Although intrinsic mechanisms that regulate arterial blood pressure (BP) are similar in men and women, marked variations exist at the molecular, cellular and tissue levels. These physiological disparities between the sexes likely contribute to differences in disease onset, susceptibility, prevalence and treatment responses. Key systems that are important in the development of hypertension and cardiovascular disease (CVD), including the sympathetic nervous system, the renin–angiotensin–aldosterone system and the immune system, are differentially activated in males and females. Biological age also contributes to sexual dimorphism, as premenopausal women experience a higher degree of cardioprotection than men of similar age. Furthermore, sex hormones such as oestrogen and testosterone as well as sex chromosome complement likely contribute to sex differences in BP and CVD. At the cellular level, differences in cell senescence pathways may contribute to increased longevity in women and may also limit organ damage caused by hypertension. In addition, many lifestyle and environmental factors — such as smoking, alcohol consumption and diet — may influence BP and CVD in a sex-specific manner. Evidence suggests that cardioprotection in women is lost under conditions of obesity and type 2 diabetes mellitus. Treatment strategies for hypertension and CVD that are tailored according to sex could lead to improved outcomes for affected patients.
Although blood pressure (BP) is lower in women than in men during the reproductive years, 50% of all cardiovascular disease (CVD)-related deaths occur in women, resulting in a greater incidence of CVD in older women than in age-matched men.
The same mechanisms that regulate BP and cardiovascular function are present in both men and women, but these systems are shifted towards cardioprotective pathways in women between puberty and menopause.
Sex hormones such as oestrogen and testosterone have a role in cardioprotection by modulating vasodilator and vasoconstrictor pathways, including the renin–angiotensin–aldosterone system (RAAS) and the endothelin system.
The sex chromosome complement can act independently of sex hormone effects, which results in sex-specific, age-specific and tissue-specific differences in gene transcription.
Obesity affects more women than men; as obesity is associated with a loss of cardioprotection, CVD occurs at an earlier age in obese women than in lean women.
Women have a longer lifespan than men and develop age-related and CVD-related pathologies later in life; these beneficial outcomes might be due in part to sex differences in cell injury and repair pathways that delay the chronic accumulation of senescent cells, end-organ damage and the progression of CVD.
Hypertension is a major risk factor for cardiovascular disease (CVD), which is the most common cause of death worldwide1. Although premenopausal women have a lower incidence and severity of hypertension — and therefore a lower incidence of CVD — than men, the risk increases sharply after menopause2,3. Multiple factors — including biological age, sex hormones, sex chromosomes and lifestyle as well as environmental factors and tissue-specific gene expression — influence blood pressure (BP) and the cardiovascular phenotype across the course of life. Biological sex and gender differences exist in the prevalence of CVD risk factors, mortality and the response to treatment following a cardiovascular event4,5. Concerns have been raised about these differences, as it has become apparent that women do not respond to treatment as well and have a higher risk of death following a cardiovascular event6. However, the mechanisms underlying this sexual dimorphism remain poorly understood (Fig. 1). One reason for the lack of investigation into sex differences in preclinical studies was the belief that larger sample sizes would be required for females than for males owing to hormone cycle-mediated data heterogeneity, but data from both rats and mice indicate that this hypothesis is likely to be unfounded7,8. Increasing evidence shows that findings of studies in males do not always apply to females. Nevertheless, females remain under-represented in preclinical animal studies and clinical trials in humans9,10. Hence, more studies are required to understand sex differences in hypertension and associated CVD. Since the US National Institutes of Health mandated that applications for funding needed to take into consideration sex as a biological variable, the number of studies investigating health disparities between the sexes has greatly increased11,12. This Review explores the current findings and knowledge gaps (Box 1) of sex differences in the causes and consequences of hypertension and CVD.
Women have a longer lifespan than men — living 5 years longer on average — and are less likely to die at any age than men: 95% of centenarians are women, and female neonates are more likely to survive to childhood13. However, more women than men die of CVD, although these deaths are delayed by ∼10 years3. Biological age probably has a major role in sexual dimorphism relating to BP regulation and cardiovascular function, which contributes to sex differences in cardiovascular risk.
