Proficiency of balance in children and youth who have had acute lymphoblastic leukemia.

Background and Purpose. As the survival rate for acute lymphoblastic leukemia (ALL) in childhood increases, long-term sequelae are a growing concern. This cross-sectional, descriptive study compared the balance skills of children and youth who have had ALL with those of comparable subjects and explored associations with demographics, therapy, physical activity, and health-related quality of life (HRQL). Subjects. Ninety-nine subjects treated previously for ALL and 89 comparison subjects were examined. Methods. Measures included the Bruininks-Oseretsky Test of Motor Proficiency (BOTMP) balance subtest, the Children's Self-perceptions of Adequacy in and Predilection for Physical Activity Scale (CSAPPA), and the Health Utilities Index (HUI), a measure of HRQL. Results. The children and youth who had ALL had poorer balance than the comparison subjects (BOTMP = 10.55 and 16.30, respectively) and lower CSAPPA scores (57.72 and 63.72, respectively) and HUI scores (0.86 and 0.97, respectively). Regression analyses identified exposure to cranial irradiation, being overweight, lower CSAPPA scores for adequacy, and lower HUI single-attribute scores for cognition as predictors of lower balance scores in subjects who had ALL. Discussion and Conclusion. Balance abilities in subjects treated for ALL were compromised, and several factors were associated with this deficit. [Wright MJ, Galea V, Barr RD. Proficiency of balance in children and youth who have had acute lymphoblastic leukemia. Phys Ther. 2005;85:782-790.]

Key Words: Balance, Oncology, Pediatrics.

Cancer is the most common cause of disease-related death in children in industrialized societies, and acute lymphoblastic leukemia (ALL) is the most common form of cancer in this age group. (1) The incidence of ALL is 3 to 4 cases per 100,000 children; incidence peaks among children 2 to 5 years of age. (1) Acute lymphoblastic leukemia is a malignant proliferation of white blood cells that begins in the bone marrow and spills over into the bloodstream, involving other organs in the process. The leukemic cells crowd out normal blood cells, resulting in anemia, susceptibility to infection, and bruising. Cure rates exceeding 80% have been achieved with the use of protocols using multiagent chemotherapy and "prophylactic" central nervous system (CNS) interventions. (2) Because of this success, a major emphasis is being placed on the costs and consequences of cure, which may have a lifelong effect on children with ALL. (3-6)

Balance, the process by which postural stability is maintained, is an integral component of functional motor abilities. (7-9) To maintain balance, the CNS processes visual, somatosensory, and vestibular input from the sensory system and sends output to the motor system. (7) In addition to neurological mechanisms, other factors--such as biomechanics, attention, motivation, obesity, environment, experience, and genetics--affect balance capabilities. (7-10) Children and youth who have had ALL can have problems in many of these areas (5,11-15) (Figure); however, the respective burdens of morbidity vary widely, even among patients receiving identical treatment. Thus, many impairments, restrictions in participation, and environmental factors can contribute to balance impairment and functional activity limitations.

[FIGURE OMITTED]

Corticosteroids and cranial irradiation, therapeutic components of ALL intervention, may result in cataracts, although these tend to be asymptomatic and nonprogressive. (4) Impaired conduction of sensory and motor pathways, due to vincristine-induced neuropathy or methotrexate-induced demyelination, may occur. (16,17) White matter damage and perfusion defects in the brain and neuropsychological disorders associated with prophylactic CNS interventions (intrathecal and intravenous methotrexate with or without cranial irradiation) have been identified. (5,18) In addition, these interventions have been found to cause problems that could affect the biomechanical aspects of balance, including decreased muscle "strength" (11) (the term used by the authors and in the assessment tool used for that study), impaired range of motion, (11,12) and obesity. (13) Learning difficulties, cognitive deficits, (4) attention problems, (14) and lack of participation in physical activities (15) also have been documented. Preschool children who receive CNS prophylaxis at the same time as chronological postnatal events of brain development, including glial cell proliferation and myelination, may be more sensitive to the neurotoxicity of radiation and chemotherapy; therefore, they are at greater risk for CNS damage than children with fully developed neurological systems. (5,14,16) Females are more susceptible than males to these treatment-related effects. (5,19) Thus, a number of mechanisms and factors could contribute to problems with balance in this population, given the complexity of this dynamic motor process.

A previous study of people who were long-term survivors of ALL demonstrated problems in gross motor functioning as measured by the Bruininks-Oseretsky Test of Motor Proficiency (BOTMP). (11) In that study, mean subtest scores were 10.2 for running speed and agility, 9.6 for balance, and 11.4 for strength. Standardized subtest means are 15 with a standard deviation of 5. Associations between gross motor proficiency and demographic and therapeutic factors were identified; however, the numbers were insufficient to show statistical significance. Balance was investigated further in a subsequent study (20) because impairment in balance can affect the quality and quantity of participation in activities of daily living, sports, and recreation. Balance measured by static sway showed people who had ALL to be comparable to those who had not had ALL, except when visual input and the base of support were decreased. (20) No component of dynamic balance (eg, walking along a line), however, was included in these tests.

