ARTICLE IN PRESS

Public Health (2007) 121, 510–517

www.elsevierhealth.com/journals/pubh

Original Research

A systematic review of controlled trials of

interventions to prevent childhood obesity and
overweight: A realistic synthesis of the evidence
J.B. Connellya,, M.J. Duasob, G Butlerc

a
Institute of Health Sciences, The University of Reading, London Road, Reading RG1 5AG, UK
b
School of Health and Social Care, The University of Reading, Reading, UK
c
School of Psychology, The University of Reading, Reading, UK

Received 16 May 2006; received in revised form 10 October 2006; accepted 15 November 2006
Available online 30 April 2007

KEYWORDS Summary Background: Preventing childhood overweight and obesity has become a
Obesity; major public health issue in developed and developing countries. Systematic reviews
Prevention; of this topic have not provided practice-relevant guidance because of the generally
Child; low quality of research and the heterogeneity of reported effectiveness.
Controlled clinical Aim: To present practice-relevant guidance on interventions to reduce at least one
trials; measure of adiposity in child populations that do or do not contain overweight or
Exercise obese children.
Design: Systematic review of eligible randomized, controlled trials or controlled
trials using a novel approach to synthesizing the trial results through application of
descriptive epidemiological and realistic evaluation concepts. Eligible trials involved
at least 30 participants, lasted at least 12 weeks and involved non-clinical child
populations.
Results: Twenty-eight eligible trials were identified to 30 April 2006. Eleven trials
were effective and 17 were ineffective in reducing adiposity. Blind to outcome, the
main factor distinguishing effective from ineffective trials was the provision of
moderate to vigorous aerobic physical activity in the former on a relatively
‘compulsory’ rather than ‘voluntary’ basis.
Conclusions: By using a novel approach to synthesizing trials, a decisive role for the
‘compulsory’ provision of aerobic physical activity has been demonstrated. Further
research is required to identify how such activity can be sustained and transformed
into a personally chosen behaviour by children and over the life course.
& 2007 The Royal Institute of Public Health. Published by Elsevier Ltd. All rights
reserved.

Corresponding author. Tel.: +44 118 378 5979; fax:+44 118 378 6808.
E-mail address: jbconn7@aol.com (J.B. Connelly).

