A real, recent, peer-reviewed study underlies this post. In a prospective longitudinal analysis of the Hisayama Study, 1344 Japanese community residents aged 65 and over without dementia at baseline were followed for a median of 11.2 years, from 2012 to 2023, with serum ergothioneine measured by liquid chromatography mass spectrometry and divided into quartiles . During follow-up 273 participants developed all-cause dementia, of whom 201 had Alzheimer's disease and 72 had non-Alzheimer's dementia, and the age and sex adjusted hazard ratios for all three outcomes decreased progressively across increasing quartiles of serum ergothioneine, with all P values for trend below 0.05 .

On the vegetable question, the paper addressed this directly in two ways. The associations remained significant after adjustment for a wide range of cardiovascular, lifestyle, and dietary factors including daily vegetable intake, and in subgroup analyses stratified by daily vegetable intake, higher serum ergothioneine was consistently associated with lower dementia risk irrespective of vegetable consumption .

The authors' own conclusion is stated in conditional terms. They concluded that higher serum ergothioneine levels were associated with a lower risk of developing all-cause dementia, Alzheimer's disease, and non-Alzheimer's dementia, and that since ergothioneine cannot be synthesized in the human body, a diet rich in ergothioneine may be beneficial in reducing dementia risk .

METHODOLOGY AND CONTEXT

Design: Prospective observational cohort. Not a trial. No intervention, no randomization, no ergothioneine administered.

Population: 1344 Japanese community residents aged 65 and over, dementia-free at baseline . Narrow demographic. Findings do not automatically transfer to younger or non-Japanese populations with different dietary patterns.

Statistical adjustment: Three models were used: model 1 adjusted for age and sex; model 2 added low education, systolic blood pressure, antihypertensive medication, diabetes, total cholesterol, BMI, ECG abnormalities, stroke history, smoking, alcohol, and exercise; model 3 added daily vegetable intake . This is a reasonably thorough adjustment set for a nutritional epidemiology paper.

The chart in the post: The axes described in the image, log of serum ergothioneine on the x-axis against hazard ratio on a doubling scale, correspond to the paper's figure of restricted cubic splines for the association between serum ergothioneine levels and risk of all-cause dementia . A spline curve shows a continuous dose-response shape with confidence bands that widen substantially at the extremes where few participants sit. The y-axis extending down to 0.0625 is a strong indication that the curve's lower tail carries very wide uncertainty. Reading a single headline percentage off the far end of a spline is not the same as the study's primary quartile result.

Funding and conflict of interest: Four of the twelve authors are affiliated with the Research Institute, Suntory Global Innovation Center Ltd. , a commercial entity with a direct interest in ergothioneine as an ingredient. This does not invalidate the work, but it is material context the post omits.

Ergothioneine food sources: Analysis of common foods found that only some foods contain ergothioneine, with the highest concentrations detected in specialty mushrooms, kidney, liver, black and red beans, and oat bran . Mushrooms contain the highest levels, especially king oyster, maitake, oyster, shiitake, and porcini, while other ergothioneine-rich foods include liver, black beans, red beans, and oat bran , and it is produced by certain types of bacteria and fungi .

Supporting evidence elsewhere: The direction of association is not unique to this cohort. A Singapore study of 470 elderly subjects attending memory clinics examined whether ergothioneine status could predict subsequent cognitive and functional decline, noting that prior associations with cognitive impairment and cerebrovascular disease had come from cross-sectional studies . All of this remains observational.