This week the media has been awash with news about the perilous dangers of eating red meat, and processed meat in particular. The cause of this consternation was a study published in the Archives of Internal Medicine which analysed data from two large ‘epidemiological’ studies [1]. These studies basically followed people over many years monitoring (infrequently) their diets and lifestyle habits and seeing if there was any association between these and risk of things like heart disease, cancer and overall risk of death.
An additional serving (82 g/3 oz) of unprocessed red meat was associated with a 13 per cent increased risk of death. Although it should be pointed out that many potentially processed forms of red meat were allowed into the ‘unprocessed’ category, including shop-bought hamburgers and meats used in sandwiches (this includes burgers, again).
An additional serving of processed red meat (e.g. two rashers of bacon or one hot dog) each day was associated with a 20 per cent increased risk of death.
Now, one major problem with this sort of evidence is what is termed ‘confounding’. Just because two things (e.g. red meat and death) are associated, does not mean one is causing the other. Sometimes, such an association can be down to what are called ‘confounding factors’: other associated factors that are the real cause of the association. For example, people who eat more red and processed meat might also smoke more. Or perhaps they are more likely to be obese and sedentary.
As it happens, in the populations used in this recent study, those who ate more red meat were less likely to be physically active, more likely to smoke and drink alcohol, and were generally heavier too. In other words, those eating more red meat were generally less healthy and, seemingly, less health-conscious.
Of course, this is exactly what you’d expect, because for decades now we’ve been warned of the dangers of red meat. So, individuals who have made a concerted effort to limit red meat in the diet are going to tend to be more health-conscious and healthier as a result.
In this week’s study, the authors have attempted to ‘control’ (take account for) for confounding factors. However, it’s well known that this is an imprecise science, and there’s always the potential for ‘residual confounding’ (confounding factors that were not addressed or not adequately controlled for). Interestingly, the authors of the study do not even mention the possibility of residual confounding, which seems like quite an omission to me.
This omission does smack a bit of bias, I think. Might the authors have been committed (even unconsciously) to finding a link between red meat and worse health outcomes, and were perhaps disinclined to let anything get in the way of that finding and the message that we should be eating less meat? Is there any other evidence for bias in this review? I believe there is.
In the ‘comment’ section of the review the authors refer to “Several studies [that] have suggested that vegetarians have greater longevity compared to nonvegetarians”, though they admit this might not be ascribed to the absence of red meat alone.
But is it true that vegetarians live longer than nonvegetarians? The studies which the authors refer to both date from 1999. One was a review of evidence in Seventh-day Adventists [2], while the other is an analysis of 5 separate studies [3]. There is considerable doubt about how appropriate it is to general the findings of Seventh-day Adventists to the general population, on the basis that their lifestyle habits tend to be quite different from the general population (e.g. low rates of smoking, low intakes of meat and alcohol).
This evidence also highlights the fact that any health issues associated with meat eating may not come from the meat per se, but a relative absence of other healthier foods. For example, in this study [2], vegetarians generally ate more vegetables, fruits and nuts. In fact, this study found evidence that the apparent health benefits of vegetarianism were tied not so much to the absence of meat, but to an increased consumption of perhaps healthier foods.
The other review [3] cited by the authors of this week’s study was interesting in that of the five studies analysed, two were performed in Seventh-day Adventists. Both found vegetarians to be at lower risk of death compared to non-vegetarians (see caveat above). However, the other three studies did not. And, actually, when all the results were put together, vegetarianism did not confer any benefit in terms of overall risk of death. In other words, the overall result of this review runs counter to the contention that vegetarianism is linked with improved longevity.
Another more recent study which compared mortality in vegetarians and nonvegetarians found, again, no difference in mortality between these groups [4].
In referring to evidence regarding vegetarianism, the authors of this week’s review appear to have been extremely selective and actually cited evidence which does not support their view. This smacks of bias.
This review is accompanied by an invited commentary from Dr Dean Ornish – a proponent of low-fat eating and a noted vegetarian [5]. He concludes, in a word, that the answer to the question of whether red meat is bad for us is ‘yes’. But he can’t possibly do that on the basis of this type of evidence. Nowhere in his commentary does he refer to fact that associations do not prove causality – it’s simply not mentioned.
Dr Ornish does refer, however, to human studies in which relatively high-protein, low-carbohydrate diets have led to improvements in health markers such as weight, blood pressure and blood fat levels. In fact these human studies (and there are many) have been found to improve a range of health markers across the board. But Dr Ornish dismisses these studies by referring to one single study which was, wait for it, performed in mice bred to have a genetic glitch which makes them particularly susceptible to heart disease. How this applies to humans is anyone’s guess, and one thing we know for sure it’s not relevance to humans is tiny compared to studies done in actual, err, humans.
Dr Ornish sweeps aside ton of good evidence is swept aside in favour of a small amount of quite useless science instead. More bias? You decide.
References:
1. Pan Am ,et al. Red Meat Consumption and Mortality: Results From 2 Prospective Cohort Studies. Arch Int Med epub 12th March 2012
2. Fraser GE. Associations between diet and cancer, ischemic heart disease, and all-cause mortality in non-Hispanic white California Seventh-day Adventists. Am J Clin Nutr. 1999;70(3)(suppl):532S-538S.
3. KeyTJ, et al. Mortality in vegetarians and nonvegetarians: detailed findings from a collaborative analysis of 5 prospective studies. Am J Clin Nutr. 1999;70(3)(suppl):516S-524S.
4. Key TJ, et al. Mortality in British vegetarians: results from the European Prospective
Investigation into Cancer and Nutrition (EPIC-Oxford). Am J Clin Nutr. 2009;89(5):1613S-1619S.
5. Ornish D, et al. What’s good for you is good for out planet. Arch Int Med epub 12th March 2012
