Severe calorie restriction has been shown to effectively prolong life in almost every species that has been studied so far. Science hasn’t yet shown whether it will work for humans, but preliminary evidence suggests that it might. Human calorie restrictors have lower weight, blood pressure, and cholesterol levels. Despite those indicators, Marshall Deutsch apparently thinks science tells us that calorie restriction ought to shorten human life rather than prolong it. Readers may have come away from his article “Hungry for Life: Calorie Restriction and Science” with the impression that higher blood pressure, weight, and cholesterol levels are good for you and that calorie restriction has only been shown to prolong the life of mice because they were prevented from exercising. These are all misconceptions. Let’s set the record straight by looking at what the science really says.
Blood Pressure
Deutsch cites a 1988 study showing that lower blood pressure was correlated with decreased survival in the very old. This is true, but it only applies to the very old and it doesn’t mean what he seems to think it means. A later study[1] showed that controlling for indicators of poor health made the apparent correlation vanish. A meta-analysis in 2010[2] showed that lowering high BP in the very old had no effecton total mortality but did help keep the elderly healthy by significantly decreasing the incidence of stroke and heart failure. Another 2010 study[3] shows that high blood pressure is a strong risk factor for all-cause mortality in the elderly.
Weight
A 2010 analysis[4] of 19 studies involving 1.46 million white adults found that overweight and obesity (and possibly underweight) are associated with increased all-cause mortality and that all-cause mortality is lowest at a body mass index (BMI) of 20.0 to 24.9, not the 25-26 that Deutsch cites from an older study. Some of the excess mortality at lower BMIs can be attributed to illnesses that interfere with appetite or cause weight loss (for instance, the cachexia associated with cancer). BMIs over 25 are clearly associated with excess mortality, the risk increasing as the BMI increases.
Cholesterol
He claims that low cholesterol levels are correlated with higher mortality rates in children.[5]This may well be true, but he doesn’t have credible evidence for it, and he forgets that correlation doesn’t prove causation. His only evidence is his own graphing of 6 data points that are not adequate to demonstrate a correlation (see figure). The data were not systematically gathered or shown to be comparable, but were a sample of convenience culled partly from an encyclopedia. 4 data points are from developed countries with higher cholesterol levels and lower mortality; they could be interpreted as an essentially flat line showing no significant correlation (no test for statistical significance was applied). 2 outlying data points are from Ghana and the Philippines where cholesterol levels are lower and child mortality is higher.
Nothing can be concluded from this graph: it certainly can’t be taken as evidence that lower cholesterol causes higher mortality. It more likely indicates that underdeveloped countries have higher child mortality (due to factors like poor hygiene, poverty, higher rates of infectious disease, poorer nutrition, and substandard medical care) and that inadequate nutrition results in lower cholesterol levels. Deutsch has responded to criticism of his graph by saying that within each group of countries (developed and underdeveloped) the relationship held true, and that he had presented data from every country for which he could find data. I’ll be polite and just say this is not good science.
Deutsch praises Uffe Ravnskov’s The Cholesterol Myths. In The Skeptic’s Dictionary, Robert Carroll demolishes that book.[6] He shows how Ravnskov cherry picks data, presents facts out of context, misuses statistics, ignores other possible explanations for the data, exaggerates bombastically, and reaches unwarranted conclusions.
Deutsch claims that people with higher cholesterol live longer, especially in the older age groups, but he can only support that claim by cherry-picking from the literature: plenty of studies show just the opposite. It is true that studies have shown higher mortality with lower cholesterol in the very old, but the authors of those studies have suggested that cholesterol levels in people over 85 may not reflect lifelong levels and may have dropped due to concurrent disease or poor nutrition. One study concluded “physicians may want to regard very low levels of cholesterol as potential warning signs of occult disease or as signals of rapidly declining health.”[7]
Deutsch tells us that the 2001 Schatz study “showed an inverse relationship between blood cholesterol level and mortality.” But he doesn’t disclose that that study only looked at the elderly and the results were not applicable to people under the age of 65. Neither does he tell us that Schatz and some of his co-authors published a later study where they concluded that “elevated lipid levels should continue to be treated in healthy elderly individuals, as they are in those who are younger, although pharmacologically lowering lipids to excessively low levels in the elderly may warrant further study, as does the contribution of subclinical frailty to the relationship of lipids to CHD risk.”[8]
There is now a huge body of evidence showing that LDL cholesterol levels are strongly correlated with increased risk of cardiovascular disease and overall mortality, and that lowering LDL levels decreases the risk. The benefit of cholesterol-lowering drugs is clear for those who already have cardiovascular disease, although the degree of benefit for preventing disease in patients who do not yet have it and for treatment of the very old is less clear and is the subject of ongoing debate in the scientific literature.
Calorie Restriction
Even Deutsch’s information on calorie restriction and longevity is questionable. He mentions rodent, fruit fly, nematode, and yeast studies; but for some reason he fails to mention that calorie restriction has been shown to prolong life in animals more comparable to humans, such as dogs and non-human primates (admittedly, the primate studies are on-going and the trend to fewer deaths has not yet reached statistical significance). He cites probably irrelevant information from a 1968 book showing that rats that exercised less ate more and got fatter. Based on this, he speculates that the mice in the calorie restriction study lived longer only because they were prevented from exercising. There is no evidence to support that hypothesis, and it doesn’t even make sense. Were the mice in the study really prevented from exercising? I don’t know; but even if they were, what about the studies of other animals? Were dogs, primates and hamsters also kept from exercise? What about fruit flies – were they prevented from flying? I don’t think so. Did they prevent yeast from exercising? Do yeast exercise? We have good evidence that exercise is correlated with longevity, so it seems likely that preventing exercise would tend to shorten life span, not prolong it. Human calorie restriction devotees are typically also fanatical exercisers.
Calorie restriction has been shown to lower BP, BMI, and cholesterol levels in humans; but what does that really mean? The majority of scientists accept that these are valid markers for cardiovascular risk, but that doesn’t necessarily mean that a severely calorie restricted diet will actually prevent heart attacks or prolong life or will not cause unacceptable adverse effects. The data aren’t in yet. So far, calorie restriction looks more promising than any other life-extending intervention; but even if it works, how many people would be willing to live with constant hunger in order to extend their lives by a few years? Enjoyment of food is one of the things that make life worthwhile. There’s an old joke where the doctor tells a patient that if he avoids tobacco, alcohol, and sex, and restricts his diet, he’ll live to be 100. The patient retorts “You call that living?”
A Word of Caution
I hope no Skeptical Briefs readers were persuaded to forgo weight control or treatment for high blood pressure or high cholesterol by Deutsch’s idiosyncratic interpretation of the scientific evidence. We don’t yet know if calorie restriction works; but we do know that modifying modifiable risk factors saves lives.
[1]http://www.bmj.com/content/316/7147/1780.full
[2]http://www.ncbi.nlm.nih.gov/pubmed/20574244
[3]http://www.ncbi.nlm.nih.gov/pubmed/21178779
[4]http://www.ncbi.nlm.nih.gov/pubmed/21121834
[5]http://www.thincs.org/unpublic.Deutsch-Ravnskov.htm
[6]http://www.skepdic.com/refuge/bunk28.html
[7]http://www.ncbi.nlm.nih.gov/pubmed/12834520
[8]http://www.ncbi.nlm.nih.gov/pubmed/15571530
This article was originally published in Skeptical Briefs, a publication of the Committee for Skeptical Inquiry.
