eTable 1. Characteristics of Participants and Nonparticipants Among Childhood Cancer Survivors
eTable 2. Characteristics of Participants With and Without Exercise Intolerance Related to Treatment Exposures (N=1,041)
eTable 3. Characteristics of the Study Population (N=1,327)
eTable 4. Social Roles, Emotional Distress and HRQOL Among Survivors and Controls (N=1,327)
eTable 5. Associations Between Exercise Intolerance and Emotional Distress, Adjusting for Treatment Exposures, Chronic Health Conditions (N=1,041)
eTable 6. Associations Between Exercise Intolerance and Social Roles, Adjusting for Treatment Exposure, Chronic Health Conditions (N=920)
eTable 7. Associations Between Exercise Intolerance and Quality of Life, Adjusting for Treatment Exposure, Chronic Health (N=1,041)
eTable 8. Associations Between Exercise Intolerance and Quality of Life, Adjusting for Treatment Exposure, Chronic Health Conditions (N=1,041)
eTable 9. Characteristics of Participants With and Without Exercise Intolerance Including Demographic Characteristics, Health Behavior and Chronic Conditions (N=1,260)
eTable 10. Characteristics of Participants With and Without Exercise Intolerance for Treatment Exposures (N=1,260)
eTable 11. Associations Between Exercise Intolerance and Emotional Distress, Adjusting for Treatment Exposures, Chronic Health Conditions (N=1,260)
eTable 12. Associations Between Exercise Intolerance and Social Roles, Adjusting for Treatment Exposure, Chronic Health Conditions (N=1,127)
eTable 13. Associations Between Exercise Intolerance and Quality of Life, Adjusting for Treatment Exposure, Chronic Health (N=1,260)
eTable 14. Associations Between Exercise Intolerance and Quality of Life, Adjusting for Treatment Exposure, Chronic Health (N=1,260)
eFigure 1. Consort Diagram
eFigure 2. Non-Linear Association Between VO2 Max and Self -Report Physical Function
eFigure 3. Prevalence of Exercise Intolerance by Primary Diagnosis
eFigure 4. Geographic Distribution of Participants in the St. Jude Lifetime Cohort (SJLIFE)
Customize your JAMA Network experience by selecting one or more topics from the list below.
Hayek S, Brinkman TM, Plana JC, et al. Association of Exercise Intolerance With Emotional Distress, Attainment of Social Roles, and Health-Related Quality of Life Among Adult Survivors of Childhood Cancer. JAMA Oncol. 2020;6(8):1194–1202. doi:10.1001/jamaoncol.2020.2054
Is exercise intolerance associated with perceived well-being among adult survivors of childhood cancer?
In this cross-sectional study of 1041 adult survivors of childhood cancer, exercise intolerance, after accounting for lifestyle, treatment, and chronic health condition, was associated with emotional distress, attainment of social roles, and health-related quality of life. Specifically, anxiety, unemployment, and self-reported physical health were more prevalent in survivors with exercise intolerance when compared with those without exercise intolerance.
Data from this study suggest that exercise capacity among adult survivors of childhood cancer is associated with emotional health, ability to participate optimally in social roles, and health-related quality of life; improving exercise capacity may be a factor in these other important indicators of health.
Exercise intolerance is associated with increased risk for morbidity and mortality in childhood cancer survivors. However, an association between exercise intolerance and psychosocial outcomes has not been fully explored.
To examine the associations between exercise intolerance and emotional distress, attainment of social roles, and health-related quality of life in childhood cancer survivors.
Design, Setting, and Participants
A cross-sectional study including 1041 adult survivors of childhood cancer and 286 community controls in the St Jude Lifetime Cohort was conducted at St Jude Children’s Research Hospital. The study was performed from April 1, 2012, to March 15, 2020.
Exercise intolerance was defined as relative peak oxygen uptake less than 85% of age- and sex-estimated levels from maximal cardiopulmonary exercise testing.
Main Outcomes and Measures
Emotional distress was measured with the 18-item Brief Symptom Inventory-18, which includes overall Global Severity Index and depression, anxiety, and somatization subscales. Participants with T scores greater than or equal to 63 were classified as having elevated levels of distress. Social attainment was evaluated using patient-reported educational, employment, and marital status. Health-related quality of life was examined with the Medical Outcomes Survey Short Form-36. Participants with T scores less than or equal to 40 were classified as reporting poor health-related quality of life.
