To be active through indoor-climbing: an exploratory feasibility study in a group of children with cerebral palsy and typically developing children

Seventeen children from the age of 11 to 13 years with (N = 11) and without (N = 6) cerebral palsy were recruited to this non randomized, non-blinded feasibility and intervention study. The children with CP were recruited according to their CP diagnosis regardless of CP type or functional level. The only exclusion criteria was severe functional dysfunction that made climbing impossible. Four of the participants knew each other prior to project (TD + CP, CP + CP). Seven of the children with CP attended mainstream school and 4 attended a school with a special program for children with physical and cognitive disabilities. All children in the TD group attended normal school.

The parents gave written and informed consent before the children participated in the study. The study was approved by the local ethics committee of the capital region of Copenhagen (H-B-2009-017). The study was performed in accordance with the Helsinki Declaration.

The participants were recruited through online advertisements. All the participants were unfamiliar with climbing training as an organized sport, a few of the participants had tried an indoor climbing wall before, but none were regular users.

Table 1 provides an overview of the participants including a note on the group that they were assigned to, based on the medical evaluation performed.

The climbing facility was a climbing gym with approximately 600 m² of climbing walls with walls and routes up to 12 m of height. The gym was reserved for this particular purpose during the hours that this intervention took place, which meant that the instructors and participants were the only ones present in the gym during the intervention. The gym was formerly the main climbing gym for the Danish Mountain and Climbing Club, but the club had moved to new larger (2000 m²) facilities. This older and smaller gym was mainly used for climbing activities with school classes. The first climbing day was performed in the larger facilities of a commercial climbing gym belonging to the Danish Mountain and Climbing Club, where the participants were provided with climbing shoes, which they used throughout the climbing period. During the first day other climbers were present in the commercial climbing gym. Nine days of climbing were planned within a period of 17 days, week 1 & 2: climbing on Monday, Tuesday and Friday, week 3: Climbing on Monday, Tuesday and Wednesday. All climbing was performed in the afternoon after school hours. Each climbing day consisted of approximately 2.5 h of physical activity. The 2.5 h were each time split into approx. 30 min of warm up exercises specifically focused on climbing. Subsequently the participants were split into two half with an approximately equal number of children with CP and TD participants in each group, where one group started with bouldering exercises, The other half of the participants were engaged in wall climbing with a top rope as safety. In order to avoid long waiting times during the top rope session, 3–4 instructors were allocated for the top rope practice, and one instructor was responsible for the bouldering session. After approximately 1 h, the two groups swapped activity. Each climbing day ended with a wrap up where the participants told what they had learned during the day. The intervention was planned as a sports activity, where the focus for the participants was to improve their climbing performance and experience that they learned new skills. Without focusing on specific needs of the individual participant with disabilities, we hoped that the focus of climbing performance also would give rise to therapeutical benefits, but intentionally without focusing on these specifically. This intentional lack of focus on therapeutical benefits was made, because we believe such a focus would remove the intrinsic enjoyment of climbing when applying external goals, see for instance [25]. The mixture of bouldering and top rope climbing provided a good foundation for a great diversity of dynamic and static movements, that challenged the participants motor skills and muscle endurance, demanding their full attention. Implementing both types of climbing also helped getting the participants to stay active and engaged longer, both physically by climbing and mentally by figuring out the routes and bouldering problems.

