EPIC-Oxford 2007 + BMC Medicine 2020
Vegan vs Vegetarian Calcium: Closing the Fracture Risk Gap
The EPIC-Oxford cohort found that vegans break more bones than vegetarians or meat-eaters. The published difference disappears once vegan calcium intake exceeds about 525 mg per day. This page works through the data, the food sources, and what the published clinical guidance recommends.
What EPIC-Oxford actually found
EPIC-Oxford is the European Prospective Investigation into Cancer and Nutrition arm based at Oxford, recruiting over 65,000 UK participants from 1993 to 1999 with deliberate over-representation of vegetarians and vegans. The 2007 fracture paper (Appleby P et al., Eur J Clin Nutr 2007; 61: 1400-1406) reported on 1,555 incident fractures across about 5 years of follow-up. Vegans had a 30% higher rate of total fractures than meat-eaters and vegetarians, who were broadly similar to each other. The crucial subgroup analysis found that vegans consuming 525 mg of calcium per day or more had no excess fracture risk.
The 2020 follow-up by Tong and colleagues (BMC Medicine 2020; 18: 353) extended the analysis with longer follow-up and many more events, and confirmed the pattern. Vegans had higher rates of total fractures (around 43% higher), hip fractures (around 130% higher, or 2.3x), and leg fractures than meat-eaters. Lower BMI and lower calcium and protein intakes accounted for much of the excess risk in mediation analysis. After adjusting for BMI, calcium, and protein, the excess hip fracture risk for vegans was reduced but did not disappear entirely.
The honest reading: the fracture excess in vegans is real, but it is modifiable. The cohort included many people who became vegan in the 1980s when fortified plant milks and calcium-set tofu were not widely available. A modern vegan with reliable access to fortified soy milk and calcium-set tofu can clear the 525 mg threshold without effort. The fracture story is not a reason to avoid the diet, but it is a reason to take calcium intake seriously.
Plant calcium sources, ranked by bioavailability
| Food | Serving | Calcium (mg) | Bioavailability |
|---|---|---|---|
| Calcium-set tofu (firm) | 100 g | 350 | High (~31%) |
| Fortified soy milk | 250 ml | 300 | ~75% of cow milk |
| Fortified oat milk | 250 ml | 300 | ~75% of cow milk |
| Fortified almond milk | 250 ml | 300 | ~75% of cow milk |
| Fortified orange juice (US) | 250 ml | 300 | Similar to cow milk |
| Bok choy (cooked) | 1 cup (170 g) | 160 | High (~54%) |
| Collard greens (cooked) | 1 cup (190 g) | 270 | High (~52%) |
| Kale (cooked) | 1 cup (130 g) | 95 | High (~49%) |
| Watercress | 1 cup (35 g) | 40 | High |
| Almonds | 30 g | 75 | Moderate (~21%) |
| Tahini | 1 tbsp (15 g) | 64 | Moderate |
| Dried figs | 50 g | 80 | Moderate |
| Spinach (cooked) | 1 cup | 245 | Low (~5%, oxalate) |
| Swiss chard (cooked) | 1 cup | 100 | Low (~5%, oxalate) |
| Rhubarb (cooked) | 1 cup | 105 | Very low (oxalate) |
| Cow milk | 250 ml | 300 | Reference (~32%) |
| Cheddar cheese | 30 g | 200 | High (~32%) |
| Plain yogurt | 170 g | 200 | High (~32%) |
Bioavailability data from Weaver and Plawecki 1994 (Am J Clin Nutr 59: 1238S) and the Heaney calcium absorption studies. The low-oxalate brassicas (kale, bok choy, collards) absorb at higher efficiency than dairy on a per-mg basis, which is why they punch above their headline calcium number. Spinach is dense with calcium but most of it is not absorbed.
Calcium-set tofu and the label question
Tofu's calcium content depends entirely on the coagulant used. Tofu set with calcium sulfate (gypsum) contains around 350 mg calcium per 100 g. Tofu set with nigari (magnesium chloride) contains around 100 mg per 100 g. Tofu set with glucono-delta-lactone (silken tofu) contains very little calcium. The carton or pack will list the coagulant in the ingredients; look for calcium sulfate or E516.