Blood pressure regulation
Transition to life ex utero. Sex differences in vascular function exist in utero and at birth: for example, male preterm babies had a higher death rate (7.1%) than female preterm babies (4.7%) in a cohort of 12,816 infants, and this difference was associated with greater peripheral microvascular flow in males14. Increased peripheral vasodilatation is associated with augmented nitric oxide production, circulatory collapse and death in preterm infants15,16,17. Preclinical studies show that the renin–angiotensin–aldosterone system (RAAS) plays a major part in the cardiovascular transition to life ex utero, with a rapid decline in the renal expression of type 2 angiotensin II receptor (AT2R), which via the production of nitric oxide promotes vasodilation, and an increase in vasoconstrictor AT1R levels18. These changes likely result in an increase in total peripheral resistance, contributing to the rise in BP that occurs during the transition from fetal to ex utero life. Thus, manipulation of this pathway by inhibiting AT2R signalling could be a therapeutic option for maintaining peripheral vascular resistance in preterm infants. Hence, the contribution of AT2R to renal and microvascular function warrants further investigation in this setting.
Age-related changes in blood pressure. Arterial BP increases with age in mammals, including humans19,20. However, the pattern of increase is different in males and females in both humans21 (Fig. 2; Table 1) and other mammalian species19,20 (Table 2). At birth, BP is low (40–50 mmHg)22 and rises steadily until adolescence, at which point BP values diverge in males and females23. In males, BP continues to rise, increasing approximately 25 mmHg from 10 to 40 years of age, but in females, the rise in BP over the same period is only ∼15 mmHg (Ref. 24). However, from ∼60 years of age onwards, BP starts to rise steeply in women and eventually surpasses that of age-matched men25,26.
Hypertension. Hypertension has traditionally been defined as BP ≥140/90 mmHg, but 2017 guidelines27 have lowered the threshold to ≥130/80 mmHg, which is in line with evidence that aggressive lowering of BP decreases the risk of CVD. The prevalence of hypertension as well as cardiovascular and renal diseases is lower in premenopausal women than in age-matched men and postmenopausal women, and these differences are seen across different ethnic populations21,28,29. For example, the 2017 National Health and Nutrition Examination Survey (NHANES) shows that the prevalence of hypertension in individuals aged 45–54 years is ∼33% in women compared with 36% in age-matched men3. However, this cardio-renal protection is lost postmenopause: the incidence of hypertension in postmenopausal women is fourfold higher than in premenopausal women, whereas the incidence in age-matched men increases only threefold3. Moreover, the proportion of women with hypertension increases markedly with age, such that over the age of 75 years, ∼81% of women and ∼73% of men have hypertension3. This marked reversal of sex difference in the prevalence of hypertension in elderly people may in part be attributable to the observation that men with hypertension-related CVD often die before the age of 75. However, loss of cardio-renal protective mechanisms with age may also contribute to the sharp increase in hypertension and CVD in postmenopausal women. Moreover, defects in these protective pathways may contribute to the development of vascular diseases that are unique to younger premenopausal women (such as hypertensive disorders of pregnancy, gestational diabetes, polycystic ovary syndrome (PCOS) and hypothalamic hypoestrogenaemia).
Although RAAS inhibition is a common therapeutic option for hypertension in adulthood, evidence suggests that dimorphic responses to these anti-hypertensive agents exist30. For example, a cohort study of 30 men and women aged 30 years reported that an angiotensin II-mediated reduction in renal blood flow is inhibited at a lower dose of the angiotensin receptor blocker (ARB) irbesartan in women than in men, which is associated with a reduction in AT1R expression in females31. Another study reported better survival after congestive heart failure for women treated with an ARB (n = 1,596) than for those treated with an angiotensin-converting enzyme inhibitor (ACEi; n = 8,267), whereas no difference was observed in males (ARB n = 991; ACEi n = 8,484)32. Thus, sex differences in BP control and responses to treatment are present at all stages of life, and understanding these differences may improve treatment in both men and women.
