Those whose magnetic resonance imaging (MRI) files were technically inadequate and not analyzed for the MRI-related variables were still included in analyses of other (non-MRI) variables. BMI indicates body mass index.
eTable 1. Inclusion and Exclusion Criteria
eTable 2. Run-in Training Program
eTable 3. Aerobic Training Program
eTable 4. Resistance Training Program
eTable 5. Resistance Training Exercises
eTable 6. Participants on Medications by Group Assignment
eTable 7. Overall Adverse Events by Study Group
eTable 8. Adverse Events by Study Group Related and Unrelated to Participation in the HEARTY Trial
eTable 9. Per-Protocol Analysis: Percent Body Fat and Lean Body Mass at Baseline and Changes After 6 Months
eTable 10. Per-Protocol Analysis: Anthropometry and Blood Pressure at Baseline and Changes After 3 and 6 Months
eTable 11. Cardiometabolic Risk Markers at Baseline and Changes After 3 and 6 Months
eTable 12. Changes in Cardiorespiratory Fitness (VO2peak and Treadmill Time) and Muscular Strength (Leg Press, Bench Press and Seated Row) in the Aerobic, Resistance, Combined and Control Groups
eTable 13. Per-Protocol Analysis: Changes in Energy Intake (3-Day Food Diaries) and Background Physical Activity (7-Day Pedometer Logs) in the Aerobic, Resistance, Combined and Control Groups
Customize your JAMA Network experience by selecting one or more topics from the list below.
Sigal RJ, Alberga AS, Goldfield GS, et al. Effects of Aerobic Training, Resistance Training, or Both on Percentage Body Fat and Cardiometabolic Risk Markers in Obese AdolescentsThe Healthy Eating Aerobic and Resistance Training in Youth Randomized Clinical Trial. JAMA Pediatr. 2014;168(11):1006–1014. doi:10.1001/jamapediatrics.2014.1392
Copyright 2014 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.
Little evidence exists on which exercise modality is optimal for obese adolescents.
To determine the effects of aerobic training, resistance training, and combined training on percentage body fat in overweight and obese adolescents.
Design, Setting, and Participants
Randomized, parallel-group clinical trial at community-based exercise facilities in Ottawa (Ontario) and Gatineau (Quebec), Canada, among previously inactive postpubertal adolescents aged 14 to 18 years (Tanner stage IV or V) with body mass index at or above the 95th percentile for age and sex or at or above the 85th percentile plus an additional diabetes mellitus or cardiovascular risk factor.
After a 4-week run-in period, 304 participants were randomized to the following 4 groups for 22 weeks: aerobic training (n = 75), resistance training (n = 78), combined aerobic and resistance training (n = 75), or nonexercising control (n = 76). All participants received dietary counseling, with a daily energy deficit of 250 kcal.
Main Outcomes and Measures
The primary outcome was percentage body fat measured by magnetic resonance imaging at baseline and 6 months. We hypothesized that aerobic training and resistance training would each yield greater decreases than the control and that combined training would cause greater decreases than aerobic or resistance training alone.
Decreases in percentage body fat were −0.3 (95% CI, −0.9 to 0.3) in the control group, −1.1 (95% CI, −1.7 to −0.5) in the aerobic training group (P = .06 vs controls), and −1.6 (95% CI, −2.2 to −1.0) in the resistance training group (P = .002 vs controls). The −1.4 (95% CI, −2.0 to −0.8) decrease in the combined training group did not differ significantly from that in the aerobic or resistance training group. Waist circumference changes were −0.2 (95% CI, −1.7 to 1.2) cm in the control group, −3.0 (95% CI, −4.4 to −1.6) cm in the aerobic group (P = .006 vs controls), −2.2 (95% CI −3.7 to −0.8) cm in the resistance training group (P = .048 vs controls), and −4.1 (95% CI, −5.5 to −2.7) cm in the combined training group. In per-protocol analyses (≥70% adherence), the combined training group had greater changes in percentage body fat (−2.4, 95% CI, −3.2 to −1.6) vs the aerobic group (−1.2; 95% CI, −2.0 to −0.5; P = .04 vs the combined group) but not the resistance group (−1.6; 95% CI, −2.5 to −0.8).
