Tonstad 2009/2013 AHS-2 + Barnard PCRM trials

Vegan vs Vegetarian Type 2 Diabetes: 62% Lower Risk

Type 2 diabetes is the chronic disease where plant-based eating has perhaps the cleanest published evidence base. The Adventist Health Study cohort data is striking, the randomised controlled trial evidence is consistent, and the mechanisms are reasonably well characterised. This page presents the numbers and the practical guidance for someone using diet to prevent or manage type 2 diabetes.

The short answer. Vegans have 49 to 62% lower type 2 diabetes risk than non-vegetarians (AHS-2). Vegetarians 39% lower. Semi-vegetarians 26% lower. RCT evidence (Barnard PCRM trials) shows vegan diet reduces HbA1c by approximately 1.2 percentage points vs 0.4 for the ADA diet over 22 weeks. Mechanisms include lower BMI, higher fibre, lower intramyocellular lipid. Coordinate medication changes with your diabetes team.

The Adventist Health Study-2 numbers

The AHS-2 Tonstad 2009 cross-sectional analysis (Diabetes Care; 32: 791-796) examined the relationship between vegetarian diet pattern and type 2 diabetes prevalence in 22,434 adult men and 38,469 adult women in the cohort. After adjustment for age, sex, ethnicity, education, BMI, physical activity, sleep, alcohol use, and television viewing, the diabetes prevalence odds ratios were:

Diet group	BMI-adjusted odds ratio for T2D	BMI-unadjusted odds ratio
Non-vegetarian (reference)	1.00	1.00
Semi-vegetarian	0.74 (26% lower)	0.50 (50% lower)
Pesco-vegetarian	0.49 (51% lower)	0.46 (54% lower)
Lacto-ovo vegetarian	0.61 (39% lower)	0.39 (61% lower)
Vegan	0.38 (62% lower)	0.22 (78% lower)

The BMI-adjusted and unadjusted columns matter because part of the diabetes benefit of vegan eating runs through weight loss, but a meaningful portion is independent of weight. The 62% reduction after BMI adjustment in vegans indicates that the dietary composition itself (not just the weight loss it produces) is protective. The 2013 follow-up paper (Tonstad et al., Nutr Metab Cardiovasc Dis; 23: 292-299) reported the prospective incidence of new-onset diabetes over 2 years, with similar effect sizes confirming the cross-sectional finding.

The Barnard PCRM trials

Neal Barnard and the Physicians Committee for Responsible Medicine published a series of randomised trials comparing a low-fat vegan diet to the standard American Diabetes Association diet in adults with type 2 diabetes. The 2006 paper (Barnard ND et al., Diabetes Care; 29: 1777-1783) and the 2009 follow-up (Am J Clin Nutr; 89: 1588S-1596S) report 74 weeks of follow-up in 99 adults randomised between the two interventions.

Findings: the vegan group had greater HbA1c reduction (mean change about -0.84 percentage points vs -0.40 for ADA diet over 22 weeks, expanding to -1.23 vs -0.38 at 74 weeks among medication-stable participants). Greater weight loss (-6.5 kg vs -3.1 kg at 22 weeks). Greater LDL reduction. More participants on the vegan arm reduced or discontinued diabetes medications. The trials used a fairly strict low-fat vegan protocol (under 10% energy from fat, mostly whole foods, low glycaemic index emphasis).

Subsequent trials (Wright 2017 BroadStudy in BMJ Open, Kahleova 2018 Nutr Diabetes) have replicated the directional finding with various plant-based protocols. The systematic reviews (Toumpanakis 2018 in BMJ Open Diabetes Research and Care; Viguiliouk 2019 meta-analysis) conclude vegetarian and vegan diets produce clinically meaningful improvements in HbA1c, fasting glucose, body weight, and lipids in adults with type 2 diabetes.

The mechanisms in more detail

Lower BMI. Vegans average BMI 22 to 24 vs 26 to 28 for omnivores in matched populations. Weight loss is the single largest predictor of insulin sensitivity improvement; 5 to 10% weight loss can produce 25 to 50% improvement in HOMA-IR.

Higher fibre intake. Typical vegan diets supply 35 to 50 g fibre per day, vs 15 to 20 g for typical Western omnivore. Fibre slows glucose absorption, feeds gut bacteria that produce short-chain fatty acids (butyrate, acetate, propionate) that improve insulin signalling, and reduces the postprandial glucose excursion. Soluble fibre from legumes, oats, and apples has the largest effect.