Decline in function during ageing. As the kidney plays a key part in the determination of BP, loss of renal function with age or disease can drive an increase in BP33. Ageing is associated with renal structural alterations, including reduced cortical mass owing to tubular atrophy, interstitial fibrosis and glomerular sclerosis34, as well as a functional decline in glomerular filtration rate (GFR), which decreases by ∼10% for every decade of life after the age of 30 years in humans and at an equivalent age in other mammals35,36,37. This reduction in GFR is accompanied by a fall in renal blood flow and an increase in renal vascular resistance and tubular dysfunction38. This reduction in renal function with age is associated with a loss of nephrons35,39, which likely contributes to the twofold higher prevalence of chronic kidney disease among people aged >60 years than among those <60 years of age40,41. The risk of renal disease and CVD is further increased by the presence of comorbidities such as hypertension, obesity and diabetes mellitus, which accelerate cellular ageing.
Sex differences in renal function. Sex differences in renal function exist, particularly in regard to mechanisms of salt and water handling, which are processes that play a key part in BP regulation42,43. Studies in rats and mice have shown that females excrete the same amount of sodium at a lower arterial pressure than males44,45. This greater ability to excrete sodium likely contributes to the lower basal BP (∼5–10 mmHg) in female than in male adult rodents. By contrast, BP is more salt-sensitive in aged than in adult female mice, as characterized by a rightward shift in the pressure–natriuresis relationship45. Consistent with these findings, a small study of 12 hypertensive and 7 normotensive postmenopausal women reported an increased salt sensitivity in the hypertensive postmenopausal women46. Furthermore, in women with a history of severe preeclampsia (n = 22), BP salt sensitivity was increased before menopause as compared with controls (n = 7) matched for age and parity47. The mechanistic basis for the increased salt sensitivity in preeclamptic and hypertensive postmenopausal females compared with controls is poorly understood, but preclinical and clinical studies suggest that sex hormones (see below) contribute in part by modulating tubular sodium reabsorption. Evidence shows that sodium transporter profiles along the renal tubules differ between the sexes in both mice and rats48. Expression levels of sodium/hydrogen exchanger 3 (NHE3), the main transporter present in the proximal portion of the renal tubule, are lower in females than in males, whereas in distal portions of the renal tubule, sodium/chloride co-transporter (NCC) and epithelial sodium channel (ENaC) abundance is greater in female than in male rats48. These differences may lead to reduced reabsorption in early segments of the tubule in females but increased reabsorption in the distal tubules and collecting duct, ultimately facilitating sodium excretion in females. Further research is required to investigate the role of these phenotypic differences in renal sodium transporters to determine whether this phenotype is lost postmenopause and to test whether diuretics that target the distal tubule and collecting duct are more efficacious in females than in males.
Senescent cells, which are induced by cell injury or developmental cues, generate coordinated, beneficial effects in regard to tissue remodelling and repair under physiological conditions (Fig. 3). These cells, which are characterized by a senescence-associated secretory phenotype (SASP), are cleared from the tissue by macrophages after exerting their beneficial effects, a process followed by repopulation with progenitor cells and regeneration of the damaged tissue. By contrast, chronically senescent cells, which arise during ageing, are not efficiently eliminated and accumulate, resulting in a vicious cycle of tissue fibrosis and organ damage owing to chronic inflammation49 (Fig. 3). The molecular mechanisms underlying the age-related decline in renal function, as well as those underlying the ageing process in general, are not well understood, but DNA damage during replication and cellular senescence have emerged as important contributors to ageing and disease49. Related factors, such as sirtuins, oxidative stress, mitochondrial dysfunction and the anti-ageing gene KL (encodes klotho), have been implicated, as each interacts with cellular pathways that direct cellular senescence50. Also of interest, RAAS inhibition reduces the age-related loss of renal function by modulating senescence-associated pathways50. Chronic activation of AT1R by angiotensin II has been postulated to activate mechanistic target of rapamycin (mTOR) as well as to induce oxidative stress, mitochondrial dysfunction and cell senescence, actions that RAAS inhibition prevents50. Moreover, RAAS inhibition increases signalling of sirtuins and klotho, which are known pro-survival factors51,52. Although age suppresses the systemic RAAS53, intrarenal sensitivity to angiotensin II is increased at least in males54, which may promote renal cellular senescence. Whether these effects also occur in females is unknown. Moreover, expression of the pro-inflammatory cytokines tumour necrosis factor (TNF) and IL-1β — signature cytokines of SASPs — was higher in mesangial cells obtained from male rats than in those obtained from female rats, indicating a greater presence of senescent cells in males. This differential cytokine expression was influenced by AT2R and sex hormones55,56. Therefore, disparities in the RAAS and senescence-related pathways might differentially contribute to the accumulation of senescent cells, the SASP and/or the pro-fibrotic response between males and females (Fig. 3).
