Associating Physical Activity Levels with Motor Performance and Physical Function in Childhood Survivors of Acute Lymphoblastic Leukemia

Abstract

Purpose: This cross-sectional, observational study investigated whether physical activity (PA) levels are associated with motor performance and physical function in children after treatment for acute lymphoblastic leukemia (ALL). Method: Participants aged 8–13 years who had completed treatment for ALL (3–36 months post-treatment) were tested at their oncology long-term follow-up appointment at the British Columbia Children's Hospital. PA level was measured using the Physical Activity Questionnaire for Older Children (PAQ-C). Motor performance was measured using the Bruininks-Oseretsky Test of Motor Proficiency, Second Edition, Short Form (BOT-2 SF), and physical function was measured using the 6-minute walk test (6MWT). Results: Thirteen children completed testing. PAQ-C scores were not associated with BOT-2 SF or 6MWT performance. Eleven children (85%) performed below the norm for the 6MWT. Children with elevated body mass index had poorer 6MWT but similar PAQ-C scores. Conclusion: PA was not found to be associated with motor performance and physical function. Participants who were overweight or obese had poorer 6MWT performance, which may indicate the need for closer monitoring of post-treatment weight status and physical function in the oncology follow-up setting.

Acute lymphoblastic leukemia (ALL) is the most common childhood malignancy in Canada, accounting for 25% of all newly diagnosed childhood cancers between 2006 and 2010.¹ With recent treatment advances, children diagnosed with ALL currently have a 5-year survival rate of 91% in Canada¹ and higher than 85% in other developed countries.² As a result, research is now focused on addressing treatment-related side effects.

Physical activity (PA) levels in children who are undergoing or who have completed treatment for ALL are lower than those of their healthy counterparts.³ In addition, they typically have post-treatment physical impairments, particularly reduced motor performance⁴ and reduced physical function, as measured with the 6-minute walk test (6MWT).^5–8 PA is essential for children to facilitate normal growth and development and particularly to promote general health and disease prevention,⁹ as well as motor¹⁰ and psychosocial¹¹ development. However, a cancer diagnosis disrupts opportunities for normal development and PA participation because of hospitalization and treatment side effects, both of which contribute to physical impairments and reduced motivation.¹² Specific to ALL survivors, evidence has suggested that vincristine^13,14 and methotrexate¹⁵ treatments directly relate to motor performance impairments during and after treatment. Anthracycline treatment has well-documented, potentially long-term cardiotoxic side effects, and it potentially limits physical function.^16,17

Few studies have investigated whether motor performance or physical function impairments are related to PA levels in children who have completed treatment for ALL,^5,6,18,19 and to date, these studies have included childhood cancer survivors with a wide range in age and time since completing treatment. Given how important it is that children participate in PA for their normal growth and development, and recognizing that a suitable time to reintroduce habitual PA is once treatment for ALL is completed,²⁰ a better understanding of the factors associated with higher or lower PA levels in children shortly after completing treatment for ALL is needed.

To our knowledge, no study has focused specifically on children who were in early childhood at the time of ALL diagnosis and treatment and also examined whether both motor performance and physical function are associated with PA levels in those children who recently completed treatment for ALL. The purpose of this study was to investigate whether motor performance and physical function are associated with PA levels in children aged 8–13 years who completed treatment for ALL in the preceding 36 months. We hypothesized that children with higher PA levels would have higher motor performance and physical function than children with lower PA levels.

Methods

This cross-sectional, observational study was approved by the University of British Columbia Children's and Women's Research Ethics Board. Participants were recruited from September 2013 to April 2014 at the British Columbia Children's Hospital (BCCH) oncology long-term follow-up clinic.

