Discussion Post – Childhood Obesity

PEDIATRIC REVIEW

Effective behaviour change techniques in the prevention and management of childhood obesity J Martin1, A Chater2 and F Lorencatto3

Rates of childhood obesity are increasing, and it is essential to identify the active components of interventions aiming to prevent and manage obesity in children. A systematic review of behaviour change interventions was conducted to find evidence of behaviour change techniques (BCTs) that are most effective in changing physical activity and/or eating behaviour for the prevention or management of childhood obesity. An electronic search was conducted for randomised controlled trials published between January 1990 and December 2009. Of 4309 titles and abstracts screened, full texts of 135 articles were assessed, of which 17 published articles were included in this review. Intervention descriptions were coded according to the behaviour-specific CALO-RE taxonomy of BCTs. BCTs were identified and compared across obesity management (n¼9) vs prevention (n¼8) trials. To assess the effectiveness of individual BCTs, trials were further divided into those that were effective (defined as either a group reduction of at least 0.13 body mass index (BMI) units or a significant difference in BMI between intervention and control groups at follow-up) vs non-effective (reported no significant differences between groups). We reliably identified BCTs utilised in effective and non-effective prevention and management trials. To illustrate the relative effectiveness of each BCT, effectiveness ratios were calculated as the ratio of the number of times each BCT was a component of an intervention in an effective trial divided by the number of times they were a component of all trials. Results indicated six BCTs that may be effective components of future management interventions (provide information on the consequences of behaviour to the individual, environmental restructuring, prompt practice, prompt identification as role model/position advocate, stress management/emotional control training and general communication skills training), and one that may be effective in prevention interventions (prompting generalisation of a target behaviour). We identified that for management trials, providing information on the consequences of behaviour in general was a feature of non-effective interventions and for prevention trials, providing information on the consequences of behaviour in general, providing rewards contingent on successful behaviour and facilitating social comparison were non-effective. To design effective behaviour change programmes for the prevention and management of childhood obesity, we would recommend utilising the BCTs identified as effective in this review. The impact on intervention effectiveness of combining BCTs should be the topic of further research.

International Journal of Obesity (2013) 37, 1287–1294; doi:10.1038/ijo.2013.107

Keywords: childhood obesity; prevention; management; behaviour change techniques; taxonomy; effectiveness

INTRODUCTION Given the dramatic rise in childhood obesity over the last decade,1–3 it is essential to design interventions that are effective in preventing and managing childhood obesity. Behaviour modification or lifestyle change has become a burgeoning avenue of investigation in this area. Systematic reviews of interventions and clinical guidelines clearly indicate that successful interventions for preventing and managing obesity in children are complex and multi-component—aimed at changing both physical (or sedentary) activity and diet or healthy eating (for example, references 4–8) and comprise multiple, potentially interacting methods of changing behaviour. Despite the vast amount of active investigation in this area, little research has deconstructed interventions and identified which individual components are most successful in changing obesity- related health behaviour in children. A lack of consistent terminology for defining intervention components has also hindered interpretation and replication. It is currently unclear what the active ingredients for bringing about obesity-related behavioural change are and thus, evidence to support the content

of effective obesity prevention and management interventions in children is still relatively weak.4,9–11

There is no lack of evidence that positive changes in behaviours linked to obesity can be effective in preventing and managing obesity in children. Hill et al.12 report that even small changes in behaviour that amount to a decrease in calorie intake of only 100 Kcal per day can effectively prevent weight gain. Randomised controlled trials (RCTs) of lifestyle interventions have reported moderate success in changing obesity-related behaviour in prevention trials4 and produce potentially clinically significant reductions in overweight in management trials.6,13 It is important that these positive findings can be replicated, thus, there is clear value in identifying the specific behaviour change techniques (BCTs) used in such interventions that are effective in achieving and sustaining behavioural change. The complex and multi- component nature of such interventions further underlines the need for developing methods to systematically deconstruct intervention content. This is a particularly imperative task for the successful prevention and management of obesity in childhood and adolescence, as it is important to identify specific BCTs that

1Department of Psychology, University of Sheffield, Sheffield, UK; 2School of Pharmacy, University College London, London, UK and 3Department of Clinical, Educational and Health Psychology, University College London, London, UK. Correspondence: Dr J Martin, Department of Psychology, University of Sheffield, Western Bank, Sheffield S10 2TN, UK. E-mail: [email protected] Received 8 November 2012; revised 29 May 2013; accepted 2 June 2013; accepted article preview online 12 June 2013; advance online publication, 2 July 2013

International Journal of Obesity (2013) 37, 1287 – 1294 & 2013 Macmillan Publishers Limited All rights reserved 0307-0565/13

www.nature.com/ijo

are particularly effective for this age group that may be distinct from those used in adult interventions. There may also be an important distinction between BCTs that are effective in preventing childhood obesity and those that manage obesity in children who are already overweight or obese.