Telomere attrition. Telomere length is similar in both sexes at birth but shortens more rapidly throughout life in men than women57. Once telomere shortening reaches a critical level, telomeres can induce permanent cellular senescence58. Thus, reduced telomere attrition may contribute to a longer lifespan in women, although the reasons for these differences are not fully understood. Reduced longevity has been associated with changes in cell senescence pathways, including a 30% increase in short telomere count, a twofold to threefold upregulation of GTPase HRAS and cellular tumour antigen p53 (cellular markers of senescence burden) and a greater loss of renal function at 15 months of age in male than in female rats59. Mechanistic evidence suggests that increased renal oxidative stress drives telomere shortening and cellular senescence in male Wistar rats59. Interestingly, the increase in telomere shortening is region-specific, being greater in the renal cortex than the renal medulla59, and increased oxidative stress in the cortex compared with the medulla has been postulated to underlie this heterogeneity. In addition, in postmenopausal women (n = 17), arterial telomere uncapping is 2.5-fold greater than that in premenopausal women (n = 11), which has been associated with a greater expression of cellular markers of senescence60. Furthermore, telomere uncapping is twofold greater in arterial biopsy samples from subjects with hypertension (n = 29) than in those from normotensive (n = 26) subjects61. Telomere shortening has also been linked to adiposity in humans, which is of particular interest given that obesity has increased deleterious effects on cardiovascular health in women compared with men (see below)62. Greater telomere attrition in obese females may therefore contribute to the increased risk of CVD in this patient population.
Role of mitochondria. Mitochondrial dysfunction has become an area of intense interest with regard to its role in the pathophysiology of CVD and ageing63. Mitochondria are important for the production of ATP and reactive oxygen species, steroid hormone synthesis and cell death. Importantly, mitochondrial DNA is exclusively inherited from the mother, which likely contributes to sex differences in lifespan owing to greater penetrance of mutations in males, resulting in sub-optimal mitochondrial function64,65. Indeed, marked sex differences in mitochondrial function have been reported that could contribute not only to the increased longevity of women but also to cardioprotection66. For example, mitochondria isolated from female rat hearts produce less reactive oxygen species and have a greater antioxidant capacity than those from males, which has been associated with post-translational modification of mitochondrial proteins owing to activation of oestrogen receptors located on the mitochondrial membrane67. Another study found sex-specific differences in renal mitochondrial gene expression (as assessed by PCR gene array) in growth-restricted baboon fetuses, which are induced by maternal nutrient restriction during pregnancy68; for example, genes encoding proteins that are important in mitochondrial respiration (such as cytochrome c and ATP synthase) are downregulated to a greater extent in males68. Maternal nutrient restriction alters fetal development, particularly of the kidney, predisposing the offspring to renal dysfunction and hypertension later in life, and females are more protected from these outcomes than males, at least during the reproductive years69. Thus, mitochondrial dysfunction caused by an altered energy balance may contribute to renal disease and CVD.
Taken together, an improved understanding of the mechanisms that drive age-related changes in kidney structure and function, including senescent pathways, the effect of telomere length and mitochondrial function — and the influence of sex differences on these pathways — will form the basis of therapeutic opportunities to target fibrosis and improve renal survival in patients with hypertension and chronic kidney disease.
