The Effect of Adaptive Sports on Individuals with Acquired Neurological Disabilities and Its Role in Rehabilitation: A Systematic Review

Abstract

The aims of this systematic review were to report on the feasibility of adaptive sports for individuals with acquired central neurological lesion; to analyze the effects of this approach according to the domains of the International Classification of Functioning, Health and Disability (ICF); and to emit guiding points for future research. Two authors searched PubMed, Scopus, Cochrane, Pedro, and SPORTdiscus for eligible trials. Data concerning demographics, outcome measures, results, and conclusions were extracted, and a qualitative synthesis was performed. Adaptive sports seem to be a feasible, efficient, and cost-effective complement to conventional rehabilitation. Significant effects were found on all domains of the ICF, except “environmental factors.” Key factors, such as intervention volume, intensity, and type, play a determining role. This review is the first to expose the beneficial effects of adaptive sports practice among individuals with neurological lesions by relying on prospective evidence.

Introduction

Sedentary behaviors are high among individuals with an acquired central neurological lesion (¹). Less than 20% of this population manage to achieve the 2010 World Health Organization recommendations, namely to engage in minimum 150 min of moderately intense physical activity per week (²). Inactivity further increases with age and degree of mobility impairment (³). However, regular engagement in physical activity decreases morbidity and mortality rates following motor disability, and even induces disease-modifying effects (^4,5). By improving the individuals' physical conditions, activities of daily living, social participation, and quality of life improve (⁶). Engaging in physical activity therefore has the ability to decrease disability, as defined by the International Classification of Functioning, Health and Disability (ICF) (⁷). However, long-term adherence to physical activity remains low (⁸). Overcoming this challenge is crucial to sustain improvements (⁹).

Physical activity programs require engagement in energy expenditure through bodily movements (¹⁰). Such programs, which have become essential components of neurological rehabilitation, lack motivational stimuli, and patients often stop engaging in these activities after discharge from rehabilitation (¹¹). Sport practice, however, focuses on performance and involves rules, competition, and team work. These factors increase sense of enjoyment and adherence (¹²).

Although the literature on the benefits of sports among the physically abled has been extensive, little has focused on sports adapted for populations with physical disabilities. Nevertheless, authors believe that by requiring task-specific, repetitive, and intense training, sport practice stimulates motor recovery and improves motor function following neurological lesion (¹³). Adaptive sports could, therefore, be used as a tool to improve rehabilitation. As populations of individuals with central neurological impairments will substantially expand in the coming years (¹⁴), it is crucial to research efficient and cost-effective treatments to integrate in rehabilitation. However, to date, no clinical reviews have explored the topic at hand.

The aims of this systematic review are to: 1) report on the feasibility of adaptive sports as a rehabilitative tool; 2) investigate its effects according to the on motor disability following acquired neurological lesion ICF framework (⁷); and 3) analyze the available literature in order to emit guiding points for future research.

Methods

Research Strategy

This systematic review was registered in PROSPERO (CRD42019125534). The research process was carried out in accordance to the 2009 PRISMA guidelines (¹⁵). The population, namely, adults with acquired central neurological disability, was specified into three subgroups: spinal cord injury (SCI), stroke, and multiple sclerosis (MS) as these are among the costliest neurological disorders (¹⁶).

Literary Search

Keywords were identified using the Populations, Intervention, Comparison, Outcome (PICO) method. Combining keywords in relation to the populations (stroke, multiple sclerosis OR MS, spinal cord injury OR SCI, wheelchair users, physical disability, neurological disability) and the intervention (adaptive sport, wheelchair sports, swimming, basketball, rugby, handbike, dance, horse-riding, golf, etc.) using boolean operators, allowed to construct different research equations, introduced into PubMed, Scopus, Cochrane, Pedro, and SPORTdiscus. All references were exported from the databases to Endnote. Reference lists of relevant trials and reviews were examined.

Study Selection

Articles were included when 1) the reported language was English or French; 2) the trial followed a longitudinal and prospective design; 3) the investigated populations were adults with stroke, SCI, or MS; 4) the intervention was an adaptive sports program; and 5) the primary outcome measured physical and/or psychological variables. Sample groups referred to as “wheelchair users” were included when the three investigated disabilities made up more than 50% of the sample. Articles were excluded if 1) the article was a review or case-report; 2) the investigated population included cognitive, mental, and sensorial disabilities, or was specific to veterans; 3) the intervention was a physical activity program (could not lead to competition) or required the use of any type of technology; and 4) the primary outcome focused on immediate physiological effects.

