Phenotypic Age: a novel signature of mortality and morbidity risk

Abstract

Background:A person’s rate of aging has important implications for his/her risk of death and disease, thus, quantifying aging using observable characteristics has important applications for clinical, basic, and observational research. We aimed to validate a novel aging measure, “Phenotypic Age”, constructed based on routine clinical chemistry measures, by assessing its applicability for differentiating risk for morbidity and mortality in both healthy and unhealthy populations of various ages.Methods:A nationally representative US sample, NHANES III, was used to derive “Phenotypic Age” based on a linear combination of chronological age and nine multi-system clinical chemistry measures, selected via cox proportional elastic net. Mortality predictions were validated using an independent sample (NHANES IV), consisting of 11,432 participants, for whom we observed a total of 871 deaths, ascertained over 12.6 year of follow-up. Proportional hazard models and ROC curves were used to evaluate predictions.Results:Phenotypic Age was significantly associated with all-cause mortality and cause-specific mortality. These results were robust to age and sex stratification, and remained even when excluding short-term mortality. Similarly, Phenotypic Age was associated with mortality among seemingly “healthy” participants—defined as those who were disease-free and had normal BMI at baseline—as well as the oldest-old (aged 85+)—a group with high disease burden.Conclusions:Phenotypic Age is a reliable predictor of all-cause and cause-specific mortality in multiple subgroups of the population. Risk stratification by this composite measure is far superior to that of the individual measures that go into it, as well as traditional measures of health. It is able to differentiate individuals who appear healthy, who may have otherwise been missed using traditional health assessments. Further, it can differentiate risk among persons with shared disease burden. Overall, this easily measured metric may be useful in the clinical setting and facilitate secondary and tertiary prevention strategies.

Quantitative Characterization of Biological Age and Frailty Based on Locomotor Activity Records

“…BAA itself is a sensible biological variable associated with mortality risk or disease status, and therefore refining the correlation to the chronological age often comes at the expense of losing biologically significant information. For example, some popular biological age models fail to fully capture signatures of all-cause mortality [5,6]. This is consistent with the conclusions of a recent study where Frailty Index better predicted mortality rates compared to DNAm age [40].…”

“…Since mortality in human populations increases exponentially with age, the log-hazard ratio prediction is roughly a linear function of age and therefore represents a sensible supervised predictor (i.e. trained using the death registry information) of biological age [5,6]. In the present work, we demonstrate that the BAA of the predictors of biological age using a log-hazard ratio correlates with the BAA of the principal component score.…”

“…These linear predictors, however, often fail to fully capture signatures of mortality and the incidence of diseases. This deficiency can be addressed with the help of log-linear mortality risk predictors, which have been used as proxies to quantify aging progress [5,6]. It remains to be seen, however, if and how any of these measures of aging are related to each other in human populations, and whether the same associations hold true and therefore can be reliably examined in animal models.…”

Biological age is a universal marker of aging, stress, and frailty

“…Proportional hazards mortality models are increasingly common tool for biological age predictions. Most recent examples include the "PhenoAge" based on blood sample data (Liu et al 2018). The "PhenoAge" prediction was further used to train a DNA-methylation-based aging marker DNAm PhenoAge .…”

“…The first choice was the Cox proportional hazards model with blood markers as covariates only, with no explicit age or sex labels. The log-scaled hazards ratio is a linear combination of blood features and can be calibrated in years as outlined in (Liu et al 2018, Pyrkov et al 2018b and was used as an alternative biological age predictor, the "MORTAL-bioage". Although the correlation of the predicted age with the chronological age was lower (r = 0.35 with RM SE = 17.5 years), its association with mortality was notably improved (HR = 1.08, p = 4.5E−138) relative to that of the models based on the chronological age only, see Table I.…”