Metrics that matter: Identifying endpoints for capturing the broad health impacts of prevention of obesity

Abstract

Aims

The link between health and economic prosperity is well established, yet quantifying the value of illness prevention remains challenging due to lack of comprehensive metrics. Obesity exemplifies this issue, having widespread societal and health care impacts but limited prevention funding. Existing metrics often fail to capture broader effects of prevention of obesity. This study aimed to identify key components for a holistic metric to assess obesity and cardiometabolic health progression, better articulating value of prevention.

Materials and Methods

We conducted a targeted literature review to identify existing individual and composite metrics for prevention in adults, agnostic of disease or risk factor. We categorized endpoints into: (i) holistic composite measures; (ii) user‐centred composite measures; (iii) health outcome composite measures; (iv) single clinical endpoints. Metrics were evaluated against eight criteria including holistic outcome, cardiometabolic relevance, data feasibility, geographical generalizability, established outcome, economic modelling methodology, stakeholder relevance and preventive value; alongside semi‐structured review from six clinical, health economics and policy experts.

Results

Cardiometabolic endpoints numbering 74 were identified: 7 holistic composites, 16 user‐centred composites, 31 health‐based composites and 20 single clinical endpoints. Five endpoints were shortlisted according to assessment criteria and expert input: waist‐to‐height ratio, low‐density lipoprotein cholesterol, systolic and diastolic blood pressure, blood glucose and the 12‐Item Short Form Health Survey (SF‐12). These endpoints collectively reflected physiological cardiometabolic risk factors with an adverse association with obesity, while SF‐12 provided health‐related quality‐of‐life measurement.

Conclusions

We developed a novel framework that shortlisted five key cardiometabolic outcome measures for assessing obesity primary prevention benefits, with potential applications in health economic modelling and public health.

1. INTRODUCTION

Obesity is one of the biggest public health challenges of recent decades. The prevalence of adult obesity has doubled over the last 30 years, reaching 12.5% in 2022. ¹ Multiple factors, including environmental, psycho‐social, and genetic influences, contribute to the pathogenesis of obesity. In turn, obesity affects multiple cardiometabolic pathways and increases the risk of obesity‐related complications (ORCs) such as cardiovascular diseases (CVD), cerebrovascular disease including stroke, type 2 diabetes mellitus (T2DM) and non‐alcoholic liver disease. ² , ³ The recognition of obesity as an independent risk factor for CVD dates back to the 1980s in a follow‐up study of the Framingham Heart Study. ⁴

Beyond medical consequences, obesity and ORCs negatively impact physical function, psychological wellbeing and social participation. A decline in all these areas subsequently reduces productivity, leading to lower health‐related quality of life (HRQoL). ⁵ On a societal scale, these individual impacts of rising healthcare expenditures, increased absenteeism, and reduced workforce productivity ⁶ accumulate and place a growing burden on healthcare systems and the economy. It is anticipated that overweight and obesity will have a worldwide economic impact of US$4.32 trillion annually by 2035. ⁷ Globally, higher‐income countries and certain geographical regions exhibit higher prevalence of obesity. ⁸

Both treatment and prevention are important to reducing the burden of obesity. For individuals already affected by obesity, interventions such as lifestyle changes, pharmacotherapy and surgery serve as treatment as well as tertiary prevention ⁹ that lowers the risk of obesity‐related complications. By contrast, primary prevention targets the root causes of obesity in the healthy population by promoting healthy behaviour through environmental changes and public health policies. Preventive interventions are especially impactful in specific populations such as children and adolescents, where early interventions can reduce the risk of lifelong health complications. ¹⁰

Recent research and clinical guidance have shed light on the components of holistic evaluation and management of obesity. The Lancet Diabetes & Endocrinology Commission ¹¹ and the European Association for the Study of Obesity (EASO) ¹² both recommended reducing the reliance on body mass index (BMI) and including additional measures in the diagnosis of obesity, such as waist‐to‐height ratio and ability to carry out daily activities. Other analyses have highlighted the psychological considerations ¹³ , ¹⁴ and the range of endpoints for measuring ORC risk. ¹⁵ The need for holistic modelling to drive public health policies is also recognized. ¹⁶ However, there is still a methodological gap to identify comprehensive and robust metrics, especially in the domain of primary prevention.

We aimed to develop a structured framework that could systematically identify key endpoints to be used in the holistic evaluation of obesity primary prevention, focusing on cardiometabolic health as an example.