Two researchers (L.D. and G.S.) independently screened titles and abstracts of the retrieved studies to assess for eligibility, after removal of duplicates. Articles free from exclusion criteria were selected for full-text lecture, and corresponding authors were contacted when full-text was not available. Articles containing all inclusion criteria were included for qualitative synthesis. Disagreements were resolved through consensus.

Data extraction and synthesis

Results were extracted into a piloted table (Table 1) by a single author (L.D.), with verification from a second author (G.S.). Articles were divided according to study design and ranked in a descending order according to quality. Sample demographics, including size, motor disorder, age, sex ratio, and phase of rehabilitation were extracted. Early phase referred to participants within the first 6 months of rehabilitation. The outcome measures, interventions (type, duration, frequency, and intensity if available), results (difference in means within or between groups in points or percentage when P < 0.05, effect size if available), and conclusions were extracted. Results were synthesized in a qualitative and narrative format, according to the ICF domains and components, with their corresponding code.

Table 1: Summary of included trials characteristics — demographics, protocol, outcome measures, and results.

Level of evidence and quality appraisal

Level of evidence was attributed according to study design. Randomized controlled trials (RCT) represented the highest level of evidence (level I), followed by controlled trials (level II) and single-cohort studies (level III). Study quality was appraised by a single author (LD) using items of the Downs and Black scale, which has high internal consistency, reliability, and accuracy (⁴⁷). Items assessing external validity, internal validity, and power were used, whereas reporting items were excluded as these did not differ according to study-design. Moreover, as the modality of the intervention prevented the masking of subjects to attribution, item 14: “attempt to blind subjects to the intervention” was removed. Sixteen items were, thus, checked. Studies scoring above 10 were of good quality (or excellent if above 14), whereas studies scoring less were of mediocre quality (or poor if below 5).

Results

The search was conducted from October 2018 to March 2019. From the 4101 articles screened, 30 were selected (Fig. 1). Among the included studies, 16%, or n = 5 trials, followed an RCT design (level I), 31% (n = 9) had a control group which did not receive adaptive sports, without random allocation (level II), and 53% (n = 16) followed a single cohort (level III). Average methodological quality was mediocre (8/16) and ranged from 3 to 16 on the modified Downs and Black scale (Table 2). Item 27, concerning power to detect clinically important effects, was the most absent, deteriorating internal validity.

Figure 1: Prisma 2009 flowchart.

Table 2: Items present or absent on the modified Downs and Black scale.

Outcome measures concerned all domains of the ICF, except “environmental factors,” commonly studied through cross-sectional designs. “Bodily functions and structures” was the most studied domain, with n = 24 trials investigating this domain, followed by “activity and participation” (n = 9), grouped due to close interaction between both domains, and finally “personal factors” (n = 6). Five trials investigated quality of life, and six studied feasibility. A wide range of components of the ICF domains were studied through heterogeneous tests and scales. Lack of RCT, together with the variety of scales used to measure a range of outcomes, limited the analysis of the literature to a qualitative synthesis.

A total of 741 participants were recruited, ranging from samples of 4 to 123. More than 65% were male participants. Average age was 42 years, with stroke patients being the oldest group (averaging 58.6 years) and SCI subjects, the youngest one (averaging 32.6 years). The latter population also was the most researched, with 44% of the included studies having samples solely made up of individuals with SCI. Studies enrolling only participants with stroke or MS made up 23% and 20% of the studies, respectively. The remaining 13% were referred to as “wheelchair users” and had heterogeneous groups in terms of neurological causes of motor disabilities (spina bifida, lower-limb amputations, etc.). Finally, 16% of the included studies were performed during the early phase, whereas 84% evaluated late phases.

