PEDIATRIC REVIEW Efficacy of exercise for treating ... - Nature

International Journal of Obesity (2006) 30, 1027–1040 & 2006 Nature Publishing Group All rights reserved 0307-0565/06 $30.00 www.nature.com/ijo

PEDIATRIC REVIEW Efficacy of exercise for treating overweight in children and adolescents: a systematic review E Atlantis1, EH Barnes2 and MA Fiatarone Singh3,4,5 1 School of Exercise and Sport Science, Faculty of Health Sciences, University of Sydney, NSW, Australia; 2National Health and Medical Research Counsel: Clinical Trials Centre, University of Sydney, Australia; 3Faculty of Medicine and Faculty of Health Sciences, School of Exercise and Sport Science, University of Sydney, NSW, Australia; 4Jean Mayer USDA Human Nutrition Center on Aging at Tufts University, Boston, MA, USA and 5Hebrew SeniorLife, Boston, MA, USA

Background: Overweight prevalence among children/adolescents is increasing, while adult obesity may potentially cause a decline in life expectancy. More exercise is uniformly recommended, although treatment efficacy remains unclear. Objective: To determine the efficacy of exercise alone for treating overweight in children/adolescents. Design: A systematic review and meta-analysis of randomized trials published in English were completed following multiple database searches performed on December 10, 2004. Studies of isolated or adjunctive exercise/physical activity treatment in overweight/obese children or adolescents which reported any overweight outcome were included. Literature searches identified 645 papers which were manually searched, of which 45 were considered for inclusion, of which 13 papers which reported 14 studies were included (N ¼ 481 overweight boys and girls, aged B12 years). Two reviewers independently identified relevant papers for potential inclusion and assessed methodological quality. Principal measures of effects included the mean difference (MD) (between treatment and control groups), the weighted MD (WMD), and the standardized MD (SMD). Results: Few studies were of robust design. The pooled SMD was 0.4 (0.7, 0.1, P ¼ 0.006) for percent body fat, and 0.2 (0.6, 0.1, P ¼ 0.07) for central obesity outcomes, whereas the pooled WMD was 2.7 kg (6.1 kg, 0.8 kg, P ¼ 0.07) for body weight, all of which favored exercise. Pooled effects on body weight were significant and larger for studies of higher doses, whereas nonsignificant and smaller effects were seen for studies of lower doses of exercise (155–180 min/weeks vs 120–150 min/ weeks). Conclusions: Based on the small number of short-term randomized trials currently available, an aerobic exercise prescription of 155–180 min/weeks at moderate-to-high intensity is effective for reducing body fat in overweight children/adolescents, but effects on body weight and central obesity are inconclusive. Recommendations for future study designs are discussed. International Journal of Obesity (2006) 30, 1027–1040. doi:10.1038/sj.ijo.0803286; published online 14 March 2006 Keywords: youth; meta-analysis; body fat; body weight; BMI; waist circumference

Introduction Global trends over the past two decades show the prevalence of overweight/obesity among children and adolescents to be increasing at an alarming rate (twofold).1–5 In adults, the morbidity and mortality burden of obesity is well documented. Of major concern is the growing body of evidence reporting overweight-related morbidity in childhood and adolescence, both in the psychosocial6–8 and physiological

Correspondence: Dr E Atlantis, School of Exercise and Sport Science, Faculty of Health Sciences, University of Sydney, PO Box 170, Lidcombe, NSW 1825, Australia. E-mail: [email protected] Received 31 August 2005; revised 11 January 2006; accepted 19 January 2006; published online 14 March 2006

domain.9–14 Recently, the prevalence of ‘metabolic syndrome’ was reported to be 39 and 50% in moderate and severely obese children and adolescents, respectively, indicating higher morbidity burden with greater obesity severity.10 Unless this trend is curtailed, obesity-related disease, disability and death worldwide is projected to increase well into the 21st century,15 whereas life expectancy at birth and at older ages may begin to decline by the middle of the 21st century.16 Fundamentally, the etiology of overweight in children and adolescents, as in adults, is attributable to energy (caloric) intake in excess of energy expenditure. Genetic, environmental, socio-cultural, and family characteristics have been identified as key influences on energy expenditure and dietary intake, and subsequent overweight prevalence.17–19 Physical activity and resting metabolic rate are obvious targets for interventions aimed at inducing energy deficits

Efficacy of exercise for treating overweight in youth E Atlantis et al

1028 for the treatment and prevention of obesity, as they collectively account for most of total daily energy expenditure. Thus, interventions that increase physical activity and/or decrease time spent in sedentary activities may be important methods to counter the increasing burden of overweight in youth. Prescribed exercise is one such intervention, as it increases energy expenditure, preserves fat-free mass (including bone and skeletal muscle), and has been shown to be effective in decreasing body fat in adults.20 In contrast to adult studies, the evidence for treatment efficacy of exercise in overweight children and adolescents is not yet established. Previous reviews have not critically assessed the quality of studies included, all of which also included nonrandomizedcontrolled trials, or studies which did not isolate the effects of exercise, or studies in normal weight cohorts.21–27 In this systematic review and meta-analysis, we critically examined the quality of studies included and their effects on overweight outcomes in overweight children and adolescents, and we focused exclusively on randomized-controlled trials (RCTs) in which the isolated effects of exercise could be determined.

Materials and method Search protocol Systematic database searches for abstracts and full-length manuscripts were performed on Ovid MEDLINEs (MEDL) (1966–2004, November week 3), Ovid MEDLINEs Daily Update, November 17, 2004), PREMEDLINE (most recently published, December 4, 2004), CAB ABSTRACTS (CABA, 1973–2004, October), PsycINFOs (PSYC) (1967–2004, November week 5), SPORTDiscus (1830–2004, November), and Evidence Based Medicine Reviews (Cochrane Central Register of Controlled Trials, 4th Quarter, 2004) on December 10, 2004. Firstly, four categorical searches were conducted using the following ‘keywords’: (1) ‘weight training,’ ‘strength training,’ ‘exercise,’ ‘exercise therapy,’ ‘weightlifting,’ ‘physical fitness,’ or ‘physical activity’ (yielded 293 106 studies), (2) ‘child’, ‘adolescent’, or ‘pediatrics’ (yielded 1 795 359 studies), (3) ‘body fat,’ ‘body weight,’ or ‘body composition’ (yielded 325 759 studies) and (4) ‘overweight,’ ‘obese’ or ‘obesity’ (yielded 53 875 studies). Secondly, categories 1–4 were combined to obtain a final total of 625 studies. Finally, all 625 titles and or abstracts of relevant studies were manually searched for potential inclusion. Bibliographies of these studies were also searched to extract further studies from those already retrieved, including a number of recent reviews.21,23,26,27

Inclusion and exclusion criteria All studies published in English (because funding was not available for translation) that met the following criteria were included in this review: (1) studies were RCTs, (2) cohorts International Journal of Obesity

were of children or adolescents (aged o18 years) defined as being overweight/obese, (3) pre- and post-test or change in any overweight outcome was reported, (4) at least one exercise or physical activity treatment arm was investigated either in isolation or as an adjunct to an alternative treatment simultaneously prescribed to the control/comparison group (e.g., exercise plus caloric intake of 1000 kcal/days vs caloric intake of 1000 kcal/days). Studies which reported previously published data, or which included normal weight subjects in either treatment and/or comparison/ control groups were excluded, as well as nonrandomized trials. Studies of behavioral interventions which offered either education and/or advice on increasing physical activity, or structured supervised/unsupervised physical activity exercise programs were considered for inclusion. More than eight experts in the field were contacted via e-mail to enquire whether they knew of any ‘published/ unpublished RCTs,’ to identify additional studies for potential inclusion.

Validity assessment Two reviewers independently assessed studies to determine whether they met all inclusion criteria (EA and MFS). Studies were defined as RCTs if the allocation of subjects to treatment and control/comparison groups was reported to have been randomized. Studies which did not report randomization or group assignment methodology were assumed to have used a nonrandomized protocol. For methodology and quality assessment, we focused on: (1) technique used for assessment of overweight outcomes, (2) study design characteristics, (3) exercise treatment characteristics and (4) subject characteristics. Data extraction Two reviewers independently identified relevant studies for potential inclusion, whereas EA conducted the database searches and extraction of all relevant data from studies included. Quantitative data synthesis Outcomes considered for data extraction and synthesis include any overweight outcome assessed by imaging technology (computed tomography and magnetic resonance imaging), percent body fat by any method (i.e., dual energy X-ray absorptiometry (DXA), bioelectrical impedance analysis, hydrodensitometry, or predictions from skin folds), skinfold measures (mm), percent over ideal body weight (according to height/weight charts), and body weight (kg/lbs), body mass index (kg/m2), and waist girths. Principle measures of exercise effects include the mean difference (MD) between treatment and control groups, the standardized MD (SMD) for each study, the weighted MD (WMD) and the pooled SMD for summary effects, including 95% confidence intervals (95% CIs).