Of the 1041 participants, 528 were women (50.7%). The prevalence of exercise intolerance among survivors (mean [SD] age, 35.5 [9.2] years) was higher than that among controls (age, 34.5 [10.0] years) (survivors: 634 [60.9%] vs controls: 75 [26.2%], P < .001). After adjusting for age at diagnosis and cardiopulmonary exercise testing, sex, race/ethnicity, smoking, physical activity, and exercise intolerance were associated with an increased risk for anxiety (prevalence rate ratio [PRR], 1.95; 95% CI, 1.20-3.16), somatization (PRR, 1.86; 95% CI, 1.23-2.80), and unemployment (PRR, 1.76; 95% CI, 1.23-2.52); an inverse association was noted with having a college degree (PRR, 0.67; 95% CI, 0.50-0.88). Exercise intolerance was associated with an increased the risk for scoring less than or equal to 40 on the physical component summary of the Medical Outcomes Survey Short Form-36 (PRR, 3.69; 95% CI, 2.34-5.84). These associations persisted when either cancer treatment exposures or chronic health conditions were added to the model.
Conclusions and Relevance
The findings of this study suggest that exercise intolerance is independently associated with emotional distress, attainment of social roles, and health-related quality of life of long-term survivors of childhood cancer. The results also suggest that improving physiologic capacity may benefit general health and wellness, as well as emotional health, ability to participate in social roles, and health-related quality of life.
Five-year survival among children with cancer exceeds 80%,1 resulting in an increasing population of adults at risk for long-term adverse outcomes resulting from their diagnosis and treatment. The burden of long-term morbidity is substantial; by age 50 years, survivors of pediatric cancer will, on average, experience 17.1 (95% CI, 16.2-18.0) chronic health conditions.2,3
Exercise intolerance is an important marker of disease severity and decreased well-being in persons with heart,4 lung,5 renal,6 musculoskeletal,7 and neurologic disease.8 Recent data indicate that more than 50% of childhood cancer survivors are at risk for exercise intolerance.9 Because exercise intolerance in noncancer populations appears to affect emotional health, participation in social roles, and quality of life,10,11 understanding this association in childhood cancer survivors who demonstrate high rates of emotional distress, suboptimal attainment of social roles, and poor health-related quality of life (HRQOL)8,9 is important. Moreover, when exercise intolerance is remediated, improvements in symptoms of depression, anxiety, and somatic health; engagement in educational and work opportunities; and perceived well-being have been reported.12-14 Thus, to inform possible future interventions, the aim of this study was to estimate associations between exercise intolerance and emotional distress, attainment of social roles, and HRQOL among adult survivors of childhood cancer in the context of both treatment exposures and the presence of chronic disease.
Participants were members of the St Jude Lifetime Cohort, a study designed to characterize health outcomes among survivors of childhood cancer treated at St Jude Children’s Research Hospital.15-17 This study was conducted from April 1, 2012, to March 15, 2020. Study documents were approved by the St Jude Children’s Research Hospital Institutional Review Board. Participants provided written informed consent prior to assessment. Participants received financial compensation. This study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline for cross-sectional studies.
Participants for this analysis received cardiopulmonary exercise testing (CPET) as part of a study designed to evaluate the outcomes related to use of anthracycline and/or chest-directed radiation for organ system function, and the association of organ system function and lifestyle with exercise tolerance.9 Eligibility from among cohort members who had survived 10 years or longer, were treated between 1962 and 2007, and were aged 18 years or older (n = 5302) was determined before recruitment (March 1, 2012-August 15, 2017) across 4 strata of treatment exposure; the number recruited was powered to address the primary aims (n = 1260). Members of the St Jude Lifetime Cohort were stratified by anthracycline exposure without radiation, radiation without anthracycline exposure, both anthracycline and radiation exposure, or neither anthracycline nor radiation exposure. Within each stratum, participants were approached in random order for recruitment until the desired sample size was achieved. A total of 1369 individuals were approached. eFigure 1 in the Supplement shows the flow of participant enrollment and CPET completion. eTable 1 in the Supplement reports the demographic- and treatment-related variables for participants and nonparticipants. Those unable to complete exercise testing were placed in the nonparticipant group. The cohort also included a community-based comparison group (controls), with frequency matched to the overall St Jude Lifetime Cohort participant population by 5-year age group, sex, and race/ethnicity. Individuals serving as controls were recruited from among friends and relatives of patients currently treated at St Jude Children’s Research Hospital and non–first-degree relatives of St Jude Lifetime Cohort participants. The current analyses included survivors who did not have a history of congenital heart disease. Women were included if they were not pregnant and were 3 months or more postpartum. Control group inclusion criteria were identical to those of survivors, except they had no history of childhood cancer.
