Hormesis, Aging and Longevity Determination

Leonard Hayflick, Ph.D.

Professor of Anatomy, Department of Anatomy

University of California, San Francisco

School of Medicine

P.O. Box 89

The Sea Ranch, CA 95497

Tel: 707 785 3181

Fax: 707 785 3809

Email: len@gene.com

Because hormesis, like biological aging, cannot be defined adequately (Turturro at al., 2000, Boxenbaum, 2000), any effort to assess the effect of an indefinable circumstance on an indefinable phenomenon surely must approach foolhardiness. Yet, this will not be the first time in the history of science that the ridiculous has been attempted. Nevertheless, one might take refuge in the belief that hormesis and aging, like beauty, cannot be defined but we know each when we see it.

Dr. Rattan has made a commendable attempt to explain the role of hormesis in aging but his efforts would benefit from distinguishing between aging and longevity determination. Unless this distinction is understood, the effect of any intervention on aging or longevity determination cannot properly be determined.


The random systemic molecular disorder that, after reproductive success eventually exceeds repair capacity, defines aging only as a phenomenon because it cannot be defined at any other level with precision. The molecular disorder of aging has multiple etiologies including specific damage by reactive oxygen species and cross-linking, but more generally it occurs from the diminishing energy states that in youth were sufficient to maintain molecular fidelity. From conception until reproductive success, animals are busy ordering their molecules. No survival advantage accrues to a species by having individuals live much beyond the critical period of reproductive maturity. Thus, natural selection for processes that might maintain adult molecular order indefinitely diminishes and the disorder characteristic of aging begins. The disorder is not programmed but is a stochastic process.

Longevity determination is, on the other hand, less affected by random processes. Although chance events that compromise molecular fidelity certainly do occur during development, unlike aging, the capacity for repair during development exceeds the damage wrought or most individuals would not survive long enough to achieve reproductive success and the species would vanish. Longevity determination is governed by the excess physiological capacity reached at the time of sexual maturation that, through natural selection, was achieved to better guarantee survival. Thus, longevity is only indirectly determined by the genome.

Greater physiological capacity, beyond the minimum required for life, increases the chances for animals to survive long enough to achieve reproductive success just as redundant vital systems in complex machines, better insures that they will maintain function. The amount of excess physiological capacity, like the amount of redundancy engineered into space vehicles, provides the potential for continued function beyond the primary goal (Hayflick, 1996, 1998).

Longevity determination in higher animals has been a profoundly neglected area of research. One class of animals that may provide some answers to the determination of longevity are those animals that do not reach a fixed size in adulthood and age slowly or not at all. If these animals do age, the process is either negligible or it occurs below the limits of detection. Animals in this class include some tortoises, many sport and cold-water deep-sea fish, some amphibians and the American lobster. Even telomerase expression, the hallmark of immortal cells, has been found at extraordinary high levels in all the cells of negligibly aging animals like the American lobster (Homarus americanus) and the rainbow trout (Onchorhynchus mykiss)(Klapper, et al, 1998a, 1998b). Whether these animals age at all, and the reasons for this, have been almost entirely neglected. They are not immortal because, like animals that do age, there is a constant threat of disease, predation and accidents (Hayflick 1996, 2000). Probably because of the conundrum that it presents, there are no reports of a hormesis effect on animals that age negligibly or not at all.

In animals that do reach a fixed size in adulthood, molecular disorder exceeds the capacity for perfect repair after reproductive success and thus reveals the universal process of aging. The physiological conditions that determine longevity might be thought of as the conditions against which the opposing forces of aging act.


When hormesis is interpreted to affect the aging process, a decrease in the rate of molecular disorder that characterizes aging, must be shown. Yet, there is no convincing evidence that this has been done. Where hormesis has been shown to have an effect, it has not been on the aging process but on the longevity determinants of the animals studied.

An example of this important difference would be the effect on life expectation if some cause of death in old age were to be delayed or resolved. If cancer were to be eliminated as a cause of death it would result in an increase of about 2.5 years in life expectation at virtually all ages (Hayflick, 1996, Anderson, 1999). Yet, it would be absurd to conclude from this extension of average longevity that cancer causes aging. Similarly, if hormesis or any other intervention is to be found to increase life expectation, a similar conclusion also would be spurious.