2. MATERIALS AND METHODS

2.1. Design

A structured approach was developed to construct a holistic metric for cardiometabolic health and primary prevention of obesity based on existing endpoints of outcome measures in the adult population. To capture a broad range of endpoints, we initially defined a classification system for endpoints. Relevant endpoints were identified and extracted through a targeted literature review and were subsequently classified. All extracted endpoints were assessed against predefined criteria and reviewed by subject matter experts to generate a shortlist. The study was set in the UK as a pilot country with the aim of being generalizable to other countries.

2.2. Classification of endpoints

Based on the types of existing outcome measures in general clinical research ¹⁷ , ¹⁸ and the objective to develop a holistic metric, we defined four mutually exclusive groupings to classify endpoints identified during the targeted literature review:

1. Holistic composite: These metrics address a broad range of aspects of good cardiometabolic health and primary prevention of obesity, including the potential societal impacts of obesity prevention. This could include population‐level metrics.

2. User‐centred composite outcome measures: These metrics measure health and wellbeing from the user's perspective and include self‐reported measures relating to quality of life, cardiometabolic or general physical health, and mental health. This aims to capture broader impacts of primary prevention of obesity at an individual level.

3. Health‐based composite: A metric that combines multiple clinical or health‐related measures of cardiometabolic health. For example, this could include risk factors, comorbidity components, and survival components.

4. Single clinical endpoints: Individual clinical cardiometabolic outcomes to measure the impact of primary prevention of obesity on specific clinical endpoints. Examples of single clinical endpoints include blood glucose, lipids, prevalence of type 2 diabetes mellitus (T2DM) and cardiovascular disease mortality.

2.3. Targeted literature review

We used search engines (Google, Google Scholar and PubMed) to identify literature relevant to the four outcome groupings. Search terms used included ‘cardiometabolic health outcomes’, ‘composite score for cardiometabolic health’, ‘cardiovascular health outcomes’, ‘holistic cardiometabolic health outcomes’, ‘measuring improvements in cardiometabolic disease’, ‘measuring prevention of obesity’, ‘measuring prevention of cardiometabolic disease’, ‘holistic composite measures of health’, ‘health wellbeing scales’, ‘mental wellbeing scales’, ‘mental health scales’, ‘mental health metrics’ and ‘holistic health metrics UK’. Relevant sources cited in the results of the literature search using the terms above were also included in the literature review.

A single reviewer screened articles in the search output to ensure comprehensive identification of outcome measures without explicit exclusion, followed by classification into the four outcome groups. All outcome measures identified in included literature were extracted and recorded, along with any information related to the quality assessment criteria described below.

2.4. Quality assessment and expert panel review

We evaluated the quality of the endpoints identified in the literature review using predefined evaluation criteria developed in line with our study objectives. A first reviewer performed the quality assessment followed by quality checking by a second reviewer for all endpoints. Any discrepancies were resolved by a third reviewer, finalizing the ratings.

We created the quality assessment criteria based on domain knowledge of factors considered in the evaluation of outcome measures as well as alignment with the study objective specific to cardiometabolic health and primary prevention of obesity. Cardiometabolic health referred to both the clinical aspects and to wider outcomes such as quality of life. Primary prevention of obesity was targeted to non‐obese adults.

Each endpoint was assigned a rating of ‘good’, ‘intermediate’ or ‘poor’ for each of the evaluation criteria, to reflect the extent to which it met those conditions. An overall rating was assigned for each endpoint based on the rule that if an endpoint had a ‘poor’ rating for any one of the criteria, the overall rating would be ‘poor’. Otherwise, the most frequently assigned rating was used as the overall rating. The initial definitions of the ratings reflected the range of details included by the endpoints identified in the targeted literature review and were described for each criterion.

An initial quality assessment was conducted, followed by a semi‐structured review by a panel of six clinical health economics, and policy experts. The expert panel aimed to review: (i) the assessment criteria, (ii) the endpoints longlist from the targeted literature review, and (iii) the assessment of the endpoints.

The expert panel discussed the scope of the quality assessment criteria and refined definitions of their ratings where appropriate. The final eight assessment criteria were holistic outcome, cardiometabolic health specific, data feasibility, geographically generalizable, established outcome, established methodology in economic modelling, stakeholder relevance (including policymakers, payers, healthcare providers, physicians, patients and caregivers) and represents preventive value (Table 1). Based on the updated criteria, the quality assessment was updated for all the endpoints in the longlist for further review by the expert panel.

