Background Balance function is central in walking, and impaired balance function may be related to walking deterioration in adults with spastic bilateral cerebral palsy (CP).
Objectives The purposes of this study were: (1) to compare balance confidence, fear of falling, and balance ability in adults with spastic bilateral CP, with and without self-reported walking deterioration; (2) to characterize balance confidence, fear of falling, and balance ability across all participants; and (3) to examine the relationship between balance confidence and balance ability across all participants.
Design A case-control design was used.
Methods Sixteen adults from a 7-year follow-up study who had spastic bilateral CP and were under 40 years of age in the 2006 survey participated. Eight participants reported walking deterioration (cases), and 8 participants did not report walking deterioration (controls). Outcome variables were: the Activities-specific Balance Confidence (ABC) Scale, the Falls Efficacy Scale–International (FES-I), and the Balance Evaluation Systems Test (BESTest).
Results No differences in any of the outcome variables were found between the cases and the controls. Across all participants, the ABC Scale and FES-I scores were 62% and 24 points, respectively. Reduced ABC Scale scores and increased FES-I scores were found when using escalators, walking in crowds, and walking on slippery surfaces. The BESTest subscale scores were 60% to 79% of the maximum score, but only 31% and 42% of the maximum score in postural responses and anticipatory adjustments, respectively. Balance confidence correlated positively with postural responses, sensory orientation, stability in gait, and BESTest total score.
Limitations The lack of reliability and validity tests for the outcome variables in this study population and the small number of participants were limitations of the study.
Conclusions Self-reported walking deterioration in this group could not be explained by differences in balance confidence, fear of falling, or balance ability. Across all participants, most balance problems seemed related to reduced postural responses and anticipatory adjustments.
High levels of fatigue, pain, and deterioration of walking have been reported in adults with cerebral palsy (CP).1–7 In our 7-year follow-up study of people with spastic unilateral and bilateral CP in 2006,7 a substantial increase in the proportion of people reporting walking deterioration was seen in the ages of 35 to 40 years in people with spastic bilateral CP. Those reporting walking deterioration also reported significantly higher levels of both musculoskeletal pain and physical fatigue. Seventy-eight of the 149 people with spastic CP reported deterioration in walking in adult age, and 51 of the 78 reported that this deterioration was characterized by reduced balance.7 These findings led us to focus on balance as an important factor for early walking deterioration in people with spastic CP.
Postural stability, also referred to as balance, is the ability to control the center of mass in relationship to the base of support.8 In order to maintain balance during a certain task, complex processes take place involving vestibular, somatosensory, visual, cognitive, and musculoskeletal systems, controlling the relationships among the different body segments and between the body and the environment.9,10 Balance has been described as one of the prerequisites for walking11 and was found to be impaired in children with CP.12,13 Burtner et al14 found that children with spastic bilateral CP had a proximal-to-distal muscle activation pattern, more co-contraction, and slower muscle activation compared with children with typical development. To the best of our knowledge, studies of balance in adults with CP have not been carried out.
In order to gather more information about balance in adults with spastic bilateral CP, we invited 3 adults to share their experiences in a group discussion. They revealed that walking on uneven or slippery ground, walking in crowds, using escalators and public transportation, and walking down stairs were difficult situations that challenged their balance reactions. They often felt too slow and unable to prevent a fall. Impaired vision and reduced lighting conditions presented additional problems. These difficulties often led to compensatory strategies such as holding on to railings or other objects. Some participants in the group discussion regarded themselves as “expert fallers” (ie, they often fell but had not had serious injuries and were not afraid of falling). Others reported that they were afraid of falling.
These described difficulties may lead to an increased risk of falling, injuries, and a gradual reduction in walking frequency and distance, thus contributing to walking deterioration. This deterioration might reduce activity participation in the community at a far earlier age than would be expected due to normal aging. However, studies of balance ability and its relationship to walking deterioration in adults with spastic bilateral CP are lacking.
Our previous 7-year follow-up study7 indicated an association between impaired balance and walking deterioration, as 51 of 78 people regarded the deterioration as a result of reduced balance. Of those with unchanged or improved walking (n=71), 26 regarded their walking ability to be a result of improved balance.7 Walking deterioration was more common in individuals with spastic bilateral CP.7 Given these results, the lack of research in the area, and the information from the group discussion, the first purpose of this study was to compare balance confidence, fear of falling, and balance ability in adults with spastic bilateral CP, with and without self-reported walking deterioration. The second purpose was to describe balance confidence, fear of falling, and balance ability across all participants. The third purpose was to examine the relationship between balance confidence and balance ability across all participants.
