March 25, 2009

That Meat Study Also Found Red Meat Associated with Accidental Death in Older Men

A study published this week is garnering a lot of attention in the press because it confirms what the health establishment would like to believe is true, that red meat--especially fatty red meat--is dangerous. But as usual, a closer look at the study calls this into question.

This study is currently available for free online. You can read it or download the PDF here:

Meat Intake and Mortality: A Prospective Study of Over Half a Million People. Rashmi Sinha, et al.Arch Intern Med. 2009;169(6):562-571.

It's worth looking at the whole study, because without reading the whole thing, you won't really understand what this study did and did not find.

The researchers in this study administered a dietary questionnaire to 617,119 AARP members aged between 50 and 71 years old in 1995. The questionnaire asked them to try to remember how many portions of various kinds of foods they'd eaten over the past month. The respondants were then followed up for ten years and the number of deaths and causes of death were determined by looking people up in the Social Security Administration Death Master File. Verification of vital status, and the cause of death information was provided by follow-up searches of the National Death Index (NDI).

After adjusting their data for factors known to affect mortality, such as smoking, physical activity, social indicators, age and BMI the authors write:
CVD mortality, as well as all other deaths in both men (Table 2) and women (Table 3) in the highest compared with the lowest quintile of red meat intake in the fully adjusted model. There was an increased risk associated with death from injuries and sudden death with higher consumption of red meat in men but not in women.
You may notice that this is not quite what you read in your newspaper. Your paper did not report, "Eating red meat increases likelihood of injury in older men but not women" because your response to that conclusion would have been to say, "Sounds like red meat must be a marker for something this study neglected to measure."

Later on, in the discussion section, the authors brush off this latter finding by explaining that the number of accidental and sudden deaths was low so perhaps the finding wasn't that significant.

After that, they explain why eating red meat causes the excess heart disease and cancer deaths they they believe they found. It is important to note that nothing in this study points to any of these causes being involved. These are, basically, religious beliefs common in the nutritional community. Here's what they wrote:
There are various mechanisms by which meat may be related to mortality. In relation to cancer, meat is a source of several multisite carcinogens, including heterocyclic amines and polycyclic aromatic hydrocarbons, which are both formed during high-temperature cooking of meat, as well asN-nitroso compounds. Iron in red meat may increase oxidative damage and increase the formation of N-nitroso compounds.Furthermore, meat is a major source of saturated fat, which has been positively associated with breast and colorectal cancer.
This last part of the study has been quoted in the press as if it had been proven in the study. This is not true.

In fact, as Dr. Eades points out in his blog today two other significant, well-conducted studies released at the same time as the red meat study debunked the meat/colorectal cancer connection. One found no relationship between colorectal cancer and meat consumption, the other found no significant differences in mortality between vegetarians and meat eaters in the UK.

You can read the details of those studies in Dr. Eades' blog with links to the studies so I won't reproduce them here.

What Dr. Eades did not mention is the glaring omission in the red meat study. The participants were asked to remember how much meat of various kinds they had eaten over the past month. They were asked about their fruit and vegetable consumption. But they were not asked anything about their carbohydrate consumption.

So we don't know whether the people eating the most meat were also those eating the most french fries with their red meat, or the most hot dog buns with their processed meat, or the most supersized Cokes with their burgers.

This is not a trivial omission. We know there is a very strong relationship between A1c and heart disease and without the carbohydrate intake figures, it is impossible to know if it is the higher portions of carbohydrates consumed along with higher portions of meat that is raising those mortality figures.

We also know that high blood sugars turn off the immune system at the same time as they provide glucose to tumors, which may help explain the connection found in this study between a slight increase in cancers and eating more red meat. Since the other studies Dr. Eades cites found no correlation when meat eating is more carefully quantified, it's possible that this study really found something completely different from what it's researchers thought they were finding because they did not ask the right questions. Remember that in 1995 when this study began it was an article of religious belief that carbohydrates were the healthiest kind of food a person could eat and that fat and meat caused disease.

I don't know about you, but I'm going to keep eating red meat and I'll keep eating it without fries or sugary soda. And if steak prices go down because people who read health news uncritically avoid it, all the better!

The two studies that Dr. Eades points to confirm what I've observed in my own circle which is filled with "health food" fanatics who eat nothing but vegetarian organic foods. Over the decades I've known them, they have developed an alarming number of autoimmune diseases and cancers.

What you want to do with this data is up to you. But I hope it includes writing a letter to the editor of your local newspaper asking why they reported on the study that linked cancer and meat but didn't mention the other two equally well conducted studies that found no such links.


Pubsgal said...

Here was a red flag for me: "The questionnaire asked them to try to remember how many portions of various kinds of foods they'd eaten over the past month." Unless participants were measuring and logging their consumption, they could be wildly off on their estimates and their memory of their consumption. (Even after nearly a year of weighing and measuring, I still get "portion creep" when I do not measure regularly. And I know I'd be hard pressed at the end of the month to give an accurate summary of what I ate, although if I'm reading the study correctly, they did some compensating for that with their food questionnaire + 2 24-hour food diaries' correlation.)

And I think you've got a good point. My cholesterol panel improved considerably after I was diagnosed and started "eating to the meter," especially my triglycerides. While I decreased my overall consumption of just about everything but vegetables, red meat still comprises most of my protein.

Scott S said...

Once again, a brilliant discussion on how the media report the big headlines, but often omit critical elements in their discussion. The journal editors should also be doing a better job of policing these studies, as it sounds like the authors got a bit carried away with the PR for the story (self-promotion, I guess).

I believe this type of reporting was once called "yellow journalism", but we know that the media's goal is to sell newspapers, or more specifically advertising space (not necessarily printed papers), so the facts are less critical to them as a group. But I do question the role of their editors, however, whose job is supposed to be to ensure that the story is reasonably accurate while also being different from the 30 other newspapers who are carrying the same story from the major news wires!

Bottom line: another flim flam study which proves nothing other than its very easy to make headlines as long as the story runs on a day when nothing else is going on!

Monica said...

"But they were not asked anything about their carbohydrate consumption. "

Bingo. I made this same point yesterday when I blogged on this absurd article:

Digby said...

Most low-carbers are not eating red meat every meal, anyway. When I do eat red mean, I try to eat lean cuts most of the time; but even eating bacon I don't think I get as much saturated fat as when I'm on the normal American diet.

Studies are only as good as their controls, so say all good scientists.

Jenny said...

I eat red meat several times a week and the amount of fat I eat depends on the level of carbohydrate I'm eating. If I'm eating at a very low carb level--under 60 grams a day, I don't worry about saturated fat. My conviction from all the research I've read is that natural animal fats are quite healthy for us. The bad reputation of fat comes from very badly conducted research (as documented by Gary Taubes) quite a bit of which confused trans fat with saturated fat.

Unknown said...

My initial reaction to the "accidental death" correlation with men who eat more red meat was: maybe tough guys who are way more concerned with looking manly than good health are likely to do both these things, do something irresponsible and then die and eat way too much meat (and other crap) and then die.

Jenny said...


It might be what you suggest though another study published in Lancet today finds that the thinner you are the more likely you are to die in an accident.

These are what we call "correlations" and they are very different from causation. People who live in expensive houses have more diamonds than people who don't. But you won't increase your diamond collection simply by buying a bigger house.

. said...

Some comments on this study and my estimate of absolute risks: