Longitudinal trajectories, correlations and mortality associations of nine biological ages across 20-years follow-up.
Level 3 - non-randomized controlled study
Prospective longitudinal cohort study with repeated measures over 20 years
PubMed 32041686 · doi:10.7554/eLife.51507
What was done
Analyzed 845 Swedish adults (aged 50–90 years) with 3,973 repeated measurements from a population-based cohort over 20 years of follow-up. The authors examined longitudinal trajectories, correlations, and mortality associations of nine biological age measures, including DNA methylation estimators (Horvath, GrimAge), telomere length, and a frailty index, relative to chronological age.
What was found
Longitudinal growth of functional biological age measures accelerated around age 70, with baseline trajectory levels differing by sex across ages 50–90. Inter-marker correlations were largely driven by chronological age. Individually, all biological age measures except telomere length were associated with mortality risk independently of chronological age, with GrimAge and the frailty index showing the largest effects. In joint models, Horvath methylation age, GrimAge, and the frailty index remained mutually predictive of mortality. The abstract reported no numerical effect sizes or risk ratios.
Why it matters
This study shows that different biological aging markers capture distinct, non-redundant aspects of physiological aging. Combining epigenetic clocks with clinical frailty assessments improves mortality risk prediction beyond chronological age or any single biological metric.
Limits
No specific numerical risk estimates, hazard ratios, or confidence intervals were provided in the abstract. The study population was restricted to a Swedish cohort, which may limit generalizability to other geographic or ancestral populations. As an observational study, it cannot establish causal mechanisms.
Cited by
- supports Epigenetic clocks such as GrimAge and PhenoAge correlate weakly with telomere length at approximately r = 0.1.