. Why dietary restriction substantially increases longevity in animal models but won't in humans. Ageing Res Rev. 2005 Aug;4(3):339-50. PubMed.


Please login to recommend the paper.


  1. The authors place much emphasis on the role of reproduction in longevity. I can’t see what relevance reproduction has to human aging. There is no difference in longevity between humans who reproduce and those who don’t. Furthermore, the studies of the effects of CR on longevity of rats and mice were done on animals that did not reproduce (the rats were usually singly housed). CR may or may not substantially increase longevity in humans. However, I do not think that calculations such as those performed by Phelan and Rose have any validity or provide any meaningful insights or information regarding this question, as they are based on erroneous assumptions. For example, sumo wrestlers who die at an early age don’t die of old age, but because of the complications of diabetes and/or atherosclerosis, that is, no direct relevance to normal aging.

  2. Can Low Calorie Intake Dramatically Improve Health and Longevity in Humans?
    John Phelan of UCLA and Michael Rose of UC-Irvine have developed a mathematical model that attempts to predict the relationship between caloric intake and longevity in humans, using representative data from controlled experiments with rodents, as well as published studies on humans, diet, and longevity. They used information on calorie intake and lifespan in Okinawan men (low calories, long lifespan) and sumo wrestlers (high calories, short lifespan) to construct a mathematical model that might predict lifespan in humans who have very low caloric intake.

    Their mathematical model shows that people who consume the most calories have a shorter lifespan, and that if people severely restrict their calories over their lifetimes, their lifespan increases by between 3 percent and 7 percent. This is much less than the up to 60 percent lifespan increase, roughly equivalent to the amount of calorie reduction, seen in studies of animals on caloric restriction (CR).

    In the author’s words, "the trade-off between calories and longevity appears to be close to a linear relationship, but the slope isn't very steep." Humans will not have rodent-like results (up to 60 percent increase in lifespan) from dramatically restricting calories," Phelan said. "Caloric restriction (CR) is not a panacea. While caloric restriction is likely to be almost universal in its beneficial effects on longevity, the benefit to humans is going to be small, even if humans restrict their caloric intake substantially and over long periods of time."

    What are we to take home from this study? Several conclusions can be made upon careful consideration of the data. First, hats off to the authors since the study addressed one of the most fundamental unanswered questions in gerontology. That is, will CR in humans add equivalent increases in lifespan to the animal studies, which show increases of 10 to 60 percent? Notably, the lifespan benefits are also accompanied by dramatic reductions in age-related pathology, including markedly lower cardiovascular disease, cancer, and cognitive impairment.

    The CR paradigm entails “undernutrition without malnutrition,” which translates into eating 10 to 60 percent fewer calories than would normally be required to maintain one’s usual body weight. It also entails avoiding any nutritional deficiencies by careful choice of foods and use of nutritional supplementation where necessary.

    The author’s choice of Okinawan men and sumo wrestlers as representative of CR and overnutrition, respectively, and use of their lifespans to plot potential gains from CR is creative but suffers from several flaws that limit the conclusions that can be made. The following problems can be identified that limit the study’s validity and generalizability:

    • Our research shows that Okinawan men did not undergo a true CR paradigm since they did not voluntarily attempt to eat fewer calories than required to maintain usual body weight, nor did they actively monitor their diets in an attempt to avoid nutritional deficiencies from low calorie intake.

    • Calories in the Okinawan population were approximately 5-15 percent fewer than what would be required (for their body weight and activity levels), and then only for about half their lives, not equivalent to the large calorie reductions seen over the entire life course in animal studies. Even under this limited “CR paradigm,” Okinawans outlive Americans by about 10 percent, similar to that seen in animal studies (i.e., 10 percent calorie reduction and 10 percent increase in lifespan).
    • Their study model is more realistic for looking at the effects of obesity than CR since comparisons are made between a very obese group and a thin, but only mildly CR population.
    • Okinawans suffered large population losses from World War II and infectious disease due to lack of adequate health care, limiting potential gains in life expectancy due to CR.
    • The sample size of sumo wrestlers is not large enough to make any meaningful conclusions about their lifespan, and sumo wrestlers eat an excessively high caloric intake only during their very short careers, not over the entire lifespan.
    • The authors did not control for the multitude of other factors that confound studies of caloric intake and human lifespan (e.g., food choices, exercise, health care).
    • Studies in progress with nonhuman primates (who share over 95 percent of our genes and have similar reproductive physiology) on a CR regimen, while not yet conclusive, are showing early results consistent with other animal data.
    • Short-term studies of humans under a true CR paradigm have shown dramatic changes in physiology similar to other animals, which would predict markedly lower cardiovascular risk, although it is too soon to make conclusions about lifespan.

    Perhaps the best way to view the conclusions of this paper is summarized by the authors themselves when they state ”given the complexity of the physiology underlying reproductive costs and other mechanisms affecting life history, it is difficult to construct a simple model for the relationship between the particulars of the physiology involved and patterns of mortality.”

    While the study itself is interesting, in the end, a simple mathematical model is only as good as the data and assumptions that go into it. Since the assumptions cannot be supported by the evidence, neither can the conclusions. Much more research needs to be done before we can make any firm conclusions about the effects of CR on human lifespan.

Make a Comment

To make a comment you must login or register.