Conclusions and Relevance
Aerobic, resistance, and combined training reduced total body fat and waist circumference in obese adolescents. In more adherent participants, combined training may cause greater decreases than aerobic or resistance training alone.
clinicaltrials.gov Identifier: NCT00195858
Childhood obesity1,2 and physical inactivity3 are serious public health concerns. Obesity is associated with adverse health outcomes in youth,4 and 80% of obese adolescents become obese adults.5 Although exercise is recommended for obese adolescents, the optimal exercise prescription to reduce adiposity and its comorbidities is unclear.
We designed the Healthy Eating Aerobic and Resistance Training in Youth (HEARTY) trial to determine the effects of aerobic training, resistance training, or their combination on percentage body fat and cardiometabolic risk markers in previously inactive postpubertal overweight and obese adolescents. The primary outcome was percentage body fat measured by magnetic resonance imaging (MRI). Secondary outcomes included total body fat, anthropometric indexes of body fat distribution, and risk markers for cardiovascular disease and diabetes mellitus (DM).
This study was reviewed and approved by research ethics boards of the Children’s Hospital of Eastern Ontario and the Ottawa Hospital, Ottawa, Ontario, Canada. All participants (and parents for participants younger than 16 years) provided written informed consent. This was a randomized clinical trial with a parallel-group design. Details on the rationale, design, inclusion and exclusion criteria, and methods have been described previously.6 Outcome assessors (A.S.A. and P.P.) were blinded to the study group of participants.
Participants (eTable 1 in the Supplement) were postpubertal (Tanner stage IV or V)7,8 adolescents aged 14 to 18 years with body mass index (BMI) at or above the 95th percentile for age and sex (http://www.cdc.gov/growthcharts) or at or above the 85th percentile plus an additional DM or cardiovascular risk factor. Exclusion criteria included habitual exercise more than twice weekly for more than 20 minutes per session, DM, or any illness or disability rendering study exercise programs inadvisable or unfeasible. Routine school physical education classes were not an exclusion criterion. For participants taking medication likely to affect body composition (metformin, oral contraceptives, or stimulants), the dosage was required to have been stable during the previous 2 months and to remain unchanged throughout the trial.
Participants entered a run-in period that included supervised moderate-intensity exercise training 4 times weekly for 4 weeks. They performed 15 to 30 minutes of aerobic exercise at 65% of measured maximum heart rate and 1 to 3 sets of 15 repetitions for each of 7 resistance exercises at each session. Details of exercise programs (eTable 2 in the Supplement) were described previously.6
To qualify for randomization, participants needed to attend at least 13 of 16 prescribed exercise sessions (≥81.3% adherence) during the run-in period. The exercise specialist used a telephone-based central randomization program (IVRS, VBvoice, version 5.3; Pronexus) and informed participants of their group assignments, allowing the research coordinator (P.P.) to remain blinded to the study group of participants. Participants (N = 304) were randomized into the following 4 groups for 22 weeks: aerobic training (hereinafter the aerobic group, n = 75), resistance training (hereinafter the resistance group, n = 78), combined aerobic and resistance training (hereinafter the combined group, n = 75), or nonexercising control (hereinafter the control group, n = 76). Randomization was in permuted blocks, stratified by sex and degree of overweight (85th to 94th percentile vs ≥95th percentile).
Participants in all 4 groups received counseling at baseline, 3 months, and 6 months by a registered dietitian to promote healthy eating, with a daily energy deficit of 250 kcal. In addition to the dietary counseling, the 3 exercising groups were asked to attend gymnasiums 4 times weekly. Exercise training progressed gradually in duration and intensity. The aerobic group exercised on treadmills, elliptical machines, or bicycle ergometers. Heart rate monitors (Polar FS1; Polar Electro Oy) were used to adjust workloads to achieve target heart rates. Participants gradually progressed in exercise duration (from 20 to 45 minutes per aerobic exercise session) and intensity (from 65% to 85% of maximum heart rate). The resistance group performed 7 exercises using weight machines or free weights, progressing from 2 sets of 15 repetitions at moderate intensity to 3 sets of 8 repetitions at the maximum resistance that could be moved 8 times (8-RM). The combined group performed the full aerobic training program plus the resistance training program during each session. Details of the exercise programs are in eTables 3, 4, and 5 in the Supplement.