The primary purposes of this study were to compare balance proficiency in children and youth who had ALL in childhood with a group of children and youth without disease and to explore the relationships between balance and demographic and therapeutic factors. Based on the available literature (4,5,16,17) and findings from our previous studies, (11,20) we hypothesized that children and youth who had ALL would have poorer balance than subjects in the comparison group and that their balance scores would be associated negatively with the following factors: a younger age at diagnosis, having received cranial irradiation, and being overweight.

A secondary purpose was to investigate whether balance was associated with self-perceptions of physical activity and health-related quality of life (HRQL) in subgroups of subjects. We hypothesized that higher balance scores would be associated with better self-perceptions of physical activity and HRQL, based on the rationale that balance contributes to functional capacity and participation in physical activity.

Method

Subjects

A cross-sectional study was conducted of children and youth with ALL who were at least 1 year after treatment. Subjects had completed Dana-Farber Cancer Institute (DFCI) protocols 85-01, 87-01, 91-01, or 95-01. (2) The DFCI protocols involve a month-long period of intensive multiagent chemotherapy to induce remission, after which the children receive a 2-year course of systemic multiagent chemotherapy and prophylaxis to reduce the risk that leukemic cells would continue to proliferate in the CNS. Treatment protocols include the use of asparaginase, corticosteroids (prednisone and dexamethasone), an anthracycline (doxorubicin), mercaptopurine, methotrexate, and vincristine. Central nervous system prophylaxis includes intrathecal chemotherapy (injection of drugs into the spinal fluid) and cranial irradiation in some children. (2) At diagnosis, the children are assigned a risk for relapse (standard or high) according to DFCI criteria based on age and severity of disease at diagnosis. (2) Children at high risk for relapse receive more intense chemotherapy, particularly with steroids and anthracyclines.

Subjects were enrolled over a 3-year period. Data were collected at a single point in time for each subject. Children and youth who had ALL and were at least 5 years of age during this study (old enough to participate in the testing) and off therapy for more than 1 year (to allow for a recovery period from acute comorbidities experienced during treatment) were considered potentially eligible for the study (n = 129). Patients with an additional diagnosis that could affect balance (eg, Down syndrome or head injury) (n = 9) and those followed at other facilities (n = 7) were considered ineligible, resulting in 113 eligible subjects.

A group of subjects without oncological disease, 5 to 31 years of age and matched for age and sex, was tested to provide comparative data. This sample consisted of siblings of the subjects in the ALL cohort, children and youth recruited through advertisement, and acquaintances. They were considered to be free of any known disorder, current or past, that could affect balance (ie, the same exclusion criteria as the ALL group) based on questioning of their parents or, for older subjects, of the subjects themselves. Written informed consent to participate in the study was obtained.

The group treated previously for ALL consisted of 99 subjects (88% of those eligible). Those subjects who were not examined included 3 who declined participation and 11 who could not be scheduled. The comparison group comprised 89 subjects (13% of whom were siblings). We wanted a comparison group that was similar to people with whom the subjects with ALL would be interacting in their everyday activities at home, at school, and in the community. We did not account for interfamily correlation. The comparative clinical characteristics of the subjects are presented in Table 1.

Procedure

All testing was done during a single session. One examiner, a pediatric physical therapist with 20 years of experience, who was aware of group designation, administered the balance subtest of the BOTMP. (21) Few measures of postural stability in children are available that have acceptable documentation of reliability and are sufficiently difficult to challenge a group such as the one we studied. (7) The BOTMP was designed for children with mild motor impairment, which, therefore, limited the problem of a ceiling effect. It is a preferred test used by therapists to describe balance in children with mild motor problems .(10)

Moderate test-retest reliability for the balance subtest was determined during development of the tool (reliability coefficient with Fisher Z transformation = .56). (21) Intrarater reliability of .97 has been determined by intraclass correlation coefficients (3,1) over 2 trials. (8) Of greater importance to our study, the balance subtest has been shown to discriminate among children and youth treated for ALL; among children with and without minor motor problems, clumsiness, learning disabilities, or cognitive delays; and among those without these diagnoses. (11,21,22) The BOTMP has construct validity relevant to the balance subtest. (21) The test includes both static (one-leg stance skills) and dynamic (walking along a line and on a balance beam) items. Scores have been standardized for children without disabilities and performance correlated with age for 4 1/2 to 14 1/2 years. Standardized subtest means are 15 with a standard deviation of 5. Norms for the oldest age group were used for subjects older than 14 1/2 years, because there were no between-group differences for balance subtest scores in study subjects less than or older than 14 1/2 years or in 2 groups of youth with mean ages of 14.51 and 17.74 years in a study of adolescents. (23) Although test-retest reliability is only moderate for the BOTMP balance subtest, its discriminative abilities, its age standardization for developmental variation, a degree of difficulty sufficient to avoid a ceiling effect, and the exclusion of activities not related to postural stability made it the tool of choice.

