Scolaris Content Display Scolaris Content Display

Cochrane Base de datos Cochrane de Revisiones Sistemáticas

Interventions for preventing obesity in children

Podcast

disponible en

Interventions for preventing obesity in children

Obesity is a growing problem around the world and an updated Cochrane Review in December 2011 summarizes the evidence on interventions to prevent obesity in children. Elizabeth Waters from the University of Melbourne in Australia describes the latest findings.

Quorom statement flow diagram ‐ Interventions for preventing obesity in children
Figuras y tablas -
Figure 1

Quorom statement flow diagram ‐ Interventions for preventing obesity in children

Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.
Figuras y tablas -
Figure 2

Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.

Funnel plot of comparison: 1 Childhood obesity interventions versus control by age groups 0‐5, 6‐12 and 13‐18 years, outcome: 1.1 Standardised mean change in Body Mass Index (BMI/zBMI) from baseline to post‐intervention.
Figuras y tablas -
Figure 3

Funnel plot of comparison: 1 Childhood obesity interventions versus control by age groups 0‐5, 6‐12 and 13‐18 years, outcome: 1.1 Standardised mean change in Body Mass Index (BMI/zBMI) from baseline to post‐intervention.

Forest plot of comparison: 1 Childhood obesity interventions versus control by age groups 0‐5, 6‐12 and 13‐18 years, outcome: 1.1 Standardised mean change in Body Mass Index (BMI/zBMI) from baseline to post intervention.
Figuras y tablas -
Figure 4

Forest plot of comparison: 1 Childhood obesity interventions versus control by age groups 0‐5, 6‐12 and 13‐18 years, outcome: 1.1 Standardised mean change in Body Mass Index (BMI/zBMI) from baseline to post intervention.

Forest plot of the standardised mean change in body mass index (BMI/zBMI) from baseline to post intervention for childhood obesity interventions versus control grouped by intervention type (physical activity, dietary, combined physical activity/dietary)
Figuras y tablas -
Figure 5

Forest plot of the standardised mean change in body mass index (BMI/zBMI) from baseline to post intervention for childhood obesity interventions versus control grouped by intervention type (physical activity, dietary, combined physical activity/dietary)

Forest plot of the standardised mean change in body mass index (BMI/zBMI) from baseline to post intervention for childhood obesity interventions versus control grouped by setting
Figuras y tablas -
Figure 6

Forest plot of the standardised mean change in body mass index (BMI/zBMI) from baseline to post intervention for childhood obesity interventions versus control grouped by setting

Forest plot of the standardised mean change in body mass index (BMI/zBMI) from baseline to post intervention for childhood obesity interventions versus control grouped by duration of intervention (short term and long term)
Figuras y tablas -
Figure 7

Forest plot of the standardised mean change in body mass index (BMI/zBMI) from baseline to post intervention for childhood obesity interventions versus control grouped by duration of intervention (short term and long term)

Forest plot of the standardised mean change in body mass index (BMI/zBMI) from baseline to post intervention for childhood obesity interventions versus control grouped by risk of bias based on randomisation (high risk=non‐randomised; unclear risk=method of randomisation or sequence generation unclear; low risk=randomisation occurred appropriately)
Figuras y tablas -
Figure 8

Forest plot of the standardised mean change in body mass index (BMI/zBMI) from baseline to post intervention for childhood obesity interventions versus control grouped by risk of bias based on randomisation (high risk=non‐randomised; unclear risk=method of randomisation or sequence generation unclear; low risk=randomisation occurred appropriately)

Comparison 1 Childhood obesity interventions versus control by age groups 0‐5, 6‐12 and 13‐18 years, Outcome 1 Standardised mean change in Body Mass Index (BMI/zBMI) from baseline to postintervention.
Figuras y tablas -
Analysis 1.1

Comparison 1 Childhood obesity interventions versus control by age groups 0‐5, 6‐12 and 13‐18 years, Outcome 1 Standardised mean change in Body Mass Index (BMI/zBMI) from baseline to postintervention.

Table 1. Study Design

Study

Type

Country

Guiding theoretical frameworks

Setting

Child age (at Baseline)

Intervention period

Care

Education

Health Service

Community

Home

0‐5 years

6‐12 years

13‐18 years

12 weeks‐1 year

>1 year‐2 years

>2 years

Dennison 2004

PA

USA

NR‐behaviour change

X

X

X

X

Fitzgibbon 2005

Diet & PA combined

USA

SCT

X

X

X

Fitzgibbon 2006

Diet & PA combined

USA

SCT

X

X

X

Harvey‐Berino 2003

Diet & PA combined

USA

NR‐behaviour change

X

X

X

Jouret 2009

Diet & PA combined

France

NR‐behaviour change theory

X

X

X

X

Keller 2009

Diet & PA combined

Germany

NR‐behaviour change

X

X

X

X

Mo‐Suwan 1998

PA

Thailand

NR‐environmental change

X

X

X

Reilly 2006

PA

Scotland

NR‐environmental change & behavioural

X

X

X

Amaro 2006

Diet

Italy

NR

X

X

X

Baranowski 2003

Diet & PA combined

USA

SCT and family systems theory

X (summer camp)

X

X

X

Beech 2003

Diet & PA combined

USA

SCT and family systems theory

X

X

X

Caballero 2003

Diet & PA combined

USA

Social learning theory & principles of American Indian culture and practice

X

X

X

Coleman 2005

Diet & PA combined

USA

NR

X

X

X

Donnelly 2009

PA

USA

NR‐environmental model

X

X

X

Epstein 2001

Diet

USA

NR

X

X

X

Fernandes 2009

Diet

Brazil

Learning through play

X

X

X

Foster 2008

Diet & PA combined

USA

Settings based, CDC guidelines to promote lifelong HE and PA

X

X

X

Gentile 2009

Diet & PA combined

USA

Socio‐ecological theory

X

X

X

X

X

Gortmaker 1999a

Diet & PA combined

USA

SCT

X

X

X

Gutin 2008

PA

USA

Environmental change

X

X

X

Hamelink‐Basteen 2008

Diet & PA combined

Netherlands

NR

X

X

X

Harrison 2006

PA

Ireland

SCT

X

X

X

James 2004

Diet

UK

NR

X

X

X

Kain 2004

Diet & PA combined

Chile

NR

X

X

X

Kipping 2008

Diet & PA combined

UK

SCT & behavioural choice

X

X

X

Lazaar 2007

PA

France

NR

X

X

X

Macias‐Cervantes 2009

PA

Mexico

NR

X

X

X

Marcus 2009

Diet & PA combined

Sweden

NR

X

X

X

Müller 2001

Diet & PA combined

Germany

NR

X

X

X

X

Paineau 2008

Diet

France

NR

X

X

X

X

Pangrazi 2003

PA

Mexico

Behavioural

X

X

X

Reed 2008

PA

Canada

socio‐ecological model

X

X

X

Robbins 2006

PA

USA

The Health Promotion Model and the Transtheoretical Model

X

X

X

X

Robinson 2003

Diet & PA combined

USA

Social cognitive theory

X

X

X

Rodearmel 2006

Diet & PA combined

USA

NR

X

X

X

Sahota 2001

Diet & PA combined

UK

Multicomponent health promotion programme, based on the Health Promoting Schools concept

X

X

X

Sallis 1993

PA

USA

Behaviour change and self‐management

X

X

X

Salmon 2008

PA

Australia

SCT and behavioural choice theory

X

X

X

Sanigorski 2008

Diet & PA combined

Australia

Socio‐ecological model

X

X

X

X

Sichieri 2009

Diet

Brazil

NR

X

X

X

Simon 2008

PA

France

Behaviour change and socio‐ecological model

X

X

X

Spiegel 2006

Diet & PA combined

USA

Theory of reasoned action, constructivism

X

X

X

Stolley 1997

Diet & PA combined

USA

NR

X

X

X

Story 2003a

Diet & PA combined

USA

SCT, youth development, and resiliency

X

X

X

X

Taylor 2008

Diet & PA combined

New Zealand

NR

X

X

X

Vizcaino 2008

PA

Spain

NR

X

X

X

Warren 2003

Diet & PA combined

England

Social learning theory

X

X

X

X

Ebbeling 2006

Diet

USA

NR

X

X

X

Haerens 2006

Diet & PA combined

Belgium

Theory of planned behaviours & transtheoretical model

X

X

X

NeumarkSztainer 2003

Diet & PA combined

USA

SCT

X

X

X

Pate 2005

PA

USA

Socio‐ecological model & SCT

X

X

X

Patrick 2006

Diet & PA combined

USA

Behavioural determinants model, SCT & transtheoretical model

X

X

X

Peralta 2009

Diet & PA combined

Australia

SCT

X

X

X

Singh 2009

Diet and PA combined

Netherlands

Intervention mapping protocol, behaviour change & environmental

X

X

X

Webber 2008

PA

USA

Socio‐ecological framework

X

X

X

TOTALS

2

43

2

6

14

8

39

8

40

7

8

Figuras y tablas -
Table 1. Study Design
Table 2. Results 0‐5 years

Study ID

Primary Outcomes

Secondary Outcomes

Dennison 2004

1. Fatness assessed by repeat measures of height and weight (and calculated BMI) at baseline and after 1 year (end of intervention):
OUTCOME: No differences between intervention and control.
2. Skinfolds:
OUTCOME: No differences between intervention and control.
3. Waist circumference:
OUTCOME: No differences between intervention and control.
4. Television Viewing:
OUTCOME: television viewing was significantly reduced in intervention group on weekdays and Sundays. The percentage of children watching > 2h per day was also significantly decreased in intervention group.

