Nuts, blood sugar and the glycaemic index

Most nuts do not have a useful glycaemic index, because they do not contain enough carbohydrate for the test to work properly. The method needs a 50 g dose of available carbohydrate from a single food, and a handful of almonds or walnuts is nowhere near that. Where nuts have been tested, the values are very low. That tells you as much about the test as it does about the food.

The short answer

• The glycaemic index ranks carbohydrate foods by how far they raise blood glucose over two hours, compared with glucose at 100.
• Peanuts and cashews are the only common nuts with published values, and both sit at the bottom of the scale. Most other nuts contain too little carbohydrate to test in any realistic way.
• A low glycaemic index does not make a food healthy, and a high one does not make it harmful. Watermelon scores high. Chocolate scores low.

This article sits with the rest of our nutrition coverage, which looks at what is in foods rather than what is claimed for them.

What does the glycaemic index measure?

The index came out of work at the University of Toronto at the start of the 1980s, as an attempt to give people with diabetes a better guide to carbohydrate foods than the exchange lists then in use (the original papers are on PubMed). The method is narrow and specific. Volunteers eat a portion of a single food containing 50 g of available carbohydrate. Their blood glucose is then measured repeatedly over the next two hours. The area under that curve is compared with the same person's response to 50 g of pure glucose, which is set at 100.

Foods are then grouped as low at 55 or below, medium from 56 to 69, and high at 70 or above. Diabetes UK explains the classification and its caveats in its own food guidance.

Glycaemic load is a later adjustment. It multiplies the index by the grams of carbohydrate in a realistic portion. That is why watermelon can have a high index and a trivial load. For nuts, the load is close to zero on both counts.

Why is the glycaemic index limited?

Because it answers one narrow question, and not the one many people assume it does. It measures the shape of a blood glucose curve after a single food eaten on its own. It says nothing about fat, salt, fibre, energy density or how much of the food people actually eat.

That is why the scale can sort foods in ways that do not match overall diet quality. Crisps and some confectionery score low, because fat slows gastric emptying. Boiled potatoes and wholemeal bread score high. Ripeness, processing, cooking time and what else is on the plate all change the number. So does the person. The same food tested in different people, or in the same person on different days, can produce noticeably different results.

The UK's Scientific Advisory Committee on Nutrition reviewed the evidence on carbohydrates in 2015 and did not adopt glycaemic index as a population dietary target, recommending instead an increase in fibre and a reduction in free sugars (SACN, Carbohydrates and Health). The NHS guidance on starchy foods and carbohydrates takes much the same line, treating fibre and portion as more useful handles. The index is a research tool that escaped into food marketing. The World Health Organization's healthy diet advice does not mention it either, framing carbohydrate quality in terms of whole grains, pulses, fruit and vegetables.

What is the glycaemic index of nuts?

Only two common ones have been tested often enough to publish a mean.

Values are rounded means drawn from the international tables of glycaemic index and glycaemic load, which pool results from many separate studies. Individual test results for a single food often span twenty points or more. See the compiled tables and the studies behind them on PubMed.

Two points matter here. Peanuts are legumes rather than tree nuts, which is why they sometimes appear in a different row of a food composition table — we cover the difference and what it means in practice separately. And the blank row is not an oversight. Carbohydrate contents in the UK composition of foods dataset are low enough for most tree nuts that a 50 g carbohydrate test dose would mean eating several hundred grams at once, which nobody does.

Do nuts change the blood glucose response to a meal?

This is the more useful question. The direction of effect is fairly consistent. The size of it is not settled. Trials that add nuts to a carbohydrate meal generally report a smaller rise in blood glucose afterwards than the same meal without them, which is what you would expect from added fat, protein and fibre slowing digestion (the trial literature is indexed here).

What that means over months rather than hours is less clear. Longer trials measuring HbA1c have been small, short and mixed in design, and the changes reported are modest. Research suggests nuts may support steadier glucose handling as part of a wider dietary pattern. It does not show them doing anything on their own. The same caution applies to the related question of nuts and body weight, where the energy density on the label and the energy people actually absorb turn out not to match.

What if you have diabetes?

Then this article is background reading, not guidance. Nuts are not a treatment for diabetes, and nothing here should be used to change medication, adjust a meal plan or interpret a glucose reading. Blood glucose targets, carbohydrate counting and the effect of any dietary change are matters for your GP, your diabetes team or a registered dietitian, who can see your results and your prescriptions. The NHS overview of type 2 diabetes is a reasonable starting point, and the British Dietetic Association food facts pages are written for the public rather than for clinicians.

What does the evidence not show?

It does not show that eating nuts lowers blood glucose in any clinically meaningful, durable way. The short-term meal studies are real, but they measure a curve over a few hours in small groups, often in healthy volunteers. Short-term postprandial changes are a poor guide to long-term outcomes.

It does not show that a low glycaemic index diet is better than a high-fibre diet for most people. Those two things overlap heavily, and trials rarely separate them cleanly.

The published index values themselves are also less tidy than they look. They are means from small studies, with wide variation between people, and the same food can test twenty or thirty points apart depending on variety, ripeness, processing and the laboratory. Treating a single figure as a fixed property of a food misreads the data.

And no health claim about nuts and blood glucose is authorised on the Great Britain nutrition and health claims register. Where we describe evidence as mixed or weak, that is a statement about the studies rather than a polite hedge — our approach to sourcing and uncertainty explains how we decide.

The broader case for nuts, such as it is, rests on their nutrient content and on cohort evidence about overall dietary patterns, not on the glycaemic index. We set out what that case does and does not support in more detail.

Where this came from

• SACN, Carbohydrates and Health (2015)
• NHS, Starchy foods and carbohydrates
• NHS, Type 2 diabetes
• Diabetes UK, food and nutrition guidance
• British Dietetic Association, Food Facts
• Composition of foods integrated dataset (CoFID)
• Great Britain nutrition and health claims register
• International tables of glycaemic index and glycaemic load values, on PubMed
• Trials of nuts and post-meal blood glucose, on PubMed
• WHO, Healthy diet fact sheet