Formally evaluating the content of interventions is inherently difficult. Recent guidelines in the field of health psychology aim to establish more open and detailed reporting of interventions to aid the scientific development of behaviour change interventions (CONSORT;14 Workgroup for Intervention Development and Evaluation Research: WIDER15), though these are not universally adhered to and current reporting of intervention content is generally poor and inconsistent,16 thus limiting what we can learn about behaviour change.17

Categorising the components of behaviour change interven- tions in the field of childhood obesity is complex and attempts to define BCTs are generally idiosyncratic (and potentially non-replicable). For example, some authors have developed their own classifications for extracting such data: Kamath et al.4

extracted data concerning ‘informational, cognitive, behavioural, environmental and social support components’ (p 4607) of interventions, and Sargent et al.18 extracted information on ‘strategies to achieve behaviour change’ (p 4). Such classifications lack the benefit of precise and standard definitions that allow authentic replication and useful comparisons of BCTs across intervention trials. More wholesome descriptions of common behaviour change methods in obesity interventions do exist (see, for example, Bennett and Sothern5, Johnston et al.19 and Hardeman et al.20) but it is not always clear how (or if) these descriptions apply to interventions aimed at children.

As part of the advancement of a science of behaviour change, Abraham and Michie21 developed a taxonomy of 26 conceptually distinct component BCTs, which are described using consistent terminology and standard definitions. This taxonomy has been reliably applied to identify and categorise BCTs featured in obesity prevention programmes involving parents and children22 and in healthy eating and physical activity interventions with adults.23

More recently, the original taxonomy21 has been revised and extended into a behaviour-specific taxonomy of 40 BCTs for physical activity and healthy eating behaviours (CALO- RE24). The CALO-RE taxonomy was developed by three teams of researchers to provide standard definitions to reliably identify techniques used for (1) increasing physical activity and healthy eating in obese adults with additional risk factors for morbidity and (2) increasing self-efficacy to promote lifestyle and recreational physical activity. A total of 72 studies across a variety of populations, behaviours and settings were used as a basis of developing the taxonomy. The authors argue that the CALO-RE taxonomy is more comprehensive than the original, with fewer conceptual problems and less overlap between items, as well as clearer labels and definitions. This taxonomy can be used to analyse the content of behaviour change interventions in depth and also provides a means of improving reporting and aiding replication attempts by specifying BCTs. For the purposes of this review, this taxonomy was used to identify BCTs in physical activity and healthy eating interventions for the prevention and management of childhood obesity. To our knowledge, this taxonomy has not yet been applied to obesity-related behaviour change interventions with children and adolescents.

We conducted a systematic review to select RCTs of childhood obesity interventions that utilised BCTs as a means to prevent or manage obesity in children and adolescents. From this, we coded intervention descriptions using the CALO-RE taxonomy to identify BCTs that were used in such interventions. To work towards aiding intervention designers in the field to build effective BCTs into childhood obesity interventions, this paper aims to (1) identify and code BCTs used in a sample of prevention and management interventions for childhood obesity using the CALO-RE taxonomy

and (2) establish which individual BCTs are components of effective interventions to manage and prevent childhood obesity.

METHODS Study selection A systematic review was conducted of RCTs of obesity management or prevention conducted with children and adolescents (aged 2–18 years) that assessed the impact of interventions including at least one BCT from the CALO-RE taxonomy compared with a no-treatment control group. Interventions that solely tested the impact of physical activity, education and/or calorie controlled diets with no behaviour change element were excluded, as were interventions that combined drug treatment or surgery with BCTs. Interventions could be carried out in any setting (for example, school, clinic and community) and be delivered by any professional (for example, health professional, teacher and therapist) over any length of time. It is important that desired health outcomes are maintained over time therefore to assess the maintenance of intervention effects; we selected only interventions that reported results 6 months beyond the point when active intervention ended. The Transtheoretical Model25

argues that long-term changes in behaviour may be assessed only after this time period has elapsed. The primary outcome measure was body mass index (BMI). BMI is the only indirect measure of adiposity in childhood that has been shown to be associated with future risk of mortality from coronary heart disease in adulthood10 and can be used as a physiological proxy measure of later health outcomes.26–28 BMI was also the only single comparative outcome reported across trials.

Search strategy and results An electronic search was conducted in the following electronic databases: MEDLINE, EMBASE, PsycINFO, Cochrane library (Cochrane Central Register of Controlled Trials), HMIC (Health Management Information Consortium), AMED (Allied and Complementary Medicine Database) and CINAHL (Cumulative Index to Nursing and Allied Health Literature) for RCTs published between January 1990 and December 2009. Search terms (available as a Supplementary electronic file) covered the concepts of ‘obesity’, ‘children’, ‘behaviour change interventions’ and ‘BMI’ and closely followed the search strategies of similar reviews (for example, Kamath et al.4 and Oude Luttikhuis et al.6) and published guidelines for identifying RCTs.29

One reviewer (JM) screened 4333 unique titles and abstracts for eligibility and a second reviewer (AC) screened a random (10%) sample of records, yielding 100% agreement between reviewers. After initial screening of title and abstract, full texts of 135 potentially relevant studies were screened for eligibility independently by the two reviewers. Three unpublished dissertations were unavailable for further screening and were excluded. We explored the unpublished literature and received a good response from experts, but no further studies were included in the review from these avenues. Agreement regarding inclusion between reviewers was 70%, and disagreements were resolved through careful discussion. Seventeen published manuscripts were included in this review.30–46

The selection process for studies is presented in Figure 1.

Coding of BCTs To obtain more complete intervention descriptions and assist data extraction for BCTs, where intervention protocols were published or available elsewhere, the manuscript was located (n¼11;47–57) and the intervention characteristics were coded from both sources. The authors of the further six intervention studies were contacted (on up to two occasions) with a request for copies of the corresponding intervention protocol or any additional documents detailing the intervention content. Three authors did not reply; one author could not locate the original protocol; one author informed us that the protocol was available in Finnish language only, and another author provided further details of the intervention mechanisms, published elsewhere on which the intervention characteristics were coded.