1029 Statistical analysis For meta-analysis, we pooled and weighted studies according to the inverse variance method using both the fixed effect model and the random effects model where appropriate.28 Individual studies which reported outcomes on the same scale were combined to estimate the WMD between treatment and control groups. Standardized MDs (i.e., effects sizes) were calculated using Glass’s method.28 To explore the potential for heterogeneity between studies we firstly carried out a meta-analysis using the fixed effect model to combine the WMDs and SMDs, followed by Q-tests for heterogeneity. Statistical significance of Q-tests was obtained using w2 analyses. Linear regressions were used to determine the strength of association between dose of exercise prescribed and size of effects (MD and SMD), including 95% CIs. P-values o0.05 were considered statistically significant. Where necessary, standard errors were converted to s.d. to permit SMD calculations. Only study completers used in weighting for meta-analysis were included in the sample sizes reported. All data were analyzed using Microsoft Excel 2000 (Microsoft Corp., Redmond, WA, USA) and Statistical Package for Social Sciences (SPSS for Windows, version 11.5 SPSS Inc., USA).

Missing data Pre-test error data for means were used to calculate post-test s.d. from two studies which did not report measurement errors at post-test.29,30 In one of these studies, the largest s.e. was selected from those reported in four subgroups (by gender), and was used to calculate a conservative SMD.30

Manually searched N = 625 Potentially relevant papers retrieved N = 45

Excluded papers Non-randomized Non-overweight cohorts Non-overweight control/ comparison group Mixed weight cohorts Previously published data

N = 32 n=5 n=4 n=6 n=8 n=9

Included papers N = 13 Included studies N = 14

Aerobic exercise n=5

Aerobic exercise + diet n=5

Aerobic + weight-training exercise + diet n=4

Figure 1 Flow and characteristics of studies included/excluded for review.

Studies included/excluded The process of study inclusion for narrative review and metaanalysis is presented in Figure 1. Of 45 potentially relevant papers considered, we excluded 32 for the following reasons: eight were conducted in mixed weight cohorts,31–37 four were specifically conducted in normal weight cohorts,38–41 six studies compared overweight subjects with normal weight control subjects,42–47 five studies were nonrandomized trials48–52 and nine studies53–61 published data which were previously published, in three studies already included in this review.51,62,63

studies. Among these 14 studies, two compared the effects of aerobic exercise with untreated controls,29,62 one study compared the effect of a device which permitted television viewing contingent to exercise (TV cycling) with that of a control group which received a similar device but differed in that television viewing was not contingent to exercise,65 one investigated the effects of ‘extra gymnastics’ sessions/weeks reported to entailed 45 min of intense physical activity performed in a small group setting compared to untreated controls,66 whereas the remaining nine studies compared the effects of various dietary/behavioral treatments with/without aerobic exercise30,67–71 or combined aerobic and weighttraining exercise.63,64,72 The mean (7s.d.) treatment duration of these studies, at post-test, was 16 (77) weeks. Only two studies investigated long-term treatment effects beyond the intervention phase, one of which reported follow-up data at 26 and 52 weeks,67 and another at 52 weeks.63 Eight studies were conducted in the US,29,30,62,65,67,68,70,71 one in Australia,69 one in Austria,72 one in Canada,64 one in Sweden66 and one in Hong Kong.63

Treatment/intervention All studies were of supervised exercise. One study prescribed exercise in a clinical laboratory setting,63 whereas the remainder used various outdoor/indoor settings. One study compared the effects of two different exercise plus behavioral treatments with that of a control group which received the behavioral treatment alone (three arm study) providing two datum points,64 and was therefore included as two separate

Cohorts Of 14 studies, one was in boys only,66 one was in girls only,67 11 were in boys and girls,29,30,62–65,68,70–72 whereas gender characteristics were not reported in another.69 In total, the 14 studies pooled yielded a sample of 481 boys and girls, and the mean age reported in nine studies was 10.971.5 years, while age ranges of 13–16 years was reported in one study,30 of 8–12 years in one study,67 and of 12–15 years in two

Results

International Journal of Obesity


1030 studies (two treatments from one study),64 while another did not report age (mean or range).69 Sample sizes were reduced for review and meta-analysis from those reported in three studies68,70,71 to omit subjects from non-treated control groups which were not included in treatment contrasts, in one study to omit subjects in the additional exercise treatment arm which was already considered as a separate study,64 and in one study69 to omit normal weight subjects who were allocated to the no-treatment control arm. No study reported cohorts to have been targeted for clinical disorders during the recruitment of overweight/obese subjects. No study reported outcomes for boys and girls separately. Characteristics of these 14 studies are presented in Table 1.

Summary effects Percent body fat. The SMD of each study is presented in Figure 2. Significant heterogeneity in SMDs for percent body fat was found between the nine studies (n ¼ 369) pooled (Q ¼ 15.5, P ¼ 0.05). The pooled SMD was 0.4 (95% CI: 0.7 to 0.1, P ¼ 0.006) using the random effects model, and 0.3 (95% CI: 0.6 to 0.1, Po0.001) using the fixed effect model, following 18 (78) weeks treatment. The dose of exercise (min/session multiplied by frequency/weeks) reported was B169 (725) min/weeks. Body weight. The MD of each study is presented in Figure 3. Significant heterogeneity in MDs for body weight was found between the 11 studies (n ¼ 334) pooled (Q ¼ 17.9, P ¼ 0.07). The pooled estimate of effect (WMD) on body weight was 2.7 kg (95% CI: 6.1 to 0.8 kg, P ¼ 0.07) using the random effects model and 1.0 kg (95% CI: 3.2 to 1.3 kg, P ¼ 0.2) using the fixed effect model following 14 (75) weeks treatment. The dose of exercise reported was B135 (742)min/weeks. Central obesity. The SMD of each of the four studies, which reported central obesity-related outcomes, is presented in Figure 4. There was no significant heterogeneity between these four studies (n ¼ 156), while the pooled SMD was found to be 0.2 (95% CI: 0.6 to 0.1, P ¼ 0.07) using the fixed effect model following 16 (713) weeks treatment. The dose of exercise reported was B125 (765)min/weeks. Sensitivity analyses. The above meta-analysis was repeated while systematically removing subsets of studies of similar design and/or characteristics to evaluate the robustness of summary effects after pooling. Percent body fat. After removing three studies of low exercise frequency (o3 days/weeks, B153725 min/weeks) the pooled SMD was increased to 0.6 (95% CI: 0.8 to 0.3, Po0.001) for studies which prescribed a higher dose of exercise (B177723 min/weeks). After removing three studies which did report changes in dietary intake, and three International Journal of Obesity