Three to five climbing instructors were present each day at the climbing gym during the intervention to supervise the 17 participants. The instructors were: one exercise physiologist (BSc), climbing instructor (level 3) and route setter; one neuroscientist, climbing instructor and researcher in this study, one school teacher and climbing instructor, one exercise physiologist student (M.Sc.), climber and researcher in this study, and one physiotherapist and PhD student on a project on CP and climber. The instructors did not have previous experience working with children with CP, but most of them had experience with climbing with children. The supervision consisted of constructive feedback and technical advice (e.g. active use of the legs and toes, advice on how to approach the different forms of holds such as jugs (large positive grips) and crimps (grips only held by the fingertips)) on how the participants could position their body, hands and feet in order to overcome obstacles and hard passages on the wall. Participants were also encouraged to help each other performing the different drills, and point out challenges to each other, when the instructors were not available for 1:1 coaching. During the 2 days with performance measurements (see next section) participants were cheering each other and encouraging one another to do their best. The participants were also engaged in observational learning when watching each other climbing, which have been shown to enhance motor skill learning by initiating formation of cognitive representation in memory that can be enacted and refined during overt practice, further enhancing consolidation of as much information as possible [26, 27]. Furthermore, we hoped that the vicarious experiences through observation of peers (e.g. the children with CP watching other children with CP climb) could promote a sense of personal efficacy for climbing, as it has been shown that observing or visualizing people similar to oneself perform successfully often raise efficacy beliefs in the observer [28].

To enhance social learning and strengthen group dynamics the instructors created a relaxed and playful atmosphere. It has been observed in previous studies that the feeling of relatedness (the feeling of being connected to others) enhances motivation and promote engagement in physical therapy [29‐31]. With that in mind a high extent of peer socialization between the children with CP and TD group was emphasized to promote positive attitudes and acceptance instead of alienation between groups. We hoped that through positive relations the groups could inspire each other to a higher degree of task persistence and thereby facilitate the participants’ intrinsic motivation (the inherent tendency to seek out new challenges, to explore and to learn [29]) towards the climbing intervention. Furthermore, it was of great importance that the climbing instructors acted as relational key figures for the children, to establish a secure relational basis that provided the foundation for growth-orientated activity participation. Therefore, the instructors implemented and supervised games with emphasis on social interaction among the children as a fundamental part of the climbing sessions to break down physical and mental barriers by having fun.

The warm-up exercises included group- and pair-plays, where all participants were playing with each other. The exercises included chain-catch, were initially two persons had to catch the others, and when one had been captured (he/she) had to hold hands thereby acting as a unity or a chain capturing the remaining participants. The participants also had to climb on each other’s bodies and roll across a line of participants.

The climbing exercises included top rope climbing, which is an individual endeavor. However, the instructors strongly encouraged the participants to cheer each other. The instructors continuously adjusted the selection of routes to the different participants according to whether they needed a challenge (choosing a route more difficult than the instructor and participant expected was possible or a good experience (choosing an easy route to gain confidence)).

The bouldering exercises included climbing on problems defined by the instructors. The instructors tried to select problems that were just above the abilities of the participant, in order to challenge them. Often the participants were put into small groups of 3–4 participants working on the same problem, which encouraged them to share ideas on how to perform the movements. The bouldering exercises also included purely movement based tasks, like holding the hands on two pre specified holds and then trying to touch as many holds as possible with the feet. The instructors also designs exercises where specific aspects of climbing techniques were challenged, such as focus on foot-placement on small holds and holding on to edges in various orientations.

Each climbing session was also finished with all participants sitting together and telling what they had learned during the session, and explaining whether they had been able to climb better than expected.

All the physiological, cognitive and psychological tests were performed at the Elsass Institute by an experienced physiotherapist specialized in child neurology or one of the two exercise physiologists that performed all assessments. The 17 participants were tested three times in all of the mentioned tests, unless otherwise stated. Two test rounds were performed prior to the intervention separated by approximately 14 days, and the 17 participants performed each test round within a period of 4–5 days on normal working/school days. The third test round was conducted the following week immediately after the last day of intervention. All the participants attended school and testing was performed such that it fitted within the participant’s normal week schedules.

Each test day lasted approximately 3 h in total for each participant. During each test day the order of tests were mixed.

Analysed data from the different tests were imported into R-studios (RStudio Team (2015) RStudio: Integrated Development for R. RStudio, Inc., Boston, MA URL http://www.rstudio.com /.) for further statistical analysis. A linear mixed model was used to investigate the effect of the climbing intervention on coherence by testing the specific interactions between group and intervention time and the differences between the values for pretest2 vs. the posttest in the CP and TD group. The large variability between participants was accounted for by including the ‘participant’ in the model as random effect, reparametrized interaction between group and time as fixed effects and age, height, gender as fixed effects. For all linear mixed models, Normal Quantile-Quantile plots were made, and analyses were rejected if these did not visually display an approximate linear relationship. In addition, residual plots were inspected visually in order to ensure near normal distributions.

