New AI Neural Networks Predict Human Biological Age Using Routine Blood Markers and Gut Microbiota Species

New research in Aging-US uses neural networks and routine blood tests to calculate biological age, offering a new way to track the success of anti-aging treatments.

By: AXL Media

Published: Mar 24, 2026, 4:52 AM EDT

Source: Information for this report was sourced from Aging-US

A Dual Approach to Mapping the Aging Process

A research team led by Anastasia A. Kobelyatskaya has introduced a sophisticated method for determining human biological age by integrating hematological data with microbiome analysis. Published on March 12, 2026, the study presents two neural network models designed to provide a more accurate reflection of a person's physical decline than chronological age alone. By examining a cohort of 637 individuals, the scientists created a gender-specific biochemical model and a separate microbiota model, both of which achieved a mean absolute error of approximately six years. This dual-model approach allows clinicians to view aging through two different biological lenses, capturing both systemic metabolic shifts and the evolving state of the internal microbial environment.

Refining the Biochemical Clock for Clinical Ease

To ensure the biochemical model remained practical for real-world clinical use, the researchers limited the predictor set to just seven routine markers. These include well-known indicators such as cystatin-C, IGF-1, and DHEAS, alongside sex-specific data points. By focusing on a small number of accessible markers, the team aims to ease the transition from laboratory research to standard medical check-ups. The high correlation coefficient, with R² values exceeding 0.8, suggests that these seven markers provide a remarkably reliable snapshot of an individual's biological trajectory, potentially serving as an early warning system for age-related diseases.

The Microbiota Signature of Biological Time

The second model focuses on the gut microbiome, utilizing full-length 16S sequencing to identify 45 specific bacterial species that correlate with aging. One such species, Blautia obeum, was highlighted for its significant role in shifting the predicted age of a subject. This "microbiota clock" tracks how the abundance of certain bacteria fluctuates over decades, reflecting the long-term impact of diet, lifestyle, and environmental exposures. While this model requires more advanced sequencing resources than the blood-based version, it offers a deep dive into the "second genome" of the human body, revealing how the microbial ecosystem ages alongside its host.