Eleven different types of adaptive sports were investigated (Fig. 2). The most researched adaptive sport was handbiking. Two studies investigated a mixed approach, where various sports were practiced. The modal intervention span was 8 wk (range of 5 d to 5 years) and modal frequency was twice weekly (range of once weekly to daily). Median volume of the adaptive sport intervention was 21 h (range of 6 to 982 h). Intensity was rarely reported. Finally, 87% of the studies analyzed postinterventional changes only, and 13% integrated follow-up measures, varying from 1 to 6 months posttrial.

Figure 2: Adaptive sports evaluated and percentage of included trials investigating it.

Feasibility of adaptive sports

A 5-wk or 8-wk group kickboxing program among 15 individuals with MS was estimated to be safe. Compliance was maintained in the 90%, and adverse events did not occur (^36,38). A likewise study taking place in a community-setting found equally high compliance to a dance intervention among eight individuals with MS (³⁴). Similarly, nine stroke patients undergoing 4 wk of dance and 45 SCI patients undergoing 8 wk of handcycling displayed excellent compliance and safety (^37,41). Both were performed in hospital settings during early rehabilitation.

The Effects of Adaptive Sports According to the ICF Framework

Bodily functions and structures

Cognitive and mental performance (b110-b139). Two level II (^22,24) and one level I (²⁰) trials found improvements in cognitive and mental performances following sports. Golf and social communication equally improved visual spatial memory of 14 individuals with stroke, though golf was superior at improving visual imagery, with a high effect size (d = 1.49) (²⁴). Ballroom dancing improved global cognitive function of six participants with MS by 0.9 points (²²). Hatha yoga, but not sports climbing, improved selective attention among 20 individuals with MS. Executive functions did not alter in either group but sports climbing significantly reduced chronic fatigue by 32.5%, whereas yoga (²⁰) and ballroom dancing did not (²²).

Cardiovascular function (b410-b429). Level II evidence demonstrates that among 17 men with SCI, the group engaged in various sport activities over a 5-year course had significantly decreased overall carotid intima-media thickness and diameter, by 0.18 and 0.044 mm, respectively. The control group did not show improvements in cardiovascular function (³⁰).

Pulmonary function (b440-b449). Forced vital capacity, forced expired volume after 1 s and maximal voluntary ventilation of 15 male tetraplegic subjects significantly increased after a year of wheelchair rugby, by 9.12%, 9.05%, and 6.5% respectively, in a level II trial. Correlations between total training time and forced vital capacity (r² = 0.97) and maximal voluntary ventilation (r² = 0.58) were found (²⁶). However, handcycling did not induce significant improvements in pulmonary function of 56 SCI participants in one level III (³²) and one level II trial (²⁹).

Physical endurance (b450-b469). Four level III trials (n = 136 subjects) found that engaging in adaptive sports significantly improved V˙O_2peak and PO_peak of wheelchair users, by percentage increases ranging from 7% to 9.6% and 17.7% to 36.4%, respectively (^32,37,45,46). Another trial of equal evidence-level found that self-guided handcycling had a moderate effect size on PO_peak among 18 subjects with SCI (d = 0.51) (⁴²). However, no significant improvements in physical fitness were observed among 24 tetraplegic subjects who had participated in at least 6 months of wheelchair rugby training in a level II study (²⁵).

Body composition (b598-b599). Three level III trials found significant changes in body composition following sports. Eighteen individuals with SCI preparing for the “Handbike battle” during 4 months underwent significant improvements in body composition but effect size was very low (d = 0.12) (⁴²). Waist circumference reduced significantly by 4% among 57 wheelchair users preparing for the “Handbike Battle” (⁴⁶). Thirteen tetraplegic subjects significantly improved trunk, arms, legs, and total body lean mass, fat mass, and bone mineral content, measured by dual-energy X-ray absorptiometry, after 8 months of wheelchair rugby (³³).

Muscle function (b730-b749). Level II evidence found handcycling was more efficient than conventional care at improving isometric peak strength of elbow flexors and shoulder rotators of 34 SCI subjects, with more than a 10% difference in change between the groups (²⁹). This is verified by a level III trial who demonstrated 5% increased isometric peak abductor strength following handcycling (³²). Two level II trials found conflicting results regarding the impact of wheelchair rugby on muscle strength. One reported significant improvement greater than 10 Newtons in isometric upper limb strength among untrained tetraplegic player (²⁵), while the other reported no significant increases in upper-limb muscle strength, assessed manually, following wheelchair rugby or conventional care among 40 tetraplegic subjects (²⁷). Sports climbing decreased spasticity by 25% in one level I trial, measured by the Expanded Disability Status Scale—Pyramidal Functions among 20 individuals with MS (²⁰).