Height and weight were measured using calibrated equipment. Body mass index (BMI) was calculated from these 2 measures (weight/height (2)). Body mass index for age and sex has been validated for predicting overweight status. (24) Subjects were considered to be at risk for overweight status if their BMI for age and sex was greater than the 85th percentile and considered to be overweight if their BMI for age and sex was greater than the 95th percentile. These designations were based on BMI data standardized for age and sex by the National Center for Chronic Disease Prevention and Health Promotion. (25) Information on vincristine-induced neurotoxic effects, cataracts, and CNS damage was collected retrospectively by chart review of problem lists and summary notes. Because of the increasing interest in the effects of impairment on the quality of life, physical self-worth, and participation in life activities of people with cancer in childhood, 2 further measures were included in the study for appropriate subgroups.

Subgroup 1. Participants attending elementary or secondary school who were 8 years of age or older completed the Children's Self-perceptions of Adequacy in and Predilection for Physical Activity Scale (CSAPPA). (26) Sixty-seven children and youth who had ALL and 64 children and youth in the comparison group were eligible to complete the CSAPPA questionnaire. This tool was designed to measure children's and youths' self-perceptions of their ability for physical activity and the likelihood that they would choose a physical activity rather than a sedentary activity. The tool is reported to yield data with good test-retest reliability (test-retest correlation coefficients ranging from .70 to .91 [statistic type not reported] for all subtests and total scores) and item partial-total correlations ranging from .65 to .85 for appropriate factors with school-aged children and youth 8 years and older. (26) Positive relationships between the CSAPPA and motor abilities, as measured by the BOTMP, have been described. (27) The CSAPPA has been shown to differentiate between patients who have had ALL and children and youth without disease. (15) Associations between self-perceptions of physical activity and actual physical activity and fitness in children and adolescents have been described. (28)

Subgroup 2. If present at the time of the examination, the subject's parent completed the Health Utilities Index (HUI) questionnaire HUI23.15Q. Parents were available to complete the HUI questionnaires for 77 (78%) of the ALL group and 71 (81%) of the comparison group, respectively. The HUI is a family of multi-attribute, generic, preference-based measures of health status and HRQL. These measures have been used in several studies of children and adults with cancer. (3,6,29,30) Overall HRQL and single-attribute utility scores for vision, hearing, speech, ambulation, dexterity, emotion, cognition, and pain were determined from the Mark 3 (HUI3) system. The HUI instruments are not burdensome and have been shown to distinguish between groups known to have clinically important differences in health status. (6,29) The one-way single-measure intraclass correlation coefficient for test-retest reliability is .767 for overall health status. (30)

Data Analysis

Between-group statistical analyses included 2-sample t tests of means for interval-scale variables and chi-square tests for ordinal-scale variables. Factors that might be associated with balance included the demographic variables of age at diagnosis, age at time of testing, time off treatment, and sex and the therapeutic variables of cranial irradiation (dichotomous), risk for relapse (dichotomous), and BMI (3-level categorical). First, univariate linear regression analyses for each of these variables were examined independently on the dependent variable of the BOTMP balance scores, which were normally distributed. Those who met an initial statistical level of P<.25 were retained for multiple linear regression using stepwise entry (criteria: probability of F to enter [less than or equal to].050, probability of F to remove [greater than or equal to].100). Rasch-Wright latent trait modeling was used to develop the point score scales. (21)

Associations between the balance scores and the CSAPPA and HUI variables were analyzed in a similar manner, but separately, because the use of different subgroups were based on age restrictions (CSAPPA) and parental reports (HUI). Analyses were performed using the Statistical Package for the Social Sciences (SPSS), version 10.0.* A level of P [less than or equal to].05 was considered statistically significant.

Results

Balance subtest scores on the BOTMP were lower (t = -6.893, P<.001) for the ALL group ([bar.X] = 10.55, SD = 5.89, range = 1-28) than for the comparison group ([bar.X] = 16.30, SD = 5.54, range = l-29). Scores on the BOTMP were similar to the subgroup scores on the CSAPPA (ALL group: [bar.X] = 10.61, SD = 5.75; comparison group: [bar.X] = 16.46, SD = 5.53) and HUI (ALL group: [bar.X] = 10.97, SD = 5.97; comparison group: [bar.X] = 16.48, SD = 5.73). On each subscale of the CSAPPA (except freedom from injury), the ALL subgroup had lower scores than the comparison subgroup. The scores (Tab. 2) for the comparison group were very similar to reported normative data (mean total CSAPPA score = 63.7 versus 65.2). (28) The overall HRQL and single-attribute utility scores for emotion, cognition, and pain were lower in the ALL group than in the comparison group (Tab. 3).