Lower intramyocellular lipid (IMCL). Fat deposited inside muscle cells impairs insulin signalling; this is one of the key mechanisms of insulin resistance. The Bach 2008 and Goff 2005 muscle biopsy studies have shown that plant-based diets reduce IMCL within weeks. The McMaster muscle physiology lab has separately demonstrated that the IMCL reduction precedes weight loss in time course, suggesting it is an independent dietary effect.

Lower advanced glycation end-products (AGEs). AGEs form on cooked animal-protein foods, particularly with browning and char. Dietary AGEs contribute to inflammation and may worsen insulin resistance. The Uribarri 2010 work measured serum AGE markers and found substantial reductions on plant-based diets compared to mixed Western diets.

What to eat for diabetes prevention or remission

Foundation foods. Legumes at every meal where possible: lentils, chickpeas, beans, split peas. The Sievenpiper 2019 meta-analysis found pulse-rich diets reduce HbA1c by approximately 0.3 percentage points over 6 weeks. Whole grains: oats, brown rice, quinoa, whole-grain bread, whole-grain pasta. Non-starchy vegetables: leafy greens, cruciferous vegetables, peppers, courgettes, mushrooms. Nuts and seeds for fat and protein density.

Foods to limit. Refined grain products (white bread, white rice, regular pasta); sugar-sweetened beverages including fruit juice; sweet desserts (vegan or otherwise); ultra-processed plant-based meat substitutes if they are high in sodium and refined ingredients; tropical oils used heavily (coconut oil, palm oil).

Fruit is fine. Whole fruit despite its sugar content does not raise diabetes risk and is associated with lower risk in cohort data. Fruit juice (which removes the fibre matrix) is different and is associated with higher risk. A piece of fruit is not a sugary drink.

Coordinating with medication

If you have diagnosed type 2 diabetes and start a vegan or WFPB diet, your glucose readings will likely drop quickly. People on insulin or sulfonylureas (gliclazide, glibenclamide) are at risk of hypoglycaemia if medication is not titrated down. Monitor glucose frequently in the first weeks of dietary change, and discuss with your diabetes team before starting; they may want to pre-emptively reduce sulfonylurea doses.

Metformin does not typically cause hypoglycaemia and usually does not need urgent adjustment, but doses may be reduced or discontinued over months as HbA1c falls. GLP-1 agonists (semaglutide, liraglutide) and SGLT-2 inhibitors (empagliflozin, dapagliflozin) interact with dietary patterns; discuss with your diabetes team. Statins are unaffected by dietary change in terms of safety but the LDL changes from diet may allow dose reduction over time.

Remission is real but supported. The DiRECT trial (Lean ME et al., Lancet 2018) showed that structured intensive dietary intervention can achieve type 2 diabetes remission in 36 to 46% of recently-diagnosed patients. Achieving remission requires sustained calorie and metabolic improvement; it is not a one-shot intervention. Work with a registered dietitian who knows both plant-based diets and diabetes management. See the WFPB page for the dietary framework, the cardiovascular page for parallel benefits.

Frequently asked questions about vegan vs vegetarian type 2 diabetes

How much lower is type 2 diabetes risk on a vegan diet?

The Adventist Health Study-2 cross-sectional analysis (Tonstad 2009, Diabetes Care) found vegans had 62% lower prevalence of type 2 diabetes than non-vegetarians, lacto-ovo vegetarians 39% lower, semi-vegetarians 26% lower. The prospective AHS-2 incidence analysis (Tonstad 2013, NMCD) confirmed the cross-sectional pattern: vegans had 49% lower incident diabetes over 2 years. The effect was partly but not fully mediated by lower BMI, suggesting both weight and diet composition contribute independently. The EPIC studies and other cohorts show consistent direction with varying effect sizes.

Can a vegan diet reverse type 2 diabetes?

Reverse is the operative word here. The Barnard Group's Physicians Committee for Responsible Medicine trials (PCRM 2006 in Diabetes Care, 2009 follow-up in Am J Clin Nutr) compared a low-fat vegan diet to the standard American Diabetes Association diet in adults with type 2 diabetes. Over 22 weeks, the vegan group had greater HbA1c reduction (about 1.2 percentage points vs 0.4 for ADA), greater weight loss, greater LDL reduction, and reduced or discontinued diabetes medications more frequently. Subsequent trials (Wright 2017, Kahleova 2018) have replicated the pattern. The DiRECT-style total diet replacement programmes show similar metabolic improvements with very low calorie diets; the vegan WFPB approach achieves comparable but more sustainable outcomes through whole-food substitution.

