How cooking methods affect blood sugar: boil, bake or cool?

Cooking changes how quickly starch turns into glucose: boiling, cooking pasta firm and cooling cooked starches usually give a lower glycaemic index (GI) than baking, roasting or mashing. In the international GI tables, boiled sweet potato averages 46 against 86 to 88 baked or roasted, and potato cooked then cooled averages 49 against 73 eaten hot.

The same food can sit in a different GI band depending on how you cook it. Filter the list, sort it, or pick what you usually cook to find a lower GI way to make it.

Why does cooking change blood sugar?

Cooking softens starch so your gut can digest it faster. Raw starch sits in tight granules that enzymes struggle to reach. Heat and water make those granules swell and burst, a process called gelatinisation, and the more fully gelatinised the starch, the faster it becomes glucose in your blood.

Three other things shift the result:

• Structure: whole pieces, intact cell walls and firm pasta slow digestion. Mashing, blending and long cooking speed it up.
• Cooling: as cooked starch cools, some of it re-forms into tighter chains, known as resistant starch, which digests slowly. This is called retrogradation.
• What comes with it: fat, protein, fibre and acid such as vinegar or lemon juice slow stomach emptying, so glucose arrives more gradually.

Boiling vs baking vs roasting

For potato and sweet potato, boiling gives the lowest GI and baking or roasting the highest. The gap is biggest for sweet potato. Across 13 studies, boiled sweet potato averaged GI 46. Baked and roasted sweet potato averaged 86 to 88, across two sets of 11 studies. A Jamaican study by Bahado-Singh and colleagues (2011) tested ten varieties and found boiling gave 41 to 50, while baking gave 82 to 94.

Potatoes are high GI whichever way you cook them, but the method still moves them. In a Canadian study, Fernandes and colleagues (2005) found boiled red potatoes had a GI of 89 eaten hot and 56 eaten cold the next day. Roasted white potatoes came in at 72 and baked Russets at 77. Variety mattered as much as method: the international tables list boiled potatoes from 58 for Nicola to 101 for Desiree.

Frying is a special case. Fat slows stomach emptying, so oven fries test at a GI of about 64 to 65, lower than a baked potato. They still carry more kilojoules, and a lower GI does not make chips a better choice for your health. Our guide to the GI of potato goes through varieties in more detail.

Does cooling and reheating lower the GI?

Yes, modestly. Cooling cooked starch overnight forms resistant starch, and reheating does not undo all of it.

The best rice study is small but well run. Sonia and colleagues (2015) gave 15 healthy adults white rice that was freshly cooked, or cooked, chilled at 4°C for 24 hours and reheated. The chilled rice had more than twice the resistant starch (1.65 vs 0.64 g per 100 g) and an 18% lower glucose response. For potato, the international tables average a GI of 49 for potato cooked then cooled, against 73 for boiled potato eaten hot. The University of Sydney's GI Foundation notes that a vinaigrette potato salad made the day before has a much lower GI than hot potatoes, partly because cold storage raises resistant starch by more than a third.

Cool it safely. Cooked rice can carry spores of Bacillus cereus, which survive cooking. HealthyWA advises keeping cooked rice above 60°C or cooling it as fast as possible: spread it in shallow containers less than 10 cm deep, keep them apart rather than stacked, and store below 5°C. Reheating does not destroy toxins that formed while rice sat warm. Our list of low GI rice covers which varieties to start with.

Does overcooking pasta raise its GI?

Yes. Pasta is low GI because its starch is trapped in a network of gluten protein, and long cooking loosens that network. The GI Foundation advises cooking pasta al dente, firm to the bite, and checking it a few minutes before the packet time, because overcooking raises the GI.

Boiled white spaghetti averages GI 47 across 11 tested foods and wholemeal spaghetti 50, both low. Our list of low GI pasta choices has more options.

What our members' CGM data shows

In everyday meals, cooking method mattered less than the lab numbers suggest. Members who logged boiled, baked or mashed potato saw similar typical rises.

Roast potato in a meal (725 members) and potato salad (274 members) also came in at a typical 1.3 mmol/L. Fried rice, often made with day-old rice, rose 2.0 mmol/L when eaten alone, slightly more than plain white rice at 1.8. Bigger servings, oil and sauce may outweigh any benefit from cooling.