The primary outcomes for this study were emotional distress, attainment of social roles, and HRQOL. Emotional distress was evaluated by having participants complete the Brief Symptom Inventory-18. The Brief Symptom Inventory-18 is validated among pediatric and adult cancer survivors18-20 and among healthy adults.21,22 The Brief Symptom Inventory-18 includes the Global Severity Index and subscales for depression, anxiety, and somatization. Raw scores were transformed to T scores based on normative data (mean [SD], 50 ) and dichotomized such that participants with T scores greater than or equal to 63 (scores ≥90th percentile) were classified as having elevated levels of distress.
Social roles were categorized based on questionnaire responses and included educational attainment (college degree compared with less than college degree), employment (employed compared with unemployed, student, retired, or other), and marital status (married or living with a partner compared with unmarried or separated). Participants who reported graduating from college, being employed, and being married or living with a partner were categorized as having higher social role attainment.
Participants completed the Medical Outcomes Survey Short Form-36 (SF-36)23 to measure general health status over the past 4 weeks. The SF-36 provides subscale scores for physical function, role physical, bodily pain, general health, vitality, role emotional, social function, and mental health, and 2-component summary scales (physical and mental).23 The SF-36 has been widely used among adult survivors of pediatric cancer.24-26 Age- and sex-specific normative data were used to calculate T scores (mean [SD], 50 ).23 Participants with T scores less than or equal to 40 were classified as reporting poor HRQOL.20
Exercise capacity, included as an independent variable, was assessed with maximal CPET, using a modified Bruce treadmill protocol.27 Participants unable to walk (eg, those with lower extremity paralysis, amputations without prostheses, and poor balance) were tested with leg or arm cycle ergometers (n = 109). Continuous 12-lead electrocardiogram and breath by breath gas analysis (Ultima CardioO2, MCG Diagnostics) data were collected.28 Cardiopulmonary exercise testing was terminated at a participant’s request, maximal exertion, or for safety, including symptoms of ischemia (>2 mm ST depression on electrocardiogram), arrhythmias, hypertensive-level blood pressure (>250/115 mm Hg), angina, shortness of breath, wheezing, or failure of heart rate to increase with increased exercise intensity. Peak rating of perceived exertion was reported at test termination. Exercise intolerance was defined as relative peak oxygen uptake (oxygen consumption per unit time, peak in milliliters per kilogram per minute) less than 85% of age and sex estimated.9,29-32
Participant characteristics, including age at assessment, sex, and race/ethnicity, were obtained by self-report. Diagnosis and treatment information was abstracted from medical records and included diagnosis, age at diagnosis, time since diagnosis, chemotherapy exposure and cumulative doses, radiation exposure and region-specific dosimetry, and surgery. Treatment exposures included in these analyses were previously reported to be related to emotional distress, social role attainment, and HRQOL.20,33 Treatments significant in bivariate analysis are reported in eTable 2 in the Supplement and include alkylating agents, high-dose methotrexate, methotrexate, brain radiation, chest radiation, abdominal/pelvic radiation, and amputation. Physical activity and smoking status were obtained by self-report. Physically active participants were those who reported at least 150 minutes per week of moderate-intensity, 75 to 150 minutes per week of vigorous-intensity, or an equivalent combination of moderate- and vigorous-intensity aerobic activity.34 Current smokers were those who reported smoking 100 cigarettes or more in their lifetime and smoking in the past month. Those who reported ever smoking but not currently smoking were classified as nonsmokers.35,36 Chronic health conditions included cardiac conditions, evaluated via echocardiography and categorized as impaired if the ejection fraction was less than 5337; pulmonary conditions, evaluated via prebronchodilator spirometry and categorized as impaired according to a modified version of the Common Terminology for Adverse Events17,38 if forced expired volume in 1 second was less than 80% estimated for race/ethnicity and sex17; neurosensory conditions, evaluated by a Modified Total Neuropathy Score where greater than or equal to 5 is considered impaired39-41; and cognitive impairment, evaluated with Wechsler Abbreviated Scale of Intelligence,42 classifying those with an age-adjusted z score less than or equal to −2 as impaired.
