Dietary and Exercise Interventions for Pediatric Oncology Patients: The Way Forward

Abstract

This review focuses on diet and exercise interventions that have been conducted in pediatric cancer and pediatric stem cell transplant patients. It examines the different reasons for conducting lifestyle interventions with attention to the different outcome measurements and feasibility of these measures with an argument toward a need for standardization to move the field forward.

Pediatric cancer now has a 5-year survival rate of more than 80%. Thus, there is a growing number of long-term survivors at risk for therapy-related complications. These include obesity and predisposition to the metabolic syndrome in those exposed to cranial radiation and steroids (1,2). Many pediatric cancer patients are deconditioned at diagnosis; fitness and loss of lean body mass often worsen on therapy and persist into long-term follow-up (3–5). Additionally, malnutrition and/or obesity from disease, treatment exposure, and illness-related poor eating habits potentially impact chances of oncologic cure (6). Finally, the pediatric cancer experience can cause psychological strain, negatively impacting health-related quality of life (HRQOL) (3). Fortunately, these disease- and therapy-related conditions are potentially responsive to diet and exercise interventions. Attempts to find unifying best practices for lifestyle interventions in childhood cancer patients are hampered by lack of large randomized data, diversity of outcomes, and heterogeneous treatment strategies, making it difficult to draw solid conclusions (7–11). This review describes existing interventions and outcome measures for pediatric oncology patients, with special attention intervention feasibility as it relates to setting and timing, and provides suggestions for future trials.

Dietary Interventions in Pediatric Cancer

Modifying nutritional intake has benefits for everyone including maximizing chances of disease cure, preventing infection, and managing body composition before it perpetuates as obesity during survivorship. Unfortunately, surveys indicate that nutritional services are inconsistent across pediatric oncology centers; not all patients get assistance in maintaining a healthy diet (12,13) Recent evidence suggests that being overweight during treatment for acute lymphoblastic leukemia increases risk of persistent disease and relapse (6,14), which really requires a paradigm shift for nutrition management during therapy. Because children typically lose weight during initial intense phases of treatment, the focus has traditionally been on optimizing calories. However, 36% of children are overweight when diagnosed with acute lymphoblastic leukemia, and 79% consume more than recommended total calories (1,15). Maintenance of normal body mass index (BMI) and weight gain are accompanied by gains in body fat percentage, often with corresponding loss in lean mass (sarcopenic obesity) (16) due to the toxicity of specific treatment modalities and compounded by inactivity. In fact, emerging evidence indicates that caloric intake actually decreases in the most common pediatric cancer populations (acute lymphoblastic leukemia [ALL] and central nervous system tumors) during therapy but fat intake increases, thus a focus on the quality of nutritional intake should be prioritized over total calories (1,17,18).

Nutritional interventions during therapy have yielded only modest results, most have been limited to children with ALL, and none have evaluated the impact of the intervention on survival. One study implemented a nutritional counseling program in 22 pediatric ALL patients during maintenance chemotherapy (12 in intervention group) over a year, with a resultant decrease in caloric intake but no changes in body weight or waist circumference (19). Another trial used a phone counseling intervention promoting healthy eating and physical activity, targeting caregivers of pediatric cancer patients (n = 53; 27 in intervention group) (20). Modest reductions in caloric intake among caregivers and in sugary beverage consumption among children were reported. Both caregivers and children had modest reductions in BMI. Two other randomized studies that added nutritional education to physical activity interventions, one during maintenance chemotherapy for ALL (21) and one after completion of ALL therapy (22), had no impact on weight, even in the presence of increased physical activity. Patients with obesity caused by tumors involving the hypothalamus often struggle to lose weight despite dietary interventions. One comprehensive clinic demonstrates that weight gain can be mitigated by a multidisciplinary approach focused on social skills classes, cognitive behavior therapy, and motivational interviewing (23).

