Breast cancer and breastfeeding: collaborative reanalysis of individual data from 47 epidemiological studies in 30 countries, including 50302 women with breast cancer and 96973 women without the disease.
Level 1 - systematic review of randomized trials
Pooled individual participant data meta-analysis of 47 observational epidemiological studies
PubMed 12133652 · doi:10.1016/S0140-6736(02)09454-0
What was done
Individual participant data from 47 epidemiological studies in 30 countries were collected, checked, and analyzed centrally. The dataset included 50,302 women with invasive breast cancer and 96,973 controls. Relative risk estimates for breast cancer associated with breastfeeding in parous women were obtained after stratification by fine divisions of age, parity, age at first birth, study, and menopausal status.
What was found
Women with breast cancer had fewer births than controls (2.2 vs 2.6), fewer parous women with cancer had ever breastfed (71% vs 79%), and their average lifetime breastfeeding duration was shorter (9.8 vs 15.6 months). The relative risk of breast cancer decreased by 4.3% (95% CI 2.9-5.8; p<0.0001) for every 12 months of breastfeeding, in addition to a decrease of 7.0% (95% CI 5.0-9.0; p<0.0001) for each birth. The risk reduction did not differ significantly by developed versus developing countries, age, menopausal status, ethnicity, parity, age at first birth, or nine other characteristics. It was estimated that cumulative breast cancer incidence in developed countries would decline from 6.3 to 2.7 per 100 women by age 70 if women adopted the parity and breastfeeding patterns historically prevalent in developing countries, with breastfeeding accounting for roughly two-thirds of the reduction.
Why it matters
This pooled analysis establishes that lifetime duration of breastfeeding independently reduces breast cancer risk in a dose-dependent manner. It demonstrates that low parity and short duration of breastfeeding largely explain the higher incidence of breast cancer in developed nations.
Limits
The underlying studies were observational epidemiological designs, making self-reported breastfeeding history subject to recall bias. The abstract does not report data on exclusivity or patterns of breastfeeding, nor does it report results broken down by molecular or histological breast cancer subtypes.
Cited by
- partial Giving birth before age 40 reduces the long-term risk of developing breast cancer.