exercise only studies the pooled SMD remained relatively unchanged (0.4 (95% CI: 0.9 to 0.0, P ¼ 0.02) and 0.4 (95% CI: 0.8 to 0.1, P ¼ 0.005), respectively). Conversely, after removing three studies which did not report exercise compliance or change in physical activity levels, the pooled WMD was decreased to 0.3 (95% CI: 0.7 to 0.2, P ¼ 0.1). Finally, after removing two studies which included weight training, the pooled SMD was increased to 0.5 (95% CI: 0.8 to 0.1, P ¼ 0.003). Body weight. After removing five studies of low exercise frequency (o3 days/weeks, B117746 min/weeks) the WMD was increased to 4.9 kg (95% CI: 9.1 to 0.7 kg, P ¼ 0.01) for studies which prescribed a higher dose of exercise (B156725 min/weeks), whereas the WMD was decreased to 1.2 kg (95% CI: 6.0 to 3.7 kg, P ¼ 0.3) after pooling the six studies of low exercise frequency (lower dose of B117746 min/weeks). After removing two studies which did report changes in dietary intake, and two exercise only studies, the WMD was increased to 5.1 kg (95% CI: 8.6 to 1.6 kg, P ¼ 0.002), and to 3.6 kg (95% CI: 8.1 to 0.9 kg, P ¼ 0.06), respectively. Conversely, after removing three studies, which did not report exercise compliance or change in physical activity levels, the WMD was decreased to 2.3 kg (95% CI: 6.8 to 2.1 kg, P ¼ 0.2). Finally, after removing four studies which included weight-training, the pooled WMD was increased to 3.1 kg (95% CI: 6.1 to 0.1 kg, P ¼ 0.02). Dose–response effects. No significant dose–response relationships were found between the volume of exercise prescribed (min/weeks), or study duration (weeks), or total dose completed (min/weeks study duration (weeks) product), and size of effects for percent body fat, body weight, and central obesity outcomes. However, there was a trend for additional weeks of study duration to be associated with greater reductions in body weight (Figure 5). Each additional week of study duration was associated with greater reductions in body weight of 0.5 kg (1.1 to 0.2 kg, P ¼ 0.1), which explained 22% of the variation in the WMD. Synthesis of studies by scale of assessment. Studies were pooled according to method of assessment used in data collection and the WMDs or MDs generated are presented in Table 2. Quality assessment (1) Technique used for assessment of overweight outcome: Techniques used for assessment of overweight outcomes were broad ranging, and in no study were coefficients of variation reported; although such criticism is less directed at studies that used DXA body scanning due to safety concerns regarding repeated measures in the young. For central obesity measures, only one study used precise imaging technology30 whereas three studies used anthropometry.63,64 In no study were the assessors reported to have been blinded to subjects’ group assignment during data collection.

Table 1

Characteristics of 14 exercise studies included

Author, year (country)Ref

Sample (no., mean age (s.d.) or range in years, %boys)

Criteria used to define overweight/ obesity

Treatment/ intervention

Control/comparison condition

Change/compliance in EX/PA reported?

Dietary assessment?

Overweight outcome

Blomquist, 1965 (Sweden)66

43 T: 9.2 (NR) years C: 9.3 (NR) years Boys 100%

Body weight X2 s.d. of the mean for height

No treatment

Yes (improved aerobic fitness, PWC 170, T4C, +75% attendance)

Yes (T: 255 kcal/ days, vs C: 89 kcal/ day, via 4 24 h recalls

Skin folds (mm), body weight

17

Epstein, 1985 (US)67

23 8–12 years Boys 0%

Body weight X20% of ideal weight for height and age

Diet-BMa

Yes (improved aerobic fitness, PWC 170, T4C)

No

% over ideal body weight, body weight

8

Becque, 1988 (US)68

Girls: 12.8 (0.3) years Boys 42%

Diet-BMb

Yes (improved VO2max, NS)

No

Hydrodensitometry (% BF), body weight

20

Hills, 1988 (Australia)69

20 Children (age NR) Boys %NR

Body weight plus triceps skin fold 475th percentile for age and sex Body weight 495th percentile for age plus BMI 425 kg/m2

Diet: education/ counseling

No

No

Skin folds (mm), body weight

16

Katch, 1988 (US)70

27 T: 12.5 (0.6) years, C: 12.8 (0.3) years Boys 72% 50 12.6 (3) years Boys 47%

‘Extra gymnastics:’ 45 min, intense physical activity, mostly in groups, 2 days/weeks AEX: weeks 1–6 supervised, 10 min+walk/run 3 miles, 3 days/ weeks+diet-BMa; AEX: weeks 7-52 walking 3 miles 3 days/weeks with parents+Diet-BMa AEX: 50 min, 60– 80% HRmax, 3 days/ weeks+ Diet-BMb AEX: 50 min, 1 day/ weeks supervised+20 min, 3–4 days/weeks unsupervised+diet: education/ counseling AEX: 50 min, 60– 80% HRmax, 3 days/ weeks+diet- BMb

Diet-BMb

Yes (improved VO2max, NS)

No

Hydrodensitometry (%BF), body weight, BMI

20

AEX: 40 min, 470– 75% HRmax, 3 days/ weeks+diet-BMb

Diet-BMb

No

No

Hydrodensitometry (% BF), body weight

20

EX-1: 48 sessions AEX: 20–30 min+10– 15 min strength training (not defined)+12 sessions nutritional education EX-2: 24 sessions AEX: 20–30 min 20– 30 min+10–15 min strength training (not defined) +12 sessions nutritional education +28

12 sessions nutritional education+48 sessions PA counselingc

Yes (improved aerobic fitness NS, PWC 170)+attendance: EX-1: 79%, C: 81%)

No

% over ideal body weight, body weight, BMI, waist circumference

12

12 sessions nutritional education +48 sessions PA counselingc

Yes (improved aerobic fitness NS, PWC 170)+attendance: EX-2: 53%, C: 81%)

No

% over ideal body weight, body weight, BMI, waist circumference

12

22 12–15 years Boys 27%

Emes (2), 1990 (Canada)64

22 12–15 years Boys 27%

Body weight X20% of ideal weight

Rocchini, 1988 (US)71


Emes (1), 1990 (Canada)64

Body weight plus triceps skin fold 475th percentile for age and sex Body weight for height 475th percentile plus triceps and subscapular skin folds 480th percentile for age and sex Body weight X20% of ideal weight

Post-test (weeks)

1031


1032


Table 1 (continued) Author, year (country)Ref

Sample (no., mean age (s.d.) or range in years, %boys)

Criteria used to define overweight/ obesity

C: 9.5 (0.3) years Boys 34%

Not reported

Ferguson, 1999 (US)62

79 9.5 (1.0) years Boys 33%

Schwingshandl, 1999 (Austria)72

30 T: 11.0 (2.5) years C: 12.2 (2.7) years Boys 43%

Triceps skin fold 485th percentile for sex, ethnicity, and age Not reported

Faith, 2001 (US)65

10 T: 10.2 (1.5) years C: 10.0 (1.6) years Boys 70% (80) 13–16 years Boys 33%

Gutin, 2002 (US)30

Woo, 2004 (Hong Kong)63

BMI 485th percentile for sex and age

Triceps skin fold 485th percentile for sex, ethnicity, and age

82 (54 obese, 28 overweight) 9.9 (1.0) years Boys 66%

BMI X21 kg/m2

Control/comparison condition

Change/compliance in EX/PA reported?

Dietary assessment?

Overweight outcome

No treatment

Yes (88% attendance)

No

DXA (% BF)

17

No treatment (wait list)

Yes (233748 kcal/ session, 80% attendance)

Yes (change NS, via 2 24 h recalls)

DXA (% BF)

17

EX+Diet: 60 min, 2 days/weeks, (AE: NR, WT, 13 exercises, 3– 4 sets, 1st 50% 10RM, last 100% 10RM+diet as controls) AEX: apparatus for television viewing contingent to cycling, ‘TV cycling’

Diet (4180 kJ/days or girls 414 years: 5016 kJ/days, or boys 414 years: 5852 kJ/ days)

No

No

Bioimpedance (% BF), body weight

12

Non-contingent apparatus, that is, TV viewing7cycling

Yes (min cycled/ weeks, T4C)

No

DXA (% BF), body weight

12

AEX-1+LEd: expend 250 kcal/session, 55– 60% VO2peak, 5 days/ weeks+LEd, or AEX-2+LEd: expend 250 kcal/session, 75– 80% VO2peak, 5 days/ weeks+LEd EX: 10 min warm-up, 30 min WT: ‘circuit’, 10 min AEX: 60–70% HRmax, 10 min ‘agility,’ 10 min cooldown; weeks 1–6, 2 days/weeks, 4weeks 6, 1 day/week+diet (as controls)

LEd

Yes (improved VO2max T4C+attendance: 51% AEX-1+LEd, 56% AEX-2+LEd)

Yes (increased caloric intake from baseline, C4T, via 2 24 h recalls)

DXA (% BF), MRI (visceral adipose tissue, cm3)

35

Diet (education for 900–1200 kcal/days)

Yes (83% attendance)

Yes (80% ‘dietary advice’, via parental assisted 3 days diary)

DXA (% BF), body weight, waist-to-hip ratio

6

sessions PA counselingc AEX: 40 min, 4150 bpm, 5 days/ weeks AEX: 40 min, HR 15777 bpm, 5 days/ weeks

Post-test (weeks)

Abbreviations: T ¼ treatment group; C ¼ control/comparison group; AEX ¼ aerobic exercise; WT: weight-training; EX ¼ exercise; HR ¼ heart rate; VO2peak/max ¼ peak/maximal oxygen uptake; PWC 170 ¼ physical work capacity at heart rate of 170 beats/min; DXA ¼ dual energy X-ray absorptiometry; MRI ¼ magnetic resonance imaging; NR ¼ not reported; NA ¼ not significant. aDiet-BM ¼ behavior modification including record keeping, stimulus controlling to restrict external cues for eating, changing the topography of eating and reinforcing altered behavior. bDiet-BM ¼ behavior modification and dietary education (‘ traffic light diet:’ green ¼ ‘o 20 calories,’ yellow ¼ ‘staples,’ red ¼ ‘high calories,’ target caloric intake of 900–1200 kcal/days+behavior modification of individual ‘self-monitoring,’ ‘praise’ and modeling,’ ‘therapist contact’ and ‘contracting’). cPA counseling ¼ physical activity counseling to increase leisure time physical activity level. dLE ¼ lifestyle education, biweekly 60 min classes by clinical psychologist for behavior modification targeting nutrition and physical activity level.


