UK Researchers Develop Obscore Tool to Predict Obesity Risks and Prioritise NHS Weight-Loss Drugs 2026

Key Points

• Researchers from University of Cambridge and Queen Mary University of London have developed a new data tool called Obscore to identify individuals most at risk of obesity-related diseases.
• The tool uses interpretable machine learning on data from nearly 200,000 UK Biobank participants with BMI of 27 or higher (overweight or obese).
• Obscore predicts 10-year risk for 18 obesity-related conditions, including type 2 diabetes, stroke, and gout, based on 20 health, lifestyle, and demographic factors like age, sex, total cholesterol, and creatinine levels.
• It categorises people into five risk groups (low to high) for each condition, showing varied risks even among those with same age, sex, and BMI.
• Tool could help NHS prioritise limited weight-loss medications like tirzepatide, moving beyond simple BMI thresholds.
• Validation used UK Biobank data and two independent health studies; also tested on tirzepatide trial data, confirming similar weight loss benefits across risk groups.
• About two-thirds of adults in England are overweight or obese, per recent government data.
• Prof Nick Wareham emphasises rational resource allocation for those most likely to benefit.
• Kamil Demircan highlights overweight individuals in high-risk categories who might be overlooked by BMI alone.
• Prof Naveed Sattar praises the holistic approach but notes needs for further development due to interrelated conditions and non-routine metrics.

Cambridge (Britain Today News) April 30, 2026 – Scientists at the University of Cambridge and Queen Mary University of London have launched Obscore, a pioneering data tool designed to pinpoint individuals most vulnerable to obesity-related diseases, potentially transforming how the National Health Service allocates scarce weight-loss medications.

The innovation arrives at a critical juncture, with recent government statistics revealing that approximately two-thirds of adults in England are overweight or obese. This alarming trend has sparked widespread concern among health experts, who argue that much of the NHS does not yet treat obesity with the urgency it demands.

Obscore employs interpretable machine learning, a form of artificial intelligence that provides transparent insights into its predictions. Applied to data from almost 200,000 participants in the long-running UK Biobank project—all with a body mass index (BMI) of 27 or greater—the tool identifies 20 key features. These include age, sex, total cholesterol levels, and creatinine measurements, among others.

How Does Obscore Predict Risks for 18 Obesity-Related Conditions?

By analysing these factors, Obscore forecasts the 10-year risk of 18 specific complications linked to obesity, ranging from gout and type 2 diabetes to stroke and certain cancers. For each condition, the tool stratifies participants into five equal-sized risk categories, from lowest to highest. Researchers then calculated the proportion of individuals in each category who developed the condition over a decade.

This granular approach reveals stark differences: people sharing the same age, sex, and BMI can exhibit wildly varying risks. As reported in the journal Nature Medicine, the study’s authors demonstrated this variability through rigorous testing on UK Biobank data, supplemented by two independent health studies.

The tool’s validity holds firm across datasets, underscoring its potential reliability.

“These constitute a population of individuals who may be overlooked if we only look at BMI and not other risk factors,”

said Kamil Demircan, a co-author from Queen Mary University of London, highlighting how overweight (rather than obese) people often populate the highest-risk groups for conditions like type 2 diabetes.

Why Is Current NHS Access to Weight-Loss Jabs Limited by BMI Alone?

NHS eligibility for weight-loss injections, such as tirzepatide or semaglutide, currently hinges on elevated BMI combined with specific comorbidities. Demand far outstrips supply, leaving many without access. Obscore aims to refine this by prioritising those with the greatest clinical need and likely benefit.

Prof Nick Wareham, of the University of Cambridge and a co-author, clarified the tool’s intent.

“It’s about developing and validating a score that can help with more rational resource allocation. So, can we prescribe therapy to those people who are most likely to need it and most likely to benefit from it – which is what we should do within the NHS,”

he stated.

To test real-world applicability, the team adapted Obscore for data from a randomised controlled trial of tirzepatide. Results showed that high-risk individuals experienced weight loss comparable to others, suggesting the tool identifies candidates who respond well without excluding broader groups.

What Makes Obscore More Accurate Than Traditional BMI Measures?

