Ovaries and 17β-estradiol
It has been difficult to disentangle the effects of ovarian function on arterial pressure in women from those of aging because ovarian hormone loss is an age-associated event. Cross-sectional studies (Figures &) suggest the rate of the age-associated rise in arterial pressure increases around the onset of menopause (i.e., 51 years of age) and prospective population studies show that postmenopausal women have higher arterial pressure than age-matched premenopausal women [102
]; however, inferences from these studies are limited because the populations within the groups compared are not likely to differ solely on the basis of age of menopause onset.
Comparing arterial pressure between young premenopausal women and women with premature ovarian failure provides another way to assess ovarian hormone loss on blood pressure since premature ovarian failure typically presents at 27 years of age. Women with premature ovarian failure have a higher incidence of hypertension than age-matched premenopausal women suggesting ovarian hormone loss raises arterial pressure independently of the effect of aging on blood pressure. These studies, however, are confounded by the possibility that premature ovarian failure represents an acceleration of the aging process (see review by Pal & Santoro [105
Turner syndrome provides a window into the blood pressure effects of ovarian function at an even earlier age since girls with Turner syndrome exhibit ovarian hormone deficiency during puberty. Compared to published population standards in girls, girls with Turner syndrome have higher blood pressure, a higher incidence of abnormal blood pressure circadian rhythms, and a higher incidence of idiopathic hypertension even in the absence of aortic coarctation, renal abnormalities or chronic urinary track infections [106
]. Turner girls do not have a complete double X chromosomal complement. Therefore, comparing arterial pressures in girls with and without Turner syndrome is confounded by sex chromosome complement differences.
Longitudinal studies investigating the relationship between arterial pressure and the age of menopause onset have not demonstrated an effect of menopause on arterial pressure that is distinct from the effect of aging. Hjortland et al. [108
] conducted a 5 year longitudinal study in a cohort of 1686 women who were 40-41 years old. van Beresteyn et al. [109
] followed 193 perimenopausal women between 49 to 56 years of age for over 10 years and Mathews [110
] examined blood pressure in 541 healthy premenopausal women 42-50 years of age for approximately 2 & 1/2 years. While none of these longitudinal studies detected an effect of menopause on arterial pressure that was distinct from aging, following changes in blood pressure in women longitudinally could have been confounded by antihypertensive medications since it is unethical to withhold antihypertensive medicine in hypertensive subjects. While it is clear that arterial blood pressure increases both with aging and ovarian hormone dysfunction in women, it is not known to what extent the mechanistic pathways of these two hypertension-associated factors overlap. Increased salt-sensitivity may be a mechanism that contributes to postmenopausal hypertension; however, whether or not postmenopausal women are more salt-sensitive than premenopausal women remains unresolved. As noted above, it is difficult to tease out age effects in women from effects due to changes in the ovarian hormonal milieu; however, much can be learned from experimental animal models.
The mRen2.Lewis rat exhibits salt-sensitive hypertension, which was not found to be magnified by ovariectomy [111
]; however, SBP was measured by tail-cuff in this study and thus, the possibility remains that restraint stress due to the measurement method confounded these findings. Furthermore, the SBP is so high in the ovariectomized mRen2.Lewis rat maintained on a HS diet for 15 weeks (approximately 230 mm Hg) that the SBP may have reached a ceiling effect (i.e., the SBP just couldn't get any higher without the animal's demise). The effects of ovariectomy on salt-sensitivity have also been examined in the SHR [112
] and DS rat [57
]. In these models, the salt-sensitive component (ΔMAP HS-LS) was found to be amplified by ovariectomy (Figure ) indicating that the loss of ovarian hormones increases salt-sensitivity. It will be important to determine how similar these findings are to other animal models of salt-sensitivity.
Figure 10 Effect of ovariectomy on salt-sensitivity. Shown is the difference in MAP between HS (8% NaCl) and LS diets 0.15-0.60% NaCl in the intact (white bar) and ovariectomized (Ovx) (striped bar) SHR  and DS rat . §p < 0.05 vs Intact, (more ...)
Much can be learned by comparing models in which ovariectomy had and did not have an effect on arterial pressure. While ovariectomy increased arterial pressure in several animal models of hypertension including Ang II- [38
] and DOCA- [54
] infusion and the DS [113
] and mRen2.Lewis [111
] rats (Figure ), ovariectomy did not raise arterial pressure in the SHR on a normal sodium chloride diet [112
] or the Wistar rat treated with the nitric oxide synthase inhibitor, L-NAME [44
]. For instance, does the finding that ovariectomy has no effect on arterial pressure in the L-NAME Wistar model suggest that ovariectomy elicits its arterial pressure raising effects by reducing nitric oxide in key target tissues such as the kidney? This possibility could be addressed by comparing bioavailable nitric oxide levels and activity in the kidneys in these various models before and after ovariectomy. Taken together, these studies indicate that the presence or absence of ovarian hormones is not the only factor that underlies sex differences in blood pressure given the mixed findings regarding the effect of ovariectomy on arterial pressure (Figure ).