TABLE 1.

Quality assessment criteria for cardiometabolic health outcome endpoints.

Cardiometabolic health criteria	Descriptions of quality assessments in relation to each criterion
	Good	Intermediate	Poor
Holistic outcome	Captures CMH, subjective wellbeing and broader societal determinants	Captures some but not all aspects for ‘Good’ rating	Captures a single or limited outcome, insufficient to capture broader impact
Specific to cardiometabolic health	Specifically measures CMH	Captures broader risk factors of CMH	Unrelated to CMH
Data feasibility	Data readily available or published	Limited availability	Data rarely collected or published
Geographically generalizable	Metric validated for use across different geographical locations	Metric validated in a limited set of populations	Use limited to population in which developed
Established outcome	Either used widely in research or included in for example, NICE recommendations, ONS/OHID publications	Evidence of some validation, but not widely used	Study‐specific outcome, not used elsewhere in research, policymaker publications or recommendations.
Established methodology for economic modelling	Evidence of widespread use	Limited use, requires validation	No methodology for economic modelling
Stakeholder relevance	Relevant to broad range of stakeholders	Relevant to specific stakeholders such as clinicians, payers, etc.	Unlikely to be relevant to any stakeholders
Represents preventative value	Enables measurement of good CMH	Used to measure poor health, requires modifications to measure good health	Focus solely on poor CMH, for example, risk equations

Abbreviations: CMH, cardiometabolic health; NICE, National Institute for Health and Care Excellence; OHID, Office for Health, Improvement, and Disparities; ONS, Office for National Statistics.

All endpoints in the longlist were reviewed, and endpoints could be aggregated into a composite where appropriate. Endpoints rated as ‘poor’ overall were excluded from the shortlisting process, which was driven by an endpoint's suitability for capturing the holistic outcomes of cardiometabolic health in the context of obesity primary prevention. Second‐tier endpoints, those conceptually similar to the shortlisted endpoints but less suitable, were listed to highlight the specific factors that influence an endpoint's suitability for measuring prevention outcomes. There were no restrictions on the overall ratings of second‐tier endpoints, nor on how many endpoints could be shortlisted or included in the second tier.

3. RESULTS

3.1. Endpoints from targeted literature review

A total of 74 possible cardiometabolic endpoints were identified in the targeted literature review (Supplementary information). Across the four predefined cardiometabolic health outcome groupings, 7 were holistic composites, 16 were user‐centred composites, 31 were health‐based composites and 20 were single clinical endpoints.

Out of the 74 endpoints, 33 (45%) were rated as ‘good’. Of the 16 user‐centred composites, 7 (44%) were rated as ‘good’. Similarly, 8 of the 31 health‐based composites (26%) and 18 of the 20 single clinical endpoints (90%) were rated as ‘good’. None of the holistic composites were rated as ‘good’ (Table 2).

TABLE 2.

Summary of quality assessment.

Endpoint grouping	Overall rating (%)			Total number of endpoints
	Good	Intermediate	Poor
Holistic composites	0 (0%)	0 (0%)	7 (100%)	7 (100%)
User‐centred composites	7 (44%)	0 (0%)	9 (56%)	16 (100%)
Health‐based composites	8 (26%)	1 (3%)	22 (71%)	31 (100%)
Single clinical endpoints	18 (90%)	0 (0%)	2 (10%)	20 (100%)

The most common assessment criterion for ‘poor’ ratings was holistic outcome, with 21 of all 74 endpoints receiving a ‘poor’ rating. This included over half of the health‐based composites (19 out of 31), indicating their specificity. The second‐most common criterion was established methodology in economic modelling, with 17 out of 74 receiving a ‘poor’ rating. Based on the quality assessment and expert panel review, five endpoints were selected for the final shortlist (Table 3).

TABLE 3.

Suitability of shortlisted endpoints for measuring primary prevention outcomes, compared with second‐tier endpoints.