0033-3506/$ - see front matter & 2007 The Royal Institute of Public Health. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.puhe.2006.11.015
 ARTICLE IN PRESS
Prevention childhood obesity and overweight
Introduction
The prevalence of overweight and obesity in
children and adolescents is an increasing
problem in the UK,1 Europe,2 Australia3 and the
USA.4 The World Health Organization, using its
criterion of a prevalence above 15%, has declared
overweight and obesity in children a global epi-
demic.5
Previous systematic reviews of controlled
trials of interventions to prevent overweight
and obesity in childhood and adolescents have
not supplied practice-relevant guidance. Summer-
bell et al.6 reviewed 22 controlled trials and
concluded that ‘‘the mismatch between the
prevalence and significance of the condition
and the knowledge base from which to inform
preventions activity continues to be remarkable’’.
Reilly and McDowell7 excluded all but three
trials on the grounds that the design quality
of the rest was too low for their results to be
valid and therefore relevant as guides to practice.
They concluded that ‘‘the evidence base for
interventions in childhood activity, with the aim
of prevention on treatment of obesity, remains
limited. Simple, effective and generalizable inter-
ventions are lacking’’. In contrast, Thomas et al.8
concluded that increasing physical activity rather
than skill development classes should be advo-
cated, and that the number of physical activity
classes in which a student could enrol could be
increased.
Three reasons justify the current systematic
review and distinguish it from those already
published. First, we used both descriptive
epidemiological and realistic evaluation concepts
and procedures to cross-classify and synthesize
the controlled trials we identified. We believe
this approach offers more analytic power in
identifying and explaining heterogeneity of trial
results than a more traditional trial by trial
narrative description. Second, most of the existing
reviews, as with more than half of the Cochrane
reviews on all topics, fail to draw practice-relevant
conclusions,9 as shown by Summerbell et al6 and
Reilly and McDowell.7 Thomas et al.8 are excep-
tional in putting forward a more positive conclu-
sion, but their advice is not very specific. Finally,
realistic evaluation specifies the particular impor-
tance of thoroughly describing the putative me-
chanisms that are expected to underlie the changes
brought about by the intervention.10,11 Thus,
specifying elements that on realistic analysis are
expected a priori to cause a change grounds the
reasons for choosing sub-groups to explore before
synthesizing the evidence.
511
Methods
Search strategy
We used the search strategy and inclusion criteria
described by Summerbell et al.6 applied to the
following electronic databases: Medline, Embase,
Cinhal, PsycINFO up to 30 April 2006. This identified
randomized, controlled trials or controlled trials of
interventions to prevent overweight or obesity in
populations that included non-overweight children
with or without overweight or obese children. Trials
had to include an outcome that measured an index
of adiposity. The ages included were 0–18 years and
the follow-up was at least 12 weeks. We also
searched the reference lists of included trials and
published reviews for potentially relevant studies.
We noted reasons for excluding papers that had
seemed eligible but were found ineligible.
Quality appraisal of trials
Each included paper was read by two researchers
who independently scored them using the quality
checklist devised by Downs and Black;12 any
disagreements were settled by discussion.
Intensity score
Four elements a priori and independent of the
effectiveness of each trial were considered to be
important in providing an estimate of the potential
efficacy of the intervention used in each trial
(Table 1). These elements were specified because
of indications from published reviews or from a
realist perspective on mechanisms. In particular,
‘compulsory’ provision of aerobic physical activity
was distinguished from ‘voluntary’ provision.
Agreement of ‘compulsory’ judgement
Two researchers (JC, MD) classified the physical
activity intervention as ‘compulsory’ or ‘voluntary’
(Table 1) blind to outcome. Chance-corrected
agreement (Kappa statistic)13 was estimated.
Data extraction from the trials
A proforma was constructed to allow a systematic
recording of data from each trial. Data items were
as follows: reference number; first author; year of
publication; country of the trial, design; unit of
analysis; age group; ethnicity; numbers in each
group; description of intervention; results; re-
searcher’s comments; appraisal comments; quality
 ARTICLE IN PRESS
512 J.B. Connelly et al.

Table 1 Elements of an effective intervention: the intensity score.

1. Intervention based on an explicit theory or on pilot studies or formative research Yes ¼ 1

No ¼ 0

2. Intervention implemented by the researchers Yes ¼ 1

No ¼ 0

3. Physical activity judged to be compulsory rather than voluntary: ‘compulsory’ is Yes ¼ 1

defined by more than one of the following: integrated into the curriculum; expectation No ¼ 0
of adherence and attendance; moderate to vigorous aerobic activity, personal goals set
and monitored; course credits given; supervised on level of intensity. ‘Voluntary’ is
defined by more than one of: encouragement to take part; opportunities given to take
part; muscle strengthening or other non-aerobic activity; use of role-modelling; non-
supervised on level of intensity
4. Multi-component intervention Yes ¼ 1
No ¼ 0

score; intensity score; and effectiveness. Consid-

Table 2 Studies included in the systematic review,
erations based upon realism produced four ele- classified by effectiveness on adiposity.
ments that comprised a measure of the ‘intensity’
of the intervention (see above). Effective Ineffective

