Soy, isoflavones and the breast-cancer myth: what the evidence shows

The widespread worry that soy foods feed breast cancer is not supported by the human evidence; the largest studies of survivors find that soy intake is, if anything, associated with lower recurrence and death [s1][s2]. The fear came from a plausible-sounding mechanism — soy isoflavones are phytoestrogens, and some breast cancers are driven by oestrogen — but the studies designed to test it in people point the other way [s1].

Where the myth came from, and what survivors' data show

Because isoflavones can bind oestrogen receptors, laboratory and rodent work raised the concern that they might behave like oestrogen and promote tumour growth, especially alongside the drug tamoxifen [s1]. The Shanghai Breast Cancer Survival Study tested that directly, following 5,042 women diagnosed with breast cancer in China and tracking soy intake as it changed over time [s1]. Over a median 3.9 years, among 5,033 surgically treated patients, 444 died and 534 had a recurrence or breast-cancer death [s1]. Women in the highest quartile of soy protein intake had a hazard ratio of 0.71 for death and 0.68 for recurrence, compared with the lowest quartile [s1].

The absolute figures are the ones to hold onto: four-year mortality was 10.3 percent in the lowest soy quartile versus 7.4 percent in the highest, and four-year recurrence 11.2 percent versus 8.0 percent [s1]. Critically, the inverse association held for both oestrogen-receptor-positive and -negative tumours, and among both users and non-users of tamoxifen — the specific interaction the mechanism predicted did not appear [s1].

Part of why the scare took hold is that isoflavones genuinely act in two directions: they have antiestrogenic and anticancer properties as well as weak estrogen-like ones, and which effect dominates depends on the tissue and the receptor [s2]. Cell-culture and rodent studies, which use isolated compounds at doses far above what diet delivers, tended to surface the estrogenic side; the human studies, measuring people who eat ordinary amounts of soy food, do not [s1][s2]. That gap between the petri dish and the plate is the crux of the whole controversy.

Replication across countries

A single cohort in one country would not settle it. A 2012 pooled analysis, the After Breast Cancer Pooling Project, combined 9,514 survivors from two US cohorts and one Chinese cohort and followed them a mean 7.4 years, recording 1,171 deaths and 1,348 recurrences [s2]. Despite very different baseline soy intakes between the countries, consuming at least 10 milligrams of isoflavones a day was associated with a statistically significant reduction in recurrence, a hazard ratio of 0.75, with no heterogeneity between cohorts [s2]. The associations with all-cause mortality (0.87) and breast-cancer-specific mortality (0.83) pointed the same way but were not statistically significant [s2]. The consistent, safe finding on recurrence is what has moved cancer bodies away from telling patients to avoid soy.

The claims that don't hold up as well

Soy's reputation runs ahead of its evidence in the other direction too — as a heart and menopause remedy. A 2006 American Heart Association science advisory reviewed the trials and found that isolated soy protein with isoflavones lowered LDL cholesterol by only about 3 percent, and did so using an average of 50 grams of soy protein, roughly half a day's total protein [s3]. There was no significant effect on HDL, triglycerides or blood pressure, and across 19 isoflavone studies the average effect on LDL was nil [s3]. The advisory also found soy did not relieve menopausal hot flushes and gave mixed results for bone loss, and it specifically recommended against isoflavone supplements in food or pills, citing meagre and cautionary trial evidence for cancer prevention [s3].

That last point is the useful distinction: the reassuring data concern soy foods, not concentrated isoflavone supplements, which the AHA advises against and which the survivor studies did not test [s2][s3]. For readers weighing soy as a protein source, the broader trade-offs sit within our coverage of plant-based diets and how much protein you actually need. All of the survivor evidence is observational, and none of it is dietary advice — but the specific claim that soy foods endanger breast-cancer patients is one the data do not support.

Sources

• Soy Food Intake and Breast Cancer Survival — JAMA, 2009
• Soy food intake after diagnosis of breast cancer and survival: combined US and Chinese cohorts — The American Journal of Clinical Nutrition, 2012
• Soy Protein, Isoflavones, and Cardiovascular Health — Circulation (AHA Science Advisory), 2006