Gutin, 1997 (US)29

Treatment/ intervention


1033 Study

Weight SMDs %

(95% CI)

Gutin29

11.8

0.0

(-0.7, 0.6)

Ferguson62

15.8

-0.8

(-1.3, -0.3)

Faith65

3.0

1.4

(-0.3, 3.1)

Gutin30

12.2

-0.4

(-1.1, 0.2)

Woo63

16.8

0.1

(-0.3, 0.5)

Becque68

8.1

-0.8

(-1.7, 0.1)

Katch70

9.3

-0.8

(-1.6, 0.0)

13.0

71

Rocchini

Schwingshandl72

Total pooled

-0.5

(-1.1, 0.1)

10.1

-0.7

(-1.5, 0.0)

100.0

-0.4

(-0.7, -0.1)

-3

-2

-1

0

1

Favors treatment

2

3

4

Favors control

Heterogeneity test, Q=15.5, P=0.05 Random effects model for overall effect (pooled), P=0.006

Figure 2 Standardized mean differences (SMDs) between treatment and control groups of nine exercise studies in overweight children/adolescents (n ¼ 369): Effects on percent body fat.

Study

Weight MDs %

Faith29

(95% CI)

3.4

2.6

Woo

18.7

4.3

(0.5, 8.1)

Becque68

4.9

-10.5

(-24.6, 3.6)

63

70

(-14.7, 19.9)

9.7

-6.8

(-15.5, 1.8)

Rocchini71

11.0

-0.9

(-8.6, 6.8)

Blomquist66

17.3

-1.3

(-5.8, 3.2)

11.1

-6.7

(-14.4,1.0)

Katch

69

Hills

67

4.5

-1.5

(-16.1, 13.2)

64

5.9

0.2

(-12.3, 12.6)

Emes (2)

64

7.6

-11.8

(-22.1, -1.4)

Schwingshandl72

5.9

-6.1

(-18.5,6.3)

100.0

-2.7

(-6.1, 0.8)

Epstein

Emes (1)

Total pooled

-30

-20

-10

Favors treatment

0

10

20

30

Favors control

Heterogeneity test, Q =17.9, P=0.06 Random effects model for overall effect (pooled), P=0.07

Figure 3 Weighted mean difference (WMD) between treatment and control groups of 11 exercise studies in overweight children/adolescents (n ¼ 334): Effects on body weight.

(2) Study design characteristics: Most study designs were problematic. For example, between-group differences at baseline were substantial in most studies, such as 3.2 and 7.5% in two DXA studies for percent body fat,29,65 6.3% in one hydrodensitometry study for percent body fat,68 12.2

and 15.3% in two studies for percent over ideal body weight,64 while differences of 2.8 and 1.5 kg/m2 were found in two studies for body mass index,64,70 and there was a mean (7s.d.) difference between groups of 7.5 (73.1) kg across six studies for body weight.63–65,68,70,72 In one study of International Journal of Obesity


1034 Study

Weight SMDs (95% CI) %

Gutin30

25.4

-0.2

(-0.8, 0.4)

53.8

-0.2

(-0.6, 0.2)

Emes (1)

13.2

-0.1

(-1.0, 0.8)

Emes (2)64

7.6

-0.9

(-2.1, 0.2)

100.0

-0.2

(-0.6, 0.1)

63

Woo

64

Total pooled

-3

-2

-1

Favors treatment

0

1

2

Favors control

Heterogeneity test, Q=1.5, P=0.7 Fixed effect model for overall effect (pooled), P=0.07

Figure 4 Standardized mean differences (SMDs) between treatment and control groups of four exercise studies in overweight children/adolescents (n ¼ 156): Effects on central obesity-related outcomes.

8

Mean difference in body weight (kg)

6 4 2 0 -2 -4 -6 -8 -10 -12 -14 Rsq = 0.2185

-16 4

6

8

10

12

14

16

18

20

22

Study duration (weeks) Figure 5 Linear regression showing the association between study duration and mean difference (MD) between treatment and control groups for body weight of 11 exercise studies in overweight children and adolescents (n ¼ 334).

central obesity, the mean waist circumference at baseline was 8.11 cm larger in the control group than the treatment group, and this was associated with a large SMD spuriously favoring the treatment group (0.9).64 Randomization protocols were clearly described in only four studies.62,65–67 The proportion of subjects recruited which completed each respective study, defined as the number of subjects with pre- and post-test data, divided International Journal of Obesity

by the number of subjects recruited, ranged between 50 and 100% (median 96%). Only six studies analyzed post-test data on all subjects who entered as study participants,29,63,68–70,72 whereas eight studies analyzed post-test data in study completers only. In no study was intention-to-treat reported as the analytic strategy ‘a priori’. (3) Exercise treatment characteristics: Exercise treatment, compliance, and modality were also likely sources of heterogeneity in effects between studies. Most studies were of aerobic exercise at moderate-to-high intensity and of wide-ranging durations, whereas four studies also included weight training, only one of which was clearly defined,72 and all but three studies,29,62,66 were adjuncts to dietary interventions. Direct or indirect methods of determining exercise level of compliance (i.e., self-reported and/or observed attendance/compliance, or change in maximal aerobic capacity or strength) were reported in 11 of 14 studies,29,30,62–68,70 four of which reported no significant improvement in maximal aerobic capacity.64,68,70 No study reported changes in habitual activity or sedentary levels beyond the prescribed exercise, and no study reported changes in energy expenditure via objective measures such as accelerometers, or doubly labeled water. Of only four studies,30,62,63,66 which reported change in dietary intake, one study66 used 24 h recalls over 4 days, two studies30,62 used 24 h recalls over 2 days, and one study63 used parentalassisted 3-day diaries (Table 1). (4) Subject characteristics: All of the 11 studies in mixed gender cohorts reported combined results for boys and girls. Only three of these 11 studies30,62,63 reported that subjects were randomized into groups by gender, and only three of


1035 Table 2 Weighted mean difference (WMD) or mean difference (MD) between treatment and control groups for overweight outcomes on the same scale of 14 exercise studies in overweight children/adolescents (n ¼ 481) Method of assessment (outcome units) 29,30,62,63,65

DXA (% BF) Bioimpedance (% BF)72 Hydrodensitometry (% BF)68,70,71 Skin folds (mm)66,69 %4Ideal body weight (%)64,67 Body weight (kg)63–72 Body mass index (kg/m2)64,70 MRI (VAT cm3)30 Waist-to-hip ratio63 Waist circumference (cm)64

No. studies

WMD or MDa

(95% CI)

5 1 3 2 3 11 3 1 1 2

0.1 3.6a 4.0 19.6 3.0 2.7 1.2 31.0a 0.01a 5.0

(2.1, 2.0) (8.1, 0.9) (7.2, 2.2) (60.1, 21.7) (6.1, 12.2) (6.1, 0.8) (3.1, 0.8) (12.1, 64.1) (0.04, 0.02) (13.0, 3.0)

Pooling method, P-value Random effects model, 0.5 NA, 0.06 Fixed effect model, o0.001 Random effects model, 0.2 Fixed effect model, 0.7 Random effects model, 0.07 Random effects model, 0.1 NA, 0.3 NA, 0.3 NA, 0.1

a

MD ¼ mean difference between treatment and control groups. NA ¼ not applicable.

14 studies62,63,69 reported the pubertal status of subjects. One study included specific pubertal stage in their eligibility criteria,63 one study reported that boys and girls were ‘prepubertal,’69 and another reported that girls were ‘premenarcheal’ at study entry and that the pubertal status of boys was not obtained.62 Thus, differential exercise effects on overweight outcomes in boys and girls were not investigated, while between-group differences in gender distribution and maturation stage were considered as potential confounders in only three of 11 studies30,62,63 and three of 14 studies,62,63,69 respectively.