Simonetti D’Arca et al. Sallis et al. (1993)16

Procedures for synthesizing studies (1986)15
Flores (1995)17a Leupker et al. (1996)19
We used both epidemiological14 and realistic Vandongen et al. Donnelly et al. (1996)20
evaluation10,11 concepts and procedures. Descrip- (1995)18
tive epidemiology clarifies and classifies units of Mo-Suwan et al. Stolley and Fitzgibbon
analysis (here the units are individual studies) by (1998)22a (1997)21
using categories of time, place and person. Robinson (1999)23 Sahota et al. (2001)25
Realistic evaluation examines characteristics be- Gortmaker et al. Epstein et al. (2001)27
(1999)24a
tween studies reporting effective and ineffective
Muller et al. (2001)26 Robinson et al. (2003)28
outcomes. The resulting classifications may show Sallis et al. (2003)29b Caballero et al. (2003)30
patterning, which indicates causal mechanisms and James et al. (2004)37 Newmark-Sztainer et al.
cause-relevant contextual features.10 (2003)31
Kain et al. (2004)39 Harvey-Berino (2003)32
Results Fitzgibbon et al. Warren (2003)33
(2005)40
Baranowski et al.
We identified 28 eligible trials published up to 30
(2003)34
April 2006.15–42 Concerning at least one measure of Beech et al. (2003)35
adiposity, 11 trials reporting a statistically signifi- Pangrazi et al. (2003)36
cant decrease between the intervention compared Dennison et al. (2004)38
with the controlled condition (‘effective’) and 17 Coleman et al. (2005)41
trials that did not show such a difference (‘in- Ramirez-Lopez et al.
effective’) (Table 2). (2005)42
a
Effective in girls only.
Summary and synthesis of the trials b
Effective in boys only.

No apparent association was observed between a

trial’s quality score and its reported effectiveness
regarding a measurement of adiposity. Table 3
summarizes the 28 trials and may be read for
summary purposes or for individual trial design
features and results. For example, the mean age of
children was between 6 and 10 years in 19 trials,
five of these showed an effect on a measure of
adiposity, one trial included boys only and eight
trials included girls only. Of the effective trials,
 ARTICLE IN PRESS
Prevention childhood obesity and overweight
two15,22 included pre-school children (these trials
both used a school setting and one used a physical
intervention),15 and the other used physical educa-
tion information and compulsory physical activity.22
Table 4 (online version only) shows the trials in the
traditional narrative style.
Table 3 shows the results of a descriptive
epidemiological and realist comparison between
‘effective’ and ‘ineffective’ trials. The most strik-
ing difference is in the ‘compulsory physical
activity’ category, which shows that five trials,
which provided relatively compulsory, weekly,
aerobic, physical activity, were effective and no
trial proved ineffective. The probability of such a
5–0 distribution, on a null hypothesis of equal
chances of (in) effectiveness is about 3%. Given the
a priori importance of ‘compulsory’ aerobic physi-
cal activity based upon realist considerations, two
researchers independently classified interventions
as ‘compulsory’ or ‘voluntary’ aerobic physical
activity or no physical activity. The agreement
level was 26/28 (93%), with a chance-corrected
Kappa coefficient (95% confidence interval) or 0.81
(0.36, 1.00), indicating excellent agreement.
Discussion
By applying a novel approach to synthesizing the
results of the 28 controlled trials identified for
inclusion in this systematic review, one character-
istic identified before synthesis and blind to out-
come of the interventions emerged as being
associated with an effective outcome on one or
more measure of adiposity. This characteristic is
the relatively compulsory nature of the provision of
aerobic physical activity. In the five trials that were
classified as ‘compulsory aerobic physical activity’,
five were classified as effective and none ineffec-
tive concerning a measure of adiposity. Compulsory
provision in these trials was defined by more than
one of the following: physical activity integrated
into the weekly curriculum; expected participation
in sessions with or without course credits; expected
participation in sessions with or without individual
goals regarding measures of aerobic fitness; and
specification of moderate to vigorous physical
activity as aerobic activity rather than another
type (e.g. muscle strengthening). ‘Voluntary’ pro-
vision was characterized as more than one of the
following: encouragement to take part in
physical activity; provision of opportunities for
such activities; possible inclusion of role modelling;
and a specification of physical activity other than
aerobic.
513
Realism was used in this systematic review to
‘open up the black box’ of the intervention(s).
Before classifying studies on effectiveness, we
devised four components that we considered to
be relatively independent measures of the ex-
pected effectiveness of the intervention, the
intensity score. These components were: (1) the
theoretical basis of the intervention, if any,
including pilot and formative research; (2) whether
the researchers themselves provided the inter-
vention on the assumption that if they did this
would increase conformity of effort and comp-
liance with the protocol; (3) whether the
provision of physical activity was judged as ‘com-
pulsory’ or ‘voluntary’; here, we believed
that ‘compulsory’ provision would increase adher-
ence and thus contribute to impact; and (4)
whether the intervention had more than one
element (e.g. nutritional education and training
in nutritional skills). This meant that all multi-
element interventions were assured a ‘floor’
intensity score of at least 1. Constructing a score
was achieved by simply adding 1 for presence or 0
for absence of the component. Overall (Table 3) no
association was observed between the intensity
score and the intervention’s effectiveness on
adiposity. When individual elements were ex-
plored, however, ‘compulsory’ provision emerged
as highly associated with effectiveness, and was
found to have an excellent level of non-chance
agreement. On the basis of realism, this association
is here argued to be causal and rational. Causal
because moderate to vigorous aerobic physical
activity is biologically an effective means of body
fat reduction, and rational as the compulsory
nature of this provision is very likely to ensure
adherence.
Our findings support and explain those of
Thomas8 and the differential effectiveness by
gender points to the importance of tailoring the
aerobic physical activity to gender-acceptable
varieties. On inspecting Table 3, the study by
Sallis16 seems to disconfirm the role of relatively
compulsory provision of physical activity.
However, the amount of moderate-to-vigorous
activity per week was only 13–16 min greater in
the intervention group compared with the control
group, and the physical activity required included
muscle and bone strengthening, with aerobic
activity being only one of three activities during
the 30-min session. Consequently, this study did not
demand a relatively high expenditure of aerobic
activity.
Our approach to synthesis reported here is
similar to other approaches. In particular,
Pawson43 has attempted to apply realist concepts
 ARTICLE IN PRESS
514 J.B. Connelly et al.

