Glycaemic index: does choosing low-GI foods actually help?

The glycaemic index (GI) ranks foods by how sharply they raise blood sugar, but using it to pick foods does not reliably improve health: in a tightly controlled feeding trial, swapping high-GI for low-GI carbohydrate within an already healthy diet did not improve insulin sensitivity, cholesterol or blood pressure [s1]. The broader evidence rates the case for judging carbohydrates by glycaemic index as weak — well below the evidence for fibre and whole grains [s2].

What GI and glycaemic load are

The glycaemic index scores a carbohydrate food from 0 to 100 by how much it raises blood glucose relative to pure glucose; glycaemic load adjusts that score for how much carbohydrate a typical portion contains. The intuition behind low-GI eating is that gentler blood-sugar rises should mean better insulin sensitivity and lower risk of diabetes and heart disease. It is a plausible mechanism, and it has spawned a large industry of GI tables and product labels. The question is whether choosing foods by that number actually changes the outcomes that matter.

The trial built to test it

The most rigorous test is OmniCarb, a randomised controlled feeding trial in which every meal was provided to participants. It enrolled 163 overweight adults with above-optimal blood pressure and had them eat four complete diets, each for five weeks: combinations of high or low glycaemic index (65 versus 40 on the glucose scale) with high or low total carbohydrate [s1]. Every diet was built on a healthful DASH-style pattern, so the comparison isolated the glycaemic index itself rather than overall diet quality [s1].

The results undercut the low-GI premise. In the main comparison — the low-GI, low-carbohydrate diet versus the high-GI, high-carbohydrate diet — there was no improvement in insulin sensitivity, systolic blood pressure, LDL cholesterol or HDL cholesterol; the only change was a fall in triglycerides, from 111 to 86 mg/dL [s1]. More pointedly, at high carbohydrate intake, switching from high to low GI actually reduced insulin sensitivity, from 8.9 to 7.1 units (a 20 percent drop), and raised LDL cholesterol from 139 to 147 mg/dL [s1]. The authors' conclusion was blunt: in the context of an overall healthy diet, using glycaemic index to choose specific foods may not improve cardiovascular risk factors or insulin resistance [s1].

Why the mechanism can mislead

OmniCarb is a clean illustration of a recurring lesson in nutrition: a sensible-sounding mechanism does not guarantee a real-world benefit. Because the trial fed people a good diet either way, it shows that once the broad pattern is healthy, the incremental value of fine-tuning by GI is small to nonexistent [s1]. That does not mean blood-sugar responses are meaningless — it means the single number on a GI chart is a poor lever for improving health in people already eating reasonably.

The wider literature agrees the GI evidence is soft. A large Lancet series of systematic reviews and meta-analyses on carbohydrate quality, spanning 185 prospective studies and 58 trials, graded the certainty of evidence as moderate for dietary fibre but only low to very low for glycaemic index and glycaemic load [s2]. In other words, the strongest carbohydrate-quality signals attach to fibre and whole grains, not to GI — the reviewers found 15 to 30 percent lower rates of major outcomes among the highest fibre consumers, with benefit greatest at 25 to 29 grams of fibre a day [s2]. That low grading for GI held across both the prospective cohort studies and the trials the series pooled, so it is not a case of observational data and experiments pointing in different directions — the evidence for glycaemic index is simply thin on both sides [s2].

What it means

The honest reading is that glycaemic index is a real physiological property that has been oversold as a dietary rule. A well-run feeding trial found no benefit from choosing low-GI foods within a healthy diet, and the graded evidence puts GI at the bottom of the carbohydrate-quality hierarchy [s1][s2]. For most people the useful signal is simpler and better supported: more fibre and more whole grains, less refined starch.

Related pieces develop that theme: the dose-response case for whole grains, the target for how much fibre you actually need, what continuous monitors do and don't reveal in glucose monitors for people without diabetes, and the popular claim that meal sequence blunts blood sugar in why food order affects glucose. This article is informational and not dietary advice.