Participants

A convenience sample was taken on the basis of the number of children with anticipated appointments at the follow-up clinic during the recruitment period. The inclusion criteria were children who were aged 8–13 years at the time of study assessment; treated for ALL at BCCH; 3–36 months after treatment for ALL; attending the follow-up clinic; and able to understand both verbal and written English (including parents or guardians). Children were excluded if they had any symptoms or impairments unrelated to the cancer diagnosis and treatment that could influence performance in physical testing sessions, such as Down syndrome or limb abnormalities; had received cranial radiation; had had a relapse; or were participating in other research studies.

Procedure

Eligible patients returning to the follow-up clinic for appointments were identified by the hospital clinical research associates. Clinic nurses and physicians pre-screened and approved eligible patients to be contacted. Parents of approved patients were invited to participate in the study by mail approximately 1 month before their next scheduled follow-up appointment. Follow-up phone calls regarding participation were made 10–14 days later. Study appointments were scheduled as part of the next follow-up appointment.

Informed consent and assent forms were sent by mail or e-mail (by preference) for parents to review before testing. The child and parent met with the student investigator (S. H. Hung) at the follow-up clinic. Informed consent and assent were obtained, and the study session was completed in a nearby physiotherapy examination room. Each study appointment lasted approximately 60 minutes.

Demographics

Current age, age at diagnosis, time since diagnosis, time since treatment completion, risk level of treatment protocols for ALL (as defined by the National Cancer Institute of Canada),²¹ and total dose of anthracyclines (mg/m²) were recorded from patient charts. Measured height and weight as well as body mass index (BMI) were obtained by the follow-up clinic staff on the study assessment day.

Physical activity

PA was measured using the Physical Activity Questionnaire for Older Children (PAQ-C).²² The PAQ-C is a self-administered, 7-day recall questionnaire containing nine items that assess moderate to vigorous PA (MVPA) levels of children aged 8–14 years during a regular school week. Children rate their frequency of participation in common leisure and school-related activities during and after school. Each item is scored on a scale ranging from 1 to 5, and the average score for all items is calculated, with a final score ranging from 1 to 5. A score of 1 indicates low MVPA levels, and 5 indicates high MVPA levels. In a prior study, children without a cancer diagnosis achieved a mean score of 3.2, and the PAQ-C has acceptable test–retest reliability (rs=0.75–0.82).²² Cronbach's α coefficient of internal consistency was reported as r=0.73,²³ and convergent, divergent, and construct validity ranged from r=0.28 to r=0.63.²⁴

Fatigue

Cancer-related fatigue was measured using the chronic version of the Pediatric Quality of Life Inventory (PedsQL) Multidimensional Fatigue Scale.²⁵ This 18-item self-administered questionnaire captures three dimensions of fatigue: General Fatigue, Rest/Sleep Fatigue, and Cognitive Fatigue; each dimension has six items. Children grade their experience with fatigue issues in the previous month using a Likert scale ranging from 0 to 4 (0=never, 1=almost never, 2=sometimes, 3=often, 4=almost always). Total fatigue and dimension scores range from 0 to 100; because this scale is reverse-scored, higher scores relate to lower fatigue. Construct validity has been tested for pediatric cancer patients against healthy children (p<0.025), and internal consistency reliability has been tested in a pediatric cancer population, with Cronbach's α coefficients ranging from r=0.74 to r=0.92.²⁵

Motor performance

Motor performance was assessed using the Bruininks-Oseretsky Test of Motor Proficiency, Second Edition, Short Form (BOT-2 SF).²⁶ The BOT-2 SF is a standardized, norm-referenced test with fine and gross motor skills assessments for individuals aged 4–21 years. It uses 14 items to assess fine manual control, manual coordination, body coordination, and strength and agility. Raw performance scores for each item were converted and combined into a total standardized score ranging from 0 to 100. Standardized scores correspond to descriptive categories ranging from well above average to well below average, and scores between 41 and 59 were categorized as average.²⁶ The BOT-2 SF has been used in previous studies that included childhood cancer patients,^27,28 and it has been tested in non-cancer clinical populations with acceptable levels of reliability (Cronbach's α coefficient of r>0.80).²⁹