Outcome	Overall rating	Description
Shortlisted endpoints
Waist‐to‐height ratio	Good	Waist‐to‐height ratio is a recognized measure of abdominal adiposity, which is a key risk factor for metabolic syndrome.
Cholesterol	Good	Low density lipoprotein cholesterol is a key indicator of risk of atherosclerotic disease.
Blood pressure	Good	Systolic and diastolic blood pressure are internationally recognized indicators of cardiovascular health and metabolic syndrome. While these measures alone do not capture the full spectrum of cardiometabolic risk, they serve as critical components of the metric.
Blood glucose	Good	HbA1c measures long‐term glycaemic control, as opposed to acute measurements of plasma glucose. It is useful for measuring the longer‐term impact of preventive interventions as part of the metric.
SF‐12	Good	The 12‐Item Short Form Health Survey (SF‐12) is a 12‐item condensed of the SF‐36 Health Survey (SF‐36), which is used to assess generic health outcomes from the patient's perspective. It covers 8 domains: physical function, role‐physical, bodily pain, general health, energy/fatigue, social functioning, role‐emotional, mental health.
Second‐tier endpoints
The American Heart Association (AHA) Life's Simple 7/Essential 8	Good	Life's Simple 7 as defined by the AHA to measure cardiovascular health. Scores on each item are categorized into ‘Ideal’, ‘Intermediate’ or ‘Poor’, which can be combined across items to provide an overall score of cardiovascular health. In Life's Essential 8, the AHA have added an 8th component to the initial 7 metrics (sleep health) and updated existing components to adopt a new scoring algorithm ranging from 0 to 100 for each metric. This allows the calculation of a new composite cardiovascular health score (the unweighted average of all components) that also varies from 0 to 100 points.
World Health Organization (WHO) definition for metabolic syndrome	Good	Metabolic syndrome refers to a cluster of inter‐related metabolic risk factors which have been found to increase the risk of atherosclerotic cardiovascular disease and type 2 diabetes.
EuroQol 5 Dimension (EQ‐5D)	Good	The EQ‐5D is a generic measure of self‐reported health in five domains: mobility, usual activities, self‐care, pain & discomfort and anxiety & depression. It measures health status or health‐related quality of life (HRQoL). For each domain, users are asked to select from three or five levels of problem they experience in that domain.
Warwick‐Edinburgh Mental Wellbeing Scale (WEMWBS)	Poor	WEMWBS is developed to enable the monitoring of mental wellbeing in the general population and the evaluation of projects, programmes and policies which aim to improve mental wellbeing. Subjective ratings collected based on subjective agreement with 14 items (or 7 items if using the 7‐item scale). Agreement is rated on a scale of 1–5, where 1 means none of the time and 5 means all the time.
Impact of Weight on Quality of Life‐Lite Clinical Trials version (IWQOL‐Lite‐CT)	Poor	The IWQOL‐Lite is a validated, 31‐item, self‐report measure of obesity‐specific quality of life in adults. Contains 7 items concerning physical health and 13 items concerning psychosocial health.

Note: The description is Green shade is labelled as good and red shade is labelled as poor.

3.2. Shortlisted endpoints

The shortlisted endpoints consist of four single clinical outcomes and one user‐centred composite outcome. All holistic composite metrics were excluded from the shortlist due to poor ratings in key domains such as cardiometabolic specificity and lack of established methodology for economic modelling.

The four single clinical outcomes are waist‐to‐height ratio, blood cholesterol, blood pressure and blood glucose. Waist‐to‐height ratio, the ratio between the waist circumference and height, measures central adiposity. A waist‐to‐height ratio ≥0.5 generally indicates higher cardiometabolic risk. ¹² , ¹⁹ Low‐density lipoprotein (LDL) cholesterol, systolic and diastolic blood pressure, and glycated haemoglobin (HbA1c) are used in the diagnosis and monitoring of hypercholesterolaemia, hypertension and type 2 diabetes, respectively.

The user‐centred outcome, the 12‐Item Short Form Health Survey (SF‐12), ²⁰ is a self‐reported questionnaire to assess generic health outcomes from users' perspectives. It is a subset of the Medical Outcomes Study (MOS) 36‐item Short‐Form Health Survey (SF‐36) ²¹ and covers eight domains on how individuals feel, and their ability to perform day‐to‐day activities, namely physical function, role‐physical, bodily pain, general health, energy/fatigue, social functioning, role‐emotional and mental health.

The expert panel proposed that the four clinical endpoints would be combined with weights using a binary or dichotomous/trichotomous scoring system. The proposed thresholds for the scoring system are based on the UK setting (Table 4). Under this system, values exceeding the threshold receive a score of 1 (indicating worse health) and those below receive a score of 0 (indicating better health), except for HbA1c, where values above a secondary threshold are assigned a score of 2. The experts highlighted the need for different thresholds depending on population and geographical location.

