# Economic evaluation of lifestyle interventions to treat overweight or obesity in children

## Abstract

### Objective:

To estimate lifetime cost effectiveness of lifestyle interventions to treat overweight and obese children, from the UK National Health Service perspective.

### Design:

An adaptation of the National Heart Forum economic model to predict lifetime health service costs and outcomes of lifestyle interventions on obesity-related diseases.

### Setting:

Hospital or community-based weight-management programmes.

### Population:

Hypothetical cohorts of overweight or obese children based on body mass data from the National Child Measurement Programme.

### Interventions:

Lifestyle interventions that have been compared with no or minimal intervention in randomized controlled trials (RCTs).

### Main Outcome Measures:

Reduction in body mass index (BMI) standard deviation score (SDS), intervention resources/costs, lifetime treatment costs, obesity-related diseases and cost per life year gained.

### Results:

Ten RCTs were identified by our search strategy. The median effect of interventions versus control from these 10 RCTs was a difference in BMI SDS of −0.13 at 12 months, but the range in effects among interventions was broad (0.04 to −0.60). Indicative costs per child of these interventions ranged from £108 to £662. For obese children aged 10–11 years, an intervention that resulted in a median reduction in BMI SDS at 12 months at a moderate cost of £400 increased life expectancy by 0.19 years and intervention costs were offset by subsequent undiscounted savings in treatment costs (net saving of £110 per child), though this saving did not emerge until the sixth or seventh decade of life. The discounted cost per life year gained was £13 589. Results were broadly similar for interventions aimed at children aged 4–5 years and which targeted both obese and overweight children. For more costly interventions, savings were less likely.

### Conclusion:

Interventions to treat childhood obesity are potentially cost effective although cost savings and health benefits may not appear until the sixth or seventh decade of life.

## Introduction

Obesity in children and adolescents is associated with a range of adverse metabolic and cardiovascular risk factors.1 Obese children are more likely to be obese as adults2, 3, 4 and also have an increased risk of coronary heart disease (CHD) in adulthood.5 It is a public health priority to prevent and treat obesity in children and adults, in order to reduce morbidity and premature mortality.6, 7

The prevalence of obesity in children aged between 2 and 10 years in England for 2007 was estimated to be 14 and 13% for girls and boys, respectively (using the 95th percentile of the UK 1990 growth reference to define obesity).8 Analysis of Health Survey for England data suggests that the increased prevalence of childhood obesity may be levelling off, but it is too early to know whether the peak in prevalence has been reached or if this apparent plateau is general variation around an on-going increase.9, 10

In the United Kingdom, the percentage of National Health Service (NHS) costs attributable to elevated body mass index (BMI) across all age groups was 6% in 2007, and is predicted to rise to 11.9% in 2025 and 13.9% in 2050.11 A variety of childhood weight-management interventions are already commissioned by UK health-care authorities. If these interventions are effective in reducing childhood obesity they have the potential to be cost effective over the lifetime of the child. The effectiveness of many of these interventions is unknown because they have not been evaluated in randomized controlled trials (RCTs). Even where RCTs have been conducted, follow-up is often limited and economic analysis is absent making it impossible for health-care payers to precisely estimate the lifetime costs and benefits for their population. Previous studies have attempted to determine the potential long-term cost effectiveness of interventions aimed at treating overweight or obese children, but these studies have largely assessed interventions that have not been evaluated in RCTs.12 However, without RCT evidence of public health effectiveness the meaning of the cost-effective analysis is unclear.

The aim of our study is to model the potential cost effectiveness of RCT-evaluated lifestyle interventions (that is, non-surgical, non-pharmaceutical) to treat overweight or obesity in primary school aged children over their lifetime in terms of NHS costs/savings, life years saved and incidence of obesity-related disease.

## Materials and methods

### Defining cohorts of overweight/obese children

As part of the NCMP (National Child Measurement Programme) in England,13 authorities are required to collect BMI data on all eligible children aged 4–5 and 10–11 years in state schools. The NCMP is an ‘opt-out’ study where guardians explicitly have to deny permission if they do not wish their child to be measured. We used data from four regions in South West England (Bristol, Bath and North East Somerset (BaNES), South Gloucestershire and North Somerset) for the year 2008/2009. In these four regions, the prevalence of overweight or obesity in the 4–5-year and 10–11-year age groups (23% and 31%, respectively) was representative of the national prevalence (23% and 33%, respectively). Based on these data, we simulated four cohorts of children, representing two weight groups (obese only or overweight and obese) and two age groups (4–5 or 10–11 years). We used the age- and gender-specific clinical thresholds recommended by NICE (National Institute of Health and Clinical Excellence)14 to define overweight and obesity (91st and 98th UK90 percentiles, respectively).

