“Biological age reflects the current state of the body, considering the aspects of lifestyle, environment, and hereditary component.”
Why do some people appear to age faster than others, even when they are the same age? Researchers increasingly believe that chronological age tells only part of the story. Biological age attempts to capture how well the body’s systems are functioning and may provide a more meaningful picture of overall health.
A research paper on this topic was published in Volume 18 of Aging titled “Blood biochemical and gut microbiotic neural network models forecasting human biological age.” In the study, Russian researchers explored whether information from routine blood tests and the gut microbiome could be used to estimate biological age.
Looking Beyond the Calendar
For decades, researchers have searched for reliable ways to measure biological aging. Some of the most well-known aging clocks rely on DNA methylation patterns, but these approaches often require specialized laboratory equipment and can be difficult to implement in routine clinical practice.
The researchers aimed to develop alternatives to DNA methylation clocks using blood biomarkers and gut microbiome characteristics. They investigated whether blood chemistry measurements and gut microbiome profiles could be used to estimate biological age with high accuracy.
To do this, they analyzed data from 637 adults ranging in age from 18 to 99 years, combining laboratory blood measurements with microbiome sequencing data obtained from stool samples.
Building an Aging Clock From Blood Markers
The first model focused on biochemical indicators measured in blood. After evaluating dozens of laboratory parameters, the researchers identified a small set of biomarkers that showed strong associations with age.
Three markers were important for both men and women:
- Cystatin C
- Insulin-like growth factor 1 (IGF-1)
- Dehydroepiandrosterone sulfate (DHEAS)
Additional sex-specific markers were incorporated for each group. In women, the model included homocysteine, urea, glucose, and zonulin. In men, the model included HbA1c, NT-proBNP, free testosterone, and high-sensitivity C-reactive protein (hs-CRP).
Using these biomarkers as inputs, the team trained neural-network models designed to predict biological age. The resulting models predicted age with an average error of roughly six years and showed strong agreement with chronological age.
The Aging Signature Hidden in the Gut Microbiome
The second model focused on the trillions of microorganisms that inhabit the human digestive tract.
Previous studies have shown that the gut microbiome changes with age, leading researchers to investigate whether these microbial shifts could serve as indicators of biological aging. Some bacterial species become more abundant with age, while others decline. Because the microbiome influences metabolism, immune function, inflammation, and gut barrier integrity, researchers have increasingly viewed it as a potential window into the aging process.
After analyzing microbial sequencing data, the investigators selected 45 bacterial species that were associated with age and used them to train a microbiome-based aging model.
Despite relying on a very different set of biological measurements, the microbiome-based model also showed strong predictive performance. Its estimates closely tracked chronological age and showed substantial agreement with both the blood-based model and an established aging measure known as PhenoAge.
Making Artificial Intelligence Explainable
Because neural networks are often difficult to interpret, the researchers also examined which variables contributed most to the predictions. To do this, they used an explainable AI approach called SHAP (SHapley Additive exPlanations). This method allowed them to determine how much each blood biomarker or bacterial species contributed to an individual’s biological age estimate.
DHEAS, a hormone known to decline with age, emerged as one of the most influential predictors of biological age in both sexes, with its contribution varying substantially across age groups. In older individuals, markers such as cystatin C and NT-proBNP became particularly important indicators of aging-related physiological changes.
The microbiome model showed a more complex pattern. Rather than relying on a single dominant bacterial species, the model incorporated information from dozens of microbes whose collective behavior reflected age-related shifts in gut health and metabolism.
What Changes in the Body Are Being Captured?
According to the authors, the blood-based model appears to capture aging-related changes across multiple biological systems, including metabolism, hormone regulation, inflammation, cardiovascular health, and kidney function. Age-related increases in glucose, HbA1c, hs-CRP, homocysteine, and NT-proBNP were associated with biological aging, while declines in IGF-1, DHEAS, and testosterone reflected reduced anabolic and endocrine function.
The microbiome model identified a different but interconnected aspect of aging. As people grow older, some beneficial bacteria involved in producing metabolites such as butyrate and acetate decline, while certain potentially harmful or inflammatory species become more abundant. These microbial shifts can influence immune responses, metabolic regulation, and intestinal barrier function.
The researchers suggest that common biological pathways may link the two models, including chronic low-grade inflammation, metabolic dysregulation, insulin resistance, and changes in gut barrier integrity. Rather than being independent processes, these mechanisms may interact to drive biological aging throughout the body.
Why These Findings Matter
A practical advantage of the study is that biological age could be estimated using a relatively small number of biomarkers. The blood-based model required only seven laboratory measurements, while the microbiome model relied on 45 bacterial species. Both approaches achieved strong predictive accuracy while remaining more interpretable than many previous aging clocks.
Although additional validation in diverse populations will be needed, these tools could eventually help researchers monitor the effects of lifestyle interventions, medical treatments, or anti-aging therapies. Because the models provide information about which factors contribute most to an individual’s biological age estimate, they may also offer insights into the specific biological processes driving accelerated aging.
Looking Ahead
The authors conclude that both blood biochemistry and gut microbiome composition contain valuable information about biological aging. Their neural-network models achieved strong predictive performance and showed substantial agreement with each other, suggesting that different aspects of human biology may converge on common aging pathways.
As biological age becomes an increasingly important concept in longevity research and preventive medicine, practical and interpretable aging clocks may help clinicians move beyond simply counting years and toward understanding how well the body is truly aging. The findings highlight how advances in laboratory medicine, microbiome research, and artificial intelligence may help researchers better understand why people age differently and how healthy aging can be measured more precisely.
Click here to read the full research paper published in Aging.
___
Aging is indexed by PubMed/Medline (abbreviated as “Aging (Albany NY)”), PubMed Central, Web of Science: Science Citation Index Expanded (abbreviated as “Aging‐US” and listed in the Cell Biology and Geriatrics & Gerontology categories), Scopus (abbreviated as “Aging” and listed in the Cell Biology and Aging categories), Biological Abstracts, BIOSIS Previews, EMBASE, META (Chan Zuckerberg Initiative) (2018-2022), and Dimensions (Digital Science).
Click here to subscribe to Aging publication updates.
For media inquiries, please contact [email protected].