First let me apologize if this story has all ready been covered today, and for my awkward question formatting. I've been a lurker for a few months here, but this is my first post.
This recently published longitudinal study by Frank Hu of the Harvard School of Public Health was caught by NPR and is making it's daily rounds.
To give NPR credit, they cover the article accurately and include the contrasting viewpoints and alternative explanations offered by other researchers known in the field. There were a few points that I took away from both the article and the news coverage that I thought might be interesting talking points.
Correlation of Intake of Dietary Heme and Myocardial Infarction (MI) I had never heard of this! The studies they list to support this statement (references 17-20 in the scholarly article) seem to suggest that the interaction of intake on heme iron and rates of MI is only significant when the highest and lowest quartile of participants were compared, and that overall iron consumption had no association with frequency of MI. What does this mean? Heme iron is bound, not free, and it simply doesn't make sense to me that it should increase the risk of MI. Thoughts?
Listed Covariables The authors also report that increased consumption of red and processed meat also correlated with increased seditivity, decreased intake of fruits, vegetables and whole grains, increased likelihood for smoking, to be obese, etc. It seems to me more likely that these covariables are also confounds to their study. Correct me if I'm wrong, but it seems to me that the authors were unable to adjust their analysis to control for these confounds, or if they did, their findings were non-significant and thus not reported. I find this a bit of an egregious oversite! Again, thoughts?
Quality of Red Meat and PUFA ratio Though the authors state the found mortality rate was slightly attenuated when they controlled for saturated fatty acids (SFAs), and thus suggest that SFAs might act as a mechanism for decreased coronary health. However there is no mention given to the average quality of the meat consumed (was it factory farmed? pastured?) and the possible interaction of oxidized poly-unsaturated fatty acids (PUFAs) after cooking. I think at least addressing the quality of the meat would be a fair point to make, and am frankly surprised the researchers didn't, though given that this study began more than 20 years ago, perhaps it wasn't something that was thought of at it's initiation.
I suppose that this post is really more a wish to start a discussion than pose a direct question. Does this study raise worrying points? Aside from the potential association of Heme Iron and increased risk of MI, the greatest worry I take away is that doctors will be reading this and informing their opinions on what I believe to be poorly conducted and badly substantiated research.
Reading this table of lifestyle data from the study is mind blowing: http://archinte.ama-assn.org/cgi/content-nw/full/archinternmed.2011.2287v1/IOI110027T1
Essentially, those with highest rates of red meat have the lifestyle of someone who goes to bars for lunch and dinner to have a smoke and beer/martini with their hamburger w/fries and no veggies, which substitutes for their social life because they aren't getting out for exercise.
But the researchers rush out to pin the blame on red meat.
Quotes from the study: "The association between red meat and CVD mortality was moderately attenuated after further adjustment for saturated fat and cholesterol, suggesting a mediating role for these nutrients." Which basically says saturated fat and cholesterol was mildly protective to whatever else was causing the CVD. [EDIT: my interpretation is disputed - it could have meant that by attempting to adjust out the influence of sat fat and cholesterol, the result appears as mortality moderately going down] This is a mind blowing statement, because people have always accused those two items in red meat of causing CVD but here they essentially gloss over their own point that it is mildly protective. Well, if thats so, then WHAT in red meat could be causing CVD???
Or...again could it be the unhealthy lifestyle, again quoting from the study:"Men and women with higher intake of red meat were less likely to be physically active and were more likely to be current smokers, to drink alcohol, and to have a higher body mass index (Table 1). In addition, a higher red meat intake was associated with a higher intake of total energy but lower intakes of whole grains, fruits, and vegetables." Which plays into the profile I stated above.
EDIT: Also, the table above shows the people consuming the most red and processed meat have the lowest % of high cholesterol. Yet another confounder because we all know our paleo lifestyle doesn't reduce cholesterol necessarily, but that we will see reduced cholesterol in half of those that do have heart disease. Again a clear lifestyle marker that goes against the conventional wisdom here in the data.
EDIT: List of bloggers analyses that, way better than I explained above, shred the study.
Well just really quickly my take:
The relative risk was small, in epidemiology this is hardly anything to talk about. But controlling for saturated fat is an interesting bit. Thanks for that.
Cooking tempature and technique matters. They didn't control for it. Just an example of health effects (well not effects because it's epidemiology...associations) of overcooking, in Australia there was no association between any red meat and colon cancer risk, however baking it was associated with a substantial reduction in risk compared with other techniques http://www.ncbi.nlm.nih.gov/pubmed/21364608 Pretty important if we're talking about meat. I think that the toxins formed from high heat cooking can potentially oxidize cholesterol in the bloodstream, and can definitely cause systemic inflammation.
In a subset of the population heme iron can accumulate and cause all sorts of oxidative stress. That is a subset. In general it is damaging but it isn't really when there is sufficient vitamin e from food sources and other antioxidants. Context, right? The general population is deficient in many nutrients, especially vitamin e, probably more for the people eating the most red meat. Those who tend to store too much iron should either cut down on their red meat intake, increase their antioxidants, or give blood. Probably all of those. The rest of us...eh attend to your iron and antioxidant levels as you see fit. Lower isn't necessarily better. Tannins in coffee or tea and calcium also inhibit the absorption. Actually there's lots one can do. Chlorophyll in the same meal can protect the GI tract, but general antioxidant intake is probably sufficient. Meat + vegetables + calcium is my policy.
There are factors that can't be controlled for. Red meat is very calming and satisfying. People who eat more of it might be more stressed out and seeking it as a remedy.
Etc etc. Given the cooking thing from the evidence I have seen where cooking technique greatly attenuates or even reverses an association with cancers, I can't see this being good evidence for anything. Not knocking the authors, though, they have provided a good study that should be taken for what it actually shows.
