Back

Estimation of biological age acceleration based on NMR metabolomics and other risk factors in the Estonian Biobank

Delesa-Velina, M.; Kronberg, J.; Estonian Biobank research team, ; Fischer, K.

2025-08-19 epidemiology
10.1101/2025.08.17.25333849 medRxiv
Show abstract

A high-throughput platform of nuclear magnetic resonance (NMR) spectroscopy blood metabolite data from 30-90 years old participants of Estonian Biobank (N = 150, 023) is used to develop and validate a biomarker score for mortality and the corresponding biological age estimate. We define biological age as the age where an individuals survival probability given their covariate profile matches the survival probability of an average individual in the population. We estimate the survival probabilities parametrically based on Gompertz model using the newly developed metabolite biomarker score and common risk factors. The resulting biological age estimate, SurvMetaboAge, is a powerful predictor of both overall and cause-specific mortality. One year of biological age acceleration (BAA, difference between individuals biological and chronological age) is associated with 17% (95% CI 15%-18%) and 12% (95% CI 11%-13%) increase of hazards for overall mortality in the validation set for men and women, respectively. An appealing interpretation of BAA can be provided, demonstrating that it is an easily interpretable predictor of mortality encompassing metabolite profile and common risk factors in a single measure with a potential in risk stratification and risk communication.

Matching journals

The top 9 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.