TABLE 4.

Proposed thresholds for scoring clinical endpoints in the cardiometabolic health composite metric.

Risk factor	Proposed threshold	Score
Blood glucose (HbA1c) 22 , 23 , 24	HbA1c < 42 mmol/mol (6%)	0
	HbA1c ≥42 mmol/mol (6%) and ≤47 mmol/mol (6.4%)	1 (Prediabetes)
	HbA1c ≥ 48 mmol/mol (6.5%)	2 (Diabetes)
Cholesterol (low density lipoprotein) 23 , 25	LDL ≤ 3 mmol/L	0
	LDL > 3 mmol/L	1
Waist‐to‐height ratio 12 , 26 , 27	Waist‐to‐height ratio < 0.5	0
	Waist‐to‐height ratio ≥ 0.5	1
Blood pressure 28	Systolic/diastolic blood pressure < 140/90 mmHg	0
	Systolic/diastolic blood pressure ≥ 140/90 mmHg	1

Abbreviations: HbA1c, glycated haemoglobin; LDL, low‐density lipoprotein.

3.3. Second‐tier endpoints

Second‐tier endpoints included the American Heart Association (AHA) Life's Simple 7, ²⁹ AHA Life's essential 8, ²⁹ the World Health Organization (WHO) criteria for Metabolic Syndrome, ³⁰ a combination of the EuroQol 5 Dimension (EQ‐5D) ³¹ and Warwick‐Edinburgh Mental Wellbeing scale (WEMWBS) ³² and the Impact of Weight on Quality of Life‐Lite Clinical Trials version (IWQOL‐Lite‐CT) (Table 3). ³³

The AHA introduced ‘Life's Simple 7’ which outlined 7 metrics for defining cardiovascular health ³⁴ and subsequently expanding and updating this to ‘Life's Essential 8’ in 2022. ²⁹ Components of Life's Essential 8 include diet, physical activity, nicotine exposure, BMI, blood lipids, blood glucose, blood pressure and a new component of sleep health. ²⁹ While the AHA's proposed metrics closely align with some of the single clinical endpoints in the shortlist, the use of BMI over waist‐to‐height ratio and the larger number of components made the AHA measures less suitable.

The WHO criteria for Metabolic Syndrome include components including waist‐to‐hip ratio, blood glucose, blood pressure and blood lipid levels. ³⁰ However, some of the ‘absolutely required’ criteria, such as diagnosis of impaired glucose intolerance, impaired fasting glucose or T2DM are not applicable to individuals in good health.

The EQ‐5D and WEMWBS cover many of the aspects of HRQoL addressed by the SF‐12. The EQ‐5D is a generic measure of self‐reported health in five domains: mobility, usual activities, self‐care, pain/discomfort and anxiety/depression. ³¹ However, its mental health coverage is limited, focusing only on anxiety and depression. ³¹ By contrast, the SF‐12 captures these aspects while also reflecting broader elements of mental health and wellbeing, like, feeling calm and peaceful versus feeling down‐hearted and blue, having a lot of energy and the extent to which physical health or emotional problems interfere with social activities. ²⁰ The WEMWBS can complement the EQ‐5D to address gaps in mental wellbeing measurement, as it was developed to monitor mental wellbeing in the general population and measures subjective agreement across 14 items. ³² However, WEMWBS is not widely or routinely used, which may limit data availability. It may also be more burdensome for individuals to complete both the EQ‐5D and WEMWBS compared with the single SF‐12 survey.

The IWQOL and the shorter IWQOL‐Lite are questionnaires that aim to specifically assess the effects of obesity on health‐related quality of life. ³³ However, the focus of IWQOL on obesity restricts its use as a measure of primary prevention and cardiometabolic health more broadly.

4. DISCUSSION

Prevention often involves diverse and long‐term benefits which cannot be fully captured by single outcome measures. To address this gap, we developed a structured framework to identify key endpoints to be used in the holistic evaluation of obesity primary prevention, focusing on cardiometabolic health as an example. This metric integrates five clinical and user‐centred endpoints, including waist‐to‐height ratio, blood cholesterol, blood pressure, blood glucose, and SF‐12. This structured framework highlighted the conjoint impact of the shortlisted endpoints in the context of obesity primary prevention.