### Determining effect sizes and costs of RCT-evaluated treatments for overweight/obese children

We estimated a plausible range for intervention effect size at 12 months based on the RCTs of lifestyle interventions targeted at overweight or obese primary school aged children. We identified RCTs published before May 2008 from a Cochrane review of obesity treatments.15 We modified and updated the Cochrane search strategy (web reference 1) to identify RCTs published between May 2008 and February 2010 and recorded in the OVID Medline database. Our eligibility criteria (web reference 2) restricted our analysis to trials conducted in predominantly overweight or obese primary school aged children where a lifestyle intervention was compared with ‘no’ or ‘minimal’ intervention over a period of at least 6 months.

Most published reports of RCTs did not perform a detailed cost analysis or provide enough information about the resources used to facilitate a retrospective estimation of cost. To provide an indicator of the likely resource use of each intervention, we ranked interventions according to the number and duration of intervention sessions (that is, professional contact time). As most lifestyle interventions did not require expensive materials or equipment, contact time is likely to be strongly correlated with the cost of intervention, with the exception of interventions delivered by high earners such as General Practitioners. Where the intervention cost per child was estimated in the trial or reported in a separate source, we recorded this (web reference 4). Costs were converted into UK currency (£) using Organisation for Economic Co-operation and Development purchasing power parity values for 2009.16 Based on the range of costs reported in the literature (£108–£662 per child; web reference 4), we modelled a range of interventions including low-cost (£100), moderate-cost (£400) and high-cost (£800) interventions. The cost of weight-management treatment in the absence of a lifestyle intervention was assumed to be zero as currently very little is offered to most primary school children who are overweight or obese.

### Determining incidence of overweight/obesity-related disease over the lifetime of the cohort children

We used the National Heart Forum (NHF) ‘Obesity Model’, originally developed for the Foresight ‘Tackling Obesities’ project and subsequently enhanced, to predict the incidence of weight-related diseases over the lifetime of the four cohorts of children in the absence of any interventions. Technical details of the NHF model have been published previously.11 Many other obesity simulation models have been developed;17 however, the NHF model is the only one calibrated to UK data predicting the health and economic consequences of weight gain. The NHF model simulates a hypothetical cohort of 50 000 children by randomly sampling, with replacement, the age, sex and BMI of actual children measured in schools from the four regions of South West England. The cohort is then ‘aged’ by the model, gaining weight and developing weight-related disease over time in line with projections based on the Health Survey for England. This is estimated through a BMI growth equation simulated using Monte Carlo techniques for every year until death for all individuals in the cohort.11 The model tracks BMI and the incidence, prevalence and costs of weight-related diseases of individuals over their lifetime.

We compared the natural progression of BMI in the four cohorts with several scenarios where children have access to weight-management interventions. A total of nine interventions (sensitivity analyses) were simulated in the model, representing all permutations of BMI standard deviation score (SDS) reduction (minimal, median or maximal effect size) and intervention cost (low, moderate and high). Our effect size estimates are based on the BMI SDS data reported in the RCTs. For studies that reported only BMI, we converted to BMI SDS based on the gender and baseline mean age of the children enrolled in the RCT using the UK90 growth charts.18 We repeated this calculation at follow-up to estimate change in BMI SDS. We used the effect sizes reported in the RCTs at 12 months to define (i) median effect size (equal to the median effect size reported in the RCTs); (ii) maximal effect size (equal to the largest effect size reported in the RCTs) and (iii) minimal effect size (equal to the lowest effect size reported in the RCTs). We assumed that these 12-month changes in BMI SDS were not transitory and defined an individual's lifetime weight trajectory, an important assumption that we return to in our discussion. Likewise, we defined ‘low’ (£100) and ‘high’ (£800) cost interventions to encompass the minimum and maximum intervention costs estimated in the literature and ‘moderate’ cost (£400) to equal the approximate cost of a typical lifestyle intervention delivered to groups of children in a community setting.