The content of interventions was assessed by two reviewers (JM and AC) who independently coded the descriptions of each intervention using the CALO-RE taxonomy24 for inclusion of BCTs. Intervention descriptions were read line-by-line and assigned a BCT label from the taxonomy where appropriate. A stringent coding method was applied so that in cases where further information was required to assess whether a BCT was present or

Effective BCTs in prevention and management of childhood obesity J Martin et al

1288

International Journal of Obesity (2013) 1287 – 1294 & 2013 Macmillan Publishers Limited

absent, it was coded as absent. We chose not to seek further clarification from the authors, as we wanted to assess published information only.

The two coders practised coding on eight intervention studies not selected for the review, and discussed the disagreements. All interventions were then coded independently and inter-rater reliability, assessed using percent agreement, was high (93%). Disagreements between the coders were discussed at length and a final decision on which BCTs were assigned to interventions was agreed. On the basis of the nature of disagreements, we refined BCTs in the taxonomy (see electronic Supplementary material Table 1 for a summary of our revisions to the CALO-RE taxonomy). To summarise the revision process, we took the following steps: (1) we revised descriptions of techniques where agreement was reduced due to misapplication of the code; (2) we added examples specific to our sample of childhood obesity trials within technique definitions; (3) we clarified the difference between similar codes where we had encountered the disagreements and (4) one additional technique was identified and defined. We verified that the revised taxonomy was also effective in categorising BCTs in the same set of papers by having a third independent coder (FL) repeat the coding task using the revised taxonomy. Agreement remained high (88%) showing that these constructs exist independently in the selected set of papers.

Assessing intervention and BCT effectiveness We divided coded interventions into those that aimed to prevent (N¼8) vs manage (N¼9) childhood obesity. Prevention trials included both overweight and normal weight participants and management trials included overweight participants only.

We then divided up prevention and management trials into effective vs non-effective using BMI outcome data. Effective management trials (N¼6) were defined as trials in which the standardised difference in the mean value of BMI between groups at follow-up was at least X�0.13 (this was the average effect size demonstrated from meta-analysis data58). Less stringent criteria were applied to prevention studies to take into consideration that not all of these trials targeted weight loss measured by a reduction in BMI. The criteria for assessing effectiveness in prevention trials was defined as a significant difference (Po0.05) in BMI at follow-up between groups (N¼4).

To assess the effectiveness of BCTs, and to illustrate the relative weight of each BCT taking into consideration it potentially being a component of both effective and non-effective trials, a percentage effectiveness ‘ratio’ was calculated as the ratio of the number of times each BCT was a component of an intervention in an effective trial divided by the number of times they were a component of all trials, including non-effective trials.

RESULTS First, we present the BCTs coded in effective and non-effective prevention and management interventions. Second, we present

differences in BCTs and effectiveness ratios between effective and non-effective interventions for prevention and management trials.

BCTs in effective and non-effective prevention and management interventions Out of the 40 BCTs in the taxonomy, we agreed that there was no (or insufficient) evidence that 11 of them were present in any of the interventions (representing 100% agreement between coders). For the remaining 29 BCTs, inter-rater reliability was good59 for 17 (average kappa value¼0.71 (range ¼0.485–1.00), average percentage agreement¼92% (range¼71–100%)) and sub- optimal for the remaining 12 due to missing data and low counts of instances of BCTs across the studies. In the light of this, disagreements were discussed at length between the coders before final codes were applied. The coders also agreed a new code, which was added to the CALO-RE taxonomy (exposure to healthy choices) from our revisions (n¼41 BCTs in total in revised CALO-RE taxonomy). This BCT was encountered as an additional method of achieving behaviour change within the pool of studies (see Supplementary file). The coding exercise did not identify any BCTs in any of the control groups.

Table 1 presents BCTs (and their frequency) coded in effective and non-effective prevention and management trials. BCTs that were unique features of management trials (that is, were not used in prevention trials) were prompt review of behavioural goals, relapse prevention/coping planning, shaping, provide information on where and when to perform the behaviour, environmental restructuring, prompt self-talk and motivational interviewing. BCTs that were unique to prevention trials were provide normative information about others’ behaviour, prompt rewards contingent on effort or progress towards behaviour, action planning, provide feedback on performance, teach to use prompts/cues and facilitate social comparison.

Effective BCTs in obesity management All but three out of the nine obesity management inter- ventions selected for review were effective according to our criteria.30–32,40,42,43 An average of 7.5 (range¼3–15) BCTs from the CALO-RE taxonomy were identified at least once across these interventions (effective and non-effective) and there was little difference in mean number of BCTs coded in effective (M¼8) vs non-effective (M¼7) interventions. Twenty-four out of forty-one (59%) BCTs were identified in at least one of the interventions. Thirteen of these techniques were unique to effective trials, two were unique to non-effective interventions and nine appeared at least once in both effective and non-effective trials. Figure 2 presents the ratio of effectiveness for BCTs appearing two or more times in trials. In order that the ratio was meaningful and to be satisfied that there was ‘evidence’ of effectiveness for individual BCTs, we required that a BCT must be a feature of two or more trials; therefore, the five unique BCTs appearing only once in effective trials were excluded. These criteria have been used in previous studies assessing BCT effectiveness.60,61 Six BCTs achieved 100% effectiveness ratios (provide information on consequences of behaviour to the individual, environmental restructuring, prompt identification as role model/position advocate, stress management/emotional control training, general communication skills training and prompt practice) and one BCT (provide information on consequences of behaviour in general) had a 100% non-effective ratio.

