Much further work is, of course, still needed to fully understand the aging process and how we age, but research in the ensuing decade has resulted in the field emerging strong, and advancing to the point that applying the available knowledge for the betterment of humankind is now a realistic goal; clinical testing of the geroscience hypothesis has become a major focus of efforts in the field. This advancement has also been made possible thanks to big strides in our ability to measure biological aging: initially, by the development of epigenetic and other so-called clocks that are trained to predict chronological age and, more recently, by the emphasis on developing measurable and tractable biomarkers and clocks, not of age but of aging trajectories or the pace of aging (in some cases combined with physiological and clinical parameters).

The accumulation of macromolecular damage is generally believed to be a central driver of how we age. This is extensively represented in both the hallmarks and geroscience papers. That macromolecular damage occurs with aging in the form of DNA mutations, telomere shortening, protein aggregation, lipid peroxidation and others is undisputable, and the theoretical consequences of that damage suggest that they might be directly responsible for the cellular dysfunction that occurs with aging. Although this assumption is reasonable and has received considerable empirical support, some conceptual gaps remain and some data point to the absence of a direct connection between macromolecular damage and lifespan (at least in some instances). For example, although increased DNA damage does indeed lead to a shortening of lifespan, to date there is no conclusive evidence that lifespan or healthspan can be extended by improving this activity.

A critical question is why macromolecular damage increases with age in the first place. Living organisms are constantly exposed to a barrage of insults, both internal (for example, free radicals and cellular waste products) and external (for example, toxins and ultraviolet radiation), and as we know that repair is not 100% efficient or accurate, it is to be expected that, if not removed, damage will accumulate as a function of age. Importantly, however, damage should accumulate more or less linearly throughout life, as both external and internal insults -- although variable on a short time scale -- should remain relatively constant as we age, at least at the population level. In this 'damage accumulation' scenario, therefore, damage produced by both external and internal insults would indeed increase with age and, in most cases, the increase should be linear: we should not see an acceleration in damage accumulation late in life, unless we assume that the internal machinery to deal with this damage starts failing with advancing age. Figure 1 provides an example (among many) of how damage accumulation is not linear with age: at the population level, amyloid accumulation in the brain is less than 10% at the age of 40 (ref. ), and although not often shown, extrapolation of the data (by the author) emphasizes the negligible accumulation of damage before the age of 40. Similar observations have been made for the accumulation of DNA damage, oxidized proteins, lipids and many other macromolecules.

So we can conclude that the accumulation of macromolecular damage with age is not really the result of passive accumulation or an intrinsic inefficiency in the repair machinery throughout all of an organism's life, but rather it is probably the result of an age-dependent failure in the protective machinery that we have defined as the hallmarks of aging. In this context, we need to consider that the hallmarks of aging merely represent a set of 'reactionary' functions that are necessary to sustain life in the face of both external and internal insults, and I have argued that the decreased activity of these defense mechanisms with age results in a loss of 'molecular resilience' (defined as the ability of every cell in the organism to respond to a variety of molecular challenges). Accordingly, a loss in molecular resilience represents the composite outcome that combines the failures of specific mechanisms of aging identified as 'hallmarks' or 'pillars'. At this point it is important to emphasize that there is a close interaction, in the form of a systems biology network, between the hallmarks of aging and all other life-sustaining functions, and therefore a deterioration in the resilience of any of the nodes will result in the deterioration of the entire system (which explains why the loss of resilience occurs in multiple domains at the same time).