Breasts in humans are highly sexually dimorphic compared to other clos
ely related primate species, One possible explanation for this increas
e in breast size is that breast asymmetry contains information regardi
ng women's phenotypic quality, The purpose of this work was to conside
r the relationship between asymmetry and measures of body size, body c
omposition, and fertility, We found significant associations between b
reast asymmetry and height, body mass, and breast volume (positive ass
ociations), breast ''density'' (negative association), and the woman's
age and her age at first childbirth (positive associations), The rela
tionship between asymmetry and breast volume showed evidence of negati
ve allometry, that is, women with large breasts had smaller asymmetry
than expected for their breast size, Asymmetry was lower in married wo
men compared to unmarried women and negatively related to offspring nu
mber, A multiple regression analysis, with breast asymmetry as the dep
endent variable, showed a strong significant relationship with breast
volume (positive association) and significant relationships with age a
t first childbirth (positive association) and number of offspring (neg
ative association), Asymmetric women therefore had fewer children late
r in life that did symmetric women, Simple linear and multiple regress
ion analysis showed that breast volume was a strong predictor of body
size and composition (weight, height, and breast density) and a weak p
redictor of age and age at menarche, It was not related to offspring n
umber or age at first childbirth, Our data support the thesis that mam
motropic hormones, particularly estrogen, are harmful because they sup
press the immune system, Heavy women with high levels of body fat prod
uce more estrogen and, therefore, bigger breasts, However, more estrog
en leads to an increase in breast asymmetry, Women with ''good genes''
are able to produce symmetric breasts despite the presence of large q
uantities of estrogen. Large and symmetric breasts are therefore hones
t signals of high phenotypic quality in women. (C) 1997 Elsevier Scien
ce Inc.