Effective BCTs in obesity prevention According to our effectiveness criteria, four trials reported in favour of the intervention33–35,46 and four trials reported no difference between groups.36–39 An average of eight BCTs (range¼1–12) from the CALO-RE taxonomy were identified at

6668 records identified through electronic database search

24 additional records identified through experts and unpublished sources

4333 records after duplicates removed

4333 records screened 4198 records excluded

135 of full-text articles assessed for eligibility

118 articles excluded: Insufficient follow-up length = 54 No BCTs = 15 Non randomised = 31 No control group = 15 Unavailable dissertations = 3

17 studies included in this review

Figure 1. Flow chart of trials selected for review.

Effective BCTs in prevention and management of childhood obesity J Martin et al

1289

& 2013 Macmillan Publishers Limited International Journal of Obesity (2013) 1287 – 1294

Table 1. Frequency of Behaviour Change Techniques coded in effective and non-effective obesity prevention and management trials

Behaviour Change Techniques Prevention Management

Effective Non-effective Effective Non-effective

Provide information on consequences of behaviour in general 0 2 0 1 Provide information on consequences of behaviour to the individual 3 2 5 0 Provide normative information about others’ behaviour 0 1 0 0 Goal setting (behaviour) 2 3 4 3 Action planning 0 1 0 0 Barrier identification 1 1 2 3 Set graded tasks 1 0 1 0 Prompt review of behavioural goals 0 0 3 1 Prompt rewards contingent on effort or progress towards behaviour 1 1 0 0 Provide rewards contingent on successful behaviour 0 2 3 3 Shaping 0 0 1 0 Prompting generalisation of a target behaviour 2 0 1 0 Prompt self-monitoring of behaviour 4 3 3 1 Provide feedback on performance 1 1 0 0 Provide information on where and when to perform the behaviour 0 0 1 0 Provide instruction on how to perform the behaviour 1 2 3 2 Model/demonstrate the behaviour 2 1 1 0 Teach to use prompts/cues 0 1 0 0 Environmental restructuring 0 0 3 0 Agree behavioural contract 1 0 1 3 Prompt practice 3 1 2 0 Facilitate social comparison 0 2 0 0 Plan social support/social change 2 3 2 1 Prompt identification as role model/position advocate 1 0 3 0 Prompt Self talk 0 0 1 0 Relapse prevention/coping planning 0 0 3 1 Stress management/emotional control training 1 0 2 0 Motivational interviewing 0 0 0 1 General communication skills training 2 2 2 0 Exposure to healthy choices 1 1 1 0

Figure 2. Percentage effectiveness of BCTs in obesity management trials.

Effective BCTs in prevention and management of childhood obesity J Martin et al

1290

International Journal of Obesity (2013) 1287 – 1294 & 2013 Macmillan Publishers Limited

least once across these interventions and there was again little difference in the mean number of BCTs coded between effective (M¼8.5) vs non-effective (M¼7.5) trials. Twenty-four out of forty- one (59%) BCTs were identified in at least one of the effective and non-effective interventions. Only one of these BCTs (prompt generalisation of a target behaviour) was unique to effective trials, seven were unique to non-effective interventions and sixteen appeared at least once in both effective and non-effective trials. Again, an effectiveness ratio was calculated to illustrate the relative weight of each BCT appearing in two or more trials (Figure 3). Only prompt generalisation of a target behaviour achieved a 100% effectiveness ratio. Three BCTs were shown to be 100% non-effective (provide information on the consequences of behaviour in general, provide rewards contingent on successful behaviour and facilitate social comparison).

DISCUSSION This review is the first to utilise a behaviour-specific taxonomy of BCTs to formally and systematically identify the components of childhood obesity interventions. We linked individual BCTs with positive health and behavioural outcomes, by assessing effective- ness ratios, thus providing some evidence for the inclusion of particular BCTs as active agents of change in interventions aiming to prevent and manage childhood obesity in the long term. We reliably identified BCTs utilised in prevention and manage- ment trials and identified BCTs that were unique features of each. Effectiveness ratios demonstrated an evidence base for the inclusion of BCTs that were unique features of effective interventions. Six BCTs were identified as uniquely effective (that is, achieved 100% effectiveness ratios) in obesity management interventions and one BCT for obesity prevention interventions. Effectiveness ratios also demonstrated BCTs that were compo- nents of non-effective trials. One BCT was identified as uniquely non-effective for obesity management interventions and three BCTs for prevention interventions.

There is currently little evidence supporting the inclusion (or not) of BCTs in childhood obesity programmes. It is unknown in particular (1) whether individual BCTs are more effective in the prevention or management of obesity, (2) whether (and how) the content of programmes may differ to those applied to adults and (3) if particular BCTs may be more or less effective when applied to specific target groups (for example, parents vs children) or intervention modes (for example, delivered in school vs hospital). The only general guidance of this nature we have identified was included in the UK National Institute for Health and Clinical Excellence (NICE)8 guidance on the prevention, identification, assessment and management of overweight and obesity in adults and children. BCTs such as self-monitoring, goal setting and providing rewards are recommended in this guidance and no further information is offered in relation to the above issues. In our sample of papers, we found inconclusive evidence that self- monitoring and goal setting were consistently effective techniques in both prevention and management trials. Our findings suggest that providing rewards was actually ineffective for obesity prevention. The potential difference in content between programmes aimed at adults vs children, and the effectiveness of BCTs applied to particular groups or modes of intervention delivery are important topics of further research in this area. From our work here however, we would recommend intervention designers in the field build into interventions the BCTs shown here in relation to specific prevention and management programmes for children and adolescents.

There are a number of issues we would like to highlight from this review. First, formally defining the content of trials is inherently problematic due to reporting inconsistencies and differences in terminology which may have resulted in providing an incomplete picture of some interventions and potentially led to exclusion of eligible studies in the selection stages of this review. We applied a stringent coding strategy where BCTs were not coded as present if there was insufficient description, therefore it may be the case that we coded as absent techniques that the authors of the intervention would argue were present. This issue

Figure 3. Percentage effectiveness of BCTs in obesity prevention trials.

Effective BCTs in prevention and management of childhood obesity J Martin et al

1291

& 2013 Macmillan Publishers Limited International Journal of Obesity (2013) 1287 – 1294

has been identified in previous attempts to categorise interven- tion content.20 To advance and implement a science of behaviour change in the field of childhood obesity, we would echo calls for improving the quality of reporting of trials.14–17 Likewise, given that only a small pool of studies reached our stringent criteria, we would acknowledge the need for more studies in this field that are conducted as RCTs, contain control conditions and report data at least 6 months after the intervention has ended.

Second, this is the first time to our knowledge that the CALO-RE taxonomy has been applied to interventions for children, and we made a number of revisions to the descriptions of BCTs. The authors of the original taxonomy themselves identify that the development of behaviour-specific taxonomies is an iterative process and revisions are inevitable and indeed welcome.24

However, we would stress that, even with our revisions, the CALO- RE taxonomy may not characterise every strategy used in childhood obesity interventions. In a larger sample, we would anticipate that the CALO-RE taxonomy may be revised further, and new BCTs added. For this reason, we would not recommend discarding BCTs in the CALO-RE taxonomy that we agreed were not present in any of the interventions in this review.

The third issue relates to the ability to isolate and assess the effect of individual BCTs in interventions. Our findings are limited to the impact of individual BCTs and we did not assess the effect of combinations of BCTs. We therefore cannot assess whether BCTs demonstrated to be individually effective did not contribute to effective outcomes only as a combination, or whether, when combined differently, their effectiveness would alter. It is worth noting that the BCTs we have shown to be non-effective may well prove to be effective in (alternative) specific combinations, or applied to specific populations or different modes of intervention delivery. Dumbrowski et al.62 assessed the effect of combining theory-congruent clusters of BCTs on outcomes of behaviour change interventions with obese adults and found that interventions that included BCTs congruent with Control Theory were associated with an increase in weight loss. Exploring combinations of BCTs according to other theories of behaviour change may also prove fruitful, for example, goal setting might be more effectively paired with action planning to enable the initiation and enaction of behavioural change according to the theory of implementation intentions.63 Our findings demonstrate the BCTs we have the best available evidence possible for, but this should not be at the expense of investigating other BCTs, individually or in combination. Evidence demonstrates that the most successful interventions contain combinations of BCTs therefore addressing these issues is of prime importance.

Finally, we would like to comment on the clinical value of using BCTs as components of childhood obesity interventions, since the effects of our sample of studies reviewed here were variable and there is currently little consensus on the clinical reductions that may reasonably be expected from behavioural modification. Unfortunately, evidence suggests that the effects of interventions to promote healthy eating are variable,17 and this variability was reflected in our sample of trials, potentially limiting our findings in relation to effective vs non-effective BCTs. Smaller effect sizes may be explained by our measurement time frame (that is, at least 6 months after the intervention had ended) as we would anticipate that differences between intervention and control groups would diminish over this time period, after initial weight loss in intervention groups had stabilised. However, we would argue that this pattern of weight loss (and subsequent gain or stabilising) represents an accurate picture of weight loss maintenance and it is essential to be clear about the impact of behaviour change interventions over time. Observed effect sizes for trials in this review were small, and our method of classifying trials as effective vs ineffective may have limitations compared with other methods of assessing effectiveness. However, our findings clearly illustrate that the effectiveness of interventions is

currently being hampered by the inclusion of BCTs that are ineffective in achieving clinically valuable reductions in BMI, and this may explain why such interventions fare so poorly in achieving (and maintaining) larger reductions in BMI over time. We have reflected on the clinical meaning of our observed effect sizes and concluded that although small, such unit reductions in BMI can be related to reduced risks of adverse health outcomes. A linear relationship between BMI in childhood and risk of coronary heart disease26 and diabetes64 in adulthood has been demonstrated, for example, and reductions in BMI also impact upon ‘secondary’ health outcomes such as cholesterol levels, blood pressure and psychological well-being that exist along with obesity.5

There is currently no clear evidence upon which to include (or not include) BCTs in successful childhood obesity prevention and management programmes. It is absolutely essential to develop such knowledge so that intervention designers in the field can apply these findings to practice. We outline here a novel approach for establishing the effectiveness of BCTs in childhood obesity interventions: this is the first study of its kind (to our knowledge) in this behavioural domain to relate BMI outcomes at 6 months to intervention content.