Extensive evidence indicates that endogenous oestrogen is associated with lower BP in women70,71. Administration of exogenous oestrogen might therefore be expected to also lower BP. However, inconsistent results regarding the effect of exogenous oestrogen treatment on BP in humans have been reported, with various studies indicating that BP is decreased72,73,74, increased75,76,77 or unchanged78,79,80. The response seems to be dependent on the type of oestrogenic compound administered, the dose of oestrogen and the method used to measure BP81. Animal studies provide stronger evidence that oestrogen influences BP and CVD82,83,84,85,86; although not all studies demonstrate BP-lowering effects of oestrogen, other beneficial effects on cardiovascular function, such as vasodilatation, were reported87,88. As with studies in humans, disparate findings with regard to the effects of oestrogen on BP are likely to be dependent on dose, route of administration, animal strain and age and timing of replacement or may reflect differences in the sensitivity of BP measurement methods (for example, radiotelemetry versus tail-cuff plethysmography)89. In particular, the dose of oestrogen administered varies widely among studies (>60-fold). Hence, determining the therapeutic range over which oestrogen has beneficial cardiovascular outcomes would be useful to advance this field.
Studies in which oestrogen was shown to modify BP have provided some insight into the underlying mechanisms. For example, ovariectomy in female Sprague Dawley rats raised baseline BP (∼10 mmHg), whereas oestrogen replacement therapy with 17β-oestradiol reversed the increase by promoting vasodilatation and thereby enhancing vascular conductance90. Furthermore, administration of 17β-oestradiol in rats modulated BP directly through non-genomic effects on vascular, renal and cardiac cells by reducing calcium efflux91,92 and indirectly through genomic actions, inhibiting the synthesis of potent vasoconstrictors, such as angiotensin II, endothelin 1 (ET1; also known as EDN1) and catecholamines80,93,94.
Substantial evidence indicates that oestrogen exerts profound effects on the RAAS and the endothelin system and that these effects might largely underlie the effects of endogenous oestrogen on the long-term regulation of BP (Fig. 4a). Oestrogen modulates the expression of most components of the RAAS, shifting the balance towards the so-called protective depressor RAAS pathways (that is, the AT2R–angiotensin-converting enzyme 2 (ACE2)–angiotensin (1–7)–MAS axis), which oppose the pressor actions of AT1R (Fig. 4b). For example, oestrogen increases the synthesis of angiotensinogen and decreases the synthesis of the RAAS enzymes renin and ACE in various animal models80,84,95,96,97. AT2R expression in rat kidneys is decreased by ovariectomy and increased by 17β-oestradiol administration98,99,100. Furthermore, 17β-oestradiol administration augments plasma angiotensin (1–7) levels in ovariectomized rats84. MAS expression is also higher in female than in male kidneys101. By contrast, oestrogen acts on the pressor pathway by downregulating AT1R expression in the kidney, adrenal cortex and vascular smooth muscle cells100,102,103.
Similar to its effects on the RAAS, oestrogen decreases the pro-hypertensive effects of ET1 by modulating not only the production of ET1 but also the expression of the endothelin receptor type A (ETAR) and endothelin receptor type B receptor (ETBR). Oestrogen inhibits ET1 mRNA expression in various cell types from multiple species104,105,106,107,108, including human umbilical vein endothelial cells109. In addition, 17β-oestradiol reduces circulating ET1 levels105,110 as well as endothelin-converting enzyme activity and expression levels in ovariectomized rats110,111. 17β-oestradiol can also inhibit ET1 synthesis indirectly in vitro by increasing nitric oxide synthase104 and/or decreasing angiotensin II production107,108. Last, 17β-oestradiol alters cardiac, renal and vascular expression of the endothelin receptors. In ovariectomized rats, 17β-oestradiol decreases cardiac expression of ETBR112,113, which elicits negative chronotropic effects114. Within the vasculature, 17β-oestradiol reduces ETAR expression in ovariectomized rabbit thoracic aortae. Whereas ETAR is expressed on vascular smooth muscle and is known to elicit vasoconstriction, ETBR is expressed in both the endothelium and vascular smooth muscle, where the receptor may mediate vasoconstriction or vasodilation, respectively. In female deoxycorticosterone acetate (DOCA)-salt hypertensive rats, ovariectomy decreased ETBR expression and increased the vasoconstrictor response to the ETBR agonist IRL-1620. These effects were reversed by 17β-oestradiol, suggesting that oestrogen increases expression of endothelial vasodilatory ETBR115. Within the kidney, 17β-oestradiol increases both ETAR and ETBRexpression116. ETBR is the dominant endothelin receptor (∼70%) within the kidney117, and its activation induces natriuretic and diuretic effects via nitric oxide synthase 1 (NOS1) and cGMP-dependent protein kinase (PRKG) in both sexes118,119. Furthermore, in females but not males, ETAR contributes to ET1-dependent natriuresis, and this effect is also mediated via NOS1.