Movement function (b750-b789). Level I evidence supports the use of sports to enhance movement functions. Six weeks of sitting-boxing resulted in significantly greater Manual Function Test scores of 26 stroke patients than the control group receiving conventional therapy (²¹). A 7-wk group volleyball program improved measures of upper limb function, such as the Box-and-Block Test and the Wrist Position Test, significantly more than conventional rehabilitation on a sample of 48 stroke participants (¹⁸).

Physical performance (b798-b799). Level III trials found physical performance did not significantly improve over the course of a season among 20 elite wheelchair basketball players when measured through field tests (^43,44). Level I (^17,21,23) and III (^{24,34,36,38,40,41}) data show balance significantly improves following golf, dance, horse-riding, sitting boxing, or kickboxing interventions among 242 participants with stroke or MS. Improvements in dynamic balance were found among MS individuals after a month-long dance intervention in one level III trial. Improvements remained significant at a 3-month follow-up though not at 6 months postintervention (³⁴). Ballroom dancing did not significantly improved balance of 12 participants with MS in one level II trial (²²).

Activity and Participation

Mobility (d450-d469)

Level III evidence showed that a 10-wk dance program did not significantly improve spatiotemporal gait parameters of 20 stroke patients (⁴⁰). This opposes level I evidence demonstrating improved Timed Up and Go scores which were maintained at a 6-month follow-up following 12 wk of horse-riding therapy among 123 individuals with stroke (¹⁷), and improved 10 m walk tests scores by 5.78 s following 6 wk of boxing among 20 stroke participants in the early phase (²¹). Level II (^23,27) and III (^34,36,38) evidence verify these improvements. Eight and 5 wk of kickboxing improved Timed Up and Go measures and gait speed of 15 subjects with MS, with a moderate effect size (r > 0.45) (^36,38). A 4-wk dance intervention improved Timed Up and Go scores of a similar group of patients posttraining and at a 3-month follow-up (³⁴). Horse riding improved stride time and ground-reaction forces, of 27 individuals with MS (²³). The Wheelchair Skill Test improved significantly more following 2 years of wheelchair rugby training than conventional therapy among 40 tetraplegic subjects (²⁷).

Independence (d510-d599)

Level I (¹⁷) and level II (²⁸) trials support the use of sports to enhance independence. Paraplegic and tetraplegic individuals assigned to a 4-month swimming intervention improved their Functional Independence Measure score by 7.75 points, which was significantly more than the control group receiving usual care (²⁸). Perception of recovery measured by the Stroke Impact Scale improved by 23.8%, which was significantly more after a horse-riding program than standard care among stroke patients (¹⁷). On the other hand, 6 months of horseback riding or conventional physiotherapy did not induce significant improvements in performance of activities of daily living, measured by the Barthel Index, among 27 individuals with MS in a level II trial (²³).

Contextual factors

Mood. Golf and social communication interventions equally improved mood of 14 participants with stroke (²⁴). A level IIII study found significantly reduced depression following sport participation of 40 SCI patients (³¹). Conversely, mood did not augment after sports climbing or ballroom dancing programs among 32 individuals with MS (^20,22).

Self-perception. Level and stability of self-perception increased following a 5-d skiing program in a trial including 10 individuals with SCI of level III evidence. These increases remain significant at a 4-wk follow-up (³⁵).

Motivation for leisure time. A 1-month dance intervention increased levels of physical activity at a 3-month follow-up, measured by the Godin Leisure Time questionnaire, of eight individuals with MS. These increases correlated to scores on the Motives for Physical Activity Measure-Revised (³⁴).

