Objectives. The purpose of the study was to determine whether breast-f
eeding is associated with increased longevity or cause-specific surviv
al. Methods. Teachers throughout California identified intellectually
gifted children as part of a prospective study begun in the 1920s by L
ewis Terman. Information on breast-feeding was available on 1170 subje
cts, who have been followed for more than 65 years. Results. Survival
analysis (Cox proportional hazards model) indicated that breast-feedin
g was associated with increased longevity, even after adjustment for a
ge at baseline, birthweight, infant health, and childhood socioeconomi
c status, but only among men, and the association was not significant
(P = .15). Neither cardiovascular disease nor cancer survival was sign
ificantly associated with duration of breast-feeding for either sex Su
rvival from deaths due to injuries was positively associated with brea
st-feeding after adjustment (P = .03) and demonstrated a clear gradien
t with duration, but only among men. Conclusions. Overall, the present
study does not provide strong evidence that breast-feeding is associa
ted with adult longevity. The reduced risk of death from injury may re
flect chance, in that the association was significant only for men, or
it may reflect psychosocial correlates of breast-feeding practices.