Can diabetics eat bananas, rice, potatoes? The large cohort studies answer each one — and the answers are not the ones most people expect.
Can diabetics eat this or that is one of the most searched health questions there is, and the premise behind it is usually wrong. Diabetes guidance stopped working from lists of forbidden foods a long time ago.
NIDDK says it plainly: you may worry that having diabetes means giving up the foods you enjoy, but you can still have your favourite foods and drinks — you might just need smaller portions, or to have them less often (1).
So rather than a yes or no, here is what the large studies actually measured for the foods people ask about most.
Key Takeaways
- Whole fruit, including bananas, was associated with lower diabetes risk. Fruit juice with higher (3).
- French fries carried 20% higher risk. Baked, boiled and mashed potato carried none (5).
- White rice matters above about 300 g a day; brown rice went the other way (4).
- Eating protein and vegetables before carbohydrate cut the glucose peak by over 40% (6).
- NIDDK: there is no right or wrong way to choose foods, and no list of bans (1).
Can Diabetics Eat Whatever They Want? What The Guidance Says
The official position is less restrictive than the internet’s. Worth setting the ground first: in type 1 diabetes the body does not make insulin, in type 2 it does not make or use it well, and prediabetes means glucose is higher than normal but not yet in the diabetes range (2).
NIDDK states there is no right or wrong way to choose healthy foods and drinks for managing diabetes, and lists dairy or plant-based dairy, fruits, non-starchy vegetables, protein foods and whole grains among what a healthy plan may include (1). What it suggests having less of is specific: foods high in saturated fat, foods high in sodium, and sugary foods and drinks (1). Green beans are one of the non-starchy vegetables, though most of the blood sugar research is on dry beans rather than the pods.
The two methods it describes — carb counting and the plate method — are both about quantity and proportion, not prohibition (1). The plate method divides a 9-inch plate into sections so portions are controlled without weighing anything (1).
One genuine caution that gets less attention than it should: if you take insulin or certain diabetes medicines, alcohol can drive blood glucose too low. NIDDK advises eating when you drink and checking your glucose afterwards (1).
Fruit: The Answer Surprises People
| Fruit | Risk of type 2 diabetes per 3 servings a week | Reading |
|---|---|---|
| Blueberries | 0.74 (0.66–0.83) | Lowest risk of any fruit measured |
| Grapes and raisins | 0.88 (0.83–0.93) | Lower risk, despite the sugar reputation |
| Apples and pears | 0.93 (0.90–0.96) | Lower risk |
| Bananas | 0.95 (0.91–0.98) | Lower risk — the opposite of the folklore |
| Oranges | 0.99 (0.95–1.03) | No clear effect either way |
| Cantaloupe | 1.10 (1.02–1.18) | The one whole fruit linked to higher risk |
| Fruit juice | 1.08 (1.05–1.11) | Higher risk. Juice is not fruit |
This comes from three prospective cohorts — the Nurses’ Health Study, Nurses’ Health Study II and the Health Professionals Follow-up Study — covering 187,382 people over 3,464,641 person-years, during which 12,198 developed type 2 diabetes (3).
Whole fruit overall was associated with a hazard ratio of 0.98 per three servings a week (3). Broken down, the differences between fruits were statistically significant, and the ranking is not what the “fruit is sugar” argument predicts: blueberries 0.74, grapes and raisins 0.88, apples and pears 0.93, bananas 0.95 (3).
Only one whole fruit went the other way: cantaloupe, at 1.10 (3).
And the clean separation: fruit juice, at 1.08, was associated with higher risk (3). Whole fruit and juice behave differently, which is the single most useful thing in the paper.
An honest limit: these are people who did not have diabetes being followed to see who developed it. That is not the same as measuring what a banana does to blood glucose in someone who already has diabetes. Portion still matters day to day.
Rice And Potatoes: The Detail Everyone Misses
| Food | What the cohort data show | Practical reading |
|---|---|---|
| White rice | Relative risk 1.16 comparing highest with lowest intake. Above about 300 g a day, each extra 158 g serving carried 13% higher risk (4) | Amount is the variable, not the grain itself |
| Brown rice | Associated with lower risk of type 2 diabetes (4) | A straightforward swap that the data support |
| Potatoes, total | 5% higher rate per three servings a week, across 205,107 people (5) | Small, and driven almost entirely by one preparation |
| French fries | 20% higher rate per three servings a week (5) | This is where the risk actually sits |
| Baked, boiled or mashed potato | Not significantly associated with risk (1.01, 0.98–1.05) (5) | The potato is not the problem. The fryer is |
| Swapping potatoes for whole grains | Estimated 8% lower rate (5) | The single clearest substitution in the study |
SaveThe potato finding is the one worth reading twice. A 2025 analysis of 205,107 people over 5,175,501 person years, with 22,299 cases of type 2 diabetes, separated potatoes by preparation (5).
Total potato intake carried a 5% higher rate per three servings a week. French fries carried 20%. Baked, boiled or mashed potatoes were not significantly associated with risk at all — hazard ratio 1.01, confidence interval 0.98 to 1.05 (5).
So the answer to “can diabetics eat potatoes” is not about potatoes. It is about the fryer. For the separate question of whether a sprouting or green potato is safe at all, see what the guidance actually says.
Rice splits the same way, on amount rather than method. Pooling 8 cohort studies, white rice carried a relative risk of 1.16 comparing highest intake with lowest, with a dose-response appearing only above roughly 300 grams a day, where each additional 158 gram serving added 13% (4). Brown rice, in the same review, was associated with lower risk (4).
