Photo by Irwin-Scott
I’ve had about a zillion people ask me about a new study that came out in the Archives of Internal Medicine this week linking red meat consumption to increased mortality.
Naturally some people are afraid their carne asada habit may be dooming them to an early death, and who could blame them with headlines like these?
On the other hand, I suspect many of you have dismissed the study out of hand because it conflicts with your world view that animal foods only make good things happen.
But in the interest of science and being grown ups, let’s take a look at the study and see what we can learn.
First, it is worth mentioning that the study was fairly well-designed and conducted by a respectable team of scientists at Harvard. They reanalyzed data from two large prospective cohort studies: The Health Professionals Follow-Up Study (HPFS, 1986-2008) and the Nurses Health Study (NHS, 1980-2008).
Both cohorts were large groups of health care professionals, which would presumably limit differences in education and income that can often confound health studies. The participants filled out regular food frequency questionnaires that have been previously validated as decently reliable (though food frequency questionnaires are notoriously unreliable).
Importantly, all the participants were eating Western diets during what have come to be known as the least healthy decades in US history. Also important, during the course of the study both red and processed meat consumption declined in both men and women.
“The mean daily intake of unprocessed red meat dropped from 0.75 to 0.63 servings from 1986 to 2006 in men and from 1.10 to 0.55 servings from 1980 to 2006 in women.”
The authors never comment on what this reduced consumption means for their analysis, however, since they “created cumulative averages of food intake from baseline to death from the repeated food frequency questionnaires.”
According to the report, people who ate the most red meat were more likely to smoke, drink, eat far more calories and be overweight. They were also less likely to exercise and eat healthy foods like fruits, vegetables, whole grains and fish. Basically they were less healthy people with less healthy habits.
The authors claim to have controlled for such lifestyle factors by doing statistical corrections for these variables, which is the industry standard for this type of analysis. They also performed a sensitivity analysis to see if any other dietary variable (including glycemic load) may have impacted their results. They did not detect any significant differences when controlling for these factors, which I admittedly find surprising.
To their credit, the authors made an effort to distinguish between processed and unprocessed meats. Given the time during which the study took place, however, it’s unlikely that any of the participants were eating non-industrial, grass-fed and pastured meat. I think this is an important point, particularly when considering cancer mortality, since toxic compounds tend to accumulate in the fat of animals.
In their analysis the authors estimated that for every one serving of red meat per day (defined as 3 oz), total mortality risk increased by 12% (20% for processed red meat alone, 13% for unprocessed). Heart disease risk increased by 16% for total red meat (21% for processed red meat, 18% for processed), and cancer risk increased 10% for total red meat (16% for processed, 10% unprocessed).
To help put this in perspective, in the Nurses Health Study (the larger of the two) the group that ate the least meat consumed about a 1.5 oz (half a deck of cards) of meat per day and the group that ate the most consumed around 6.5 oz of meat per day (here’s the data I’m pulling from, using the 3 oz serving size for conversion).
Remember, these numbers are for daily consumption. For the highest group, that’s nearly 3 pounds per week (45.5 oz). For the lowest group, under 1 pound (10.5 oz). Realistically, the lowest group probably ate red meat 1-2 times per week, while the highest group ate it once or twice a day. How we got from here to “all red meat will kill you” isn’t exactly clear.
Interestingly, when they did an analysis to see the specific effect of saturated fat in meat it accounted for only 4% of the 16% estimated risk. This is fairly low considering that saturated fat is supposedly what makes meat so bad for us by raising cholesterol. But since the authors say that saturated fat could account for some of the increased risk, can we at least assume that those eating the most meat were more likely to have higher cholesterol? Not so fast. It turns out that in both cohorts, those in the lowest group of meat consumption were the most likely to have high cholesterol. (Thanks Denise Minger for making this astute observation).
So what about the meat is killing us exactly? In addition to saturated fat, the authors also estimated that heme iron in meat (assumed to be a risk factor for some diseases) can account for another 5% of the risk, but they do not elaborate on how this might work. It is unclear what else about red meat may be increasing mortality risk, though preservation methods are suspected for the higher risk associated with processed meats.
The authors also used some fancy statistical magic to estimate what would happen if the participants theoretically replaced one “daily” meat serving with an equal portion of either fish, poultry, nuts, legumes, low-fat dairy or whole grains and determined that mortality risk would decline 7%, 14%, 19%, 10%, 10% and 14%, respectively. It is important to remember though that *ahem* these are not real experiments but merely estimations based on the numbers and statistical models used in this study. At best an analysis like this can generate a hypothesis that could then be tested in a controlled trial.
Fortunately controlled studies replacing meats (oh, and all the other crap in the Western diet) with other nutritious, whole foods have already been done. For example, in the Lyon Diet Heart Study (1988) a group of patients who had already had a heart attack were instructed to change their diets. One group went on the low-fat American Heart Association diet, the other group adopted a Mediterranean style diet that included lots of green and root vegetables, fruits, legumes, more fish and poultry, less red meat, olive oil and no cream. After only 3 years the study was stopped by the ethics and safety committee because the Mediterranean diet group had a 70% reduced risk of death compared to those on the low-fat diet.
Studies have consistently shown that replacing some dietary meat with fish and other sources of omega-3 fatty acids reduces the risk of heart disease. However, replacing red meat with anything else (even olive oil) seems to be pointless. For this reason I’ve always been skeptical that red meat is uniquely bad when the simpler explanation would be that eating more fish is uniquely good. I don’t see how this new analysis of old studies changes anything.
Lastly, although the authors included controls for lifestyle factors I’m highly suspicious that people with so many unhealthy habits are at an increased risk of death primarily because of meat consumption. Consequently, all that I’d feel comfortable concluding from the new analysis is that in the context of a Western diet, eating something other than meat every once in awhile is probably a good idea. Outside of the Western diet? It’s much harder to say.
What are your thoughts on the study?