We have outlined a number of issues remaining to be addressed; however, the methods used in this review represent an important first step towards establishing a method by which it is possible to distinguish between effective and non-effective components of interventions. The results of this study offer the first evidence for the inclusion of effective BCTs in childhood obesity prevention and management programmes, and may contribute to our theoretical understanding of the mechanisms under-pinning obesity-related behaviour change. Identifying and utilising effective BCTs for childhood obesity interventions will aid the development of potentially more effective, replicable and targeted interventions, policies and guidelines for health professionals in the field, thus building a strong evidence base to support obesity prevention and management in children.

CONFLICT OF INTEREST The authors declare no conflict of interest.

REFERENCES 1 Aylott J, Brown I, Copeland R, Johnson D. Tackling Obesities: The Foresight Report

and Implications for Local Government 2008; Government Office for Science: London, UK.

2 National Obesity observatory (NOO). National Child Measurement Programme. Changes in Children’s Body Mass Index Between 2006/7 and 2008/9, NOO: UK, 2010.

3 WHO. World Health Organisation. Population-Based Prevention Strategies for Childhood Obesity: Report of the WHO Forum and Technical Meeting. WHO: Geneva, 15–17 December, 2009.

4 Kamath CC, Vickers KS, Ehrlich A, McGovern L, Johnson J, Singhal V et al. Clinical review: behavioral interventions to prevent childhood obesity: a systematic review and metaanalyses of randomized trials. J Clin Endocrinol Metab 2008; 93: 4606–4615.

5 Bennett B, Sothern MS. Diet, exercise, behavior: the promise and limits of lifestyle change. Semin Pediatr Surg 2009; 18: 152–158.

6 Oude Luttikhuis H, Baur L, Jansen H, Shrewsbury VA, O’Malley C, Stolk RP et al. Interventions for treating obesity in children. Cochrane Database Syst Rev 2009; CD001872.

7 Hillier F, Pedley C, Summerbell C. Evidence base for primary prevention of obesity in children and adolescents. Bundesgesundheitsblatt Gesundheitsforschung Gesundheitsschutz 2011; 54: 259–264.

8 NICE. National Institute for Health and Clinical Excellence. Obesity Guidance on the Prevention, Identification, Assessment and Management of Overweight and Obesity in Adults and Children, NICE: UK, 2006.

9 Bogle V, Sykes C. Psychological interventions in the treatment of childhood obesity: What we know and need to find out. J Health Psychol 2011; 16: 997–1014.

Effective BCTs in prevention and management of childhood obesity J Martin et al

1292

International Journal of Obesity (2013) 1287 – 1294 & 2013 Macmillan Publishers Limited

10 Kipping RR, Jago R, Lawlor DA. Obesity in children. Part 2: Prevention and management. BMJ 2008; 337: a1848.

11 Van Cauwenberghe E, Maes L, Spittaels H, van Lenthe FJ, Brug J, Oppert JM et al. Effectiveness of school-based interventions in Europe to promote healthy nutrition in children and adolescents: systematic review of published and ’grey’ literature. Br J Nutr 2010; 103: 781–797.

12 Hill JO, Wyatt HR, Reed GW, Peters JC. Obesity and the environment: Where do we go from here? Science 2003; 299: 853–855.

13 Kitzmann KM, Dalton WT, Stanley CM, Beech BM, Reeves TP, Buscemi J et al. Lifestyle interventions for youth who are overweight: a meta-analytic review. Health Psychol 2010; 29: 91–101.

14 Schulz KF, Altman DG, Moher D. CONSORT 2010 statement: updated guidelines for reporting parallel group randomised trials. J Pharmacol Pharmacother 2010; 1: 100–107.

15 WIDER. Workgroup for Intervention Development and Evaluation Research. Available at: http://interventiondesign.co.uk. Accessibility verified 29 October 2012.

16 Michie S, Fixsen D, Grimshaw JM, Eccles MP. Specifying and reporting complex behaviour change interventions: the need for a scientific method. Implement Sci 2009; 4: 40.

17 Atkins L, Michie S. Changing eating behaviour: what can we learn from beha- vioural science? Nutr Bul 2013; 38: 30–35.

18 Sargent GM, Pilotto LS, Baur LA. Components of primary care interventions to treat childhood overweight and obesity: a systematic review of effect. Obes Rev 2011; 12: e219–e235.

19 Johnston CA, Tyler C, Foreyt JP. Behavioral management of obesity. Curr Atheroscler Rep 2007; 9: 448–453.

20 Hardeman W, Griffin S, Johnston M, Kinmonth AL, Wareham NJ. Interventions to prevent weight gain: a systematic review of psychological models and behaviour change methods. Int J Obes Relat Metab Disord 2000; 24: 131–143.

21 Abraham C, Michie S. A taxonomy of behavior change techniques used in interventions. Health Psychol 2008; 27: 379–387.

22 Golley RK, Hendrie GA, Slater A, Corsini N. Interventions that involve parents to improve children’s weight-related nutrition intake and activity patterns - what nutrition and activity targets and behaviour change techniques are associated with intervention effectiveness? Obes Rev 2011; 12: 114–130.