Barriers to Successful Nutrition Interventions

There are several barriers that prevent successful implementation of nutritional interventions. First, although there are many national recommendations for healthy eating, there are no standard programs for recommended nutritional intake during cancer therapy. As a result, it is challenging to set benchmarks for ideal intake and to know which components of dietary intake should be targeted during treatment (12). Secondly, most dietary intake interventions to date have focused on the patient only (10). Because a patient’s dietary intake is often dependent on what their family is eating, patients might not have the opportunity to make healthy choices if they are not available in the family’s meal and/or snack repertoire. Thus, family-based approaches are the most likely to result in successful change and/or choice. Finally, dietary intake varies greatly between individuals, with reasons for suboptimal (too little, too much, the wrong foods) dietary intake dependent on age, preference, religion, altered taste related to chemotherapy administration, nausea, and parent’s desire to provide something for their sick child. Interventions will have to incorporate individualized nutrition plans in the context of reproduceable and evaluable nutritional guidelines.

Specific Nutritional Modifications

Prescribing a neutropenic diet (avoiding raw vegetables and fresh deli meats), conceived to prevent infections in pediatric oncology patients by limiting introduction of bacteria into the gastrointestinal tract (24), is not uncommon in practice (25), even though data from a randomized study (n = 150) indicate that there is no difference in infection rates in patients using a neutropenic diet compared to those who followed Food and Drug Administration food safety guidelines (24). Furthermore, emerging evidence indicates that bacteria in the microbiome may actually help prevent infections. There is little data to support other diet modifications during therapy other than a case series indicating feasibility of a high-fat, low-carbohydrate diet, often used to treat refractory seizures, in pediatric brain tumor patients (26). A summary of key nutritional recommendations for the way forward is given in Table 1.

Table 1.

Key recommendations for dietary intake studies

Maintaining a healthy weight is important for both general heath but also can impact the chance of cancer cure. Body mass index should not be the only assessment of healthy weight because cancer patients often have sarcopenic obesity (increased fat mass percentage with decrease in lean muscle mass). Current published dietary interventions studies have only shown a minor impact in weight changes. Ideal nutrition goals for pediatric cancer patients need to be defined. Nutritional interventions that target the whole family may increase the chances of success. When designing an intervention, balancing the individual needs of a given patient with creating a reproducible and evaluable intervention is needed.

Barriers to Successful Exercise Interventions

Exercise is good for everyone including children with cancer, given their increased risk of premature cardiovascular disease and mortality (27). However, there are some barriers that make exercise difficult. A pediatric cancer diagnosis is a major life event. Appropriate focus on cure stresses families and leads to complications such as unemployment, divorce, and home relocation, all which make it difficult to incorporate exercise into a daily routine (28). Although parents often recognize that their children are overweight, most do not remember that their oncologist discussed potential complications of being overweight (29). Other barriers reported by patients and families include being too tired, too busy, afraid of injury, and lack of access to exercise resources and having negative thoughts and feelings about themselves (30). Exercise trials remove some of these barriers (access, fear of injury) but are difficult to conduct in families disrupted and consumed by the intensity of cancer therapy (30,31). Another potential barrier to exercise is asking the patient to be active when their family is not. As with nutritional interventions, physical activity interventions are most likely to be successful when the entire family is included. Designing a feasible and successful intervention is challenging and requires input from family stakeholders in terms of when to intervene, where to intervene (venue), and what an acceptable time commitment would be.

Exercise Interventions Targeting Body Composition

Weight loss is a common aim in exercise interventions. However, in 12 studies, no statistically significant weight loss occurred in the intervention group, and in 6 randomized trials delivered during active ALL therapy, weight gain occurred regardless of group assignment (21,22,32–41), In addition, inclusion criteria, intervention timing, and measurement techniques are not uniform. Positive findings include one study among patients whose BMI was above the 85th percentile, where rate of BMI gain (height and weight measurements) during therapy was less in those in the exercise group than in the control group (22), another that demonstrated loss of fat mass (dual x-ray absorptiometry measurement) during the first 2 years after therapy among children with ALL who were randomly assigned to exercise during therapy when compared to control subjects (35), and one that showed statistically significant decreases in overall and abdominal body fat (skinfold and waist circumference measurements) in 17 ALL survivors after a 16-week exercise intervention (36). Further study is needed to establish inclusion criteria (targeting those at greatest risk), to identify the most appropriate timing for intervention, and to determine the best body mass assessment tool.