Exercise training took place at 6 community-based facilities in Ottawa (Ontario) and Gatineau (Quebec), Canada. Exercise was supervised by personal trainers twice weekly during the run-in period, weekly from randomization to 3 months, and biweekly from 3 to 6 months. Personal trainers monitored attendance and exercise progression by reviewing sign-in sheets and exercise logs. Intervention adherence was calculated as the total number of exercise sessions the participant attended divided by the total number of sessions prescribed.
Body composition was assessed by MRI with a 1.5-T system (EchoSpeed, signal 11 version; GE Medical Systems) at baseline and 6 months. Participants lay prone for whole-body cross-sectional images using protocols by Ross and colleagues.9,10 The MRIs were analyzed using a software program (Slice-O-Matic, version 4.3; Tomovision).
Weight, height, BMI, waist and hip circumferences, and blood pressure were measured as described previously.6 Participants were asked to wear pedometers (DIGI-WALKER SW-700; Yamax Corporation) and maintain step-count logs for 7 days at baseline and 6 months to assess background physical activity. Energy intake was assessed using 3-day food diaries and analyzed using food composition analysis software (Food 324 Processor SQL 2006; ESHA Research) at baseline, 3 months, and 6 months.
Fasting insulin, glucose, glycated hemoglobin, and lipid levels were measured at baseline, 3 months, and 6 months as described previously.6 Samples were obtained 2 to 10 days after the last exercise session to avoid potentially confounding acute effects of exercise. A 75-g oral glucose tolerance test (fasting and 2-hour postload glucose) was performed at baseline and 6 months.
Peak oxygen consumption (V̇o2peak, in milliliters of oxygen per kilogram of body weight per minute) was measured at baseline and 6 months by indirect calorimetry using a metabolic cart (MOXUS Modular Metabolic System; AEI Technologies) during a Balke-Ware treadmill test.11 Criteria for test termination included volitional exhaustion or the participant’s desire to stop. Treadmill time was defined as the time from the start of the test until the time the participant reached V̇o2peak.
An 8-RM test was performed on leg press, bench press, and seated row machines to measure muscular strength at baseline, 3 months, and 6 months. The 8-RM test measured the maximum weight that could be lifted 8 times while maintaining proper form.
Research staff (including P.P.) reported all directly observed adverse events and those spontaneously reported by the participants. Participants were also questioned about adverse events at each study visit.
To assess the effects of exercise training modality on changes in percentage body fat (primary outcome) over time, assuming a moderate effect size of 0.6 SD for the least-powered planned comparison, we calculated a sample size of 248 participants (62 per group) completing the intervention to allow 80% power for each of 4 prespecified comparisons (aerobic group vs control group, resistance group vs control group, combined group vs aerobic group, and combined group vs resistance group) tested simultaneously (overall α = .05 overall). We randomized 304 participants to allow for 18% to 20% dropout, similar to rates observed in previous exercise trials among obese youth12 and adults with type 2 DM.13 We performed analyses on an intent-to-treat basis that included all randomly allocated participants (including those who later withdrew). For the primary analysis, we used linear mixed-effects modeling for repeated measures over time using percentage body fat as the dependent variable and effects for time, group (aerobic group, resistance group, combined group, or control group), and time by group interaction, with age and sex as covariates and an unstructured covariance matrix. Within the mixed model, we calculated 95% CIs and P values for 4 prespecified intergroup contrasts and for change in percentage body fat within each group over time. We used the same procedure for other continuous outcome variables, including 3-month values when available in addition to baseline and 6-month values. For all linear mixed-model analyses, we examined distributions of residuals and used transformations to achieve normality when necessary. We also examined the effects of the exercise intervention in per-protocol analyses, restricted to participants with complete baseline and 6-month data and at least 70% adherence throughout the exercise intervention, following the same procedures as in the intent-to-treat analyses. We used statistical software (SAS, version 9.2; SAS Institute) for all analyses.
Recruitment began March 1, 2005, and closed April 30, 2010. The final follow-up visit was in May 2011. Of 358 participants who entered the run-in phase, 304 (84.9%) were randomized. Reasons for prerandomization withdrawal included time demands, lack of interest, moving to a different town or city, or a medical condition (Figure). Seven participants (1 from the aerobic group, 3 from the resistance group, 1 from the combined group, and 2 from the control group) were excluded from MRI analyses because images were technically inadequate. At 6 months, 54 participants in the aerobic group, 55 participants in the resistance group, 55 participants in the combined group, and 57 participants in the control group completed MRI; an additional 6 completed all 6-month measures except MRI.