Here are the key takeaways, direct from the published article:
ASSESSMENT OF MEAT CONSUMPTION
In 1980, a 61-item FFQ was administered to the NHS participants to collect information about their usual intake of foods and beverages in the previous year. In 1984, 1986, 1990, 1994, 1998, 2002, and 2006, similar but expanded FFQs with 131 to 166 items were sent to these participants to update their diet. Using the expanded FFQ used in the NHS, dietary data were collected in 1986, 1990, 1994, 1998, 2002, and 2006 from the HPFS participants
So they based the study off of surveys given every four years.
Now, take a look at the charts included with the study that show cases of mortality against red meat consumption (grouped by quintiles of consumption):
http://archinte.ama-assn.org/cgi/content-nw/full/archinternmed.2011.2287v1/IOI110027T2 http://archinte.ama-assn.org/cgi/content-nw/full/archinternmed.2011.2287v1/IOI110027T3 http://archinte.ama-assn.org/cgi/content-nw/full/archinternmed.2011.2287v1/IOI110027T4
The trend for cases of mortality is actually flat (or slightly negative!) for quintiles 1 through 4. It's only Q5 which establishes the trend line for being positive! Now why do the hazard ratios keep going up in each quintile?
We estimated the associations of substituting 1 serving of an alternative food for red meat with mortality by including both as continuous variables in the same multivariate model, which also contained nondietary covariates and total energy intake. The difference in their β coefficients and in their own variances and covariance were used to estimate the hazard ratios (HRs) and 95% CIs for the substitution associations.
They determined the hazard ratios from the fitted line, which as we saw above, only has a trend because of the last quintile! It is pretty misleading the way the data is presented!
Basically what they figured out is that a lot of people have a crap diet and they don't know what to blame it on, so pick one...
EDIT TO ADD: I've just written a comprehensive response here: Always Be Skeptical Of Nutrition Headlines: Or, What "Red Meat Consumption and Mortality" (Pan et.al.) Really Tells Us
(It incorporates, and credits, a couple of the responses from this page.)
Everyone calm down and take a deep breath. Now, note that the data used is from the Nurses' Health Study. Chris Masterjohn covered this a while ago, last time this data was mined:
"Will Eating Meat Make Us Die Younger?"
"Focusing on the second questionnaire, we found that butter, whole milk, eggs, processed meat, and cold breakfast cereal were underestimated by 10 to 30% on the questionnaire. In contrast, a number of fruits and vegetables, yoghurt and fish were overestimated by at least 50%. These findings for specific foods suggest that participants over-reported consumption of foods often considered desirable or healthy, such as fruit and vegetables, and underestimated foods considered less desirable."
In other words, the already-proven inaccuracy of the data set is far greater than the negative effect claimed for red meat.
(Also, Roger C's statistical analysis is good.)
I agree with everyone here. When a person is eating grain fed hamburger with a lot of artificial fillers in it, then chowing it down with coke, then going out to take a smoke.. there are way too many factors to pin down as the root cause.
Correlation not causation
I heard about this watching the news this morning (I'm in a hotel, which means it's the only time I bother with watching the news).
On CNN, the two anchor mavens were harping about how they LOVE eating red meat and bacon... how it was some kind of forbidden fruit and they try to resist the temptation to eat it - then they segue to another newswoman who apparently eats bacon every morning, and they were making cracks about how thin she was, but how she's gonna die.
Her response was "yeah, Sanjay (Dr. Sanjay Gupta, I'm assuming) keeps harassing me about my morning bacon intake but I just ignore him." I wanted to high-five her through the TV.
The answer is the same every single time this issue comes up. These studies are worthless because correlation does not prove causation. There are way too many variables involved in human health to simply compare meat consumption to CVD events and expect a scientific outcome.
For example, when you go to a fast food restaurant the majority of the calories you get are carbs. The soda is carbs, the fries are carbs, and the bun is carbs. The only meat is the small little patty they give you, which, depending on the chain, can be as small as 1/8th a pound. Unless you're eating a steak, chances are your meat is accompanied by a large amount of carbs, which contribute to heart disease.
There is also the fact that people who are trying to be healthy and do things such as stop smoking, minimize alcohol, avoid candies and sodas, exercise more, get proper sleep, etc, in addition to eating less meat because that's what they're told they need to do to be healthy. These kinds of studies do not even make an attempt to try and isolate for these extremely influential variables, which makes their conclusions pretty worthless towards practical application.
Has anyone ever seen a study that shows a high meat/low carb diet to result in higher CVD events than a low meat/high carb diet? No? Exactly.
Primo Health Coach beat me to it above, but Chris Kresser tweeted that link and Robb Wolf indicatedhe's going to do a post about it.
I commented last night that you could read the study to say red meat causes smoking, drinking and avoidance of exercise.
I do happen to believe that intake of vegetables and fruit is very healthy, but meat is the foundation of my diet and that's not going to change.
It seemed every "news" site trumpeted this stupid study yesterday and I am disheartened by how many people--including young adults whose health could be affected for decades--will read those articles and take them for gospel.
My take from reading the NPR article is that variety is king. They say to reduce intake of processed meats. Duh! they also say that eating a variety of protein sources is healthier. Well that goes back to the "eat fish twice a week" rule that most of us learned as kids.
Several people have pointed out cooking techniques as a contributing factor. One other problem is they did not separate whole foods from conventional. The nutrient profiles and fat profiles of pastured animals, raised without hormone injections or antibiotics to make them bigger, are different from those of animals raised in feed lots. If red meat was so bad, then how did the Lakota live longer, healthier lives than white men 200 years ago when their diet was primarily red meat?