### Determining lifetime costs of overweight/obesity-related diseases

We used national programme budgeting data19 to estimate the annual NHS cost per case associated with five weight-related diseases (CHD, stroke, diabetes, breast cancer and colorectal cancer). The total NHS secondary and primary care spending in England on each of the five weight-related diseases (web reference 3) is divided by the estimated prevalence of each disease, derived from the NHS Clinical and Health Outcomes Knowledge Base compendium indicators20 or cancer registry data,21 to estimate the cost per patient per year of disease in the model. These five diseases include those with the highest BMI attributable incidence (for example, diabetes), but exclude other weight-related conditions such as osteo-arthritis for which no national expenditure data were available. NICE guidance recommends that both costs and health outcomes occurring in future years are discounted at 3.5% per annum.22 In effect, discounting devalues any future savings due to health promotion strategies in children and is controversial in this context.23 We therefore report both discounted and undiscounted results.

The incremental costs/savings and health effects of interventions in all four cohorts were calculated by comparing cost and outcomes in the intervention scenarios with those in the equivalent natural progression scenario.

## Results

In the NCMP, 9956 (92.7%) of 4–5 year olds and 9698 (90.0%) of 10–11 year olds were measured in the four regions included in this study. The mean BMI SDS (0.47 versus 0.40) and the proportion of children classified as overweight or obese (23.3% versus 14.5%; 91st UK90 percentile) or obese (9.5% versus 4.7%; 98th UK90 percentile) were higher in 10–11 year olds than in 4–5 year olds. We identified 707 potentially relevant publications (Figure 1) in our updated literature review. Five RCTs24, 25, 26, 27, 28 from the existing Cochrane review and five recently published RCTs29, 30, 31, 32, 33 met our eligibility criteria. These are summarized in tables in web reference 4.

The interventions to treat obesity could broadly be categorized into three groups: interventions aimed at modifying behaviour, diet and/or physical activity delivered to groups of children and/or parents24, 26, 29, 31, 33 or delivered to individual families;25, 27, 30 and technology-based interventions used in the home or community aimed at promoting physical activity or reducing sedentary activities.28, 32 For the group-based interventions, planned contact time with professionals ranged from 10(ref. 24) to 128(ref. 29) hours delivered over periods ranging from 5–12 months to groups of between 7 and 14 children and/or families. Where reported, intervention costs for group-based interventions ranged from £219 per child34 for a programme involving up to 45 h of professional contact time delivered by a dietitian and students to £662 per child for a programme involving up to 128 h of professional contact time delivered by sports coaches, dietitians and psychologists.35 The MEND (‘Mind Exercise, Nutrition, Do it!’) intervention33 currently commissioned by many health-care payers in England involves 36 h of contact time with children and parents conducted by health, social, education and exercise professionals and is estimated to cost £385 per child.36

In contrast, interventions delivered to individual families all had <10 h of planned contact time with participants and were delivered by General Practitioners, pediatricians or dietitians over periods ranging from 12 weeks to 24 months. Cost per child, where estimated, ranged from £108 reported by Hughes et al.25 for an intervention involving eight brief individual family contacts with a dietician, to £557 for a General Practitioners led intervention delivered over four consultations.30 Based on this indicative range of costs, we modelled low, moderate and high intervention costs to be £100, £400 and £800 per child, respectively.

The majority of studies reported outcome data at 6 or 12 months. With one exception,25 all studies suggested a trend for a relative reduction in BMI SDS in the intervention group at 6 or 12 months (Figure 2). At 12 months, the median effect on BMI SDS was −0.13(ref. 31) and ranged from 0.04(ref. 25) to −0.60,29 although the reduction in BMI SDS was only statistically significant, at the conventional 5% level, in three studies.27, 29, 31 The ‘Sea Lion Club’ intervention evaluated by Weigel et al.29 had the largest effect size and was a clear outlier. The smallest favourable effect size observed was −0.03.28 Therefore, in our model we evaluated interventions with minimal (−0.03), median (−0.13) and maximal (−0.60) reductions in BMI SDS.

In the model of obese children aged 10–11 years, all interventions led to an improvement in both life expectancy and number of individuals surviving past 75 years of age (Table 1). The median estimate of effect size results in an increase in life expectancy of 0.19 years and an increase in the percentage of individuals living past the age of 75 years of 0.7% compared with no intervention. For the most optimistic estimate of the intervention effect size, life expectancy increased by 1 year and the proportion of individuals living to be older than 75 years increases by 4.1% compared with no intervention. An intervention with a median effect size would decrease years lived with diabetes by 0.61 years, CHD by 0.09 years and stroke by 0.03 years. As the effect size increases the lifetime prevalence of these three diseases decreases (Figure 3). However, this association between intervention effect size and disease prevalence is not consistent for breast cancer, colorectal cancer, arthritis and hypertension. For these diseases, the decrease in incidence due to weight reduction in the intervention cohort is more than offset, in some scenarios, by the increase in incidence due to increased longevity.