Discussion This is the first systematic review and meta-analysis restricted to randomized trials of exercise studies in overweight children/adolescents in the English published literature. We found that exercise significantly reduced percent body fat in obese boys and girls aged approximately 12 years. Summary effects on body weight and central obesity outcomes approached statistical significance. Much of the evidence stemmed from studies of moderate-to-high intensity aerobic exercise plus behavioral interventions for dietary restriction. We found larger summary effects after pooling studies which prescribed a higher vs lower dose of exercise (155–180 vs 120–150 min/weeks), and in those pooled after removing studies which did not report exercise compliance or change in physical activity levels. However, current recommended doses (volumes) of exercise/physical activity for treating overweight in children and adolescents published by several expert committees (B30–60 min/days at moderate intensity, most days (210–360 min/weeks))73,74 as well as USDA’s Dietary Guidelines for Americans 2005,75 are substantially higher than doses which have been tested in RCTs in young overweight cohorts, to date. In fact, we found no randomized trial in the English published literature to have prescribed a dose of more than 200 min/weeks of aerobic exercise to overweight children. Although higher

doses may in fact have greater benefits on body fat/weight and central fat than those observed in the studies we reviewed, this should be subjected to rigorous trials to demonstrate the efficacy and feasibility of such a prescription in this cohort. Our findings do suggest that significant effects on percent body fat are achievable with a lower dose of prescribed exercise than that currently recommended, based on the summary effect of the empirical randomized trials we pooled. Again, the additional benefits of higher volumes of exercise for this specific outcome should be tested empirically. Our findings are of clinical importance since overweight prevalence in children is increasing, adult obesity may potentially cause a decline in life-expectancy, and nonconventional treatments (i.e., pharmacotherapy, very low calorie diets, and bariatric surgery) are not widely accepted as safe, nor have they been shown to be effective treatment options in overweight children/adolescents. Moderate dietary restriction is commonly prescribed in this cohort despite being associated with a reduction in spontaneous physical activity, and a loss of fat-free mass, potentially compromising maximal aerobic capacity; all of which may limit social interaction, as well as growth and cognitive development.76 Low levels of physical activity and greater amounts of sedentary pursuits, in particular television viewing (43 h/ days), during childhood and/or adolescence is predictive of greater future adiposity and or overweight.77–84 Thus, the rationale for prescribing exercise as an adjunct to dietary restriction is compelling given its potential to reduce overweight-related comorbidity and the hazard associated with dietary restriction alone. A quantitative comparison of exercise to other interventions was not possible because sufficient numbers of studies with similar comparisons and outcomes needed for appropriate pooling are not yet available.27 For a qualitative comparison with other interventions we refer readers to the most recent Cochrane systematic review.27 Compared with exercise effects on adult obesity, our summary effect of B3 kg reduction on body weight over 14 weeks is comparable to that reported in the only meta-analysis we found in adults of International Journal of Obesity


1036 exercise-only studies.85 However, we cannot infer that exercise is equally effective for treating both overweight children/adolescents and adults, given the few numbers of studies included in our review, and that the meta-analysis by Miller et al.85 included studies that were conducted in normal and overweight subjects, and additional methodological and validity concerns were raised in a recent quality assessment of this review.86 Nevertheless, a reduction in body weight of 1–2 kg/m is considered to be clinically important, and in accordance with the recent position statement by the American Heart Association on treatment goals for weight management of overweight in children and adolescents.74 The summary effect for percent body fat was medium, whereas small, nonsignificant, summary effects for both body weight and central obesity were found. This was largely due to substantial heterogeneity in effects across studies that were likely due to a number of design/methodological inadequacies identified. Firstly, we found large betweengroup differences at baseline in percent body fat and in body weight in many studies that were broadly equivalent to, and often larger than, the magnitude of change expected after exercise treatment. This was most evident in two studies that used DXA scanning29,65 which spuriously favored control treatments despite observing larger reductions in percent body fat in treatment vs control groups. Better randomization protocols than those used in these studies would have reduced such between-group differences at baseline, and maximized the potential for observing the true intervention effect on overweight outcomes. Secondly, in addition to low exercise frequencies (o3 days/ weeks) prescribed in some studies, exercise compliance may have also been an issue, as only seven studies29,30,62–64,66 reported attendance rates. Also, similar improvements in aerobic capacity in subjects randomized to treatment and control groups were reported in many studies, suggesting all subjects may have increased their exercise/physical activity levels outside the study, and some cross-group contamination may have occurred. Alternatively, such changes may have been due to measurement error, particularly for estimates of aerobic capacity. We recommend that studies report compliance in a comprehensive way in order to identify contributing factors for both increased and decreased efficacy of respective interventions. Thirdly, we cannot discount the possibility that some subjects may have increased or decreased dietary intake, which would have confounded the effects of exercise, because only four studies30,62,63,66 reported change in dietary intake. Studies need to measure changes in both caloric expenditure and caloric intake, to determine the energy balance achieved and the relationship to overweight outcomes. Fourthly, it is well known that substantial measurement error is associated with anthropometry (e.g., waist circumference for central obesity), and such assessment techniques are less sensitive for detecting changes in body composition International Journal of Obesity

outcomes than more precise assessment techniques, such as DXA and computer imaging. Finally, other potential confounding factors identified include the possibility of between-group differences in pubertal status of subjects, and in gender distribution. Greater activity-related energy expenditure is seen in children vs adolescents,87 and in prepubertal boys vs pubertal boys,88 while the converse is seen for resting energy expenditure.87,88 Girls have significantly more body fat than boys from early childhood to adulthood, and percent body fat increases in girls while a decline is seen in boys between the ages of 12–18 years.89 Treatment groups which are allocated more boys than girls may be vulnerable to floor effects, and may be confounded by gender differences in body composition shifts during developmental growth. Thus, both maturation and age exert major effects on human metabolism, whereas gender differences in body composition suggest that studies should report data in boys and girls separately. All RCTs in children and adolescents should use stratified randomization procedures and adjust statistical models for maturation stage and gender, as baseline covariates, to minimize their potential confounding effects on overweight outcomes. Although we identified numerous methodological inadequacies in most of the studies reviewed, future studies should undertake costeffectiveness analyses, given the small to medium summary effects on overweight outcomes observed following our meta-analysis of the best available empirical evidence in the English published literature. Our sensitivity analysis showed that studies which included some weight training contributed little to the summary effects. Similar to aerobic exercise studies, 2/4 studies which included weight training, prescribed an exercise frequency of o3 days/week,63,72 two studies did not report exercise frequency (two treatments from one study),64 and only one study reported exercise intensity.72 Weight-training regimes in these studies were a small component of combined treatments, most of which were performed at unspecified intensities. Based on the scarcity of studies in the literature, the role of isolated or adjunctive weight training for overweight treatment in children cannot be determined at present. In adults, weight training has been shown to be superior to aerobic exercise for the maintenance of fat-free mass during dietary restriction,90–92 while aerobic and weight-training exercise in combination was found to be superior to either modality alone for decreasing whole body fat during dietary restriction.90,93 Theoretically, a similar rationale for preserving fat-free mass would apply when selecting an optimal treatment for overweight in youth.

Central obesity Despite the paucity of studies investigating effects on central obesity, we found small exercise effects after pooling four studies which approached statistical significance. There is


1037 sufficient evidence in adult studies to show that reductions in central obesity can be seen following either aerobic94,95 or weight-training96 exercise without weight loss, whereas combined aerobic and weight-training exercise is more effective than aerobic exercise alone,97 and exercise plus dietary restriction is more effective than dietary restriction alone.98 As in adults, excessive visceral fat (central obesity) is associated with an increased disease risk in children and adolescents, particularly for components of the ‘metabolic syndrome.’9–14 The waist circumference measure, a proxy of visceral fat, is uniformly included in the diagnosis of the ‘metabolic syndrome’ in both children and adults.99 Safe and effective treatments for central obesity in youth are urgently needed, given the increasing prevalence of overweight in youth,1–5 and that the prevalence of ‘metabolic syndrome’ in overweight children and adolescents has been reported to be as high as 29–50%,10,13 and rising.100 Numerous studies have found that physically active boys and girls are at decreased risk of metabolic syndrome compared with those less active.101,102 Although regular exercise should be prescribed for disease prevention because it increases total physical activity level, the efficacy of exercise in the treatment of metabolic syndrome in boys and girls is yet to be established. Greater improvement in components of the metabolic syndrome, including blood pressure and blood vessel resistance of the forearm71 and multiple coronary heart disease risk68 (including dyslipidemia and abnormal blood pressure), are seen in boys and girls following exercise plus dietary weight loss vs dietary weight-loss alone. Further studies which use more precise assessment techniques, and which investigate predictors of treatment response are needed to determine the efficacy of exercise in reducing central obesity and other risk factors for the ‘metabolic syndrome.’