1. Computer games playing:
OUTCOME: No differences between intervention and control.
2. Dietary assessment:
OUTCOME: No significant changes or differences between intervention and control groups in the frequency of snacking whilst watching TV or the number of days family ate dinner together or watched TV during dinner (actual data not reported).

Fitzgibbon 2005

MEASURES: BMI

OUTCOMES: Immediately post‐intervention, changes in BMI and BMI z score were not significantly different between intervention and control children.

Intervention children had significantly smaller increases in BMI compared with control children at 1‐year follow‐up (0.06 vs 0.59 kg/m2; difference ‐0.53 kg/m2 (95%CI: ‐0.91 to ‐0.14), P = 0.01), and at 2‐year follow‐up (0.54 vs 1.08 kg/m2; difference ‐0.54 kg/m2 (95% CI: ‐0.98 to ‐0.10), P = 0.02), with adjustment for baseline age and BMI.

MEASURES: dietary intake

OUTCOMES: Reported intake of total fat and dietary fibre was similar between children in the control and intervention groups at all assessment points.

Saturated fat intake was significantly lower in intervention children at Year 1 (P = 0.002) but not post‐intervention or at Year‐2 follow‐up.

MEASURES: Physical activity

OUTCOMES: No significant differences between groups in reported frequency and intensity of exercise.

MEASURES: Television viewing

OUTCOMES: No significant differences between groups in TV viewing at any assessment point.

Fitzgibbon 2006

MEASURES: BMI

OUTCOMES: Post‐intervention changes in BMI and BMI z score were not significantly different between intervention and control children

MEASURES: dietary intake

OUTCOMES: Reported intake of total and saturated fat and dietary fibre was similar between children in the control and intervention groups at Year 2 follow‐up.

MEASURES: Physical activity

OUTCOMES: No significant differences between groups in reported frequency and intensity of exercise.

MEASURES: Television viewing

OUTCOMES: No significant differences between groups in TV viewing at any assessment point.

Harvey‐Berino 2003

1. Maternal fatness assessed by repeat measures of height and weight (and calculated BMI) at baseline and end of pilot:
OUTCOME: No differences between intervention and control.
2. % WHP scores > 85th and 95th percentile:
OUTCOME: No differences between intervention and control.
3. % WHZ scores > 85th and 95th percentile:

1. Diet 3‐day food records:
OUTCOME: No differences between intervention and control.

2. Physical activity: CSA accelerometer,
OUTCOME: No differences between intervention and control.
3. Psychological variables: Outcomes Expectations
Self‐efficacy
Intentions
Child Feeding Questionnaire
OUTCOME: No differences between intervention and control.

Jouret 2009

MEASURES: Weight, height

OUTCOMES:

Prevalence of overweight (BMI ≥ 90th percentile)

1          At end of study, 12.6% in EPIPOI‐1 group was overweight, 11.3% in EPIPOI‐2 group, and 17.8% in control (EPIPOI‐1 vs control P = 0.02; EPIPOI‐2 vs control P =0.003)

2         There was no difference between groups if the schools were not in underprivileged areas, however there was a significant intervention effect in school s in underprivileged areas

At end of study, 12.2% in EPIPOI‐1 group was overweight, 17.0% in EPIPOI‐2 group, and 36.8% in control (EPIPOI‐1 vs control P <0.01; EPIPOI‐2 vs control P = 0.001)

BMI z‐score

1          At end of study and among schools not in underprivileged areas, median change in BMI z‐score in EPIPOI‐1 group was +0.39,  +0.22 in EPIPOI‐2 group, +0.41 in control (EPIPOI‐2 vs control P = 0.01)

3         At end of study and among schools in underprivileged areas, median change in BMI z‐score in EPIPOI‐1 group was +0.35, +0.50 in EPIPOI‐2 group, and +1.35 in control (EPIPOI‐1 vs control P < 0.001; EPIPOI‐2 vs control P < 0.001)

Keller 2009

MEASURES: Height, Weight

OUTCOMES:  This study population stabilized their BMI SDS (P < 0.025).   The children randomised in the intervention group who were not interested to participate, and the children of the control group increased their BMI SDS within the observation period of one year (P < 0.001, P = 0.002).

MEASURES: Diet

OUTCOMES:  According to nutrition diaries a decrease energy intake of the participants of the intervention group was detected. The percentage of protein intake was particularly remarkable, amounting to 363% fulfilment of demand at the beginning of the study and 274% at the end.

Mo‐Suwan 1998

1. Fatness assessed by weight, height (BMI, WHCU weight (kg)/height cubed), and triceps skinfold thickness at baseline, twice during intervention and at 29.6 weeks (end of intervention).

OUTCOME: No statistically significant change between intervention and control at 29.6 weeks (end of intervention). The prevalence of obesity, using 95th percentile National Center for Health Statistics triceps skinfold‐thickness cutoffs, of both the intervention and control groups decreased. The intervention group decreased from 12.2% at baseline to 8.8% (P = 0.058) and the control group decreased from 11.7% to 9.7% (P = 0.179). A sex difference in the response of BMI to exercise was observed. Girls in the exercise group had a lower likelihood of having an increasing BMI slope than the control girls did (odds ratio: 0.32; 95% CI: 0.18 to 0.56).

Follow‐up data on (overall prevalence of obesity, using 95th percentile National Center for Health Statistics triceps‐skinfold thickness cut‐offs in the control group )

Prevalence of obesity
Baseline Intervention 12.9 Control 12.2
Post‐intervention (29.6 wks) Intervention 8.8 Control 9.4
Six months later Intervention 10.2 Control 10.8

Data for follow‐up 29.6 wks + 6 months.

School I
Baseline Intervention 16.2 Control 12.5
Post‐intervention (29.6 wks) Intervention 8.1 Control 8.3
Six months later Intervention 13.5 Control 8.3.

School II
Baseline Intervention 11.8 Control 12.1
Post‐intervention (29.6 wks) Intervention 9.2 Control 9.9
Six months later Intervention Intervention 9.1 Control 12.1.

It is not known (information not available) if the changes at 29.6 weeks plus 6 months are statistically significant . But small changes are unlikely to be clinically significant.

None reported

Reilly 2006

MEASURES: BMI

OUTCOMES: No significant differences between intervention and control groups.

MEASURES: physical activity and sedentary behaviour by accelerometry

OUTCOxMES: No significant differences between intervention and control groups.

MEASURES: fundamental movement skills

OUTCOMES: Children in the intervention group had significantly higher performance in movement skills tests than control children at 6 month follow‐up (i.e. immediately post‐intervention) after adjustment for sex and baseline performance.

BMI: body mass index
BMIz: standardised body mass index
CI: confidence interval

Figuras y tablas -
Table 2. Results 0‐5 years
Table 3. Results 6‐12 years

Study ID

Primary Outcomes

Secondary Outcomes

Amaro 2006

MEASURES:  Height, weight

OUTCOMES:  No significant difference in zBMI between treated group and control group at post‐assessment controlling for baseline values. Adjusted means were 0.345 (95% CI 0.299 to 0.390) for the intervention group and 0.405 (95% CI 0.345 to 0.465) for the control group. 

MEASURES: Nutrition knowledge

OUTCOMES: Intervention group had significant increase in nutrition knowledge (P < 0.05) compared to control.  Adjusted means were 11.24 (95%CI 10.68 to 11.80) for the intervention group and 9.24 (95% CI 8.50 to 9.98) for the control group.

MEASURES: Dietary Intake

OUTCOMES: Intervention group had significant increase in weekly vegetable intake (P < 0.01) compared with control.  Adjusted mean number of servings per week was 3.7 (95% CI 3.5 to 4.1) for the intervention group and 2.8 (95% CI 2.4 to 3.3) for the control group.

MEASURES: Physical activity

OUTCOMES: No significant difference between groups post‐intervention

Baranowski 2003

1. Fatness assessed by repeat measures of height and weight (and calculated BMI) at baseline and end of pilot:
OUTCOME: No differences between intervention and control.
2. Waist circumference:
OUTCOME: No differences between intervention and control.
3. Dual X‐Ray Absorptiometry (DEXA) for % Body fat
OUTCOME: Not reported.

4. Physical activity: CSA accelerometer,
OUTCOME: No differences between I and C.
5. a modification of the Self‐Administered Physical Activity Checklist (SAPAC),
OUTCOME: No differences between intervention and control.
6. GEMS Activity Questionnaire (GAQ) computerised
OUTCOME: No differences between intervention and control.

7. Dietary intake measured by two 24 hour recalls using Nutrition Data System computer programme (NDS‐R).
OUTCOME: No differences between intervention and control.

