A person’s body can age faster than the calendar suggests, and that gap may carry important clues about dementia risk.

In a study of more than 220,000 UK Biobank participants, researchers at King’s College London found that people whose biological age appeared older than their chronological age were more likely to develop dementia over time. They were also more likely to develop it sooner. The pattern was especially strong for vascular dementia, a form linked to reduced blood flow in the brain.

The work points to a simple idea with large consequences. Two people may be the same age on paper, but one may show signs of faster internal aging in the blood. That difference, the researchers say, could help identify people who face a greater chance of dementia before symptoms begin.

“Our findings suggest that biological ageing data can help identify individuals at risk of dementia before clinical symptoms emerge,” said lead author Dr. Julian Mutz, King’s Prize Research Fellow at the Institute of Psychiatry, Psychology & Neuroscience at King’s College London. “By combining genetic factors with potentially modifiable factors captured in biological ageing, we may be able to develop preventative strategies, potentially based on a simple blood test.”

Dementia already affects an estimated 982,000 people in the UK, and that number is projected to rise to 1.4 million by 2040. In 2023, dementia and Alzheimer’s disease together were the leading causes of death in England and Wales. They accounted for 11.6% of all registered deaths.

The team focused on what is known as a metabolomic aging clock. It is built from metabolites, small molecules found in blood that reflect how the body processes energy and carries out other chemical functions. By comparing a person’s blood-based biological age with their actual age, the researchers calculated a value called MileAge delta. A positive value meant the body appeared older than expected.

The study included 223,496 adults after exclusions. Participants were drawn from UK Biobank, which recruited people aged 37 to 73 at baseline between 2006 and 2010. Over follow-up, 3,976 participants developed dementia. That total included 1,881 cases of Alzheimer’s disease, 933 cases of vascular dementia, 512 cases of dementia in other diseases, and 988 cases of unspecified dementia.

People whose biological age was at least one standard deviation above the average had a 24% higher hazard of all-cause dementia than those whose biological age was at least one standard deviation below it, after full adjustment for factors including age, sex, education, income, deprivation, fasting time, and APOE genotype. For vascular dementia, the increase was much larger, 61%.

The same blood-based aging measure was also tied to earlier onset. Higher MileAge delta was associated with an earlier age of onset for Alzheimer’s disease, vascular dementia, unspecified dementia, and all-cause dementia. The link held even after adjustment for social and genetic factors.

That detail stands out. The aging measure was not significantly associated with incident Alzheimer’s disease itself, yet it still tracked with earlier Alzheimer’s onset among people who did develop the disease.

The analysis also looked at genetic risk, including APOE genotype and dementia polygenic scores. APOE is a well-known dementia risk gene, and the ε4 version is especially important. People who carried two copies of APOE ε4 faced the greatest genetic risk in the study.

When advanced biological aging and high genetic risk appeared together, the numbers became much sharper. For all-cause dementia, individuals whose biological age exceeded their chronological age and who were in the high APOE risk group had a hazard ratio of 10.30 compared with the study’s reference group. The researchers said the two influences appeared largely additive and mostly independent. They were not strongly interacting with each other.

That matters because it suggests dementia risk is not traveling down a single biological road. Genetic susceptibility and blood-based signs of aging may reflect separate pathways that both shape who develops disease and when.

The team also examined hundreds of individual metabolites. Lipids, lipoproteins, and branched-chain amino acids played an important role in the aging clock and were also linked to dementia outcomes. GlycA was associated with a higher hazard of vascular dementia. Meanwhile, a composite glucose-lactate measure was associated with higher hazards of unspecified and all-cause dementia.

Other metabolites moved in the opposite direction. Several lipid-related markers and branched-chain amino acids such as leucine and valine were associated with lower hazards for some dementia outcomes. The pattern was complex. The authors noted that not all metabolites that mattered most for the aging clock were the ones most strongly tied to dementia.

The strongest association in the study involved vascular dementia, not Alzheimer’s disease. That may reflect the kinds of biological changes captured by the metabolomic platform. It is heavily focused on lipids and lipoproteins.

The researchers were also careful not to overstate what the findings prove. This was an observational study, so it cannot show that faster biological aging causes dementia. Reverse causality cannot be ruled out. However, the team ran sensitivity analyses that excluded cases diagnosed in the first two years of follow-up.

There were other limits. Dementia diagnoses came from routine care records, death registries, primary care data, and self-reported physician diagnosis, which can introduce misclassification. Some participants carried more than one dementia label over time. Most people also had metabolomics measured at only one point. Therefore, it was impossible to track how biological aging changed over the years.

The work was also done within UK Biobank, meaning any cohort-specific bias could affect both the development of the MileAge clock and the dementia analyses. The authors said replication in independent cohorts will be important.

Still, the overall pattern held up across several sensitivity checks, including analyses excluding self-reported diagnoses, restricting the sample to older participants, and removing people with multiple diagnostic labels.

Professor Marian Knight, Scientific Director for NIHR Infrastructure, said earlier detection could have broad effects beyond the people diagnosed. “Finding ways to detect and treat it earlier will help not only those affected but the families, friends and medical staff who support them.”

This study suggests that a blood-based measure of biological aging could become a useful tool for spotting dementia risk in midlife, before memory problems or other clinical symptoms appear.

That does not mean a screening test is ready now. The findings still need confirmation in other groups, and the researchers said future work should compare this metabolomic clock with other aging measures, such as epigenetic or proteomic clocks. Repeated blood sampling, brain imaging, and cognitive testing could also help show whether the measure is tracking very early brain changes.

Even so, the research offers a practical direction. Because plasma-based metabolomic tests are scalable and minimally invasive, they could eventually help identify people for prevention studies or for trials of disease-modifying therapies. And since biological aging may capture modifiable risk alongside inherited risk, the work raises the possibility that dementia prevention could become more targeted, more personalized, and earlier than it is today.