Fig. 3

Associations of a having a first-degree relative with breast cancer before age 50 and b breastfeeding duration with mammographic density by province and menopausal status. Province-specific OR (95% CI) for having dense breast (BI-RADS, 3–4) compared with non-dense breast (BI-RADS, 1–2) was estimated by multivariable logistic regression. Meta-analysis was performed based on the random-effect model. a Group without family history of either breast or ovarian cancer was used as the reference. Model was adjusted for age (per 2 years), education (none to elementary, middle school, high school, college+), BMI (per 2 kg/m²), age at menarche (<13, 13-14, 15+years old), and parity (parous vs. nulliparous). Model for postmenopausal women was additionally adjusted for age at menopause (<49, 49-50, 51+). b Breastfeeding duration (no breastfeeding, 1–6 months, 7–12 months, 13 months+) was treated as an ordinal variable among parous women with the adjustment for age (per 2 years), family history (no history, first- and/or second-degree relatives with breast and/or ovarian cancers except for first-degree with breast cancer before age 50 years, and first-degree relative with breast cancer before age 50 years; premenopausal women only), education (none to elementary, middle school, high school, college+), BMI (per 2 kg/m²), age at menarche (<13, 13–14, 15+years old), age at full-term birth (<25, 25–26, 27–28, 29+), and age at menopause (<49, 49–50, 51+; postmenopausal women only)