Physical function

Physical function was assessed using the 6MWT, a sub-maximal test used to assess an individual's overall aerobic endurance.³⁰ Participants were instructed to walk at a self-selected pace between two pylons as many times as they could in 6 minutes. 6MWT performance reflected the total distance covered during the 6 minutes (6MWD), and it was calculated by counting the number of full laps completed and measuring the distance of partial laps walked. American Thoracic Society³⁰ methodological guidelines were followed, with two exceptions: A pre-existing 25-metre course at BCCH was used instead of the recommended 30-metre course, and heart rate and oxygen saturation were not measured. Test–retest reliability for the 6MWT in chronic pediatric conditions has an interclass correlation ranging between 0.84 and 0.98.³¹ Criterion validity has been reported against estimated VO_2max (r=0.34) and measured VO_2max (r=0.76).³¹ 6MWD of healthy children of the same age, gender, and height, drawn from published data,³² was used to calculate the associated SD score (6MWT SDS) for normative data comparison and to allow for normalized comparison among participants.

Statistical analysis

Demographic, anthropometric, and medical data were summarized as means and SDs. Participants were categorized into different groups: anthracycline cardiotoxicity risk groups, based on total anthracycline dosage (mg/m²), with risk measured as low (0–100), moderate (101–250), or high (>250);³³ BMI weight status, based on the World Health Organization's Child Growth Standards³⁴ and further categorized as overweight or obese (OWO) and healthy weight (HW); and the BOT-2 SF descriptive categories. Outcome variables were summarized as means, SDs, and 95% CIs.

We used IBM SPSS Statistics, version 22.0 (IBM Corporation, Armonk, NY), with accepted levels of significance at p<0.05. Spearman's correlation coefficient was used to determine the association between PAQ-C scores and BOT-2 SF standardized scores, 6MWD, and 6MWT SDSs. We used the Mann–Whitney U test to test the difference in outcome variables among ALL risk level, BMI weight status, and anthracycline cardiotoxicity risk groups. Demographic, anthropometric, and medical covariates were analyzed to test their effect on primary and secondary outcomes using Spearman's correlations, and a simple linear regression model was used in cases in which associations were found. We used G*Power, version 3.1.9.2 (Statistical Consulting Group, University of California at Los Angeles), to perform a post hoc power analysis³⁵ for Spearman's correlation among PAQ-C and BOT-2 SF standardized scores, 6MWD, and 6MWT SDS.

Results

Twenty-one potential participants were identified and contacted. Of these, 13 children agreed to participate (62%). Of those not recruited, three children were not interested, three families could not be contacted, one family moved away and was not returning for follow-up, and one child had a conflicting appointment. No study-related adverse events occurred. Participant characteristics are summarized in Table 1.

Table 1.

Participant Characteristics

Characteristic	Mean (SD)	No. (%) of participants
Sex
Female		4 (31)
Male		9 (69)
Age at study date, y	9.6 (1.4)
Age at diagnosis, y	5.0 (1.7)
Time from diagnosis, y	5.1 (0.9)
Time after treatment completion, y	2.1 (0.7)
BMI (kg/m2)	19.2 (3.6)
Healthy weight		7 (54)
Overweight		3 (24)
Obese		3 (24)
SD score	0.9 (1.2)
Percentile	72.2 (28)
Anthracycline dosage, mg/m2	105.7 (44.7)
Low cardiotoxicity risk		9 (69)
Moderate cardiotoxicity risk		4 (31)
ALL risk level
Standard risk		9 (69)
High risk		4 (31)

BMI=body mass index; ALL=acute lymphoblastic leukemia.

