The angiotensinogen gene is one of the very few related by linkage ana
lysis to human hypertension, but the linkage has been consistently sho
wn only among males. Moreover, polymorphisms in this gene predict an a
bnormal renal responsiveness to angiotensin II, a feature of non-modul
ation, but again, only among males. To pursue these related bridges be
tween genetics and physiology, we evaluated the effects of sex on a se
cond feature of non-modulation, the aldosterone response to infused an
giotensin II during low sodium balance. We tested the resultant hypoth
esis-that non-modulation would be less frequent in women-by conducting
identical protocols on 225 hypertensive inpatients (70 women, 155 men
). Non-modulation was strikingly less frequent among women (26%; 95% c
onfidence interval, 16% to 37%) than men (49%; 95% confidence interval
, 40% to 57%) (P=.001). We tested the hypothesis that sex steroids pla
y a role by comparing young, premenopausal women (<35 years) with wome
n who were perimenopausal (45 to 55 years) and postmenopausal (>55 yea
rs). Among the youngest women, the frequency of nonmodulation was only
7%, significantly less than in young men (41%, P=.02). A steady incre
ase in non-modulation frequency accompanied advancing age in women, re
aching 47% in those older than 55 years, equal to the fraction of men
affected. Age influenced non-modulation frequency in men far less. We
conclude that a striking sex difference underlies the nonmodulation ph
enotype and that female sex hormones may confer protection against a g
enotypic predisposition in women. This ''override'' of genotype, manif
est by a very low frequency of non-modulation in young women, may part
icipate in their known protection against cardiovascular disease.