In order to test specific hypotheses concerning the effect of the climbing intervention on the two groups we tested differences between pretest 2 and postest in the children with CP and in the TD group, and in order to test for unspecific learning effects due to multiple exposure to the same test, we tested the differences between pretest 1 and 2 in the children with CP and in the TD group.

Correlations were tested using a Pearson’s product-moment correlation.

Statistical trends ^o) p < 0.1, as well as statistical significant results *) p < 0.05, **) p < 0.01, ***) p < 0.001 are indicated when tested.

Four children with CP had tendon lengthening surgery 2–6 years prior the intervention and no children received antispastic medication.

All children with CP were diagnosed with bilateral CP except one participant that was diagnosed as unilateral CP. All children with CP were ambulant and categorized with GMFCS 1 except in one case with GMFCS 2. No walking aids were used by any of the children with CP but in 6 of the children ankle orthoses were used during daytime, but not during the climbing training. No hand or elbow orthoses were used in the group of children with CP.

The neurological examination revealed generally reduced ability to produce force bilaterally in dorsal flexors, plantar flexors, elbow flexors and elbow extensors. Spasticity was identified (>/= 1 on the MAS scale) in the plantar flexors in 7 cases and elbow flexors in 3 cases. Reduced range of motion was identified (dorsal flexion <20 degrees) in the ankle joint in 6 cases whereas no reduction in elbow was found in any of the cases.

The TD children all had normal age related strength, normal range of motion in all joints-, reflex activity and walking ability.

The primary outcome of this study is that children with CP perform an equal amount of physical activity as their TD peers without having to perform training interventions that are specialized for children with physical disabilities.

Table 3 summarizes the amount of time that was spent doing physical activity during the climbing training sessions. As the results indicate, both groups wore the armband approximately the same duration (within 2 min) and performed the same amount of physical activity.

Nine of the 11 children with CP took part in all 9 climbing days, one child with CP missed 1 day due to a logistic problem, another child with CP cancelled one climbing day due to a school arrangement. Three of the TD participants took part in all climbing days and three took part in 6 of the 9 climbing days. Cancellations in the TD group were due to illness, logistics or other arrangements, not lack of interest in the training.

Using the individual’s own time spent doing physical activity, which on average was 1:45 h. We calculated the total time of physical activity performed during the climbing sessions. The children with CP performed on average 16:05 h of physical activity during the climbing specific training and the TD group performed on average 13:01 h of physical activity taking into consideration how often each individual was present at the training. The IPAQ questionnaire disclosed that all children except two (one child with CP and one TD) made the climbing training as addition to their normal activities.

The measures of the performance during the climbing intervention test climbing route is summarized in Table 3. These results show that the children with CP improved climbing performance by the fraction of the test route climbed. However, the children with CP also increased their number of erroneous hand and footholds used during the test climb. The TD group improved climbing speed.

During a warm up play on one of the last days of climbing one of the TD participants accidentally sprained the wrist in a collision with one of the other participants.

Significant improvement was seen after the climbing intervention in number of times the children with CP were able to perform a sit-to-stand movement (pre intervention mean 20.2 SD 4.6, median 19 IQR 17; post intervention mean 23.5 SD 5.8, median 25 IQR 18.5; p < 0.01). In order to investigate whether the improvement in Sit-to-Stand was related to climbing performance, we performed correlation analyses of the change in Sit-to-Stand with the change in climbing performance (Climbing speed and fraction of route climbed) from pretest 2 to posttest in the children with CP. Neither of the two correlation analyses revealed significant relationships. No significant improvements were observed in the Romberg 30 s tests. Further details can be found in the Additional file 1: S1.

