Soy and Testosterone at Food Doses

The internet still treats a block of tofu like a hormone intervention. In 2021, Reed and colleagues pooled 41 clinical studies on soy and male hormones, with total testosterone measured in 1,753 men, free testosterone in 752, and estradiol in 1,000. Soy protein and isoflavones had no statistically significant pooled effect on those outcomes, including when the trials were split by dose and study duration.¹ That pooled result does not mean every individual trial was null: one small randomized crossover trial at 120 mg of isoflavones a day reported a modest fall in total testosterone.²⁰ The men whose hormones changed in published case reports were taking in roughly 310 to 360 mg of isoflavones a day.⁴⁵¹²

01Isoflavones bind estrogen receptor beta more weakly than estradiol

Isoflavones are the compounds that put soy in the phytoestrogen category. Genistein, daidzein, and glycitein can occupy estrogen receptors. Kuiper and colleagues showed that these compounds bind estrogen receptor beta with much lower affinity than 17β-estradiol, and they bind receptor alpha more weakly still.⁹ Binding affinity in a lab assay cannot tell you what a serving does inside a person. That answer depends on the dose and on human trials that measured hormones.

Traditional soy foods deliver about 3.5 mg of isoflavones per gram of soy protein. Alcohol-washed concentrates can sit far lower. Product-to-product spread is wide, so the table below counts milligrams per serving.²³

Values are aglycone equivalents drawn from USDA isoflavone data and later compilations. Minima and maxima in the raw database are wider than this table, especially for soy milk made from isolate versus whole beans.²³

A typical U.S. diet supplies under 5 mg of isoflavones a day. Japanese and Chinese patterns often land between 25 and 55 mg.³ By the table, a 200 g tofu dinner, an 85 g tempeh lunch, and a cup of soy milk reach roughly 95 to 160 mg. That is a high-soy athlete day, and it still reaches at most about half of the 310 to 360 mg intakes in the case reports below.

02Forty-one clinical studies showed no statistically significant pooled shift

Hamilton-Reeves and colleagues first pooled this question in 2010. Fifteen placebo-controlled treatment groups showed no significant change in testosterone, SHBG, free testosterone, or free androgen index.¹⁰ Reed's 2021 update kept those trials, added a decade of new work, and reached 41 studies. Total testosterone in 1,753 men, free testosterone in 752, estradiol in 1,000, estrone in 239, and SHBG in 967 showed no statistically significant pooled effect. Sub-analyses by isoflavone dose and by study duration were also null.¹ That conclusion describes the pooled estimate, not every individual trial.

That is the paper to send the person who is sure soy crashed their T. When testosterone does come back low, the usual diet and lifestyle suspects are energy availability, sleep, alcohol, body fat, and a real zinc shortfall, the same levers already on the testosterone page.

Soy isolate also holds up as a training protein when the dose is matched. Hevia-Larrain and colleagues compared habitual vegans on plant diets plus soy isolate with habitual omnivores on mixed diets plus whey, both at 1.6 g/kg during 12 weeks of supervised lifting. The groups showed no between-group differences in lean mass or strength gains.¹¹ This was a comparison of habitual diets, not a randomized diet assignment. Source quality still matters for leucine and meal size, which is the job of the DIAAS versus PDCAAS and whey versus casein versus plant guides. Hormones are not a reason to pass over soy as the plant default.

03The hormone case reports involved over 300 mg of isoflavones a day

Three published male case reports get passed around as if they describe a serving of tempeh. Each one involved a daily isoflavone load that ordinary soy eating does not reach.

Martinez and Lewi described a 60-year-old man with six months of bilateral gynecomastia, low libido, and erectile dysfunction. Estrone and estradiol sat about fourfold above the reference range. Imaging was unremarkable. He had been drinking three quarts of soy milk a day for lactose intolerance, later estimated at about 361 mg of isoflavones.⁴¹² Tenderness resolved and estradiol fell after he stopped. Total and free testosterone stayed in range the entire time.

Siepmann and colleagues described a 19-year-old man with type 1 diabetes who developed loss of libido and erectile dysfunction on a vegan diet built around large soy loads, later estimated near 360 mg of isoflavones a day. Free and total testosterone were low. Both hormones and sexual function recovered over a year after the diet stopped.⁵ A 2022 Japanese report sat in the same neighborhood: a 54-year-old man drank about 1.2 liters of soy milk a day, roughly 310 mg of isoflavones, while also restricting carbohydrate intake. His weight fell from 85 kg to 60 kg, and his hormones improved after he stopped both soy and the carbohydrate restriction.¹²

Those are real patients, and the stop-and-recover pattern makes soy a possible contributor. The Japanese case also had concurrent carbohydrate restriction and major weight loss, so it cannot isolate soy as the cause. A case report still answers a narrow question. It shows what an extreme load did in one man. It cannot say how often that happens, and it says nothing about a dose a third the size. The 41-study meta-analysis is the evidence for ordinary intakes.

If someone is drinking soy milk by the quart as a meal replacement, the move is to stop the liquid load and recheck labs. Someone using tofu, tempeh, edamame, and a scoop of isolate to hit a vegetarian cut protein target is eating well below every intake in these reports.