Limitations This systematic review did not include studies published in any language other than English, or unpublished studies, or thesis dissertations, and therefore the true treatment effects of exercise may have been overestimated due to publication and/or search biases. Additionally, subjects included in our meta-analysis were of study completers, and effects (MDs, WMDs and SMDs) generated in studies following intentionto-treat analysis would likely be lower. Finally, generalizations from our results are limited to short-term treatment effects in the few populations studied, most of which were comprised of a small number of overweight children and adolescents, community dwelling in the US. Implications Future studies will contribute to the urgently needed empirical knowledge base in this field if they: (1) use precise techniques in the assessment of body composition, (2) choose well-defined, robust, single modality exercise treat-

ments, (3) measure changes in energy balance (i.e., caloric intake and caloric expenditure), (4) monitor intervention compliance (both attendance and dose), (5) follow the CONSORT guidelines for reporting RCTs,103 (6) include long-term interventions and follow-up (at least 12 months), (7) use stratified randomization procedures to minimize between-group differences in age, pubertal status, gender and overweight outcomes at baseline, (8) include factorial studies of exercise alone and in combination with multimodal treatments (individual and family behavioral/dietary/ physical activity interventions), (9) measure clinical correlates associated with obesity during exercise treatments (e.g., components of the ‘metabolic syndrome,’ C-reactive protein, self-efficacy for physical activity, sleep apnea), and (10) conduct cost-effectiveness analyses to compare economic benefits of exercise treatments with other therapeutic modalities. Further research is needed to clarify dose–response effects, long-term sustainability, relationship to health outcomes, and optimal modality of exercise for the treatment of overweight and metabolically important visceral fat depot in children and adolescents.

References 1 Chinn S, Rona RJ. Prevalence and trends in overweight and obesity in three cross sectional studies of British Children, 1974–94. BMJ 2001; 322: 24–26. 2 Moreno LA, Sarria A, Fleta J, Rodriguez G, Bueno M. Trends in body mass index and overweight prevalence among children and adolescents in the region of Aragon (Spain) from 1985 to 1995. Int J Obes Relat Metab Disord: J Int Assoc Study Obes 2000; 24: 925–931. 3 Freedman DS, Srinivasan SR, Valdez RA, Williamson DF, Berenson GS. Secular increases in relative weight and adiposity among children over two decades: the Bogalusa Heart Study. Pediatrics 1997; 99: 420–426. 4 Troiano RP, Flegal KM, Kuczmarski RJ, Campbell SM, Johnson CL. Overweight prevalence and trends for children and adolescents. The National Health and Nutrition Examination Surveys, 1963 to 1991. Arch Pediatr Adolesc Med 1995; 149: 1085–1091. 5 Booth ML, Chey T, Wake M, Norton K, Hesketh K, Dollman J et al. Change in the prevalence of overweight and obesity among young Australians, 1969–1997. Am J Clin Nutr 2003; 77: 29–36. 6 Britz B, Siegfried W, Ziegler A, Lamertz C, Herpertz-Dahlmann BM, Remschmidt H et al. Rates of psychiatric disorders in a clinical study group of adolescents with extreme obesity and in obese adolescents ascertained via a population based study. Int J Obes Relat Metab Disord: J Int Assoc Study Obes 2000; 24: 1707–1714. 7 Erickson SJ, Robinson TN, Haydel KF, Killen JD. Are overweight children unhappy?: Body mass index, depressive symptoms, and overweight concerns in elementary school children. Arch Pediatr Adolesc Med 2000; 154: 931–935. 8 Zametkin AJ, Zoon CK, Klein HW, Munson S. Psychiatric aspects of child and adolescent obesity: a review of the past 10 years. J Am Acad Child Adolesc Psychiatr 2004; 43: 134–150. 9 McMurray RG, Harrel JS, Levine AA, Gansky SA. Childhood obesity elevates blood pressure and total cholesterol independent of physical activity. Int J Obes Relat Metab Disord: J Int Assoc Study Obes 1995; 19: 881–886.



1038 10 Weiss R, Dziura J, Burgert TS, Tamborlane WV, Taksali SE, Yeckel CW et al. Obesity and the metabolic syndrome in children and adolescents. N Engl J Med 2004; 350: 2362–2374. 11 Reich A, Muller G, Gelbrich G, Deutscher K, Godicke R, Kiess W. Obesity and blood pressure – results from the examination of 2365 schoolchildren in Germany. Int J Obes Relat Metab Disord: J Int Assoc Study Obes 2003; 27: 1459–1464. 12 Dietz WH. Overweight in childhood and adolescence. N Engl J Med 2004; 350: 855–857. 13 Cook S, Weitzman M, Auinger P, Nguyen M, Dietz WH. Prevalence of a metabolic syndrome phenotype in adolescents: findings from the third National Health and Nutrition Examination Survey, 1988–1994. Arch Pediatr Adolesc Med 2003; 157: 821–827. 14 Pinhas-Hamiel O, Dolan LM, Daniels SR, Standiford D, Khoury PR, Zeitler P. Increased incidence of non-insulin-dependent diabetes mellitus among adolescents. J Pediatr 1996; 128: 608–615. 15 Murray CJ, Lopez AD. Alternative projections of mortality and disability by cause 1990–2020: Global Burden of Disease Study. Lancet 1997; 349: 1498–1504. 16 Olshansky SJ, Passaro DJ, Hershow RC, Layden J, Carnes BA, Brody J et al. A potential decline in life expectancy in the United States in the 21st century. N Engl J Med 2005; 352: 1138–1145. 17 Dietz WH, Gortmaker SL. Preventing obesity in children and adolescents. Ann Rev Publ Health 2001; 22: 337–353. 18 Kumanyika SK. Minisymposium on obesity: overview and some strategic considerations. Ann Rev Public Health 2001; 22: 293–308. 19 Blair SN, Nichaman MZ. The public health problem of increasing prevalence rates of obesity and what should be done about it. Mayo Clin Proc 2002; 77: 109–113. 20 Jakicic JM, Clark K, Coleman E, Donnelly JE, Foreyt J, Melanson E et al. American college of sports medicine position stand. Appropriate intervention strategies for weight loss and prevention of weight regain for adults. Med Sci Sports Exerc 2001; 33: 2145–2156. 21 National Health and Medical Research Council. Draft National Clinical Guidelines for Weight Control and Obesity Management in Children and Adolescents. Canberra: NHMRC, 2002. pp 1–94. 22 Campbell K, Waters E, O’Meara S, Summerbell C. Interventions for preventing obesity in childhood. A systematic review. Obes Rev 2001; 2: 149–157. 23 Campbell K, Waters E, O’Meara S, Kelly S, Summerbell C. Interventions for treating obesity in children. Cochr Database Syst Rev 2004; 4. 24 LeMura LM, Maziekas MT. Factors that alter body fat, body mass, and fat-free mass in pediatric obesity. Med Sci Sports Exerc 2002; 34: 487–496. 25 Coates TJ, Thoresen CE. Treating obesity in children and adolescents: a review. Am J Publ Health 1978; 68: 143. 26 Lobstein T, Baur L, Uauy R. Obesity in children and young people: a crisis in public health. Obes Rev 2004; 5: 4–85. 27 Summerbell CD, Ashton V, Campbell KJ, Edmunds L, Kelly S, Waters E. Interventions for treating obesity in children. Cochr Database Syst Rev 2003; 3, C0001872. 28 Egger M, Smith GD, Altman DG. Systematic Reviews in Health Care: Meta-Analysis in Context. BMJ Publishing: London, 2001. 29 Gutin B, Owens S, Slavens G, Riggs S, Treiber F. Effect of physical training on heart-period variability in obese children. J Pediatr 1997; 130: 938–943. 30 Gutin B, Barbeau P, Owens S, Lemmon CR, Bauman M, Allison J et al. Effects of exercise intensity on cardiovascular fitness, total body composition, and visceral adiposity of obese adolescents. Am J Clin Nutr 2002; 75: 818–826. 31 Dwyer T, Coonan WE, Leitch DR, Hetzel BS, Baghurst RA. An investigation of the effects of daily physical activity on the health of primary school students in South Australia. Int J Epidemiol 1983; 12: 308–313.


