Paleo Diet for Type 2 Diabetes: What Trials Show

By Paleo Life Coach Editorial Team · Last updated September 25, 2026 · Evidence from randomized trials indexed on PubMed

If you’re weighing a paleo diet for type 2 diabetes, you’ve probably seen claims that going paleo can “fix” blood sugar. You may also have heard a clinician say there’s no evidence for it. Neither is quite right.

The paleo diet has been tested for type 2 diabetes in a handful of small, well-run trials, and the results are more interesting than either side suggests. This guide summarizes what those trials found, what a 2025 study reveals about why paleo seemed to work, and how to try it safely with your care team.

The quick version: In small trials lasting 3 weeks to 3 months, paleo diets improved blood sugar, waist size and triglycerides more than standard diabetes diets. But a 2025 trial that held weight and macronutrients equal found no HbA1c difference. The benefit likely comes from weight loss and fewer processed carbs, not from banning grains, dairy and legumes.

What does the research actually test?

The trials compare a paleo-type diet (lean meat, fish, eggs, vegetables, fruit, root vegetables and nuts; no grains, dairy, legumes or added sugar) against a guideline-based diabetes or “healthy” diet that includes whole grains, low-fat dairy and legumes. Most are small, short and conducted by a few research groups in Sweden and California. That matters: they show what can happen under close supervision for weeks or months, not what happens over years of real life.

What did the paleo diabetes trials find?

Here is every major trial in plain numbers, from the original abstracts.

What did the paleo diabetes trials find?
Trial	Participants	Design and length	Comparison diet	Main result
Lindeberg et al., 2007	29 people with heart disease plus glucose intolerance or type 2 diabetes	Randomized, 12 weeks	Mediterranean-like “consensus” diet	Glucose curve (AUC) fell 26% on paleo vs 7% on control; improvement was independent of waist change
Jönsson et al., 2009	13 adults with type 2 diabetes (most on metformin)	Crossover, two 3-month periods	Diabetes diet per guidelines	HbA1c −0.4 percentage points, triglycerides −0.4 mmol/L, diastolic BP −4 mm Hg, weight −3 kg, waist −4 cm vs diabetes diet
Masharani et al., 2015	24 adults with type 2 diabetes	Metabolically controlled, about 3 weeks	American Diabetes Association-based diet	Both groups improved; paleo group had greater benefits in glucose control and lipids
Otten et al., 2019	22 adults with type 2 diabetes and overweight or obesity	Randomized, 12 weeks	Paleo alone vs paleo plus supervised exercise	Both groups had major metabolic improvements; exercise added cardiac benefits
Fontes-Villalba et al., 2025	14 adults with type 2 diabetes	Crossover, two 4-week periods, weight kept stable	Diabetes diet matched for macronutrients and glycemic load	No difference in HbA1c or fructosamine
Manheimer et al., 2015 (meta-analysis)	4 trials, 159 people with metabolic syndrome components	Pooled analysis	Guideline-based diets	Fasting glucose −0.16 mmol/L, with a confidence interval crossing zero

Sources (PubMed abstracts): Lindeberg 2007, Jönsson 2009, Masharani 2015, Otten 2019, Fontes-Villalba 2025, Manheimer 2015.

Read across the rows and a pattern appears. When people ate paleo freely, they tended to eat less energy, lose weight and waist, and cut glycemic load; blood sugar improved. Jönsson’s team reported that their paleo diet was lower in total energy, energy density, carbohydrate and glycemic load than the diabetes diet. The Lindeberg trial is the notable exception suggesting some effect beyond waist size, but it was small and in a specific group of heart patients.

Why does the 2025 weight-stable trial matter?

Because it tested the paleo idea directly. If grains, dairy and legumes themselves harm glucose control, removing them should help even when everything else is equal. In the 2025 crossover, researchers adjusted calories to keep body weight stable and matched the two diets for macronutrients and glycemic load. The paleo version excluded cereal grains, dairy and legumes; the diabetes version included substantial whole grains, low-fat dairy and legumes. After four weeks on each, there was no difference in HbA1c or fructosamine among the 14 participants (Fontes-Villalba et al., J Nutr Sci 2025).

The authors concluded that the characteristic food-group differences between the diets do not cause an effect on glycemic control. It’s one small, short trial, so it isn’t the final word. But it points to a practical conclusion: the parts of paleo that likely help blood sugar are weight loss, fewer refined and processed carbohydrates and a lower glycemic load, all of which you can also get from other well-planned diets. The Mediterranean-style diet used as a comparison in the Lindeberg trial is one example; our paleo vs Mediterranean comparison covers how the two differ.

What does mainstream diabetes guidance say?

Official guidance doesn’t endorse one named diet. The National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) says there is no right or wrong way to choose healthy foods and drinks that may help manage diabetes, and that a healthy meal plan, with physical activity, sleep and other healthy behaviors, may help people with overweight or obesity reach and keep a healthy weight (NIDDK, Healthy Living with Diabetes). NIDDK notes that healthy diabetes meal plans may include dairy or plant-based dairy, which paleo omits.

So a paleo-style plan can fit within mainstream advice, as long as it is nutritionally complete and agreed with your care team. Harvard’s review of the paleo diet raises the caveats: cutting whole grains and dairy can lower fiber, calcium and vitamin D intake if you don’t plan replacements (Harvard T.H. Chan School of Public Health). Our guide to paleo calcium and nutrient gaps shows how to close them.

Is the paleo diet safe if you take diabetes medication?

