In a new paper published in Nature, a group of scientists led by Peter Fedichev explores human longevity through the concept of loss of resilience, by which the researchers mean the gradual age-related expansion of time that the body needs to recover from stress, such as illness. According to the findings, this progressive loss of resilience might be the factor that puts a cap on human lifespan .
Is there a maximum lifespan?
Our scientific knowledge seems to confirm that most species have maximum lifespans, which usually correlates with certain parameters such as body weight, except for a small number of extremely long-lived species. This includes us as humans: despite the immense scientific achievements that led to a drastic increase in life expectancy in just a few decades, humans are still unable to break the age barrier of 120-ish years. Medical interventions and a healthy lifestyle can fight off age-related diseases for quite some time, but even the healthiest individual eventually succumbs to a short period of extreme frailty and illness followed by death – an effect called “compressed morbidity” . Scientists are divided on the question of whether any existing or future technology can overcome this sad predicament.
Variance is the key
In this paper, the researchers introduce an index called DOSI (dynamic state organism index). At the heart of DOSI is a biological age clock based on CBC (complete blood count). CBC is a routinely performed blood test, which means CBC records are plentiful, and they provide a trove of data. This is a plus, even if a CBC-based clock may not be as precise as some other biological age clocks that use more esoteric markers, such as DNA methylation.
DOSI’s most important feature is that its dynamics over time are intended to measure resilience – the metric of how fast the body overcomes various types of stress and returns to homeostasis.
The data shows that with age, DOSI variance increases (meaning the body gets more affected by stress) while the ability to return to the age-adjusted baseline representative of a healthy organism dwindles, until such return is no longer possible. Extrapolation of this data hints at a maximum lifespan: the baseline and DOSI are expected to irreversibly diverge at the age of 120-150 years.
This theoretical prediction might explain the common experience of aging, as well as the phenomenon of compressed morbidity and extreme frailty near life’s end. We all know that with age, it gets harder to recover from such types of stress as illness, physical exertion, lack of sleep, or alcohol consumption. Near the end of life, any disease or trauma strikes harder than before, and recovery becomes less and less possible; this is how the age-related decline in resilience manifests itself. Dr. Fedichev explains it as follows:
We are a quite long-lived species, and so, our body is more or less dynamically stable for the most of our lifespan. Stress factors and shocks are absorbed, and the organism’s state recovers its homeostatic position within a few weeks in healthy individuals in mid-life. As we age and get sick, resilience is reduced. Fluctuations of the organism’s state increase in magnitude in such a way that the probability of losing the dynamic stability increases exponentially. Recovery rate increases with age and eventually runs away at around 120+ no matter how you measure it. Low resilience and increased variance can be an early sign of an upcoming period of extreme frailty.
DOSI can be used to interpret dynamic changes in various biological age clocks. According to Fedichev, the high variance we see in many biological age clocks is due to the fact that their readings are heavily affected by the current fleeting state of the organism. When stressed, the organism looks much less healthy in many aspects, which can be interpreted by biological clocks as premature aging. However, the same organism can give a very different reading upon recovery. If this is the case, biological age is better derived not from a single reading of any particular set of biological parameters but from a longitudinal analysis of the organism’s ability to recover.
The fact that a high sampling rate is crucial for calculating resilience is the reason the researchers chose CBC for their model in the first place. CBC is so ubiquitous that the data set they used, which was collected from a Russian clinic, contained numerous measurements for each individual.
Understanding resilience from exercise
According to Fedichev, since the states of various components of living systems are highly correlated with each other, especially at the point of catastrophic failure (i.e, death), any variable associated with all-cause mortality could be used to estimate DOSI, including physical activity.
This led Fedichev’s company, Gero, to attempt to detect resilience using data collected by wearable fitness devices. In another paper, the researchers describe a model that derives resilience measurements from the dynamics of physical activity or a lack of it . For instance, if you begin to exercise less, and the periods between bouts of exercise widen, Gero’s app, which is already available for download, might interpret it as a decline in your resilience potential.
Even if data from wearable devices might not be a very good indicator of biological age, frequent measurements might be able to compensate for this disadvantage. “Due to the ubiquity of wearable trackers”, explains Fedichev, “physical activity data is available for multitudes of users over long periods of time in quantities sufficient to compute resilience”.
We have asked Dr. Fedichev a few questions about his work.
How does this benefit people, what does it tell them about their biological age, and how will they be able to adjust their lifestyle in order to age slower?
Right now, we report resilience as a purely informative feature. We plan to test if resilience can be associated with lifestyle choices, supplements, or drugs in our further studies.
How is your app different from other lifestyle apps?
Resilience is not something other lifestyle apps can measure. I think most people who tried tracking weight, sleep duration, or the number of steps taken daily know how difficult it can be to distinguish an emerging trend from a temporary fluctuation.
“Am I just a bit tired, or is this aging knocking at my door?” I guess this is an important distinction.
Yes, personally, I have always felt that measuring the rate of recovery would be an extremely useful feature that today’s health apps lack.
How would you describe the major takeaway from your study?
As resilience is being lost with age, the magnitude of fluctuations in biological age becomes too big at the chronological age of around 120+. This means that any intervention aimed solely at already existing health conditions, that is, any attempt to fix the current state, would at some point quickly become insufficient.
Even the best interventions of this kind could only compress morbidity. Meanwhile, quality of life will continue to dwindle due to diminishing resilience, increasing fluctuations in the organism’s state, and loss of coherence between its organs and subsystems.
On the positive side, we may speculate that yet-to-be-discovered interventions aimed specifically at the loss of resilience, rather than at specific health conditions, might have a major effect on lifespan. Hence, this should be the future of anti-aging therapeutics.
 Pyrkov, T. V., Avchaciov, K., Tarkhov, A. E., Menshikov, L. I., Gudkov, A. V., & Fedichev, P. O. (2019). Longitudinal analysis of blood markers reveals progressive loss of resilience and predicts ultimate limit of human lifespan. bioRxiv, 618876.
 Fnnzs, J. F. (1980). Aging, natural death, and the compression of morbidity. The New England journal of medicine, 303(3), 130-250.
 Pyrkov, T. V., Sokolov, I. S., & Fedichev, P. O. (2021). Deep longitudinal phenotyping of wearable sensor data reveals independent markers of longevity, stress, and resilience. Aging (Albany NY), 13(6), 7900.