Figure 11 Effect of gonadectomy on blood pressure in experimental models of hypertension. Top panel, Shown is the difference in blood pressure between ovariectomized (Ovx) and intact (Intact) females in animal models of hypertension including Ang II infusion in (more ...)
To address ovarian hormone regulation of blood pressure, most investigators have compared intact animals with ovariectomized animals treated with and without 17β-estradiol (E2
) treatment. E2
treatment prevented the blood pressure raising effects of ovariectomy in the Ang II- [114
] and DOCA- [54
] infused models as well as in the mRen2.Lewis rat, the DS rat on both LS [82
] and HS [116
] diets and the SHR on a HS diet [112
]. The consistency of the ability of E2
to prevent the blood pressure raising effects of ovariectomy across these distinct experimental models of hypertension, suggests the mechanisms by which E2
loss raises blood pressure are robust and highly conserved and thus support the conclusion of cross-sectional studies showing that menopause increases arterial pressure in women.
Most experimental studies of E2
depletion and replacement on arterial pressure have been conducted in young animals. Therefore, it has been difficult to isolate effects of E2
deficiency from aging effects. Hinojosa Laborde et al. [82
] addressed this question by following MAP by telemetry for a year in DS rats maintained on a LS diet. This study found that ovariectomy at a young age accelerated the age-associated increase in MAP whereas E2
replacement in the ovariectomized rats markedly attenuated this effect. The fact that MAP was significantly lower in the E2
replaced group compared to the intact group after one year suggests E2
treatment can protect against the age-associated increase in MAP [82
]. At 1 year of age, plasma E2
in the ovariectomized DS rats treated chronically with E2
was 33% < levels found in young 4 mo DS rats [82
]. Therefore, it would be informative to modify the E2
replacement regimen in the ovariectomized rats so that similar E2
levels to the young rats were achieved even after 1 year and then determine if these higher plasma E2
levels further improved upon or even totally prevented the age-associated increase in MAP. Mimicking the cycling of E2
that is found in the young intact female is another important variable that needs to be investigated. A constant replacement dose of E2
might desensitize ERs in a manner similar to how super agonists of the leutinizing hormone (LH) receptor cause receptor down regulation [117
]. Thus, it is possible that mimicking the estrus cycle would result in greater attenuation of the age-associated increase in arterial pressure by preventing chronic ER desensitization.
Far less is known regarding the effects of progesterone loss on blood pressure since few studies have investigated the role of ovarian hormones other than E2
. In 1997, Crofton et al. [115
] found that progesterone had no effect in and of itself on SBP in ovariectomized rats in the DOCA-NaCl model; however, this study did find that progesterone slowed the ability of E2
treatment to attenuate the blood pressure raising effects of ovariectomy. Studies are clearly needed to investigate the role of progesterone in diverse experimental models of hypertension to fully understand the role of progestins in blood pressure control. This is particularly relevant to women on hormone replacement regimens. In fact, clinical studies have shown that some progestins impair endothelial function in women while others do not, suggesting that certain progestins like depot-medroxyprogesterone acetate can inhibit the beneficial effects of E2
while others like drospirenone are without effect [118
Ten years ago, studies of estrogen receptor alpha and beta (ERα & ERβ) null mice suggested that ERβ protects against the age-associated increase in arterial pressure since at 6-7 months of age compared to wild type (WT), SBP was shown to be elevated by 17 and 28 mm Hg in the female and male KO mice, respectively [121
]. In contrast, the ERα null mouse did not show a similar age-associated increase in SBP. Pharmacological studies using ER subtype selective ligands in the SHR revealed that ligand-dependent activation of ERβ lowers arterial pressure to a greater extent than E2
alone or the ERα agonist [122
]; however, these studies were conducted solely in the female so this study did not assess if sex differences in the activity of these receptors contributes to sex differences in arterial pressure.
The gonadal hormone milieu changes markedly throughout the life of a woman. After puberty, E2
levels markedly increase by 5-10-fold [123
]. Five years or more after menopause, E2
levels drop by greater than 10-fold compared to levels in normal cycling women [124
]. The vast majority of research in menopause [125
] has centered on the role of the precipitous drop in E2
secretion from the ovarian follicles that occurs at this life transition [126
]; however, other significant hormonal changes are also occurring. By the onset of menopause, T declines by approximately 40% [124
]. The fact that the decline in T is far less dramatic than the drop in E2
means that the T:E2
ratio markedly increases as a woman transitions into menopause.