Descriptive statistics were used to characterize the study population. Differences between participants and nonparticipants, as well as survivors and controls, were evaluated using χ2 statistics for categorical variables and 2-sample t tests for continuous variables. Percentages for social roles, Brief Symptom Inventory-18, and SF-36 categories were compared between participants with and without exercise intolerance with χ2 statistics. For attainment of social roles outcomes, analyses were conducted on participants aged 25 years or older. Generalized linear models (3 per outcome) with a binomial distribution and a log link43 were used to evaluate associations between exercise intolerance and emotional health, attainment of social roles, and HRQOL among childhood cancer survivors. Because the prevalence of our outcomes of interest was low among control participants, we did not evaluate this association in controls. The first model was adjusted for demographics and lifestyle factors, the second for demographics, lifestyle factors, and treatment exposures, and the third for demographics, lifestyle factors, and chronic health conditions. Treatment exposures and chronic conditions were not included in the same models because chronic conditions are likely on the causal pathway between treatment exposures and emotional distress, HRQOL, and social attainment.44-46 Prevalence rate ratios (PRRs) and 95% CIs are presented as measures of these cross-sectional associations. We used a false discovery rate less than or equal to 10% and adjusted 95% CIs accordingly in our generalized linear models to account for type 1 errors related to multiple comparison.47 Although examination of the data (eFigure 2 in the Supplement) indicated a nonlinear association between exercise intolerance and emotional distress, HRQOL, and social attainment, these associations were also evaluated in a sensitivity analysis, treating relative peak oxygen uptake as a continuous measure. In an additional sensitivity analysis, participants who did not complete CPET were classified as having exercise intolerance, and generalized linear models were constructed. Two-sided P values <.05 were considered significant. SAS, version 9.4 (SAS Institute Inc), was used to conduct all analyses.
Among 1369 eligible adult survivors of childhood cancer, 24 completed a survey only, 67 declined participation, and 18 were lost to follow-up. Of the 1260 who completed a campus visit, 1041 individuals (528 [50.7%] women) completed CPET (eFigure 1 in the Supplement). Compared with nonparticipants, participants were more likely to be female, younger at diagnosis, younger at assessment, and treated for acute lymphoblastic leukemia (eTable 1 in the Supplement).
The characteristics of the study participants (survivors and controls) are summarized in eTable 3 and eFigure 4 in the Supplement. Compared with survivors, controls were more likely to be non-Hispanic white (258 [90.2%] vs 886 [85.2%]), younger (mean [SD], 34.5 [10.0] vs 35.5 [9.2] years), and to report 150 minutes/week of at least moderate physical activity (179 [62.6%] vs 531 [51.0%]). Controls were less likely than survivors to have exercise intolerance (75 [26.2%] vs 634 [60.9%]) and cardiac (19 [6.6%] vs 239 [23.0%]), pulmonary (20 [7.0%] vs 316 [30.4%]), and neurosensory (15 [5.2%] vs 268 [25.7%]) conditions. Survivors with central nervous system tumors reported the highest prevalence of exercise intolerance (128 [81.5%]), followed by those with bone (78 [70.3%]) and Wilms (50 [58.8%]) tumors (eFigure 3 in the Supplement).
Social attainment, emotional distress, and HRQOL outcomes among survivors and controls are provided in eTable 4 in the Supplement. Survivors were less likely than controls to report college graduation (406 [44.1%] vs 141 [60.5%], P < .001), employment (684 [77.4%] vs 192 [84.6%], P < .001), or being married/living with a partner (523 [56.9%] vs 189 [81.1%], P < .001), and more likely to report depression (11.9% vs 14 [4.9%], P < .001), anxiety (99 [9.8%] vs 13 [4.6%], P < .001), somatization (139 [13.7%] vs 13 [4.6%], P < .001), and global distress (9.9% vs 10 [3.5%], P < .001). Survivors were also more likely than controls to report poor HRQOL (Physical Component Summary: 151 [15.1%] vs 9 [3.2%], P < .001; Mental Component Summary 178 [17.8%] vs 33 [11.8%], P < .001).
After adjusting for age at diagnosis, age at CPET, race/ethnicity, smoking, and physical activity, survivors reported more anxiety (PRR, 1.95; 95% CI, 1.20-3.16) and somatization (PRR, 1.86; 95% CI 1.23-2.80) than those without exercise intolerance (Table 1). Adding treatment exposures (alkylating agents, high-dose methotrexate, methotrexate, brain radiation, chest radiation, abdominal/pelvic radiation, and amputation), and chronic conditions (cardiac, pulmonary, and neurosensory conditions) to models attenuated the associations but did not fully explain the contributions of exercise intolerance to anxiety and somatization. Exercise intolerance was not associated with depression or global distress.
