Wei · European journal of epidemiology 2020 · prospective cohort study and dose-response meta-analysis · n=>300000

Soy intake and breast cancer risk: a prospective study of 300,000 Chinese women and a dose-response meta-analysis.

Cited 83 times in the scientific literature.

Level 3 - non-randomized controlled study

Prospective cohort study and meta-analysis of prospective observational cohorts

PubMed 31754945 · doi:10.1007/s10654-019-00585-4 · record verified 2026-08-28

What was done

The authors evaluated the association between soy intake and breast cancer incidence using data from the prospective China Kadoorie Biobank (CKB) cohort and conducted an accompanying dose-response meta-analysis of prospective cohort studies. The CKB cohort followed over 300,000 women aged 30–79 across 10 regions in China enrolled between 2004 and 2008 through December 31, 2016. Soy intake was assessed using baseline assessments, two resurveys, and twelve 24-hour dietary recalls. Prospective cohort studies evaluating soy isoflavone intake and breast cancer were pooled in a dose-response meta-analysis.

What was found

In the CKB cohort, mean soy isoflavone intake was 9.4 (SD 5.4) mg/day, and 2,289 incident breast cancer cases occurred over 10 years of follow-up. The multivariable-adjusted relative risk comparing the fourth quartile (19.1 mg/day) to the first quartile (4.5 mg/day) of intake was 1.00 (95% CI 0.81–1.22). In the dose-response meta-analysis of prospective studies, each 10 mg/day increment in soy isoflavone intake was associated with a 3% reduction in breast cancer risk (95% CI 1–5%).

Why it matters

This study provides large-scale prospective evidence that moderate soy intake is not associated with breast cancer risk in Chinese women, with broader pooled prospective data suggesting a modest protective effect with higher intakes.

Limits

The abstract does not report the number of studies or total participants in the meta-analysis component. Dietary intake was self-reported via surveys and recalls, introducing potential measurement error. As an observational study, residual confounding cannot be excluded, and the intake range in the CKB cohort was moderate, limiting evaluation of very high soy intake.

Cited by