36-fold higher estimate of deaths attributable to red meat intake in GBD 2019: is this reliable?

The Nutritional Recommendations (also known as NutriRECS) international consortium performed four parallel systematic reviews of randomised trials and observational studies.4–7 The consortium reported finding low to very low certainty evidence that diets lower in unprocessed red meat might result in very small reductions in risk of cardiovascular disease, stroke, myocardial infarction, type 2 diabetes, and overall lifetime cancer mortality.8 The 2018 World Cancer Research Fund's Continuous Update Project Expert Report judged the evidence for a link between red meat intake and breast cancer to be limited and that no conclusion could be reached regarding a causal or protective relationship.9 Furthermore, there appears to be a considerable disparity between the updated relative risk curves of GBD 2019 (appendix 1 of GBD 2019 [p 349]) and the dose–response curves of peer-reviewed cohort studies that examine the relationship between red meat intake and adverse outcomes. According to the GBD 2019 analysis,1 the relative risk of suffering an ischaemic stroke, an intracerebral haemorrhage, or a subarachnoid haemorrhage for people aged 55–59 years who are consuming 50 g per day of red meat versus those consuming no red meat is 1·20 [95% CI 1·11–1·26], 1·20 [1·1–1·28], and 1·20 [1·1–1·28], respectively. [...]for protective dietary components, the TMREL was set using the 85th percentile of levels of exposure included in the published cohort studies or randomised controlled trials. By contrast, the TMREL for risk factors viewed as harmful was, by default, set to zero. [...]the red meat TMREL changed from 22·5 g per day to 0 g per day.