A case-control design was chosen to compare people who reported walking deterioration (cases) with people who did not (controls).
The testing took place during the fall of 2009 in the Motion Analysis Laboratory at Sunnaas Rehabilitation Hospital, Norway, and was conducted by trained and experienced multidisciplinary personnel.
The participants were recruited from a 7-year follow-up study of 149 adults with spastic unilateral and bilateral CP in 2006.7 In a mailed letter, we invited all people with spastic bilateral CP who reported Gross Motor Function Classification System (GMFCS) levels I through III15 and were under 40 years of age in the 2006 survey to participate. Individuals who were unable to walk at least 10 m without walking aids were excluded. A flow chart of the inclusion process is presented in Figure 1. Sixteen adults (5 men, 11 women) with documented spastic bilateral CP, GMFCS levels II (n=15) and III (n=1), agreed to participate (Tab. 1).
The Activities-specific Balance Confidence (ABC) Scale was administered in a telephone interview by a physical therapist 2 to 3 weeks before the semistructured interview and the tests of balance ability in the laboratory. This physical therapist did not take part in the other assessments. In the semistructured interview, carried out by the principal investigator (A.O.), the changes in walking function over the years (since the end of adolescence) were reassessed and categorized as improved, unchanged, or deteriorated. Each participant's Falls Efficacy Scale–International (FES-I) and Functional Mobility Scale (FMS) scores, GMFCS level, and history of falls also were obtained during this interview. All assessments were carried out in the same order: first the semistructured interview, then tests of muscle tone and muscle strength, followed by the tests of balance ability. The principal investigator led and administered all balance tests, assisted by at least one other person (physical therapist or biomechanist) who observed and registered the performance of the tests of balance ability and aided in participants' safety. The principal investigator was blinded to the ABC Scale scores until after the tests of balance ability. When there were doubts about the scorings on the tests of balance ability, these doubts were discussed, and the principal investigator had the final word.
The GMFCS15 was used to assess gross motor function. The GMFCS categorizes gross motor function into 5 levels, with level I being the highest and level V being the lowest functional level. The need for railings when descending stairs was used to distinguish between levels I and II. The GMFCS has been found to be valid and reliable in adults with CP.16,17
The FMS18 was used to describe the level of independent walking on a scale of 1 to 6 points (1=wheelchair user, 6=fully independent on all surfaces) over 5, 50, and 500 m, respectively. The FMS has been found to be both reliable and valid in children with CP.18
Muscle tone in the triceps surae, rectus femoris, hamstring, and adductor muscle groups was assessed bilaterally using the 6-level modified Ashworth scale (MAS) (0–5),19 which assesses resistance to rapid, passive movements. Higher scores indicate higher muscle tone. The median MAS score for these muscle groups was calculated. Bohannon and Smith19 found an agreement of 87% between 2 raters who used the MAS to assess muscle tone in the elbow flexors, but the reliability of the MAS was found to variable in later studies.20 The validity of the MAS in the assessment of muscle spasticity has been questioned,20,21 but it is still the most common clinical scale for assessing muscle tone by grading the resistance to passive movements,22 acknowledging that this assessment also includes mechanical compliance in muscle and joints.23
Muscle strength in the hip flexors, extensors and abductors, knee flexors and extensors, and ankle plantar flexors and dorsiflexors, bilaterally, was assessed using the 0–5 grade manual muscle test scale.24 Higher scores indicate stronger muscles. The median muscle strength score for these muscles was calculated. Manual muscle test scores have been found to correlate (r=.768, P<.001) with the results of dynamometer testing.25
Balance confidence and fear of falling were assessed with the ABC Scale and the Norwegian version of the FES-I, respectively. Balance ability was tested with the Balance Evaluation Systems Test (BESTest) and the Four-Square Step Test (FSST). These scales and balance ability tests were chosen because they cover several of the balance problems mentioned in the group discussion, such as fear of falling, maintaining balance while performing different activities, balance reactions, and stepping over obstacles.
