Giessen Raw Food Study + Academy of Nutrition position 2016

Raw Vegan Diet: Benefits, Risks, and the B12 Problem

Raw veganism is the strictest mainstream plant-based diet, limiting cooking to around 40 to 46 degrees Celsius. The published evidence base is small but consistent: weight loss, low B12, low HDL, and meaningful fertility risk in women. This page presents the diet honestly, including where its theoretical underpinnings hold and where they do not.

The short answer. Raw veganism is plant-only food with no cooking above 40 to 46 C. The largest published cohort (Giessen Raw Food Study) found low BMI, B12 deficiency, low HDL, and 30% amenorrhoea in women. The Academy of Nutrition and Dietetics does not list raw veganism among diets appropriate for all life stages. As a short-term elimination diet in specific clinical contexts (some RA trials) raw food has evidence; as a whole-life diet it carries documented risks not present in cooked vegan or vegetarian diets.

The theoretical case for raw food

The raw food movement traces its modern roots to Maximilian Bircher-Benner's Zurich sanatorium in the early 20th century and Ann Wigmore's Hippocrates Health Institute work in the 1960s. The two central theoretical arguments are enzyme preservation (cooking denatures food enzymes which the proponents claim are useful for human digestion) and reduction of cooking byproducts (advanced glycation end-products, acrylamide, heterocyclic amines).

The enzyme argument is weak. Human digestion proceeds with our own pancreatic and intestinal enzymes; dietary plant enzymes are denatured by stomach acid (pH 1 to 2) and are not significant contributors to nutrient breakdown. Whatever survival of dietary enzymes into the small intestine exists, the evidence does not support a meaningful contribution to absorption.

The cooking-byproducts argument has some validity. Acrylamide is formed when starchy foods are heated above 120 C (browning of potatoes, bread); high intake has been linked to cancer risk in animal models, less clearly in humans. Heterocyclic amines and polycyclic aromatic hydrocarbons form on charred meat and to a lesser extent on charred plant foods. Advanced glycation end-products (AGEs) form during high-heat cooking of protein-rich foods and have been associated with inflammation. A moderate-heat plant-based diet (steaming, boiling, low-temperature roasting under 180 C) avoids most of these byproducts without going fully raw.

The Giessen Raw Food Study, in detail

The Giessen Raw Food Study (Koebnick C, Strassner C, Hoffmann I, Leitzmann C; published primarily in Annals of Nutrition and Metabolism 1999 and follow-up papers through 2005) recruited 297 long-term raw food eaters in Germany and Switzerland between 1996 and 1998. Average duration on the diet was 3.7 years. The sample comprised four sub-groups by strictness: people eating 70 to 100% raw, by percentage of food.

Findings: substantial weight loss with mean BMI of 20.1 for women and 21.4 for men, and about a third of participants in the underweight range (BMI under 18.5). Serum vitamin B12 was below the deficiency threshold (118 pmol/L) in 38% of participants. HDL cholesterol was low. Amenorrhoea (absence of menstrual cycles) affected approximately 30% of women under 40, compared to under 5% in age-matched general population. The amenorrhoea was strongly associated with low body weight rather than the raw diet per se, but raw veganism was the proximate cause of the underweight.

The strictly raw subset (95 to 100% raw) had the most consistent deficiencies. The 70 to 79% raw subset (which included some cooked food) had nutrient profiles much closer to standard vegan eaters. The take-away from the cohort is not that raw food is uniquely toxic, but that strict raw veganism is hard to sustain at adequate energy and nutrient intake.

