Vegan Diet

What it is

A vegan diet excludes all animal products: no meat, fish, eggs, dairy or honey [1]. Most versions are built on vegetables, fruit, pulses, whole grains, nuts and seeds [2]. Trials have tested low-fat vegan diets [3], vegan versions of the Mediterranean diet [4], and vegan diets delivered with classes or meal deliveries [5][6].

How it might work

Proposed, not established. Vegan diets tend to be lower in saturated fat and higher in fibre than typical Western diets, and reviewers suggest the weight difference comes partly from eating fewer calories and partly from a different mix of macronutrients [7]. In a 16-week trial, the vegan group's thermic effect of food - the energy used to digest a meal - rose by about 14% alongside a 5.9 kg weight loss, and insulin sensitivity improved [3]. These mechanisms are plausible but not established [7].

What the research says

Weight. The vegan-specific evidence pools 11 randomised trials of at least 12 weeks in 796 adults with overweight or type 2 diabetes: body weight fell by 4.1 kg more than in the control groups, and BMI by 1.38 kg/m2 [7]. A 2023 umbrella review that recalculated the meta-analyses put the general-population difference at 2.52 kg across 8 trials, rated moderate certainty [8]. Trials of vegetarian diets point the same way: 12 trials found 2.02 kg more weight loss overall and 2.52 kg in the vegan subgroup [9], and 20 trials in 1,878 people with or at high risk of heart disease found 3.4 kg more weight loss than comparison diets [10]. The largest single vegan trial followed 244 adults with overweight for 16 weeks; the low-fat vegan group lost 5.9 kg while the control group, asked to make no diet changes, lost none [3]. Among 22 pairs of identical twins, the twin on a vegan diet lost 1.9 kg more than the twin on a healthy omnivorous diet over 8 weeks [6].

Cholesterol. A 2023 meta-analysis of 30 randomised trials found that vegetarian and vegan diets lowered total cholesterol by 0.34 mmol/L and LDL ("bad") cholesterol by 0.30 mmol/L (about 12 mg/dL) compared with an omnivorous diet, with similar effects across age, continent, health status and study duration [11]. In people with overweight or type 2 diabetes, LDL fell by 0.24 mmol/L [7]. Cochrane's 2021 review of vegan diets found total cholesterol 0.24 mmol/L and LDL 0.22 mmol/L lower than with no or minimal intervention, rated moderate certainty [1], and the umbrella review found apolipoprotein B, another blood fat, lower across 7 trials [8]. In the twin study, LDL fell 13.9 mg/dL more in the vegan twins [6].

Blood sugar. In people with type 2 diabetes, vegetarian and vegan diets lowered HbA1c - a marker of average blood sugar over recent weeks - by about 0.2 to 0.4 percentage points across four meta-analyses, two of which rated the evidence moderate [12][13][10][7]. One review also found body mass index 0.96 kg/m2 lower and, in some trials, reduced diabetes medication [12].

Blood pressure. The picture is mixed. Cochrane found no clear effect on systolic or diastolic pressure, and a 2024 review of vegetarian diets found no significant difference [1][14]. Two older meta-analyses found small reductions of roughly 1 to 3 mmHg, and in people whose systolic pressure started at 130 mmHg or above, a vegan diet lowered it by about 4 mmHg [15][16].

Heart disease. No vegan trial has measured heart attacks, strokes or deaths [1]. Prospective cohort studies that followed 844,175 people link vegetarian diets to about 15% lower risk of cardiovascular disease and 21% lower risk of ischaemic heart disease; for vegans specifically, ischaemic heart disease was 18% lower, with a confidence interval that just included no difference [17]. An earlier analysis of 10 cohorts found 25% lower ischaemic heart disease risk with vegetarian diets [18], and in the Adventist Health Study 2 vegans had 15% lower all-cause mortality than non-vegetarians, with a confidence interval of 0.73 to 1.01 [19].