Results of the EMG-EMG coherence analysis during the 2 min of static ankle dorsi flexion are summarized in Fig. 2. We have displayed the pooled coherence for the children with CP (2A) and the TD group (2B). Comparisons of pre and post intervention in the ~10 - ~30 Hz range shows a trend towards a difference in pooled coherence for the children with CP but not for the TD group. The χ² difference confirms a slight, above the confidence interval, increase in EMG-EMG coherence for the children with CP, especially around the 20 Hz bin. To further investigate the changes in coherence from pretest2 to posttest we calculated the log normalized area under the coherence curve and tested it in a mixed linear model in the 15–30 Hz beta band. This confirmed the trend (z = 1.95, p = 0.0994) of coherence increases in the children with CP, but did not obtain significant difference.

Ankle strength was measured as torque around the ankle joint using the PSAD. Only data from pretest 1 and posttest was acquired and compared. The linear mixed model of TA strength did not reveal a significant difference between the pretest and posttest in the children with CP or the TD group. Further details can be found in the Additional file 1: S5.

CogState computerized tests was used to assess speed and accuracy in a detection task, and identification task (two choice reaction time task), in a 1-back working memory task, in a continuous paired association learning task and in a maze learning task.

The linear mixed model of the combined detection score did not reveal a significant difference between the pretest 2 and post test in the cardbased CogState computerized tests (the DETECT, IDENT and 1-BACK tasks) or the spatial location based tasks (the CHASE test, the MAZE test, the PAL test and the MAZEdelay test) in any of the groups. Further details can be found in the Additional file 1: S6.

The results of the self-reported questionnaire “Sådan er jeg”, where the participant should self-report their experience of themselves in 72 questions in 5 different domains did not show significant differences in any of the groups when comparing pre2 and post. Further details can be found in the Additional file 1: S7 (Table 4).

We have shown that children with CP can engage in climbing training on equal terms with their TD peers. The children with CP performed equally many hours of physical activity as their TD peers, which amount to more than 5 h per weeks of moderate or vigorous physical activity.

Furthermore, four children with CP and three of the TD children and their parents expressed a wish to continue the climbing training after the 3-weeks ended. On a side note it is also worth mentioning, that three of the children after the 3-week period continued climbing as an organized sport in a local climbing club.

In the present study we have shown that children in the age range of 11 to 13 with and without cerebral palsy can improve climbing abilities with less than 3 weeks of intense climbing training. Furthermore, we have demonstrated that some aspects of physiological measures as well as functional measures can be improved in children with CP with less than 3-weeks of climbing.

In summary, children with CP improved the fraction of the test route climbed, but at the expense of producing more hand placement mistakes while climbing. The children in the TD group did not improve the fraction of route climbed but increased their climbing speed.

With respect to functional improvements, children with CP improved their Sit-to-stand score, a measure used as a general measure of motor abilities.

The physiological tests revealed a significant increase in EMG-EMG coherence in the pinch task for the children with CP after the climbing intervention. Furthermore, the children with CP increased RFD in their least affected hand after the climbing intervention and had increased range of motion in the ankle joint.

None of the cognitive computer tests or the psychological questionnaire revealed any significant changes in either of the two groups when comparing before and after climbing training.

This study extends upon previous studies of biomechanical and physiological effects of sport climbing [7, 46‐50]. Previous studies of psychological effects of rock climbing has previously focused on performance anxiety [51‐53], attention [11] or the experience of flow [25] but to our knowledge no other studies has focused on effects of climbing training upon other psychological domains.

We have shown that a short intense intervention period of climbing training can improve climbing performance in children with CP GMFCS level I and II. We have also shown that a very basic test of general functional performance reveals improvements after the climbing intervention. Furthermore we have shown that fine motor control of the index and thumb is changed as reflected by an increase in coherence in the most affected hand of the children with CP. In the least affected hand the children with CP improved RFD from 0 to 100 ms. Finally, we have shown that climbing training increase the ankle joint ROM in the children with CP measured on the most affected lower limb. These findings suggest that climbing training have physiological impact on both upper and lower limb physiology and that climbing training can improve aspects of general functionality as measured by the Sit-to-stand test.