Table 3 Studies classified by effectiveness on adiposity and by study characteristics.

Characteristic Effective Ineffective

Number of Reference Number Reference number

studies number of studies

Age group (years)a

0–5.9 2 15, 22 2 38, 31
6.0–10.0 5 23, 26, 37, 39, 14 16, 19, 20, 21, 25,
40 27, 28, 30, 33, 34,
35, 36, 41, 42
11.0–13.9 4 17,18, 24, 29 0 –
14.0–17.0 0 – 1 32
Ethnicity
Caucasian 5 15, 18, 23, 26, 3 20, 27, 38
37
Caucasian and black or minority ethnic 3 17, 24, 29 5 16, 19, 25, 32, 34
group
African American 0 – 5 21, 28, 31, 35, 36
Native American 0 – 2 30, 33
Other 3 22, 39, 40 2 41, 42
Gender
Boys and girls 7 15, 18, 23, 26, 12 16, 19, 20, 25, 27,
37, 39, 40 30, 33, 34, 36, 38,
41, 42
Boys only 1 29 0 –
Girls only 3 17, 22, 24 5 21, 28, 31, 32, 35
Setting
Home only 0 – 1 33
Home and School (including pre-school) 3 18, 23, 26 2 30, 31
School only (including pre-school) 8 15, 17, 22, 24, 10 16, 19, 20, 25, 32,
29, 37, 39, 40 34, 36, 38, 41, 42
Clinic 0 – 0 –
Community/other 0 – 4 21, 28, 35, 27
Intervention
Nutrition education only 1 37 1 42
Nutrition education and nutritional skills 0 – 0 –
Physical education information (PE-I) only 1 15 0 –

Physical education information and 0 – 0 –

physical activity (PA-V) (voluntary)

