Just because something isn’t a lie does not mean that it isn’t deceptive. A liar knows that he is a liar, but one who speaks mere portions of truth in order to deceive is a craftsman of destruction.
― Criss Jami
- Eating protein will kill you in middle age, and then somehow–magic, I think we can assume–becomes protective after age 65. Variations of this article came out all over the news about two weeks ago. It was such a headline that I had several people ask me about it. It took me about a week to get the time to find the actual study, and then more time to get around to writing about it. But here we go.
As you probably already guessed from the idiotic statement that the same protein that will kill you if you’re 64 will suddenly protect you from death once you turn 65, this is a load of baloney. This contention alone ought to be enough to put every reasonable person on alert.
But let’s look at some details from the article:
Eating too much protein could be as dangerous as smoking for middle-aged people, a scientific study has found.
Research which tracked thousands of adults for nearly 20 years found that people who eat a diet rich in animal protein are four times more likely to die of cancer than someone with a low protein diet.
The risk is nearly as high as the danger of developing cancer by smoking 20 cigarettes each day.
It goes on and on, but the important “facts” they report are these:
1. Eating “a lot” (as they decided to define “lot”) of protein from animal sources seemed to be associated with a gigantic risk of death in the 50-65 set. This association is so dangerous it’s just as bad as smoking every day.
2. If those people ate lots of protein from plant and not animal sources, they didn’t die.
3. Mysteriously, once those people turn 65, a high-protein diet has exactly the opposite effect. Perhaps the American Association of Retired Persons sends them an injection in the mail that causes this change.
Let’s delve right into taking this idiocy apart, shall we? (You can find the actual study here.)
- The most important thing to know is that the lead researcher on this project, a Prof. Longo, is the founder of Pro-Lon. Pro-Lon is a “Fasting Mimicking Enhancing Diet” of non-animal protein that will help you lose weight and also live forever. And also beat cancer. And for a modest sum, which they don’t disclose until you contact them directly, they can send you their bars and shakes and energy drinks. Priya Balasubramanian, Sebastian Brandhorst, and Luigi Fontana are three other researchers into this protein=death study who also work for Pro-Lon. It’s quite a shock, I’m sure, to discover that people who make large amounts of cash off of plant-based protein products are claiming that plant-based protein is really safe and healthy, while animal protein will kill you.
2. The second important thing to know is that the data here is very suspect.
6,381 people were followed for 18 years. They reported what they ate, telling researchers they ate an average of 1,823 calories per day. Since the average American eats 2,534 calories per day, we just encountered our first problem. Either these were America’s 6,381 most conscientious dieters–on an 18-year diet–or they weren’t accurate about what they ate, for whatever reason.
We can also expect this to confound results because we all know human nature. What does the average person on the street think is most healthy? Fruits and vegetables and whole grains. What does the average person on the street think is most unhealthy? Fat and meat, especially eggs, red meat, processed meat, and fried meat. Maybe sugar, but usually only sugar in fatty desserts. What is the most common reason for people to lie? To protect themselves in some way–physically, emotionally or from punishment, rejection or conflict. So…if you choose to lie to someone about what you ate yesterday, by about 700 calories, what are you most likely to lie about? 700 calories worth of salad you just “forgot” about? Or wouldn’t it be more likely that what you lie about is that fried chicken or that cake?
So then our heroic researchers were out looking for the equation “animal protein = death.” They divided up their group into three. Normally, you divide groups into equal numbers. You can’t compare a group of 1,146 with a group of 4,798 with a group of 437 and expect any really useful results that can be applied to everyone. Or, maybe you can, because that’s what our researcher friends did.
What they found was nothing.
That’s right. No association that was worth mentioning.
So when the numbers don’t come out the way you want, what do you do? Well, naturally you start fiddling with them! And that’s just what these folks did. Their next trick was to divide these already weirdly divided groups into two age groups: 50-65 and 65+.
Once they did that, they found the following association:
Among those ages 50–65, higher protein levels were linked to significantly increased risks of all-cause and cancer mortality. In this age range, subjects in the high protein group had a 74% increase in their relative risk of all-cause mortality (HR: 1.74; 95% CI: 1.02–2.97) and were more than four times as likely to die of cancer (HR: 4.33; 95% CI: 1.96–9.56) when compared to those in the low protein group.
Of course what they don’t bother to mention is that there was an equal and opposite result for the over–65ers. Interestingly, I didn’t see even one headline that ran:
EATING LOTS OF ANIMAL PROTEIN AFTER AGE 65 RESULTS IN 74% REDUCTION IN MORTALITY!
Nope. They only choose to mention one half of their association.
And here’s where dividing these people into such small groupings causes all kinds of issues. Not only are we dealing with groups of wildly different sizes, but the numbers we’re dealing with here are tiny.