Among survivors aged 25 years or older, after adjusting for age at diagnosis, age at CPET, race/ethnicity, smoking, physical activity, and full-scale IQ, survivors with exercise intolerance were less likely to have a college degree (PRR, 0.67; 95% CI, 0.50-0.88) and more likely to be unemployed (PRR, 1.76; 95% CI, 1.23-2.52) than those without exercise intolerance (Table 2). Adding treatment exposures and chronic conditions to the models did not alter these associations. Exercise intolerance was not associated with marital status.
After adjusting for age at diagnosis, age at CPET, race/ethnicity, smoking, and physical activity, survivors with exercise intolerance were more likely than those without exercise intolerance to score less than or equal to 40 on the physical function (PRR, 5.54; 95% CI, 3.20-9.53), role physical (PRR, 2.62; 95% CI, 1.74-3.94), bodily pain (PRR, 2.14; 95% CI, 1.43-3.22), general health (PRR, 2.31; 95% CI, 1.65-3.23), role emotional (PRR, 1.86; 95% CI, 1.21-2.87), and social function (PRR, 1.79; 95% CI, 1.20-2.68) subscales, as well as the physical component summary (PRR, 3.69; 95% CI, 2.34-5.84) of the SF-36 (Table 3 and Table 4). Adding treatment exposures and chronic conditions to the models provided similar results. Results from generalized linear models for all outcomes treating exercise intolerance as a continuous variable were consistent with dichotomous findings (eTables 5-8 in the Supplement); models that included survivors unable to complete CPET, classifying them as exercise intolerant, did not change our results (eTables 9-14 in the Supplement).
In this large population of young adult survivors of childhood cancer, we found independent associations between exercise intolerance and emotional distress, attainment of social roles, and HRQOL. These associations persisted after accounting for age, sex, lifestyle factors, previous treatment exposures, and chronic health conditions. As previous data in noncancer populations suggest that improving exercise tolerance results in improvements in emotional health,48-50 participation in social roles,51 and HRQOL,52 these findings suggest support for interventions to improve exercise tolerance in this vulnerable population. Potential intervention approaches will have to accommodate limitations associated with chronic health conditions prevalent among adult survivors of childhood cancer.
Our findings that survivors with exercise intolerance have emotional distress are consistent with previous data from several large, noncancer cohort studies in which investigators reported associations between lower levels of cardiorespiratory fitness and depressive symptoms53-55 and provide a potential target for intervention among survivors with both exercise intolerance and poor mental health. One randomized clinical study among patients with congestive heart failure found that improved exercise capacity following 8 weeks of aerobic training reduced depression and anxiety,56 and a meta-analysis supports an association between 12 weeks of exercise training and cardiorespiratory fitness among persons with depression.57 Exercise training may have the potential to improve both emotional health among the approximately 61.0% of survivors with impaired exercise capacity and exercise capacity among the 10.0% of survivors with poor mental health.
Exercise intolerance in our population also appears to interfere with attainment of social roles. Studies from other populations have suggested that individuals with impaired cardiorespiratory fitness are less likely than those not impaired to participate in physical and social activities.58 Both disability59 and fatigue60 interfere with everyday performance. For example, in a study of 683 individuals aged 20 to 69 years employed in occupations that require some physical exertion, investigators found that cardiorespiratory fitness was associated with work performance.61 Furthermore, individuals with exercise intolerance become less physically active and have reduced social contacts.62 Leisure time activities are often the first to be eliminated, as they generally require efforts not critical to daily life. Less time participating in leisure time activities might increase the risk of social isolation. Some evidence suggests that exercise training, in addition to its known cardiorespiratory benefits, may enhance participation. For example, aerobic exercise training among persons with stroke improves cardiovascular fitness and performance on daily life activities,63 and cardiac rehabilitation reduces deconditioning and disability in older adults.64 Exercise training may even mitigate lack of participation in social activities; cardiac rehabilitation decreases feelings of social isolation among older adults when performed in a group setting.65
