Investor Karl Pfleger: lifespan is the wrong way to measure progress against aging, for individual therapies and for the field as a whole
Investor Karl Pfleger: lifespan is the wrong way to measure progress against aging, for individual therapies and for the field as a whole
On September 25, Karl Pfleger published an essay on why the standard progress metric is misleading: a final-outcome metric poorly captures progress on any multi-component problem, including aging. The following day, biologist Max Unfried pushed back: decomposing aging into subproblems risks creating a hydra with an unknown number of heads, and the list of currently recognized subproblems may be arbitrary. Pfleger explained why he nonetheless considers the list the single most important element.
Karl Pfleger, an investor and the creator of AgingBiotech.info (a public directory of anti-aging companies), argues that the final-outcome metric of a multi-component problem does not reveal where work is actually progressing, whether for a single therapy or for the entire field. Aging, in his view, is exactly such a problem. He illustrates this with a toy example: a boat has two independent failures, a leaky hull and algae on the propeller. Individually, patching the hull or cleaning the propeller does not extend the boat's useful life, but both repairs together extend it several times over. A more modest third intervention (lightening the cargo) falsely "wins" in single-factor tests. This happens because progress accumulates invisibly and does not move the overall metric until a critical mass of subproblems has been solved, while the field tests each intervention in isolation against total lifespan.
By the same logic, the complaint that no intervention has beaten the longstanding mouse lifespan record set by strict caloric restriction is beside the point: by Pfleger's count, more than 300 biotech companies have appeared, over 60 have raised rounds of 3 million dollars or more, and more than a dozen have demonstrated efficacy in humans in Phase II trials, yet none of this registers on a graph of "mouse lifespan by year." He proposes measuring progress differently: by evidence of each subproblem's causal role in age-related disease, by the efficacy of interventions targeting each one, and by biomarkers that track them. The National Institute on Aging's mouse testing program, widely considered the gold standard, is in his view a misleading indicator of progress.
Biologist Max Unfried of The Thalion Initiative praised the essay but objected to the subproblem list, the hallmarks, the recognized causes of aging:
This could simply create a hydra with even more heads, and so far we only see a few arbitrary ones, the so-called hallmarks, and even solving them is unlikely to produce a serious effect.
Pfleger agreed that undiscovered subproblems almost certainly exist but disagreed that the known list is arbitrary: the subproblems differ greatly in importance, and the ones already identified are the most important precisely because the evidence linking them to age-related disease is strongest. He called this a detailed version of the Longevity Escape Velocity idea, the scenario in which medicine adds more than one year of life for every calendar year that passes, and supported it with a number: 7 to 12 subproblems already have strong preclinical data, and no one has tried combining them in a single organism. In his reply, he turned the hydra metaphor around:
This is not a case where all heads are the same. It is more like a version where there are several large, especially dangerous heads and a succession of progressively smaller and less threatening ones, so you do not necessarily have to kill every last one to live long and well after defeating the most fearsome.
The only organization attempting to combine these interventions is the Longevity Escape Velocity Foundation, whose second experiment has stalled for lack of funding. The debate leaves the next step unresolved: whether to search for new subproblems or to try combining what is already known.