Exercise Interventions Targeting Physical Deconditioning

Physical deconditioning in children with cancer includes loss of strength, flexibility, fitness, and mobility, all which result from or contribute to inactivity. Evaluating the effects of interventions on these outcomes is challenging, because studies (n = 36; 30 in cancer, 6 in stem cell transplant [SCT]) are small (mean = 28, subjects range = 6–97) and use heterogeneous measurement tools and outcome assessment timing (21,22,32–65). In addition, adherence to different exercise interventions varies from 25% to 100%, making it difficult to determine if null results are associated with poor adherence or the result of inadequate exercise dose. Adherence was highest in studies where the exercise was directly observed by study personnel, or when conducted in an inpatient setting as opposed to in-home settings. Nevertheless, home or community-based interventions are necessary, because it is difficult to translate labor-intensive and costly supervised or hospital-based programs into most clinical settings.

Increasing physical activity is a focus for many exercise interventions. Sedentary and active behavior can be measured objectively using accelerometers, now incorporated into most smart phones. Accelerometer data are hard to evaluate. Complete data require full time wear, and children frequently forget to wear them, or lose them. Nevertheless, they do provide objective data. Self-report or exercise logs are also used to evaluate activity but provide less reliable estimates (37,53,65) or have low adherence (46,65), respectively. Of 11 studies that used accelerometers, 10 showed no statistically significant change in moderate-to-vigorous activity for the intervention group when compared to control subjects (21,22,32,34,48–50,52,53,57,63). Although these data are discouraging, sensor technology is improving. Newer devices (66) that stick to the body and have prolonged durability, or data captured on smart phones typically carried everywhere, may be a better way to capture movement-derived data.

Other goals of exercise interventions are to improve flexibility, strength, and motor function. Ankle range of motion, a common problem for children with ALL, has been evaluated with goniometry in 10 studies, with 3 showing some improvement in the intervention group (32,51,61), 1 showing worsening (35), and the rest showing no significant change (32,33,35,38–40,46,51,61,64). General flexibility was assessed in three studies (one improvement) using the sit-and-reach test (21,33,64,67). Handheld dynamometers or ability to lift increasing weight loads are used to measure strength. Knee extension (six studies used, three showing improvement); hip abduction (three studies, one improvement); handgrip (six studies, three improvement); shoulder flexion (two studies, no change) (32,33,40–42,46,50,51,56); and increases in weight load (four studies, three improvement) (34,38,39,54) have been measured. The Bruininks-Oseretsky Test of Motor Proficiency was used to test motor function in three studies. Statistically Significant improvement in response to exercise was reported in two (33,55,57). Psychometric assessment to determine an optimal set of tests to evaluate flexibility, strength, and motor function in children with cancer would improve interpretability of studies.

Fitness and functional mobility are also exercise intervention targets. Fitness assessment by evaluating peak oxygen uptake on cycle ergometers (four studies, one showing improvement) or treadmills (two studies, both showing improvement) is the gold standard but is time intensive and may be difficult for children who are deconditioned (32,38–40,55,68). Walk tests—distance walked in a predetermined number of minutes—are also used to assess fitness (eight studies, six improvement) (33,41,42,51,53,59,60,62). Walk tests require an open hallway clear of obstacles, are effort dependent, and again may be difficult for deconditioned children. Functional mobility is evaluated with the Timed Up and Go Test—the time to stand from a chair, walk 10 feet and return to sitting (seven studies, three improvement) (34,38–41,60,64), or the Timed Up and Down Stairs—the time required to ascend and descend one flight of a standard staircase (seven studies, two improvement) (34,38–41,51,60). Like the walk tests, these measures of functional mobility are easy to perform but are effort dependent and require a motivated child.

Exercise Interventions Targeting HRQOL

HRQOL is also a target of pediatric oncology exercise interventions (69,70). The most commonly used measure to date is the 23-item Pediatric Quality of Life Inventory (PedsQL) 4.0 Generic Scale (69). It was designed for ages 2–18 years and uses both child reporting and parent proxy to assess physical, emotional, social, and school functioning. Fourteen exercise intervention (five randomized) studies in children with cancer have used this measure. These vary by sample size (6–69 participants), type of exercise (11 traditional, 3 yoga), encompass all phases of treatment (newly diagnosed, long-term survivorship), use of proxies, and frequency of reporting (time between assessments 3 weeks–18 months) (32,37,43,45–47,50,51,54,59,60,64,71). Seven studies showed improvement in at least 1 domain. In one study, the exercise intervention was delivered early in leukemia therapy. HRQOL decreased as therapy became more intense (47). In another study, authors found no difference in HRQOL between exercise and control groups as they progressed further out from end of therapy (32), illustrating the difficulty of evaluating effects of exercise interventions on HRQOL, an outcome largely influenced by cancer treatment intensity.