Survivors in our study with exercise intolerance reported poor HRQOL, which is in agreement with previous literature describing this association among persons with chronic obstructive pulmonary disease,66 those at risk for cardiovascular disease,67 and women at risk for gestational diabetes.68 Because HRQOL is a stronger predictor of morbidity and mortality than many objective measures of health,69,70 targeted interventions to improve HRQOL are important. Improving cardiorespiratory fitness appears to benefit quality of life in adult cancer survivors and other populations with chronic disease.71 Cardiac rehabilitation that included aerobic and resistance training improved cardiorespiratory fitness by 14% among 152 breast cancer survivors and significantly improved quality of life in 7 of 8 SF-36 domains,72 as well as exercise capacity (27% change, P < .001) and overall quality of life (14% change, P < .001) among persons with a mean (SD) age of 59 (10) years with obesity and coronary artery disease.73 In addition, data from a recumbent cycling intervention among 65 (mean age, 45 years; range, 37-48 years) persons receiving hemodialysis suggest significant improvement (P < .01) in exercise capacity and in all 8 domains of HRQOL measured with the SF-36.74
These findings should be considered in the context of several study limitations. First, these are cross-sectional data; the temporal associations between exercise tolerance and social and emotional outcomes could not be assessed. Longitudinal studies among childhood cancer survivors to assess the directionality of these associations are needed. Second, the study does not include information about the childhood cancer survivors with the most impairments who were too ill to complete CPET, potentially resulting in an underestimate of association between exercise intolerance, emotional distress, social participation, and HRQOL. An additional sensitivity analysis was conducted to address this limitation (eTables 9-14 in the Supplement), with those who did not complete CPET classified as having exercise intolerance. Results were consistent with the original analysis, although the magnitude of the association was larger. Third, only 2% of participants were Hispanic or other race/ethnicity. Because it is plausible that the social and emotional significance of exercise intolerance may differ by race/ethnicity, our findings may not apply to those who identify as Hispanic, Asian, or other. Fourth, our results are based on a population treated at a single institution, and even though previous data indicate that participants are largely representative of the source population (eFigure 4 in the Supplement shows the geographic distribution of childhood cancer survivors previously treated at St Jude Children’s Research Hospital),16 results may not be generalizable to survivors treated elsewhere. Fifth, not every eligible survivor (2%-4%) completed the emotional distress, social role, and HRQOL components of the questionnaires. If questionnaire completion was related to or dependent on exercise tolerance, our estimates of the association between exercise tolerance and these outcomes may be biased toward the null.
The findings of this study suggest that exercise intolerance has an independent association with emotional distress, attainment of social roles, and HRQOL in childhood cancer survivors. Because exercise intolerance may be remediated with training, even in populations with significant chronic disease, this finding provides an opportunity for intervention with broad targets. Enhancing physiologic capacity may not only improve general health and wellness, but also emotional health, the ability to participate optimally in social roles, and perception of quality of life among adult survivors of childhood cancer.
Accepted for Publication: April 17, 2020.
Corresponding Author: Kirsten K. Ness, PhD, Department of Epidemiology and Cancer Control, St Jude Children’s Research Hospital, 262 Danny Thomas Pl, Mail Stop 735, Memphis, TN 38105 (email@example.com).
Published Online: June 25, 2020. doi:10.1001/jamaoncol.2020.2054
Author Contributions: Dr Ness had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Concept and design: Hayek, Brinkman, Durand, Srivastava, Hudson, Armstrong, Ness.
Acquisition, analysis, or interpretation of data: Hayek, Brinkman, Plana, Joshi, Luepker, Green, Partin, Santucci, Howell, Srivastava, Hudson, Robison, Armstrong, Ness.
Drafting of the manuscript: Hayek, Durand, Santucci, Ness.
Critical revision of the manuscript for important intellectual content: Hayek, Brinkman, Plana, Joshi, Luepker, Green, Partin, Howell, Srivastava, Hudson, Robison, Armstrong, Ness.
Statistical analysis: Hayek, Srivastava, Ness.
Obtained funding: Hudson, Robison, Armstrong.
Administrative, technical, or material support: Hayek, Durand, Partin, Santucci, Robison, Armstrong.
Supervision: Plana, Srivastava, Armstrong, Ness.
Conflict of Interest Disclosures: None reported.
Funding/Support: Support for this study was provided to St Jude Children’s Research Hospital by grants from the National Cancer Institute (U01 CA195547, Drs Hudson and Robison, principal investigators; R01 CA157838, Dr Armstrong, principal investigator), a Cancer Center Support grant (P30 CA21765, Charles Roberts, MD, principal investigator), and the American Lebanese-Syrian Associated Charities.
Role of the Funder/Sponsor: The funding sources had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Additional Contributions: Tracie Gatewood (St Jude Children’s Research Hospital, Memphis) assisted with manuscript preparation; there was no financial compensation outside of salary.