Rheographic biological age defined per hemodynamic indicators using the age equivalent of hemodynamic risk

The aim of this study was to assess the feasibility of estimating rheographic biological age as an age equivalent of he-modynamic risk and to internally validate its association with long-term mortality.

The study included data from 2,655 examined individuals, including 1,235 women and 1,420 men. Survival analysis was performed in a subsample of 2,342 observations with available follow-up data. During the follow-up period, 347 deaths were registered. Anthropometric, hemodynamic and rheographic parameters, as well as derived indices, were used as input features. For comparison, age approximation models were applied, including ridge regression, elastic net, an ensemble of extremely randomized trees and KDM-PC. The final RheoBA-Cox model for rheographic biological age was developed as an age equivalent of hemodynamic risk based on the Cox proportional hazards model. Internal validation was performed using five-fold cross-validation.

Adding RheoBA-Cox acceleration to the model including chronological age and sex increased the C-index from 0.773 to 0.815. Each 5-year increase in ΔRheoBA-Cox was associated with a 36 % increase in the risk of death. The hazard ratio was 1.36; 95 % confidence interval 1.29–1.44; p = 2.1•10-29. RheoBA-Cox enables biological age to be estimated as an age equivalent of hemodynamically determined risk rather than as a simple reconstruction of chronological age. Acceleration of RheoBA-Cox was associated with long-term mortality independently of chronological age and sex, which allows the model to be considered a promising tool for risk-oriented assessment of the functional state of the cardiovascular system.