The true effect of caloric restriction

A response on Nature 2012 and Science 2009

Understanding and postponing the aging process is a challenge of current medical research. Researchers are capable to increase the lifespan in yeast, worm, flies and rodents. The challenge of the coming decade is to translate these results to humans and increase human lifespan. One possible way to increase the lifespan in animals is by caloric restriction (CR). CR is a reduction in food intake while preserving optimal levels of nutrients with supplements.

The life prolonging effect of CR was first described in 1935 by McCay et al., who saw an increase in life expectancy in male rats. Fifty years later Weindruch stated in Science that: "lifelong dietary restriction beginning at 3 to 6 weeks of age in rodents is known to decelerate the rate of aging." From this point onwards research into the mechanisms underlying CR and its life prolonging effect in different animals expanded. Two independent studies were initiated with long lived primates, the National Institute on Aging (NIA) and the Wisconsin National Primate Research Center (WNPRC), to test the hypothesis whether CR could increase human lifespan.

In 2009 the WNPRC study was the first to report on the effect of CR on primates. They reported an increased survival for age-related deaths for the rhesus monkeys on CR. Last week the NIA study published their results, and in contrast to the WNPRC study, reported no increase in life expectancy for the monkeys on CR. This reopened the debate on CR and its effect on life expectancy.

In response to the data published in Nature by the NIA study, Steven N. Austad investigated the difference between the two studies in order to explain the contradicting results. And he questions whether CR might ever extent human lifespan, or "is it just the elimination of excess fat?"
The biggest difference between the WNPRC and the NIA study is the feeding protocol of the controls. In the WNPRC study the control monkeys can eat whenever they want for as much as they desire, i.e. true ad libitum feeding. In the NIA study the control monkeys received regulated proportioned food, which means that even the controls were slightly restricted. As I shall argue, this slight restriction might explain the different results of the two studies.

Therefore, we will argue that the true effect of CR on lifespan is likely an artifact. Since the effect of CR is measured as a percentage of the life span reported in the control group, a poor health status due to the increased food intake leading to decreased survival also causes the same effect on lifespan seen in the animals on CR. Thus the effect of CR is possibly an artifact observed because of the poor health status caused by excess food consumption in the control group (Cherkin, 1979; Cutler, 1982; Hayflick, 1994; Austad, 2001).

In humans, obesity is associated with increased prevalence of and mortality from age related diseases (Manson et al., 1995; Kopelman et al., 2000; Kopelman, 2007). Not surpringly, obesity and CR (WNPRC study) affect the same age-related diseases, including diabetes, coronary heart disease and cancer (Willet et al., 1995; Calle et al., 2003; Kopelman, 2007; Ray & Cleary, 2011). It has been suggested that insulin resistance is the link between obesity and these adverse clinical syndromes (Reaven, 2009). Insulin sensitivity and glucoregulation were preserved in the monkeys on CR in the WNPRC study, and are also known to improve in humans upon weight loss. Interestingly however, not all obese individuals are insulin resistant, and the health benefits of weight reduction may also differ among individuals. When the causes of death are compared between the animals on CR and the ad libitum fed controls, a similarity with finding in humans appears. Cancer incidence is lower in animals on CR compared to the control group (Birt et al., 2004; Harper et al., 2006). However, an opposite interpretation of the observed difference is also possible. Instead of a lower incidence in the restricted animals, there is a higher incidence in the control group. Since the poor health status of the ad libitum fed controls can be the cause of the high cancer incidence. This is supported by findings in humans, were obesity is associated with increased death rate for all cancers (Calle et al., 2003). Recently, body weight measurements have been suggested for risk evaluation and prevention for breast cancer (Ray & Cleary, 2011). Thus providing support for the relation between obesity and increased cancer incidence, and consequently the poor health state of the ad libitum fed controls.

The relation between high body weight and increased mortality due to diabetes, coronary heart disease and cancer is also reported in primates. As rhesus macaques on CR did not show a significant difference with the controls on overall survival (Bodkin et al., 2003; Mattison et al., 2003; Colman et al., 2009). However, when Colman et al., selectively divided overall mortality in age-related causes and all other causes, they reported an increased survival for the restricted animals (Colman et al., 2009). They regarded diabetes, cardiovascular disease and cancer as age-related causes of mortality. Due to this division, CR appeared the have improved survival in restricted rhesus macaques because more controls died due to age-related causes than restricted animals. However, the increased mortality due to diabetes, cardiovascular disease and cancer amongst the ad libitum fed controls is in fact, an indication of their poor health status. The controls had higher body weights, body fat percentages and reduced muscle mass, the ad libitum fed controls had a poorer health status.

To paraphrase, the WNPRC study which had ad libitum fed controls found an increased survival for age-related deaths for the rhesus monkeys on CR and the NIA study found no difference in life expectancy between the primates on CR and their controls, which were on slight restriction and hence had a improved health status compared to the ad libitum fed controls of the WNPRC study. This explains the different results of the two studies and makes us rethink about the effect of CR.

To conclude and to answer the question raised by Steven N. Austad. "Is CR anything more than the elimination of excess fat?" No, overfeeding compresses health and longevity is promoted not by CR but by overall healthy lifestyle with an energy intake in balance with energy expenditure, similar to the situation experienced in the wild, which leads to reduced overfeeding related diseases. Therefore, his ‘laboratory glutton’ hypothesis still holds. Thus CR does not retard the aging process nor does it promote longevity. It merely illustrates the potential for animal and human lifespan by diminishing the negative effects of overfeeding on health and survival. Similar effects may be achieved by other lifestyle interventions, such as those aimed at increasing physical activity, reducing stress and improving sleep. Therefore, questioning the relevance of CR studies for our understanding of the aging process.

Alexander Pothof, student assessor
Diana van Heemst, scientific staff
Leyden Academy on Vitality and Ageing


Remarks and/or questions on this text can be mailed to the website editor of Leyden Academy.