What is the difference between glycaemic index and glycaemic load?

What GI measures

Glycaemic index is measured by giving participants a portion of food containing 50 g of available carbohydrate (or sometimes 25 g for foods where a larger portion would be impractical). Blood glucose is measured over the following two hours, and the incremental area under the glucose response curve is compared with the response to an equivalent amount of carbohydrate from a reference food, typically glucose, which has a GI of 100.

"Available carbohydrate" refers to carbohydrate that can be digested and absorbed, excluding dietary fibre. On UK and EU nutrition labels, fibre is already listed separately from carbohydrate. On US labels, fibre is included within total carbohydrate and generally needs to be accounted for when calculating available carbohydrate.

Glycaemic load accounts for portion size

GI tells you nothing about how much of the food someone eats. Glycaemic load accounts for that by scaling the GI to the carbohydrate in a real portion.

GL = GI × available carbohydrate in the portion (g) ÷ 100

Same food, two portions
Jasmine rice, white, cooked (GI 89, about 28 g available carbohydrate per 100 g)
  180 g serving: about 49.99 g carbohydrate
  GL = 89 × 49.99 ÷ 100 = 44.5  (high)
  90 g serving: about 24.99 g carbohydrate
  GL = 89 × 24.99 ÷ 100 = 22.2  (high)

High GI, low GL
Pineapple, raw (GI 82 in one study, about 11.72 g available carbohydrate per 100 g)
  100 g serving: about 11.72 g carbohydrate
  GL = 82 × 11.72 ÷ 100 = 9.6   (low)

Low GI, low GL
Watermelon (GI 50), 150 g serving: about 10.73 g available carbohydrate
  GL = 50 × 10.73 ÷ 100 = 5.4   (low)

The rice GI stays at 89 whatever the portion, but halving the serving halves the GL. It is still high, so a smaller portion alone may not be enough for a client who needs a lower glycaemic impact. Pineapple shows the reverse. Its GI is high, but a small serving holds little carbohydrate, so its GL is low. A larger serving of 150 g would reach about 14, which is medium. This is why GI alone can be misleading when comparing typical portions.

GI values are taken from the 2021 international tables (Atkinson et al.), which list rice as low as 54 for Doongara and 60 for basmati, and show watermelon at a mean of 50. Variety, ripeness and cooking all shift GI, so look up the specific food. The pineapple GI of 82 comes from a single study in the tables, so don't treat it as a universal value. GL values are rounded to one decimal place from unrounded carbohydrate figures. Available carbohydrate is carbohydrate minus fibre, from USDA FoodData Central entries for cooked white long-grain rice (28.2 g less 0.4 g fibre) raw pineapple (13.1 g less 1.4 g fibre) and raw watermelon (7.6 g less 0.4 g fibre) per 100 g. The rice figures are for generic cooked long-grain white rice, not jasmine specifically, so check the label or database entry for the product you use.

Why a meal is harder than a food

GI values come from single foods eaten on their own. Real meals mix them, and the other parts of a meal change the response.

Meal GI = Σ (GI of each food × its available carbs) ÷ total available carbs
Meal GL = Meal GI × total available carbs ÷ 100

• Fat and protein. Fat and protein can influence gastric emptying, insulin secretion and the timing of the post-meal glucose response. Adding them to a carbohydrate-containing meal may reduce or delay the initial glucose rise, although the effects vary depending on meal composition and individual metabolic health.
• Acidity. Acidity may influence post-meal glucose responses. Some studies suggest that vinegar consumed with carbohydrate-rich meals can reduce the initial glucose response, although the magnitude and consistency of this effect vary.
• Processing and cooking. Processing and cooking can change a food's GI. Rolled oats often have a lower GI than instant oats, and pasta cooked al dente often has a lower GI than softer pasta, but values vary by product, brand and cooking time. Check the specific food rather than assuming a type.
• Cooling cooked starches. Cooling cooked starches such as potatoes, rice and pasta can increase resistant starch formation through retrogradation. This may modestly reduce the post-meal glucose response, although the effect depends on the food, preparation and portion size.
• Individual responses. Individual responses vary. Factors such as insulin sensitivity, physical activity, sleep and meal composition can influence post-meal glucose responses. For clients who already monitor their glucose, these readings can provide additional context alongside GI and GL values.

To estimate a mixed meal, weight each food's GI by its share of the carbohydrate, then work out the GL from the total available carbohydrate.

These calculations provide theoretical estimates rather than precise predictions of post-meal glucose responses, as interactions between meal components and individual physiological factors can influence the outcome.

How to use them with clients

GI is most useful for swaps inside the same food group: rolled oats for cornflakes, grainy bread for white, a firm waxy potato for a floury one. GL is particularly useful when discussing portion sizes, as it demonstrates how the amount of carbohydrate consumed contributes to the overall glycaemic impact of a food. Both carbohydrate quantity and quality remain important considerations.

Don't use either measure on its own to judge a food. Chocolate can have a relatively low GI because of its fat content, while some fruits, such as pineapple, may have a higher GI but a low GL in typical portions. Neither measure captures the overall nutritional quality of a food. A low GI or GL doesn't automatically make a food nutritionally superior or more suitable for weight loss, because neither accounts for energy density or micronutrient content.

For clients with diabetes, carbohydrate quantity, food quality and overall dietary patterns remain central to nutritional management. GI and GL can provide additional guidance when selecting carbohydrate-rich foods, particularly when choosing between otherwise nutritionally similar options. How to build a meal plan for type 2 diabetes puts it in context.