The ABC Scale (with scores expressed as a percentage) was used to quantify, from 0 to 100 (0=not at all), the confidence a person had in not losing balance while performing 16 activities in daily living.26 The scale has been used extensively in both balance and fall prevention research in the elderly population and has good psychometric properties.27 It was found to be valid and reliable in people who were healthy and frail elderly people, in people with a history of falls, and in patients with Parkinson disease.26,28,29
The Norwegian version of the FES-I was used to grade fear of falling during different activities in daily living on a 4-point scale (1=not at all concerned, 4=very concerned), giving a total scoring range of 16 to 64 points.30–32 It is widely used internationally in the field of fall prevention and has been found to be reliable and valid for community-dwelling elderly people and elderly patients who were treated for fall-related fractures.33,34
The BESTest was used to assess different subsystems related to balance.35 It contains 6 subscales covering these different subsystems: I—biomechanical constraints, II—stability limits/verticality, III—anticipatory postural adjustments, IV—postural responses, V—sensory orientation, and VI—stability in gait. The BESTest consists of 27 items, some of which are divided into 2 to 4 sub-items (eg, for left and right sides), resulting in a total of 36 tasks. All items are scored on a 4-category ordinal scale from 0 (not able or absent) to 3 (normal), and the scores are summed for each subscale. The sum of the 6 subscale scores is the BESTest total score. The BESTest subscale and total scores are reported as a percentage of the maximum score.
The FSST (in seconds) was used to test the participant's ability to step over a low obstacle in all 4 directions in a given sequence as fast as possible.36 The FSST was found to have good test-retest reliability, it correlated well with other gait measures,37 and it was feasible and valid for testing dynamic balance in patients who were ambulatory after stroke.38
Bias was controlled for by having the principal investigator (A.O.) administer the FES-I and FMS, assess GMFCS level, and lead all tests of balance ability. At least 2 people were always present during all balance tests. The principal investigator was blinded to the scores on the ABC Scale during the tests of balance ability. He was not blinded to whether the participants had reported deteriorated walking. All participants gave written informed consent.
We used SPSS version 15.0 statistical software (SPSS Inc, Chicago, Illinois) to analyze the data. Because of the small number of participants and the explorative nature of the study, nonparametric statistics were used throughout the study. Descriptive variables are presented with frequencies for history of falls and GMFCS levels and with medians and interquartile ranges (Q1–Q3) for muscle strength and MAS. All outcome variables (ABC Scale, FES-I, BESTest, and FSST) are presented with medians and interquartile ranges (Q1–Q3).39 For the case-control study, the between-group differences in ABC Scale, FES-I, BESTest (subscale and total), and FSST scores were tested with the Mann-Whitney U test.39 The relationship between balance confidence and the BESTest subscale scores, BESTest total scores, and FSST scores was tested with scatter plots and Spearman rho (rs) correlation coefficient. The 95% confidence interval (CI) for the correlation coefficient was calculated with VassarStats.40 The correlation coefficient was interpreted according to Domholt41: 1.00–0.90=very high; 0.89–0.70=high; 0.69–0.50=moderate; 0.49–0.26=low; and 0.25–0.00=little, if any correlation.
The descriptive variables of the participants are presented in Table 1. Eight participants reported deteriorated walking over the years (cases), and 8 participants reported either improved (n=2) or unchanged (n=6) walking (controls) (Tab. 1). The FMS showed that 15 participants were independent walkers on even surfaces over 50 m, and 12 participants were independent walkers over 500 m (data not shown). There were frequent falls among the participants, but no serious injuries were reported. There were no significant differences between the cases and the controls for the descriptive variables (Tab. 1). There were no differences in ABC Scale and FES-I scores between the participants who reported deteriorated walking and those who did not, nor were there any differences in BESTest subscale scores, BESTest total scores, or FSST scores (Tab. 2).
For the whole group, the scores on the ABC Scale showed the lowest confidence in balance during the activities “stepping on or off an escalator without holding the railings,” “walking on an icy surface,” and “stand on chair and reach” (Fig. 2). The overall ABC Scale score across all activities was 62% (Q1–Q3=37–70) confidence in not losing balance.
The FES-I showed the greatest fear of falling during the activities “walk on a slippery surface” and “walk up or down a slope,” with median scores of 3 (Q1–Q3=2–4) and 2 (Q1–Q3=2–4), respectively. The median FES-I summed score was 24 points (Q1–Q3=21–34). For 12 of the 16 items, the participants reported no or little fear of falling (Tab. 2).
The BESTest showed the lowest scores (percentage of the maximum score) for anticipatory postural adjustments (subscale III), with a median score of 42 (Q1–Q3=39–54), and for postural responses (subscale IV), with as median score of 31 (Q1–Q3=18–58). The median scores for the other subscales ranged from 60% to 79% of the maximum score (Tab. 2). The BESTest total score (percentage of the maximum score) showed a median score of 60 (Q1–Q3=55–68) (Tab. 2). The FSST showed a skewed distribution, with 2 distinct outliers. The median time to complete the stepping sequence was 16 seconds (Q1–Q3=11–22) (Tab. 2).
