J Clin Endocrinol Metab. Mika Ala-Korpela et al.

Context: Ageing varies between individuals with profound consequences for chronic diseases and longevity. One hypothesis to explain the diversity is a genetically regulated molecular clock that runs differently between individuals. Large and long enough human studies to test the hypothesis are rare due to practical challenges, but statistical models of ageing are built as proxies for the molecular clock by comparing young and old individuals cross-sectionally. These models remain untested against longitudinal data.

Objective: We applied novel methodology to test if cross-sectional modelling can distinguish slow versus accelerated ageing in a human population.

Design: We trained a machine learning model to predict age from 153 clinical and cardiometabolic traits. The model was tested against longitudinal data from another cohort.

Patients or other participants: The training data came from cross-sectional surveys of the Finnish population (n = 9,708; ages 25-74 years). The validation data included three time points across 10 years in the Young Finns Study (YFS; n = 1,009; ages 24-49 years).

Intervention(s): Predicted metabolic age in 2007 was compared against observed ageing rate from the 2001 visit to the 2011 visit in the YFS dataset.

Main outcome measure(s): Correlation between predicted versus observed metabolic ageing.

Results: The cross-sectional proxy failed to predict longitudinal observations (R2 = 0.018%, P = 0.67).

Conclusions: The finding is unexpected under the clock hypothesis that would produce a positive correlation between predicted and observed ageing. Our results are better explained by a stratified model where ageing rates per se are similar in adulthood but differences in starting points explain diverging metabolic fates.