Soybean Intake and Bone Density

Part III of a Series on Phytoestrogens

March 1999; Volume 1: 25-28

By Adriane Fugh-Berman, MD

Phytoestrogens from soy or flax have promise in treating hot flashes (see Alternative Therapies in Women’s Health, Premiere issue, pp. 1-4), but what effect do they have on bone? Conjugated estrogens help preserve bone density, but phytoestrogens are much weaker in binding assays than pharmaceutical or endogenous estrogens. There is some evidence that phytoestrogens have a beneficial effect on bone, and intriguing animal studies indicate that these effects may involve non-estrogenic mechanisms. There are few clinical trials, however, and long-term human studies are lacking.

Epidemiologically, bone density is lowest in Asian women and highest in African-American women, with white women in the middle. Interestingly, however, hip fracture rates are lower in Asians than in whites,1 despite the fact that Asians have thinner bones. Compared to white women, Asian women have a 40-50% lower risk of hip fracture, and African-Americans a 50-60% lower risk of hip fracture.2 The reasons for these differences are unknown, although body mass, differences in hip axis length (distance from the greater trochanter to the inner pelvic brim), and bone quality may all play a role. Fracture risk is also influenced by muscle function; however, this has not been quantified among ethnic groups in the same level of detail as bone density. It is possible that high soy intake contributes to improved bone quality in Asians compared to whites, but this is entirely speculative, since there are many differences between these two groups. It bears noting that bone density measurements are only one indicator of fragility; disordered trabecular microarchitecture probably contributes to fracture risk independently of bone mass.3

Clinical Trials

The most thorough human study to date is a double-blind trial in 66 postmenopausal, hypercholesterolemic women, aged 49-73.4 Subjects were randomly assigned to 40 g protein a day from one of three sources: milk (nonfat dried milk and casein), soy protein with medium isoflavone content (equivalent to 55.6 mg isoflavones daily), or isolated soy protein with high isoflavone content (equivalent to 90 mg isoflavones daily). All women also followed a low-fat, low-cholesterol diet. Dual-energy X-ray absorptiometry (DEXA) bone density studies of the lumbar spine, proximal femur, and total body were done at the beginning of the six-month study (after a two-week lead-in) and at the end of the study. (See Table for isoflavone content of soyfoods.)

    Estimated Isoflavone Content of Soyfoods                                                      
    Serving Size
    Isoflavone mg/serving*
    Mature soybeans, uncooked
    1/2 cup
    Roasted soybeans
    1/2 cup
    Green soybeans, uncooked
    1/2 cup
    Tempeh, uncooked
    4 oz
    Soy isolate, dry
    1 oz
    Soy flour
    1/4 cup
    Tofu, uncooked
    4 oz
    Textured soy protein, dry
    1/4 cup
    Soy milk
    1 cup
    Soy concentrate, dry
    1 oz

    *Values obtained from published literature and analyses of selected products. Isoflavone content varies widely among soybean varieties and from product to product based on manufacturing process and sources of soy protein. 

    Source: Indiana Soybean Board

No differences were seen among the three groups in bone density studies of the hip or total body, but subjects receiving the high isoflavone preparation experienced a significant increase in lumbar bone density and mineral content (2%), compared to the milk protein group. This is consistent with an estrogenic effect; conjugated estrogens have a more pronounced effect on trabecular than on cortical bone, and thus affect bone density more in vertebrae than in femoral bone density. Six months is relatively brief for an osteoporosis study. A greater effect may have been observed with a longer study.

An extremely problematic one-year pilot study in eight white women with low bone mass (mean age 56) has been reported in abstract form.5 This study enrolled subjects in a six-day course emphasizing lifestyle, exercise, intake of whole-grain products, and sufficient calcium, magnesium, and vitamin D intake. Women also received a vitamin and mineral supplement containing calcium, magnesium, silica, zinc, boron, copper, vitamin C, pyridoxine, vitamin D, and vitamin K, and apparently received or were encouraged to drink calcium-enriched soy milk in unspecified amounts. However, six women were concurrently taking HRT or bisphosphonate drugs, and two participants were allergic to soy and were not included in the final analysis. Although researchers noted that five of the six women demonstrated a gain in BMD/BMC (the sixth subject discontinued HRT during the study), DEXA measurements of bone mass were not significantly different from baseline at the end of the study. It is not noted how many of the six subjects analyzed were using HRT or bisphosphonate drugs.

This is an excellent example of how not to do a study. No effort seems to have been made to screen subjects to assure even a minimally homogenous study population. Exclusion criteria obviously should include allergy to the compound to be tested. Allowing the concurrent use of drugs known to affect the endpoint of bone density in a small study is unjustifiable. And the multimodal, individualized approach would have rendered any positive results uninterpretable. It is interesting, however, that even with this panoply of pharmaceutical and non-pharmaceutical interventions, there were no positive results.

Animal Studies

Both soybean intake and genistein (the predominant isoflavone phytoestrogen in soybeans) have been tested on bone density in the oophorectomized rat model. A rat study noted that genistein protects against loss of trabecular bone volume and bone mineral density and appears to increase osteoblastic activity without affecting osteoclasts.6 As the mechanism of estrogen appears to be suppressing bone turnover (i.e., osteoclastic activity) rather than stimulating bone growth, the effect of genistein on bone may not be an estrogen-like mechanism.

One study compared four groups: oophorectomized rats, sham-operated rats, oophorectomized rats fed a milk protein diet with 17-beta-estradiol, and oophorectomized rats fed soybean protein isolate.7 Oophorectomized rats showed significant decreases in bone density; both 17-beta-estradiol and soybean protein isolate protected against this bone loss.