The results of the linear regression equations are presented in Table 4. Models were confirmed with equations using backward regression methods. The variables of age, risk for relapse, cranial irradiation, time off treatment, and overweight were entered into a stepwise regression analysis based on their level of significance using univariate linear regression. This calculation indicated that cranial irradiation, overweight, and longer time off treatment were predictors of poorer balance, but accounted for only 18.7% of the variability in the BOTMP scores. The adjusted [R.sup.2] was 10.6 for cranial irradiation, 15.5 for cranial irradiation and overweight, and 18.7 for cranial irradiation, overweight, and time off therapy. The addition of the other variables did not improve the fit of the equation.

The cranial irradiation groups did not differ in being overweight ([chi square] = 1.224, P = not significant). The groupings did vary according to BOTMP scores (nonirradiated group = 14.32 versus irradiated group = 9.47; t = 3.607, P<.001) and BMI (not overweight = 11.57, at risk for overweight = 8.88, overweight = 8.33; F = 3.196, P = .045; post hoc difference between not overweight and overweight, P = .033).

A separate regression analysis was performed on the subgroup of children and youth who had ALL and who were appropriate subjects for completing the CSAPPA (8 years of age or older and attending school, n = 67). The CSAPPA total scores of the ALL group were not statistically significant in the univariate analysis for balance (unstandardized coefficient = 0.094, SE = 0.053, P = .080). Adequacy was a predictor of balance (Tab. 4). Enjoyment of physical education was a significant variable in the univariate analysis, but was not when combined with the adequacy scores.

Balance scores associated positively with the overall HUI3 scores (unstandardized coefficient = 12.164, SE = 3.339, P<.001) in the subgroup of subjects who had parental proxy HUI results (n = 77). Only the single-attribute score for cognition was a predictor of balance within the ALL group.

The unadjusted regression coefficients of variables not included in the final regression models are listed in Table 5. Risk for relapse approached statistical significance; however, it was excluded in the stepwise regression equation calculations. Risk for relapse is associated highly with cranial irradiation (r = .508, P<.001). There was no difference in balance proficiency between patients at standard risk for relapse and patients at high risk for relapse when only those who received cranial irradiation were considered (standard risk = 9.93 versus high risk = 9.36; t = 0.198, P = not significant), but there was a difference in patients at standard risk for relapse that depended on whether they received cranial irradiation (irradiated group = 9.63 versus nonirradiated group = 14.48; t = 3.402, P = .001).

Four subject files had documentation of vincristine-induced neuropathy while on treatment, which had resolved clinically by the time of testing. There were 3 reported cases of cataracts; however, only 1 subject had a visual deficit. One child had white matter damage, and 1 child had had a stroke with transient hemiparesis as a complication of treatment. The results of the chart review regarding these factors could not be considered valid enough to be included in any analyses in the current study. Two subjects had passive ankle dorsiflexion less than neutral at the time of testing. Their BOTMP balance scores were 11 and 13.

Discussion and Conclusions

Balance skills result from the dynamic interaction of many subsystems in a task-specific context. The children and youth treated previously for ALL had variable scores, but their mean balance proficiency was substantially poorer than the group of young people without disease, confirming our primary hypothesis. The etiology of balance problems in the ALL group is most likely multifactorial and variable, reflecting the various systems that contribute to balance and possibly individual differences in the ability to adapt to balance demands through practice and compensations. As we hypothesized, cranial irradiation and being overweight were significant predictors of impaired balance proficiency, but accounted for only a small proportion of variance. The probable explanations for the within-group differences are as diverse as the severity and individual interactions of the long-term complications of ALL with numerous factors that contribute to balance proficiency.

The association between cranial irradiation and poorer balance may reflect the CNS damage that can result from this therapeutic intervention and that has been demonstrated with outcomes other than balance. (4,5,14,16,18) These include white matter changes, cortical atrophy, calcification, and slowing of cortical processing. All subjects in the ALL group received both intrathecal and intravenous methotrexate, interventions that are known to contribute to CNS toxicity. However, those who also received cranial irradiation had poorer balance scores than those who did not. Of the 20 children and youth who had balance scores of 5 or less (2 or more standard deviations below the standardized norms), 19 had received cranial irradiation. Chemotherapeutic and radiation components of CNS "prophylaxis" interact synergistically, appearing to result in impaired balance.

Contrary to our hypothesis, balance was not associated with age at diagnosis in our study. The trend of poorer balance being associated with high risk for relapse appears to be the result of the greater proportion of children and youth at high risk for relapse who received cranial irradiation. Obesity was more prevalent in the subjects treated for ALL and was a significant predictor of poor balance proficiency in this group. The causal mechanism of obesity remains unclear, but is probably multifactorial and includes previous cranial irradiation. (13) Obesity has been shown to be associated with poorer static and dynamic stability in adolescents because of poorer control of sway in the mediolateral directions caused by excess weight rather than underlying postural instability. (9) Excess weight also may affect the child's confidence and willingness to participate in strenuous physical activities that could challenge and improve postural control and expend calories.

