Life-extending interventions do not necessarily result in compression of morbidity: a case example offering a robust statistical approach.

Despite extensive research on life-extending interventions, rigorous statistical techniques to determine their impact on compression of morbidity (CoM) are rarely used. We present a case example of an analytical method for examining the effect of life-extending dietary interventions on CoM by comparing the rates of decline in vitality and survival toward the end of life. Using data from previous experimental studies in mice, we calculated the average rate of vitality decline by fitting exponential decay models to individual vitality trajectories and compared this rate with the rate of survival decline estimated from Cox proportional hazards model. The results showed that, surprisingly, the life-extending interventions failed to compress morbidity. Instead, the models suggested a potential expansion of morbidity with the interventions, particularly for chronic caloric restriction(definition), as evidenced by a greater difference between the rates of vitality decline and survival decline in the intervention group than in the control group. The results further showed age-dependent effects, with interventions likely to demonstrate CoM at very old ages. These findings challenge the assumption that lifespan extension necessarily compresses morbidity, highlighting the need to consider lifespan, healthspan(definition), and CoM as endpoints when evaluating anti-aging interventions. We do not claim that life-extending interventions categorically fail to achieve CoM; rather, we demonstrate how a rigorous statistical approach can be applied to test the CoM hypothesis. Our framework offers a valuable tool for future studies, and further refining this method will be crucial to determine under which circumstances lifespan extension leads to CoM. The potential to choose from multiple analyses of CoM calls for leadership in the geroscience community both to standardize the nomenclature so that different CoM analysis approaches can be communicated clearly and to establish a "default" CoM analysis for everyone to use in addition to their analyses of choice, thus facilitating comparisons and meta-analyses across studies.