One size never fit all

The idea that nutrition is a settled science — that every body responds to every food the same way — has not aged well. The clinical trials that shaped the food pyramid were powered to detect group averages, not individual response. They told us, on average, what happens when populations eat certain things. They did not promise that any one person would land near the average. A · NEJM, 2018

The newer literature, much of it driven by continuous glucose monitoring and stool metagenomics, suggests the variation between individuals can be larger than the variation between foods. Two people eating the same banana can record glucose curves that differ by a factor of three. A · Cell, 2015

And then it stopped being a footnote. The deeper the research went, the less the population average predicted any one person's response.

What the research suggests

A growing body of evidence points to response heterogeneity as the rule, not the exception. The implications for clinical guidance are uncomfortable: a recommendation that helps one patient may be inert or counterproductive for another. The honest stance is not to abandon recommendations but to soften them — to present what the evidence supports on average, and what an individual could observe in their own data. B · BMJ, 2019

Reading your own signal

For members curious about where their own response sits, three measurements tend to be tractable: post-prandial glucose, sleep quality the following night, and energy at the 90-minute mark after the meal. None of these are diagnostic on their own. Together, over enough meals, they sketch a shape. B · Lancet, 2020

Studies show that members who log this kind of data over four weeks often surface patterns their physician would not have predicted from labs alone. The point is not to replace clinical guidance — it is to bring more signal into the conversation.