Two studies published this week in Cell Metabolism, Solon-Biet et al. 2014 and Levine et al. 2014, are telling us that having a diet rich in animal proteins could considerably reduce our lifespan.

Here is what Science magazine comments about the studies.

Quote Originally Posted by Science
A new theory about the foods that can extend life is taking shape, and it’s sure to be a controversial one. Two studies out this week, one in mice and another primarily in people, suggest that eating relatively little protein and lots of carbohydrates—the opposite of what’s urged by many human diet plans, including the popular Atkins Diet—extends life and fortifies health.

The research challenges other common wisdom, too. The authors of both studies believe that calorie restriction, a drastic diet that helps mice and other species live much longer than normal, may work not because it slashes calorie intake, but mostly because it cuts down on protein. They also speculate that the low-protein/high-carbohydrate balance that appears to extend life in the two studies, published in Cell Metabolism, could clarify why slightly plump people live longer on average than skinny ones—something epidemiologists have been hard-pressed to explain.

“If these two studies are really correct, what people in general are trying to do” to get and stay thin “might be completely wrong in terms of maintaining health and even longevity,” says Shin-ichiro Imai, a molecular biologist at Washington University in St. Louis who studies aging.
The mice whose diets included 5% to 15% protein and 40% to 60% carbohydrates lived the longest, up to 150 weeks compared with 100 weeks for those on a diet of about 50% protein. By comparison, Americans on average take in about 16% of their calories from protein. The animals on the low-protein/high-carb plan also had lower blood pressure, better glucose tolerance, and healthier cholesterol. (Levels of fat in their diet didn’t seem to make much difference.)

One problem with the Solon-Biet et al. study is that they tested the diets exclusively on mice. As we all know mice and other rodents are largely herbivorous animals with a particular predisposition to consume high-carb aliments like cereals. Humans, on the other hand, have been carnivorous since Homo Erectus started hunting 2 million years ago, and high carbs diet have not become common until the Neolithic, a few thousands years ago.

Dietary recommendations cannot work across species, and especially not between two species with such different ancestral diets and mice and humans. It's pretty obvious that mice shouldn't eat a lot of proteins, being herbivorous, and that they should be in better health if they got their proteins from plant sources. It's highly unnatural for feed mice diets consisting of 50% of proteins. Mind you, it's even ridiculous for humans to have a diet made of 50% of proteins. Americans, who are among the world's biggest meat eaters, only get 16% of their energy from proteins in average. And proteins are also found in dairy products, beans, and so on. No wonder that mice with high-protein diets die younger. There is no need to spend time and money researching that. It's common sense. It would be like trying to put a cat on a vegetarian diet ! They would die quickly since most carnivorous animals they lack the gut bacteria to digest vegetables and cereals.

As for humans, we may be omnivorous, be we are not all equal in the way we metabolise food. People in region where agriculture was adopted early have lower genetic predisposition to develop gluten intolerance (Celiac disease) or diabetes from high-carb diets than societies that abandoned the hunter-gathering lifestyle recently. The fast evolution of genes for metabolism after the adoption of cereal agriculture is eloquently explained in The 10,000 Year Explosion. Therefore it is doubtful that there exists a universally ideal diet for all human beings. It ultimately depends on what genes one inherited.

The bottom line is: the Solon-Biet is pointless for nutritional recommendation on humans as it only studies mice. Additionally, any similar study even conducted on humans would meaningless because it fails to take into account interpersonal genetic variables in humans.

The second study, led by gerontology researcher Morgan Levine, focused on 6,381 people over 50 years old, who were interviewed once about their diet as part of NHANES, a national survey of health and nutrition. They concluded that people aged 50 to 65 on high-protein diets (over 20% of calories from protein) were at higher risks of developing cancer and diabetes or dying during 18 years of follow-up, compared to those who consumed less proteins. But the interesting part is that the trend was reversed for the risk of cancer and death after 65 years old ! In other words, eating more proteins increased lifespan and lowered chances of having cancer in people over 65.

The first question that springs to mind is: how about other age groups under 50 years old ? Well, unfortunately younger people weren't part of the study. So all we know from this limited data is that apparently eating more proteins is bad between the ages of 50 and 65, but good later. That doesn't make much sense either since all kinds of people age very differently, depending on their lifestyle, social class, physical activity, stress level, environment, and of course their genes. A sample of only 6,381 people is also much too small to draw conclusions about 7 billion of human beings. It's statistically irrelevant. Any sample size under 100,000 for dietary studies is too small to be meaningful, and can just be discarded.

Now let's go back to the initial claim that the low-protein/high-carbohydrate diet appears to extend life and that slightly plump people live longer on average than skinny ones. This sounds ridiculous from the onset if we look at the countries with the highest life expectancy on the planet. It is undeniable that in countries with the longest lifespans, like Japan, Hong Kong, Australia, Canada, Sweden, Iceland, Italy, France and Switzerland people tend to have high-protein diets, be it by eating more meat (Australia, Canada, Europe), fish and seafood (Japan, Iceland, Australia, Italy, France), beans (Japan) and dairy products (Scandinavia, France, Italy, Switzerland), and therefore have higher than average intake of proteins as a percentage of the total calories consumed. Western countries where vegetarianism is more common, like Germany (9% of vegetarians), the Netherlands (4.5%), Britain (3-5%) or Brazil (5%), fare worse in terms of life expectancy than countries like France, Sweden or Australia, where 1-2% of the population is vegetarian. Italy is the only exception, with 10% of vegetarians (the highest percentage in Europe). But Italians are heavy consumers of cheese, eating twice more of it than British people, so that even vegetarians in Italy tend to have a protein-rich diet. Note that vegetarianism is almost unheard of in Japan or Hong Kong. Japan has been dubbed as one of the "Top 5 Countries for Carnivores", which I can attest, having lived there for 5 years.

It would be interesting to study life expectancy between religious groups in India and see if, once adjusted for socio-economic background, vegetarian Hindus and vegan Jains live longer than meat-eating Muslims and Christians or not. We could try to find out by comparing life expectancy by state with the percentage of religious affiliations by state and adjusting for the states' GDP per capita. The state of Kerala has by far the highest life expectancy in India, but also happens to have one of the highest percentages of Muslims (24%, against a national average of 13%) as well as the highest percentage of Christians (19%) nationwide (source). The second highest life expectancy in India is in Punjab, a state that is 60% Sikh (who are not vegetarian), and only 32% Hindu or Jain. In contrast, the state of Gujarat has over 92% of Hindus and Jains and is generally considered the most vegetarian state in India. It is also one of the richest, with a similar GDP per capita as Kerala. Yet the life expectancy in Gujarat is 7.4 years lower than in Kerala and 5.1 years lower than in Punjab. The Indian state with the lowest life expectancy is Madhya Pradesh, where 93% of the population is Hindu or Jainist. Based on that data it doesn't look like consuming less animal proteins increases lifespan at all. More the other way round.

Now if we look at the average body-mass index by country, we immediately see that Japan has one of the lowest percentage of obese (3%) and overweight people (23%) in the world - lower in fact that many African and other Asian countries. 69% of Japanese have a normal BMI. Only three countries (Laos, Ghana, Madagascar) have a higher percentage, and not by much. Switzerland, France, Italy and Scandinavian countries also rank among the thinnest countries within Europe, not only for lacking obese people but also slightly overweight ones. The conclusion that can be drawn from these nationwide statistics is that thinner people with a normal BMI live longer than obese and even slightly overweight people.