https://www.acpjournals.org/doi/10.7326/M19-1621
Means only "low" or "very low" quality evidence exists to support idea that meat causes heart disease, cancer, T2 diabetes. So if you want to eat red meat, v. unlikely to cause harm. That low-quality data is from epidemiological studies which show association only 2/
Nutritional epidemiology has been shown to be highly unreliable. When tested in clinical trials, it's confirmed only 0-20% of the time. Therefore 80-100% of the time it's wrong. This weak science is what's responsible for flip-flopping advice: eggs (cholesterol), total fat etc 3/
Epidemiologists, mainly from Harvard @HSPH, were predictably v. upset that their science had been contradicted by more rigorous evidence. Said study would "confuse" the public. Complex motives here-Not just professional but also financial conflicts of interest are significant. 4/
Willett, Katz et al. claim methodology in Annals paper, called GRADE, not designed for nutrition, yet Nat'l Acad. of Sciences recommended GRADE as one of only a few methods for US nutrition guidelines. bit.ly/2nnewru. GRADE also used for US nutrient guidelines (DRIs) 8/
Same Nat'l Acad report says our nutrition guidelines "lack scientific rigor." We've had so many reversals in diet advice that was originally based on weak epidemiology. Any scientist should try to welcome rigorous analysis, even if result is contrary to long-held beliefs. 9/
Last! Annals paper simply says: currently, there's no strong evidence for meat -> ill-health. Maybe more trials will be done and maybe those will show something different. Science evolves! (disclosures: i get no $ from any industry for my work) fin. amazed if u got this far.