32 Vandongen R, Jenner DA, Thompson C, Taggart AC, Spickett EE, Burke V et al. A controlled evaluation of a fitness and nutrition intervention program on cardiovascular health in 10- to 12-year-old children. Prev Med 1995; 24: 9–22. 33 Sallis JF, McKenzie TL, Alcaraz JE, Kolody B, Faucette N, Hovell MF. The effects of a 2-year physical education program (SPARK) on physical activity and fitness in elementary school students. Sports, Play and Active Recreation for Kids. Am J Public Health 1997; 87: 1328–1334. 34 Morris FL, Naughton GA, Gibbs JL, Carlson JS, Wark JD. Prospective ten-month exercise intervention in premenarcheal girls: positive effects on bone and lean mass. J Bone Mineral Res 1997; 12: 1453–1462. 35 Mo-suwan L, Pongprapai S, Junjana C, Puetpaiboon A. Effects of a controlled trial of a school-based exercise program on the obesity indexes of preschool children. Am J Clin Nutr 1998; 68: 1006–1011. 36 McKay HA, Petit MA, Schutz RW, Prior JC, Barr SI, Khan KM. Augmented trochanteric bone mineral density after modified physical education classes: a randomized school-based exercise intervention study in prepubescent and early pubescent children. J Pediatr 2000; 136: 156–162. 37 Warren JM, Henry CJ, Lightowler HJ, Bradshaw SM, Perwaiz S. Evaluation of a pilot school programme aimed at the prevention of obesity in children. Health Promotion Int 2003; 18: 287–296. 38 Williford HN, Blessing DL, Duey WJ, Barksdale JM, Wang N, Olson MS et al. Exercise training in black adolescents: changes in blood lipids and Vo2max. Ethnicity Dis 1996; 6: 279–285. 39 Tolfrey K, Campbell IG, Batterham AM. Exercise training induced alterations in prepubertal children’s lipid-lipoprotein profile. Med Sci Sports Exerc 1998; 30: 1684–1692. 40 Eliakim A, Makowski GS, Brasel JA, Cooper DM. Adiposity, lipid levels, and brief endurance training in nonobese adolescent males. Int J Sports Med 2000; 21: 332–337. 41 Nichols DL, Sanborn CF, Love AM. Resistance training and bone mineral density in adolescent females. J Pediatr 2001; 139: 494–500. 42 Kahle EB, O’Dorisio TM, Walker RB, Eisenman PA, Reiser S, Cataland S et al. Exercise adaptation responses for gastric inhibitory polypeptide (GIP) and insulin in obese children. Possible extra-pancreatic effects. Diabetes 1986; 35: 579–582. 43 Huttunen NP, Knip M, Paavilainen T. Physical activity and fitness in obese children. Int J Obes Relat Metab Disord 1986; 10: 519–525. 44 Nichols JF, Bigelow DM, Canine KM. Short-term weight loss and exercise training effects on glucose-induced thermogenesis in obese adolescent males during hypocaloric feeding. Int J Obes Relat Metab Disord 1989; 13: 683–690. 45 Hayashi T, Fujino M, Shindo M, Hiroki T, Arakawa K. Echocardiographic and electrocardiographic measures in obese children after an exercise program. Int J Obes Relat Metab Disord 1987; 11: 465–472. 46 Treuth MS, Hunter GR, Figueroa-Colon R, Goran MI. Effects of strength training on intra-abdominal adipose tissue in obese prepubertal girls. Med Sci Sports Exerc 1998; 30: 1738–1743. 47 Widhalm K, Maxa E, Zyman H. Effect of diet and exercise upon the cholesterol and triglyceride content of plasma lipoproteins in overweight children. Eur J Pediatr 1978; 127: 121–126. 48 Pena M, Barta L, Regoly-Merei A, Tichy M. The influence of physical exercise upon the body composition of obese children. Acta Paediatr Acad Sci Hungari 1980; 21: 9–14. 49 Reybrouck T, Vinckx J, Van den Berghe G, VanderschuerenLodeweyckx M. Exercise therapy and hypocaloric diet in the treatment of obese children and adolescents. Acta Paediatr Scand 1990; 79: 84–89. 50 Jette M, Barry W, Pearlman L. The effects of an extracurricular physical activity program on obese adolescents. Canad J Publ Health Rev Canad Sante Publ 1977; 68: 39–42.


1039 51 Gutin B, Cucuzzo N, Islam S, Smith C, Moffatt R, Pargman D. Physical training improves body composition of black obese 7to 11-year-old girls. Obes Res 1995; 3: 305–312. 52 Pena M, Bacallao J, Barta L, Amador M, Johnston FE. Fiber and exercise in the treatment of obese adolescents. J Adolesc Health Care 1989; 10: 30–34. 53 Kang HS, Gutin B, Barbeau P, Owens S, Lemmon CR, Allison J et al. Physical training improves insulin resistance syndrome markers in obese adolescents. Med Sci Sports Exerc 2002; 34: 1920–1927. 54 Barbeau P, Gutin B, Litaker M, Owens S, Riggs S, Okuyama T. Correlates of individual differences in body-composition changes resulting from physical training in obese children. Am J Clin Nutr 1999; 69: 705–711. 55 Barbeau P, Gutin B, Litaker MS, Ramsey LT, Cannady WE, Allison J et al. Influence of physical training on plasma leptin in obese youths. Canad J Appl Physiol 2003; 28: 382–396. 56 Mitchell BM, Gutin B, Kapuku G, Barbeau P, Humphries MC, Owens S et al. Left ventricular structure and function in obese adolescents: relations to cardiovascular fitness, percent body fat, and visceral adiposity, and effects of physical training. Pediatrics 2002; 109: e73. 57 Gutin B, Ramsey L, Barbeau P, Cannady W, Ferguson M, Litaker M et al. Plasma leptin concentrations in obese children: changes during 4-mo periods with and without physical training. Am J Clin Nutr 1999; 69: 388–394. 58 Gutin B, Cucuzzo N, Islam S, Smith C, Stachura ME. Physical training, lifestyle education, and coronary risk factors in obese girls. Med Sci Sports Exerc 1996; 28: 19–23. 59 Gutin B, Owens S, Okuyama T, Riggs S, Ferguson M, Litaker M. Effect of physical training and its cessation on percent fat and bone density of children with obesity. Obes Res 1999; 7: 208–214. 60 Humphries MC, Gutin B, Barbeau P, Vemulapalli S, Allison J, Owens S. Relations of adiposity and effects of training on the left ventricle in obese youths. Med Sci Sports Exerc 2002; 34: 1428–1435. 61 Sung RY, Yu CW, Chang SK, Mo SW, Woo KS, Lam CW. Effects of dietary intervention and strength training on blood lipid level in obese children. Arch Dis Child 2002; 86: 407–410. 62 Ferguson MA, Gutin B, Le NA, Karp W, Litaker M, Humphries M et al. Effects of exercise training and its cessation on components of the insulin resistance syndrome in obese children. Int J Obes Relat Metab Disord J Int Assoc Study Obes 1999; 23: 889–895. 63 Woo KS, Chook P, Yu CW, Sung RY, Qiao M, Leung SS et al. Effects of diet and exercise on obesity-related vascular dysfunction in children. Circulation 2004; 109: 1981–1986. 64 Emes C, Velde B, Moreau M, Murdoch DD, Trussell R. An activity based weight control program. Adapt Phys Activity Quart 1990; 7: 314–324. 65 Faith MS, Berman N, Heo M, Pietrobelli A, Gallagher D, Epstein LH et al. Effects of contingent television on physical activity and television viewing in obese children. Pediatrics 2001; 107: 1043–1048. 66 Blomquist B, Borjeson M, Larsson Y, Persson B, Sterky G. The effect of physical activity on the body measurements and work capacity of overweight boys. Acta Paediatr Scand 1965; 54: 566–572. 67 Epstein LH, Wing RR, Penner BC, Kress MJ. Effect of diet and controlled exercise on weight loss in obese children. J Pediatr 1985; 107: 358–361. 68 Becque MD, Katch VL, Rocchini AP, Marks CR, Moorehead C. Coronary risk incidence of obese adolescents: reduction by exercise plus diet intervention. Pediatrics 1988; 81: 605–612. 69 Hills AP, Parker AW. Obesity management via diet and exercise intervention. Child Care Health Dev 1988; 14: 409–416. 70 Katch V, Becque MD, Marks C, Moorehead C, Rocchini A. Basal metabolism of obese adolescents: inconsistent diet and exercise effects. Am J Clin Nutr 1988; 48: 565–569.

