Dementia is a medical term for several age-related diseases that affect cognitive functions, such as memory, reasoning, and problem-solving activities. The prevalence of dementia is sharply increasing worldwide, with more than 57 million people currently living with the condition.
Chronological age is the most common risk factor for dementia. However, dementia is not an inevitable consequence of natural aging. Up to 45 % of dementia cases can be prevented, or at least delayed, by improving modifiable lifestyle risk factors, including unhealthy diet, physical inactivity, smoking, alcohol drinking, and social isolation. Identification of dementia biomarkers during its long preclinical period, i.e., the mid-life period, could be an effective strategy for early risk stratification, targeted prevention, and modification of disease trajectory.
The pathogenesis of most age-related diseases, like dementia, is associated with changes in the metabolome, the total pool of metabolites in the body, including lipids, sugars, vitamins, and amino acids. High-throughput analysis of blood metabolites can therefore serve as a promising strategy for early disease detection.
While blood metabolite profiles enable risk stratification, aging clocks provide an opportunity to identify patterns in biological data to predict a person's functional age or biological age using machine learning. The current study, conducted by researchers at King’s College London, aimed to investigate whether metabolite-predicted age can predict the risk of dementia and its age of onset.
Blood metabolite profiles analyzed alongside genetic risk
The researchers analyzed data from 223,496 UK Biobank participants who were registered with the UK National Health Service (NHS) and provided sociodemographic, behavioral, and medical information, underwent physical examinations, and gave blood and urine samples.
The researchers specifically analyzed participants’ levels of blood metabolites, presence of genetic risk factors for dementia (APOE ε4 alleles), and incidence and age of onset of dementia.
The metabolite data were used to calculate participants' metabolomic or biological age. The difference between this metabolite-predicted age and the chronological age was defined as the Metabolomic age (MileAge) delta, with positive values indicating an older biological aging profile.
APOE ε4 and MileAge jointly raised dementia risk
The researchers identified a total of 3976 cases of dementia during a median follow-up period of 13.7 years, including 1881 cases with Alzheimer's disease, 933 with vascular dementia, 512 with dementia in other diseases, and 988 with unspecified dementia.
Participants with a metabolite-predicted biological age greater than their chronological age, i.e., higher MileAge delta, exhibited significantly higher hazards of developing vascular, unspecified, and all-cause dementia, as well as earlier onset. Specifically, participants with a MileAge delta greater than one standard deviation above the mean exhibited about 20 % to 24 % higher hazards of developing all-cause dementia and 60 % higher hazards of developing vascular dementia, respectively.
The metabolome-wide association analysis identified lipids, lipoproteins, and amino acids as major determinants of the MileAge delta. These metabolites were significantly associated with dementia, with many lipids, lipoproteins, and branched-chain amino acids linked to lower dementia hazards, while markers such as glycoprotein acetyls (GlycA) and glucose-lactate were associated with higher hazards for certain dementia subtypes.
The analysis of genetic risk factors revealed that participants with a higher MileAge delta and two APOE ε4 alleles were at a 10.30-fold higher hazard of developing all-cause dementia over time.
Blood-based aging clock may aid early detection
The study found that individuals having a biological age older than chronological age and carrying known genetic risk factors for dementia are at substantially higher risk of developing the disease.
The study found the strongest association between MileAge delta and vascular dementia risk, whereas the association with Alzheimer’s disease risk was not statistically significant. Among cases of vascular dementia, glycoprotein acetylation was linked to higher disease risk, while lipids and lipoproteins were associated with lower risk. In Alzheimer’s disease, branched-chain amino acids were the only metabolites robustly associated with lower disease risk.
The researchers also identified an inverse association between MileAge delta and dementia in other diseases. However, they cautioned that this category included a diverse group of secondary dementias with different underlying biological mechanisms and clinical features.
Importantly, a higher MileAge delta was associated with earlier dementia onset, suggesting that metabolite-derived aging clocks may help predict not only dementia risk but also when the disease is likely to develop. The researchers observed a similar association with earlier Alzheimer’s disease onset, despite not finding a statistically significant link between MileAge delta and incident Alzheimer’s disease risk itself. These findings are particularly relevant given the long preclinical phase of neurodegenerative diseases such as dementia.
The study also found that genetic risk factors and MileAge delta independently contributed to dementia risk. This suggests that metabolite-predicted biological age does not explain the relationship between genetic susceptibility and dementia. Instead, genetic risk and MileAge delta appear to reflect separate but additive pathways influencing dementia development.
Overall, the study findings highlight the potential relevance of implementing MileAge delta in clinical settings to identify at-risk individuals before the appearance of clinical symptoms of dementia. Future risk stratification that incorporates both genetic and modifiable factors, such as metabolite-predicted age, may enable more targeted preventive strategies.
The researchers cautioned that the study was observational and could not establish causality. They also noted that dementia diagnoses were based on routine healthcare records, which may include some diagnostic misclassification, and that most participants had metabolomic measurements taken at only one time point. The authors emphasized that replication in independent cohorts and additional longitudinal studies are needed before MileAge-based approaches could be broadly implemented in clinical practice.
Download your PDF copy by clicking here.
Emergency settings more likely to initiate risky medications for seniors