4.1. Rationale and existing applications

The structured approach in this study is novel but shares conceptual parallels with the evaluation framework for obesity prevention efforts proposed by the Institute of Medicine ³⁵ and the recommendations from the core outcome measures in effectiveness trials (COMET) initiative ³⁶ guiding the development of core outcome sets for clinical trials. Both aim to create a reproducible set of endpoints that can be widely adopted, use a stepwise process and pre‐specified scope and criteria, and emphasise stakeholder consensus. Our approach for endpoint selection leverages existing, well‐validated measures of cardiometabolic risk factors and HRQoL.

The shortlisted endpoints can be used in assessing, comparing, or monitoring the impacts of obesity prevention interventions across healthcare providers in various settings.

4.1.1. Single clinical endpoints

Waist‐to‐height ratio was included in the new framework for the diagnosis, staging and management of obesity in adults from the EASO, which recognized the association between abdominal adiposity and increased risk of developing cardiometabolic complications. ¹² Waist‐to‐height ratio is superior to waist circumference as a cardiometabolic disease risk marker. ¹² , ²⁶ In addition, it is easier to implement than waist‐to‐hip ratio, especially for preventive interventions aimed at healthy individuals who may perform self‐measurement.

Metabolic syndrome is an ORC and refers to a set of risk factors associated with a higher risk of T2DM and atherosclerotic cardiovascular disease. Measures of blood lipids, blood pressure, and blood glucose are commonly included in the diagnostic criteria for metabolic disease, where the presence of a greater number of risk factors generally indicates poorer cardiometabolic health. Widely recognized definitions include those by the WHO in 1998, ³⁰ subsequent definitions by the AHA/National Heart, Lung and Blood Institute (NHLBI) ³⁷ and the International Diabetes Federation (IDF). ³⁸

The shortlisted single endpoints are established outcome measures for cardiovascular risk that also capture metabolic aspects. Systolic and diastolic blood pressure are the standard measures for diagnosis of hypertension and metabolic syndrome. LDL cholesterol reflects the broader spectrum of cardiometabolic health, compared with triglycerides and high‐density lipoprotein for metabolic syndrome. HbA1c is a measure of longer‐term glycaemic control, compared with fasting plasma glucose in the IDF definition of metabolic syndrome. ³⁷

The scoring system proposed to combine the single clinical endpoints illustrated the operationalization of using the endpoints in evaluation. Rather than evaluating the endpoints individually by categorising them by primary and secondary or ordering by a hierarchy of importance, our approach was more adaptive. The various thresholds and weights of the component endpoints could be adjusted to various settings such as intervention evaluation, health economic modelling, and decision‐making in population health.

4.1.2. User‐centred composite outcome measure

Although SF‐36 contains more questions than SF‐12 and allows more reliable conclusions to be made about specific domains, ³⁹ SF‐36 and SF‐12 are highly correlated. The majority of the variation in QOL is captured in the shorter SF‐12, based on studies conducted across different clinical areas from obesity to osteoarthritis. ²⁰ , ³⁹ , ⁴⁰ In some cases, SF‐12 was viewed as a better measure of HRQoL, such as when measuring differences associated with BMI. ³⁹ The SF‐12 is also widely used by healthcare professionals, researchers, and policy makers.

4.2. Wider implications

There is growing recognition that health and the value of good health are linked to broader economic and societal benefits, particularly for non‐communicable diseases. ⁴¹ , ⁴² The Lancet Diabetes & Endocrinology Commission created the classification of ‘clinical obesity’ and ‘pre‐clinical obesity’, and recommended risk reduction management strategies for the latter category. ¹¹ The EASO recommended staging obesity ‘as a chronic, relapsing disease, according to the severity of its medical, mental and functional complications’. ¹² While these recent recommendations have shown a broadening of scope in managing obesity, there are still gaps in primary prevention.

Validated and established endpoints were more suitable for our cardiometabolic health outcomes framework due to practical reasons. The shortlisted endpoints are commonly recorded in the electronic health records, meaning healthcare providers can easily work with familiar endpoints. Moreover, all endpoints are already used or likely suitable for economic modelling with established methodology. There are several examples of metabolic syndrome being used in models of cost‐effectiveness, modelling economic burden of metabolic syndrome and forecasting of future costs. ⁴³ , ⁴⁴ , ⁴⁵ The overlap between our proposed metric and established definitions of metabolic syndrome supports its potential use. The SF‐12 is widely used to measure self‐reported HRQoL and can be used for constructing generic health outcome measures such as quality‐adjusted life years (QALYs). Collectively, the endpoints can be used for risk stratification, and further research would help expand its use by linking the metric with specific outcomes.