Dr. Rattan has described the many putative genes (gerontogenes?) that allegedly drive aging processes. However, these genes are better understood as governing longevity determination and not aging.

Aging is not a programmed process governed directly or entirely by genes (Hayflick, 1996, 1998, 2000). The interpretation of experiments in which it is concluded that age changes are governed by genes comes mainly from those who work with invertebrates. Genes are not involved in aging because they have not been shown to cause, affect, reverse or arrest the inexorable expression of molecular disorder that is the hallmark of aging. The studies on invertebrates are more accurately interpreted to have impact on longevity determination because the experimental results usually affect the development of physiological capacity that precedes the aging process. By the same reasoning caloric restriction and the cryogenic preservation of living cells are not believed to cause age changes simply because they have been found to delay the occurrence of the aging process compared to controls.

Just as a blueprint is vital to manufacture a complex machine and contains no information to cause the aging of that machine, the genome is critical for biological development but unnecessary to cause the aging of that animal. Both machine and animal eventually fail as a result of increasingly irreparable loss of molecular fidelity, which in living systems increases vulnerability to predation, accidents or disease and in inanimate objects increases vulnerability to analogous failures. There is no need for genes to govern aging processes and the principle of parsimony in biological systems is maintained by not having them.

Another argument against the direct role of genes in programming the aging process is that animals do not age at the same rate, even when inbred, nor are the patterns of age changes identical in identical twins. When the random events characteristic of aging are compared with the orderly, virtually lock-step, changes that occur during genetically driven embryogenesis and development, the orderliness and precision stands out in stark contrast to the quantitative and qualitative disorder of age changes (Hayflick, 2000a). The variability in the manifestations of aging differs greatly from animal to animal but the variability in developmental changes differs trivially. Humans from conception to adulthood are virtually identical in respect to the stages and timing of biological development but from about thirty on, age changes make humans more heterogeneous (Hayflick, 2000a).


If hormesis is to be understood as a stress event then its' effect on longevity determination might be understood in light of what we know about the effect of caloric restriction on longevity. The stress of caloric restriction is frequently interpreted to result in increased longevity. However, it is just as reasonable to conclude that, because control animals usually feed ad libitum, over eating reduces longevity. That is, the experimental animals actually have become the controls while the controls are actually the experimental animals. Because the most likely feeding pattern for feral animals is not ad libitum feeding but rather feast or famine, calorically restricted laboratory animals come closest to resembling the life style of their feral cousins.

One might also conclude from this observation that stress is more likely to mimic feral conditions than does the unnatural pampered lives of caged controls. Thus, the puzzle of why an increase in longevity is observed in hormesis experiments would be better understood by interpreting the effect of the stress as having mimicked the feral state compared to the pampered controls. What then becomes revealed is the increased longevity that is naturally characteristic for that species' universally stressed feral life style.

Hormesis experiments like caloric restriction experiments can be interpreted not to have increased longevity but to have simply recreated the more stressful conditions of a feral life style over that of caged, pampered animals where over feeding and protection from stress has reduced longevity.


Anderson R.N. U.S. Decennial Life Tables for 1989-91, United States life tables eliminating certain causes of death. Hyattsville, Maryland, National Center for Health Statistics, Vol. 1, No 4, 1999.

Boxenbaum, H. Commentaries: Does Caloric restriction induce hormesis? BELLE Newsletter 8:12-13, 2000.

Hayflick L. How and Why We Age. New York City, Ballantine Books, 1996.

Hayflick L. The future of ageing. Nature 408: 267-269, 2000.

Turturro, A., Hass, B.S., Does Caloric restriction induce hormesis? BELLE Newsletter 8: 2-12, 2000.

Klapper W., Heidorn K., Kuhne K., Parwaresch R., Krupp G. Telomerase activity in "immortal" fish. FEBS Lett 434: 409-412, 1998a.

Klapper W., Kuhne K., Singh K. K., Heidorn K., Parwaresch R., Krupp G. Longevity of lobsters is linked to ubiquitous telomerase expression. FEBS Lett 439: 143-146, 1998b.