An intervention with a minimal effect size of −0.03 BMI SDS only saves £125 in undiscounted costs for treating weight-related diseases over the lifetime of a cohort of obese 10–11 year olds (Table 2). Therefore, assuming a moderate intervention cost of £400, this intervention would not be cost saving. The minimal effect size (−0.03 BMI SDS) has an incremental cost of £275 and <0.02 life years gained, which equates to an undiscounted cost per life year gained of £16 898. A moderate cost intervention with larger effect sizes (−0.13 and −0.60) results in undiscounted lifetime savings of £110 and £1728 per child, respectively.

All interventions are much less cost effective if future costs and health benefits are discounted. For example, an intervention that costs £400 and produces a −0.13 reduction in BMI SDS is no longer cost saving; instead discounted costs per life year gained are £13 589. However, this would still be considered cost effective at current national (£20 000–£30 000 per quality adjusted life year)22 and international37 cost effectiveness thresholds assuming that weight reduction does not have a detrimental effect on quality of life. Results were largely similar, for cost and health outcomes, between the two school years and between the obese and overweight and obese groups (web reference 5).

At a cost of £100 per child, similarly to the cost estimated by Hughes et al.,25 all of the intervention effect sizes become cost saving within the lifetime of the children if future treatment costs are not discounted (Table 3). For a BMI SDS reduction of −0.03, the cost of the intervention would never become cost saving for interventions over £200. An intervention producing a −0.13 BMI SDS reduction and costing £400 breaks-even 58 years post intervention. An intervention that results in a −0.60 BMI SDS reduction produces net savings unless the intervention is very expensive.

## Discussion

### Principal findings

Substantial proportions of children are overweight or obese. The RCT literature demonstrates that interventions can reduce BMI SDS over the course of a year among overweight and obese children in developed nations, but the nature of those interventions varies widely. For example, the target (parent, child or both), the setting (individual versus group, outpatient versus community), intensity (frequency and duration of contacts), practitioner (physician, dietician, exercise specialist) all differed substantially. It is, therefore, unsurprising that the observed treatment effect and reported costs of interventions were not uniform. Based on a median effect size (−0.13 reduction in BMI SDS) and moderate intervention cost (£400), we estimate that an intervention would lead to a small increase (0.19 life years) in the life expectancy of obese 10–11 year olds. Furthermore, years of life lived with common weight-related diseases such as diabetes, CHD and stroke would decrease by 0.61, 0.09 and 0.03 years, respectively. This in turn will lead to savings in the treatment costs spent on these diseases; however, these will not pay back the initial intervention costs until the cohort reaches the sixth or seventh decade of life. The findings were broadly similar if the interventions were extended to both overweight and obese children and children aged 4–5 years.

### Strengths and weaknesses of the study

The main strengths of this study are its importance, uniqueness and our systematic approach. The limitations largely result from the lack of available evidence and hence the assumptions that we have to make. Evidence is sparse on the extent to which short-term reductions in BMI SDS are maintained post-intervention. The longest follow-up for trials included in our review was 2 years.32 Long-term follow-up of an RCT of parent versus child focussed treatments for childhood obesity suggested that the additional benefit of the parent focussed approach peaked at 3 years and declined, but was still evident, up to 8 years after the intervention.38 We assumed that 12 month changes in BMI SDS observed in the RCTs then defined an individual's weight trajectory for the rest of their life. In actuality, a sustained lifestyle change could lead to continued reductions in weight eventually back to the normal range or may only achieve a transitory change in habits with no long-term health benefits. Furthermore, the NHF model uses cross-sectional data from the Health Survey for England to construct BMI trajectories for individuals. However, children in 2010 may not follow the same BMI trajectory as previous generations. If weight loss in the first 12 months after intervention is not sustained throughout life or if children in 2010 follow steeper (increasing) gradient of BMI trajectory than previous generations our estimated cost savings will be exaggerated; with the converse true if our assumptions are incorrect in the opposite direction.