1. Participation in summer camp
OUTCOME: I: 91.5% and C: 80.5%
2. Monitoring website usage (log‐on rates).
OUTCOME: Intervention: child mean 48%, parent mean 47%; Control: child mean 25%, parent mean 16%.

Beech 2003

1. Fatness assessed by repeat measures of height and weight (and calculated BMI) at baseline and end of pilot:
OUTCOME: No differences between intervention and control.
2. Waist circumference:
OUTCOME: No differences between intervention and control.
3. Dual X‐Ray Absorptiometry (DEXA) for % Body fat
OUTCOME: Not reported.

4. Physical activity: accelerometer CSA,
OUTCOME: No differences between intervention and control.
5. a modification of the Self‐Administered Physical Activity Checklist (SAPAC),
OUTCOME: Not reported.
6. GEMS Activity Questionnaire(GAQ) computerised
OUTCOME: No differences between intervention and control.

7. Dietary intake measured by two 24 hour recalls using Nutrition Data System computer programme (NDS‐R).
OUTCOME: Intervention parent group significantly lower for sweetened drinks compared with intervention child group and controls.

1. Psychological variables:
Body silhouettes McKnight Risk Factor Survey, and Stunkard et al. 1983.
OUTCOME: No differences between intervention and control
2. Over concern with weight or shape:
OUTCOME: Intervention significantly better than control.
3. Parental food preparation practices
OUTCOME: Intervention significantly better than control.
4. Self‐Perception Profile for Children
OUTCOME: No differences between intervention and control
5. Healthy Growth Study for physical activity expectations, and a self‐efficacy measure.
OUTCOME: No differences between intervention and control.

Caballero 2003

1. Fatness assessed by repeat measures of height and weight (and calculated BMI), at baseline and after 3 years (end of intervention):
OUTCOME: No differences between intervention and control
2. Triceps and subscapular Skinfolds.
OUTCOME: No differences between intervention and control
3. Bioelectrical impedance.
OUTCOME: No differences between intervention and control.

1. Lunch Programme:
OUTCOME: Intervention school's lunches had significantly less energy from fat (4%), P = 0.005. 24 hour dietary records showed significant reduction in energy P = 0.003 and total fat P = 0.001.
2. Physical Activity
OUTCOME: Tri Trac R3D accelerometer showed no significant differences, but trends were in the desired direction. 24 hour recalls were significantly higher in I P = 0.001.
3. Knowledge, attitudes and beliefs:
OUTCOME: significant improvements were found in I, especially in the 3rd grade (8‐9 years), but Self efficacy to be physically active was higher in I schools but choosing healthy foods was not.
4. Family Programme
OUTCOME: families attending events was 58%.

Coleman 2005

MEASURE: Risk of overweight and overweight

OUTCOME: The rate of increase in the percentage of students at risk of overweight or overweight from 3rd to 5th grades was 13% in control girls compared with 2% in intervention girls and 9% in control boys compared with 1% in intervention boys.

MEASURE: Anthropometry

OUTCOME: No effect of intervention on height, weight, waist‐to‐hip ratio or BMI. All children had increases in these measures from year to year.

MEASURE: Aerobic fitness

OUTCOME: Results for passing Fitnessgram standards were similar between intervention and control schools for the 3rd grade. In the 4th grade, control schools had higher rates than intervention schools, while in the 5th grade, intervention schools had higher rates than control schools.

MEASURE: PE outcomes

OUTCOME: For part of the 3rd and 4th grades, intervention schools had higher MVPA than control schools. By the end of the 4th grade, control schools had reached similar values to intervention schools, with a similar pattern for the 5th grade.

Intervention schools has higher vigorous physical activity (VPA) than control schools in the fall of 4th grade and for both 5th grade semesters.

MEASURE: Cafeteria outcomes

OUTCOME: At the beginning and end of the 4th grade, intervention schools had a lower percentage of fat than control schools. This difference disappeared by the 5th grade. Intervention schools met programme goals for fat content in school lunches during the 2nd year of the programme, while control schools did not at any time.

No schools reached programme sodium recommendations.

Donnelly 2009

MEASURES: BMI

OUTCOMES: No significant differences for change in BMI or BMI percentile (baseline to 3 year) for intervention vs control (not influenced by gender).

Schools (n = 9) with ≥ 75min of PAAC/wk showed significantly less increase in BMI at 3 years compared to schools (n = 5) with < 75min (1.8 ± 1.8 vs  2.4 ± 2.0; P = 0.02)

MEASURES: Daily PA (accelerometer)

OUTCOMES: Over a 4‐day average (consecutive), children in PAAC schools had greater levels of PA (13%>) compared to children in control schools (P = 0.007).

Children in PAAC schools had greater levels of PA during the school day (12%>; 0.01) and on weekends (17%>; 0.001) compared with children in control schools.

Children in PAAC schools exhibited 27% higher levels of moderate to vigorous intensity PA (?4 METS) compared with children in control schools (P = 0.001).

MEASURES: Academic achievement measured using Weschler Individual Achievement Test 2nd edition.

OUTCOMES: significant improvement in academic achievement from baseline to 3 years were observed in the PAAC compared with the control schools for the composite, reading, math and spelling scores (all P < 0.01)

Epstein 2001

1. Fatness assessed by percentage of overweight (established by comparing the BMI of the subject with the relevant 50th BMI percentile based on the gender and age of the subject) at baseline and at one year (end of intervention).
OUTCOME: Children showed no significant differences in percentage of overweight with either intervention: increase fruit and vegetable intervention (‐1.10 + 5.29) or decrease high fat/high sugar intervention (‐2.40 + 5.39).

2. Dietary intake:
OUTCOME: High fat/high sugar intake significantly decreased across all children independent of group. Children also showed trends toward greater increases in fruit and vegetable intake for the Increase Fruit and Vegetable group through the one year study.

Fernandes 2009

MEASURES: Nutritional status defined on the basis of BMI for age and sex.

OUTCOMES: No significant changes from baseline in the prevalence of overweight/obesity (BMI?85th percentile) were observed in either group (both p=1.0) with no difference between groups. The percentage of overweight/obese children increased from 21.8 to 23.6% in the intervention group and from 33.7 to 35.0% in the no‐intervention group (P > 0.05).

MEASURES: Frequency of eating foods (either 0‐1 day or 2‐3 days) prohibited by School Canteens Act by self‐report for two 3‐day dietary recalls

OUTCOMES: The percentage of children who ate foods prohibited by the Act on 2‐3 days decreased in both intervention and control groups (not significant)

MEASURES: Distribution of children eating certain foods on the two 3‐day dietary recalls

OUTCOMES: In the control group, the percentage eating mass‐produced snacks increased (P = 0.008), while in the intervention group, this decreased (p=0.016).

In both groups, there was a significant reduction in the intake of artificial juice (P < 0.001) and chocolate (I: P < 0.001; C: P = 0.031).

There was an increase in the percentage of children drinking soda in both intervention and control groups (P = 0.002 and P = 0.016).

For the percentage of children eating yoghurt, there was an increase in the intervention group (P = 0.016) and a decrease in the control group (P = 0.016).

There was a decrease in the percentage of children drinking natural juice in the control group (P < 0.001) and a numerical increase in the intervention group (P = 0.063).

There was an increase in the percentage of children eating fruit in both groups, but this was only significant in the control group (P = 0.016 vs P = 0.25).

Foster 2008

MEASURES: Incidence of overweight and obesity

OUTCOMES: Fewer children in the intervention schools (7.5% [unadjusted mean]) than in the control schools (14.9% [unadjusted mean]) became overweight after 2 years (adjusted odds ratio: 0.67 [0.47 to 0.96]; P = 0.03 [adjusted for gender, race/ethnicity and age]).

No differences between control and intervention groups in the incidence of obesity at 2 years (P = 0.99).

The predicted odds of incidence of either overweight or obesity were 15% lower for the intervention group (odds ratio: 0.85 [0.74 to 0.99]; P <0.05)

MEASURES: Prevalence and remission of overweight and obesity

OUTCOMES: After 2 years, the predicted odds of overweight prevalence were 35% lower for in the intervention group (odds ratio: 0.65 [0.54 to 0.79]; P < 0.0001 [adjusted for gender, race/ethnicity, age]).

Effect was slightly greater in black students who, if receiving the intervention, were 41% less likely to be overweight than those in control schools after 2 years (after controlling for gender, age and baseline prevalence).

No differences between intervention and control groups in the prevalence of obesity after 2 years (P = 0.48).

No difference between intervention and control groups for combined prevalence of overweight and obesity (P = 0.07).

No differences between groups with respect to the remission of overweight or obese after 2 years.

MEASURES: Dietary intake and physical activity sedentary behaviours, potential adverse effects

OUTCOMES: Reported decreases in both intervention and control schools in self‐reported consumption of energy, fat, and fruits and vegetables over 2 years with no differences between groups.

Decreases in self‐reported amounts of physical activity in both intervention and control groups with no differences between groups.

MEASURES: Sedentary behaviours

OUTCOMES: Inactivity was 4% lower after 2 years in the intervention group compared with the control group after adjusting for gender, age, race/ethnicity and baseline inactivity (odds ratio: 0.96 [0.94 to 0.99]; P < 0.01).