Physical education Information and 1 22 1 16

physical activity (compulsory)
Nutrition education and physical education 5 17, 18, 29, 39, 0 –
Information and physical activity 40
(compulsory)
Nutrition education and nutrition skills and 1 26 12 19, 20, 21, 25, 27,
physical education information and 30, 31, 32, 34, 35,
physical activity (voluntary) 36, 41
Advice and skills for reducing television 2 23, 24 2 28, 38
viewing & PE-I & PA-V
Active parenting 0 – 1 33
Intervention score components
Theory or research based 7 17, 18, 22, 23, 13 16, 19, 20, 21, 23,
24 26, 40 28, 30, 31, 32, 34,
35, 36. 39
Researcher Implemented 4 17, 22, 39, 40 5 21, 31, 32, 34, 35
 ARTICLE IN PRESS
Prevention childhood obesity and overweight 515

Table 3 (continued )

Characteristic Effective Ineffective

Number of Reference Number Reference number

studies number of studies

Compulsory physical activity 6 17, 18, 22, 29, 1 16

39, 40
Multi-component intervention 9 17, 18, 23, 24, 13 19, 20, 21, 25, 28,
26 27, 29, 39, 30 31, 32, 34, 35,
40 36, 38, 41
Quality score
0–4 0 – 0 –
5–9 2 15, 26 1 33
10–14 2 17, 29 4 16, 20, 36, 42
15–19 7 18, 22, 23, 24, 9 21, 25, 27, 30,
37 39, 40 31,34 35, 38, 41
20–24 0 – 3 19, 28, 32
25 and above 0 – 0 –
a
Age group contains the mean age of trial participants but age-group range may overlap.

to synthesis by incorporating earlier work on cing television viewing is supported by two