The Effect of Adaptive Sports on Quality of Life

A 6-wk boxing program significantly increased quality of life by 12% on the Stroke Specific Quality of Life questionnaire, which was significantly more than following physical therapy, among 26 stroke patients (²¹). Level III evidence showed improvements of in the Short-Form health survey scores by 7% among 16 paraplegics following 1 year of wheelchair basketball. Significant changes occurred in the functional capacity, general health, and emotional aspects categories while decreasing in the mental health (³⁹). Level I evidence shows that horse-riding produced significantly greater improvements on the functional capacity, physical aspects, and mental health among 24 stroke patients than conventional physiotherapy (¹⁹). Quality of life did not significantly improve after 5 wk of group kickboxing program for 11 individuals with MS (³⁶) but did improve by 42 points after 6 wk of ballroom dancing among a similar sample (²²).

Discussion

To our knowledge, this systematic review is the first to investigate the effect of adaptive sports on individuals with acquired neurological disabilities by relying solely on prospective evidence.

Feasibility

Adaptive sports are feasible in both early and late phases (^34,36,38,41), allowing for continuity. Participants are compliant and satisfied with this intervention (^34,36,38,41). Sports can be practiced on a regular and long-term basis. Its practice is cost-effective by increasing the instructor to participant ratio (³⁴) which could reduce the high medical costs associated to rehabilitation (⁴⁸).

Efficiency: Bodily Functions and Structures

Bodily functions and structures improve following adaptive sport interventions, as shown by trials of mostly II to III evidence-level. Sport practice shields individuals from the harmful effects of a sedentary lifestyle (^30,33) and cross-sectional data indicates that athletes with SCI require fewer medical appointments than their nonactive peers (⁴⁹). However, adaptations of bodily functions and structures seem to depend on key factors and should be measured through sensitive testing. Field tests and manual assessments should be avoided (⁵⁰).

Training volume (= frequency × duration) and intensity are two determining factors. Higher volumes of training, which exceed 100 h, are more efficient at inducing changes in aerobic fitness and pulmonary function than shorter trainings (^29,32,45). Moreno et al. (²⁶) demonstrated that tetraplegic subjects undergoing larger volumes of rugby training experienced greater pulmonary improvements. Further research is needed to investigate the exact volume required. Similarly, though the exact intensity needed to induce bodily adaptations still remains unknown, moderate to high intensity sports are preferred. Adaptive rugby seems less efficient at improving physical endurance than handbike training (^{25,32,37,42,46}). This sport has been found to be less intense and result in lower energy expenditure than other wheelchair sports (⁵¹). Improvements in mental and cognitive functions, such as fatigue, similarly depend on training-intensity. Whereas sports climbing significantly reduced chronic fatigue of subjects with MS, yoga did not (²⁰).

A third factor is the presence of competition, which enhances the benefits achieved from sport practice (¹⁸). The anticipation of competition also increases commitment to training, inducing improvements in physical fitness and health outcomes (^42,46).

Finally, physical effects of sport-training largely depend on the type of sport being practiced. Handbiking results in increased upper-limb strength (^29,32), which correlates to activities of daily living among paraplegics and tetraplegics (⁵²). Balance improves following golf (²⁴), dancing (^34,40,41), boxing (^21,36,38) and horse-riding (^17,23) among stroke or MS participants. However, Ng et al. (²²) did not find that ballroom dancing significantly improved balance of individuals with MS. This type of dancing requires physical contact and participants may have leaned on non-MS partners. It is thus preferable, for skill-improvement, to rely on as little outside assistance as possible.

Efficiency: Activity and Participation

Participating in adaptive sport programs allows individuals to recover performance in activities of daily living (^18,21) and in mobility, as shown by trials of strong evidence level (^{23,27,34,36,38,40}). Furmuniak et al. (²⁷) demonstrated that participants with inferior ASIA scores at baseline underwent greatest increases in wheelchair skills after wheelchair rugby, as these subjects had more room for improvement. However, sensitive testing methods are required to note improvements in activity and participation. Ceiling effects may account for a lack of significant changes, especially among groups with adequate levels of activity at baseline (²³). Global measures of mobility, such as the Dynamic Gait Index and the Timed Up and Go, are appropriate to objectify improvements (^17,30).

Adaptive sport participation increases independence and perception of recovery more efficiently than conventional care (^17,28). Improving autonomy decreases the economic burden of disability and the burden placed on the participant’s informal care giver (^17,53). Cross-sectional data further found correlations between adaptive sport practice and improved autonomy in everyday life activities, social participation, integration in the community, and rate of employment (⁵⁴). No longitudinal study has measured these variables thus far.