What is the mechanism for the diabetes benefit?

Multiple mechanisms. Lower body weight on vegan diets reduces insulin resistance directly. Higher fibre intake from legumes, vegetables, and whole grains slows glucose absorption and improves the gut microbiome's short-chain fatty acid output, which improves insulin sensitivity. Lower intramyocellular lipid (fat stored inside muscle cells) on vegan diets improves muscle glucose uptake; the McMaster muscle biopsy studies have demonstrated this directly. Reduced advanced glycation end-products from cooked animal foods may reduce inflammatory contributions to insulin resistance. Lower iron stores (vegans have lower ferritin) may also matter because high ferritin is itself a diabetes risk factor.

How fast does the diabetes effect appear after going vegan?

Insulin sensitivity improvements can show up within weeks; HbA1c (which reflects 3 months of average glucose) shows clear change at 12 weeks; clinical resolution criteria (HbA1c under 48 mmol/mol or 6.5%) for diabetes typically requires 3 to 6 months. The Barnard trials showed average HbA1c reduction of 1.2 percentage points at 22 weeks. Weight loss is the largest single predictor of remission magnitude; vegans typically lose 5 to 10% of body weight over the first 6 to 12 months of dietary transition, which is enough to drive substantial metabolic improvement. Insulin requirements often drop within weeks; medications may need active reduction under clinical guidance to avoid hypoglycaemia.

Are some plant foods better than others for blood glucose?

Yes. Legumes have the strongest evidence: lentils, chickpeas, beans, and split peas all have low glycaemic index and high fibre, and pulses-rich diets reliably improve HbA1c and fasting glucose in trials (Sievenpiper 2019 meta-analysis). Whole grains are better than refined for glucose control. Nuts (Mediterranean-diet trial data) lower cardiovascular risk and may modestly improve glucose. Berries, leafy greens, and cruciferous vegetables are high in polyphenols that may improve insulin sensitivity. Unhealthy plant foods (sugar-sweetened beverages, fruit juices, refined grain breads, vegan sweets) work against glucose control just as their omnivore counterparts do. The Satija healthy vs unhealthy plant-based distinction applies here too.

What about gestational diabetes?

The cohort evidence is suggestive but limited. The Chen 2017 EPIC-Oxford-style analysis found that vegetarian and vegan diets pre-conception were associated with lower gestational diabetes risk. The mechanism is plausibly similar to type 2 diabetes (lower BMI, higher fibre, better insulin sensitivity). Pregnant women with diagnosed gestational diabetes should follow their obstetric and dietetic teams' guidance, which may include carbohydrate counting and structured meal patterns; a vegan diet is compatible with all standard GDM management approaches but should be set up with a registered dietitian. See the pregnancy page.

Sources cited. Tonstad S et al. Type of vegetarian diet, body weight, and prevalence of type 2 diabetes, Diabetes Care 2009; 32: 791-796; Tonstad S et al. Vegetarian diets and incidence of diabetes in the Adventist Health Study-2, Nutr Metab Cardiovasc Dis 2013; 23: 292-299; Barnard ND et al. A low-fat vegan diet improves glycemic control and cardiovascular risk factors in a randomized clinical trial in individuals with type 2 diabetes, Diabetes Care 2006; 29: 1777-1783; Barnard ND et al. A low-fat vegan diet and a conventional diabetes diet in the treatment of type 2 diabetes: a randomized controlled, 74-wk clinical trial, Am J Clin Nutr 2009; 89: 1588S-1596S; Lean MEJ et al. Primary care-led weight management for remission of type 2 diabetes (DiRECT): an open-label, cluster-randomised trial, Lancet 2018; 391: 541-551; Viguiliouk E et al. Effect of vegetarian dietary patterns on cardiometabolic risk factors in diabetes: a systematic review and meta-analysis of randomized controlled trials, Clin Nutr 2019; 38: 1133-1145; Sievenpiper JL et al. Effect of non-oil-seed pulses on glycaemic control: a systematic review and meta-analysis of randomised controlled experimental trials, Diabetologia 2009; 52: 1479-1495. All values as of May 2026.