Table 1 and eTable 6 in the Supplement list baseline characteristics and potentially weight-altering medications, respectively. No significant intergroup baseline differences were observed. Most participants (282 of 304 [92.8%]) were obese (BMI ≥95th percentile for age and sex).
From baseline to 26 weeks, the median exercise training adherence was 62% (interquartile range, 36%-81%) in the aerobic group, 56% (interquartile range, 37%-75%) in the resistance group, and 64% (interquartile range, 39%-75%) in the combined group, with no significant differences between groups. Seventy-five participants (24.7%) withdrew between randomization and 6 months, including 18 (24.0%) from the aerobic group, 21 (26.9%) from the resistance group, 17 (22.7%) from the combined group, and 19 (25.0%) from the control group. Reasons for withdrawal are shown in the Figure.
Details of adverse events are listed in eTable 7 and eTable 8 in the Supplement. Two participants, both in the aerobic group, withdrew because of adverse events unrelated to study interventions, including hip fracture from a fall at school (1 participant) and a forearm fractured when playing football (1 participant). Overall, adverse events occurred in 49 of 228 exercise group participants (21.5%) and in 18 of 76 controls (23.7%). Musculoskeletal injury or discomfort requiring exercise program modification or temporary activity restriction occurred in 27 of 228 exercise group participants (11.8%) and in 2 of 76 controls (2.6%). Adverse events definitely, probably, or possibly related to the intervention occurred in 18 of 228 exercise group participants (7.9%) and in 1 of 76 controls (1.3%). Almost all related adverse events involved musculoskeletal injury or discomfort. Two medical adverse events were possibly related to the intervention, including postexertional malaise (1 participant) and noncardiac chest pain (1 participant).
Body fat and lean body mass results are listed in Table 2. Percentage body fat did not change in the control group (−0.3; 95% CI, −0.9 to 0.3) but decreased in all exercising groups, by −1.1 (95% CI, −1.7 to −0.5) in the aerobic group (P = .06 vs control group), −1.6 (95% CI, −2.2 to −1.0) in the resistance group (P = .002 vs control group), and −1.4 (95% CI, −2.0 to −0.8) in the combined group (not significantly different from the aerobic group or resistance group). In per-protocol analyses (≥70% adherence) (eTable 9 in the Supplement), percentage body fat changes in the combined group (−2.4, 95% CI, −3.2 to −1.6) were greater than those in the aerobic group (−1.2; 95% CI, −2.0 to −0.5; P = .04 vs combined group) but not significantly different from changes in the resistance group (−1.6; 95% CI, −2.5 to −0.8).
Table 3 lists the effects of the exercise interventions on anthropometric indexes. Waist circumference changes were −0.2 (95% CI, −1.7 to 1.2) cm in the control group, −3.0 (95% CI, −4.4 to −1.6) cm in the aerobic group (P = .006 vs control group), −2.2 (95% CI, −3.7 to −0.8) cm in the resistance group (P = .048 vs control group), and −4.1 (95% CI, −5.5 to −2.7) cm in the combined group. In per-protocol analyses (eTable 10 in the Supplement), BMI decreased in the aerobic group vs the control group, and decreases were greater in the combined group vs the aerobic group and in the combined group vs the resistance group. Waist circumference decreased in the aerobic group and the resistance group vs the control group and decreased significantly more in the combined group compared with the aerobic group.
Results for cardiometabolic risk factors are listed in eTable 11 in the Supplement. Fasting insulin and triglycerides levels were log transformed for analysis; geometric means are presented. No significant intergroup differences were observed in the levels of fasting insulin, fasting or 2-hour glucose, triglycerides, glycated hemoglobin, high-density lipoprotein cholesterol, low-density lipoprotein cholesterol, or total cholesterol.
Treadmill time increased in all 3 exercise groups (eTable 12 in the Supplement). Increases in V̇o2peak in the aerobic group were greater than those in the control group. Although within-group V̇o2peak and treadmill time increased in the combined group, these changes were not significantly different from those in the other groups. Treadmill time increased in the aerobic group vs the control group.
The aerobic group and the resistance group showed larger increases in leg press than the control group (eTable 12 in the Supplement), and the combined group showed larger increases in leg press, bench press, and seated row machines compared with the aerobic group. The resistance group showed greater increases in bench press than the control group. Changes in muscular strength in the combined group did not differ from changes in the resistance group on leg press, bench press, or seated row machines.