UK supermarket firm tofu (Cauldron, Tofoo, Clearspring) is typically calcium-set. US supermarket firm tofu varies; Nasoya, House Foods, and Wildwood usually use calcium sulfate for firm and extra-firm grades. If calcium intake is a priority, choose accordingly. A 100 g portion at lunch and dinner from calcium-set tofu delivers 700 mg of calcium alone, which clears the UK SACN target by itself.
Cooking method does not meaningfully affect tofu calcium content. Pressing, baking, frying, and freezing all leave the calcium intact.
The vitamin K2 question, briefly
Vitamin K2 (menaquinone) activates osteocalcin, the protein that incorporates calcium into bone. K2 comes primarily from fermented foods (natto is the densest source, supplying about 1,000 mcg per 100 g) and from some animal products. Vegans on a low-natto diet may run low on K2, and there is published if not yet conclusive evidence (the Geleijnse 2004 Rotterdam Study and the Knapen et al. 2013 trials) that K2 supports bone mineral density and reduces fracture risk.
The Academy of Nutrition and Dietetics 2016 position paper does not yet make a K2 supplementation recommendation for vegans because the evidence base is still developing. The pragmatic position: if you eat sauerkraut, kimchi, miso, tempeh, or natto regularly, you are probably fine. If you do not, a K2 (MK-7) supplement at 90 to 180 mcg per day is low-risk insurance, particularly if your fracture-risk profile is otherwise elevated. Vitamin K1 (phylloquinone, from leafy greens) is also necessary but plays a different role; both forms matter, and most vegan diets are rich in K1.
Salt, caffeine, and the calcium-loss myths
High sodium intake increases urinary calcium loss; this is well established. The effect size in published balance studies is around 60 mg additional calcium lost per gram of sodium ingested above maintenance. A typical Western diet with 8 g of salt is therefore losing several hundred mg more calcium than a 6 g salt diet would. The largest single win for bone calcium retention on a plant-based diet (or any diet) is to keep added salt moderate. SACN recommends adults consume no more than 6 g of salt per day.
Caffeine's effect on calcium loss is real but small (about 4 mg additional calcium loss per 100 mg caffeine). At normal coffee intakes (2 to 3 cups per day, around 200 to 300 mg caffeine) this is not clinically meaningful as long as overall calcium intake is adequate.
The "high protein leaches calcium" claim is largely outdated. Older balance studies suggested high-protein diets increased urinary calcium loss, but subsequent isotope studies (Kerstetter and Insogna et al.) showed the loss was offset by increased intestinal calcium absorption driven by the protein-stimulated rise in IGF-1. Net calcium retention on adequate-protein diets is similar to or better than on low-protein diets. This is reassuring for vegan athletes and older vegans both groups need adequate protein for muscle and bone.
Related nutrient and outcome pages
Keep reading
Frequently asked questions about calcium
Do vegans break more bones than vegetarians or meat-eaters?
How much calcium do vegans and vegetarians need?
Is the calcium in plant milks the same as in cow milk?
What plant foods are highest in calcium?
Does dairy actually prevent fractures?
Do I need a calcium supplement on a vegan diet?
Sources cited. Appleby P et al. Comparative fracture risk in vegetarians and nonvegetarians in EPIC-Oxford, Eur J Clin Nutr 2007; 61: 1400-1406; Tong TYN et al. Vegetarian and vegan diets and risks of total and site-specific fractures: results from the prospective EPIC-Oxford study, BMC Med 2020; 18: 353; NIH ODS Calcium fact sheet; BDA Calcium food fact sheet; Weaver CM, Plawecki KL. Dietary calcium: adequacy of a vegetarian diet, Am J Clin Nutr 1994; 59: 1238S-1241S; Heaney RP. Calcium intake and disease prevention, Arq Bras Endocrinol Metab 2006; 50: 685-693; Knapen MH et al. Three-year low-dose menaquinone-7 supplementation helps decrease bone loss in healthy postmenopausal women, Osteoporos Int 2013; 24: 2499-2507. All values as of May 2026.