Weekday television watching was 5% lower in the intervention group than in the control group (odds ratio: 0.95 [0.93 to 0.97]; P < 0.0001) after 2 years.

MEASURES: Potential adverse effects

OUTCOMES: The intervention showed no evidence of an adverse impact with respect to a worsening body image or changes in incidence, prevalence and remission of underweight.

Gentile 2009

MEASURES: Height and Weight

OUTCOMES: No significant difference in BMI between groups post‐intervention or at 6 months follow‐up

MEASURES: Screen time

OUTCOMES: Child report (hours/week):  No significant difference between groups post‐intervention or at 6 months follow‐up

Parent report (hours/week)  Significantly lower in intervention group post‐intervention  (I: 22.8(0.7), C: 24.6(0.3), P <0.05) and at 6 months follow‐up (I: 23.7(0.5), C: 25.7(0.5), P <0.05) ) compared with control group

MEASURES: Fruit and vegetable consumption

OUTCOMES: Child report (servings/week):  Significantly lower in intervention group post‐intervention  (I: 4.4(0.2), C: 4.2(0.1), P < 0.05) and at 6 months follow‐up (I: 4.1(0.2), C: 4.0(0.1), P < 0.05) compared with control group

Parent report (servings/week)  Significantly lower in intervention group post‐intervention  (I: 24.9(0.7), C: 22.6(0.4), P < 0.05) and at 6 months follow‐up (I: 22.5(0.7), C: 21.3(0.3), P < 0.05) ) compared with control group

MEASURES: Physical activity (steps/day)

OUTCOMES:  No significant difference on pedometer measures of physical activity

Gortmaker 1999a

1. Fatness assessed by repeat measures of height, weight, (and calculated BMI), and triceps skinfold thickness, at baseline and after 18 months (end of intervention) :
OUTCOME: The prevalence of obesity among girls in intervention schools was reduced compared with controls, controlling for baseline obesity (odds ratio, 0.47; 95% confidence interval, 0.24‐0.93; P = 0.03), with no differences found among boys. There was greater remission of obesity among intervention girls vs control girls (odds ratio, 2.16; 95% confidence interval, 1.07‐4.35; P = 0.04).

Reestimated regressions that excluded observations with missing data and got similar results with both approaches.

1. Television viewing time:
OUTCOME: Both girls and boys in the intervention group spent less time viewing television.
2. Dietary intake:
OUTCOME: Intervention girls reported eating more fruit and vegetables and reduced their increase in dietary energy over the two years of the intervention.

Behavioural variables as explanations for intervention effect: Regression indicated that only change in television viewing mediated the intervention effect.

Gutin 2008

(see Notes in Included Studies table)

MEASURES: Percent body fat (%BF)

OUTCOMES:

1 year: %BF decreased in intervention participants with no change in control participants (adjusted change: ‐0.76 [‐1.42, ‐0.09]; P = 0.027). No significant differences between groups for ITT analysis.

Significant relationship between level of programme attendance and change in %BF in intervention group, with greater decreases in %BF observed with higher programme attendance (P = 0.0004).

3 year: Significant group by time interaction (P < 0.05). Intervention group reduced their body fat during school months and this returned to levels similar to those of the control group after the summer months (school vacation).

MEASURES: Bone Mineral Density (BMD)

OUTCOMES:

1 year: Compared with control, intervention participants showed significantly greater gains in BMD (adjusted change: 0.008 [0.001, 0.015]; P = 0.023).

In intervention group greater increases in BMD were observed with higher programme attendance (P = 0.029).

3 year: Significant group by time interaction in favour of intervention participants (P < 0.01).

MEASURES: Fat mass

OUTCOMES:

1 year: No significant differences between groups post‐intervention.

In intervention group greater decreases in fat mass were observed with higher programme attendance (P = 0.0004).

3 year:  No significant differences between groups post‐intervention (data not reported).

MEASURES: Fat‐free soft tissue (FFST)

OUTCOMES:

1 year: No significant differences between groups post‐intervention.

3 year: Significant group by time interaction in favour of intervention participants (P < 0.01).

MEASURES: cardiovascular fitness (CVF)

OUTCOMES:

1 year: Compared with control, intervention participants showed significantly greater gains in CVF (adjusted change: ‐4.4 [‐8.2 to ‐0.6]; P = 0.025).

In intervention group greater increases in CVF were observed with higher programme attendance (P = 0.029).

3 year: Significant group by time interactions in favour of intervention participants (P < 0.01). The intervention group improved in fitness during school months and this returned to levels similar to those of the control group after the summer months.

MEASURES: BMI

OUTCOMES:

1 year: No significant differences between groups post‐intervention.

3 year: Significant group by time interaction, with the increase in BMI being greater in the intervention group than in the control group (P < 0.05).

MEASURES: waist circumference

OUTCOMES:

1 year: No significant differences between groups post‐intervention.

MEASURES: CV risk factors

OUTCOMES:

1 year: No significant differences between groups post‐intervention.

Hamelink‐Basteen 2008

MEASURES: Height, weight

OUTCOMES: BMI increase did not differ between groups post‐intervention

MEASURES: Obesity

OUTCOMES: Prevalence of obesity did not differ between groups post‐intervention

MEASURES: Nutrition knowledge

OUTCOMES:  Higher level of knowledge about importance of vegetables and fruit and a healthy diet

MEASURES: Lifestyle and behaviours

OUTCOMES: Intervention group walked more frequently to school, watched less television, drank less soft drinks and ate less sweets than control group.

Harrison 2006

MEASURES: Height, weight

OUTCOMES: No significant difference between groups post‐intervention

MEASURES:  Physical activity

OUTCOMES: MVPA in 30 min blocks was significantly higher (by 0.84 blocks; 95%CI 0.11, 1.57) in the intervention group post‐intervention (P = 0.03)

MEASURES: Screen time

OUTCOMES: No significant difference between groups post‐intervention

MEASURES: Physical activity self‐efficacy

OUTCOMES: Significantly higher self efficacy (by 0.86 units; 95%CI: 0.16, 1.56) post‐intervention (P = 0.03)

MEASURES: Aerobic fitness (20m shuttle test)

OUTCOMES:  No significant difference between groups post‐intervention

James 2004

MEASURE: BMI at 1 year (end of intervention) and 3 years post‐baseline (or 2‐year follow‐up)
OUTCOME: No differences between intervention and control in the change in BMI from baseline

MEASURE: Proportion of children overweight or obese at 1 year (end of intervention) and 3 years post‐baseline (or 2‐year follow‐up), based on proportion above 91st centile
OUTCOME: At 1 year, the mean percentage of overweight and obese children increased in the control clusters by 7.5%, compared with a decrease in the intervention group of 0.2% (mean difference 7.7%, 95%CI: 2.2% to 13.1%). At 3 years, this difference was smaller and no longer significant (Odds ratio: 0.79 (95%CI: 0.52 to 1.21)).

MEASURE: Carbonated drink consumption at 1 year (end of intervention):
OUTCOME: Children in intervention classes reported fewer carbonated drinks (0.6 glasses fewer compared with an increase in controls of 0.2 (95% CI: 0.1 to 1.3).
MEASURE: Water consumption at 1 year (end of intervention)
OUTCOME: No differences between intervention and control.

Kain 2004

Fatness assessed by repeat measures of height and weight (and calculated BMI) at baseline and after 1 year (end of intervention):
OUTCOME: No differences between intervention and control.
Skinfolds:
OUTCOME: No differences between intervention and control.
Waist circumference:
OUTCOME: decreased significantly in intervention group by a mean of 0.9cm and increased in controls by same amount.
Physical Fitness:
OUTCOME: Shuttle run test and lower back flexibility both improved for boys and girls in the intervention group compared with controls.

1. Dietary assessment: food frequency questionnaire of 16 key items:
OUTCOME: Not reported.
2. Attitudes and behaviours (14 questions about physical activity and some about fruit and vegetable consumption):
OUTCOME: Not reported.

Kipping 2008

MEASURES: Time spent doing screen‐based  activities

OUTCOMES: No statistically significant differences between intervention and control groups.

MEASURES: BMI

OUTCOMES:  No statistically significant differences between intervention and control groups.

MEASURES: Obesity (BMI > 95th percentile)

OUTCOMES:  No statistically significant differences between intervention and control groups. However, subgroup analysis by gender showed that the odds of being overweight post‐intervention were higher in females (1.52; 95%CI: 0.37 to 6.25) than males (0.28; 95% CI: 0.06 to 1.33)

MEASURES: Walks/cycles to and from school

OUTCOMES:  No statistically significant differences between intervention and control groups.

Lazaar 2007

MEASURES: Obesity status

OUTCOMES: A larger proportion of obese children (BMI > 97th percentile) became overweight (90th <BMI <97th percentile) in the intervention group compared with control (16.3%, P < 0.05 versus 9.3%, P < 0.05).

The proportion of non obese children becoming obese or overweight was greater in controls than in the intervention group (14.8%, P < 0.05 versus 2.6%, P= ns)

MEASURES

OUTCOMES

MEASURES: BMI

OUTCOMES: Average BMI remained unchanged over time in both groups overall.

