Soy – A Review of the Literature

ALPHARAJHealth, Nutrition, SupplementsLeave a Comment

Vegan Nutrition

Review of Human Estrogens

Our bodies make several types of estrogens. You can see the synthesis below; the estrogens are highlighted in the pink triangle.  Synthesis starts with cholesterol.


Estrone, estriol, and estradiol (E2) are the three highlighted estrogens produced endogenously (inside our bodies). Estrogens are the ligands for estrogen receptors, meaning estrogens bind to estrogen receptors and have effects on receptor function. There are two forms of estrogen receptors – ER Alpha (ERα) and ER Beta (ERβ). The affinity of E2 for ERα is 0.05 nM, whereas the affinity of E2 for ERβ is 0.09 nM.

Determining the Affinity of a Ligand for a Receptor

To determine how well a ligand binds to its receptor, we can measure the affinity. One way to measure the affinity is to use radiolabeled compounds and unlabeled compounds together, and you can measure the amount of radiolabeled compound displaced from the receptor by unlabeled compounds. For example, a compound with a 5 nM affinity binds the receptor more than a compound with a 50 nM affinity. In an experiment, we could take the 5 nM compound and label it with a radioactive isotope. We mix the radiolabeled compound with the receptors in a test tube, then later we add unlabeled compound at increasing concentrations and determine how much radiolabeled compound remains bound to the receptors. We can plot this data on a graph (Scatchard plot) and determine the affinity. The lower affinity compound (our 50 nM compound) requires more compound to be added to the test tube to displace the radiolabeled compound (our 5 nM compound). In our example, the 5 nM compound binds 10 times more than 50 nM compound.


Does Soy Have Estrogens?

Yes, soy does have compounds with estrogen-like structures [1]. The two most researched compounds are the isoflavones genistein and daidzein.


From this data, we can compare the binding affinities of these compounds for the ERα and ERβ. IC50 is the concentration where the binding of the radiolabeled compound is reduced by 50% due to the competing compound. It takes ~1 nM of unlabeled E2 and 85 nM to 400 nM of unlabeled isoflavones to displace half of the radiolabeled E2.  This means the isoflavones are much, much weaker binders than E2. If E2 is set to 100%, genistein binds to ERα at 0.7% and to ERβ at 13%, whereas daidzein binds to ERα at 0.2% and to ERβ at 1%.


In the same paper, the ability of the phytoestrogens to stimulate ER-dependent transcription was examined. Each estrogen receptor was added to cells, the compounds were added at 1000 nM, and the ability of the added receptors to activate a reporter gene was measured. This dose was chosen because consumption of a soy-containing meal can produce this concentration in serum in studies cited. This system is helpful for identifying compounds that can produce a response, but it is limited by its cell culture nature. Your endogenous estrogens will be in circulation and will be able to compete against phytoestrogens for binding the receptor, so the phytoestrogens may not produce such a robust response in your body.


Genistein and daidzein bind to ERβ much more than to ERα. When someone is diagnosed with ER+ breast cancer, the ERα is the receptor being measured. It is largely ERα that drives estrogen-dependent growth, whereas ERβ tends to oppose ERα action and inhibit growth [2].

Does Soy Affect Sex Hormones?

The only case report of a man growing breasts while consuming soy involves a 60-year-old man consuming 3 quarts of soy milk a day and this went away when he stopped consuming soy milk [3].

Many people claim that soy product consumption will alter sex hormones in men. In several studies examining soy protein or isoflavone supplementation, men did not have significant changes in testosterone, free testosterone, estrogen, sex hormone binding globulin protein, or semen quality [4-12]. A decrease in prostate specific antigen (PSA), which could be beneficial for prostate cancer, was seen in men consuming soy supplements [9, 10]. Interestingly, one study showed no changes in T levels, but decreased androgen receptor (AR) expression was observed; decreased AR could be beneficial in hormone-sensitive prostate cancer growth [8]. Another surprising finding was that soy milk consumption was associated with reduced circulating estrone concentrations [13]; estrone’s affinity for estrogen receptors isolated from rat uteri is ~7% of the E2 binding [14].

