All the studies analyzed the cost-effectiveness of childhood obesity primary prevention programs. Two of them examined cost-effectiveness of community-based intervention programs, four were only school-based programs while four of them examined both community-based and school-based intervention programs. The studies were different in study design, characteristics of population and in health and economic outcomes. In addition, there was heterogeneity about the type of intervention program used. Some articles did not clarify the childhood age range evaluated.

An intervention study considered the economic evaluation of the URMEL-ICE (Ulm Research on Metabolism, Exercise and Lifestyle Intervention in Children), an overweight school-based prevention program, based on developed teaching materials including health education (consumption of sweetened beverages and media use), physical activity and parent involvement in the primary schools. The basic study was a school-based, cluster-randomized intervention trial conducted in the region of Ulm and Günzburg, in Germany [31].

Differences in BMI, WC and WHtR were calculated. Intervention costs were €24.09 per child. The maximum willingness to pay (MWTP) was €35. The ICER (Incremental cost-effectiveness ratio) for WC and WHtR was calculated. ICER was €11.11 per cm WC growth inhibited and €18.55 per unit (0.01) WHtR increase avoided, so the study showed favourable cost-effectiveness ratios. The results proved that prevention of overweight and obesity in a school setting could be cost-effective.

The systematic review of Erdöl Ş et al. included eight studies, five of which concerning community-based primary intervention programs, while three of them were school-based programs. [23] Health outcome measures were evaluated by DALYs, QALYs, BMI scores, kilograms (kg) weight gain prevented and percentage of body fat reduction. The economic measures consisted in Cost Effectiveness (CER) and Incremental Cost Effectiveness ratios for DALY and QALY and cost of intervention per kg weight gain prevented. In summary, three school-based and one community-based primary prevention programs reported were cost effective: two of them considered cost per QALY, one cost for DALY and the last cost per percentage point body fat reduction, while the other studies were not cost effective. Thus, fifty-percent of the studies had positive results. In addition, we have to consider that these studies are low-medium quality and heterogeneous in study designs and outcome measures.

The “Healthy Beginnings” (HB) was an early childhood community-based program, a randomized clinical trial, delivered to families in socio-economically disadvantaged areas of Sydney, Australia during 2007–2010. [28] The HB program consisted of eight home visits by specially trained community nurses, from the 30–36 weeks gestational age up to age two years, with age-appropriate education and advice on feeding, nutrition and physical activity. Health measures were BMI and BMI z-score. Net costs evaluated direct and healthcare costs, in particular the cost of HB intervention in the clinical trial over two years was $1309 per child. Cost-effectiveness analysis was estimated by ICER per BMI unit avoided and cost per 0.1 BMI z-score reduction. The incremental cost-effectiveness ratio was $4230 per unit BMI avoided and $631 per 0.1 reduction in BMI z-score. If the trial reduced the travel time for home visits, it was estimated that the program could costs $709 per child; with incremental cost-effectiveness ratios of $2697 per unit BMI avoided and $376 per 0.1 reduction in BMI z-score. The analyses showed a BMI reduction among the first 2 years of life. The economic analyses showed only short-term benefits of the program using 2 years outcome data, without benefits on longer term. Thus, HB was a moderately priced intervention, with an increased cost-effectiveness if the travel time for home visits was reduced.

A recent overview paper reported the initial results from the Childhood Obesity Intervention Cost-Effectiveness Study (CHOICES), a model to estimate cost-effectiveness of interventions to reduce childhood obesity in the U.S. [25] There were four initial interventions, both community-based and school-based, like childhood obesity intervention programs.

The interventions were the following:

- an excise tax of $0.01 per ounce sugar-sweetened beverages (SSB);

- elimination of the tax deductibility of advertising costs of TV advertisements for “nutrient-poor” food and beverages seen by children and adolescents (TV AD);

- physical education in public elementary school ≥ 50% of physical education class time to moderate and vigorous physical activity (Active PE);

- early child educational setting by increasing physical activity, improving nutrition and reducing screen time (ECE).