Where cooking helps and where raw helps

Nutrient or compound	Effect of cooking	Implication
Vitamin C	Lost (30 to 60% from boiling)	Eat some raw fruit and vegetables for vitamin C
Folate	Lost (30 to 50% from boiling)	Raw or steamed leafy greens preserve more
Lycopene (tomatoes)	Bioavailability rises 2 to 3x	Cooked tomatoes are better for lycopene
Beta-carotene (carrots)	Bioavailability rises 3 to 4x	Cooked carrots are better source
Sulforaphane precursors	Brief blanch preserves; long boil destroys	Light steam or brief blanch of broccoli
Protein digestibility (legumes)	Rises from 60 to 70% raw to 85 to 95% cooked	Cook beans for protein utilisation
Phytate	Reduced 20 to 40% by cooking	Cooked grains and legumes have better mineral availability
Lectins, antinutrients in beans	Destroyed by boiling	Raw or undercooked beans cause GI distress
Polyphenols and antioxidants	Mixed; some lost, some bioavailability up	Variety matters more than raw vs cooked
Acrylamide	Forms above 120 C in starchy food	Avoid charring; moderate-heat cooking is fine

The optimal eating pattern from a pure nutrient-availability perspective is a mix: some raw (leafy greens, fruit, salads, sprouts) and some lightly cooked (steamed broccoli, boiled potatoes, simmered legumes). A diet that is exclusively one or the other loses ground in different places.

Clinical trials where raw food has shown benefit

The strongest published clinical data for raw or near-raw food is in rheumatoid arthritis. The Kjeldsen-Kragh trial (Lancet 1991) randomised RA patients to a one-year programme starting with seven days of fasting, then four months of vegan diet followed by lacto-vegetarian transition. The intervention group showed sustained improvement in joint tenderness, morning stiffness, and several inflammatory markers compared to controls. Subsequent trials by Hanninen and others using uncooked vegan diets in RA showed similar short-term improvements with high attrition rates.

The mechanism is thought to involve gut microbiome shifts away from inflammatory profiles, reduction in food-derived inflammatory triggers, and elimination of cooked-protein AGEs. The improvements in these trials are real but most patients return to standard diet after the trial because compliance is hard. As a treatment trial used under medical supervision, raw food has a place. As a long-term whole-life diet for healthy adults, the evidence does not support raw food over a varied cooked vegan diet.

If you decide to try it, do it safely

Despite the risks, raw food can be done safely for limited periods with care. Essential precautions: supplement B12 at 25 to 50 mcg daily; monitor body weight weekly and stop if you fall below BMI 18.5; ensure adequate calorie intake (raw diets are often inadvertently low-calorie because the food is bulky and filling); include sprouted legumes (sprouting partially activates antinutrient-reducing enzymes) and soaked nuts and seeds; supplement vitamin D and iodine; use a basic blood panel (B12, ferritin, vitamin D, lipid profile, full blood count) at 3 and 6 months.

Do not attempt raw veganism during pregnancy, lactation, childhood, adolescence, or if you are over 65 and have any baseline risk of sarcopenia. Stop if you develop amenorrhoea, frequent illness, persistent fatigue, hair thinning, or unintentional weight loss below your healthy range. Reintroduce cooked legumes and grains if you do.

This page is not advocacy. The Academy of Nutrition and Dietetics 2016 position paper lists well-planned vegan and lacto-ovo vegetarian diets as appropriate for all life stages; it does not extend that endorsement to raw veganism. The intent of this page is to give an honest picture of what the published evidence says, not to recommend the diet. If you are considering raw food, discuss with a registered dietitian, not just an internet community. See getting started for safer entry points to plant-based eating.

Related diet-type pages

Keep reading

Frequently asked questions about raw veganism

What is a raw vegan diet?

A raw vegan diet excludes all animal products (the vegan part) and limits cooking temperature to around 40 to 46 degrees Celsius (104 to 115 F), on the theory that this preserves naturally occurring enzymes and reduces formation of advanced glycation end-products. Common foods include fresh fruit and vegetables, sprouted grains and legumes, nuts and seeds (often soaked), cold-pressed oils, and dehydrated raw crackers and bars. Excluded foods include all cooked grains and legumes, baked goods, pasteurised products, and most fermented foods that involve heat.

Is the enzyme preservation argument scientifically valid?

