Scientists Can Now Predict Obesity Risk in Early Childhood

A large-scale genetic analysis involving over five million people has revealed new insights into obesity risk and enabled reliable predictions from early childhood through adulthood.

Led by researchers from the Universities of Copenhagen and Bristol and published in Nature Medicine, the study introduces a new polygenic risk score (PGS) capable of identifying individuals, including children, at high genetic risk for obesity – offering opportunities for earlier, targeted intervention.

Obesity: A growing global crisis

By 2035, over half of the world’s population is expected to be overweight or obese – a trend with alarming implications for global health and economic stability.

While investing in the prevention and treatment of obesity may require substantial financial resources, the long-term cost of inaction – including rising healthcare expenditures, reduced workforce productivity and increased burden of chronic disease – will be significantly greater.

While lifestyle changes, bariatric surgery and pharmacological treatments are available, they remain inaccessible or ineffective for many people. Scalable, equitable and personalized strategies for prevention and treatment are, therefore, urgently needed.

Building a genetic tool to predict obesity

To help address this gap, researchers developed a new PGS using genetic data from more than five million individuals – making it the most diverse and comprehensive dataset of its kind to date. The data came from sources such as the Genetic Investigation of ANthropometric Traits (GIANT) consortium and 23andMe, a consumer DNA testing company.

A PGS estimates an individual’s susceptibility to a condition based on the combined effects of thousands of genetic variants. In this study, the PGS was evaluated in a cohort of over 500,000 individuals, including long-term participants in the Children of the 90s study – a UK-based birth cohort that has tracked health outcomes over decades.

According to the researchers, the new PGS was twice as effective at predicting a person’s risk of developing obesity compared to previous models. It accounted for nearly 17% of the variation in body mass index (BMI) across individuals – substantially more than earlier versions, which typically explained less than 10%.

Early identification, early intervention

The strength of the new PGS lies in its ability to predict obesity risk from a very young age. Patterns linking the genetic score to BMI were evident before the age of five and remained consistent into adulthood.

“What makes the score so powerful is the consistency of associations between the genetic score and BMI before the age of five and through to adulthood – timing that starts well before other risk factors begin to influence weight,” explained Dr. Roelof Smit, assistant professor at the University of Copenhagen and lead author of the study. “Intervening at this point could theoretically make a huge impact.”

Dr. Kaitlin Wade, associate professor in epidemiology at the University of Bristol, added: “Obesity is a major public health issue with multiple contributing factors, including genetics, environment and behavior. Many of these factors begin to take shape in childhood. This tool could help identify at-risk individuals early, improving prevention and outcomes.”

The study also investigated how genetic risk affects response to lifestyle-based weight loss interventions, such as changes in diet and exercise. The researchers found that individuals with a higher genetic predisposition to obesity lost more weight during interventions – but were also more likely to regain it once the intervention ended.

This finding reinforces the idea that genetics is not destiny. While genetic risk may shape a person’s susceptibility, behavioral and environmental interventions can still have a meaningful impact – particularly when they are sustained over time.

The importance of inclusive genetic research

Despite its use of the most diverse genetic dataset assembled to date, the PGS showed variation in predictive power across ancestral backgrounds. It was significantly more accurate for individuals of European ancestry than for those of African ancestry.

This gap underscores a long-standing issue in genetic research: underrepresentation of non-European populations. To build tools that are truly equitable and clinically useful for all populations, the researchers say future studies must prioritize diversity and inclusion in genomic research.

Toward a healthier future

The development of a more accurate and accessible genetic risk tool for obesity may mark an important step toward personalized prevention strategies. By identifying individuals at high risk earlier in life, public health systems could implement targeted interventions – potentially reducing the long-term burden of obesity and its related diseases.

“We hope this work will contribute to detecting individuals at high risk of developing obesity at an earlier age,” said Wade. “This could have a vast clinical and public health impact in the future.”

Reference: Smit RAJ, Wade KH, Hui Q, et al. Polygenic prediction of body mass index and obesity through the life course and across ancestries. Nat Med. 2025:1-18. doi: 10.1038/s41591-025-03827-z

This article is a rework of a press release issued by the University of Bristol. Material has been edited for length and content.