The mean PAQ-C score was 3.1, which is lower than published data (a mean of 3.2) for children aged 9–15 years without a cancer diagnosis (see Table 2).²² Individual BOT-2 SF standardized scores illustrate that the majority of participants achieved average scores (see Figure 1). Mean BOT-2 SF standardized and percentile scores were 50.9 and 52.9, respectively. One participant's performance was below average, nine participants' performance was average, and three participants' performance was above average. Individual 6MWT SDSs show that the majority of participants (n=11, 85%) scored below healthy normative values (see Figure 2). Mean 6MWD was 544.42 m, and the mean 6MWT SDS was −1.62 (Table 2). All measured PedsQL Multidimensional Fatigue Scale scores were lower than healthy normative data (see Table 2).²⁵

Table 2.

Summary of Outcome Measures

Measure	Mean (SD)*	95% CI
PAQ-C score	3.1 (0.5)	2.8, 3.4
6MWT
6MWD, m	544.4 (95.4)	486.8, 602.1
Calculated norm distance, m	628.4 (29.0)	610.8, 645.9
6MWT SDS	−1.62 (1.50)	−2.53, –0.71
No. (%) Below norm	11 (85)
No. (%) Above norm	2 (15)
BOT-2 SF
Standardized score	50.9 (6.5)	47.0, 54.9
Percentile score	52.9 (22.5)	55.5, 89.0
BOT-2 descriptive category, no. (%)
Below average	1 (8)
Average	9 (69)
Above average	3 (23)
PedsQL Multidimensional Fatigue Scale
Total Fatigue	74.6 (11.1)	67.9, 81.3
General Fatigue	82.7 (10.5)	76.4, 89.0
Sleep/Rest Fatigue	70.2 (15.3)	61.0, 79.4
Cognitive Fatigue	74.6 (11.1)	58.5, 83.1

^*Unless otherwise indicated. PAQ-C=Physical Activity Questionnaire for Older Children; 6MWT=6-minute walk test; 6MWD=6-minute walk distance; SDS=SD score; BOT-2 SF=Bruininks-Oseretsky Test of Motor Proficiency, Second Edition, Short Form; PedsQL=Pediatrics Quality of Life Inventory.

Figure 1.

Scatterplot of participants' BOT-2 SF standardized scores. Note: Motor performance descriptive categories are as follows: 41–59 (dashed lines)=average; ≤40=below average; ≥60=above average. BOT-2 SF=Bruininks-Oseretsky Test of Motor Proficiency, Second Edition, Short Form.

Figure 2.

Scatterplot of participants' 6MWT SDS. Note: Scores below the dashed line (0 SDS) refer to children who performed below published normative data of healthy children 6MWT SDS=6-minute walk test SD score.

The association of PAQ-C scores with BOT-2 SF standardized score and 6MWT SDS is shown in Figure 3. PAQ-C scores were not associated with BOT-2 SF standardized scores (r_s=0.282, p=0.35), 6MWD (r_s=−0.429, p=0.14), or 6MWT SDS (r_s=−0.094, p=0.76). Outcomes were not associated with anthracycline dosage, anthracycline cardiotoxicity risk group, ALL risk level, time since treatment completion, or fatigue (data not shown). According to the post hoc power analysis for Spearman's correlation between PAQ-C and BOT-2 SF standardized scores, 6MWD, and 6MWT SDS, the power obtained was 0.54, 0.81, and 0.19, respectively. The sample sizes required to achieve 80% power were 23, 13, and 78, respectively.

Figure 3.

Scatterplots of participants' PAQ-C scores: (a) BOT-2 SF standardized scores; (b) 6MWT SDS. Note: Both scatterplots show no relationships. PAQ-C=Physical Activity Questionnaire for Older Children; BOT-2 SF=Bruininks-Oseretsky Test of Motor Proficiency, Second Edition, Short Form; 6MWT SDS=6-minute walk test SD score.