Multiple stakeholders would be able to use this consistent set of measures to evaluate the value of obesity primary prevention. The endpoints could be incorporated into clinical practice and public health programmes to facilitate data collection across the population. They may be further integrated into clinical prediction programmes to stratify patient risk and monitor temporal trends. Individuals eligible for preventive interventions could better understand the health outcomes that matter to them, such as functional abilities and quality of life. Payers and policymakers could use the endpoints to inform reimbursement decisions and shape public health strategies.

4.3. Strength and limitations

This study utilized a structured approach to generate a holistic cardiometabolic health metric. The pre‐defined elements in our approach ensured broad inclusion of endpoints and reduced selection biases during the shortlisting process. Stakeholders across the society, including policymakers, payers, healthcare providers, physicians, patients, and caregivers were considered in the assessment of stakeholder relevance.

The proposed framework guided the creation of a novel, intervention‐agnostic metric in representing the overall preventive value for obesity in cardiometabolic health that can be applied to any preventive intervention. This ensures the frequency, ease, and cost of measurement, and supports the use in economic assessments.

While the UK was considered a pilot country in this study, geographical generalizability was included as one of the assessment criteria. The thresholds for clinical endpoints may be adjusted based on different genetic backgrounds and local clinical practices. In addition, the interpretation of the endpoints might vary between at‐risk populations, the general population, and populations of different ages. While the single clinical endpoints are age‐independent, SF‐12 is used only in the adult population. Future research should consider the underlying characteristics of the study populations when applying the endpoints.

A limitation at the targeted literature review stage was the absence of formal inter‐rater reliability metrics. However, consistency of assessment was maintained by using a structured rating system and incorporating deliberation throughout the process.

The expert panel suggested that equity and mental health/wellbeing may be considered more systematically in the future. Although these two criteria are not immediately operationalisable, they are significant in the primary prevention of obesity and are covered by some of the endpoints assessed including the shortlisted SF‐12. Recent wellbeing measures such as WELLBY ⁴⁵ could be useful in future research.

A scoring system was proposed to combine the shortlisted clinical endpoints but not the SF‐12. The expert panel agreed that the ratio of specific levels of clinical risk factors may not always correlate with quality‐of‐life measures. Integration of the SF‐12 into this study's scoring system may involve more complex adjustments to the original SF‐12 scoring algorithm involving population average and variance. ²⁰ Nevertheless, the importance of a user‐centred outcome measure for HRQoL was recognized in the outcomes framework.

5. CONCLUSION

We developed a structured framework for generating a cardiometabolic health metric that captures the clinical and wider outcomes of primary prevention of obesity. The five shortlisted endpoints are applicable in obesity prevention across clinical practice, health economic evaluation and public health research. This framework and the endpoints can be adapted to different geographical regions and be tested in further research to articulate the impact of preventive initiatives.

AUTHOR CONTRIBUTIONS

All authors critically reviewed and approved the final version of the manuscript and accept responsibility to submit for publication. A medical writer (Florence Ma) contracted by the sponsor provided assistance in preparing the manuscript.

FUNDING INFORMATION

This study was funded by Novo Nordisk A/S.

CONFLICT OF INTEREST STATEMENT

JP‐S is Partner and Head of Health Analytics at Lane Clark & Peacock, Chair of the Royal Society for Public Health and reports personal fees from Novo Nordisk A/S outside of the submitted work. SH and JK are employees of Lane Clark & Peacock LLP. IP, MG and LG report personal fees from Lane Clark & Peacock LLP. RR is an employee of Novo Nordisk Health Care. KB is an employee of Novo Nordisk Ltd. JKHS is an employee of Novo Nordisk A/S.

PEER REVIEW

The peer review history for this article is available at https://www.webofscience.com/api/gateway/wos/peer-review/10.1111/dom.70016.

ETHICS STATEMENT

Research Ethics Committee was not required as the study did not involve patients nor individual‐level data.

Supporting information

Data S1. Supporting information.

Pearson‐Stuttard J, Holloway S, Kettle J, et al. Metrics that matter: Identifying endpoints for capturing the broad health impacts of prevention of obesity. Diabetes Obes Metab. 2025;27(11):6275‐6283. doi: 10.1111/dom.70016

DATA AVAILABILITY STATEMENT

Data sharing is not applicable to this article as no datasets were generated or analysed during the current study.