Consistent with these findings, BP is lower in women during the luteal phase — when high oestrogen levels are maintained — than during the follicular phase of the menstrual cycle120. Furthermore, hormonal changes during the luteal phase of the menstrual cycle are associated with increased arterial compliance, reduced endothelial reactivity, increased vascular smooth muscle cell sensitivity to nitric oxide and activation of the RAAS and endothelin system121,122,123,124. Evidence from clinical studies suggests that BP salt sensitivity in postmenopausal women is linked to the loss of oestrogen46. Studies in animal models demonstrate that oestrogen promotes sodium excretion by modulating the effects of the sympathetic nervous system (SNS), RAAS, ET1 and nitric oxide in the kidney125,126, which together act to reduce renal sodium reabsorption. Conversely, during pregnancy when oestrogen is elevated, ET1 expression is reduced and the sensitivity to the pressor effects of angiotensin II is lower than that in non-pregnant women, which enables blood volume to double without increasing BP during pregnancy124,127,128,129. However, large-scale clinical trials have failed to demonstrate that hormone replacement therapy (HRT) is cardioprotective: for example, HRT in the Women's Health Initiative trial comprising 16,608 postmenopausal women aged 50–79 years resulted in increased cardiovascular risk130. Re-examination of this study led to the 'timing hypothesis', which postulates that discrepancies in cardiovascular outcome can be explained by differences in time elapsed between the onset of menopause and the start of HRT as well as the form of oestrogen therapy used131,132. The Kronos Early Estrogen Prevention Study (KEEPS; n = 713) and Early Versus Late Intervention Trial (ELITE; n = 643), which were designed to address these issues, showed trends for reduced cardiovascular risk but the differences did not reach statistical significance; these trials have been discussed in detail elsewhere132. A subsequent clinical trial (n = 1,006) reported that women who started HRT at the onset of menopause had a statistically significant reduction in all-cause mortality risk, heart failure and myocardial infarction without an apparent increase in the risk of cancer, venous thromboembolism and stroke133. Although this trial suggests that oestrogen is cardioprotective in postmenopausal women, this conclusion is not definitive, as the study was not blinded and a placebo group was not included134. In addition, a meta-analysis showed that early-onset menopause is associated with an increased risk of coronary heart disease and all-cause mortality, indicating that early loss of oestrogen is detrimental for cardiovascular health135. Last, an observational 18-year follow-up study of the Women's Health Initiative trial reported that compared with women receiving placebo, women receiving HRT for a median of 5.6 years are not at increased risk of mortality from CVD136. In addition to oestrogen playing an important part in female cardiovascular health, compelling evidence demonstrates that oestrogen also contributes to cardiovascular health in males137. For example, in Cyp19a1−/− mice, which lack the enzyme aromatase that converts androgens to oestrogen, a marked phenotype of increased adiposity is observed in males, demonstrating an important role for oestrogen in glucose and lipid metabolism138,139 associated with an increase in BP in males. These effects of oestrogen in males (and females) are predominantly mediated via oestrogen receptor (ER)137. An emerging area of interest is the potential of oestrogen to influence immune cell function and inflammation, which may also confer cardiovascular protection. Evidence shows that oestrogen, via ER signalling, regulates macrophage and T cell function in a sexually dimorphic pattern140. Thus, a strong interest remains in the cardioprotective potential of oestrogen, and further studies that delineate the downstream pathways through which oestrogen exerts beneficial cardiovascular actions might identify therapeutic targets with fewer off-target effects than currently available options.