Unlike BMI, which offers a blunt snapshot of weight relative to height, Obscore integrates a multifaceted profile. It draws on easily measurable biomarkers like cholesterol and creatinine—routinely checked in clinical settings—alongside demographics and lifestyle indicators.

The interpretable machine learning ensures clinicians understand the “why” behind predictions, fostering trust. For instance, in the highest-risk quintile for type 2 diabetes, a notable share comprised overweight individuals, not just the obese. This challenges BMI-centric policies, as Demircan noted.

Government data from May 2025 underscores the stakes: obesity profiles indicate escalating prevalence, straining NHS resources. Experts have long criticised insufficient attention to obesity, with calls for proactive strategies.

Who Validates Obscore and What Challenges Remain?

Independent validation bolsters Obscore’s credentials. Beyond UK Biobank’s half-million volunteers, the tool performed consistently in external cohorts. Yet, Prof Naveed Sattar, a professor of cardiometabolic medicine at the University of Glasgow and not involved in the research, offered measured praise.

“Overall, this work represents a thoughtful attempt to move towards more holistic risk prediction across multiple obesity‑related conditions,”

Sattar said. He cautioned, however, that many conditions interlink closely, and robust, simpler scores already exist for some. Metrics like creatinine may not be universally available in primary care.

“Substantial further development and validation will be required before such an approach can be translated into routine clinical practice,”

Sattar added.

Could Obscore Transform NHS Weight Management Strategies?

Proponents envision Obscore streamlining interventions. With weight-loss drugs costing thousands annually per patient, prioritisation becomes ethical and economic imperative. Wareham reiterated: the focus is not expanding therapies but targeting them effectively.

The tool’s flexibility shines in trial data: high-risk participants lost weight similarly to low-risk ones on tirzepatide, implying broad efficacy while aiding selection.

Obesity’s toll is immense—linked to heart disease, diabetes, cancers, and mental health burdens. Two-thirds of English adults affected signals a public health crisis. Obscore could empower GPs to intervene precisely, potentially averting thousands of cases.

What Lifestyle and Demographic Factors Drive Obscore Predictions?

The 20 predictors span categories: age and sex as baselines; biochemicals like total cholesterol and creatinine for metabolic insights; lifestyle proxies inferred from Biobank data, such as physical activity levels and smoking history.

Demographic elements ensure equity, accounting for variations across populations. For stroke risk, older males with elevated cholesterol might score highest, while younger females with normal BMI but poor kidney function (via creatinine) could surprise.

This nuance addresses BMI’s flaws—it ignores muscle mass, fat distribution, or genetics. Obscore’s machine learning sifts vast data to surface patterns humans might miss.

How Does UK Biobank Enable Such Breakthroughs?

UK Biobank, tracking 500,000 Britons since 2006, provided the gold-standard dataset. Participants, aged 40-69 at enrolment, underwent detailed assessments—blood tests, scans, questionnaires. With BMI ≥27 filtered, the 200,000-strong subset yielded robust statistics.

Privacy safeguards are paramount, amid ongoing debates. Yet, its scale enabled AI training without overfitting, ensuring generalisability.

Will Obscore Integrate into NHS Systems Soon?

No timeline exists, but researchers urge pilots. Sattar’s caveats highlight hurdles: data infrastructure, clinician training, equity in access. Routine metrics must expand; interoperability with electronic health records is key.

If scaled, Obscore could save lives and costs. Modelling suggests preventing one diabetes case yields £10,000+ savings over a decade.

What Broader Implications for Obesity Policy?

This tool spotlights obesity’s complexity, urging holistic policies. Beyond drugs, it complements lifestyle programmes, surgery referrals. Politicians face pressure to fund amid fiscal squeezes.

Health Secretary Wes Streeting has pledged obesity crackdowns; Obscore aligns, offering evidence-based precision.

Critics worry over-reliance on AI, but transparency mitigates this. Wareham’s vision: equitable NHS care.

In summary, Obscore marks a leap in personalised medicine for obesity. As England grapples with its weight crisis, this Cambridge-led innovation promises smarter, fairer interventions. Further trials will determine its legacy.