Within-group energy intake (3-day food diaries) decreased similarly in all groups. No significant between-group differences were observed in energy intake or background physical activity (pedometer logs, excluding exercise sessions that were part of the intervention). Background physical activity data were limited because many participants did not wear pedometers or complete pedometer logs as instructed: per-protocol analysis results are listed in Table 4 and in eTable 13 in the Supplement.
To our knowledge, this is the first randomized trial to date comparing the effects of aerobic training, resistance training, and their combination on body composition and cardiometabolic risk factors in a large sample of obese adolescents. The primary findings were that modest but clinically significant reductions in percentage body fat can be achieved through aerobic, resistance, or combined exercise training in obese adolescents. In per-protocol analyses, combined aerobic and resistance training produced greater decreases in percentage body fat, waist circumference, and BMI than aerobic training alone.
The modest decreases in percentage body fat in the HEARTY trial were comparable to those observed by Lee et al,14 who examined the effects of thrice-weekly aerobic or resistance training vs a nonexercising control group during a 3-month period. They showed decreases in aerobic-only results (−2.6%) and resistance-only results (−2.5%) vs controls. Although exercise prescriptions in the study by Lee et al were similar to those of the HEARTY trial, Lee and colleagues showed greater reductions in percentage body fat in both exercising groups, possibly attributable to higher exercise adherence (99%) or because their participants received some financial compensation. Lack of financial compensation may have contributed to lower training adherence (66%) in the HEARTY trial.
Compared with the control group, we found decreases in waist circumference in all 3 exercise groups, and per-protocol analyses showed greater decreases in the combined group vs the aerobic group. Our results agree with previous smaller studies showing decreases in abdominal fat through aerobic training,14 resistance training,14 and combined aerobic and resistance circuit training15,16 performed 2 to 3 times weekly compared with nonexercising controls. Abdominal fat accumulation is associated with increased cardiometabolic risk.17 Attenuating increases in abdominal fat during adolescence could confer important cardiometabolic protection because each additional year of abdominal obesity is associated with a 4% greater risk of developing DM.18
We found no changes in glucose or lipid levels, possibly due to mostly normal baseline values. Potential participants who had DM at baseline were excluded from our study, and 86.5% of our participants had normal baseline fasting and 2-hour postload glucose levels, leaving little room for improvement. Two smaller randomized clinical studies by Lee et al14 and Shaibi et al19 showed that resistance exercise 2 to 3 times per week increased insulin sensitivity by 27% (euglycemic clamp) in black, white, and mixed racial/ethnic obese adolescent boys and by 45% (frequently sampled intravenous glucose tolerance test) in obese Latino adolescent boys, respectively. Lee and colleagues14 found no changes in insulin sensitivity in the aerobic group. The racial/ethnic background and sex of HEARTY participants (72.0% white and 70.1% female) differed from previous research conducted with male Latino youth,15,16,19 who are more likely to be insulin resistant than white youth independent of body fat.20
Although within-group V̇o2peak and treadmill time increased in the aerobic group and the combined group, only the increases in the aerobic group were significantly greater than those in controls. In the Diabetes Aerobic and Resistance Exercise (DARE) trial,21 with a design similar to that of the HEARTY trial in adults with type 2 DM, findings were similar: aerobic training and combined training each increased aerobic fitness, but the aerobic-only intervention did so to a slightly greater extent. Similar to the present trial, DARE also reported no differences in fat-free mass changes between groups but observed decreases in body weight and fat mass in the aerobic and combined training groups.13 The HEARTY trial findings support that moderate-intensity to high-intensity aerobic or combined exercise training for at least 3 sessions per week improves cardiorespiratory fitness in overweight and obese adolescents.
All 3 types of exercise training (aerobic, resistance, and combined) increased lower body muscular strength. The greatest increase was observed in the resistance group. Adherence rates were similar, suggesting that adding aerobic training might attenuate leg strength development compared with resistance training alone. Nonetheless, combined training increased both V̇o2peak and muscular strength.
We excluded adolescents with DM, so our results do not necessarily apply to them. Our results may not be generalizable to unsupervised exercise. The estimated intervention cost per participant in 2010 was $1137.94 (Canadian dollars), including gymnasium membership ($30.00 per month times 6 months equals $180.00), personal training ($33.11 per hour times 20 sessions equals $662.20), and counseling by a dietitian ($49.29 per hour times 6 hours equals $295.74). Costs would decrease over time, assuming the required frequency of personal trainer sessions would decrease.
