Nutrition & Metabolism
Why the Glycaemic Index Is Both Useful and a Lie
The same bowl of rice can spike your blood sugar dramatically or barely move it at all — depending on who's eating it.
The Idea
The glycaemic index, or GI, ranks foods by how quickly they raise blood glucose compared to pure glucose. White bread scores high; lentils score low. The logic seems clean: eat low-GI foods, avoid blood sugar spikes, stay energised, reduce your risk of type 2 diabetes. And at a population level, that logic holds up reasonably well. But here's where it gets interesting — and where the GI starts to quietly unravel. The index was derived by testing foods on groups of people and averaging the results. That averaging hides something enormous: individual variation. Researchers at the Weizmann Institute discovered that two people can eat identical meals and have blood glucose responses so different they might as well be eating different foods. One person's blood sugar barely moves after a banana; another person's rockets. The same individual can also respond differently to the same food on different days, depending on sleep quality, gut microbiome composition, stress levels, and what they ate the meal before. So the GI is a statistical artefact — a useful population-level signal dressed up as personal dietary advice. It tells you something real about food chemistry, particularly the role of fibre in slowing glucose absorption and the way processing strips that protection away. But treating it as a precise prescription for your body specifically is like navigating with a map drawn from other people's journeys.
In the World
In 2015, Eran Segal and Eran Elinav at the Weizmann Institute of Science in Israel recruited 800 adults and fitted them with continuous glucose monitors — devices that track blood sugar in real time, every five minutes, for a week. Participants ate standardised meals and logged everything else they ate, along with sleep and exercise data. The researchers also sequenced each person's gut microbiome. What they found upended a lot of confident nutritional advice. One participant, a middle-aged woman with pre-diabetes, had been carefully following a low-GI diet. She had been told, and had every reason to believe, that tomatoes were a safe choice. But her continuous monitor showed that tomatoes sent her blood sugar soaring — higher than almost any other food she tested. For her specifically, tomatoes were a high-GI food, regardless of what the index said. Meanwhile, other participants handled white rice without drama. The team used their data to build a machine-learning algorithm that could predict an individual's glucose response to a new food using gut microbiome data and personal variables — and it worked far better than the GI alone. The study, published in the journal Cell, didn't kill the glycaemic index. But it reframed it: a useful heuristic, not a biological law, and one that needs personalising to mean much at all.
Why It Matters
Most nutritional advice floats somewhere between over-confident and flat-out contradictory, and the GI is a good lens for understanding why. It was built on real science but applied as though biology were more uniform than it is. Knowing this won't tell you exactly what to eat — that's still genuinely complicated — but it does free you from treating any single ranking system as gospel. It also nudges you toward something more useful: paying attention to how your own body actually responds. Not everyone has access to a continuous glucose monitor, but the underlying principle is broadly applicable. How do you feel two hours after a particular meal? How does your energy hold across the afternoon? These aren't precise measurements, but they're data about you, not about an averaged stranger. The deeper takeaway is about how we use population statistics in personal decisions. Averages are real and valuable; they just don't straightforwardly apply to individuals. That gap — between what's true at the group level and what's true for you — is one of the most underappreciated ideas in all of health science.
A Question to Ponder
If your body's response to food is shaped partly by your unique gut microbiome — which is itself shaped by your history, environment, and choices — how much of what we call 'healthy eating' is actually portable advice, and how much needs to be discovered rather than prescribed?
Get a new one of these every morning.
Start learning with Thinkable