Our findings are dependent on the validity of the BMI SDS changes reported in the RCTs. All RCTs had been through the peer review process, but any remaining flaws in conduct or analysis could have introduced bias. While we believe that the median effect of intervention on BMI used in our model should be achievable in practice, results from individual RCTs should be interpreted in the context of the quality of trial conduct and analysis. The cost of intervention was rarely reported alongside the RCT. Even where reported, the costing methods were often inadequately described. We selected a range of costs in our modelling exercise that encompassed those reported in the literature. However, the lack of high quality cost analyses in the literature is a clear limitation for health-care commissioners trying to judge the likely cost effectiveness of a specific intervention.

Our model possibly provides a conservative estimate of cost savings due to intervention because we do not include societal costs (for example, lost productivity) and were unable to cost all of the health states tracked by the NHF model (for example, osteo-arthritis and kidney cancer) for which programme budgeting data are unavailable. On the other hand, the NHF model, in common with most other obesity models developed internationally, does not track disease and costs associated with a healthier, longer living population (for example, increased risk of dementia and other age-related conditions). The net effect of these unvalued diseases is unclear. The model also does not track other treatment costs, for example, bariatric surgery or pharmacotherapy, which will occur if obesity persists into adolescence and adulthood. There is sparse and mixed evidence39, 40, 41 on the association between BMI and health service costs in children and young adults. We also did not adjust life expectancy for quality of life because quality of life data (utilities) are not readily available or collected using a standardized methodology for all of the weight-related diseases tracked in the model.

### Results in relation to other studies

Cecchini et al.42 assessed the health effects and cost effectiveness of public health strategies targeted at diet and physical activity in seven countries including England using a microsimulation model. The strategies aimed to prevent chronic diseases related to obesity (stroke, ischaemic heart disease and cancer) and included school-based health promotion interventions, worksite health promotion interventions, mass media health promotion campaigns, counselling of individuals at risk in primary care, fiscal measures and regulatory methods. The model showed differences in outcomes by age, with interventions targeting adults generating health effects immediately after their implementation and conversely interventions targeting children, regardless of initial BMI, having meaningful effects 40–50 years later, similar to the time lag we found when modelling interventions for child obesity. Cecchini et al. highlighted differences in health expenditure by stage of life, with almost no effects on expenditure up to 40 years of age, reduced health expenditures between ages 40 and 80 years and increased expenditure in later years because of enhanced survival and need for health care. Again, this pattern was found in our study. They conclude that the prevention interventions, which deliver the best value for money, are improved awareness and information; appropriate fiscal measures affecting the price of fruit, vegetables and foods high in fat; and enhanced regulatory mechanisms such as food advertising to children and compulsory food labelling.

Moodie et al.43 estimated lifetime cost effectiveness based on the pilot44 for the LEAP (live, eat and play) trial30 located in General Practitioners clinics in Melbourne, Australia. The trial, conducted in 5–10 year olds who were either overweight or moderately obese, compared an intervention modifying lifestyle, diet and exercise with usual care. Estimates of cost effectiveness were calculated assuming a relative BMI reduction of −0.25 (equivalent to −0.06 BMI SDS reduction in a 10-year old male of average height on the obesity threshold) in a cohort of children and adolescents aged 5–19 years in 2001. The study used Markov modelling to track BMI to 100 years of age and estimate the cost per Disability Adjusted Life Year. The authors estimated a discounted intervention cost of AUD$530 (£300 (year 2001 values)) per child offset by a disease-related cost saving of AUD$350 (£200) per child treated resulting in a cost per Disability Adjusted Life Year saved of AUD\$4670 (£2700). In our analysis of a median intervention effect size in obese children aged 10–11, the incremental discounted costs were £298 (albeit with a higher discount rate) and the discounted cost per life year gained was £13 589. The similar findings of these two independently developed models provide some cross-validation of our results.