It can be, but the switch needs supervision. Cutting carbohydrates quickly while taking insulin or certain other diabetes medicines can drop blood sugar too low. MedlinePlus explains that low blood sugar, below 70 mg/dL, may occur in people with diabetes who take insulin or certain other medicines, although many other diabetes medicines don’t cause it (MedlinePlus, Low blood sugar). Exercise can also lower blood sugar in people taking insulin.

Before you start:

Tell your doctor or diabetes educator what you plan to change, especially if you use insulin or a sulfonylurea. Your doses may need adjusting as your eating changes.
Check your glucose more often for the first few weeks, including before driving.
Know the signs of low blood sugar and keep a fast-acting carbohydrate with you.
Change one thing at a time so you and your clinician can see what’s working.
Keep lab follow-up for HbA1c, kidney function and cholesterol.

Never stop or change a diabetes medication on your own because your readings improve.

What does a paleo-style plate for type 2 diabetes look like?

Build each meal from non-starchy vegetables, a protein and a measured portion of a starchy vegetable or fruit. The swap table shows how common starches on a diabetes plate compare with paleo options, per serving, from USDA FoodData Central.

What does a paleo-style plate for type 2 diabetes look like?
Food (serving)	Carbs (g)	Fiber (g)	Digestible carbs (g)
White rice, cooked (1 cup) — typical starch	44.5	0.6	43.9
Sweet potato, baked in skin (1 cup)	41.4	6.6	34.8
White potato, baked with skin (1 medium)	36.6	3.8	32.8
Butternut squash, baked (1 cup cubes)	21.5	6.6	14.9
Carrots, raw (1 cup chopped)	12.3	3.6	8.7
Cauliflower “rice,” boiled (1 cup)	5.1	2.9	2.2
Blueberries (1 cup)	21.4	3.6	17.9
Raspberries (1 cup)	14.7	8.0	6.7

Method: USDA SR Legacy values (FDC IDs 168878, 168483, 170093, 169296, 170393, 170397, 171711, 167755) scaled to USDA portion weights. Digestible carbs = total carbohydrate minus fiber.

The table shows why “paleo” alone doesn’t guarantee low-carb meals: a cup of sweet potato carries about as much digestible carbohydrate as three-quarters of a cup of rice. Portion size and pairing matter more than the diet’s label. For daily carb targets and a full sample day, see how many carbs to eat on paleo; agree on your own number with your care team.

A sample paleo-style diabetes plate:

Half the plate: non-starchy vegetables such as broccoli, green beans, salad greens, cauliflower or zucchini
A quarter: fish, chicken, eggs or lean meat
A quarter or less: squash, sweet potato or a small fruit
Plus: a thumb of healthy fat such as olive oil, avocado or nuts

Common mistakes when trying paleo with diabetes

Assuming paleo means unlimited fruit and sweet potatoes. Both carry significant carbohydrate; measure portions.
Swapping bread for paleo baked goods. Cassava and tapioca flours are more carb-dense than wheat flour.
Using honey or maple syrup freely. They’re still added sugars.
Cutting carbs sharply without telling your prescriber. This is how avoidable low-blood-sugar episodes happen.
Expecting results beyond the evidence. The trials show improvements, not a cure, over weeks to months.
Ignoring calcium and fiber. Removing dairy and whole grains without replacements creates new problems.

Frequently asked questions about the paleo diet and type 2 diabetes

Does the paleo diet help type 2 diabetes?

Small trials suggest it can improve blood sugar, waist size and triglycerides compared with standard diabetes diets over 3 weeks to 3 months. A 2025 trial that kept weight and macronutrients equal found no HbA1c difference, suggesting the benefit mainly comes from weight loss and eating fewer processed carbohydrates rather than from excluding grains, dairy and legumes.

Can the paleo diet reverse type 2 diabetes?

The trials don’t show reversal. They show improvements in measures such as HbA1c and glucose tolerance over weeks to a few months, in small groups under supervision. Some people reach lower medication needs with weight loss and diet change, but that should be managed with your doctor, not assumed.

Is paleo better than a diabetes diet?

In several short trials, paleo improved glucose control more than guideline-based diabetes diets, but people on paleo also tended to eat less and lose more weight. When weight and macronutrients were matched in a 2025 trial, results were the same. The best diet is one you can sustain that lowers your numbers.

What can a diabetic eat on the paleo diet?

Non-starchy vegetables, fish, poultry, eggs, lean meat, nuts, seeds, olive oil, avocado, berries and measured portions of starchy vegetables such as squash and sweet potato. Limit high-sugar fruit, honey, maple syrup and paleo baked goods. Plan calcium and fiber sources because dairy and whole grains are excluded.

Is paleo safe for people on insulin?

It can be, with supervision. Reducing carbohydrates while taking insulin or certain other diabetes medicines can cause low blood sugar. Talk to your doctor before changing your diet, check glucose more often during the transition, and let your care team adjust doses if needed.

Your next step: bring the evidence to your care team

If paleo appeals to you, take this summary to your doctor or diabetes educator and agree on a trial period, a glucose-monitoring plan and when to recheck HbA1c. Focus on the parts with the best support: more vegetables, fewer processed carbohydrates, measured starch portions and gradual weight loss if you need it. For background on the diet itself, start with the complete paleo diet guide, and see paleo diet health benefits for the wider research.

Consider this general nutrition reading, not a replacement for medical advice. If you have diabetes, talk to your doctor before changing your diet, especially if you take insulin or other glucose-lowering medication.

Paleo Diet for Type 2 Diabetes: What the Trials Show (and What They Don’t)