The relationship between balance confidence and balance ability showed a moderate positive correlation between the ABC Scale scores and the BESTest subscale IV, V, and VI scores and BESTest total scores (Tab. 3). The tests of balance ability were easily understood by the participants, there were no major practical problems during the testing, and all participants completed all tests.
There were no differences in balance confidence, fear of falling, and balance ability in adults with spastic bilateral CP, with and without self-reported walking deterioration. This finding might reflect a more complex interaction between walking deterioration and balance than originally hypothesized. The variability in the outcome variables in both groups was substantial. Different social roles and exposure to different contextual factors at work, in the family, and during leisure time, as well as personal factors, put different demands on both balance and walking. A small change in balance ability, therefore, might have a considerable impact on walking for some people with high functional demands and expectations, whereas for people with smaller demands these changes might go unnoticed. The groups had similar median levels and range of both muscle strength and muscle tone, and as 12 of the 16 participants were independent walkers over more than 500 m, the sample size was too small to detect a possible between-group difference.
If there are changes in balance ability over time in adults with spastic bilateral CP, they might be the result of slow processes involving both neural and musculoskeletal structures, such as slower proprioception, signal processing and execution, contractures, and reduced muscle strength. These processes might take a long time to develop before they result in reduced balance and later affect walking. Therefore, a prospective, longitudinal study of both walking and balance ability would have been the preferred design. However, as little is known about the optimal follow-up time, the most relevant outcome variables, or the balance in this group, we decided to use a cross-sectional study design in this study.
The BESTest showed reductions in all subsystems of balance, with anticipatory adjustments and postural responses being the most affected. Difficulties with anticipatory postural adjustments points toward reduced ability to anticipate the effect of movements on balance. When anticipating a posteriorly directed push, the anterior muscles must be activated and the body center of mass directed forward. These mechanisms are result of experience and motor learning,8,42 but are known to be slow and delayed in children with CP.8
The difficulties related to postural responses are in accordance with Gage et al,11 who described the equilibrium problem as one of the primary abnormalities of gait in children with CP. Reduced postural responses in children with CP compared with children developing typically also were found by Burtner et al.12 The ability to take a rapid step to regain balance is a crucial postural response to avoid falls and is probably a primary problem in adults with CP as well. Lajoie and Gallagher28 found significantly higher reaction times in “fallers” than in “nonfallers” in a group of elderly people.
In people with spastic CP, both anticipatory adjustments and postural responses can be affected by the brain lesion itself. Increased antagonistic coactivation,14 perhaps as a strategy to cope with reduced postural control,42 a top-down muscle activation,43 a reduced ability to modulate electromyography amplitude,44 and muscle weakness and joint impairments also are likely to be contributing factors. For overall balance, the BESTest total score was 60% of the maximum score, and the BESTest subscale scores ranged from 31% to 79% of the maximum score. These scores were considerably lower than those in the study by Lord et al,45 who found BESTest scores of 80% to 90% of the maximum in elderly people who were healthy.
The overall balance confidence was 62%, which is the same level as in elderly people with a history of hip fracture46 and lower than in community-dwelling elderly people.47 Lajoie and Gallagher28 found that a cutoff ABC Scale score of 67% could predict fallers in the elderly population with a sensitivity of 84% and nonfallers with a specificity of 88%. The high number of fallers in the present study suggests an even higher cutoff point in this group. There were substantial differences in ABC Scale scores for the different activities. The ABC Scale scores were lowest when using escalators and walking on icy surfaces. In 7 of the 16 activities, the median scores showed more than 80% confidence and, consequently, did not reflect the balance problems of the study group. Not surprisingly, fear of falling showed a trend similar to that of balance confidence. Delbaere et al33 found cutoff points for low (16–19), moderate (20–27) and high (28–64) concern in community-dwelling elderly people. Fear of falling in the present study was moderate, and slightly lower than in a group of people with spinal cord injury48 and in a group of women with osteoporosis.49 It was at about the same level as in the validation study by Delbaere et al33 and as in elderly people who were treated for fall-related fractures.34 There was a high prevalence of fallers in the present group, but no reports on fractures due to falls, which might partly correspond to “expert fallers” (ie, those who fall often, but without injuries and fear of falling). However, the risk of an injurious fall, reduced balance confidence, and fear of falling, especially in outdoor and community activities, might substantially restrict both social and community participation for these relatively young people. Fatigue also is a factor that might have an impact on both balance and falls, but this relationship was not assessed in this study.