In girls, there was a significant group*time interaction (P < 0.01) and a significant effect of PA intervention between intervention and control in obese (‐1.4% vs 0.9%; P < 0.05) and non obese (‐0.2% vs 2.1%; P < 0.001) girls.

MEASURES: BMI z‐score

OUTCOMES: In boys, BMI z‐score declined significantly over time only in the intervention group and was significantly different compared with controls (P < 0.001). In boys, there was also a significant difference between intervention and control groups in both obese (‐2.8% vs 1.5%; P < 0.05) and non obese boys (‐2.4% vs 2.6%; P < 0.01).

In girls, BMI z‐score declined significantly in all groups except for obese controls. The decrease was higher in the intervention groups compared with control groups for both obese (‐6.8% vs ‐2.4%; P < 0.001) and non obese (‐3.1% vs ‐1.8%; P < 0.01) girls. Changes were greater in obese compared with non obese girls (P < 0.001).

MEASURES: Waist circumference

OUTCOMES: In girls, waist circumference was affected over time, decreasing in the intervention group and increasing in the control group (‐3.3% vs 2.8%; P < 0.001).

In boys, waist circumference was not significantly affected over time.

MEASURES: Skinfold thickness

OUTCOMES: In girls, the sum of skinfolds was significantly decreased over time in the intervention groups in both obese (‐4.4%, P < 0.05) and non obese (‐3.2%, P < 0.001) girls, with a significant difference between obese and non obese girls (P < 0.05) and no significant changes in controls.

In boys, the sum of skinfolds was not significantly altered over time.

MEASURES: Fat‐free mass

OUTCOMES: In girls, fat‐free mass increased over time, with greater increases in intervention children compared with controls for both obese (5.2% vs 2.4%, P < 0.001) and non obese (4.0% vs 0.6%, P < 0.05) girls.

In boys, fat‐free mass improved over time with higher changes in the intervention groups (obese = 6.4%, P < 0.001 and non obese = 3.4%, P < 0.001) compared with control groups (obese = 1.3%, P = ns and non obese = 0.7%, P = ns), and a higher increases in obese boys compared with non obese boys (P < 0.01)

Macias‐Cervantes 2009

MEASURES: Anthropometric measurements: height, weight, BMI, waist circumference, triceps skinfold

OUTCOMES: Differences between groups post‐intervention were not tested

MEASURES: Glucose, triglycerides, cholesterol, HDL‐C, LDL‐C, HOMA‐IR

OUTCOMES: Intervention group decreased insulin (P < 0.001) and HOMA index (4.36 vs. 2.39, P <0.001) from baseline to follow‐up, but no difference in control group. No other differences were reported.   Differences between groups post‐intervention were not tested

 

MEASURES: Physical activity (steps/day, by pedometer)

OUTCOMES:  Intervention group increased their median daily steps from baseline to follow‐up (15,329 to 19,910). Differences between groups post‐intervention were not tested

MEASURES: Food intake

OUTCOMES: Not reported

Marcus 2009

MEASURES: Height and weight

OUTCOMES: Prevalence of overweight/obesity decreased by 3.2% (from 20.3 to 17.1) in intervention schools compared with an increase of 2.8% (from 16.1 to 18.9) in control schools (P < 0.05).

No difference between intervention and control groups in change in zBMI (BMIsds) post‐intervention

A larger proportion of the children who were initially overweight reached normal weight in the intervention group (14%) compared with the control group (7.5%), P < 0.017

MEASURES: Physical activity measured by accelerometry

OUTCOMES: No significant differences between groups post‐intervention

MEASUREs: Eating habits at home measured by parental report.

OUTCOMES: Post‐intervention eating habits at home were healthier among intervention families.   Significant differences between children in intervention and control schools were found for high‐fat dairy products (P < 0.001), sweetened cereals (P<0.02) and sweet products (P < 0.002).

MEASURES: Eating disorders measured by self‐report.

OUTCOMES: No significant differences between groups post‐intervention

Müller 2001

1. Fatness assessed by repeat measures of height and weight.
OUTCOME: No significant difference between I and C from BMI data available at baseline and 1 year. The median of the BMI was 15.2 (intervention school) and 15.4 for children in control schools. At one‐year follow‐up the corresponding data were 16.1 and 16.3 respectively.

2. Triceps skinfold thickness
OUTCOME: Significant difference in favour of the intervention group at one‐year follow‐up (age‐dependent increases in median triceps skinfolds of the whole group (from 10.9 to 11.3mm in ‘intervention schools’ vs from 10.7 to 13.0mm in ‘control schools’, P < 0.01).  Also positive intervention impacts on percentage fat mass of overweight children (increase by 3.6 vs 0.4% per year without and with intervention, respectively; P < 0.05).

1. Nutrition knowledge
OUTCOME: significant increase from 48% to 60% of the children.
2. Daily physical activities
OUTCOME: significant increase from 58 to 65% of the children.
3. Daily fruit and vegetable consumption
OUTCOME: significant increase from 40 to 60% of the children.
4. Daily intake of low fat food
OUTCOME: significant increase in frequency of daily intake of low fat food from 20 to 50%.
5. Decrease in TV watching
OUTCOME: significant decrease from 1.9 to 1.6 h/day.

Paineau 2008

MEASURES: Nutritional intake

OUTCOMES: Compared with controls, participants in the intervention groups achieved their nutritional targets for fat intake and for sugar and complex carbohydrate intake, leading to a decrease in energy intake (children, P<0.001; parents, P = 0.02).

MEASURES: height and weight

OUTCOMES:  No significant differences were found between groups in BMI or zBMI, with  a trend toward negative changes in zBMI in all 3 groups.

BMI differed in parents (group A, +0.13, 95% CI, ?0.01, 0.27; group B, ?0.02, 95% CI, ?0.14, 0.11; control group,+ 0.24, 95% CI, 0.13, 0.34; P =.001), with a significant difference between group B and the control group (P = 0.01)

MEASURES:  Physical activity

OUTCOMES:  In children, changes in physical activity throughout the study did not differ between groups, either for daily screen viewing or for activities in clubs

MEASURES: Food‐related quality of life

OUTCOMES:  In parents, food‐related quality of life did not change differently between groups throughout the study

Pangrazi 2003

1. Fatness assessed by repeat measures of height and weight (and calculated BMI) at baseline and end of pilot:
OUTCOME: No differences between intervention and control.

2. Physical activity: accelerometer CSA,
OUTCOME: All students: PLAY & PE, and PLAY only groups were significantly more active than C. Girls: PLAY & PE, and PE only groups were significantly more active than controls.

None reported.

Reed 2008

MEASURES:  Cardiovascular fitness (measured by 20‐m shuttle run test)

OUTCOMES:  The intervention group demonstrated a significantly greater increase (20%)  in fitness (20‐m shuttle run) compared with the control group (P <0.05).

MEASURES: Blood pressure (systolic and diastolic)

OUTCOMES: Systolic blood pressure in the intervention group decreased significantly compared with an increase in the control group (5.7% smaller increase; P < 0.05). There was no difference for change in diastolic blood pressure.

MEASURES: Total cholesterol, HDL, LDL, Apo B, C‐reactive protein and fibrinogen

OUTCOMES: Although all serum variables in the intervention group decreased more than these same variables for the control group changes failed to reach significance

MEASURES: Weight, height

OUTCOMES: BMI not different between groups post‐intervention

Robbins 2006

MEASURES:  Physical activity variables (frequency, intensity, duration, and  readiness)

OUTCOMES: No significant differences between groups

MEASURES: Physical activity determinants (interpersonal influences, physical activity enjoyment, self efficacy, and perceived benefits and barriers of physical activity)

OUTCOMES:  The intervention group had significantly greater social support across time (P = 0.019).  No other significant differences between groups.

No other significant differences between groups

Robinson 2003

1. Fatness assessed by repeat measures of height and weight (and calculated BMI) at baseline and end of pilot:
OUTCOME: No differences between intervention and control.
2. Waist circumference:
OUTCOME: No differences between intervention and control.
3. Dual X‐Ray Absorptiometry (DEXA) for % Body fat
OUTCOME: Not done

4. Physical activity: accelerometer CSA,
OUTCOME: No differences between intervention and control.
5. a modification of the Self‐Administered Physical Activity Checklist (SAPAC):
OUTCOME: No differences between intervention and control.

6. GEMS Activity Questionnaire(GAQ) computerised
OUTCOME: Not reported

7. Dietary intake measured by two 24 hour recalls using Nutrition Data System computer programme (NDS‐R).
OUTCOME: No differences between intervention and control.

1. TV usage: TV, videotape and video games:
OUTCOME: No differences between intervention and control.
2. Total household TV usage:
OUTCOME: Intervention significantly less than control.
3. Ate breakfast with TV on:
OUTCOME: No differences between intervention and control.
4. Ate dinner with TV on:
OUTCOME: Intervention significantly less than control.

5. Over concern with weight or shape:
OUTCOME: Intervention significantly better than control.

Rodearmel 2006

MEASURES: Steps per day

OUTCOMES: Steps per day increased in all members of intervention families but not in any members of the control families. Intervention target girls/boys, mums and dads, all took significantly more steps per day on average than their control counterparts (P < 0.05).