71 Rocchini AP, Katch V, Anderson J, Hinderliter J, Becque D, Martin M et al. Blood pressure in obese adolescents: effect of weight loss. Pediatrics 1988; 82: 16–23. 72 Schwingshandl J, Sudi K, Eibl B, Wallner S, Borkenstein M. Effect of an individualised training programme during weight reduction on body composition: a randomised trial. Arch Dis Childhood 1999; 81: 426–428. 73 Saris WH, Blair SN, van Baak MA, Eaton SB, Davies PS, Di Pietro L et al. How much physical activity is enough to prevent unhealthy weight gain? Outcome of the IASO 1st Stock Conference and consensus statement. Obes Rev 2003; 4: 101–114. 74 Daniels SR, Arnett DK, Eckel RH, Gidding SS, Hayman LL, Kumanyika S et al. Overweight in children and adolescents: pathophysiology, consequences, prevention, and treatment. Circulation 2005; 111: 1999–2012. 75 USDA. Dietary Guidelines for Americans 2005. http://www. healthierus.gov/dietaryguidelines/, (accessed 20 July, 2005). 76 Bar-Or O, Foreyt J, Bouchard C, Brownell KD, Dietz WH, Ravussin E et al. Physical activity, genetic, and nutritional considerations in childhood weight management. Med Sci Sports Exerc 1998; 30: 2–10. 77 Arluk SL, Branch JD, Swain DP, Dowling EA. Childhood obesity’s relationship to time spent in sedentary behavior. Military Med 2003; 168: 583–586. 78 Crespo CJ, Smit E, Troiano RP, Bartlett SJ, Macera CA, Andersen RE. Television watching, energy intake, and obesity in US children: results from the third National Health and Nutrition Examination Survey, 1988–1994. Arch Pediatr Adolesc Med 2001; 155: 360–365. 79 Maffeis C, Talamini G, Tato L. Influence of diet, physical activity and parents’ obesity on children’s adiposity: a four-year longitudinal study. Int J Obes Relat Metab Disord: J Int Assoc Study Obes 1998; 22: 758–764. 80 Proctor MH, Moore LL, Gao D, Cupples LA, Bradlee ML, Hood MY et al. Television viewing and change in body fat from preschool to early adolescence: The Framingham Children’s Study. Int J Obes Relat Metab Disord: J Int Assoc Study Obes 2003; 27: 827–833. 81 Bouziotas C, Koutedakis Y, Nevill A, Ageli E, Tsigilis N, Nikolaou A et al. Greek adolescents, fitness, fatness, fat intake, activity, and coronary heart disease risk. Arch Dis Childhood 2004; 89: 41–44. 82 O’Loughlin J, Gray-Donald K, Paradis G, Meshefedjian G. Oneand two-year predictors of excess weight gain among elementary schoolchildren in multiethnic, low-income, inner-city neighborhoods. Am J Epidemiol 2000; 152: 739–746. 83 Hernandez B, Gortmaker SL, Colditz GA, Peterson KE, Laird NM, Parra-Cabrera S. Association of obesity with physical activity, television programs and other forms of video viewing among children in Mexico city. Int J Obes Relat Metab Disord: J Int Assoc Study Obes 1999; 23: 845–854. 84 Patrick K, Norman GJ, Calfas KJ, Sallis JF, Zabinski MF, Rupp J et al. Diet, physical activity, and sedentary behaviors as risk factors for overweight in adolescence. Arch Pediatr Adolesc Med 2004; 158: 385–390. 85 Miller WC, Koceja DM, Hamilton EJ. A meta-analysis of the past 25 years of weight loss research using diet, exercise or diet plus exercise intervention. Int J Obes Relat Metab Disord: J Int Assoc Study Obes 1997; 21: 941–947. 86 Centre for R, Dissemination. A meta-analysis of the past 25 years of weight loss research using diet, exercise or diet plus exercise intervention (structured abstract). Database Abstr Rev Effectiv 2005; 2: 2. 87 Ekelund U, Yngve A, Brage S, Westerterp K, Sjostrom M. Body movement and physical activity energy expenditure in children and adolescents: how to adjust for differences in body size and age. Am J Clin Nutr 2004; 79: 851–856. 88 Roemmich JN, Clark PA, Walter K, Patrie J, Weltman A, Rogol AD. Pubertal alterations in growth and body composition.



1040 89

90

91

92

93

94

95

V. Energy expenditure, adiposity, and fat distribution. Am J Physiol Endocrinol Metab 2000; 279: E1426–E1436. Guo SS, Chumlea WC, Roche AF, Siervogel RM. Age- and maturity-related changes in body composition during adolescence into adulthood: the Fels Longitudinal Study. Int J Obes Relat Metab Disord: J Int Assoc Study Obes 1997; 21: 1167–1175. Kraemer WJ, Volek JS, Clark KL, Gordon SE, Puhl SM, Koziris LP et al. Influence of exercise training on physiological and performance changes with weight loss in men. Med Sci Sports Exerc 1999; 31: 1320–1329. Bryner RW, Ullrich IH, Sauers J, Donley D, Hornsby G, Kolar M et al. Effects of resistance vs. aerobic training combined with an 800 calorie liquid diet on lean body mass and resting metabolic rate. J Am Coll Nutr 1999; 18: 115–121. Geliebter A, Maher MM, Gerace L, Gutin B, Heymsfield SB, Hashim SA. Effects of strength or aerobic training on body composition, resting metabolic rate, and peak oxygen consumption in obese dieting subjects. Am J Clin Nutr 1997; 66: 557–563. Sweeney ME, Hill JO, Heller PA, Baney R, DiGirolamo M. Severe vs moderate energy restriction with and without exercise in the treatment of obesity: efficiency of weight loss. Am J Clin Nutr 1993; 57: 127–134. Ross R, Dagnone D, Jones PJ, Smith H, Paddags A, Hudson R et al. Reduction in obesity and related comorbid conditions after diet-induced weight loss or exercise-induced weight loss in men. A randomized, controlled trial. Ann Int Med 2000; 133: 92–103. Pritchard JE, Nowson CA, Wark JD. A worksite program for overweight middle-aged men achieves lesser weight loss with


96

97

98

99 100

101

102

103

exercise than with dietary change. J Am Diet Assoc 1997; 97: 37–42. Treuth MS, Ryan AS, Pratley RE, Rubin MA, Miller JP, Nicklas BJ et al. Effects of strength training on total and regional body composition in older men. J Appl Physiol 1994; 77: 614–620. Park SK, Park JH, Kwon YC, Kim HS, Yoon MS, Park HT. The effect of combined aerobic and resistance exercise training on abdominal fat in obese middle-aged women. J Physiol Anthropol Appl Hum Sci 2003; 22: 129–135. Smith SR, Zachwieja JJ. Visceral adipose tissue: a critical review of intervention strategies. Int J Obes Relat Metab Disord: J Int Assoc Study Obes 1999; 23: 329–335. Cruz ML, Goran MI. The metabolic syndrome in children and adolescents. Curr Diabet Rep 2004; 4: 53–62. Duncan GE, Li SM, Zhou XH. Prevalence and trends of a metabolic syndrome phenotype among US adolescents, 1999–2000. Diabet Care 2004; 27: 2438–2443. Ku CY, Gower BA, Hunter GR, Goran MI. Racial differences in insulin secretion and sensitivity in prepubertal children: role of physical fitness and physical activity. Obes Res 2000; 8: 506–515. Brage S, Wedderkopp N, Ekelund U, Franks PW, Wareham NJ, Andersen LB et al. Features of the metabolic syndrome are associated with objectively measured physical activity and fitness in Danish children: the European Youth Heart Study (EYHS). Diabet Care 2004; 27: 2141–2148. Begg CP, Cho MP, Eastwood SELSD, Horton RMB, Moher DM, Olkin IP et al. Improving the quality of reporting of randomized controlled trials: the CONSORT statement. JAMA 1996: 276: 637–639.