In women, studies on soy are divided by the women’s menopausal status. A review noted that pre-menopausal women do not have any changes in estrogen, but changes were noticed in LH and FSH (these hormones influence menstrual cycle progression), whereas post-menopausal women had small, but not significant, increases in E2 [15]. Further studies are needed to confirm if increases occur and are significant (not due to chance). A small study in a metabolic ward showed decreased ovarian hormones, but not LH and FSH [16]. They noted, “Decreases in ovarian hormones are related to isoflavones contained in soy and also to energy intake and other components such as protein and fiber but not fat.” The soy group ate less protein and more carbs than they normally ate at home. Hot flashes are reduced in perimenopausal and post-menopausal women consuming soy [17, 18].

Is Soy an Appropriate Protein Source for Training?

In a study examining soy concentrate and soy isolate over 12 weeks in trained individuals, serum testosterone did not decrease and changes in lean body mass were not inhibited by either soy protein supplement [19]. In a rat study, mTOR was phosphorylated by soy protein isolate, but not as much as whey protein [20]. The diets were matched for total protein content, but not for leucine content (10.2 g leucine in 100 g whey, 8.2 g leucine in 100 g soy protein). Approximately 0.05 g leucine per kg body weight is needed to stimulate muscle protein synthesis (MPS), and leucine is considered of very high importance in MPS [21]. A human study showed that soy could not induce mTOR phosphorylation or MPS [22], but again, the study diets were not controlled for leucine content (9.1% leucine and 20.4% BCAA in dairy, 7.8% leucine and 17.6% BCAA in soy).

In researching the use of soy protein as a sport supplement, papers from Dr. Stu Phillips comparing milk to soy were analyzed [23-25]. Strength gains were similar between milk and soy groups, but small but non-significant increases were seen in hypertrophy with milk. In corresponding with Dr. Phillips, he noted that his studies were also controlled for total protein, not leucine content. Our email exchange is below with further comments included.

[VBBN] Were the milk and the soy given in doses that equalized the leucine content?

[Stu] No. Since leucine’s not a label ingredient (and protein is) and protein requirements are given, but Leu requirements are rarely cited (and are within protein requirements) when we planned that study, which is now quite old, we equated protein. Milk beats soy hands down – see Wilkinson et al [26]

[Comments] Both milk and soy produced a positive net protein balance, but the milk effect was greater in overall net balance and borderline significant in the fractional synthesis rate. The conclusions state either protein source coupled with resistance exercise will promote muscle mass maintenance and gains but that milk *may* promote a more rapid rate of gain.

[VBBN] Whey protein powder has slightly more leucine than soy protein isolate from what I’ve seen online, ~3 g Leu in 30 g whey vs ~2 g Leu in 30 g soy isolate.

[Stu] It depends on the whey and soy isolates (actually we can obtain a whey fraction what has – per 30g – almost 4g of Leu), but yes soy isolate generally has less Leu than whey. We did a study where we matched whey and soy for EAA content (higher Leu in whey) and whey, perhaps not surprisingly, was superior – see Tang et al [27]

[Comments] Tang study lists leucine quantities: 2.3 g in whey, 1.8 g in soy.

“We conclude that the feeding-induced simulation of MPS in young men is greater after whey hydrolysate or soy protein consumption than casein both at rest and after resistance exercise; moreover, despite both being fast proteins, whey hydrolysate stimulated MPS to a greater degree than soy after resistance exercise. These differences may be related to how quickly the proteins are digested (i.e., fast vs. slow) or possibly to small differences in leucine content of each protein.”

[VBBN] In your studies, milk was concluded to be more beneficial for trainees, so I was curious if the leucine content was controlled for in the study design.

