Am Fam Physician. 2010 Oct 1;82(7):828-836.
Background: Estrogen is thought to play a major role in breast cancer growth; therefore, blocking estrogenic exposure or effects is widely used in breast cancer therapy. Soy foods are rich in isoflavones, which have both estrogen-like and antiestrogenic actions. Although American women eat significantly less soy than Chinese women (1 to 6 mg per day of isoflavones compared with 47 mg per day), soy consumption is increasing in the United States and, with it, the concern that soy isoflavones may exert estrogenic effects and promote cancer recurrence. Reports also suggest that soy isoflavones may interact with tamoxifen. To date, there has not been a comprehensive study evaluating the effect of soy food consumption on breast cancer recurrence and survival. Shu and colleagues analyzed data from the Shanghai Breast Cancer Survival Study, a longitudinal cohort study, to assess the relationship between soy food consumption and breast cancer recurrence.
The Study: Study participants were women identified through the population-based Shanghai Cancer Registry who were 20 to 75 years of age and were diagnosed with primary breast cancer between March 2002 and April 2006. Recruitment occurred approximately six months after diagnosis. The researchers conducted in-person standardized interviews with a validated questionnaire to assess baseline demographic data, including reproductive history, disease history, lifestyle factors, diet, use of complementary medicine, and quality of life. Specific dietary habits were assessed at baseline to review intake for the preceding six months, at 18 months for the preceding 12 months, and at 36 months for the preceding 18 months. The 60-month review is ongoing. The food frequency questionnaire measured consumption of commonly used soy foods, including tofu, soy milk, and fresh soybeans. The primary end points were all-cause mortality and cancer recurrence/metastasis or death related to breast cancer. Cox proportional hazard models were used to evaluate the relationship between soy intake and the study outcomes. Soy protein and soy isoflavones were categorized by quartiles of intake amount.
Results: Of 6,299 women identified through the registry, 5,042 (80 percent) consented to participate. Nine were excluded because they did not have surgical treatment. Follow-up ranged from 0.5 to 6.2 years, with a median of 3.9 years; 444 total deaths and 534 breast cancer recurrences or cancer-related deaths occurred in the cohort. Advanced age or disease stage at diagnosis, negative estrogen or progesterone receptor status, low education level or income, presence of comorbidities, more than three pregnancies, and receiving radiation therapy were inversely associated with survival. Use of tamoxifen and previous use of hormone therapy for menopausal symptoms were directly associated with survival. Women in the highest quartile of soy intake had significantly decreased four-year mortality and recurrence rates compared with those in the lowest quartile (multivariate-adjusted mortality: 7.4 versus 10.3 percent; recurrence: 8.0 versus 11.2 percent). The survival benefit of soy followed a linear dose-response curve until soy protein intake reached 11 g per day (40 mg per day of soy isoflavone). The benefit did not vary with the presence of estrogen receptors or menopausal status. Tamoxifen use in women who were estrogen receptor positive was not adversely affected by soy intake.
Conclusion: The authors conclude that soy food intake is safe and is associated with improved survival in women with breast cancer, but no additional benefit occurs once soy protein intake exceeds 11 g per day.
AMY CRAWFORD-FAUCHER, MD
Shu XO, et al. Soy food intake and breast cancer survival. JAMA. December 9, 2009;302(22):2437–2443.
editor's note: An accompanying editorial by Ballard-Barbash and Neuhouser reminds us that, although these results show a positive benefit to soy intake and should be reassuring to survivors of and patients with breast cancer in the United States, more study is needed. Significant differences may exist between the Chinese and U.S. populations that could limit generalizability. Different rates of breast cancer screening could skew the proportion of ductal carcinoma in situ (DCIS) to invasive breast cancer cases, and subsequent survival rates. There were relatively few patients with DCIS in the Chinese study, whereas up to 20 percent of U.S. women with breast cancer have DCIS.1 The four-year median follow-up is also relatively short, which may not reflect true long-term effects.
Because outcomes were based on disease-free and survival times, accurate staging at study intake is vital; the authors note that disparities in medical record keeping and abstraction between China and the United States could limit comparison between stage- and treatment-specific results. Additionally, larger cohorts are needed to better describe the effects of soy on clinically relevant subgroups, including estrogen receptor status and tamoxifen use. Finally, not only is soy consumed in much lower quantities in the United States, but soy supplements are used more often; this study does not address the safety or effectiveness of soy components. Despite these limitations, the authors agree that women who have breast cancer and those who have survived the disease should feel comfortable eating soy foods.—a.c.f.
1. Ballard-Barbash R, Neuhouser ML. Challenges in design and interpretation of observational research on health behaviors and cancer survival [Editorial]. JAMA. 2009;302(22):2483–2484.
Copyright © 2010 by the American Academy of Family Physicians.
This content is owned by the AAFP. A person viewing it online may make one printout of the material and may use that printout only for his or her personal, non-commercial reference. This material may not otherwise be downloaded, copied, printed, stored, transmitted or reproduced in any medium, whether now known or later invented, except as authorized in writing by the AAFP. Contact firstname.lastname@example.org for copyright questions and/or permission requests.
Want to use this article elsewhere? Get Permissions