Two other common patient and/or proxy reported outcome measures are the 27-item PedsQL 3.0 Cancer Module (69), which assesses pain and/or injury, nausea, procedural anxiety, treatment anxiety, worry, cognitive problems, perceived physical appearance, and communication, and the 18-item Multidimensional Fatigue Inventory, which assesses general, sleep and/or rest, and cognitive fatigue. Just one of four exercise intervention studies in children with cancer demonstrated improvement on the PedsQL 3.0 Cancer Module or the Multidimensional Fatigue Inventory as a result of the intervention, and gains were likely not clinically signficant (32,34,37,54,56,59,60,65,71). Many other measures have been employed. In all cases, these measurements have been evaluated to be feasible, but standardization is needed across trials for these measurements to be comparable.

Exercise Interventions Targeting Other Cancer Comorbidities

Exercise has other potential benefits that could prevent or mitigate complications of pediatric cancer therapy, including immune system dysfunction, cardiac toxicity, and neural recovery, and bone mineral density. Data are preliminary, but somewhat positive. Five studies with 6–10 patients in the exercise group resulted in an association between exercise and a boost in neutrophil count among children with cancer (72–74), a small increase in natural killer cell cytotoxicity by boosting CD56^dim cells, and mitigation of fall in dendritic cell counts among children after stem cell transplantation (75,76). Among survivors exposed to anthracyclines, Smith et al. (77) demonstrated that exercise is safe among those with reduced ejection fraction. Studies also indicate that that exercise can improve endothelial structures and function and peak circumferential systolic and diastolic strain (68,78,79). Among pediatric brain tumor survivors, exercise is associated with increasing brain white matter, hippocampal volume, and cortical thickness (80,81). Studies evaluating the effects of exercise on bone mineral density and metabolic function in children with cancer and among survivors are largely null with small sample sizes and poor intervention adherence (82–84). One randomized study, using low-magnitude mechanical stimulation as an exercise mimetic, did have a statistically and clinically meaningful (0.5 standard deviations) effect on bone mineral density in children and adolescents with low bone mineral density after completion of cancer therapy (85). There is a need for more studies with larger sample sizes of the effects of exercise on immune function, cardiovascular health, neural recovery, bone density, and metabolic health in pediatric oncology patients. A summary of key exercise recommendations for the way forward is given in Table 2.

Table 2.

Key recommendations for exercise intervention studies

Exercise is good for all and can help prevent premature cardiovascular disease and mortality. During cancer, patients often feel sick and are deconditioned, therefore, it is challenging to design interventions that are feasible. Targeting exercise intervention to the whole family unit could increase success. Weight loss is difficult to achieve with exercise interventions alone. Evaluating changes in physical activity, functional mobility, and fitness as endpoints for an exercise intervention are effective outcomes to demonstrate improvement; however, both standardization and better definitions of clinically significant outcomes are needed. Measuring changes in health-related quality of life and fatigue can also give objective evidence of impact for an exercise intervention; however, standardization and establishing a metric for clinically significant outcomes is needed. Exercise has additional health benefits and further study into how it impacts the immune system, heart, brain, and bones is needed.

Diet and exercise interventions are feasible in the pediatric oncology and SCT populations, however, there are many study design issues and barriers to participation that need to be overcome. There are many compelling reasons for conducting these interventions, and when designing new interventions, consideration should be made to pick outcomes that are reliable and reproducible. Standardization of outcome measures is needed so that studies can be compared and combined.

Notes

Affiliations of authors: Vanderbilt-Ingram Cancer Center, Nashville, TN (AJE); Monroe Carell Jr. Children’s Hospital at Vanderbilt Division of Pediatric Hematology-Oncology, Nashville, TN (AJE); Department of Epidemiology and Cancer Control, St. Jude Children’s Research Hospital, Memphis, TN (KKN).

The authors have no conflicts of interest to report.