analytic induction within qualitative analysis ap-
proaches. Although we are aware of these devel-
opments, we have worked with our own
understanding of realism and epidemiology in this
review, and have not used these alternative
approaches. The rigour of this study should be
judged from the transparency of the methods for
synthesis described and from our adherence to
established and accepted methods for identifying
and quality scoring studies.
We conclude the following: first, nutritional
education, nutritional skills training and physical
education do not distinguish effective from inef-
fective interventions regarding reducing childhood
adiposity. However, compulsory rather than volun-
tary aerobic physical activity does distinguish and,
from a realist approach, may be rationally under-
stood as causally related to a decrease in adiposity
in children. Nutritional education and skills training
may reasonably be considered useful in a general
health promotion sense. Moreover, although not a
necessary and sufficient component, nutritional
education and skills may be hypothesized to be a
necessary basis for compulsory physical activity to
be effective. Here, we use a theory of component
causation put forward by Kenneth Rothman.13
Against this hypothesis is the disconfirming trial
by Mo-Suwan,22 which showed an effect of compul-
sory physical activity alone. There also seems to be
a difference in the effectiveness of physical activity
by gender, which implies tailoring the type of
aerobic activity offered. The effectiveness of
decreasing sedentary behaviour in terms of redu-
trials,23,24 and is shown ineffective by two
trials.28,38 Consequently, this approach does not
warrant full support on current evidence. We
recommend that existing policies on physical
activity in schools be reviewed to emphasize the
need for a compulsory level of moderate to
vigorous aerobic activity within the curriculum.
Further research should focus on the design and
sustainability of compulsory aerobic physical activ-
ity, its internalization as a voluntary behaviour and
its durability across the life-course.
Acknowledgements
The authors thank Mrs Chris Stannard for preparing
the manuscript. JC was funded by the Berkshire
Public Health Network.
Appendix A. Supplementary Material
Supplementary data associated with this article can
be found in the online version at doi:10.1016/
j.puhe.2006.11.015.
References
1. Reilly JJ, Dovostly AR. Epidemic in obesity in UK children.
Lancet 1999;354:1874–5.
2. Livingstone MBE. Childhood obesity in Europe: a growing
concern. Public Health Nutr 2001;4:109–16.
 ARTICLE IN PRESS
516
3. Magarey AM, Daniels LA, Boulton TJC. Prevalence of over-
weight and obesity in Australian children and adolescents.
Med J Aust 2001;175:561–4.
4. Jolliffe D. Extent of overweight among US children and
adolescents from 1971 to 2000. Int J Obes Relat Metab
Disord 2004;28:4–9.
5. World Health Organization. Obesity, preventing and mana-
ging the global epidemic. Report of the WHO consultation of
obesity. Geneva: World Health Organization; 1997.
6. Summerbell CD, Waters E, Edmunds LD, Kelly S, Brown T,
Campbell KJ. Interventions for preventing obesity in
children. In: The cochrane database of systematic reviews
2005, Issue 3. Chichester: Wiley; 2005.
7. Reilly JJ, McDowell ZC. Physical activity interventions in the
prevention and treatment of paediatric obesity: systematic
review and critical appraisal. Proc Nutr Soc 2003;62:611–9.
8. Thomas H, Ciliska DM, Micucci S, et al. Effectiveness of
physical activity enhancement and obesity prevention
programs in children and youth. Hamilton, Ontario: Effec-
tive Public Health Practice Project; 2004.
9. Petticrew M. Why certain systematic reviews reach uncer-
tain conclusions. BMJ 2003;326:756–8.
10. Pawson R, Tilley N. Realistic evaluation. London: Sage;
1997.
11. Connelly J. Realism in evidence based medicine: interpret-
ing the randomised controlled trial. J Health Organ Manag
2004;18:70–81.
12. Downs SH, Black N. The feasibility of creating a checklist for
the assessment of the methodological quality both of
randomised and non-randomised studies of health care
interventions. J Epidemiol Community Health 1997;52:
377–84.
13. Altman DG. Practical statistics for medical research.
London: Chapman & Hall; 1999.
14. Rothman KJ, Greenland S. Causation and causal inference.
In: Rothman KJ, Greenland, editors. Modern epidemiology.
2nd ed. Philadelphia, PA: Lippincott-Ravenl; 1998.
15. Simonetti D’Arca A, Tarsitani G, Cairella M, Siani V, De
Filippis S, Mancinelli S, et al. Prevention of obesity in
elementary and nursery school children. Public Health
1986;100:166–73.
16. Sallis JF, McKenzie TL, Alcaraz JE, Kolody B, Hovell MF, Nader
PR. Project SPARK. Effects of physical education on adiposity
in children. Ann N Y Acad Sci 1993;699:127–36.
17. Flores R. Dance for health: improving fitness in African
American and hispanic adolescents. Public Health Rep