23 Michie S, Abraham C, Whittington C, McAteer J, Gupta S. Effective techniques in healthy eating and physical activity interventions: a meta-regression. Health Psychol 2009; 28: 690–701.

24 Michie S, Ashford S, Sniehotta F, Dombrowski S, Bishop A, French D. A refined taxonomy of behaviour change techniques to help people change their physical activity and healthy eating behaviours - the CALO-RE taxonomy. Psychol Health 2011; 26: 1479–1498.

25 Prochaska JO, DiClemente CC. Stages and processes of self-change of smoking: toward an integrative model of change. J Consult Clin Psychol 1983; 51: 390–395.

26 Baker J, Olsen L, Sorenson T. Childhood body-mass index and the risk of coronory heart disease. N Engl J Med 2007; 357: 2329–2337.

27 Kirk S, Zeller M, Claytor R, Santangelo M, Khoury PR, Daniels SR. The relationship of health outcomes to improvement in BMI in children and adolescents. Obes Res 2005; 13: 876–882.

28 Lawlor DA, Benfield L, Logue J, Tilling K, Howe LD, Fraser A et al. Association between general and central adiposity in childhood, and change in these, with cardiovascular risk factors in adolescence: prospective cohort study. BMJ 2010; 341: c6224.

29 Glanville JM, Lefebvre C, Miles JN, Camosso-Stefinovic J. How to identify randomized controlled trials in MEDLINE: ten years on. J Med Libr Assoc 2006; 94: 130–136.

30 Golley RK, Magarey AM, Baur LA, Steinbeck KS, Daniels LA. Twelve-month effectiveness of a parent-led, family-focused weight-management program for prepubertal children: a randomized, controlled trial. Pediatrics 2007; 119: 517–525.

31 Janicke DM, Sallinen BJ, Perrie MG, Lutes LD, Huerta M, Silverstein JH et al. Comparison of parent-only vs family-based interventions for overweight children in underserved rural settings: outcomes from project STORY. Arch Pediatr Adolesc Med 2008; 162: 1119–1125.

32 Johnson WG, Hinkle LK, Carr RE, Anderson DA, Lemmon CR, Engler LB et al. Dietary and exercise interventions for juvenile obesity: long-term effect of behavioral and public health models. Obes Res 1997; 5: 257–261.

33 Fitzgibbon ML, Stolley MR, Schiffer L, Van Horn L, KauferChristoffel 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: 618–625.

34 Fitzgibbon ML, Stolley MR, Schiffer L, Van Horn L, KauferChristoffel K, Dyer A. Hip-Hop to Health Jr. for Latino Preschool Children. Obesity 2006; 14: 1616–1625.

35 Chen S, Weiss S, Heyman MB, Lustig RH. Efficacy of a child-centred and family- based program in promoting healthy weight and healthy behaviors in Chinese

American children: a randomized controlled study. J Public Health (Oxf) 2010; 32: 219–229.

36 Gentile GW, Reimer RA, Walsh DA, Russell DW, Callahan R, Walsh M et al. Evaluation of a multiple ecological level child obesity prevention program: Switch what you Do, View, and Chew. BMC Med 2009; 7: 49.

37 Nader PR, Stone EJ, Lytle LA, Perry CL, Osganian SK, Kelder S et al. Three-year maintenance of improved diet and physical activity: the CATCH cohort. Child and Adolescent Trial for Cardiovascular Health. Arch Pediatr Adolesc Med 1999; 153: 695–704.

38 Reilly JJ, Kelly L, Montgomery C, Williamson A, Fisher A, McColl JH et al. Physical activity to prevent obesity in young children: Cluster randomised controlled trial. BMJ 2006; 333: 1041.

39 Singh AS, Chin APMJ, Brug J, van Mechelen W. Dutch obesity intervention in teenagers: effectiveness of a school-based program on body composition and behavior. Arch Pediatr Adolesc Med 2009; 163: 309–317.

40 Kalavainen MP, Korppi MO, Nuutinen OM. Clinical efficacy of group-based treat- ment for childhood obesity compared with routinely given individual counseling. Int J Obes 2007; 31: 1500–1508.

41 McCallum WM, Gerner B, Baur LA, Gibbons K, Gold L, Gunn J et al. Outcome data from the LEAP (Live, Eat and Play) trial: a randomized controlled trial of a primary care intervention for childhood overweight/mild obesity. Int J Obes 2007; 31: 630–636.

42 Nemet D, Barkan S, Epstein Y, Friedland O, Kowen G, Eliakim A. Short- and long- term beneficial effects of a combined dietary-behavioral-physical activity inter- vention for the treatment of childhood obesity. Pediatrics 2005; 115(4 Part 1): e443–e449.

43 Tanofsky-Kraff DE, Wilfley P, Young JF, Mufson L, Yanovski SZ, Glasofer DR et al. A pilot study of interpersonal psychotherapy for preventing excess weight gain in adolescent girls at-risk for obesity. J Eat Disord 2010; 43: 701–706.

44 Wake M, Baur LA, Gerner B, Gibbons K, Gold L, Gunn J et al. Outcomes and costs of primary care surveillance and intervention for overweight or obese children: the LEAP 2 randomised controlled trial. BMJ 2009; 339: b3308.

