Here we go again. Another study based on faulty data and poor-quality study methods claims red meat is a killer. We have been down this road many times before, yet the same problems persist. The study is essentially meaningless, and tells low-carb eaters nothing about the effects of eating red meat on their health.
CNN: Changing your meat-eating habits could mean a longer life, study suggestsThe "new" study, published in the journal BMJ, is not a new study at all. It is based on data mined from the Nurse's Health Study and the Health Professionals Follow-up Study, both studies that started back in 1976 with an initial food frequency questionnaire. The recent version of the data started with a new, baseline food frequency questionnaire in 1994 and followed subjects with additional questionnaires every 4 years until 2010. (As we have pointed out in the past, there are many problems with the accuracy of food frequency questionnaires.)
Once again, the baseline data tells a big part of the story. The meat eaters had "less healthy diet consisting of a higher intake of energy and alcohol; they were also more likely to be less physically active, have a higher body mass index, and be current smokers." To the authors' credit, this study looked at the change in meat consumption, rather than the baseline consumption, and tried to correlate that to mortality. But once again, we are left wondering what other healthy or unhealthy habit changes occurred in participants who changed (or didn't change) their level of meat consumption. This study cannot answer these questions.
Next, let's get into the data. This is where the statistics of the article get dizzying. There is a huge table that lists results based on two different models depending on what the authors "controlled" for, i.e. age, aspirin use, smoking etc. The bottom line is that the vast majority of the statistical findings were not significant, or if they were, they had miniscule hazard ratios of 1.06 or 1.12. Remember, anything less than 2.0 is an incredibly weak finding that is likely confounded by variables and not likely to be causative. Most associations weaker than a hazard ratio of 2.0 are NOT found to hold true when studied in a higher-quality randomized controlled trial.
As is usually the case with these trials, an increase in consumption of processed meat had a slightly higher association (1.13) with mortality than an increase in unprocessed meats (1.08). Both still fall far short of the Bradford Hill criteria to suggest a true causative effect. In addition, the study did not report what else these individuals were eating. Remember, a high-meat, high-carbohydrate diet is much different than a moderate-meat, low-carbohydrate diet. Without controlling for the carbs and the quality of the food intake, these types of studies tell us very little about a healthy low-carb diet.
To their credit, in the summary section the authors admit the data is not unanimous, and they reference prior studies of meat consumption and acknowledge the discrepancy between US and European or Japanese consumers:
Unprocessed meat consumption was only associated with mortality in the US populations, but not in European or Asian populations. A recent Japanese study did not find any strong association between red meat consumption and cardiovascular disease death.
Thus, in the end, we are left with yet another low-quality study attempting to point out a link between meat intake and mortality. But when we examine it critically, this study adds little to the discussion of low-carb diets and their impact on our health. That is why we continue to promote healthy, low-carb lifestyles. And, we will continue to examine new studies to ensure the headlines match the data.
Thanks for reading,
Bret Scher, MD, FACC