Another study in oophorectomized rats compared sham-operated rats with three oophorectomized groups; two were fed casein- (milk protein-) based diets and one group was fed a 22% soybean diet.8 One of the casein-fed groups received estrogen injections. Rats in the sham-operated, soy-fed, and estrogen-treated groups had significantly higher femur and tibia ash contents than casein-fed rats. Markers of bone turnover were not affected.

Two studies compare oophorectomized rats fed casein, reduced-isoflavone soy protein, or normal isoflavone soy protein to oophorectomized or sham-operated rats fed a casein-based diet. The first study9 found that femoral density did not differ between either soy group and oophorectomized controls (sham-operated controls, as expected, had higher bone mass). The second study10 found increased femoral bone density with isoflavone-intact soy but no effect of isoflavone-reduced soy. Measures of bone turnover were not affected by any of the soy diets in either study. Both of these studies, reported by the same group, found differing results on femoral bone density but similar findings on lack of effect on bone turnover rates (which are accelerated by oophorectomy).

Another study in which rats were injected subcutaneously with genistein (5 mcg/g body weight) found less bone loss in the genistein-treated group and higher rate of bone formation, again without any effect on bone resorption.11

Genistein may have a biphasic effect. In a model of extreme bone loss, it is more effective at lower rather than higher doses. In a total assault on bone mineral stores, lactating oophorectomized rats are fed a low-calcium diet; rats so treated will lose more than 50% of bone mineral mass in two weeks. Three different doses of genistein (0.5 mg, 1.6 mg, or 5.0 mg daily) were compared to conjugated estrogen in this model,12 endpoints were scanning electron microscopy (SEM) measurements of the proximal tibia, femoral ash weights, and uterine weights. Mean femoral ash weights were significantly higher in the low-dose genistein group than the conjugated estrogen group, and higher than ash weight of those in the intermediate or high genistein group. SEM results showed low-dose genistein to be equivalent to conjugated estrogens in the retention of cancellous bone tissue. However, high doses of genistein were less effective.

A primate study found different results than the rat studies. Forty-one oophorectomized cynomolgus monkeys received one of four treatments: a milk protein diet with or without 17-beta-estradiol or a soy protein based diet with or without 17-beta-estradiol.13 Histomorphometry on mid-femur cortical bone showed that soy protein did not prevent oophorectomy-induced increase in bone turnover and actually increased bone turnover on the endosteal (inner lining of the bone) surface compared to milk-fed monkeys. On the other hand,17-beta-estradiol suppressed oophorectomy-induced increase in bone turnover.


There is no definitive evidence that dietary soy or isoflavone supplementation benefits bone, but one reasonable human study shows a modest effect of high-isoflavone soy supplementation on vertebral (not femoral) bone. There is enough intriguing animal evidence to recommend further clinical research.

Although most rat studies show beneficial effects of soy or genistein on bone, a primate study did not confirm these results. Results from animal studies are not entirely consistent in terms of preservation of bone density, but a consistent finding is that soy, unlike estrogen, does not reduce oophorectomy-induced bone turnover. Any beneficial effect of soy appears to be due to stimulating bone formation rather than reducing bone resorption.

Additional animal studies should be done to elucidate mechanisms of action. Long-term human studies comparing the effect of HRT or bisphosphonates with soy supplementation on fracture incidence should be done to determine whether phytoestrogens protect bone.


1. Lauderdale DS, et al. Hip fracture incidence among elderly Asian-American populations. Am J Epidemiol 1997;146:502-509.

2. Cummings SR, et al. Racial differences in hip axis length might explain racial differences in rates of hip fracture. Study of Osteoporotic Fractures Research Group. Osteoporos Int 1994;4:226-229.

3. Cooper C. The epidemiology of fragility fractures: Is there a role for bone quality? Calcif Tissue Int 1993;53(suppl):23-26.

4. Potter SM, et al. Soy protein and isoflavones: Their effects on blood lipids and bone density in postmenopausal women. Am J Clin Nutr 1998;68(suppl):1375-1379.

5. Olsen EL, et al. Bone gain after calcium enriched soy milk, food supplement, and lifestyle changes in women with low bone mass: A pilot project in course form. Am J Clin Nutr 1998;68(suppl):1518(abstract).

6. Fanti O, et al. Systematic administration of genistein partially prevents bone loss in ovariectomized rats in a nonestrogen-like mechanism. Am J Clin Nutr 1998;68(suppl):1517(abstract).

7. Arjmandi BH, et al. Dietary soybean protein prevents bone loss in an ovariectomized rat model of osteoporosis. J Nutr 1996;126:161-167.

8. Harrison E, et al. The effect of soybean protein on bone loss in a rat model of postmenopausal osteoporosis. J Nutr Sci Vitaminol (Tokyo) 1998;44:257-268.

9. Arjmandi BH, et al. Role of soy protein with normal or reduced isoflavone content in reversing bone loss induced by ovarian hormone deficiency in rats. Am J Clin Nutr 1998;68(suppl):1358-1363.

10. Arjmandi BH, et al. Bone-sparing effect of soy protein in ovarian hormone-deficient rats is related to its isoflavone content. Am J Clin Nutr 1998;68(suppl):1364-1368.

11. Fanti P, et al. The phytoestrogen genistein reduces bone loss in short-term ovariectomized rats. Osteoporos Int 1998;8:274-281.

12. Anderson JJ, et al. Biphasic effects of genistein on bone tissue in the ovariectomized, lactating rat model. Proc Soc Exp Biol Med 1998;217:345-350.

13. Lees CJ, Ginn TA. Soy protein isolate diet does not prevent increased cortical bone turnover in ovariectomized macaques. Calcif Tissue Int 1998;62:557-558.