Other diets. When the comparison is another healthy diet, the advantage shrinks or disappears [1][5]. Cochrane found little or no difference in cholesterol between a vegan diet and other prescribed diets [1], and a 2-year trial in 159 African American adults found the same weight and cholesterol changes on a vegan diet and a low-fat omnivorous "soul food" diet [5]. One 16-week crossover trial in 62 adults went the other way, with 6.0 kg more weight loss and 15.3 mg/dL more LDL reduction on the vegan diet than on a Mediterranean diet, although blood pressure fell more on the Mediterranean diet [20].

Bones and nutrients. Observational studies raise a bone signal [21][22]. Pooling four cohorts of 529,672 people, vegans had a 75% higher rate of hip fracture than meat-eaters and vegetarians 25% higher, with very low certainty for the vegan estimate [21]. An earlier analysis of 20 studies found lower bone mineral density at the spine and hip in vegetarians and vegans, and higher fracture rates in vegans [22], and the umbrella review found 46% higher fracture risk (low certainty) [8]. On nutrients, a meta-analysis of 17 studies found lower serum vitamin B12 and higher homocysteine in vegans than omnivores, and supplements improved every marker [23]. A review of 141 studies found lower intakes or status of vitamin B12, vitamin D, iron, zinc, iodine and calcium in plant-based eaters, with vegans lowest for B12, calcium and iodine [2]. In one German study a third of vegans had iodine excretion below the WHO threshold for severe deficiency [24].

Doses used in studies

These are the diets and durations studied, not recommendations. The vegan weight meta-analysis included trials of at least 12 weeks [7]; the twin study ran for 8 weeks with meals delivered for the first 4 [6]. The largest single vegan trial ran 16 weeks [3], the five-diet comparison 6 months [25], the Mediterranean-versus-vegan crossover 16 weeks per diet [20], and the soul food trial 2 years [5]. Support varied widely, from weekly classes [5] to provided meals [6].

Safety and interactions

Side effects. In the trials read here, side effects were absent or minor [1].

Nutrients to plan. Without supplements or fortified foods, severe vitamin B12 deficiency can occur [26]. Vegan adults in a 17-study meta-analysis had lower serum B12 and higher homocysteine than omnivores, and B12 supplements improved every marker [23]; in a randomised trial, B12 markers fell after only four weeks on an unsupplemented vegan diet [27]. Reviews also found lower intakes or status of vitamin D, iodine, calcium, iron and zinc in vegans, with vegans lowest for vitamin B12, calcium and iodine [2]. One study found a third of vegans below the WHO threshold for severe iodine deficiency [24].

Upper limits. The studies read here did not test upper limits or toxic doses of the diet itself; the nutrients listed above are the ones that need cover [1][26].

Medicines. The trials read here did not test interactions with medicines [1]. In type 2 diabetes trials some participants reduced diabetes medication as blood sugar fell, with low-certainty evidence, so people on glucose-lowering medicines should ask a doctor before changing diet [12].

Pregnancy and breastfeeding. Nutrient needs are higher and deficiency risk is a particular concern in pregnancy, breastfeeding, infancy and childhood; a vegan diet at these stages should be planned with a health professional [26].

Bones and other conditions. Observational studies link vegan diets to lower bone density and higher fracture rates, so calcium and vitamin D intake and bone health deserve attention [22][21]. The studies read here did not report separate results for kidney disease or bleeding disorders.

Caveats and open questions

Most trials were small, short and unblinded, and several had funding ties: one vegan meta-analysis reports authors employed by or holding shares in food and drug companies, and the largest single vegan trial was run by authors whose organisation promotes plant-based diets [7][3]. Participants were mostly adults with overweight, type 2 diabetes or high heart-disease risk, so results may not apply to leaner or younger people [7][10]. The trials tested supported programmes - classes, follow-up, sometimes delivered meals - not simply choosing vegan food [5][6]. No trial has measured heart attacks, strokes or deaths, so whether the cholesterol and weight changes translate into fewer events is still open [1]. Whether the bone signal is caused by the diet or by other differences between vegans and meat-eaters is unresolved [21][22]. Ongoing research includes a 6,000-person cohort of people following plant-based diets (NCT06323538) and a trial of dietary patterns and LDL behaviour (NCT07474233).