Increases in steps/day over baseline in intervention groups approached the primary goal of the intervention (an additional 2000 steps/day)

MEASURES: cereal consumption

OUTCOMES: Intervention families consumed approximately 1 serving of cereal/day, double the amount consumed by control families (P < 0.05)

MEASURES: Food Intake

OUTCOMES: No significant changes in self‐reported total energy intake or in intake of any macronutrient in either group.

MEASURES: Body weight/adiposity

OUTCOMES: Significant between‐group differences (P < 0.05) were found pre‐ to post‐study in the difference in the mean change of all body weight/adiposity measures of primary importance for overweight target children (%BMI‐for‐age, % body fat) and their parents (weight, BMI and % body fat). All trends were in favour of the intervention group.

When analysed by gender, significant between‐group differences were found in the difference in the mean change of both the child‐ and adult‐specific body weight/adiposity measures between intervention and control for target girls and mums, but not for target boys and dads.

Sahota 2001

1. Fatness assessed by repeat measures of height and weight (and calculated BMI) at baseline and after 1 year (end of intervention)

OUTCOME: No differences between groups overall (weighted mean difference between intervention and control of 0 (95% CI: ‐0.1 to 0.1) overall, or when analysed by weight status for overweight (WMD: ‐0.07, 95% CI ‐0.22 to 0.08) or obese (WMD: ‐0.05; 95% CI: ‐0.22 to 0.11) children separately.

2. Dietary intake:

OUTCOME: Vegetable consumption by 24 hour recall was higher in children in the intervention group than the control group (weighted mean difference 0.3 portions/day, 95% CI 0.2 to 0.4). Fruit consumption was lower in obese children in the intervention group ( ‐ 1.0, ‐ 1.8 to ‐ 0.2) than those in the control group. The three‐day diary showed higher consumption of high sugar foods (0.8, 0.1 to 1.6)) among overweight children in the intervention group than the control
group.

3. Physical activity:

OUTCOME: Sedentary behaviour was higher in overweight children in the intervention group (0.3, 0.0 to 0.7).

4. Psychological measures:

OUTCOME: small increase in global self‐worth for obese children in the intervention schools.

1. Nutrition knowledge:OUTCOME: Focus groups indicated higher levels of self‐reported behaviour change, understanding and knowledge.

Sallis 1993

1. Fatness assessed by weight, height, BMI, calf and triceps skinfold at baseline and 6, 12, 18 months.
OUTCOME: Little difference in BMI for boys and girls between specialist and teacher led intervention conditions (statistical significance not addressed) at 6, 12 and 18 months.
Small differences in BMI for boys and girls between specialist‐led, teacher‐led conditions and usual physical education control. (statistical significance not addressed) at 6, 12 and 18 months.

None reported

Salmon 2008

MEASURES: BMI

OUTCOMES: Significant reduction in BMI post‐intervention in the BM/FMS group compared with control (average ‐1.88 BMI units less than control; P < 0.01). This was maintained at 6 and 12 month follow‐up.

MEASURES: Weight status

OUTCOMES: On average, those in the BM/FMS group were over 60% less likely to be overweight or obese compared with control (P < 0.05). This was maintained at 6 and 12 month follow‐up.

MEASURES: Physical activity (accelerometer)

OUTCOMES: Compared with controls, FMS group children recorded higher levels and greater enjoyment of PA; and BM children
recorded higher levels of PA across all four time points.Significant average effects over time in favour of the BM and FMS groups compared with control (P < 0.05). This was maintained at 6 and 12 month follow‐up. Gender was a significant moderator, with boys showing greater increases.

MEASURES: Self‐reported screen behaviours

OUTCOMES: Children in the BM group reported 229 min/week more in TV viewing on overage over time compared with control (P < 0.05). These effects were maintained with inclusion of 6 and 12 month follow‐up data.

MEASURES: Self‐reported enjoyment of physical activity (five‐point Likert scale)

OUTCOMES: Children in the FMS group reported higher average enjoyment scores over time compared with those in the control group (P < 0.05).

MEASURES: Mastery of fundamental movement skills

OUTCOMES: No significant intervention effects on FMS z‐scores between baseline and any of the post‐intervention time points. In girls, there was a significant effect, with those in the BM (P < 0.05) and FMS (P < 0.01) groups recording higher average FMS z‐scores compared with those in the control group.

MEASURES: Body Image (five‐point Likert scale)

OUTCOMES: No effects on children’s happiness with their body shape and body weight, or eating to gain weight or lose weight in the last month. When stratified by gender, boys in the FMS group (P = 0.003) and BM/FMS group (P = 0.014) recorded significantly higher satisfaction with their body shape between baseline and all post‐intervention time points compared with control.

Sanigorski 2008

MEASURES: Body weight

OUTCOMES: Children in intervention population gained less weight than in the comparison population (‐0.92kg [‐1.74 to ‐0.11], P = 0.03).

MEASURES: Waist circumference

OUTCOMES: Children in intervention population showed lower increases in waist circumference than in the comparison population (‐3.14cm [‐5.07 to ‐1.22], P = 0.01).

MEASURES: BMI‐z score

OUTCOMES: Children in intervention population showed lower increases in BMI‐z score than in the comparison population (‐0.11 [‐0.21 to ‐0.01], P = 0.04).

MEASURES: Relationship between baseline indicators of children’s household SES and changes in children’s anthropometry.

OUTCOMES: In the comparison population, lower SES was associated with a greater weight gain (statistically significant relationship in 19 of 20 analyses).

In the intervention population, no statistically significant relationships were observed.

MEASURES: BMI

OUTCOMES: No significant difference between intervention and comparison populations (P = 0.20).

MEASURES: waist/height ratio

OUTCOMES: Children in intervention population showed lower increases in waist/height ratio than in the comparison population (‐0.02 [‐0.03, ‐0.004], P = 0.01).

MEASURES: Prevalence and incidence of ow/ob

OUTCOMES: Prevalence of overweight/obesity increased in both groups, and the incidence of overweight/obesity was not significantly different between groups.

Sichieri 2009

MEASURES: change in BMI

OUTCOMES: BMI and weight increased in both groups with no statistically significant differences between groups. Among students overweight at baseline, the intervention group showed greater BMI reduction and this difference was statistically significant among girls (P = 0.009).

MEASURES: carbonated SSB and juice intake

OUTCOMES: Mean intake of sodas per class was reduced in both groups, with reduction being about four times greater in the intervention compared with the control group (‐69ml vs ‐13ml). Carbonated beverage intake was significantly reduced in the intervention group compared with the control group (p=0.03), but fruit juice consumption was slightly increased in the intervention group (P = 0.08).

MEASURES: overweight and obesity

OUTCOMES: For both groups, obesity changed from about 4 to 4.5% with no statistically significant difference between groups.

Simon 2008

MEASURES: BMI

OUTCOMES: intervention students showed a lower increase in BMI (P = 0.01) over time than control students. The differences across groups of the adjusted (by baseline weight status) BMI changes were ‐0.33 (‐0.55 to ‐0.12) at 3 years and ‐0.36 (‐0.60 to ‐0.11) at 4 years.

Cumulative incidence of overweight was lower in the intervention group than in the control group (4.2% vs  9.8% at 4 years; P < 0.01).

Sensitvity analyses conducted using intention to treat population to compare this with analysis using data from only those participants who completed the study and similar results were observed.

MEASURES: Self‐reported leisure physical activity

OUTCOMES:  At 4 years, 79% of intervention students practised at least one supervised physical activity outside school PE classes, compared with 47% of control students (P < 0.001). Supervised leisure physical activity increased in intervention students, whereas it slightly decreased in controls, with a difference across groups of the 4‐year within‐group changes of 66min (95%CI: 34 to 98) per week (P < 0.0001).

MEASURES: TV/video viewing time

OUTCOMES: Intervention students had a greater reduction over time of TV/video viewing than controls (P < 0.01), with a difference in the 4‐year changes of ‐16min (95%CI: ‐29, ‐2) per day.

MEASURES: Active commuting to/from school

OUTCOMES: Slight increase in active commuting observed across both groups.

MEASURES: Self‐efficacy and intention towards physical activity

OUTCOMES: Intervention associated with an increase of self‐efficacy during the first 2 years (P < 0.0001 and 0.01 at 1 and 2 years respectively) and a sustained improvement of intention toward physical activity (P < 0.05).

MEASURES: Cardiovascular risk factors

OUTCOMES: Compared with controls, intervention participants had a higher increase of high‐density lipoprotein‐cholesterol concentration at 4 years and a slight decrease in blood pressure at 2 years. Other biological cardiovascular risk factors were similar between groups over time.

Spiegel 2006

MEASURES: Height, weight

OUTCOMES:  There were significant shifts in BMI in the intervention group, with a 2% reduction in overweight (BMI > 85% for age and sex) in the intervention group.  There was a significant correlation at the 0.01 level between the intervention and a reduced BMI and BMI‐for‐age data showed that 39.4% of the comparison group and 36.4% of the intervention group were either overweight or at risk for overweight when measured at the baseline interval.