Not in the way it is usually presented. Plant enzymes are denatured by stomach acid during digestion regardless of whether the food is raw or cooked, so the dietary enzyme contribution to human digestion is negligible. The human body produces its own digestive enzymes (amylase, lipase, proteases) in sufficient quantity for normal diet composition. The valid scientific point about raw foods is that cooking destroys some thermolabile vitamins (vitamin C, folate) to varying degrees, while improving the bioavailability of others (lycopene in tomatoes, beta-carotene in carrots, sulforaphane precursors in broccoli). The optimal diet contains a mix of raw and cooked foods.

What does the Giessen Raw Food Study say?

The Giessen Raw Food Study (Koebnick C et al., Ann Nutr Metab 1999 and follow-up papers) recruited 297 long-term raw food eaters in Germany and Switzerland. The published findings: substantial weight loss (mean BMI 20.1 for women and 21.4 for men, with about a third in the underweight range), low serum vitamin B12 with most participants below the deficiency threshold, low HDL cholesterol, and notably high rates of amenorrhoea in women (around 30%, compared to under 5% in age-matched controls). The study is the largest systematic look at long-term raw vegan nutrition and the picture is not flattering, particularly for women of reproductive age. The lead authors did not recommend the diet on the basis of their findings.

Can a raw vegan diet provide enough protein?

It is harder than on a standard vegan diet. Protein digestibility from raw legumes is around 60 to 70%, compared to 85 to 95% from cooked legumes (Bressani 1973, Singh 1988, and subsequent work). Raw beans also contain lectins and phytohaemagglutinins that cause GI distress and reduce nutrient absorption; this is why most legumes need cooking before consumption. Raw vegans rely on sprouting (which partially activates phytase and reduces antinutrients) and on nuts and seeds for protein. Hitting 60 to 80 g of utilisable protein per day on a raw vegan diet is possible with deliberate planning but takes more food volume than a cooked vegan diet at the same protein target.

What are the documented health risks of long-term raw veganism?

Published risks from the Giessen study and other cohorts: low body weight and BMI, B12 deficiency (unless deliberately supplemented), low HDL cholesterol, amenorrhoea and fertility issues in women, dental erosion (particularly with high citrus and acidic-fruit intake), and inadequate energy intake leading to fatigue. The diet is not recommended during pregnancy, lactation, childhood, adolescence, or for older adults at risk of sarcopenia. The Academy of Nutrition and Dietetics does not list raw veganism as an appropriate diet for all life stages, unlike well-planned standard vegan or vegetarian diets.

Are there any documented benefits of raw veganism?

Yes, with caveats. The Giessen cohort had lower serum cholesterol and lower blood pressure than meat-eating controls, although these effects are also seen on standard vegan or whole-food plant-based diets. Some clinical trials in rheumatoid arthritis have shown symptomatic improvement with raw-vegan or near-raw-vegan diets over weeks (Kjeldsen-Kragh 1991, Hanninen 2000), with effects attributed to gut microbiome shifts and reduced inflammatory food intake. The benefits, where they appear, are not unique to raw food and are also obtainable from a well-planned cooked vegan diet without the deficiency risks. Raw food has a legitimate place as a short-term elimination diet in specific clinical contexts, but the long-term whole-life version carries meaningful downsides.

Sources cited. Koebnick C, Strassner C, Hoffmann I, Leitzmann C. Consequences of a long-term raw food diet on body weight and menstruation: results of a questionnaire survey, Ann Nutr Metab 1999; 43: 69-79; Koebnick C et al. Long-term consumption of a raw food diet is associated with favorable serum LDL cholesterol and triglycerides but also with elevated plasma homocysteine and low serum HDL cholesterol in humans, J Nutr 2005; 135: 2372-2378; Kjeldsen-Kragh J et al. Controlled trial of fasting and one-year vegetarian diet in rheumatoid arthritis, Lancet 1991; 338: 899-902; Hanninen O et al. Antioxidants in vegan diet and rheumatic disorders, Toxicology 2000; 155: 45-53; Melina V, Craig W, Levin S. Position of the Academy of Nutrition and Dietetics: vegetarian diets, J Acad Nutr Diet 2016; 116: 1970-1980; Bressani R. Antinutritional factors in beans and other legumes, J Am Oil Chem Soc 1973 and subsequent work. All values as of May 2026.