A new study by researchers at Tufts University and the Cleveland Clinic and published in the American Heart Association journal Atherosclerosis, Thrombosis, and Vascular Biology claims that higher red meat intake is associated with increased risk for heart attacks and strokes (ASCVD or atherosclerotic cardiovascular disease), partly because it contains carnitine, a nutrient found only in animal source foods (ASF), particularly in red meat. More than 50 news outlets have already picked up this article. How important are these new findings? Should you worry that eating red meat could cause you to have a stroke or a heart attack?
Unfortunately, this happens to be a nutrition epidemiology study, so it is simply not capable of providing meaningful information about this question that we can use to guide our food choices.
No nutrition epidemiology study can be taken seriously because the methodology used is wholly unscientific. [To learn more, please see: The Problem with Epidemiological Studies.]
Nutrition epidemiology relies on memories, not measurements
Science requires data. Data, by definition, must be objective and quantifiable. Nutrition epidemiologists do not generate dietary data because they do not measure actual food intake. Instead, they use memory-based questionnaires that force people to estimate their intake of a select list of foods over long periods of time.
These excerpts below, taken from this new study, tell you everything you need to know about how this plant-biased sausage was made:
“Usual dietary habits over the past year were assessed in 1989 to 1990 using a validated 99-item picture-sort-food frequency questionnaire (FFQ) adapted from the National Cancer Institute, and again in 1995 to 1996 using a validated Willett semiquantitative FFQ.
For each FFQ, participants were asked to indicate how often, on average, they had eaten given amounts of various foods during the past year. [my emphasis]”
Notice:
- People were asked to remember what they had eaten over an entire year. How accurately do you think you could quantify all of the foods you have eaten since August 2021? Most people can’t remember what they ate last week, let alone what they ate 12 months ago.
- This study claims to have followed nearly 4,000 participants over more than 12.5 years, yet they only inquired about dietary intake twice: once in 1990 and once in 1996, as if intake during these two years would accurately represent intake over the entire 12-plus years of follow-up.
- One of the FFQs only contained 99 food questions (and the other probably contained no more than 131 items*), yet the typical U.S. supermarket contains thousands of food items. Don’t all of the other foods matter?
*It’s unclear which version of Professor Willett’s FFQ was used because the authors provide three different citations. Below is a description of the 131-item FFQ taken from ref. #37:
“The semiquantitative FFQ is designed to assess dietary variables that are hypothesized to affect the occurrence of heart disease and cancer. It includes 131 food items with specific serving sizes that are described using natural portions (eg, one banana, two slices of pizza) or standard weight and volume measures of servings commonly consumed in this study population. For each food item, participants indicated their average frequency of consumption over the past year in terms of the specified serving size by checking one of the nine frequency categories. The selected frequency category for each food item was converted to a daily intake. For example, a response of “two to four per week” was converted to 0.429 servings a day (three per week). [my emphasis]”
Translation: the authors “convert” ballpark guesses into concrete quantities down to the third decimal place, thereby creating the illusion of precision.
Even if the information gathered was accurate (which is not possible given the methodology), the strength of the associations (represented by the hazard ratios) between meat and heart disease risk are all well below 2.0, the threshold recommended by the Bradford-Hill criteria as one of (many) prerequisites for establishing causality (1.0 = no association). Notice also that the lower limit of the confidence intervals (noted in parentheses following each bolded value) approach 1.0:
“After multivariable-adjustment, higher intakes of unprocessed red meat, total meat, and total ASF associated with higher ASCVD risk, with hazard ratios (95% CI) per interquintile range of 1.15 (1.01–1.30), 1.22 (1.07–1.39), and 1.18 (1.03–1.34), respectively.”
Using the first “data” point as an example: the hazard ratio was 1.15 (far below 2.0) and the confidence interval ranged from 1.01 to 1.30, meaning that the actual value could fall anywhere within that range. The lower limit is 1.01…just a whisker above 1.0; these are truly trivial and therefore meaningless associations.
Furthermore, if their theory about meat and TMAO were correct (which it’s not), then fish should have been much more strongly associated with ASCVD risk, because fish naturally contains high amounts of TMAO, whereas the carnitine in meat has to first be converted to TMAO in the intestine by gut bacteria. Yet this new study found no association between fish and ASCVD:
“Processed meat intake associated with a nonsignificant trend toward higher ASCVD (1.11 [0.98–1.25]); intakes of fish, poultry, and eggs were not significantly associated. [my emphasis]”
Norwegian fishery scientists have rightfully taken issue with this illogical “fish contains lots of TMAO but don’t fear fish, fear red meat” argument in more than one scientific rebuttal, including this one:
“Microbial Trimethylamine-N-Oxide as a Disease Marker: Something Fishy?” which concludes “We suspect that the TMAO story may be a red herring.”
Fishy indeed…
Nutrition epidemiology studies will continue to pollute the stormy seas of nutrition science unless and until academic journal editors decide to categorically reject them as unscientific and unpublishable. It should be considered scientific malpractice to subject the public to these epidemi-illogical, data-free hypotheses about food and human health.
Notes:
The Cleveland Clinic first began studying the potential relationship between red meat, carnitine, TMAO, and heart disease back in 2013. If you are curious to learn more, you can read my explanation and critique of that convoluted study here: “Does Carnitine from Red Meat Cause Heart Disease?.” (I must admit I had a little fun with it).
If nutrition epidemiology interests you, my forthcoming book Change Your Diet, Change Your Mind (2023) will devote a full chapter to the topic. This is necessary because, in order to arrive at a useful definition of what a brain-healthy diet is, we must first disrobe and discard the entire field of nutrition epidemiology. Since the lion’s share of what we have been led to believe about food and human health springs from this deeply flawed methodology, the way to the truth about what foods keep us healthy must begin with a clean slate.