Longer time off treatment was a predictor of poorer balance proficiency. Differing associations with time off treatment have been reported in follow-up of cognitive functioning. (5) However, reports are often cross-sectional and not longitudinal. A longitudinal follow-up study of motor nervous system impairment by Lehtinen et al (17) showed no decline in gross motor functioning over time after treatment.

The group treated for ALL had poorer CSAPPA scores relative to the comparison group in all except one category. Associations between the BOTMP and CSAPPA scores suggest that balance is an important component of proficiency in sports, physical education, and active play and may affect the enjoyment of physical education, but does not necessarily contribute to the between-group differences in preference for sedentary pastimes. Regardless of the lack of other associations, the youth who had ALL are less likely to participate in physical activities. They have fewer opportunities to reap the potential physical and psychological benefits of this activity and are at risk of becoming increasingly inactive and more resistant to change in their activity habits with increasing age. (15,26)

The HUI findings of poorer overall HRQL and lower scores for the attributes of emotion, cognition, and pain are consistent with previous studies of people with ALL in childhood. (6) The lack of association between balance and the gross motor attribute of ambulation, a skill that could potentially be affected by balance problems, is most likely due to a ceiling effect, because there is no discrimination among motor skills beyond basic ambulation in the HUI. The positive correlation between the single-attribute utility score for cognition and balance most likely reflects the numerous neurotoxic effects of "prophylactic" CNS interventions.

The effects of altered somatosensory input and motor output could not be differentiated clearly in our study. The most probable direct cause is vincristine-induced peripheral sensorimotor neuropathies, which manifest as areflexia, sensory impairment, pain and muscle weak ness, and atrophy most notable in the distal extremities. Studies have shown prolonged latencies and decreased amplitudes of sensory and motor-evoked potentials in the peripheral nerves, indicating demyelination and loss of descending fibers, or loss of muscle fibers, which are partially but not totally reversible. (18) Reinders-Messelink et al (31) identified problems with balance in children receiving treatment for ALL, at diagnosis and during their therapy. Although a relationship between the gross motor problems and vincristine neurotoxicity seemed plausible, that explanation has not been supported statistically. A previous study investigated balance in children who received vincristine for the treatment of Wilms tumor or ALL. (32) The children treated for Wilms tumor did not have balance problems as measured by the BOTMP. However, the children treated for ALL, who received multiagent chemotherapy and cranial irradiation in addition to vincristine, did have problems.

The biomechanical system provides the background on which postural adjustments are made. In addition to obesity, 2 biomechanical factors related to postural stability in children are force output and range of motion in the lower extremities. (7) Long-term survivors of ALL have been shown to have limitations in these areas, (11,12) but the amount of force output or range of motion that is required to perform balance activities successfully remains unknown. (7) Although not tested formally, the only subjects in our study with obvious muscle weakness that could affect the balance skills items tested were the 2 subjects with long-term contractures that prevented attainment of a neutral ankle position and limited force output. Surprisingly, their balance scores were less than one standard deviation below the mean, suggesting that poor biomechanics did not affect their stability, perhaps due to the development of compensatory strategies.

Three children had cataracts. Of these subjects, the only child who was felt to have reduced visual acuity had a balance score of 1. However, this subject had white matter brain damage that could have contributed to poor balance abilities.

A limitation in this study was that the examiner was unblinded to group designation introducing the possibility of measurement bias. Adherence to BOTMP test guidelines limited subjectiveness. Only 3 patients declined involvement in the study, reducing the likelihood of participation bias. The findings are specific to children treated on DFCI protocols. However, the DFCI protocols and their complications are similar to those of other common treatment strategies. This makes the current findings generalizable to the management of ALL in children and youth.

Our results confirmed that children and youth who have had ALL have poorer balance than comparison subjects. Although many factors contribute to balance, it was possible to demonstrate that cranial irradiation and being overweight were negative predictors of balance proficiency. Balance abilities were related to perceptions of adequacy in physical activity and overall HRQL scores. According to the systems theory of motor control, therapists should identify the system or systems that are malfunctioning and provide treatment focused on those systems. (7) Clinicians, including physical therapists, should address the various late effects of cancer treatment, which may have a lifelong effect on the child. Recognition of balance problems is important to enable appropriate consultation to promote optimal balance skills, or, if necessary, compensate for balance limitations. We encourage further research to support the importance of exercise and physical activity during treatment for cancer, which is becoming recognized in practice and in the literature. (33)

All authors provided concept/idea/research design, writing, fund procurement, and consultation (including review of manuscript before submission). Ms Wright and Dr Galea provided data collection and analysis. Ms Wright and Dr Barr provided project management and subjects.

Dr Galea provided facilities/equipment.

