EAT Lancet diet adherence

Executive summary

* A paper was published in November 2023, which claimed that people who adhered most closely to the EAT Lancet diet had a lower risk of cancer incidence and mortality from all causes than those who did not closely follow this 'planetary health diet'.

* This was a population paper and thus it suffered from the usual flaws: dietary questionnaire inaccuracies; association not causation; relative risk was headlined/absolute risk is usually tiny; and the healthy person confounder.

* In this study, these were all valid issues but irrelevant compared to the single major issue – the paper claimed to have studied the Eat Lancet diet, but it didn't.

* The EAT Lancet diet was set out in specific detail in January 2019. Food items and groups were allocated by the calorie e.g., 575 calories should come from legumes and nuts for a male with a daily intake of 2,500 calories.

* This paper awarded 11 points for claimed adherence to the EAT Lancet diet.

* The single biggest intake in the EAT Lancet diet is 811 calories for whole grains. This paper awarded one point for zero cereal intake. The EAT Lancet diet allocated 153 and 120 calories to dairy foods and sugars respectively. This paper awarded one point for zero intake of dairy and another point for zero intake of sugars.

* I divided the EAT Lancet diet into the 15 component parts for which calories were allocated. This paper used UK Biobank data despite UK Biobank not having any information about intake for five line items of the EAT Lancet diet. For 4 of the 15 EAT Lancet parts, this paper allocated points for zero intake rather than the specified EAT Lancet intake. For the remaining 6 out of 15 foods, I generously allowed that most of the EAT Lancet calories matched.

* Overall, this paper's diet accounted for barely one fifth of the EAT Lancet diet calories.

* None of the claims for adherence to the EAT Lancet diet are valid because the research did not review the EAT Lancet diet.