1995;110:189–93.
18. 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.
19. Luepker RV, Perry CL, McKinlay SM, Nader PR, Parcel GS,
Stone EJ, et al. Outcomes of a field trial to improve
children’s dietary patterns and physical activity. JAMA
1996;275:768–76.
20. Donnelly JE, Jacobsen DJ, Whatley JE, Hill JO, Swift LL,
Cherrington A, et al. Nutrition and physical activity program
to attenuate obesity and promote physical and metabolic
fitness in elementary school children. Obes Res
1996;4:229–43.
21. Stolley MR, Fitzgibbon ML. Effects of an obesity prevention
program on the eating behaviour of African American
mothers and daughters. Health Educ Behav 1997;24:152–64.
22. Mo-suwan L, Pongprapai S, Junjana C, Puetpaiboon A.
Effects of a controlled trial of a school-based exercise
J.B. Connelly et al.
program on the obesity indexes of preschool children. Am J
Clin Nutr 1998;68:1006–11.
23. Robinson TN. Reducing children’s television viewing to
prevent obesity: a randomised controlled trial. JAMA
1999;282:1561–7.
24. Gortmaker SL, Peterson K, Wiecha J, Sobol AM, Dixit S, Fox
MK, et al. Reducing obesity via a school-based interdisci-
plinary intervention among youth. Arch Pediatr Adolesc Med
1999;153:409–18.
25. Sahota P, Rudolf MC, Dixey R, Hill AJ, Barth JH, Cade J.
Evaluation of implementation and effect of primary school
based intervention to reduce risk factors for obesity. BMJ
2001;323:1027–9.
26. Mueller MJ, Asbeck I, Mast M, et al. Prevention of
Obesity—more than an intention. Concept and first results
of the Kiel Obesity Prevention Study (KOPS). Int J Obes
2001;25(Suppl 1):S66–74.
27. Epstein LH, Gordy CC, Raynor HA, Beddome M, Kilanowski
CK, Paluch R. Increasing fruit and vegetable intake and
decreasing fat and sugar intake in families at risk for
childhood obesity. Obes Res 2001;9:171–8.
28. Robinson TN, Killen JD, Kraemer HC, Wilson DM, Matheson
DM, Haskell WL, et al. Dance and reducing television viewing
to prevent weight gain in African–American girls: the
Stanford GEMS pilot study. Ethn Dis 2003;13(Suppl 1):
S65–77.
29. Sallis JF, McKenzie TL, Conway TL, Elder JP, Prochaska JJ,
Brown M, et al. Environmental interventions for eating and
physical activity. A randomized controlled trial in middle
school. Am J Prev Med 2003;24:209–17.
30. Caballero B, Clay T, Davis SM, Ethelbah B, Rock BH, Lohman
T, et al. Pathways Study Research Group. Pathways: a
school-based, randomized controlled trial for the preven-
tion of obesity in American Indian schoolchildren. Am J Clin
Nutr 2003;78:1030–8.
31. Neumark-Sztainer D, Story M, Hannan PJ, Rex J. New Moves:
a school-based obesity prevention program for adolescent
girls. Prev Med 2003;37:41–51.
32. Harvey-Berino J, Rouke J. Obesity prevention in preschool
Native-American children: a pilot study using home visiting.
Obes Res 2003;11:606–11.
33. 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 Promot Int
2003;18:287–96.
34. Baranowski T, Baranowski JC, Cullen KW, Thompson DI,
Nicklas T, Zakeri IE, et al. The fun, food, and fitness project
(FFFP): the Baylor GEMS pilot study. Ethn Dis 2003;13(Suppl
1):S30–9.
35. Beech BM, Klesges RC, Kumanyika SK, Murray DM, Klesges L,
McClanahan B, et al. Child- and parent-targeted interven-
tions: the Memphis GEMS pilot study. Ethn Dis 2003;13
(Suppl 1):S40–53.
36. Pangrazi RP, Beighle A, Vehige T, Vack C. Impact of
promoting lifestyle activity for youth (PLAY) on children’s
physical activity. J Sch Health 2003;73:317–21.
37. James J, Thomas P, Cavan D, Kerr D. Preventing childhood
obesity by reducing consumption of carbonated drinks:
cluster randomised controlled trial. BMJ 2004;328:1237.
38. Dennison BA, Russo TJ, Burdick PA, Jenkins PL. An interven-
tion to reduce television viewing by preschool children. Arch
Pediatr Adolesc Med 2004;158:170–6.
39. Kain J, Uauy R, Albala, Vio F, Cerda R, Leyton B. School-
based obesity prevention in Chilean primary school children:
methodology and evaluation of a controlled study. Int J Obes
2004;28:483–93.
 ARTICLE IN PRESS
Prevention childhood obesity and overweight
40. Fitzgibbon ML, Stolley MR, Schiffer L, Van Horn L, Kaufer
Christoffel K, Dyer A. Two-year follow-up results for hip-hop
to health jr:a randomized controlled trial for overweight
prevention in preschool minority children. J Pediatr
2005;146:618–25.
41. Coleman KJ, Tiller CL, Sanchez J, Heath EM, Sy O, Milliken
G, et al. Prevention of the epidemic increase in child risk of
overweight in low-income schools. The El Paso coordinated
517
approach to child health. Arch Pediatr Adolesc Med
2005;159:217–24.
42. Ramirez-Lopez E, Grijalva-Haro MI, Valencia ME, Antonio
Ponce J, Artalejo E. Effect of a school breakfast program on
the prevalence of obesity and cardiovascular risk factors in
children. Salud Publica Mex 2005;47:126–33 [in Spanish].
43. Pawson R, Greenhalgh T, Harvey G, et al. Realist synthesis:
an introduction. ESRC Research methods programme, Uni-
versity of Manchester. RMP methods paper 2/2004.