Why the gap between lab and life? GI tests feed people 50 g of carbohydrate from one food after an overnight fast. Real portions are smaller, meals come with butter, gravy, meat and vegetables, and members logged different varieties. Cooking method is still worth using, but portion size and the rest of the plate move the result more. Members' typical responses to sweet potato and pasta are on our food pages.

“But in those who average blood glucose levels remain in the normal range, some of those people are in fact experiencing highs and lows and this in itself can be an issue.”
Dr Michelle Woolhouse, Vively Medical Director

Kitchen habits for steadier glucose

Small changes in how you cook and plate starchy food add up. Tick off the ones you already do.

Meal order helps too. Eating vegetables and protein before the starch slows the rise for many people, and adding vinegar, dairy or beans to rice lowered its GI by 20 to 40% in a Japanese study by Sugiyama and colleagues (2003).

“Eat your vegetables or salad first, followed by protein/fat and then carbohydrates.”
Dr Michelle Woolhouse, Vively Medical Director

For more ways to flatten the curve after a meal, see our guide to glucose spikes.

How Vively can help

Tested GI values come from small groups of volunteers eating one food at a time. Your response to last night's roast potatoes or this morning's leftover rice is personal. The Vively CGM program shows how your own glucose responds to the way you cook and combine food, and the app and dietitian support through the membership help you work out which changes are worth keeping.

Frequently asked questions

Does baking increase the glycaemic index of food?

For starchy foods, yes. Baking fully gelatinises the starch and dries the food, so it digests faster. Baked or roasted sweet potato averages GI 86 to 88 against 46 boiled, and baked potato tests at 83 to 93 against an average of 73 boiled. Baking makes little difference to foods with little starch, such as most vegetables.

Does cooling rice lower its effect on blood sugar?

A little. In a 2015 study of 15 adults, white rice chilled for 24 hours then reheated had more than twice the resistant starch and an 18% lower glucose response than freshly cooked rice. Cool rice quickly in shallow containers, keep it below 5°C and reheat it until steaming hot.

Is it safe to cool and reheat rice?

Yes, if you cool it fast. Cooked rice can carry Bacillus cereus spores that survive cooking and make toxins if rice sits warm. HealthyWA advises cooling rice quickly in shallow containers less than 10 cm deep and storing it below 5°C. Reheating does not destroy toxins that have already formed.

Does frying lower the GI of potato?

It can. Fat slows stomach emptying, so oven fries test at a GI of about 64 to 65, lower than baked potato. Fried sweet potato averages 71, higher than boiled at 46. Fried food carries more kilojoules and often more salt, so a lower GI does not make it a healthier choice.

Does microwaving raise the GI of food?

Not compared with other hot cooking. Microwaved potatoes test at a GI of 73 to 93, in the same range as boiled and baked potato. Microwaving is a quick way to cook vegetables with little water. What matters more is the food, its variety and whether you cool it afterwards.

Does overcooking pasta raise blood sugar?

Yes. Long cooking loosens the protein network that keeps pasta's starch digesting slowly. The University of Sydney's GI Foundation advises cooking pasta al dente and checking it a few minutes early. Boiled white spaghetti averages GI 47 across 11 tested foods, so cooked well it is a low GI choice.

Sources

1. Atkinson FS, Brand-Miller JC, Foster-Powell K, Buyken AE, Goletzke J. International tables of glycemic index and glycemic load values 2021: a systematic review. Am J Clin Nutr 2021;114(5):1625-32 (supplemental table 1).
2. University of Sydney GI Foundation. GI FAQs (pasta al dente, potato salad, fat and acid).
3. Sonia S, Witjaksono F, Ridwan R. Effect of cooling of cooked white rice on resistant starch content and glycemic response. Asia Pac J Clin Nutr 2015;24(4):620-5.
4. Fernandes G, Velangi A, Wolever TM. Glycemic index of potatoes commonly consumed in North America. J Am Diet Assoc 2005;105(4):557-62.
5. Bahado-Singh PS, Riley CK, Wheatley AO, Lowe HI. Relationship between processing method and the glycemic indices of ten sweet potato cultivars commonly consumed in Jamaica. J Nutr Metab 2011;2011:584832.
6. Sugiyama M et al. Glycemic index of single and mixed meal foods among common Japanese foods with white rice as a reference food. Eur J Clin Nutr 2003;57:743-52.
7. HealthyWA, Department of Health Western Australia. Safe cooling of cooked rice.
8. Vively. How we get our member data numbers.