This study was approved by the Research Ethics Board of Hamilton Health Sciences and the Faculty of Health Sciences, McMaster University.

Funding to support this study was received from the Research Development Fund of Hamilton Health Sciences and National Institutes of Health grant CA68484.

This article was received September 28, 2004, and was accepted February 2, 2005.

* SPSS Inc, 444 N Michigan Ave, Chicago, IL 60611.

References

(1) Margolin JF, Steuber CP, Poplack DG. Acute lymphoblastic leukemia. In: Pizzo PA, Poplack DG, eds. Principles and Practice of Pediatric Oncology. 4th ed. Philadelphia, Pa: Lippincott Williams & Wilkins; 2002:489-544.

(2) Silverman LB, Gelber RD, Dalton VK et al. Improved outcome for children with acute lymphoblastic leukemia: results of Dana-Farber Consortium Protocol 91-01. Blood. 2001;97:1211-1218.

(3) Barr RD, Feeny D, Furlong W. Economic evaluation of treatments for cancer in childhood. Eur J Cancer. 2004;40:1335-1345.

(4) Waber DP, Tarbell NJ. Toxicity of CNS prophylaxis for childhood leukemia. Oncology (Huntingt). 1997;11:259-264.

(5) Moore BD. Neurocognitive outcomes in survivors of childhood cancer. J Pediatr Psychol. 2005;30:51-63.

(6) Barr RD, Furlong W, Dawson S, et al. An assessment of global health status in survivors of acute lymphoblastic leukemia in childhood. Am J Pediatr Hematol Oncol. 1993;15:284-290.

(7) Westcott SL, Lowes LP, Richardson PK. Evaluation of postural stability in children: current theories and assessment tools. Phys Ther. 1997;77:629-645.

(8) Habib Z, Westcott S. Assessment of anthropometric factors on balance tests in children. Pediatr Phys Ther. 1998;10:101-109.

(9) McGraw B, McClenaghan BA, Williams HG, et al. Gait and postural stability in obese and nonobese prepubertal boys. Arch Phys Med Rehabil, 2000;81:484-489.

(10) Westcott SL, Murray KH, Pence K. Survey of the preferences of pediatric physical therapists for assessment and treatment of balance dysfunction in children. Pediatr Phys Ther. 1998;10:48-61.

(11) Wright MJ, Halton JM, Martin RF, Barr RD. Long-term gross motor performance following treatment for acute lymphoblastic leukemia. Med Pediatr Oncol. 1998;31:86-90.

(12) Wright MJ, Hanna SE, Halton JM, Barr RD. Maintenance of ankle range of motion in children treated for acute lymphoblastic leukaemia. Pediatr Phys Ther. 2003;15:146-152.

(13) Warner JT, Evans WD, Webb DK, Gregory JW. Body composition of long-term survivors of acute lymphoblastic leukaemia. Med Pediatr Oncol. 2002;38:165-172.

(14) Paakko E, Harila-Saari A, Vanionpaa L, et al. White matter changes on MRI during treatment in children with acute lymphoblastic leukemia: correlation with neuropsychological findings. Med Pediatr Oncol. 2000;35:456-461.

(15) Wright MJ, Galea V, Barr RD. Self-perceptions of physical activity in survivors of acute lymphoblastic leukemia. Pediatr Ex Sci. 2003;15: 191-201.

(16) Harila-Saari AH, Vainionpaa LK, Kovala TT, et al. Nerve lesions after therapy for childhood acute lymphoblastic leukemia. Cancer. 1998;82:200-207.

(17) Lehtinen SS, Huuskonen UE, Harila-Saari AH, et al. Motor nervous system impairment persists in long-term survivors of childhood acute lymphoblastic leukemia. Cancer. 2002;94:2466-2473.

(18) Harila-Saari AH, Paakko EL, Vainionpaa LK, et al. A longitudinal magnetic resonance imaging study of the brain in survivors of childhood acute lymphoblastic leukemia. Cancer. 1998;83:2608-2617.

(19) Bleyer A. CNS chemoradiotherapy of childhood leukemia: the plot thickens but the ending bodes well. J Clin Oncol. 1995;13:2480-2482.

(20) Galea V, Wright MJ, Barr RD. Measurement of balance in survivors of acute lymphoblastic leukemia in childhood. Gait Posture. 2004;19: 1-10.

(21) Bruininks RH. Bruininks-Oseretsky Test of Motor Proficiency: Examiner's Manual. Circle Pines, Minn: American Guidance Service; 1978.

(22) Wilson BN, Polatajko HJ, Kaplan BJ, Faris P. Use of the Bruininks-Oseretsky Test of Motor Proficiency in occupational therapy. Am J Occup Ther. 1995;49:8-17.

(23) Davis PL, Rose JD. Motor skills of typically developing adolescents: awkwardness or improvement? Phys Ther Occup Ther Paediatr 2000; 20(1):19-42.

