Yes, I understand correlation and causation. I need more holes to poke in this study. I'm at work, someone posted it and I want to post a quick rebuttal
asked byStephen_Aegis (22923)
Get FREE instant access to our Paleo For Beginners Guide & 15 FREE Recipes!
on August 17, 2010
at 03:46 PM
You can point out that the study likely used conventionally raised beef and explain the vast difference between grass fed animal meat/fat and grain fed. Explain that humans evolved eating ruminants. Hot dogs and hamburgers from a discount grocery store are literally a different animal than properly raised un-processed beef. Point out that in this country the average American that eats beef regularly probably eats it at a fast food restaurant half the time where they are also likely to be ingesting tons of sugar and rancid trans fats. The anti-beef campaign has been around for a long time, therefore people who try to be healthy by following mainstream nutrition advice are probably eating more lean chicken and fish and avoiding beef, but they are also unlikely to be eating fast food and junk food- therefore it would make sense that they have a lower risk of heart disease. Most people I have known in my life who were striving for a healthy diet wouldn't dream of eating beef, but they also wouldn't eat donuts/french fries/soda etc. A study like this needs to take these factors into consideration, but they don't b/c that doesn't provide nice headlines.
on August 17, 2010
at 03:29 PM
This is an epidemiological study and therefore while it may be good for coming up with hypothesies, those ideas then must be furthered studied in a proper study using control groups and showing causation. But since this is epidemiological, it can't show causation. THere are hundreds of differences between people that effect diet. For instance, people who eat less red meat might also exercise more, live in a different environment, have a different income level, have different access to medical care, have different opinions on how to live one's life, or hundreds of other differences that probably could not even be easily thought of. All these could also account for any differences in heart disease. Just because something is correlated does not mean it caused something. For instance, just because there is a correlation between car accidents and street lights does not mean that we should be cutting down on street lights. This is why epidemiological studies are weak, hard to interpret, and cannot and do not prove causation. Such studies are supposed to be followed up by more controlled studies in order to further explore causation. However, controlled and better designed followup studies, much to the frustration of some scientists who have banked their grant money on this theory, have NOT reliably shown red meat to have a causational link to heart disease.
on August 18, 2010
at 06:05 AM
I wonder what the mystery heart-killing factor is with red meat since studies have already shown NO link between heart disease and saturated fat.
Now I know this sounds like a lot of assumptions but don't you think it is possible there are conflicting factors. Perhaps women more likely to eat red meat have unhealthy habits compared to those who eat it less often exactly because red meat is promoted as unhealthy (so the health conscious women avoid it!).
on August 18, 2010
at 02:06 AM
I've uploaded the full text - http://ifile.it/pb7g129/Circulation-nurses-health-red-meat-chd.pdf
Disclosure: My mom is a nurse, enrolled in the NHS, and I have been filling out the questionnaires for 15+ years now for the NHS as I am enrolled as well.
To address cynarin's comment on how they separated meats, the food questions were a series of food items that you rated frequency of consumption from daily to never (6 options as I recall). The list of food items was fairly comprehensive, I recall usually filling out several pages of food questions.
The only real problem with this study is the framing of the results by the media & researchers "In this prospective cohort study with 26 years follow-up, we observed that a higher consumption of [processed] red meat was associated with an increased risk of CHD."
Table 2 is a good breakdown, "RRs and 95% CIs for CHD...", no big surprises.
The most significant predictor of CHD was "Ratio of red meat to poultry" with unadjusted RRs by quintile of 1.41, 2.83, 4.53, 7.32 and 13.86.
After that was "Ratio of red meat to poultry and fish" unadjusted RR's 0.77, 1.59, 2.48, 3.83, 7.14.
After those are (in no specific order) total meat, red meat, high-fat dairy and low-fat dairy.
Red meat excluding processed meat RRs are 0.28, 0.45, 0.64, 0.84, 1.17
Beans were the most protective, follow closely by nuts & fish, then eggs & poultry.
Nothing really interesting that hasn't been reported elsewhere; eating the SAD diet kills you with CHD; fish, nuts & beans, poultry & eggs are better than SAD.
on August 17, 2010
at 02:21 PM
Here's a link to the actual study abstract: http://circ.ahajournals.org/cgi/content/abstract/CIRCULATIONAHA.109.915165v1
I would be interested to read the full study to find how they separated out red meats and processed meats, since no paleo will argue that Oscar Meyer is a health food. Other than that, it is worthwhile to not that past iterations of the Nurses' Health Study have shown that diets with high glycemic loads increase heart disease risk. http://www.nejm.org/doi/full/10.1056/NEJMoa055317