Soy and ER+ Breast Cancer: What the Human Evidence Actually Shows

Why so many patients were told to avoid soy

If you have been diagnosed with estrogen receptor-positive (ER+) breast cancer, someone has probably warned you about soy. It might have been a friend, a website, or even a well-meaning clinician. The warning sounds logical. Your cancer cells carry receptors that respond to estrogen, and soy contains plant compounds called isoflavones (mainly genistein and daidzein), often called “phytoestrogens,” or plant estrogens. So it seems natural to assume that eating soy is like pouring fuel on the fire.

The worry has a real scientific origin. In 1998, Hsieh and colleagues at the University of Illinois showed that genistein made MCF-7 cells (a human ER+ breast cancer cell line) grow faster in a dish at low concentrations. They also fed genistein to a special type of mouse. These mice lack a thymus, so they cannot reject human tumor cells, and their ovaries were removed to mimic menopause. In this setting, dietary genistein increased the size of implanted human breast tumors (Hsieh and colleagues, 1998). Follow-up work from the same laboratory suggested that genistein could even blunt the effect of tamoxifen and letrozole in these mice (reviewed by Messina and Nechuta, 2025).

Those findings deserved attention. They also had serious limits when it came to people.

Why mice are not a good stand-in for you

When you eat soy, your liver and gut quickly attach small molecules (a process called conjugation) to most of the isoflavones. This tags them for removal and leaves them largely inactive. Setchell and colleagues (2011) showed that more than 98% of circulating genistein in women is in this inactive, conjugated form. Rodents conjugate isoflavones far less efficiently, so they have much higher levels of the active form in their blood than people who eat the same amount of soy.

There is a second difference. Mice convert daidzein into a more estrogen-like compound called equol very efficiently. Only about 25% of Western adults and about 50% of Japanese and Chinese adults carry the gut bacteria that can make equol (Messina and Nechuta, 2025). And the tumor model itself is extreme: a mouse with no ovaries, no working immune system, and a human tumor implanted under the skin. As Messina and Nechuta (2025) point out, much of the mouse evidence that fueled the concern came from a single laboratory. That is a useful starting point for a question, but it is a weak basis for telling millions of women what to eat.

Your body’s handling of plant compounds reflects careful design. Your liver and gut work together to process what you eat, and that is a big part of why a food can behave very differently in your body than it does in a petri dish.

What happens to women who actually eat soy after diagnosis

The most important question is simple. Among real women with breast cancer, do those who eat more soy do worse? Starting in 2009, several large studies began to answer it, and the answer has been consistent.

The Shanghai Breast Cancer Survival Study followed 5,042 women in China after a breast cancer diagnosis. Women with the highest soy intake had about a 29% lower risk of death and a 32% lower risk of recurrence compared with women with the lowest intake. The benefit was seen in both ER+ and ER-negative disease, and in women who did and did not take tamoxifen. The benefit seemed to level off at about 40 mg of isoflavones per day (Shu and colleagues, 2009).

In the United States, the Life After Cancer Epidemiology (LACE) study followed 1,954 survivors. It found no harm from soy isoflavones. Among postmenopausal women on tamoxifen, those with the highest daidzein intake (more than 1,453 micrograms a day) had a 52% lower recurrence rate (hazard ratio 0.48) than those with the lowest intake (Guha and colleagues, 2009). The Women’s Healthy Eating and Living (WHEL) study followed 3,088 American survivors for a median of 7.3 years and found no adverse effect of soy. The authors concluded that doctors no longer needed to advise women with breast cancer against soy foods (Caan and colleagues, 2011).

In 2012, researchers combined these three cohorts in the After Breast Cancer Pooling Project, which included 9,514 survivors from the US and China. Eating 10 mg or more of isoflavones a day (about what you get from a cup of soy milk or a small serving of tofu) was linked to a 25% lower risk of recurrence (hazard ratio 0.75, 95% confidence interval 0.61-0.92) compared with eating less than 4 mg a day. The trend was similar in American women and Chinese women (Nechuta and colleagues, 2012). For ER+ women specifically, the estimate leaned toward benefit but was not statistically certain. That matters, because there was no signal of harm.

The Breast Cancer Family Registry, which included 6,235 women in the US and Canada, found that women with the highest isoflavone intake had a 21% lower risk of death from any cause over about 9.4 years (Zhang and colleagues, 2017). The benefit was strongest in women with hormone receptor-negative tumors. For women with hormone receptor-positive tumors, the result was neutral. Soy did not raise their risk.