(24) Mei Z, Grummer-Strawn LM, Pietrobelli A, et al. Validity of body mass index compared with other body-composition screening indexes for the assessment of body fatness in children and adolescents. Am J Clin Nutr. 2002;75:978-985.

(25) Centers for Disease Control, National Center for Health Statistics, National Center for Chronic Disease Prevention and Health Promotion. 2000 CDC Growth charts page. Available at: http://www.cdc.gov/growthcharts. Accessed April 21, 2005.

(26) Hay JA. Adequacy in and predilection for physical activity in children. Clin J Sport Med. 1992;2:192-202.

(27) Hay JA, Missiuna C. Motor proficiency in children reporting low levels of participation in physical activity. Can J Occup Ther. 1998;65:64-71.

(28) Raudsepp L, Liblik R, Hannus A. Children's and adolescents' physical self-perceptions as related to moderate to vigorous physical activity and physical fitness. Pediatr Ex Sci. 2002;14:97-106.

(29) Furlong WJ, Feeny DH, Torrance GW, Barr RD. The Health Utilities Index (HUI) system for assessing health-related quality of life in clinical studies. Ann Med. 2001;33:375-384.

(30) Feeney DH, Torrance GW, Furong WJ. Health Utilities Index. In: Spilker B, ed. Quality of Life and Pharmacoeconomics in Clinical Trials. 2nd ed. Philadelphia, Pa: Lippincott-Raven; 1996:239-252.

(31) Reinders-Messelink H, Schoemaker M, Snijders T, et al. Motor performance of children during treatment for acute lymphoblastic leukemia. Med Pediatr Oncol. 1999;33:545-550.

(32) Wright MJ, Halton JM, Barr RD. Musculoskeletal and gross motor function in survivors of acute lymphoblastic leukemia and Wilm's tumor. J Pediatr Hematol Oncol. 1996;14:461.

(33) Galvao DA, Newton RU. Review of exercise intervention studies in cancer patients. J Clin Oncol. 2005;23:899-909.

MJ Wright, PT, BSc, MEd, is Clinical Specialist, McMaster Children's Hospital, and Assistant Clinical Professor, School of Rehabilitation Sciences and Department of Pediatrics, McMaster University, Hamilton, Ontario, Canada. Address all correspondence to Ms Wright at McMaster Children's Hospital, Box 2000, Hamilton, Ontario, Canada, L8N 3Z5 ([e-mail]).

V Galea, PhD, is Associate Professor, School of Rehabilitation Sciences, McMaster University.

RD Barr, MB, ChB, MD, is Professor, Department of Pediatrics, Hamilton Health Sciences, Hamilton, Ontario, and School of Rehabilitation Sciences, McMaster University.

Table 1.
Characteristics of Subjects Treated Previously for Acute Lymphoblastic
Leukemia (ALL) (n = 99) and Comparison Subjects Without Disease
(n = 89) (a)

                                         ALL Group

                                 [bar.X]
                                 or Ratio     SD        Range

Age (y)                             12.1      4.9     5.1-25.2
Age at diagnosis (y)                 5.2      2.6     0.3-17.0
Years off treatment                  4.7      3.5     1.0-13.6
Height (cm)                        141.9     17.7   102.0-182.0
Weight (cm)                         43.2     18.8    18.0-103.0
Females/males                      44/55
No. of standard-risk/high-risk     51/48
  patients
Overweight/at risk for           18.2/16.2
  overweight (%)

                                     Comparison Group

                                 [bar.X]
                                 or Ratio      SD      Range

Age (y)                            12.2      5.2     5.1-31.5
Age at diagnosis (y)                NA
Years off treatment                 NA
Height (cm)                       148.5     18.6   107.0-188.0
Weight (cm)                        45.3     17.4    17.9-95.9
Females/males                     45/44
No. of standard-risk/high-risk      NA
  patients
Overweight/at risk for           6.7/11.2
  overweight (%)

                                    t        p

Age (y)                          -0.241   NS (b)
Age at diagnosis (y)
Years off treatment
Height (cm)                      -2.253   .025 (b)
Weight (cm)                      -0.795   NS (b)
Females/males                             NS (c)
No. of standard-risk/high-risk
  patients
Overweight/at risk for                    .026 (c)
  overweight (%)

(a) NA = not applicable; NS = not significant.

(b) t tests.

(c) Chi-square tests.