As women approach menopause, atresia of the ovarian follicles occurs, fertility declines and serum follicle stimulating hormone (FSH) and LH start to surge in an effort to produce more ovarian follicles [126
]. After menopause, there is more than a 10-fold increase in the levels of FSH and LH, the pituitary hormones of the hypothalamic-pituitary-ovarian axis, that are secreted in response to the hypothalamic hormone, gonadotropin releasing hormone [124
]. The secretion of these hormones is tightly regulated through hypothalamic-pituitary-ovarian feedback mechanisms. Supplementation with E2
markedly reduces FSH and LH levels in postmenopausal women, although levels remain significantly higher than those found in premenopausal women [126
]. Studies investigating the role of E2
in the age-associated increases in arterial pressure are confounded by changes in these hormones of the hypothalamic-pituitary-ovarian axis and therefore, the effects of E2
deficiency can not be separated from the effects of rising FSH and LH levels.
Some studies suggest that elevation of FSH contributes to hypertension-associated disease. Chu et al. [127
] showed that elevated FSH in premenopausal cycling women is associated with increased cardiovascular risk. The authors studied 40 women between the ages of 29 to 49 with normal menstrual cycles and premenopausal levels of E2
who were not receiving exogenous hormone or statin treatment. Total cholesterol and low density lipoprotein were significantly higher in women who had an FSH level ≥ 7 IU/I compared to those with an FSH ≤ 7 IU/I and these effects were independent
of age. Patients with low FSH had E2
levels around 27 pg/ml whereas patients with high FSH had E2
levels around 52 pg/ml. If the increased prevalence of cardiovascular risk factors is solely due to lower E2
levels then the opposite findings would be expected, namely, that the premenopausal women with higher plasma E2
levels would have less cardiovascular risk factors than those women with lower plasma E2
levels. This study implicates a role for FSH and LH in and of themselves in cardiovascular risk.
Inactivating mutations in the FSH receptor gene have been reported to cause hereditary hypergonadotropic ovarian failure in women [128
]. More recently, Nakayama et al. [129
] examined 5 single nucleotide polymorphisms in the FSH receptor gene in more than 1000 essential hypertension patients and age-matched controls in a subgroup analysis of the Hypertensive Section of the Japanese Millennium Project. One single nucleotide polymorphism in the 5'-untranslated region of the FSH receptor gene occurred with increased frequency in women with hypertension. Patients with the A/A genotype in this polymorphism exhibited lower levels of FSH receptor transcriptional activity and had lower E2
levels than those without the A/A genotype (G/G or G/A). These studies suggest that this single nucleotide polymorphism is a susceptibility mutation for essential hypertension in women and underscores the value of future studies that focus on ovarian hormone changes in addition to E2
in blood pressure control. Surprisingly, the roles of T, LH and FSH and changes in their ratio to E2
) throughout a woman's life span are generally neglected in hypertension research and consequently, much less is known regarding their specific effects on blood pressure control.
Testes and testosterone
There are fewer studies investigating the influence of the testicular hormonal milieu on blood pressure compared to studies of the ovarian hormonal milieu. As in women, as men age, the prevalence of hypertension increases (Figures &). Accompanying this rise in arterial pressure is a gradual decrease in total T. By the time men reach the eighth decade, free T drops to approximately 50% of the lifetime maximum [130
]. This gradual decrease in T is misleading since bioavailable T constitutes only a small fraction (1-3%) of serum T since the majority of this circulating steroid is bound to sex-hormone binding globulin [134
]. Since sex-hormone binding globulin levels increase nearly 2-fold over the male lifespan, bioavailable T is markedly decreased by aging [136
]. A positive correlation among aging, arterial pressure and decreasing bioavailable T exists and hypertensive men have lower T [138
] and androstenedione [140
] than normotensive men. This inverse relationship between T and arterial pressure [141
] suggests that hypertension in aging men is associated with decreased T activity.