45 Hughes AR, Stewart L, Chapple J, McColl JH, Donaldson MD, Kelnar CJ et al. Randomized, controlled trial of a best-practice individualized behavioral program for treatment of childhood overweight: Scottish Childhood Overweight Treat- ment Trial (SCOTT). Pediatrics, 2008; 121(3): e539–e546.

46 Salmon J, Ball K, Hume C, Booth M, Crawford D. Outcomes of a group-randomized trial to prevent excess weight gain, reduce screen behaviours and promote physical activity in 10-year-old children: switch-play. Int J Obes (Lond) 2008; 32: 601–612.

47 Eisenmann JC, Gentile DA, Welk GJ, Callahan R, Strickland S, Walsh M et al. SWITCH: rationale, design, and implementation of a community, school, and family-based intervention to modify behaviors related to childhood obesity. BMC Public Health 2008; 8: 223.

48 Fitzgibbon ML, Stolley MR, Dyer AR, VanHorn L, Kaufer-Christoffel K. A commu- nity-based obesity prevention program for minority children: rationale and study design for Hip-Hop to Health Jr. Prev Med 2002; 34: 289–297.

49 Golley RK, Perry RA, Magarey AM, Daniels LA. Family-focused weight manage- ment program for five- to nine-year-olds incorporating parenting skills training with healthy lifestyle information to support behaviour modification. Nutr Diet 2007; 64: 144–150.

50 Janicke DM, Sallinen BJ, Perri MG, Lutes LD, Silverstein JH, Huerta MG et al. Sensible treatment of obesity in rural youth (STORY): design and methods. Contemp Clin Trials 2008; 29: 270–280.

51 McCallum Z, Wake M, Gerner B, Harris C, Gibbons K, Gunn J et al. Can Australian general practitioners tackle childhood overweight/obesity? Methods and processes from the LEAP (Live, Eat and Play) randomized controlled trial. J Paediatr Child Health 2005; 41: 488–494.

52 Perry CL, Parcel GS, Stone EJ, Nader PR, Mckinlay SM, Luepker RV et al. The Child and Adolescent Trial for Cardiovascular Health (CATCH): overview of the Intervention Program and Evaluation Methods. Cardiovasc Risk Fact 1992; 2: 36–44.

53 Perry CL, Stone EJ, Parcel GS, Ellison RC, Nader PR, Webber LS et al. School-based cardiovascular health promotion: the child and adolescent trial for cardiovascular health (CATCH). J Sch Health 1990; 60: 406–413.

54 Salmon J, Ball K, Crawford D, Booth M, Telford A, Hume C et al. Reducing sedentary behaviour and increasing physical activity among 10-year-old children: overview and process evaluation of the ’Switch-Play’ intervention. Health Promot Int 2005; 20: 7–17.

55 Singh AS, Chin APMJ, Kremers SP, Visscher TL, Brug J, van Mechelen W. Design of the Dutch Obesity Intervention in Teenagers (NRG-DOiT): systematic development, implementation and evaluation of a school-based intervention aimed at the prevention of excessive weight gain in adolescents. BMC Public Health 2006; 6: 304.

Effective BCTs in prevention and management of childhood obesity J Martin et al

1293

& 2013 Macmillan Publishers Limited International Journal of Obesity (2013) 1287 – 1294

56 Stewart L, Houghton J, Hughes A, Pearson D, Reilly J. Dietetic management of pediatric overweight: development and description of a practical and evidence- based behavioral approach. Am Diet Assoc 2005; 105: 1810–1815.

57 Tanofsky-Kraff M, Wilfley DE, Young JF, Mufson L, Yanovski SZ, Glasofer DR et al. Preventing excessive weight gain in adolescents: inter- personal psychotherapy for binge eating. Obesity (Silver Spring) 2007; 15: 1345–1355.

58 Chater A, Martin J, Smith L. Behaviour change techniques used in randomised controlled trials of obesity prevention and treatment in children: a systematic review. Annual Conference of the Division of Health Psychology. 14–16 September 2011Southampton, UK.

59 Cohen J. A coefficient of agreement for nominal scales. Educ Psychol Meas 1960; 20: 37–46.

60 Michie S, Churchill S, West R. Identifying evidence-based competences required to deliver behavioral support for smoking cessation. Ann Behav Med 2011; 41: 59–70.

61 Lorencatto F, West R, Michie S. Specifying evidence-based behaviour change tech- niques to aid smoking cessation in pregnancy. Nicotine Tob Res 2012; 14: 1019–1026.

62 Dumbrowski SU, Sniehotta FF, Avenell A, Johnston M, MacLennan G, Araujo-Soares V. Identifying active ingredients in complex behavioural interven- tions for obese adults with obesity-related co-morbidities or additional risk factors for co-morbidities: a systematic review. Health Psychol Rev 2012; 6: 7–32.

63 Gollwitzer PM. Implementation intentions: Strong effects of simple plans. Am Psychol, 1999; 54: 493–503.

64 Tirosh A, Shai I, Afek A, Dubnov-Raz G, Ayalon N, Gordon B et al. Adolescent BMI trajectory and risk of diabetes versus coronary disease. N Engl J Med 2011; 364: 1315–1325.

Supplementary Information accompanies this paper on International Journal of Obesity website (http://www.nature.com/ijo)

Effective BCTs in prevention and management of childhood obesity J Martin et al

1294

International Journal of Obesity (2013) 1287 – 1294 & 2013 Macmillan Publishers Limited

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.