Table 2.
Comparison of Children's Self-perceptions of Adequacy in and
Predilection for Physical Activity Scale (CSAPPA) Scores Between
School-Aged Children Who Had Acute Lymphoblastic Leukemia (ALL)
(n = 67) and Children Without Disease (n = 64)

                                                  ALL Group
                                  Possible
CSAPPA Item                         Range    [bar.X]    SD     Range

Total CSAPPA                        19-76    57.72     12.30   25-76
Predilection                         9-36    27.37      6.11   15-36
Adequacy                             7-28    19.64      5.23    7-28
Enjoyment of physical education      3-12     9.84      2.60    3-12
Freedom from injury                   1-4     2.54      0.94     1-4

                                   Comparison Group

CSAPPA Item                       [bar.X]    SD    Range

Total CSAPPA                      63.72     8.46   39-76
Predilection                      29.95     4.56   19-36
Adequacy                          22.97     3.45   13-28
Enjoyment of physical education   10.80     1.94    3-12
Freedom from injury                2.89     0.82     1-4

                                     t or
CSAPPA Item                       [chi square]     P

Total CSAPPA                         -3.178      .002 (a)
Predilection                         -2.729      .007 (a)
Adequacy                             -4.273      <.00
Enjoyment of physical education      -2.392      .018 (a)
Freedom from injury                   5.593      NS (b,c)

(a) t test.

(b) Chi-square test.

NS = not significant.

Table 3.
Comparison of Health Utility Index (HUI) Scores in Subgroups of
Subjects Who Had Parental Proxy Assessments Between Children Who Had
Acute Lymphoblastic Leukemia (ALL) (n = 77 and Children Without Disease
(n = 71)

                  ALL Group     Comparison Group

HUI Item          [bar.X]     SD   [bar.X]    SD       t     P (a)

Overall HUI3      0.86      0.18   0.97      0.08   -4.747   <.001
Vision HUI3       0.99      0.07   0.99      0.02   -0.801   NS
Hearing HUI3      0.99      0.08   1.00      0.00   -0.960   NS
Speech HUI3       0.97      0.07   0.98      0.07   -0.446   NS
Ambulation HUI3   1.00      0.02   1.00      0.00   -0.960   NS
Dexterity HUI3    0.99      0.03   1.00      0.00   -1.959   NS
Emotion HUI3      0.97      0.06   0.99      0.02   -2.956    .004
Cognition HUI3    0.90      0.17   0.98      0.17   -3.644   <.001
Pain HUI3         0.97      0.07   0.99      0.01   -4.119   <.001

(a) t tests, NS = not significant.

Table 4.
Multiple Linear Regression for Subjects Treated Previously for Acute
Lymphoblastic Leukemia (ALL) Using the Bruininks-Oseretsky Test of
Motor Proficiency (BOTMP) Balance Subtest as the Dependent Variable
(3 Separate Analyses) (a)

                                Unstandardized       Standardized
Variable                       Coefficient (SE)   Coefficient [beta]

Therapeutic/demographic variables (total ALL group)

  Radiation                     -4.606 (1.323)          -.321
  Overweight                    -1.801 (0.686)          -.240
  Time off treatment            -0.314 (0.158)          -.200

CSAPPA (subgroup of ALL school-age subjects 8 y and older)

  Adequacy                       0.314 (0.124)           .301

HUI3 (subgroup of ALL subjects-parental proxy)

  Cognition                     10.484 (3.407)           .302

                               Adjusted
Variable                       [R.sup.2]     t       P

Therapeutic/demographic variables (total ALL group)

  Radiation                      .187      -3.480   0.00
  Overweight                               -2.624   0.01
  Time off treatment                       -2.165   0.03

CSAPPA (subgroup of ALL school-age subjects 8 y and older)

  Adequacy                       .078       2.542   0.01

HUI3 (subgroup of ALL subjects-parental proxy)

  Cognition                      .091       3.077   0.00

(a) CSAPPA = Children's Self-perceptions of Adequacy in and
Predilection for Physical Activity Scale, HUI = Health Utilities Index.

Table 5.
Unadjusted Regression Coefficients of Variables Relating to
Bruininks-Oseretsky Test of Motor Proficiency (BOPMP) Balance Scores
Not Included in Final Models (for SuBjects Treated Previously for Acute
Lymphoblastic Leukemia [ALL]) (a)

                                       Unstandardized
Variable                              Coefficient (SE)   P

Demographic and treament/medical
  Sex                                  -0.982 (1.194)    .413
  Age                                  -0.210 (0.121)    .085
  Age at diagnosis                      0.008 (0.160)    .959
  Risk for relapse                     -2.207 (1.181)    .065
CSAPPA (school-age subgroup only)
  Predilection                          0.104 (0.112)    .359
  Enjoyment of physical education       0.534 (0.253)    .038
  Freedom from injury                   1.180 (0.705)    .099
HUI3 (parental proxy subgroup only)
  Vision                               24.940 (34.053)   .466
  Hearing                               6.786 (8.708)    .438
  Speech                                5.679 (8.977)    .529
  Ambulation                           21.201 (34.876)   .545
  Dexterity                            33.177 (24.916)   .186
  Emotion                               8.104 (9.705)    .406
  Pain                                 17.726 (9.728)    .072

(a) CSAPPA = Children's Self-perceptions of Adequacy in and
Predilection for Physical Activity Scale, HUI = Health Utilities Index.