For the general principle behind both, see our guide to a low-glycaemic diet.
A Free Trick: The Order You Eat In
This one costs nothing and has a measured effect.
Fifteen people with prediabetes ate the same meal three times in random order: carbohydrate first, protein and vegetables first, or vegetables first (6). When carbohydrate came last, the incremental glucose peak fell by more than 40%, and the area under the glucose curve was 38.8% lower than when carbohydrate came first (6).
Fifteen people is a small study and it was in prediabetes, not established diabetes. Treat it as worth trying rather than proven. But eating the salad and the protein before the rice is not a hard change to make.
What Actually Moves The Needle
Stripped of the food-by-food arguments, the guidance comes down to a few things.
Amount, not bans. Carb counting and the plate method are the two approaches NIDDK names, and both are about how much (1).
Form over ingredient. Whole fruit behaves differently from juice (3). A boiled potato behaves differently from a fried one (5).
Swaps beat sacrifices. Replacing three weekly servings of potato with whole grains was estimated to cut rates by 8% (5). Brown rice for some of the white rice points the same way (4).
Fibre and whole foods. Fibre-rich foods and a balanced diet are the frame everything else sits inside. For the glucose side specifically, see ways to lower blood sugar.
And the line that matters: diabetes is managed with a health care team. Medicines, insulin doses and targets are individual, and no article can set them.
Carrots come up often in this question because they taste sweet. Their measured glycemic index is 23, lower than banana — the detail is in our piece on whether carrots are good for you.
Frequently Asked Questions
Can diabetics eat fruit?
Yes, and the largest evidence on this points the opposite way to the folklore. Three prospective cohorts following 187,382 people for over 3.4 million person-years, with 12,198 cases of type 2 diabetes, found that whole fruit intake was associated with slightly lower risk. Broken down by fruit, blueberries showed the lowest risk at a hazard ratio of 0.74 per three servings a week, followed by grapes and raisins at 0.88, apples and pears at 0.93 and bananas at 0.95. Only cantaloupe went the other way, at 1.10. The exception that matters is fruit juice, at 1.08, which was associated with higher risk.
Can diabetics eat bananas?
The evidence says bananas are not the problem they are made out to be. In the same three-cohort analysis, every three servings a week of bananas was associated with a hazard ratio of 0.95, meaning slightly lower risk of type 2 diabetes, with a confidence interval of 0.91 to 0.98 that does not cross one. That is an association in people without diabetes rather than a blood glucose measurement in people who have it, so portion still matters for day-to-day glucose. But the idea that bananas are off-limits does not come from the data.
Can diabetics eat rice?
Amount and type both matter more than a yes or no. A systematic review and meta-analysis pooling 8 cohort studies found white rice associated with a relative risk of 1.16 comparing the highest intake with the lowest. A dose-response appeared above roughly 300 grams a day, where each additional 158 gram serving carried 13% higher risk. Brown rice, in the same analysis, was associated with lower risk. So the practical answer is that a normal portion is not the issue, very high daily intake is, and switching some white rice for brown is a swap the evidence supports.
Can diabetics eat potatoes?
This is where a 2025 analysis of 205,107 people is unusually clear. Total potato intake carried a 5% higher rate of type 2 diabetes per three servings a week. French fries alone carried 20%. Baked, boiled or mashed potatoes were not significantly associated with risk at all, with a hazard ratio of 1.01 and a confidence interval of 0.98 to 1.05. Replacing three servings a week of potatoes with whole grains was estimated to lower rates by 8%. In other words the cooking method, not the vegetable, is what the data single out.
Does the order you eat food in matter?
It appears to, at least for the glucose rise after a meal. In a crossover study, 15 people with prediabetes ate the same meal three times in different orders. When protein and vegetables came first and carbohydrate followed 10 minutes later, the incremental glucose peak was reduced by more than 40% compared with eating the carbohydrate first, and the area under the glucose curve was 38.8% lower. It is a small study, so treat it as promising rather than settled, but it costs nothing to try.
So is any food actually forbidden?
NIDDK puts it directly: you may worry that having diabetes means giving up foods and drinks you enjoy, but you can still have your favourite foods, possibly in smaller portions or less often. It also states there is no single right or wrong way to choose foods. The methods it describes are carb counting and the plate method, both of which are about amount and balance rather than bans. The foods it suggests limiting are those high in saturated fat, high in sodium, and sugary foods and drinks.
The Bottom Line
Can diabetics eat a given food is usually the wrong question, because official guidance does not work from a banned list. NIDDK says favourite foods can stay, in smaller portions or less often, and that there is no right or wrong way to choose. Where the large studies do separate foods, the splits are instructive rather than restrictive: whole fruit including bananas was associated with lower risk while fruit juice was associated with higher; French fries carried 20% higher risk while baked, boiled and mashed potatoes carried none; white rice mattered above about 300 grams a day while brown rice went the other way. And eating protein and vegetables before the carbohydrate cut the post-meal glucose peak by over 40% in a small crossover study. Amount, form and order do more work than any list of forbidden foods.
This article is for informational purposes only and is not a substitute for professional medical advice. Diabetes is managed individually with a health care team. Do not change medicines, insulin doses or your meal plan on your own, and speak to your doctor or a registered dietitian about what suits your own targets.
Related Articles
How to Lower Blood Sugar Naturally: 8 Evidence-Based Habits
Low-Glycemic Diet: GI Food List, Benefits and Limits
High Fiber Foods List and How Much Fiber You Need
Balanced Diet Guide With the 5 Food Groups