Significant shifts in BMI were noted in the intervention group, with a 2% reduction in overweight (BMI _ 85% for age and sex) youth in the intervention group. Student’s t test and Pearson correlations were used to evaluate the significance of the BMI shift. Both analyses showed a significant correlation at the 0.01 level between the intervention and a reduced BMI. Student’s t test mean for the comparison group was 0.5210 (N _ 479; SD _ 1.01610, SE _ 0.04643) and for the intervention group was 0.1606 (N _ 534, SD _ 0.89446, SE _ 0.03871). The Pearson correlation for change in BMI baseline to post‐data measure with treatment (r _ _0.186; N _ 1013) was significant at the 0.01 level (two‐tailed).

BMI‐for‐age data showed that 39.4% of the comparison group and 36.4% of the intervention group were either overweight or at risk for overweight when measured at the baseline interval. There was no significant shift in the comparison group, but there was a notable reduction in the intervention group in overweight and at risk for overweight classification, which was most significant at the at risk for overweight (BMI‐for‐age between 85% to 95%) level.  There was a 16.2% attrition rate in the comparison group (N _ 479 matched measures between baseline to post‐data) and a 13.7% attrition rate in the intervention group (N _ 534 matched).

There was no significant shift in the comparison group, but there was a reduction in the intervention group in overweight and at risk for overweight classification, which was most significant at the ‘at risk for overweight’ (BMI‐for‐age between 85% to 95%) level.

MEASURES:  Fruits and Vegetable Consumption

OUTCOMES:  Post‐intervention, there was an increase in fruit and vegetable consumption

in both groups from baseline levels, with a higher increase in the intervention group.

 

MEASURES: Physical activity levels

OUTCOMES: Physical activity levels in the intervention group increased in both school and home settings.  Post‐intervention, intervention students reported an average of 102.5 min/wk of physical activity during the school day (up from 59min/wk at baseline) and a mean level of 37.42 min/d outside of the school day (up from 22.34 min/d at baseline). Physical activity levels increased slightly in the comparison group in reported levels of light exercise

Stolley 1997

1. Fatness assessed by weight and height at baseline and at 12 weeks (end of the intervention):
OUTCOME: No statistically significant change between intervention and control.

1. Dietary Intake:
OUTCOME: Significant reductions found in intervention mothers' daily saturated fat intakes and percentage of energy from fat when compared to controls. Also intervention girls had statistically significant reductions for percentage energy from fat when compared to controls.

Story 2003a

1. Fatness assessed by repeat measures of height and weight (and calculated BMI) at baseline and end of pilot:
OUTCOME: No differences between intervention and control.
2. Waist circumference:
OUTCOME: No differences between intervention and control.
3. Dual X‐Ray Absorptiometry (DEXA) for % Body fat
OUTCOME: Not done.

4. Physical activity: CSA accelerometer,
OUTCOME: No differences between intervention and control.
5. a modification of the Self‐Administered Physical Activity Checklist (SAPAC),
OUTCOME: Not reported.
6. GEMS Activity Questionnaire(GAQ) computerised
OUTCOME: No differences between intervention and control.

7. Dietary intake measured by two 24 hour recalls using Nutrition Data System computer programme (NDS‐R).
OUTCOME: No differences between intervention and control.

Psychological variables:
1. Over concern with weight or shape:
OUTCOME: Intervention significantly better than control.

2. Diet: Healthy choice Behavioural Intentions:
OUTCOME: Intervention significantly better than control.
3. Self‐Efficacy for Healthy Eating
OUTCOME: No differences between intervention and control.
4. Diet knowledge:
OUTCOME: Intervention significantly better than control.

5. Physical Activity Outcomes Expectations, and a self‐efficacy measure.
OUTCOME: No differences between intervention and control (except physical activity preference).

6. Parental reported diet
OUTCOME: Significant differences with intervention better than control: % energy from fat and low fat food practices.
7. Parental reported physical activity:
OUTCOME: No differences between intervention and control.

Taylor 2008

MEASURES: Weight, Height

OUTCOMES:
Post‐intervention:  Adjusted mean BMI Z‐score was lower in intervention relative to control children by ‐0.12 units (95% CI: ‐0.22 to ‐0.02).

Follow‐Up: Mean BMI z score (and 95% CI) remained significantly lower in intervention children in the whole group (n = 554, ‐0.17; ‐0.25 to ‐0.08) and in the group who underwent at least 1 (n = 389;‐0.19;‐0.24 to ‐0.13) or 2 (n = 256;‐0.21;‐0.29 to ‐0.14) full years of intervention.

MEASURES: Prevalence Overweight and obesity

OUTCOMES:
Post‐intervention:  Although the risk of being overweight or obese (18) at year‐end in intervention compared with control children (odds ratio 0.55; 95% CI: 0.19 to 1.48) did not achieve statistical significance, more intervention children who were overweight at baseline tended to be classified as normal weight at year‐end (12 of 49, 24%) than control children (10 of 65, 15%). Furthermore, 10 of 158 (6%) intervention children became overweight during the year compared with 13 of 112 (12%) control children.

Follow‐up: Intervention children were less likely to be overweight, but only in those who were present for the full intervention (n = 256; Relative Risk: 0.81; 95% CI: 0.69, 0.94).  9 (10%) intervention and 10 (14%) control children became overweight during the 2 yr after the cessation of the intervention project.  12 (30%) intervention and 14 (25%) control children who were overweight at baseline were not overweight at follow‐up.  13 (10%) intervention and 18 (17%) control children became overweight.  20 (30%) intervention and 20 (24%) control children became normal weight

MEASURES: Physical activity

OUTCOMES: Post‐intervention average accelerometry counts at 1 year were 28% (95% CI: 11 to 47%) higher in intervention compared with control children after adjusting for age, sex, baseline values and school.

Intervention children spent less time in sedentary activity (ratio 0.91, P = 0.007) and more time in moderate (1.07, P = 0.001) and moderate/vigorous (1.10, P = 0.01) activity.

MEASURES: Waist circumference, blood pressure, pulse rate

OUTCOME:  No intervention effect was observed

 

Vizcaino 2008

MEASURE: BMI

OUTCOMES: No significant differences between intervention and control groups

MEASURES:  Triceps skin‐fold thickness (TST)

OUTCOMES: Significant reduction in TST in intervention children compared with controls for both boys (‐1.14mm; 95%CI: ‐1.71, ‐0.57; p<0.001) and girls (‐1.55mm; 95%CI: ‐2.38, ‐0.73; p<0.001).

MEASURES:  Percentage body fat

OUTCOMES: Significant reduction in % body fat in girls (‐0.58%; 95%CI: ‐1.04, ‐0.11; p=0.02). No significant differences between intervention and control for boys.

MEASURES:  Blood pressure, total cholesterol, triglycerides, apo A and apo B

OUTCOMES: Compared with controls, intervention children had lower apo B levels and higher apo A‐1 levels.

Intervention was not associated with any significant changes in total cholesterol, triglycerides or blood pressure, with the exception of diastolic blood pressure, which rose in intervention versus control boys.

Warren 2003

1. Fatness assessed by repeat measures of height and weight.
OUTCOME: No significant changes in the rates of overweight and obesity were seen as a result of the 3 different interventions (Be Smart, Eat Smart, Play Smart). Post‐intervention, the change in prevalence of overweight from baseline was ‐1, +5, 0 for the Be Smart, Eat Smart, Play Smart groups, respectively. Post‐intervention, the change in prevalence of obesity from baseline was ‐1, ‐2, 0 for the Be Smart, Eat Smart, Play Smart groups, respectively.

1. Nutrition knowledge:
OUTCOME: all conditions improved their knowledge, I vs C not reported. No gender differences.
2. Diet:
OUTCOME: significant increase in vegetable consumption (P<0.05) and fruit (P<0.01). However, 24h recall showed no significant differences between the groups or genders at base line or at follow‐up.
3. Physical activity:
OUTCOME: No intervention effect was found in either the children's or parents questionnaires.

Figuras y tablas -
Table 3. Results 6‐12 years
Table 4. Results 13‐18 years

Study ID

Primary Outcomes

Secondary Outcomes

Ebbeling 2006

MEASURE: BMI

OUTCOME: Change in BMI was not significantly different between groups (mean ± SE: 0.07 ± 0.14 kg/m2 for intervention group and 0.21 ± 0.15 kg/m2 for control group). This varied according to baseline BMI, with the intervention effect significant in those subjects with baseline BMI > 30 kg/m2 and a significant difference between BMI change in intervention and control subjects among those in the upper baseline‐BMI tertile (‐0.63 ± ‐0.23 kg/m2 vs +0.12 ± 0.26 kg/m2).

MEASURE: Energy intake from SSB (kJ)

OUTCOME: Energy intake from SSB decreased in intervention subjects (‐1201 ± 836 kJ) and this was significantly different from control (‐185 ± 945 kJ) (P < 0.0001)

MEASURE: Noncaloric beverage intake (mL)

OUTCOME: Significant increase in intervention subjects compared with control (p=0.002)

MEASURE: Physical activity (MET)

OUTCOME: No difference between intervention and control

MEASURE: Television viewing (hours)

OUTCOME: No difference between intervention and control

MEASURE: Total media time (hours)

OUTCOME: No difference between intervention and control

Haerens 2006; Table 3

MEASURES: BMI

OUTCOMES:

Prevalence of overweight was not different between groups (baseline:18.5 ± 38.8 and post‐intervention: 18.6 ± 38.9).