Efficiency: Contextual Factors

Personal factors such as mood improve following an adaptive sport program, especially when there is presence of a team (^19,31,35,39). Motivation to engage in physical activity increases following adaptive sports participation (^{29,32,36,38,40}). Mandelbaum et al. (³⁴) found that 4 wk of salsa class achieved to induce significant increases in leisure time activity postintervention and at the 3-month follow-up period among individuals with moderate MS. Long-term improvements in level and stability of self-perception also occur following adaptive sport participation (³⁵). Improving physical performance and motor control through sports allows participants to reconnect with their body.

Efficiency: Quality of Life

Trials of varying evidence-levels demonstrate the long-term effect of sport participation on quality of life after stroke (^19,21) and SCI (^17,39). However, a 5-wk group kickboxing program did not significantly augment quality of life among participants with MS (³⁸). A ceiling effect may account for this lack of effect as the sample was made up of individuals with mild disability who were not clinically depressed at baseline. Using more sensible testing methods to assess quality of life, such as the PROMIS-GH, yields more encouraging results among this population (²²).

Limitations

The literature available on adaptive sports is extensive, but only few studies have analyzed the effect of this approach through prospective trials, and most have not adopted an RCT design. Nonblinding of the evaluators further deteriorated methodological quality. Moreover, the frequent use of questionnaires and self-assessment evaluations predisposed the data to subjective interpretation and bias.

Numerous inclusion criteria limited the samples to small numbers. More severely affected subjects were frequently excluded, which limits the applicability of results. Participants were often recruited through volunteer-based processes thus selecting motivated individuals. Valent et al.'s 2009 study, who recruited tetraplegic individuals discharged from rehabilitation centers, found that participants dropping out had significantly lower physical capacity than those completing the intervention (³²). Conclusions drawn from individuals completing the study difficultly extend to the population. Moreover, it was not always reported whether participants had engaged in adaptive sports before beginning the trial. While n = 4 trials evaluated individuals involved in adapted sports clubs, and n = 5 trials were conducted in the early phase, thereby including sample groups which did not engage in such activities previously, this confounding factor was generally not disclosed. Bias could subsequently occur as previous experience in adaptive sports practice may influence reaction to intervention.

Heterogeneity of the sample groups, intervention protocols, and outcome measures further complicated between-study comparisons. Within-study comparisons were difficult as control groups undergoing conventional therapy received less treatment than the intervention group. As improvements occurring in rehabilitation are strongly correlated to the volume of treatment (⁵⁵), it is difficult to attribute the changes occurring in the intervention group solely to the sport.

Finally, a lack of follow-up evaluations makes it difficult to draw conclusions regarding long-term efficacy. Follow-up data may considerably differ from postintervention data, especially among individuals with degenerative deficiencies such as MS (³⁴). Partial to complete losses of exercise-induced adaptations also have been observed 4 wk after training cessation among physically-abled adults (⁹). One can assume that a detraining phenomenon also is present among individuals with physical disabilities. Maintaining benefits thus entails sustained participation.

Guidelines for Future Research

Future research should continue investigating the effects of adaptive sport programs on individuals with acquired central neurological lesion, especially during early stages. Optimal intensity and volume should be explored, as well as how these factors are best adapted according to the patient. Prospective studies should further investigate psychosocial effects of adaptive sports.

Randomization of subjects and blinding of evaluators should be assured. Objective data should complement subjective measures. The control and intervention groups should receive equal amounts of therapy and follow-up measures should not be neglected. This will allow to adopt an evidence-based approach when integrating adaptive sports programs in clinical fields.

Conclusions

Adaptive sports programs seem to be a feasible, safe, and cost-effective complement to all phases of rehabilitation. Participants are satisfied and compliant to this approach, and adverse events are scarce. Patients with motor disabilities stemming from neurological lesions experience an array of positive outcomes following adaptive sport participation, and this on all domains of the ICF. However, these effects largely depend on the volume, intensity, and type of sport, as well as the presence of a team and competition. Its feasibility and effectiveness are essential factors in making adaptive sport a valuable therapeutic tool, which can be instated as a complement to conventional therapy.