### Meaning of the study

Our analyses suggest that there is reason to believe that investment in interventions aimed at treating overweight and obese children will yield health benefits and treatment cost savings over the lifetime of the cohort, although the investment will not be recouped quickly. Although our study is based on the data from England, it is probable that these findings are also relevant for other developed nations with similar childhood overweight and obesity prevalences. Our results cannot be extrapolated to other programmes which have not been evaluated in RCTs, such as the more intensive residential weight loss programmes.45 Results were largely similar for overweight and obese children at ages 4–5 and 10–11 years. This may well be a manifestation of the well-described ‘prevention paradox’ whereby a large number of people (for example, overweight or obese) exposed to a lower risk of disease generate more cases of disease than a small number of people (for example, obese) with a higher risk of disease.46 However, given the large numbers of overweight and obese children, health-care commissioners may struggle to afford these interventions for all children who might benefit. For example, to provide a £400 intervention for the 126 000 children in England aged 4–5 years who are overweight or obese would cost in excess of £50 million and would require large numbers of staff to be trained to deliver the intervention. The initial cost would fall to £21 million if it were restricted to children who were obese. If this investment is made, then the lifestyle interventions evaluated in our study appear to be competitive with other pharmaceuticals and procedures currently recommended through NICE technology appraisals.14

### Unanswered questions and future research

Future research is needed to understand the optimum duration of lifestyle treatments for overweight and obese children and long-term follow-up to establish the effectiveness and sustainability of treatments into adolescence and adulthood. Commissioners and providers of interventions should monitor attendance and treatment outcomes for overweight and obese children referred to weight-management programmes. This would allow both a comparison of local treatment outcomes with those reported in the RCTs and an assessment of whether treatment is most effective in specific subgroups, defined by factors such as age, gender, ethnicity and treatment duration. Large observational studies describing the association between BMI and use of health services in adolescents and young adults would be valuable to determine whether the potential cost savings identified in our study can be realized more promptly.

## References

1. 1

Lawlor DA, Riddoch CJ, Page AS, Anderssen SA, Froberg K, Harro M et al. The association of birthweight and contemporary size with insulin resistance among children from Estonia and Denmark: findings from the European Youth Heart Study. Diabet Med 2005; 22: 921–930.

2. 2

Serdula MK, Ivery D, Coates RJ, Freedman DS, Williamson DF, Byers T . Do obese children become obese adults? A review of the literature. Prev Med 1993; 22: 167–177.

3. 3

Power C, Lake JK, Cole TJ . Body mass index and height from childhood to adulthood in the 1958 British born cohort. Am J Clin Nutr 1997; 66: 1094–1101.

4. 4

Whitaker RC, Wright JA, Pepe MS, Seidel KD, Dietz WH . Predicting obesity in young adulthood from childhood and parental obesity. N Engl J Med 1997; 337: 869–873.

5. 5

Owen CG, Whincup PH, Orfei L, Chou QA, Rudnicka AR, Wathern AK et al. Is body mass index before middle age related to coronary heart disease risk in later life? Evidence from observational studies. Int J Obes (Lond) 2009; 33: 866–877.

6. 6

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

7. 7

Han JC, Lawlor DA, Kimm SY . Childhood obesity. Lancet 2010; 375: 1737–1748.

8. 8

The NHS Information Centre LS. Statistics on Obesity, Physical Activity and Diet: England, 2010. NHS Information Centre for Health and Social Care, 2010 (http://www.ic.nhs.uk/pubs/opad10).

9. 9

The Information Centre. Health Survey for England 2007 Latest Trends 2008. (Archived by WebCite® at http://www.webcitation.org/5u8ekeewr).

10. 10

Wang YC, McPherson K, Marsh T, Gortmaker S, Brown M . Health and economic burden of the projected obesity trends in the USA and the UK. Lancet 2011; 378: 815–825.

11. 11

McPherson K, Marsh T, Brown M . Tackling Obesities: Future Choices—Modelling Future Trends in Obesity and the Impact on Health, 2nd edn. UK Government Foresight, 2007 (http://www.bis.gov.uk/assets/bispartners/foresight/docs/obesity/14.pdf).

12. 12

Carter R, Moodie M, Markwick A, Magnus A, Vos T, Swinburn B et al. Assessing cost-effectiveness in obesity (ACE-obesity): an overview of the ACE approach, economic methods and cost results. BMC Public Health 2009; 9: 419.

13. 13

Dinsdale H, Rutter H . National Child Measurement Programme: 2006/07 School Year. National Obesity Observatory, 2008 (http://www.ic.nhs.uk/webfiles/publications/ncmp/ncmp0607/NCMP%202006%2007.%20Bulletin%20Final.pdf).

14. 14

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 Clinical Guideline 43. NICE, 2006 (http://www.nice.org.uk/CG043).

15. 15

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; 1: CD001872.

16. 16

Organisation for Economic Co-operation and Development. Purchasing Power Parities for GDP 2010. 10-11-2010. (Archived by WebCite® at http://www.webcitation.org/5u8YikMQZ).