BMI SDS was negatively associated with 6MWD (r_s=−0.602, p=0.029) and 6MWT SDS (r_s=−0.691, p=0.009), and it approached significance with BOT-2 SF standardized scores (r_s=−0.515, p=0.07; see Table 3). Every increase in 1 BMI SDS was associated with a decrease in 0.70 6MWT SDS (constant=−0.98, β=−0.70, p=0.04, R²=0.325). A significant difference was found in the 6MWD (p=0.035) and 6MWT SDS (p=0.002) between OWO and HW BMI groups (see Table 3). Similarly, the difference in BOT-2 SF standardized scores between OWO and HW participants approached statistical significance (p=0.051). The majority of HW children achieved higher BOT-2 SF standardized scores and 6MWT SDSs, respectively, compared with OWO participants. No significant difference was found for PAQ-C scores by BMI weight status category. Significant associations with other covariates were not observed (data not shown).

Table 3.

Comparison of Means and Outcome Measures between Healthy Weight and Overweight/Obese Children

			Mann–Whitney U Statistics
	Mean		Mean Rank
Measure	HW, n=7	OWO, n=6	HW, n=7	OWO, n=6	U	Z	Effect size (r)	1-tailed p-value
6MWD, m	595.9	484.3	9.14	4.50	6.0	−2.143	0.6	0.035
6MWT SDS	−0.63	−2.78	9.86	3.67	1.0	−2.857	0.8	0.002
BOT-2 SF	54.7	46.5	9.00	4.57	7.0	−2.017	0.6	0.051
PAQ-C	3.1	3.1	6.93	7.08	20.5	−0.072	0.02	0.95

HW=healthy weight; OWO=overweight or obese; 6MWD=6-minute walk test distance; 6MWT SDS=6-minute walk test SD score; BOT-2 SF=Bruininks-Oseretsky Test of Motor Proficiency, Second Edition, Short Form; PAQ-C=Physical Activity Questionnaire for Older Children.

Discussion

To our knowledge, this is the first study to assess whether PA levels are related to motor performance and physical function specifically in children within 36 months of completing treatment for ALL. In the current study, self-reported PA was not associated with motor performance or physical function in children aged 8–13 years who had completed treatment for ALL in the previous 36 months. However, physical function, as measured by the 6MWT, was lower in 85% of children compared with published normative values. Physical function was associated with BMI, such that poorer 6MWT SDS was related to higher BMI SDS.

Previous studies investigating the relationship among PA, motor performance, and physical function included childhood cancer survivors of all ages and those who had completed treatment up to 13.4 years prior.¹⁹ Furthermore, results from these studies have been mixed, with one study⁵ reporting no significant association between 6MWD and sports participation in childhood ALL survivors, and two studies reporting that more active childhood cancer survivors had better overall physical function, as measured with the 6MWT⁶ and individual muscle performance tests.¹⁹ Another study found that less physically active childhood cancer survivors felt more fatigued and had less physical strength and endurance.¹⁸

The lack of observed association among PA, motor performance, and physical function in our study could be a result of several factors. First, the current study identified only one child with motor performance impairment. Either the motor performance outcome measure used was not appropriate to identify impairments, or motor performance was not an issue for study participants. Studies using the BOT tools with childhood ALL survivors have reported a relatively low incidence of impairment (5%–16%)^27,28 compared with another assessment tool, the Movement Assessment Battery for Children, Second Edition (MABC-2),³⁶ on which higher rates of motor performance impairments (25%–54%) have been reported for children treated for ALL.⁴ Differences between these assessment tools may stem from the differing levels of motor planning and cognition required for each test.²⁸ The BOT-2 SF was selected for use in this study because it had previously been used by members of the research team. Future work may want to consider using the MABC-2.

Second, PA levels may not be related to motor performance and physical function impairments after treatment for ALL. Children undergoing the early stages of treatment have reduced PA, mainly because of the constraints associated with hospitalization, which include medical apparatus, and side effects from intensive treatment regimes, such as fatigue and nausea, which lead to sedentary bed rest.^37,38 A spectrum of disuse, and overprotective parents and physicians monitoring children in remission, may contribute to patterns of inactivity persisting after treatment is completed.³⁹ Although five children in the current study had PAQ-C scores lower than published mean scores for children without a cancer diagnosis (mean score 3.2),²² two children had comparable scores, and six achieved higher scores. PA levels during maintenance therapy have been demonstrated to gradually increase to comparable levels in children without a cancer diagnosis,³⁸ a finding that is consistent with the PAQ-C score of the majority of children in the current study.

