Researchers developed “tissue clocks” that estimate biological age from routine histology images. The models identify organ-specific aging patterns linked to disease, suggesting that tissue structure may provide information beyond chronological age.
For the study, published in Nature Medicine, the researchers analyzed 25,712 whole-slide images from 40 tissue types collected from 983 postmortem donors. Computer models examined features such as tissue architecture, fibrosis, atrophy, and blood vessel density to predict each donor’s age.
The difference between predicted biological age and chronological age was called the tissue “age gap.” A larger gap indicated that a tissue appeared older than expected. Across all tissue types, the models estimated age with an average error of approximately five years.
Larger age gaps were associated with shorter telomeres, more comorbidities, and subclinical pathological changes. The specific features varied by organ. In the cerebellum, greater age gaps were linked to myelin loss and ischemic changes. In the aorta, they were associated with wall thickening and structural damage related to vascular disease.
The researchers tested the models in independent brain, lung, and skin cohorts involving 295 donors. Models trained on the same tissue type performed better than those trained on other tissues, indicating that aging patterns were organ specific. Performance varied between cohorts, however, partly because of differences in staining, tissue preparation, and scanning methods.
The team then combined histology findings with gene expression data to develop blood-based models that estimated age gaps in specific organs. These models were tested using 1,205 blood samples, including 577 from healthy donors and 628 from patients with seven chronic diseases or stroke.
The predicted patterns corresponded with several affected organs. Patients with stroke had the largest age gap in the brain. Crohn’s disease was associated with increased age gaps throughout the gastrointestinal tract, while vasculitis was linked to greater aging in the kidney, liver, and heart. In patients with Alzheimer’s disease, the brain was the only organ with a higher age gap.
These findings indicate that histology images and blood gene expression may contain measurable signs of tissue-specific aging. Such information could eventually support disease assessment or monitoring, but the models are not ready for diagnostic use.
The blood analysis involved patients who already had disease, so the study did not determine whether age gaps could predict disease before symptoms or diagnosis. Other limitations included the use of postmortem samples, unequal numbers of female and male donors, and the lack of matching tissue samples in the external blood cohorts. Prospective studies using prediagnostic samples are needed to assess clinical utility.
Newsletters
Receive the latest pathologist news, personalities, education, and career development – weekly to your inbox.
