Health-record aging measure linked to disease and death
An aging measure developed in 320,891 UK Biobank participants was linked to all 20 non-cancer diseases assessed in two UK datasets.
Our read
Biomarkers- Studied:
- Cohort study of 4,728,376 UK primary care patients (Independent replication)
- What changed:
- Higher ClinicalAge was linked to higher risk of death from any cause in primary care patients (adjusted hazard ratio 1·36)
- The catch:
- Existing illness and health care use could explain links with later disease or death
Editor of record: Boris Djordjevic · AgingResearch.News
Women and men in UK Biobank—320,891 participants, with a median age of 58—and a UK primary care database—4,728,376 people, with a median age of 51—formed the cohorts for a model development and validation study of ClinicalAge.
The researchers trained an artificial intelligence model on health records collected over time, using PhenoAge, an established aging measure, as its reference. ClinicalAge was associated with all 20 non-cancer diseases assessed in both datasets. It was also associated with death from any cause, cancer and cardiovascular disease after accounting for clinical and lifestyle factors. Other links included measures of function and biological aging. Midlife stood out as a key period when ClinicalAge trajectories diverged, with early acceleration among people who developed diabetes, high blood pressure, depression or lifestyle disorders.
Why it matters
Early identification of accelerated aging could help guide preventive care. The authors propose ClinicalAge as a practical tool for identifying people who might benefit.
Caveats
Differences in existing illness and health care use could contribute to the links between ClinicalAge and later disease or death. Validation in UK datasets leaves its performance in other health systems open.
Comments
Be the first to comment.