Consistent patterns of reduced 6MWD in childhood cancer survivors were observed^5–8 as well as a higher proportion of children who were OWO compared with the general population of children in Canada (46% vs. 32%⁴⁰).⁴¹ The current study also identified that higher BMI was associated with reduced physical function; this finding was consistent with those of previous studies in children with^5,19 and without a cancer diagnosis. BMI has been identified as a key predictor of 6MWD in children without⁴² a cancer diagnosis, potentially because of higher energetic cost with greater mass, attitude toward PA, altered movement patterns, and pain.⁴³ Although it is difficult to conclude that there is a causal relationship because children who complete treatment for ALL have an increased risk for weight gain and obesity,⁴¹ this pattern of elevated BMI accompanying reduced physical function indicates the need for regular follow-up for physical function assessment and weight management in pediatric oncology.^19,44

Our study has several limitations. The post hoc power analysis showed that the study was underpowered to demonstrate significant associations with PA for the BOT-2 SF and 6MWT SDS. Sampling bias potentially attracted more active children and families, which may prevent generalizability of the results. Also, identifying changes throughout cancer diagnosis, treatment, and recovery was not possible using a cross-sectional study design. There are also specific limitations related to the outcome measures used. For example, recall bias is common among children using self-report PA questionnaires.⁴⁵ The individual PAQ-C score also limited analysis by not considering PA type, intensity, and duration and seasonal changes in PA, and it lacks established normative data. In addition, the BOT-2 SF does not distinguish between gross and fine motor scores, although differences between them are important to consider in clinical assessments and designing rehabilitation interventions. Finally, the 6MWT is highly dependent on participant motivation and interest.³⁰

However, the current study had several strengths. To our knowledge, this was the first study to incorporate a validated self-report PA questionnaire to investigate whether both motor performance and physical function are associated with PA levels in children who have completed treatment for ALL.^5,19 Second, we believe that this study was the first to focus on understanding the physical factors associated with PA levels in children aged 8–13 years who have completed treatment for ALL in the previous 3–36 months, rather than using a sample that includes a wide age range and time since treatment completion.

Future studies can consider using a longitudinal study design and including objective measures for PA, such as pedometers or accelerometers, to allow for broader PA analysis⁴⁶ and to capture changes before and throughout participants' cancer treatment. Including physical and psychosocial variables, such as PA self-efficacy,⁴⁷ should also be considered because these factors have been shown to be predictors of PA levels in children diagnosed with cancer.^48,49

Conclusion

The current study observed no association among PA, motor performance, and physical function in children who had completed treatment for ALL. However, consistent with previous literature, poorer 6MWT performance was associated with higher BMI weight status. These findings have important clinical implications because pediatric⁴¹ and adult⁵⁰ childhood ALL survivors are reported to be at higher risk of developing obesity and the health conditions associated with obesity, such as cardiovascular disease.⁵¹ The study results emphasize the importance of the role of health care providers in weight management among childhood ALL survivors and the potential need for physiotherapy interventions focusing on physical function as children complete treatment.

Key Messages

What is already known on this topic

Reduced PA levels, motor performance, and physical function are common in children who have completed treatment for ALL. Previous studies investigating whether PA levels are related to motor performance and physical function include children of a wide age range and time since treatment completion, which hampers the ability to understand the different findings between studies.

What this study adds

This study is the first to investigate whether PA levels are related to motor performance and physical function specifically in children aged 8–13 years within the first 36 months after treatment completion for ALL. We observed no association among PA, motor performance, and physical function. However, poorer 6MWT performance was associated with higher BMI weight status. This study suggests the need for greater emphasis on weight management for childhood ALL survivors and potential need for physiotherapy interventions to improve physical function as children complete treatment for ALL.