Childhood Blood Proteins May Signal Heart, Liver Disease Risk Decades Early

A blood sample drawn from a 13-year-old may carry warning signs of heart, liver and metabolic disease that wouldn’t show up on a standard checkup for decades. Researchers tracked thousands of proteins in children’s blood and found patterns that echoed known markers of adult metabolic disease, patterns that went on to predict new cases of disease and death in tens of thousands of separate adults tracked for more than a decade.

Published in Nature Metabolism, the study linked protein patterns in Hispanic and Latino children and teenagers to cardiovascular-kidney-metabolic disease, or CKMD, a term researchers use for overlapping heart, kidney and metabolic problems like diabetes and fatty liver disease. Scientists applied those childhood signatures to more than 28,000 adults in the UK Biobank database, tracked for a median of nearly 14 years. Adults whose profiles matched riskier childhood patterns faced substantially higher odds of serious illness and death, even though none were the same people as the children studied.

This blind spot matters: pediatric screening often misses trouble brewing under the surface. Standard tests like fasting blood sugar can miss a large share of kids heading toward diabetes, and children can show blood pressure or cholesterol readings that look unremarkable yet are still tied to greater disease risk later in life. This research points to a more sensitive way to catch it early, written into the proteins circulating in a child’s bloodstream.

Childhood Protein Scores Predicted Death, Diabetes and Heart Disease in Adults

Researchers worked with 273 Hispanic or Latino children and teenagers, averaging about 13 years old, from a community health study in South Texas along the U.S.-Mexico border. About half were female, more than a third had obesity, and many already showed early signs of unhealthy cholesterol and insulin resistance, a condition where the body struggles to use insulin properly and which often comes before diabetes.

Scientists measured 25 health markers, covering body fat, liver fat and scarring, kidney function, cholesterol, blood sugar control and blood pressure, then checked how each related to more than 5,000 proteins in the children’s blood. They found over 1,000 proteins meaningfully connected to at least one marker, producing nearly 3,000 protein-health links. Some already had known roles in adult disease, like leptin, long tied to appetite and fat storage. Others were less expected, including BACE-1, a protein usually studied in Alzheimer’s research, which turned up linked to blood sugar problems and body fat in the kids.

Researchers then checked whether these patterns held up in adults, comparing findings with 685 adults from the same South Texas community. Of the nearly 3,000 links found in kids, 64% showed the same direction and strength in adults, with the strongest matches in body fat and liver health.

From there, the team condensed 16 measurements into six broader summary scores, covering inflammation-related body fat, liver fat and scarring, cholesterol, blood pressure, kidney function and insulin resistance. Tested against the larger UK Biobank group, scores for inflammation-related body fat, liver problems and insulin resistance predicted higher rates of death from any cause, type 2 diabetes, several heart conditions, sleep apnea and fatty liver disease. The cholesterol score moved the opposite direction, which fits how it was built, since a higher score there reflected a less adverse childhood pattern.

Semaglutide Shifted Childhood-Risk Proteins in a Healthier Direction

Researchers didn’t stop at spotting these patterns; they wanted to know whether proteins tied to childhood risk could actually be changed, suggesting they aren’t just passive warning signs but possible treatment targets. So the team turned to data from a clinical trial of semaglutide, the drug marketed for weight loss and diabetes under brand names like Wegovy and Ozempic. This class of drugs has exploded in popularity in recent years, including among children; prescriptions for young people reportedly jumped nearly 600% between 2020 and 2023.

Scientists mapped more than 1,000 childhood-linked proteins onto the semaglutide trial data. Of those examined, hundreds shifted after 68 weeks of treatment, and ones tied to worse childhood metabolic health generally moved healthier, though the authors note they cannot separate the drug’s effect from that of the weight loss it produced.

Finally, genetic data offered evidence some proteins may be involved in causing disease rather than simply appearing alongside it, linking many childhood-flagged proteins to adult heart failure, obesity, type 2 diabetes, kidney problems and fatty liver disease. A separate dataset tracking 100 people from age 4 to 24 found much of this protein variation wasn’t explained by aging or sex, suggesting diet, activity or environment help drive it, leaving room for risks to be caught early.

This Blood Test Isn’t Ready for Pediatricians Yet

None of this means every child needs a blood protein panel at their next checkup. Crucially, the children weren’t followed into adulthood to see whether their own profiles predicted their own future disease; that link came from separate, unrelated adults, since tracking the same children for decades would take that long. The study was also conducted in one community, and the authors say more research is needed elsewhere to confirm the patterns hold.

Still, the message is notable. Heart, kidney and metabolic disease has long been treated as an adult problem, but the biological groundwork appears to form far earlier than screening guidelines account for. As childhood obesity climbs and drugs like semaglutide become more common in pediatric care, a sharper way to spot children who may be at higher risk, rather than relying on body weight alone, could eventually change how early prevention works.

---

Disclaimer: This article summarizes findings from a single peer-reviewed study and is intended for general informational purposes. It is not medical advice and should not be used to diagnose or guide treatment decisions. Anyone with questions about a child’s individual health should consult a pediatrician or other qualified healthcare provider.

---