Association between duration of total breastfeeding for the first child and incident hypertension among 55,636 parous women in the Nurses’ Health Study II, United States, 1991–2005. Adjusted for age and inverse probability weight, derived from a multinomial logistic regression model for probability of each breastfeeding duration category as a function of maternal body mass index at age 18 years (linear and quadratic), year of first birth (linear and quadratic), self-reported history of preeclampsia, gestational hypertension, gestational diabetes, birth of an infant at <37 weeks’ gestation, birth of an infant weighing <2,500 g, miscarriage or stillbirth at >12 weeks’ gestation, smoking status, vigorous physical activity, alcohol consumption, Dietary Approaches to Stop Hypertension (DASH) diet score quintile, family history of hypertension, current oral contraceptive use, current nonnarcotic analgesic use, and self-reported race.