The psychological as well as cognitive tests did not reveal any significant improvements. All though climbing training has mental aspects related to management of fear and anxiety and cognitive aspects related to decision making and planning we were not able to see any effects on the employed measurements. This may be due to several reasons. The number of participants was rather small compared to the groups that the psychological test (Sådan er jeg) has been evaluated on.

Climbing training seems to be a way to engage children with CP in many weekly hours of physical activity in a fun and motivating way that may help improve climbing skills, muscle strength, balance, as well as mental and social skills. Therefore, we propose climbing as an efficient clinical training method with several benefits for children with CP.

This is also what we experienced during the intervention. Although the children had very diverse physical capabilities, they all found routes that suited their different skill-sets. With guidance from the instructors the participants could progress on these routes, often in collaboration with their peers from both the CP and the TD group. The mixture of top rope and bouldering provided a good foundation for the participants to try many different aspects of climbing, and for the instructors it increased the possibilities for setting up playful climbing games where the participants could interact independently of their physical foundation.

The participants self-reliance in relation to their climbing abilities may have improved during the 3 weeks and that the immediate feedback that the nature of climbing provided (e.g. overcoming specific obstacles on the route), may have fueled the children’s pursue of their inherent enjoyment rather than external reward or influence. Some of the children even personalized the routes, as one of the children with CP expressed it: He wanted to overcome his ‘arch enemy’ (the route he progressed on for several weeks). The participants had surprisingly few complaints about the level of their physical exertion, even though the sessions were relatively long and challenged the participants both physically and mentally. This may reflect that the climbing environment and the instructors encouraged the participants to develop the right mindset and self-determination for pushing their limits. In that context, it was enjoyable to see that the children with CP met the challenge with the same enthusiasm as the healthy controls, even though the training was not specifically adapted to the children with CP. Future studies could have even more focus on customizing the climbing training in order to accommodate the specific needs of participants with disabilities. Especially it could be of interest to investigate if climbing training can be used as an effective tool to treat people with even more severe CP. It should also be noted that the clinical significance of climbing has started to be recognized in other patient groups as well (e.g. in people with multiple sclerosis and geriatric patients) [54‐56].

Only very few studies have been conducted looking at effect of (therapeutic) climbing training in children and adolescents [21‐23]. We did not find significant changes in hand strenght. The average hand strength measurements reported by pervious studies in youth climbers at a comparable age [21] are within the range which we also observe (between 0.25 and 0.69 (kg/body weight) in the most dominant hand. However, the ranges for the study by Balas is unknown. Balas et al. used 8 weeks of training where we only look at the 3 weeks of training. That may be a reason to why we do not see changes in hand strengt. Balas only observed changes in the group that performed a high volume of climbing. Unfortunately we do not have any reliable way of estimating the amount of climbed meters as Balas et al. had.

With respect to the intervention intensity, previous studies have used climbing once or twice a week for 8 weeks [21] for children that already were climbing, A study of climbing in children with special needs Mazzoni et al. (2009) used 1 hour of climbing once a week for 6 weeks [22]. Therme et al. 1992 used 6 training session (unfortunately, only access to the abstract was available) so further details of the intervention are not possible to obtain [23]. Our intervention protocol was more densely packed, but in terms of total duration pr. day, but also in terms of number of sessions per week (three vs. one or two. However, we do not believe this relatively higher intervention intensity had a major role to play in terms how compliance. For all cases of absentee, reasons such as already planned family arrangements, missing transportation to the climbing gym and illness were given. This relatively high weekly frequency of climbing was used, because the overall duration of the project had to be limited, but future interventions would more ideally be spread over a longer total duration with max 2 training sessions per, week, with a possible gradual increase in weekly frequency.

