A number of companies and research groups are performing drug discovery by using effects on frailty in mice as a readout. To what degree is frailty an adequate measure of the harms done by aging? One way to answer that question is to assess mortality in a human study population against a measure of frailty, with and without factoring in chronological age. Researchers here show that frailty is a fair marker for age-related mortality, but it is not a reflection of every degenerative, harmful process taking place under the hood. Frailty and age combined provide a better correlation with mortality than frailty alone, indicating that there are aspects of age-related decline that contribute meaningfully to mortality without producing evident frailty.
As populations get older, the association between chronological age and health status becomes increasingly heterogeneous. To describe this heterogeneity in health status as we age, the concepts of biological age or frailty versus fitness spectrum have been proposed. The frailty index (FI) methodology was introduced to quantify the accumulation of people's health 'deficits' (i.e., symptoms, clinical signs, medical conditions and disabilities) at a given chronological age. This method has allowed for the establishment of potentially useful population norms and the study of influences of wider determinants of health on the variation in health status within people of a similar chronological age.
Since FI deficits increase with age, the FI has a statistically significant association with chronological age. However, on the account of population heterogeneity, the effect size of this association has been found to be small. It has been suggested that given the age-related nature of its constituent deficits, the FI should be interpreted jointly with age. Our aim was to utilize data from 8,174 wave 1 participants in The Irish Longitudinal Study on Ageing (TILDA) to conduct, separately by sex, supervised machine learning analyses of the ability of the individual items of an FI to predict 8-year mortality. To gain insights as to the importance of age in this prediction, we repeated the analyses including age as a feature.
By wave 5, 559 men and 492 women had died. In the absence of age, the FI was an acceptable predictor of mortality with area under the curve (AUCs) of 0.7. When age was included, AUCs improved to 0.8 in men and 0.9 in women. After age, deficits related to physical function and self-rated health tended to have higher importance scores. Not all FI variables seemed equally relevant to predict mortality, and age was by far the most relevant feature. Chronological age should remain an important consideration when interpreting the prognostic significance of an FI.