17. 17

Levy DT, Mabry PL, Wang YC, Gortmaker S, Huang TTK, Marsh T et al. Simulation models of obesity: a review of the literature and implications for research and policy. Obes Rev 2011; 12: 378–394.

18. 18

Freeman JV, Cole TJ, Chinn S, Jones PR, White EM, Preece MA . Cross sectional stature and weight reference curves for the UK, 1990. Arch Dis Child 1995; 73: 17–24.

19. 19

Department of Health. Programme Budgeting Guidance Manual 2005. 11-10-2010. (Archived by WebCite® at http://www.webcitation.org/5sIPUxPd7).

20. 20

National Centre for Health Outcomes Development. Clinical and Health Outcomes Knowledge Base 2010. 10-11-2010. (Archived by WebCite® at http://www.webcitation.org/5sIP7RBPO).

21. 21

Maddams J, Brewster D, Gavin A, Steward J, Elliott J, Utley M et al. Cancer prevalence in the United Kingdom: estimates for 2008. Br J Cancer 2009; 101: 541–547.

22. 22

National Institute for Health and Clinical Excellence. Guide to the Methods of Technology Appraisal. NICE Publications, 2008 (http://www.nice.org.uk/media/B52/A7/TAMethodsGuideUpdatedJune2008.pdf).

23. 23

Ganiats TG . Prevention, policy, and paradox: what is the value of future health? Am J Prev Med 1997; 13: 12–17.

24. 24

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.

25. 25

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 Treatment Trial (SCOTT). Pediatrics 2008; 121: e539–e546.

26. 26

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

27. 27

Nova A, Russo A, Sala E . Long-term management of obesity in paediatric office practice: experimental evaluation of two different types of intervention. Ambulatory Child Health 2001; 7: 239–247.

28. 28

Rodearmel SJ, Wyatt HR, Stroebele N, Smith SM, Ogden LG, Hill JO . Small changes in dietary sugar and physical activity as an approach to preventing excessive weight gain: the America on the Move family study. Pediatrics 2007; 120: e869–e879.

29. 29

Weigel C, Kokocinski K, Lederer P, Dotsch J, Rascher W, Knerr I . Childhood obesity: concept, feasibility, and interim results of a local group-based, long-term treatment program. J Nutr Educ Behav 2008; 40: 369–373.

30. 30

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.

31. 31

Kalarchian MA, Levine MD, Arslanian SA, Ewing LJ, Houck PR, Cheng Y et al. Family-based treatment of severe pediatric obesity: randomized, controlled trial. Pediatrics 2009; 124: 1060–1068.

32. 32

Epstein LH, Roemmich JN, Robinson JL, Paluch RA, Winiewicz DD, Fuerch JH et al. A randomized trial of the effects of reducing television viewing and computer use on body mass index in young children. Arch Pediatr Adolesc Med 2008; 162: 239–245.

33. 33

Sacher PM, Kolotourou M, Chadwick PM, Cole TJ, Lawson MS, Lucas A et al. Randomized controlled trial of the MEND program: a family-based community intervention for childhood obesity. Obesity 2010; 18 (Suppl-8): S62–S68.

34. 34

Kalavainen M, Karjalainen S, Martikainen J, Korppi M, Linnosmaa I, Nuutinen O . Cost-effectiveness of routine and group programs for treatment of obese children. Pediatr Int 2009; 51: 606–611.

35. 35

Erlangen.de. Programme for Autumn and Winter 2005/6 2010. 11-10-2010. (Archived by WebCite® at http://www.webcitation.org/5u8cT73Cw).

36. 36

Steeds J . MEND Programme Evaluation May 2008-Nov 2009. Community Sport and Active Lifestyles. South Gloucestershire Council, 2010. 10-11-2010. (Archived by WebCite® at http://www.webcitation.org/5u8d11m1Z).

37. 37

Hutubessy R, Chisholm D, Edejer TT . Generalized cost-effectiveness analysis for national-level priority-setting in the health sector. Cost Eff Resour Alloc 2003; 1: 8.

38. 38

Golan M, Kaufman V, Shahar DR . Childhood obesity treatment: targeting parents exclusively v. parents and children. Br J Nutr 2006; 95: 1008–1015.

39. 39

Janssen I, Lam M, Katzmarzyk PT . Influence of overweight and obesity on physician costs in adolescents and adults in Ontario, Canada. Obes Rev 2009; 10: 51–57.

