When avoiding appeal to authority and applying proper scientific standards of evidence, it turns out that there is insufficient proof to assert that the current intake of red meats should be reduced to avoid disease, even if some organizations and guidelines claim the opposite. This is also the case for cold cuts, even if there may be forms of processing that need a cautionary approach to avoid the formation of harmful chemicals (such as harsh curing, heating, or smoking).
The reason is that the available data from observational studies is based on weak (to very weak) associations, which are potentially biased and confounded (see below), so that causal affirmations cannot and should not be made. Association does not necessarily imply causation. One of the reasons to be suspicious is that problematic associations with health are usually obtained within a US-style context and tend to diminish, disappear, or sometimes even invert when assessed in other regions or at the global level. Moreover, as outlined elsewhere, there is no clear support from randomized and controlled intervention studies, which are better suited to demonstrate a causal relationship (even if they are very difficult to perform in the domain of nutrition and only look at indirect parameters). Taken together, the evidence shows that the problem is not meat as such, but rather the unhealthy context in which it is sometimes eaten.
We can hardly ever be confident about estimates of less than 2.0, and when estimates are much below 2.0, we are simply out of business. Epidemiologists have only primitive tools, which for small relative risks are too crude to enable us to distinguish between bias, confounding and causation” – Shapiro 2004
One of the major limitations of observational studies is that they are possibly (and likely) confounded due to “healthy user bias“. In the West, individuals eating the most meat also tend to follow more unhealthy lifestyles. Such studies essentially capture that people who obediently follow health recommendations (often limiting meat as an expression of virtuous eating within the upper-middle classes) come out as more healthy on average. Eating less meat thus serves as a marker and not as a cause of healthiness. Even if scientists sometimes (but not always) correct for socioeconomic status, smoking, alcohol, or obesity, it is impossible to statistically eliminate all lifestyle effects, leaving some residual confounding of the data. In such cases, causality cannot be stated confidently, as many studies are at critical risk of bias. IARC/WHO admitted as much by declaring that “other explanations for the observations (chance, bias or confounding) could not be ruled out, while consumption of red meat has not been established as a cause of cancer” (emphasis added).
Therefore, what matters primarily is the background diet and lifestyle. For instance, in a Canadian study, higher red meat intake paralleled increased cancer risk for people who consumed low fruit and vegetables. However, for people who consumed most fruits and vegetables, a neutral to protective association was found with more meat. Similarly, in another study the association between heme iron intake and colorectal adenoma risk was influenced by the dietary antioxidant capacity, emphasizing the need for a broader assessment of diets. One may seriously question the value of assessing the disease risks associated with single foods or food groups, when all evidence points to the predominant role of dietary patterns.