Meta-analyses pooling these studies point to the same conclusion. Chi and colleagues (2013) found that eating soy after diagnosis was linked to a 15% lower risk of death and a 21% lower risk of recurrence. Qiu and Jiang (2019) found that isoflavone intake was linked to lower recurrence, but the link to overall survival was not statistically significant. The most recent pooled analysis by van Die and colleagues (2024) in JNCI Cancer Spectrum found a 26% lower risk of recurrence with higher soy isoflavone intake (hazard ratio 0.74) and an 18% lower risk among ER+ survivors (hazard ratio 0.82). The largest benefit, about a 30% reduction, appeared at 60 mg per day, with no further gain at higher intakes.

What about tamoxifen and aromatase inhibitors?

This is the question I hear most often in clinic, and it is a fair one. Tamoxifen blocks estrogen at the receptor. Aromatase inhibitors (such as anastrozole, letrozole, and exemestane) lower the amount of estrogen your body makes. If soy acted like estrogen, it could, in theory, work against either drug.

In human data, that has not happened. In the Shanghai study, soy was linked to better outcomes whether women took tamoxifen or not (Shu and colleagues, 2009). In the pooled analysis, tamoxifen users in the highest isoflavone intake category (10 mg or more a day) had a 37% lower recurrence risk (Nechuta and colleagues, 2012). Kang and colleagues (2010) studied 524 Chinese women on endocrine therapy. Among postmenopausal women, higher isoflavone intake (above about 42 mg a day) was linked to lower recurrence in women with ER+/PR+ tumors and in those taking anastrozole.

To be fair, the evidence on aromatase inhibitors comes from fewer patients than the evidence on tamoxifen. Still, across every human study I reviewed, I found no sign that dietary soy weakens these medicines.

The plain-language science: soy is not the same as estrogen

Your body has two main estrogen receptors. Think of ER-alpha as an “accelerator”: when it is switched on in breast tissue, cells tend to divide. ER-beta acts more like a “brake,” slowing growth. Your own estrogen binds both equally. Soy isoflavones bind much more weakly overall, and they prefer ER-beta (Messina and Nechuta, 2025). That is why scientists describe isoflavones as natural selective estrogen receptor modulators. They behave differently from estrogen, and in some tissues they may even work against it.

Randomized trials that give one group soy and another a placebo support this. Messina and Nechuta (2025) identified 14 trials of isoflavones and mammographic breast density (a known marker of breast cancer risk). None showed a significant increase. An earlier meta-analysis by Hooper and colleagues (2010) found no effect in postmenopausal women. It did find a small increase in premenopausal women, but that result did not hold up in every analysis. None of the six trials that measured breast cell proliferation found a meaningful increase compared with placebo.

There are some honest wrinkles. Hargreaves and colleagues (1999) gave soy to premenopausal women for two weeks. They saw no significant change in breast cell growth, but they did see small changes in nipple fluid markers consistent with a weak estrogen-like signal. Khan and colleagues (2012) gave healthy, high-risk women a high-dose isoflavone supplement (235 mg a day, several times what a traditional Asian diet provides) for six months. It did not reduce breast cell proliferation. In premenopausal women, proliferation rose modestly from baseline, although the difference from placebo was not significant.

The study that deserves the most careful reading is Shike and colleagues (2014), a randomized trial at Memorial Sloan Kettering. It enrolled 140 women with early breast cancer, 70 of whom took a soy protein powder and 70 a placebo for 7-30 days, from diagnosis until surgery. Overall, the soy group showed no change in Ki67, a standard marker of how fast cancer cells are dividing. However, in a subset of women with the highest blood levels of genistein (above 16 ng/mL), tumors overexpressed FGFR2 and genes that drive cell division and growth. The authors concluded that, in some women, soy could adversely affect gene expression in breast cancer.

Here is what this study does and does not show. It used a concentrated soy protein supplement at a dose most people would not reach through ordinary meals. It measured gene activity over a few weeks. It did not measure recurrence or survival. Gene activity is a biological signal, not a clinical outcome, and the long-term cohort studies of soy foods have not shown more recurrences. Still, I take the finding seriously, and it is one of the main reasons I recommend food over supplements.

Prevention: why the Asian data look different

Studies in Asia, where people often eat soy from childhood, tend to show lower breast cancer risk with higher intake. Western studies are mostly neutral. Dong and Qin (2011) found an 11% lower breast cancer incidence with higher isoflavone intake, a finding driven by Asian populations, along with a 16% lower recurrence risk. Wu and colleagues (2008) found the same split between Asian and Western studies.

Timing seems to matter. The Shanghai Women’s Health Study followed 73,223 women. High soy intake during adolescence was linked to a 43% lower risk of premenopausal breast cancer, and women who ate soy consistently in both adolescence and adulthood had the lowest risk (Lee and colleagues, 2009). There was no clear link with postmenopausal breast cancer. One likely reason Western studies look neutral is that most Americans and Europeans eat less than 3 mg of isoflavones a day (Messina and Nechuta, 2025). That is too little to test the question properly.