And what about life expectancy? This is a statistical thing. At birth, an American boy will live an average of 69 years. If the boy lives to age 11, he will then, on average, live to be 70.3 years old. If he makes it alive to 20, then the average becomes 71.2. If he makes it to 40, avoiding diseases, accidents, and the stupidity of youth, now his life expectancy is 73.5. When he hits 50, he can then expect to live to be 76.1. And if he makes it all the way to 65…now you’re talking. If you managed to avoid dying all the way to 65, now your life expectancy is over 80! This kind of statistic throws a wrench in the works of studies like the ones we’re talking about.
And speaking of tiny amounts of information, the total number of deaths upon which we’re basing this screaming headline that eating protein is as dangerous as smoking is 6 deaths in 18 years.
Finally, we have the tired old error of reporting relative risk instead of absolute risk, knowing that you and I will just assume absolute risk.
Here’s what I mean: what are your odds of dying in a plane crash? If you fly once a year, every year of your life, your average over your whole lifetime is about 1 in 66,000. Let’s say you’re a daredevil who flies twice a year. Now your odds over your whole lifetime are 1 in 33,000. Your relative risk has just gone up 50%. “You are 50% more likely to die in a plane crash!” the headline screams, and you go running for your car. But that headline is only telling you the risk relative to only flying once per year. Your absolute risk of dying if you fly once every single year for your entire life is 1.52%. Your absolute risk of dying if you fly twice every single year for your entire life is still only 3.03%. 3.03% is twice as high as 1.52%, but still not high enough to stop anyone flying to Orlando for vacation, and still far less risky than driving.
And that’s the kind of shenanigans going on in this article. Zoe Harcome did a brilliant dissection of this article and demanded the raw data from the researchers to find out the true actual risk. If you want more details on the math, I recommend her analysis.
3. The third thing to know is that this isn’t a study. It’s two studies.
The first study was epidemiological. This is very different from a randomized clinical trial. It merely takes data and looks at it, trying to find a connection. This is designed to be a jumping off point for hypothesis generation and further research. It is not meant to be, and can never be, useful in making an “A caused B” statement. Yet that’s exactly what these “researchers” and the reporters talking about it have done. Some of those involved can probably claim ignorance of statistics and the nature of epidemiological vs. clinical studies. But some cannot, and it’s because of them I placed this article in the “ugly” category.
Epidemiological studies always raise more questions than they answer. When you say “animal protein,” are you differentiating between a fresh piece of roast beef or chicken breast and a breaded chicken nugget fried in toxic seed oil? Between half a pound of deli meat and spray-cheese-stuffed hot dogs, and a steak or salmon? And of course the ones who eat high protein also engage in all kinds of other bad habits, because most people who care about nutrition and their health mistakenly eat more carbohydrates and vegetation than they do meat and eggs. The people who care about their health are also the ones who exercise, who weigh less, who eat less refined sugar and carbs, who don’t smoke, and who are less likely to drink to excess. The ones who don’t care about their health may be eating lots of protein, sure. But they’re also eating lots of french fries, white bread, and snack cakes.
Once the “researchers” had found their first associations, they went about trying to verify the causation they’d assumed using rat models. Whenever any article on this study mentions tumors or tumor growth related to protein intake, bear in mind that it is talking only about the second study in rats. There is no evidence of any kind to suggest a connection in humans.
So they stuffed the rats full of casein (milk protein ripped out of its natural, fatty-filled packaging) as a stand-in for all types of animal protein, or soy as a stand-in for all vegetable protein. That’s a problem right there–unless you’re drinking all your protein in the form of reconstituted casein, this isn’t relevant to you in the slightest.
Then they implanted these poor rats with cancerous tumors. The ones who ate lots of protein had bigger tumors after a few weeks than the ones eating low protein.
Interestingly, with the rats it didn’t make any difference whether they ate animal or vegetable protein, when supposedly the data from human observations said that animal protein was evil and plant protein good. The researchers funded by Pro-Lon conveniently forgot to mention that the rat models showed no difference between animal and plant protein.
That by itself should suggest that the action of humans feeding themselves protein in the wild has no relationships to stuffing caged rats full of casein and soy.
Anyway, the rat study never measured mortality: only tumor growth. The human study never measured tumor growth: only mortality.
Let’s end this nonsense with a good quote from Zoe Harcome that sums it up:
This study has made an interesting observation and that’s it. It has not provided a plausible explanation. As for the smoking comparison – this is a stunt to grab headlines – not appropriate for researchers who want to be taken seriously. Smoking presents an absolute risk – provide the same numbers for my grass-grazing roast dinner if that’s what you’re claiming….Should you ditch meat and dairy as a result of this? Not unless you want to deprive yourself of essential fats, complete protein and invaluable quantities of vitamins and minerals.
Plato says he’s hungry
Help us keep paying for this site and feeding the dogs.