The relationship between various body size indices and breast cancer r
isk before and after menopause was elucidated by means of a case-contr
ol study conducted between lune 1991 and April 1994 in 6 Italian cente
rs on 2,569 patients aged below 75 with histologically confirmed breas
t cancer, and on 2,588 controls admitted to the hospital for a wide sp
ectrum of acute, non-neoplastic, non-hormone-related diseases. Weight
and, more consistently, body mass index (BMI, kg/m(2)) at diagnosis we
re inversely related to pre-menopausal breast cancer risk and directly
to post-menopausal risk. An 8-unit increase in BMI resulted in an odd
s ratio of 0.8 for pre-menopausal and of 1.2 (significant) for post-me
nopausal women. Risk seemed to increase gradually after menopause in t
he 7th (OR for an 8-unit BMI increase, 1.3) and 8th decades (OR, 1.6)
of life. Conversely, height, waist-to-hip ratio, bra cup size and weig
ht (or BMI) in adolescence and in young adulthood did not exert a sign
ificant or consistent influence on breast cancer risk. The apparent re
lationship with BMI at middle age and weight gain between age 30 years
and diagnosis was eliminated by allowance for BMI at diagnosis. The a
ge-related pattern of the association between BMI and breast cancer ri
sk after menopause may reflect a duration-risk relationship, and resem
bles the effect of postmenopausal estrogen use, which seems greater am
ong older women. (C) 1996 Wiley-Liss, Inc.