Hyperfunction: Aging May Not Be Wear and Tear but a Program Left Running

In 1935, a Cornell nutritionist named Clive McCay put one group of rats on a diet. The animals received adequate vitamins and minerals but sharply fewer calories, and they grew slowly, staying small long after their freely fed littermates had filled out. Then they outlived them. McCay had designed the experiment to test whether retarding growth would lengthen life, and, as a historical review of calorie restriction research recounts, it was the first to show that cutting calories without malnutrition extends both average and maximum lifespan in rats.

The finding sat inside a question aging science has never settled: is decline something that happens to the body, or something the body keeps doing? In a research perspective published in the journal Aging, João Pedro de Magalhães of the University of Birmingham traces that question from the nineteenth century to the present and argues that the field’s dominant answer may have the causality partly backwards.

The prevailing view holds that aging is accumulated damage. DNA mutates, mitochondria leak oxidants, proteins misfold, and the wreckage piles up until systems fail, a picture formalized in the widely cited “hallmarks of aging” framework. The alternative, revived in 2006 by the Russian American scientist Mikhail Blagosklonny, is stranger and simpler. The same genetic programs that build an organism, growth, development, reproduction, never fully switch off. Blagosklonny called the result hyperfunction: not a body breaking down, but a body overshooting, cells and organs continuing to execute instructions whose useful window has closed.

The Damage Theory Began Failing Its Own Tests

De Magalhães came to this view through negative results. In 2003, a study of mice bred with half the normal level of a key antioxidant enzyme found exactly what damage theory predicted in terms of damage: more oxidative lesions in their DNA, more cancer. Yet the mice lived just as long as normal animals. Other findings piled up. Metabolic rate does not predict mammalian lifespan. Rats kept in the cold ate 44% more food, burning far more energy, without dying any younger.

The evolutionary logic had been laid out decades earlier. In 1957 George Williams proposed that genes beneficial early in life but harmful later would still be favored by natural selection, an idea called antagonistic pleiotropy, and six of his nine specific predictions have since found at least partial empirical support. Comparative biology adds a telling pattern: across mammals, how long a species takes to develop strongly predicts how long it lives afterward. A mouse ages faster than a human in any environment. Whatever sets that pace is written into the genome, not scratched into the tissues.

A Theory That Arrived by Email

The perspective doubles as a memoir of an idea. Blagosklonny first contacted de Magalhães in June 2006 to compliment senescence.info, the aging-research website the younger scientist ran, and sent him a manuscript arguing that the TOR pathway, a cellular growth switch, drives a program of aging. The two corresponded for years and met once, at a conference in Sochi in 2014, but never collaborated, something de Magalhães now calls unfortunate. Blagosklonny, who has since died, had framed his theory in a single dense sentence: “a quasi-program for aging, a continuation of the developmental program that is not turned off, is constantly on, becoming hyper-functional and damaging, causing diseases of aging.”

A Drug Turned the Philosophy Into a Prediction

What moved hyperfunction from philosophy toward mechanism was a compound from Easter Island soil bacteria. Rapamycin inhibits TOR, the master regulator that tells cells nutrients are plentiful and growth should proceed. Blagosklonny predicted early that it should slow aging, and in 2009 a three-site trial run by the U.S. National Institute on Aging found that rapamycin extended both median and maximum lifespan in genetically diverse mice, by 14% in females and 9% in males measured at the age of 90% mortality, even when treatment began at 600 days of age, roughly the mouse equivalent of a 60-year-old person. It remains the most effective lifespan-extending drug yet found in mammals.

Caloric restriction and mutations that dampen growth hormone and IGF-1 signaling point the same direction: dial down growth programs, and aging slows. No damage-repair intervention has matched that breadth, which, de Magalhães argues, is hard to explain if aging were mainly wreckage.

The theory does concede one large territory to damage. Cancer is driven by mutations, full stop, and the perspective treats it as the opposite face of aging rather than part of the program. The suspicion is that aging overall is a hybrid, programs and damage intertwined, with the balance varying by tissue.

The Theory Now Has a Way to Lose

That is still a suspicion, and the author is unusually explicit about it. Programmatic theories remain abstract and largely untested; only a handful of human conditions, presbyopia, the shrinking of the thymus, plausibly fit the pattern. But the perspective names experiments that could settle the matter. If mice engineered to accumulate far fewer somatic mutations lived two to three times longer, hyperfunction would be in serious trouble. If no such effect appears, damage theory would be.

The sharper test may already be underway. Partial reprogramming, pulsing cells with the Yamanaka factors that reset gene regulation, has been shown to erase multiple cellular marks of aging and extend life in prematurely aging mice, without repairing damage in any conventional sense. If resetting regulatory programs rejuvenates tissue after tissue, the claim that aging is written in those programs stops being rhetoric. The danger is equally concrete: whole-body reprogramming can nudge cells toward cancer, so the goal is to find factors that return cells to young states without returning them to embryonic ones.

For now the theory stands where its author left it: coherent, evolutionarily grounded, and short on decisive proof. The experiment that would kill it is clear. Nobody has run it yet.

FAQ