04Sperm trials and sperm surveys disagree

Chavarro and colleagues surveyed 99 men at a fertility clinic. Men in the highest soy-food category, mean intake of half a serving a day, had sperm concentrations 41 million per milliliter lower than non-consumers after adjustment. Motility, morphology, and ejaculate volume were unrelated. The association was stronger in overweight men and at the high end of the sperm-concentration distribution.⁷ That paper is why the fertility worry exists. It is a cross-sectional sample of men already in an infertility clinic, with soy intake recalled over three months.

Beaton and colleagues then ran the cleaner test. Thirty-two healthy men completed a randomized crossover of milk protein isolate, low-isoflavone soy isolate at 1.6 mg a day, and high-isoflavone soy isolate at 61.7 mg a day, each for 57 days with 28-day washouts. Semen volume, sperm concentration, sperm count, motility, total motile count, and morphology did not change on either soy isolate relative to milk protein. Thirty-one percent of the men were equol producers, the gut-bacteria phenotype that converts daidzein to a more estrogen-receptor-active metabolite, and the null result still held.⁶ That covers about 10 equol producers at 62 mg of isoflavones a day. The result fits food-dose soy in that small group and leaves higher doses untested.

Male partner soy intake was also unrelated to fertilization, embryo quality, implantation, clinical pregnancy, or live birth among couples in infertility treatment.¹³ When a clinic survey and a randomized feeding trial disagree, the trial deserves more weight. Its size is the limit. Thirty-two men over 57 days can rule out a large drop in semen quality more convincingly than a small one.

05Soy food after diagnosis has not tracked with worse breast cancer outcomes

The female side of the fear is that isoflavones will stimulate estrogen-sensitive breast tissue. The largest survival data go the other way. These are cohort studies, so women who ate more soy may also have differed in diet, weight, or treatment in ways the adjustments did not fully capture.

Shu and colleagues followed 5,042 surgically treated breast cancer patients in the Shanghai Breast Cancer Survival Study. Highest versus lowest quartile of soy protein intake had a hazard ratio of 0.71 for total mortality and 0.68 for recurrence. The association appeared in both estrogen-receptor-positive and estrogen-receptor-negative disease, and in tamoxifen users and non-users. The dose-response flattened near 11 g of soy protein a day, or about 40 mg of isoflavones.¹⁴

Nechuta then pooled 9,514 U.S. and Chinese survivors from the After Breast Cancer Pooling Project, mean follow-up 7.4 years. Intake of at least 10 mg of isoflavones a day after diagnosis was associated with a 25% lower recurrence risk (HR 0.75, 95% CI 0.61 to 0.92). All-cause and breast-cancer-specific mortality point estimates were lower and not statistically significant.⁸ Food soy is the exposure in these cohorts. Isolated high-dose supplements are a different product, and a survivor's oncologist still owns that decision.

Hot-flash trials are a separate, smaller effect. Taku and colleagues meta-analyzed 17 randomized trials of extracted or synthesized soy isoflavones, median 54 mg aglycone equivalents. Frequency fell 20.6% versus placebo and severity fell 26.2%, with more punch when genistein exceeded 18.8 mg. Heterogeneity was high, and later, smaller reviews have been less consistent on hot flashes specifically.¹⁵ Those trials tested extracted or synthesized isoflavones at a fixed dose. A tofu dinner varies in isoflavone content and was never the product under test, so the hot-flash numbers do not transfer to ordinary food soy.

06Levothyroxine absorption is a timing problem

Bell reported a woman after thyroidectomy whose levothyroxine dose would not suppress TSH while she took a soy protein "cocktail" immediately after the pill. Separating the soy drink from the tablet brought free T4 and TSH into range on a lower dose.¹⁶ That is an absorption timing problem. Standard advice is to take levothyroxine on an empty stomach and keep soy protein, calcium, iron, and coffee away from the dose window.

Persiani and colleagues then gave 12 postmenopausal women on stable levothyroxine 60 mg of soy isoflavones either with the tablet or six hours later. Levothyroxine Cmax and AUC ratios sat at 1.02 and 0.99, inside bioequivalence limits.¹⁷ That trial tested isolated isoflavones. It clears the isoflavones more than the soy protein drink in Bell's case, so the spacing advice still applies to protein shakes. A 2023 narrative review concluded that any soy interference with levothyroxine, if present, appears to have little clinical impact for most tablet users, while still leaving the old case reports on the table.¹⁸

Thyroid hormone production is a separate question from pill absorption. Isoflavones can inhibit thyroid peroxidase in vitro, which is why iodine-deficient settings historically paired soy formula with goiter. In iodine-replete adults with intact thyroid function, clinical reviews have not found a meaningful TSH or free T4 shift from ordinary soy food.¹⁹ If the diet is dairy-free and built on sea salt, check iodine the way a vegan already has to check B12. Do not treat tofu as the thyroid diagnosis.

07Soy-forward training days stay below case-report loads

For a lifter or plant-forward athlete, soy earns its place because it is a complete protein with a usable leucine fraction, which is why it sits at the top of the plant-based protein and vegan diet pages. Most soy-forward training days land between 40 and 160 mg of isoflavones. That is above a typical Western diet and well short of the case-report intakes.

Use soy to hit the per-meal protein number. Keep levothyroxine and the soy serving a few hours apart if you take the pill. Leave the 360 mg soy-milk experiment in the case reports, where it belongs.

If testosterone is actually low, measure it, then audit energy intake, sleep, alcohol, and training load before rewriting the grocery list around a phytoestrogen story the meta-analysis already tested.