studies in this population. We have included articles which include interventions like taping whose effects on the trunk have been tested for the first time on the selected population till date. Furthermore, we have included three outcome variables that have been assessed by their respective different outcome measure scales so that enough evidence can be collected to assess the effect of these interventions on them.
There are also limitations of the literature that were included in the review. The most common limitations were lack of follow-up and a small sample size. Also, there were a few studies that did not give a detailed information on the interventions that were performed.

Conclusion
The findings of our review draw us to a conclusion that trunk targeted training should be used along with conventional therapy program to bring about an overall functional improvement in children with CP. The trunk targeted interventions also improved gross motor function and balance along with trunk control. These interventions should be provided either in the form of exercises focusing on strengthening the trunk using electrical stimulation for the trunk musculature, using taping to provide stable posture, or using external stabilizers like trunk belt which provide support to the trunk. Neuromuscular electrical stimulation is superior to kinesiotaping, when provided in adjunct to the conventional therapy. In addition, incorporating NDT/ Bobath-based protocol for a minimum duration of 6 weeks will improve balance in children with spastic CP. However, the results of this review are restricted to children with spastic CP, since majority of the study population were children with spastic CP. Therefore, future studies must include other types of cerebral palsy as well, so that the results can be generalized.

Disclosure statement
No potential conflict of interest was reported by the author(s).