Recently, people started to wonder if protein restriction may drive the effect of CR in rodents. Speakman et al. provide a beautifully succint review and meta-analysis of the CR vs PR debate. A must read for everyone in the field, since Speakman is generally known for his diligent work.
Some background: Calorie restriction (CR) was discovered in the 1920s but Clive McCay's rigoros studies from the 30s are credited with starting the field of biogerontologic CR. Nevertheless the question whether CR or PR is more important was unresolved at that time, but after a few decades the consensus emerged that CR is not due to protein restriction. Many relevant papers were published including those by Ross et al. in the 1960s and Masoro et al. in the 1980s. However, when life extending essential amino acid restriction was discovered in the late 80s/early 90s, and secondly Linda Partridge (and many others) showed that insect lifespans are more responsive to PR than CR, this reignited the debate about mammalian CR vs PR.
Recent work on mice has involved varying the protein to carbohydrate ratio in ad libitum fed mice and then plotting their responses including lifespan in the context of a two-dimensional framework defined by the balance of intake of protein to non-protein components ( Solon-Biet et al., 2014). This massive experiment involved feeding mice 25 different diets that varied in their protein to non-protein ratios, and following cohorts of them until they died. The data clearly [and controversially] indicated that the key factor influencing longevity, as was observed in insects, was the protein to non-protein (carbohydrate) ratio in the diet. As the protein to carbohydrate ratio declined, lifespan increased, recapitulating the same finding detailed above (Ross, 1961).To rephrase, Solon-Biet claim that protein restriction was the exclusive determinant of lifespan and not CR. This claim will be challenged by Speakman et al.
Methods. A shout-out to the free science crowd:
Occasionally (n = 4 papers) we could not extract data on the diet compositions, either because the full text of the papers were hidden behind pay-walls, or the specialised diet manufacturer appeared to be no longer in business, and the paper only cited the source of the diet rather than its composition.Speakman excluded studies with a null result in their analysis because they seek to explain changes in lifespan across macronutrient levels. There is nothing to explain if the experiment didn't work, even if these experiments have some relevance to CR per se. So they excluded the important studies on wild-derived mice and in the ILSXISS strains.
The idea was to derive a dose-response curve for CR w/o PR and contrast it with a curve for CR + PR.
The effect of ‘CR with PR’ [n~67]on rodent lifespan does not differ from the effect of ‘CR without PR’ [n~8]... the overlap of the fitted regressions was striking, and it is clear that this is not a situation where there is a trend for an effect that fails to reach significance because of the high variation in the responses within treatments.
Speakman and colleagues describe the problem with Solon-Biet et al. (2014) which is that in their hands 30% CR shortended lifespan in the B6 strain, in contrast with the vast majority of all studies ever performed (dozens!) Intriguingly, Solon-Biet et al. achieved CR via cellulose addition which could have reduced hunger signalling which is essential to the life extension effects. Thus it seems that Speakman et al. speculate along the same lines as I did when I first discussed Solon-Biet after a first glance review. Furthermore, "the mice ingested fewer calories, they did so while ingesting almost twice as much mass of food". Hence I would speculate that excess cellulose could be harmful, but I'm not sure about the natural cellulose content of mouse diets, much less if natural is optimal (4).
A second quick glance at the literature makes me wonder again if there is something wrong with the Solon-Biet paper. Kokkonen et al (1988) for instance claim that dietary dilution via cellulose extends lifespan (I have no full access to ref. 2 using two different university accounts, who doesn't love paywalls?). A review of all studies using cellulose to induce CR might be useful esp. when combined with macronutrient data.
"Effect of percentage protein composition of the diet on lifespan" in ad libitum animals
If 18-26E% protein is taken as the baseline an extension of LS is seen especially with lower intakes of protein. The lowest studied intake was 4E% of protein.
The authors believe that:
...the impact of CR plus PR on lifespan is a completely different phenomenon from PR alone. Hence, the effect of reducing food intake on lifespan in rodents acts only via the restriction of calorie intake. This conclusion is consistent with the fact that the morphological, physiological and behavioural responses of mice to CR with PR (between 0 and 40% restriction) are completely different to the responses of mice to PR alone over the equivalent range (20% down to 12% protein) ...
it is also clear that there is an independent impact of dietary protein reduction on lifespan, but it operates over a different range of restriction (50 to 85%: relative to a reference intake of 18–26% protein in the diet) than that over which CR is effective (10–65% relative to ad libitum intake), and has a much smaller impact. Hence, reducing protein levels by 80% (from 20% to 4%) increases median lifespan by about 15%, while reducing calories by half this amount (40%) increases median lifespan by on average twice as much (30%).The finding is a little surprising if I am reading it correctly, because PR seems to be exclusively beneficial under ad libitum conditions. This non-additivity is confusing: if CR and PR both led to a graded response via e.g. mTOR or IGF-1 reduction we would expect them to be additive. However, it is interesting to note that acute CR leads to a decrease in IGF-1 proportional to the reduction in calories (5), but protein restriction does not.
And a further caution against taking invertebrate studies too seriously:
...[there may be] a major methodological difference between studies on insects and rodents. In particular, rodents are generally provided with less food when on restriction, but insects are primarily manipulated by giving them food diluted with indigestible components, and hence, paradoxically, despite being called on ‘restriction’, they eat a greater mass of food.A cautious conclusion: low/reduced protein is probably still beneficial as are other macros
First of all, there is evidence that reduced n3 fatty acid intake is relevant to CR in rodents, which was acknowledged by Speakman. Unfortunately, there was only study so far. Furthermore, the Situation in rodents is different from humans. In rodents CR leads to reduced IGF1 but in humans moderate protein restriction is absolutely essential for this reduction (cf Fontana work).
1. Speakman, J. R., Mitchell, S. E., & Mazidi, M. (2016). Calories or protein? The effect of dietary restriction on lifespan in rodents is explained by calories alone. Experimental Gerontology.
Kokkonen, G. C., & Barrows, C. H. (1988). The effect of dietary cellulose on life span and biochemical variables of male mice. Age, 11(1), 7-9.
Solon-Biet, S. M., McMahon, A. C., Ballard, J. W. O., Ruohonen, K., Wu, L. E., Cogger, V. C., ... & Gokarn, R. (2014). The ratio of macronutrients, not caloric intake, dictates cardiometabolic health, aging, and longevity in ad libitum-fed mice. Cell metabolism, 19(3), 418-430.
Anderson, R. L., Owens, J. W., & Timms, C. W. (1992). The toxicity of purified cellulose in studies with laboratory animals. Cancer letters, 63(2), 83-92.
Mitchell, S. E., Delville, C., Konstantopedos, P., Hurst, J., Derous, D., Green, C., ... & Lusseau, D. (2015). The effects of graded levels of calorie restriction: II. Impact of short term calorie and protein restriction on circulating hormone levels, glucose homeostasis and oxidative stress in male C57BL/6 mice. Oncotarget, 6(27), 23213-23237.