The outcomes evaluated in shorter-term (2 years) and longer-term (10 years) were cost per BMI unit change for 2 years following an intervention and 10-year healthcare cost, net cost, DALYs and QALYs. The estimated cost-effectiveness of the interventions for the first two years varied more, ranging from a low of $1.16 per BMI unit change for TV AD, to $3.16 for SSB, $ 57.80 for ECE and $401 for the Active PE intervention. Substantial variations were in 10-year health outcomes (net cost saved per $ spent: $55 for SSB, $38 for TV AD, $6 for ECE). The major saving was for SSB intervention ($23.2 billion) because this impacted on all age groups. At last, SSB intervention averted DALYs and both SSB and TV AD increased QALYs. These results had a lower cost than some medical treatments (e.g. bariatric surgical intervention).

However, this review examined studies with high heterogeneity, in which the population was different in age and with an undefined starting BMI.

Similar to the previous study, other authors developed a microsimulation model to evaluate the cost-effectiveness of seven different interventions identified as potentially important strategies for addressing childhood obesity (both community-based and school-based). [32] The interventions were: an excise tax of one cent per ounce on sugar-sweetened beverages, applied nationally; the elimination of the tax deductibility of advertising costs for television ads seen by children for nutrient-poor food and beverages; restaurant menu calorie labeling; implementation of nutrition standards for federally reimbursable school meals sold; implementation of nutrition standards for all food and beverages sold in schools outside of reimbursable school meals; improved early childhood education policies and practices and a nationwide fourfold increase in the use of adolescent bariatric surgery. The study calculated costs per BMI units reduced over two years (2015–17), health-care costs, net costs, and net costs saved per dollar spent over ten years (2015–25). Moreover, it was estimated the number of obesity cases prevented and changes in childhood obesity prevalence in 2025. In brief, eliminating the tax deduction for advertising nutrient-poor food reduced a BMI unit for $0.66 per person, while increasing access to bariatric surgery could reduce a BMI unit for $1,611. Three of these interventions were projected to save more and reduce health costs (the sugar-sweetened beverage excise tax, eliminating the tax subsidy for advertising unhealthy food to children and setting nutrition standards for food and beverages sold in schools outside of school meals). For example, the beverage excise tax would save $14.2 billion in net costs, over the decade 2015–2025, primarily due to reductions in adult health care costs. Important was the prevention of 576,000, 129,100, and 345,000 cases of childhood obesity, respectively, in 2025, while the net savings to society for each dollar spent were projected to be $30.78, $32.53, and $4.56, respectively.

Conesa and colleagues carried out a cost-effectiveness analysis of a primary prevention school-based intervention of childhood obesity, the EdAI (Educació en Alimentació). [29] It was a randomized, controlled primary prevention intervention implemented in various school of Catalonia, Spain, by healthy lifestyle choices through diet and physical activity (increasing fruit, vegetables, legumes and fish intake for children aged 7–8 years over a period of 28 months). The study analyzed changes in BMI z-score and obesity prevalence and ICER for the number of obesity cases avoided, the decrease in obesity prevalence, the decrease in BMI units, and the decrease in BMI z-score units from the beginning to the end of the intervention was calculated. The ICER was performed only for boys because the intervention was not effective for obesity-related outcomes for girls. After 28 months, the obesity prevalence decreased significantly (by 2.02%) in the intervention group and increased by 0.44% in the control group. In particular for boys, the intervention group exhibited an effective reduction of − 0.24 units in the BMI z-score compared with the control group. Total cost of intervention was 15.64 € per child or 5.21 €/child/year. About ICER, 968.66 € to avoid one case of obesity in boys (1.20 € per boy); 3.56 €/child to reduce the obesity prevalence by 1% in boys; 47.39 € for a decrease of one BMI unit per boy; and 65.17 € for a decrease of one BMI z-score unit per boy. In conclusion, despite gender differences in outcome, EdAI costs 2.4 € per child per year to achieve a greater than 2% reduction in the obesity prevalence, as requested by the 2009 cost effectiveness criteria of the Spanish Health Ministry.

