The EAT-Lancet report is based on weak science.
Web URL: Read the article on Scribd here
Scientific Evidence on Red Meat and Health
The theory that red meat is bad for human health and causes obesity, diabetes, heart disease, cancer and even premature death is not substantiated by rigorous science.
The newly released EAT-Lancet report, as many reports before it, have managed to cast red meat as the nutritional boogeyman by relying on a weak kind of science: epidemiology.
A prominent example of this was the World Health Organizations 2015 designation of red meat as a carcinogen (for colorectal cancer). But this decision depended entirely upon epidemiological data which showed that the relative risk of getting this cancer for red meat eaters, compared to non-meat eaters, was only 1.17 to 1.18. Relative risks below 2 are generally considered in the field of epidemiology to be too small to establish a reliable correlation.
This is the kind of weak science upon which the EAT-Lancet report is based.
The fact is, there is no evidence to back up claims that red meat is bad for health. Randomized controlled trials on humans, considered the gold standard of scientific research, do not support the idea that red meat causes any kind of disease.
On the saturated fats in meat:
The two largest-ever NIH-funded, multi-center clinical trials (the Women’s Health Initiative and the Minnesota Coronary Survey) where saturated fats were replaced by unsaturated fats, on nearly 54,000 men and women, concluded that saturated fats had no effect on cardiovascular mortality or total mortality. A large meta-analysis of all clinical trials on saturated fats came to the same conclusion.
On red meat and cancer:
Two large NIH-funded, multi-center clinical trials on altogether more than 50,000 men and women who significantly cut back on red-meat consumption (while increasing fruits, vegetables and grains) did not see any risk reduction for polyp re-occurrence or any kind of cancer.
On red meat and heart disease:
Two meta-analyses of randomized controlled trials (in the Journal of Clinical Lipidology and the American Journal of Clinical Nutrition) both found that red meat had either neutral or positive effects on most cardiovascular outcomes (blood pressure, cholesterol and other lipids).
On red meat and type 2 diabetes:
Red meat cannot possibly cause diabetes, because glucose (sugar) is the principal driver of type 2 diabetes, and meat contains no glucose. Moreover, red meat availability has dropped dramatically as diabetes has skyrocketed, making any proposed connection between red meat and diabetes self-evidently unreasonable.
Epidemiology has given us some spectacular health failures over recent decades: hormone replacement therapy, anti-oxidant vitamins and caps on dietary cholesterol, to name a few. Read the EAT Lancet report with great caution, as it lacks any kind of scientific rigor and only serves to misguide Americans on their nutritional health.
Here are the facts about epidemiology
At best, epidemiological studies can show only association and cannot establish causation, which means that the data can be used to suggest hypotheses but not to prove them. Observational studies that link nutrition with disease generally find tiny differences in risk (relative risks of 1-2) which are not enough to generate confidence that an association is real.
Epidemiological studies rely on self-reported food surveys which can often be imprecise. Researchers from the Mayo Clinic tested "memory-based dietary assessment methods" and found that the nutritional data collected was "fundamentally and fatally flawed.”
Only a small number of nutritional related epidemiological studies are ultimately confirmed by more rigorous scientific studies. In 2005, Stanford’s John Ioannidis analyzed several dozen highly cited studies and concluded that subsequent clinical trials could only reproduce around 20% of observational findings. A 2011 paper published by Significance analyzed 52 claims made in nutritional studies, and none withstood the scrutiny of subsequent clinical trials.