Development of a Novel Multidimensional Measure of Aging to Predict Mortality and Morbidity in the Prospective MJ Cohort

Abstract Background Although biological aging has been proposed as a more accurate measure of aging, few biological aging measures have been developed for Asians, especially for young adults. Methods A total of 521 656 participants were enrolled in the MJ cohort (1996–2011) and were followed until death, loss-to-follow-up, or December 31, 2011, whichever came first. We selected 14 clinical biomarkers, including chronological age, using a random forest algorithm, and developed a multidimensional aging measure (MDAge). Model performance was assessed by area under the curve (AUC) and internal calibration. We evaluated the associations of MDAge and residuals from regressing MDAge on chronological age (MDAgeAccel) with mortality and morbidity, and assessed the robustness of our findings. Results MDAge achieved an excellent AUC of 0.892 in predicting all-cause mortality (95% confidence interval [CI]: 0.889–0.894). Participants with higher MDAge at baseline were at a higher risk of death (per 5 years, hazard ration [HR] = 1.671, 95% CI: 1.662–1.680), and the association remained after controlling for other variables and in different subgroups. Furthermore, participants with higher MDAgeAccel were associated with shortened life expectancy. For instance, compared to men who were biologically younger (MDAgeAccel ≤ 0) at baseline, men in the highest tertiles of MDAgeAccel had shortened life expectancy by 17.23 years. In addition, higher MDAgeAccel was associated with having chronic disease either cross-sectionally (per 1-standard deviation [SD], odds ratio [OR] = 1.564, 95% CI: 1.552–1.575) or longitudinally (per 1-SD, OR = 1.218, 95% CI: 1.199–1.238). Conclusion MDAge accurately predicted mortality and morbidity, which has great potential in the early identification of individuals at higher risk and therefore promoting early intervention.