Key result
Corrected vascular biological age matches Klemera-Doubal for detecting non-healthy individuals and eliminates age correlation bias.
Why the study?
Using chronological age as the training label in supervised algorithms fails to reflect individual differences in aging, causing the age delta correlation phenomenon.
Does incorporating relative individual risk difference (RIRD) improve the prediction of vascular biological age compared to uncorrected models?
Comparison
Corrected vascular biological age vs KDM and uncorrected vascular biological age
Design
Prediction model development and validation study
Authors
Loading...
Hypothesis-generating for unbiased vascular age metrics; prospective validation required before clinical or research adoption.
Does incorporating relative individual risk difference (RIRD) improve the prediction of vascular biological age compared to uncorrected models?
Incorporating individual aging heterogeneity (RIRD) is essential for constructing accurate biological age prediction models using supervised AI algorithms.
Tang et al. (2026) studied Vascular biological age. Corrected vascular biological age using relative individual risk difference (RIRD) vs. Klemera-Doubal method (KDM) and uncorrected vascular biological age was evaluated on Ability in detecting non-healthy individuals and elimination of age delta correlation (ADC) bias. Corrected vascular biological age using relative individual risk difference detected non-healthy individuals comparably to the Klemera-Doubal method and eliminated age delta correlation bias.