The HELP program was a school-based obesity prevention program in children aged 9–10 years, through primary schools in south-west England, with a follow-up at 24 months. [27] The program consisted of four steps: (1) building a receptive environment, (2) a drama-based healthy lifestyles week, (3) one-to-one goal setting, (4) reinforcement activities. The health outcomes were BMI standard deviation score (SDS) at 24 months post the beginning of the intervention, waist circumference SDS, percentage body fat SDS, proportion of children overweight at 18 and 24 months, accelerometer-assessed physical activity and food intake at 18 months. The study reported no difference in BMI SDS at 24 months or at 18 months, and no differences in waist circumference SDS, percentage body fat SDS or physical activity levels between the intervention and control groups. At 18 months, children in the intervention group consumed less energy-dense snacks compared to children in the control group. The cost of the HELP program was approximately £144,749, with a mean estimated cost per child of £214. The study showed that there were not improvements in the incidence of the weight-related health events or cost savings to the health and social care system associated with weight-related events. The study estimated cost-effectiveness analyses, cost-per-life-year and cost-per-QALY. Through this model, the cost-effectiveness of public health interventions typically resulted in the use of a long-term time horizon, often over 30–40 years, the mean incremental differences in costs and outcomes were relatively small. This is the first no cost-effective study in preventing overweight or obesity in children aged 9–10 years.

In 2019 Canaway published the economic analysis of “WAVES” (The West Midlands ActiVe lifestyle and healthy Eating in School children), a childhood obesity school-based preventing program. [24] The 12-month intervention, in children aged 6–7 years from 54 schools across the West Midlands (UK), was based on the increasing physical activity by 30 min per day and encourage healthy eating. QALYs was calculated as health outcome by using the CHU-9D quality of life questionnaire, validated for use in a pediatric population. The mean difference in QALYs between the group of children who received the intervention and the control group at follow up was 0.006. The incremental cost of the intervention compared to the control arm was £155.53 per child. The incremental cost-effectiveness at 30 months was £26,815 per QALY, less than the standard willingness to pay threshold of £30,000 per QALY, but the calculated probability of the WAVES intervention being cost-effective was only 52%. Therefore, it was not clear if the WAVES program was a cost-effective obesity childhood preventing intervention, because there was a small incremental QALY gain and a major cost when compared to the control group, so it is likely that the incremental of QALYs was due to underlying baseline differences between intervention and control arm and not to an effective health gain.

An Australian study examined the cost-effectiveness of community-based and school-based obesity prevention interventions (CBIs), defined as a community-based program to promote healthy eating and physical activity for children aged 5–18 years. [30] The CBIs consisted in capacity building, awareness raising, physical activity and nutrition strategies implemented in schools, infrastructure changes to schools, and changes to food and physical activity environments within the community. The study was a cost-effectiveness analysis of a review and meta-analysis. The change in BMI has been evaluated for individual age and sex groups from 5–18 years and the health-adjusted life years (HALYs). The meta-analysis revealed a difference in BMI z-score promoting the CBI community compared with the control community in children aged 5–18 years. The estimated net cost of implementing CBIs across all local government areas in Australia was AUD426M over three years. Incremental cost-effectiveness ratio was calculated. If ICER was less than the commonly used willingness to pay threshold for Australia of AUD 50,000 per HALY gained, the intervention was considered cost-effective. The ICER had a value of AUD8155 per HALY gained. CBIs was cost-effective obesity prevention initiatives; however, implementation across Australia would be relatively expensive when compared with current investments in preventive health. Thus, the interventions would be cost-effective over a 29-year time horizon.