The relationship between balance confidence and balance ability showed a moderate positive correlation for the BESTest subscales of postural responses, sensory orientation, and stability in gait and for the BESTest total score. The correlation between postural responses and balance confidence suggests that a reduced ability to react to external perturbations with a rapid step has a negative impact on balance confidence. The proximal-to-distal activation pattern found in children with CP43 indicates a reduced ability to use an ankle strategy for regaining balance. Thus, a reduced ankle strategy in addition to the reduced postural responses, creates a double problem, as both the most adequate balance adjustment strategy and the ability to take rapid steps to regain balance are reduced. The reduced postural responses may be confirmed by the low score of balance confidence when using escalators and walking on slippery surfaces. Sensory orientation, as evaluated in the test, can be the result not only of impaired sensory systems such as vision, the vestibular system, and proprioception, but also of an impaired motor output (ie, task execution). Both sensation and proprioception often are affected in CP.50 The group discussion highlighted balance problems during dim light conditions, when compensatory mechanisms (eg, vision for reduced proprioception) are no longer available. Stability in gait, as tested in the BESTest, includes walking under different conditions that challenge the balance. Impaired balance during walking may reduce the balance confidence and, therefore, could explain the correlation between stability in gait and balance confidence. A positive correlation between balance confidence and balance ability was found by Horak et al35 (r=.64) and by Sihvonen et al46 (r=.74), and confirmed by the present study.
The positive correlation between balance confidence and the scores on BESTest subscales IV, V, and VI and the BESTest total score suggests that the participants had a realistic perception of their own balance. The fact that there were no correlations between balance confidence and scores on BESTest subscales I, II, and III may be due to the mixture of different body functions and structures in these subscales, as well as the small number of participants. The FSST had little, if any, correlation with balance confidence, as also shown by Whitney et al37 in people with vestibular dysfunctions. The FSST includes tasks with different levels of difficulty, a variation that may have different effects on balance confidence and result in increased FSST time. For instance, the reduced ability to step backward may have a different effect on the balance confidence than a reduced ability to take a step forward or sideways.
These results confirmed the balance problems described in the group discussion and highlight the specific balance problems in this group. These problems were at the same level as in the elderly population treated for fall-related fractures, and thus they can be in need of targeted interventions. No major practical problems were encountered during the testing, indicating the practical feasibility of these tests in this population.
Future research should focus on long-term changes in balance in a prospective, longitudinal design among people who have spastic CP and on the relationship between changes in balance ability and walking. Furthermore, there is a need for studies on the reliability and validity of these balance tests in this group.
The reliability and validity of the outcome variables for use with this group were not previously tested. The number of participants in this study was relatively small; therefore, this study should be regarded as exploratory, and the results of the between-group comparisons should not be generalized to a larger group. However, the study group was not a convenience sample, but a specific subgroup of people recruited from our previous follow-up study.
Self-reported walking deterioration in this group of adults with spastic bilateral CP could not be explained by differences in balance confidence, fear of falling, or balance ability in this exploratory study. Across the whole group, most balance problems seemed related to reduced postural responses and anticipatory adjustments. Fear of falling was at the same level as in elderly people with fall-related fractures. Those individuals with higher balance confidence scored better on several of the tests of balance ability.
Mr Opheim, Dr Jahnsen, and Dr Olsson provided concept/idea/research design. All authors provided writing. Mr Opheim provided data collection. Mr Opheim and Dr Olsson provided data analysis. Mr Opheim and Dr Stanghelle provided project management. Dr Stranghelle provided fund procurement and facilities/equipment. Dr Olsson and Dr Stanghelle provided institutional liaisons. Dr Jahnsen, Dr Olsson, and Dr Stanghelle provided consultation (including review of manuscript before submission). The authors thank all of the participants and the personnel at the Motion Analysis Laboratory, Sunnaas Rehabilitation Hospital, for their respective contributions to the study.
This article was published as part of Mr Opheim's PhD thesis at Karolinska Institutet, Stockholm, Sweden.
Ethical approval for the study was obtained from the Regional Ethics Committee in Southeastern Norway (Ref. ID: 1.2006.952 and 2009/119) and the Commissioner for the Protection of Privacy in Research.
This research, in part, was presented at the 4th Journee Regionale du Reseau Breizh IMC; Rennes, France; February 1, 2011.
This study was financially supported by the Research Council of the South-Eastern Regional Health Authority in Norway.
- Received December 22, 2010.
- Accepted September 5, 2011.
- © 2012 American Physical Therapy Association