Other hormone-sensitive cancers, briefly

The data on other hormone-sensitive cancers are reassuring. Meta-analyses link higher soy or isoflavone intake to about a 19% lower risk of endometrial cancer, although prospective cohort studies alone showed a weaker result (Zhong and colleagues, 2018). For prostate cancer, an updated meta-analysis found that higher intake of soy foods was linked to a lower risk (Applegate and colleagues, 2018). The evidence for ovarian cancer is much thinner, and I would not draw firm conclusions from it.

What major cancer organizations say

The American Cancer Society’s guideline for cancer survivors concludes that soy intake, especially isoflavone intake after diagnosis, is associated with a lower risk of breast cancer recurrence, although it does not show a protective link with mortality (Rock and colleagues, 2022). The World Cancer Research Fund’s Global Cancer Update Programme graded the evidence as “limited-suggestive” that soy foods lower all-cause mortality, breast cancer mortality, and recurrence after breast cancer (Becerra-Tomás and colleagues, 2023). That grade means the evidence leans toward benefit but is not yet strong enough to call it proven.

The American Institute for Cancer Research says population studies consistently show no increased risk for breast cancer survivors who eat soy foods, and that up to 3 servings a day (up to 100 mg of isoflavones) have not been linked to increased breast cancer risk. It also cautions that studies of whole soy foods should not be used to justify isolated soy protein or isoflavone supplements. Memorial Sloan Kettering advises patients to eat whole soy foods rather than soy in supplement, powder, or pill form, and to talk with their care team before taking soy supplements, especially while on tamoxifen. MD Anderson Cancer Center also reports that soy foods may lower recurrence risk, even in ER+ disease.

Whole food soy versus supplements and isolates

Not all soy is the same. Traditional soy foods (edamame, tofu, tempeh, miso, natto, and soy milk made from whole beans) provide protein, fiber, folate, minerals, and other compounds, along with isoflavones. According to the American Institute for Cancer Research, one serving, such as a cup of soy milk, provides about 25 mg of isoflavones and about 7 g of protein. Two servings provide about 50 mg, which is close to what older adults in Japan typically eat (Messina and Nechuta, 2025). Fermented soy foods like tempeh, miso, and natto have been part of traditional diets for centuries. Soy sauce and soybean oil contain almost no isoflavones.

Isoflavone pills, by contrast, can deliver more than 100 mg in a single dose. They also lack the long record of safe use that soy foods have in survivors. The concentrated protein powder in the Shike trial is the kind of product that raised a caution flag.

You may also wonder about genetically modified or ultra-processed soy, such as some meat substitutes and protein bars. There is no good evidence that conventional versus organic soy changes cancer risk. Still, if you want to stay close to the foods that were actually studied, choose minimally processed soy, and organic if that matters to you and fits your budget.

Two brief notes on thyroid health. In people with normal thyroid function and enough iodine, soy does not appear to harm the thyroid. If you take levothyroxine, soy can reduce how much of it you absorb, so take your thyroid pill a few hours apart from soy foods and let your doctor know you eat soy (Messina and Redmond, 2006).

Honest caveats

Most of what we know about soy after breast cancer comes from observational studies. These studies show that soy eaters do well, but they cannot completely rule out other explanations. Women who eat soy may also exercise more or eat more vegetables. No randomized trial has directly tested soy against recurrence, and one may never be done. The strongest data come from Chinese women who ate soy for much of their lives, while American cohorts had fewer high-soy eaters. Questions also remain about equol-producer status, about women who start eating soy only after diagnosis, and about very high intakes. Finally, one of the most thorough recent reviews was co-written by a scientist employed by a soy-industry-funded organization (Messina and Nechuta, 2025). Its conclusions match those of independent cohorts and guideline bodies, but you deserve to know that.

What I recommend

Caring for your body well, which I see as an act of stewardship, means making food decisions based on human evidence rather than fear. Looking at that evidence as a whole, eating soy foods is not dangerous for women with ER+ breast cancer, and it may help.

If you enjoy soy, aim for about 1-2 servings a day of whole or fermented soy foods, which provide roughly 25-50 mg of isoflavones. Most studies examined up to about 3 servings a day. A serving might be a cup of soy milk, half a cup of tofu or tempeh, or half a cup of edamame. Use soy to replace processed meats or refined snacks rather than adding it on top of them. Avoid high-dose isoflavone supplements, red clover or “menopause support” phytoestrogen blends, and large amounts of soy protein isolate, unless you have discussed them with your oncologist.

And please bring this conversation to your oncology team. Your tumor biology, medicines, and overall health are unique, and this article does not replace individualized advice from the people caring for you.

References

Soy and ER+ Breast Cancer: What the Human Evidence Actually Shows

Dr. Daniel Thomas, DO, MS