MALES: No significant positive intervention effects on BMI were found.

FEMALES: After 1 year of intervention, there was a trend for a significant lower increase in BMI in the intervention group with parental support when compared with the control group (F = 3.04, P < 0.08). After 2 years of intervention, there was a significant lower increase in BMI (F = 12.52, P < 0.05) and BMI z‐score (F = 8.61, P < 0.05) in the intervention with parental support group compared with the control group. There was also a significantly lower increase in BMI z‐score (F = 2.68, P = 0.05) in the intervention with parental support group compared with in the intervention‐alone group.

MEASURES:  Physical activity

OUTCOMES:

MALES: school‐related physical activity increased significantly more in the intervention groups compared with the control group (P < 0.05).  Using accelerometry, there were significantly lower decreases in physical activity of light intensity in the intervention groups (‐6 min/day) compared with the control group (‐39 min/day, P < 0.001). Where

time spent in MVPA remained stable in the intervention group, it significantly decreased (‐18 min/day) in the control group (P < 0.05).

 

FEMALES: Time spent in physical activity of light intensity decreased significantly less in the intervention groups (‐2 min/day) compared with the control group (‐20 min/day, P < 0.05).

MEASURES: fat intake, fruit, water and soft drinks

OUTCOMES:

MALES: No differences between groups

FEMALES: Decreases in fat intake and percent energy from fat were significantly higher in the intervention groups (‐20 g/day) when compared with the control group (‐10g/day, P < 0.05).

In either males or females there were no positive intervention effects on  fruit, water and soft drink consumption

Parental involvement did not increase intervention effects

NeumarkSztainer 2003

The primary outcomes were the feasibility i.e. sustainability and satisfaction of the intervention as assessed by a various satisfaction, behaviour change, personal change and socio‐environmental support variables. All did not achieve significance except:
1. Change in Physical Activity Stage:
OUTCOME: Intervention significantly greater than controls at 8 month follow‐up only.

1. BMI

2. Diet and physical activity related behaviours

OUTCOME: No differences between intervention and control.

Pate 2005

MEASURES: % of girls who reported participating in vigorous physical activity during an average of 1 or more 30‐minute blocks per day during the 3‐day reporting period.

OUTCOMES: At follow‐up, the prevalence of vigorous physical activity was greater in the LEAP intervention schools than in control schools (45% vs 36% P = 0.05) after adjusting for baseline differences. When missing data at follow‐up were imputed by applying a regression method, this prevalence difference increased in statistical significance (P < 0.05).

MEASURES: % overweight or at‐risk for overweight

OUTCOMES: No significant differences between intervention and control schools at follow‐up.

Patrick 2006

MEASURES: Physical activity
OUTCOMES: Both groups improved in all behaviours with no significant difference between intervention and control.

Boys in the intervention group increased their number of active days per week (P = 0.01) compared with control adolescents.

MEASURES: Sedentary behaviours based on a composite self‐report measure including time spent watching television, playing computer/video games, sitting talking on the telephone, and sitting listening to music

OUTCOMES: Significant (P < 0.001) between‐group difference for the change in sedentary behaviours (intervention ‐21% versus control +4.8% in girls and intervention ‐24% versus control +2.4% in boys).

MEASURES: % of energy from fat and servings per day of fruits and vegetables

OUTCOMES: More girls in the intervention group met the guideline for maximum % of daily calories from saturated fat at 12 months. Both groups increased their daily fruit and vegetable intake with no differences between groups.

MEASURES: BMI

OUTCOMES: No differences at 12 months between groups for BMI z scores.

Peralta 2009

MEASURES: Height and weight

 

OUTCOMES: No significant differences between groups post‐intervention

 

 

MEASURES: Waist circumference

 

OUTCOMES: No significant differences between groups post‐intervention

 

MEASURES: Percentage body fat assessed using Tanita body fat analyser

 

OUTCOMES: No significant differences between groups post‐intervention

MEASAURES: Cardiorespiratory fitness (by 20‐metre Multistage Fitness Test)

OUTCOMES: No significant differences between groups post‐intervention

MEASURES: Physical activity measured using Actigraph accelerometers

OUTCOMES: Only significant difference was for intervention boys to have significantly less weekend vigorous physical activity (min/day)  than comparison boys ( ?5.3; 95% CI: ?10.4, ?0.2; P = 0.045)

 

MEASURES: Time spent using small screen recreation measured using the Adolescent Sedentary Activities Questionnaire

OUTCOMES: No significant differences between groups post‐intervention

MEASURES: Sweetened beverage and fruit consumption measured using a validated Food Frequency Questionnaire

OUTCOMES: No significant differences between groups post‐intervention

Singh 2009

MEASURE: BMI

OUTCOME: No significant differences between intervention and control groups

MEASURE: Hip and waist circumference

OUTCOME: After 8 months, there were significant differences in hip circumference for intervention compared with control (mean difference in of 0.53 cm; 95% CI 0.07 to 0.98) in females. In males, the intervention resulted in a significant difference in waist circumference (mean difference, ‐0.57 cm; 95% CI, ‐1.10 to ‐0.05).

At the 20 month follow‐up assessment, waist circumference in boys was significantly lower in the control group. In girls at 20 months, there was no significant difference between intervention and control.

MEASURE: Skinfold thickness

OUTCOME: Significant difference in sum of skinfolds for intervention females compared to control females were observed at 8 months (mean difference ‐2.31cm; 95% CI ‐4.34 to ‐0.28).

In boys, there was a significant intervention effect on triceps (‐0.7mm; 95%CI: ‐1.2 to ‐0.1mm), biceps (‐0.4mm; 95%ci: ‐0.8 to ‐0.1mm) and subscapular (‐0.5mm; 95%CI: ‐1.0 to ‐0.1mm) skinfold thickness at 20 months. In girls, there was a significant intervention effect on biceps skinfold thickness (‐0.7mm; 95%CI: ‐1.3 to ‐0.04mm) and the sum of skinfold thickness at 20 months (‐2.0mm; 95%CI: ‐3.9 to ‐0.1mm)

MEASURE: Aerobic Fitness

OUTCOME: No reported difference between intervention and control at 8 months.

MEASURE: Consumption of sugar‐containing beverages

OUTCOME: While consumption of sugar‐containing beverages was significantly lower among students of the intervention schools at both 8‐ and 12‐month follow‐ups, there were no significant differences at 20 months.

MEASURE: Consumption of high‐energy snacks

OUTCOME: No significant intervention effects.

MEASURE: Screen viewing behaviour

OUTCOME: Numerical differences in screen‐viewing behaviour consistently favoured students from intervention schools at all follow‐up measurements, with statistically significant differences in favour of boys of the intervention group at 20 months (‐25 min/d; 95%CI: ‐50 to ‐0.3 min/d).

MEASURES: Active commuting to school

OUTCOMES: No significant intervention effects.

Webber 2008

MEASURES: Physical activity

OUTCOMES: At 2 years, there was no difference in adjusted MET‐weighted minutes of MVPA between 8th‐grade girls in intervention compared with control schools. At 3 years, 8th‐grade girls in intervention schools had 10.9 more MET‐weighted minutes of MVPA than those in control schools (P = 0.03). The decrease in MET‐weighted minutes of MVPA in intervention schools from 6th grade to 8th grade was 6% compared with 15% in control schools. These differences in physical activity were seen more during weekdays than weekends.

There were differences in the number of MET‐weighted minutes of MVPA among the three largest racial/ethnic groups. After adjusting for 6th‐grade activity differences, both MET‐weighted minutes and unweighted minutes of MVPA were higher for white girls than for African‐American and Hispanic girls at both 2 years and 3 years.

MEASURES: Body composition

OUTCOMES: Changes in triceps skinfold thickness and percent body fat were similar between intervention and control groups

Figuras y tablas -
Table 4. Results 13‐18 years
Comparison 1. Childhood obesity interventions versus control by age groups 0‐5, 6‐12 and 13‐18 years

Outcome or subgroup title

No. of studies

No. of participants

Statistical method

Effect size

1 Standardised mean change in Body Mass Index (BMI/zBMI) from baseline to postintervention Show forest plot

37

27946

Std. Mean Difference (IV, Random, 95% CI)

‐0.15 [‐0.21, ‐0.09]

1.1 0‐5 years

7

1815

Std. Mean Difference (IV, Random, 95% CI)

‐0.26 [‐0.53, 0.00]

1.2 6‐12 years

24

18983

Std. Mean Difference (IV, Random, 95% CI)

‐0.15 [‐0.23, ‐0.08]

1.3 13‐18 years

6

7148

Std. Mean Difference (IV, Random, 95% CI)

‐0.09 [‐0.20, 0.03]

Figuras y tablas -
Comparison 1. Childhood obesity interventions versus control by age groups 0‐5, 6‐12 and 13‐18 years