[Stu] No and it really doesn’t make sense to do since then – using your example above – you’d have to feed people 33% more protein (and more total energy) to get a comparable Leu content. What we compared in the Wilkinson paper [26] was milk to a soy beverage designed to be ‘like’ milk. Thus, to equate leucine the total energy would have to be greater. Others who have studied whey vs. soy find soy to not even be as good as carbohydrate – see attached Volek et al. [28]

[Comments] Again, leucine appears to be a big driver of the whey response. Even though whey and soy total protein amounts were the same at 1.4 g/kg, leucine concentrations while fasting and after exercise were greater in the whey group.

“Lean body mass gains were significantly (p < 0.05) greater in whey (3.3 ± 1.5 kg) than carb (2.3 ± 1.7 kg) and soy (1.8 ± 1.6 kg). Fat mass decreased slightly but there were no differences between groups. Fasting concentrations of leucine were significantly elevated (20%) and postexercise plasma leucine increased more than 2-fold in whey. Fasting leucine concentrations were positively correlated with lean body mass responses.”

[VBBN] Studies comparing pea or rice to whey did dose the plant proteins to control for leucine content and showed similar results between the proteins.

[Stu] I can’t think of a single soy versus whey Leu-matched study (do you know of one?), but you’re right they would likely be the same if you matched Leu. Pea and rice are inferior in terms of DIAAS so they are well below whey, but yes so long as you eat enough of the protein (to get enough Leu – and more energy) you get ‘equal’ responses. However, as per the above you’re always eating more protein (and thus total energy) to equate Leu – we made this point recently in a review – see Devries et al [29]. In my view, so long as Leu drives the response then you can always plan a study to show equivalent of proteins if you feed enough of one protein, but there’s a reason all of those proteins are always compared to whey: because it’s the best and most efficient protein around!

[Comments] As a vegan, whey is off-limits since it comes from dairy. It appears that vegans can get what they need in terms of maximizing muscle protein synthesis if they eat enough plant protein to hit the leucine threshold (~3 g). The pea and rice studies both show similar gains as whey protein, and these studies did consider meeting the leucine threshold for the treatment groups [30, 31]. The concern over eating more protein, and therefore more calories, is minor. Feeding studies in trained people showed protein at 4.4 g/kg body weight (2 g/lb) in a hyper-caloric diet (where excess cals came from protein) did not produce weight gain [32].

Comparison of Leucine Content Between Pea, Rice, and Soy Isolates

The pea study [30] used 25 g pea isolate twice per day, noting that the pea isolate had 6.4 g leucine per 100 g and the whey isolate had 8.6 g leucine per 100 g. Each 25 g dose of pea isolate provided 1.6 g leucine for a daily total of 3.2 g leucine, whereas the 25 g dose of whey isolate provided 2.15 g leucine for a daily total of 4.3 g leucine.

The rice study [31] used 48 g protein isolate once per day, noting that the rice isolate had 8 g leucine per 100 g and the whey isolate had 11.5 g leucine per 100 g. The 48 g portion of rice protein isolate provided 3.84 g leucine, whereas the 48 g portion of whey protein isolate provided 5.52 g leucine. The authors noted quantities above 2-3 g leucine do not provide additional benefit to stimulating MPS.

Given these data on plant proteins and leucine content, we can calculate the doses needed to hit the leucine threshold using soy protein products. In a paper by Dr. Douglas Kalman, he measured the leucine content of soy protein isolate and soy protein concentrate (as well as brown rice isolate and concentrate) [33]. Soy isolate contained 6.783 g leucine per 100 g, whereas soy concentrate contained 4.917 g leucine per 100 g. To hit the leucine threshold of 3 g, the soy isolate dose would be 44 g and the soy concentrate dose would be 61 g. Most protein scoops are ~30 g, so you could use 1.5 scoops soy isolate and 2 scoops soy concentrate to hit the leucine threshold.

Side note: Hemp protein has 2.3-2.6 g leucine per 100 g [34, 35], and is therefore not an ideal protein source for stimulating MPS due to the required quantity.

Does Soy Negatively Impact the Thyroid Gland?