40. 40

Kuhle S, Kirk S, Ohinmaa A, Yasui Y, Allen AC, Veugelers PJ . Use and cost of health services among overweight and obese Canadian children. Int J Pediatr Obes 2010; 6: 142–148.

41. 41

Wenig C . The impact of BMI on direct costs in Children and Adolescents: empirical findings for the German Healthcare System based on the KiGGS-study. Eur J Health Econ 2012; 13: 39–50.

42. 42

Cecchini M, Sassi F, Lauer JA, Lee YY, Guajardo-Barron V, Chisholm D . Tackling of unhealthy diets, physical inactivity, and obesity: health effects and cost-effectiveness. Lancet 2010; 376: 1775.

43. 43

Moodie M, Haby M, Wake M, Gold L, Carter R . Cost-effectiveness of a family-based GP-mediated intervention targeting overweight and moderately obese children. Econ Hum Biol 2008; 6: 363–376.

44. 44

McCallum Z, Wake M, Gerner B, Baur LA, Gibbons K, Gold L 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 (Lond) 2007; 31: 630–636.

45. 45

Gately PJ, Cooke CB, Barth JH, Bewick BM, Radley D, Hill AJ . Children's residential weight-loss programs can work: a prospective cohort study of short-term outcomes for overweight and obese children. Pediatrics 2005; 116: 73–77.

46. 46

Rose G . Sick individuals and sick populations. Int J Epidemiol 1985; 14: 32–38.

## Acknowledgements

We are grateful for help provided by John Twigger, Emily Van De Venter, Helen Yeo and Helen Tapson the public health analysts at NHS Bristol, South Gloucestershire, Bath and North East Somerset and North Somerset in accessing the NCMP data. We are grateful for permission from the National Heart Forum to use the NHF obesity model and for the advice given by Tom Byatt on the model. The work was undertaken with the support of The Centre for the Development and Evaluation of Complex Interventions for Public Health Improvement (DECIPHer), a UKCRC Public Health Research: Centre of Excellence. Funding from the British Heart Foundation, Cancer Research UK, Economic and Social Research Council (RES-590-28-0005), Medical Research Council, the Welsh Assembly Government and the Wellcome Trust (WT087640MA), under the auspices of the UK Clinical Research Collaboration, is gratefully acknowledged. The research was funded by NHS Bristol.

Disclaimer

The funder had no role in the development of the study design, data analyses and interpretation, writing of the article or the decision to submit for publication. All authors had full access to all of the data and can take responsibility for the integrity of the data and the accuracy of the data analysis.

## Author information

Correspondence to W Hollingworth.

## Ethics declarations

### Competing interests

RK was employed by the South West Public Health Training Programme on secondment to the University of Bristol during this work and her employment contract was hosted by NHS Bristol. DAL works in a Centre that receives support from the UK Medical Research Council (G0600705) and the University of Bristol. RRK was receiving funds from the South West NHS Public Health training scheme during the time that she worked on this project.

## Rights and permissions

Reprints and Permissions

• #### DOI

https://doi.org/10.1038/ijo.2011.272

### Keywords

• costs and cost analysis
• primary prevention
• lifestyle

• ### Interventions for obesity among schoolchildren: A systematic review and meta-analyses

• , Shariff Ghazali Sazlina

PLOS ONE (2019)

• ### A New Model for Evaluation of Interventions to Prevent Obesity in Early Childhood

• Alison Hayes
• , Eng J. Tan
• , Thomas Lung
• , Vicki Brown
• , Marj Moodie
•  & Louise Baur

Frontiers in Endocrinology (2019)

• ### Setting Kids Up for Success (SKUFS): Outcomes of an Innovation Project for Promoting Healthy Lifestyles in a Pediatric Patient-Centered Medical Home

• Josephine M. Cole
• , Jane Weigel
• , Susan Albrecht
• , Dianxu Ren
• , Anne K. Reilly
•  & Cynthia A. Danford

Journal of Pediatric Health Care (2019)

• ### The association between body mass index, primary healthcare use and morbidity in early childhood: findings from the Born In Bradford cohort study

• B. Kelly
• , J. West
• , T.C. Yang
• , D. Mason
• , T. Hasan
•  & J. Wright

Public Health (2019)

• ### A Systematic Review of Methods, Study Quality, and Results of Economic Evaluation for Childhood and Adolescent Obesity Intervention

• Mandana Zanganeh