While it is difficult to isolate the effects of declining T levels on blood pressure from other effects of advancing age on hypertension-associated disease [142
], studies of oncology patients support the idea that T deficiency increases blood pressure in men. Surgical removal of the testes due to testicular cancer results in higher arterial pressures than that found in age-matched controls [143
]. Furthermore, there was an inverse association between T levels and arterial pressure in these cancer survivors. Thus, these studies suggest that T deficiency has an adverse effect on arterial pressure. A study of 22 prostate cancer patients found that central arterial pressure was increased after T deficiency was induced by 3 mo treatment with LH-releasing hormone agonists [144
]. Since T is converted to E2
, lower levels of T in men are also associated with lower levels of E2
. LH receptor agonists not only reduced T from 14.5 to 1.2 nmol/liter, E2
was reduced by 3-fold [144
]. Thus, it remains unclear as to whether the increased arterial pressure in these men was due to the loss of T, decreased E2
or changes in the T:E2
In contrast to the clinical data suggesting that T deficiency exacerbates hypertension, T deficiency induced by orchiectomy lowers blood pressure in several experimental models of hypertension including Ang II infusion [38
], DOCA-salt [54
], chronic L-NAME treatment [44
] and in the SPSHR [145
] and SHR on either a normal salt and HS diet [146
]. Orchiectomy also reduced MAP by 33 mm Hg in the α1 soluble guanylate cyclase KO mouse [75
]. In contrast, orchiectomy had no effect on DS rats maintained on a LS diet although on a HS diet, removing the testes reduced arterial pressure in these animals [116
] (Figure ). What makes the DS-LS model different from these above models is that the DS-LS rats were normotensive. Thus, it is worth exploring whether or not testicular hormones only affect blood pressure in hypertensive models. Investigators have compared intact with orchiectomized animals treated with and without T. Treatment with T prevented the arterial pressure lowering effects of orchiectomy in the DOCA-NaCl [115
] infused model as well as in the SHR [69
], suggesting that T deficiency due to the gonadectomy lowers arterial pressure in these models of hypertension.
The absence of the androgen receptor (AR) did not affect arterial pressure under basal conditions as no differences in MAP were found between male WT and AR null mice (ARKO) nor was there any effect of AR deficiency on the magnitude of Ang II-induced hypertension [149
]. This study suggests the AR is not a major factor in the control of arterial blood pressure. In contrast to conclusions drawn from comparing male ARKO and WT mice, the AR antagonist flutamide reduced blood pressure in the male SHR [145
] and the male SPSHR [145
]. Flutamide also attenuated the development of hypertension in male TGR(mRen2)27 rats [151
] and the male α1 soluble guanylate cyclase KO mouse [75
]. These studies with flutamide indicate that inhibiting the AR lowers arterial pressure suggesting that the AR contributes to hypertension. To address the discrepancy between the ARKO mouse data and the flutamide data, experiments are needed to ensure flutamide is acting solely through the AR and that no pharmacological differences exist in the Emax or EC50
in the Ang II-blood pressure dose response curve between WT and ARKO mice.
The cytochrome P450 enzyme aromatase produces E2
and mice with a targeted disruption of the aromatase gene are deficient in E2
. No detectable differences in SBP were reported between the aromatase KO and WT female mice [152
]. The finding that a deficiency in E2
induced by aromatase disruption had no detectable effect on SBP is consistent with studies in WT mice that showed E2
depletion by ovariectomy did not alter basal arterial pressure [38
] and experiments in female rats that showed aromatase inhibition did not alter basal arterial pressure [153
]. The aromatase KO did however exhibit a lower basal DBP than the WT and this was ascribed to greater BP variability within mid and high frequency bands and loss of autonomic control of the heart. Although the authors did not induce hypertension in the aromatase KO mice, one might predict that the KO females would have a higher magnitude of Ang II-dependent hypertension than their WT littermates since ovariectomized mice have a higher degree of Ang II-dependent hypertension than the intact females [38
]. These are important experiments to conduct since previous studies with the aromatase inhibitor 10-propargyl-androst-4-ene,3,17-dione (19-AA) showed that 19-AA attenuated arterial pressure in an inbred salt-sensitive rat strain [154
] and the SHR [155
]; however, 19-AA also inhibits non-aromatizing adrenal 19-hydroxylation - a key step in forming 19-nordeoxycorticosterone, so it was unclear whether or not the attenuation of BP was due to E2
deficiency or lowering of 10-nor-corticosteroids in these experiments.
Why then does this majority of the animal data directly conflict with the clinical data showing T deficiency is associated with blood pressure elevation? One explanation is the discrepancy in age. Animal castrations for the most part were performed on young animals while the clinical findings are primarily reported on men in their fifth to seventh decades of life. In fact, when orchiectomies in the SHR were conducted at 6 mo of age [145
], there was no effect of the gonadectomy on SBP whereas when castration was performed between 3 weeks to 4 months [69
], orchiectomy markedly attenuated the magnitude and slowed the rate of the developing hypertension. Taken together, these clinical and experimental studies suggest that T has bimodal effects. At high levels such as found at early ages, T adversely contributes to mechanisms of hypertension, while at later ages when T levels naturally decline, there is insufficient T protection. Perhaps T action follows a physiological U-shaped response in which doses that are too low or too high are equally disadvantageous to the animal.
Studies of T and blood pressure have also been investigated in the female. In women with excess T production, T has a blood pressure elevating effect. Women with polycystic ovary syndrome have elevated T levels and there is a positive association between T levels and arterial pressure [156
]. This latter observation emphasizes that T dosage or ratio to other hormones (e.g., T:E2
) may impact blood pressure. Clearly more research into of T action is needed to sort out the role of T, E2
and the ratio of T:E2
in blood pressure control in both the male and female.