Send to

Choose Destination
JNCI Cancer Spectr. 2019 Jun;3(2):pkz017. doi: 10.1093/jncics/pkz017. Epub 2019 May 21.

Soy Food Consumption, Exercise, and Body Mass Index and Osteoporotic Fracture Risk Among Breast Cancer Survivors: The Shanghai Breast Cancer Survival Study.

Author information

Yale College, Yale University, New Haven, CT.
Section of Rheumatology, Yale School of Medicine, Yale University, New Haven, CT.
Division of Epidemiology, Department of Medicine, Vanderbilt-Ingram Cancer Center, Vanderbilt University Medical Center, Nashville, TN.
Shanghai Municipal Center for Disease Prevention and Control, Shanghai, China.
Shanghai Cancer Hospital, Fudan University, Shanghai, China.



Breast cancer survivors have a high incidence of osteoporosis-related fractures; the associated factors are understudied. We investigated incidence of bone fracture and its associations with soy food consumption, exercise, and body mass index among breast cancer survivors.


This prospective study included 4139 stage 0-III breast cancer patients and 1987 pre-/perimenopausal and 2152 postmenopausal patients. Fractures were assessed at 18 months and at 3, 5, and 10 years after cancer diagnosis. Osteoporotic fractures were defined as fractures caused by falls from standing height and at sites associated with osteoporosis. Exercise and soy isoflavone intake were assessed at 6 and 18 months postdiagnosis. Weight and height were measured at baseline. Lifetable and Cox regression analyses were employed. All statistical tests were two sided.


The 10-year incidence for osteoporotic fractures was 2.9% and 4.4% for pre-/perimenopausal and postmenopausal patients, respectively. High soy isoflavone intake was associated with reduced risk among pre-/perimenopausal patients (hazard ratio [HR] = 0.22, 95% confidence interval [CI] = 0.09 to 0.53, for soy isoflavone mg/d ≥56.06 vs <31.31; P trend < .001) but not among postmenopausal patients (P interaction < .01). Overweight (vs normal weight) was a risk factor for pre-/perimenopausal patients (HR = 1.81, 95% CI = 1.04 to 3.14) but not for postmenopausal patients (HR = 0.67, 95% CI = 0.43 to 1.03; P interaction = .01). Exercise was inversely associated with osteoporotic fractures in postmenopausal patients (HR = 0.56, 95% CI = 0.33 to 0.97, for metabolic equivalents hours ≥12.6 vs <4.5) following a dose-response pattern (P trend = .035), an association not modified by menopausal status.


Our findings, especially the novel association of soy food intake with osteoporotic fractures in breast cancer survivors, if confirmed, can help guide future strategies for fracture risk reduction in this vulnerable population.

Supplemental Content

Full text links

Icon for PubMed Central
Loading ...
Support Center