There is limited evidence that people with healthy thyroids and appropriate iodine intake are negatively affected by soy consumption [36]. Some hypothyroid patients may need to adjust their medicine’s dose to compensate for soy foods, as well as other foods that contain goitrogens (uncooked cruciferous vegetables, for example). Hypothyroid patients should discuss iodine intake and medications with their doctors.

Is Soy Formula Safe for Babies?

Yes, human babies fed soy formula had similar development to breast-fed and dairy formula babies [37, 38]. Male marmoset monkeys did not show any gross adverse effects in reproduction when fed soy from infancy to 2.65 years [39].

Does Soy Affect Cancer Development?

The role of soy in cancer development is discussed in several reviews and meta-analyses.

Prostate Cancer – There is some evidence for soy and its isoflavones to reduce prostate cancer risk, but the mechanism is unclear. Effects on hormones, prostate specific antigen (PSA), and sex hormone binding globulin (SHBG) protein could not be derived due to limitations of the reviewed data sets [40]. In patients with localized prostate cancer, thyroid hormone and sex hormones were not significantly altered by the genistein treatments, and serum PSA was decreased [41].

Ovarian Cancer – There does not appear to be a significant decrease in ovarian cancer risk with phytoestrogen consumption [42, 43].

Breast Cancer – Long-term observational data shows no harm from soy for breast cancer risk or recurrence. The traditional Japanese diet has 25-50 mg isoflavones per day in 2-3 servings of soy and may be protective. E2 and estrogen-responsive tissues do not appear to be affected by soy in human trials [44]. In nearly 10,000 breast cancer survivors, those who ate more soy after diagnosis had a significant 25% reduction in recurrence at 7.4 years post-diagnosis [45].

In 11 cancer cell lines, isoflavones “induce decreased cell proliferation, increased apoptosis and cell cycle arrest” [46].

Does Soy Raise IGF-1?

IGF-1 is a peptide hormone with known actions in bone; it is required for bone development and function [47]. IGF-1 interacts with IGF-1 binding proteins (IGFBP) for transport in the body. There are concerns that elevated IGF-1 levels promote cancer development, as associations between high IGF-1 and cancer were detected [48]. Interestingly, higher levels of IGFBP reduced the risk associated with high IGF-1 [49, 50].  An abundance of binding proteins would reduce the concentration of available IGF-1 in the bloodstream. Protein intake is associated with IGF-1 and IGFBP levels, with a larger effect observed for animal-derived proteins [51].

Studies have examined whether soy or soy isoflavones can alter IGF-1 or IGFBP levels. The results are inconsistent. In the literature, 7 studies found no increase in IGF-1 levels [52-59], 13 studies found increases in IGF-1 levels [60-72], and 4 studies found decreases in IGF-1 levels [73-76]. It is worth noting that the increases of IGF-1 in these studies can be quite small, as one study noted an increase of only 4 nmol/L [54]. Several studies also measured IGFBP and found increased IGFBP [52, 62, 70, 76]. In post-menopausal women, some studies found improvements in bone health markers, but bone mineral density was not improved in all studies [66-69]. In breast cancer patients, specifically in pre-menopausal women, a negative trend was observed between soy consumption and IGF-1 levels, but a positive trend between soy consumption and IGFBP was observed [76]. Another study of breast cancer patients found soy consumers had reduced recurrence in some cases, and that women at risk of breast cancer may benefit from soy and soy isoflavone consumption [77]. A potential gene-environment interaction has been suggested for soy isoflavone consumption and reduced IGF-1 levels [78].

About The AuthorChristine

Christine Crumbley has a PhD in molecular biology. Her research interests include nuclear receptors, regulatory mechanisms of transcription, metabolism, and circadian rhythms. As a scientist and an ethical vegan, she feels she is uniquely positioned to discuss health-related topics. She helps moderate an evidence-based vegan nutrition and fitness group on facebook (VBBN), and enjoys researching members’ questions to produce new content for the site. Her hobbies include powerlifting, Olympic lifting, Crossfit, baking, and talking to cats. Her favorite inspirational quote is, “Use your strength to help others.”
















































































Leave a Reply

Your email address will not be published. Required fields are marked *