Another early community-based intervention program analysis evaluated if childhood overweight prevention intervention, like sleep intervention alone or with food, activity, and breastfeeding advice was cost-effective compared with usual care. [26] The cost-effectiveness analysis was based on Prevention of Overweight in Infancy (POI) clinical trial, that showed an important reduction risk of obesity through sleep intervention, with or without food, activity, and breastfeeding advice (FAB), within the first two years of life. The POI trial included a 3.5- and 5-years follow-up. The Sleep intervention consisted of contact by trained research staff, in which participants received advice about how to promote healthy infant sleep: identifying when infants were tired, helping infants learn how to settle themselves to sleep, and not using feeding as the first response to infant distress. The FAB intervention consisted of parent contact with information on nutrition, physical activity, and breastfeeding. The cost-effectiveness analysis began at five years and modeled cost-effectiveness and cost-utility analysis to age 15 years for both Sleep and Combination interventions, to evaluate the period between early childhood and adolescence. The health outcomes were BMI at age 5 years and QALYs at age 15 years. QALYs was based only on child weight status and the net mean difference in effectiveness was 0.008 in the Sleep group and 0.006 in the combination group for QALYs. The ICERs was calculated as incremental cost per QALY gained and incremental cost per unit BMI avoided in the intervention compared with control. The average costs of the Sleep and Combination interventions were $184 and $601 per child, respectively, with ICER for the Sleep intervention of $18,125 per QALY gained and the ICER for the Combination intervention of $94,667 per QALY gained, while the ICER was $1,022 per unit BMI avoided at age 5 years and $558 per unit BMI avoided at age 15 years in the Sleep group. ICER for combination intervention was higher ($6,678 and $5,164 respectively). Thus, we had 74% probability for Sleep intervention of being cost-effective at a willingness-to-pay threshold of $50,000 per QALY, while the Combination intervention had a 23% probability of being cost-effective. In fact, the addition of FAB in the Combination intervention greatly increased the cost per child with an ICER not cost-effective. In addition, the cost-effectiveness analyses measured in BMI avoided showed that the Sleep intervention was more cost-effective than the Combination program at all willingness-to-pay thresholds. The study concluded that only the Sleep intervention, without food, activity and breastfeeding advice, was a cost-effective approach to prevent childhood obesity, while both interventions were cost-effective over a 15-year time horizon because there was some money saving in health care costs and incremental BMI that occurs during adolescence.

Table 1 summarizes the characteristics of studies included in the systematic review.

Table 1 Description of characteristics of the studies included in the systematic review

Table 2 summarizes economic outcomes.

Table 2 Description of economic outcomesQuality of studies

We used the Drummond Checklist to test quality assessment of the studies included in the systematic review (Table 3). [21, 22] We did not use the checklist for the review by Ş. Erdöl et al., because of the presence of many studies in a single work. [23] All the studies are considered of low-medium quality. Five of the seven items are adherent to study design. Hayes et al. and J. Ananthapavan et al. did not state clearly the rationale for choosing the alternative programs. The presence of a justification for the form of economic evaluation is not clear in relation to the questions addressed in seven studies, except in the two studies by Gortmaker et al. Regarding data collection, eight items are totally adherent. Half of the studies gave details about the method of synthesis and meta-analysis of estimates. Only Wyatt et al. provided details about the subjects from whom evaluations were obtained, while only three studies reported quantities of resources separately from their unit costs. The majority of studies described methods for the estimation of quantities and unit costs, lastly only one gave some details of currency of price adjustments for inflation. Eight of them show information of the model used. There is a lack in the analysis and interpretation of results in many studies. In particular, nine of them are different in the following sections: the discount and the choice of rate, explanation if costs or benefits are not discounted, details of statistical tests and confidence intervals for stochastic data, the approach to sensitivity analysis, the choice of variables for sensitivity analysis, the ranges over which the variables are varied, comparing relevant alternatives, incremental analysis and major outcomes presented in a disaggregated as well as aggregated form.

Table 3 The Drummond checklist and the quality assessment of the studies included in the systematic review