The HEARTY trial showed that aerobic training, resistance training, and their combination decreased percentage body fat in obese adolescents. In participants adhering to the exercise protocol, combined aerobic and resistance exercise training tended to be superior to aerobic training alone in decreasing percentage body fat, waist circumference, and BMI. Combined training might have additive effects through greater exercise volume or combinations of unique effects of aerobic training (improvement in the oxidative metabolism–dependent energy system, qualitative changes in skeletal muscle fiber type, metabolic capacity, and cardiorespiratory fitness)22 and resistance training (quantitative changes in skeletal muscle mass or fiber diameter and increased muscular strength).23 Adolescents who want to maximize the effect of exercise on these variables should ideally perform both aerobic and resistance exercise, but significant benefit can be achieved through either type of exercise alone.
Accepted for Publication: June 24, 2014.
Corresponding Author: Ronald J. Sigal, MD, MPH, Faculties of Medicine and Kinesiology, University of Calgary, 1820 Richmond Rd SW, Room 1898, Calgary, AB T2T 5C7, Canada (firstname.lastname@example.org).
Published Online: September 22, 2014. doi:10.1001/jamapediatrics.2014.1392.
Author Contributions: Drs Sigal and Alberga had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Sigal, Goldfield, Prud’homme, Hadjiyannakis, Gougeon, Wells, Kenny.
Acquisition, analysis, or interpretation of data: Sigal, Alberga, Goldfield, Prud’homme, Gougeon, Phillips, Tulloch, Malcolm, Doucette, Ma, Kenny.
Drafting of the manuscript: Sigal, Alberga, Kenny.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: Doucette, Wells, Ma.
Obtained funding: Sigal, Goldfield, Prud’homme, Hadjiyannakis, Gougeon, Wells, Kenny.
Administrative, technical, or material support: Sigal, Alberga, Phillips, Malcolm, Kenny.
Study supervision: Sigal, Alberga, Goldfield, Prud’homme, Kenny.
Conflict of Interest Disclosures: None reported.
Funding/Support: The HEARTY trial was supported by grant MCT-71979 from the Canadian Institutes of Health Research. Dr Sigal is supported by a Health Senior Scholar award from Alberta Innovates–Health Solutions and was previously supported by a Research Chair from the Ottawa Hospital Research Institute during part of this trial. Dr Alberga was supported by a Doctoral Student Research Award from the Canadian Diabetes Association. Dr Goldfield was supported by a New Investigator Award from the Canadian Institutes of Health Research for part of the trial and subsequently by an Endowed Research Scholarship from the Children’s Hospital of Eastern Ontario Volunteer Association Board. Dr Kenny was supported by a University Research Chair from the University of Ottawa.
Role of 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: We thank the HEARTY trial participants. Krista Hind, BSc (deceased), Bruno Lemire, PhD, Marta Wein, BSc, Kim Robertson, BSc, Kim Fetch, BSc, Brittany Hanlon, MHA, Jane Yardley, PhD, Nadia Balaa, BSc, Karen Lopez, BSc, Pamela Martino, MSc, Kim Morin, BSc, Colleen Gilchrist, BSc, RD, Pascale Messier, BSc, RD, Kelley Phillips, MA, and students in the School of Human Kinetics, University of Ottawa, who contributed to study coordination, exercise training, and evaluation of study participants. Robert Ross, PhD (Queens University, Kingston, Ontario, Canada), Alison Bradshaw, MSc, and Jennifer Kuk, PhD (York University, Toronto, Ontario, Canada), and Yves Martel, PhD (Tomovision, Magog, Quebec, Canada) assisted with training and provided ongoing advice on body composition analysis. The Ottawa-Carleton Regional YMCA/YWCA (Ottawa, Ontario, Canada), RA Centre (Ottawa, Ontario, Canada), Children’s Hospital of Eastern Ontario, and Nautilus Plus and MRI Plus (both in Gatineau, Quebec, Canada) collaborated throughout the trial. Robert Ross, PhD, received financial compensation beyond his usual salary.
Correction: This article was corrected on June 25, 2015, to fix an SD value in Table 1.