Children with CP are often confronted with the challenge that they are not disabled enough to participate in adaptive sports, but fall short in traditional competitive sports, which can lead to a loss of motivation and self-confidence [57]. Climbing as an intervention tool has the advantage that the participant competes against him- or herself and the many different routes makes it possible to adjust the level of difficulty according to the requirements of the individual. This provides a basis for successful experiences in a social environment, independent of the level of expertise. Climbers of different levels may climb together or alongside each other with the possibility of the more experienced climbers providing positive feedback to the less experienced, further enhancing good relations between climbers

One can hope that the climbing environment in the future can be an arena for children with CP where they can develop social, mental and motor skills alongside their healthy peers, without the fear of falling short of certain competitive standards. The use of climbing as a tool in physical therapy is still a largely uninvestigated area and therefore much more research should be done on the possible physical, psychological and social benefits of climbing.

This study was designed as an exploratory study in order to investigate whether climbing training has the potential to be used as a possible therapeutic intervention, which combines physical and psychological challenges. In order to take into account possible exposure bias to the employed tests, we decided to compare two pre-intervention tests against the pre-post intervention comparison. We tested the differences between the pre1 and pre2 test and only found a significant improvement of the chase test in the CogState test battery for the TD group. All other tests did not show any significant differences between the pre1 and pre2 test. We therefore believe possible learning effect are minimal, and therefore differences between pre2 and post are more likely due to an interventional effect rather than an unspecific learning effect. Some of the employed tests, such as the questionnaire “Sådan er jeg” and the CogState tests may be very general, and not specific enough to test effects of the intervention. In particular we believe the “Sådan er jeg” questionnaire, after having employed it, is not suited for these short term intervention studies. With respect to the CogState tests we hoped to see effects on the MAZE-LEARNING test because the skills that the test resembles are similar to what you learn when trying to memorize a climbing route. The remaining tests in the CogState battery were less climbing-skill related.

The study was designed with the purpose of exploring whether children with CP can engage in climbing training together with their peers and explore possible physiological, psychological and social gains from climbing training. To explore these parameters in the most reliable way, practical issues from the surrounding activities, such as getting to the gym, organizing climbing coaches, etc. was taken care of by the experimenters and not the participant’s and their parents to the same degree as “normal” climbing training would require. Therefore, the surrounding circumstances were somewhat easier to cope with for the participants and their parents. We employed a relatively high instructor-to-participant ratio with 3–5 instructors to 17 children. This was done because we wanted to focus on both bouldering and top rope climbing, where the latter requires someone who can belay (hold the rope through a breaking device) the climber. Under normal circumstances, part of a longer climbing training course, belaying would be something that was taught the participants in the beginning or taken care of by parents. But because we wanted to focus on climbing as a physical activity (with it’s related aspects) we did not teach belaying techniques during the 3-week session. Furthermore, the activities took place in the early afternoon after school, where the parents of the children were still at work. It has to be considered that the instructors were highly motivated towards the intervention and already prepared that special precautions had to be made in order to properly adjust the climbing training to the children with CP. We therefore acknowledge that this fact potentially could bias the effect of the intervention to favor the children with CP, but when an instructor deals with a participant it would be impossible not to individualize the instruction without paying attention to the individual participant’s needs. At the same time the setting of the climbing intervention took place in a safe and inclusive environment, where the instructors facilitated interation between all participants. Therefore, in order to implement children with CP in real life climbing settings, a challenge could be the recruitment of dedicated instructors that are willing to make an extra effort, preferably with some prior knowledge of CP. However, a specialized team has been established for 6 children with CP where they climb once a week in a public climbing gym. Further, the disability level of the children with CP in this study has to be considered. The Children in our study was (with one exception) level 1 on GMFCS scale and therefore had a reasonable level of functional capability. Therefore, another consideration is how well a climbing intervention like ours could be implemented for more severely challenged children with CP.

Despite the very small group of participants included in this study, we were able to show some changes in the functional and physiological tests. However, given the large number of tests we performed none of the tests would survive strict Bonferroni correction for multiple comparisons. We have therefore decided to show all tests uncorrected in order to give a better impression of the raw results from the individual tests. We also decided not to correct for multiple comparisons because many of the tests cannot be considered completely independent. As an example, we found high correlations between performance in the computer based chase test and most of the other computer based cognitive tests. Also, tests like the four different RFD measurements are non-independent because they include data from the same time-interval.