Testosterone confers many physiological advantages in males, contributing to the increased body size, stronger bones, greater muscle mass and greater heart size of men than women141. In particular, the left ventricle is ∼30% larger and heart rate is ∼5–10 beats per minute lower in men than in women, which contributes to the greater cardiac reserve in men142. However, in regard to BP, testosterone is pro-hypertensive and likely contributes to the increase in cardiovascular risk observed with increasing age in males. Although testosterone can stimulate both vasodilatory and vasoconstrictor pathways (Fig. 4b), it shifts the balance of the RAAS towards the pressor pathway in both rats and mice, in which testosterone is associated with activation of AT1R, resulting in increased vasoconstriction, sodium retention by the kidneys and vascular and cardiac hypertrophy143,144. Numerous preclinical studies have demonstrated that a higher AT1R:AT2R ratio is found in the blood vessels and kidneys of males than in those of females96,101,145. Castration attenuates BP in various animal models of hypertension146, and testosterone increases plasma renin activity and BP in ovariectomized rats88,147. In addition, the vascular sensitivity to ET1 is increased in male rats compared with female rats owing to greater expression of ETAR, which promotes vasoconstriction148. Furthermore, testosterone has been shown to increase sodium reabsorption in male and female rats149,150.
Consistent with these findings, plasma ET1 levels are higher in men (∼40%)123 and in female-to-male transsexuals (∼90%) (that is, biological females treated with testosterone) than in premenopausal women124,151. Moreover, although inhibition of ETBRs in the cutaneous vasculature decreased cutaneous blood flow in women, the same treatment had the opposite effect in men152. Hence, cutaneous ETBRs mediate tonic vasodilation in women and tonic vasoconstriction in men152. Taken together, these studies suggest that testosterone is pro-hypertensive. However, as testosterone confers physiological advantages, blocking testosterone signalling is not an attractive approach. Moreover, inhibition of testosterone signalling in humans may have unfavourable consequences, as chronic testosterone deficiency is associated with obesity, a known risk factor for CVD. Indeed, a case for beneficial cardiovascular effects of testosterone has been put forward, in which testosterone treatment may improve cardiovascular outcomes owing to its positive effects on body composition as well as on glucose and lipid metabolism153,154. However, further investigation into the mechanistic pathways by which testosterone exerts its effects in both males and females is required. In the future, the beneficial effects of testosterone could be therapeutically harnessed without unwanted adverse effects.
Interestingly, plasma testosterone levels are elevated approximately twofold to threefold in women with PCOS — individuals who are at increased risk of developing vascular dysfunction and hypertension155. Targeting the depressor RAAS pathways could have therapeutic benefits for these women156, as selective AT2R stimulation in a rat model of PCOS normalized the hyperandrogenaemia157. Furthermore, AT2R activation has been reported to stimulate the conversion of androgen to oestrogen in the ovaries of rats and rabbits158,159. Thus, targeting AT2R signalling may be beneficial for ameliorating the detrimental effects of PCOS on vascular health and warrants further investigation.
Although other hormones — including relaxin, progesterone, oxytocin, prolactin and vasopressin, which are classically known for their roles in reproduction — are less well characterized than oestrogen and testosterone, they also have roles in vascular function and BP regulation. Furthermore, substantial interactions among these hormones are highly likely. Relaxin has vasodilatory and anti-fibrotic functions in both sexes, and these effects may be increased in females160,161. We have shown using Rln1−/− mice that relaxin modulates basal BP in females — but not in males — and that the hormone contributes to BP regulation during pregnancy162. Progesterone receptors, which are widely expressed in the vasculature, have been shown to promote both vasodilation and vasoconstriction in humans, which is dependent on the vascular bed, the dose of progesterone and whether or not it is coadministered with oestrogen163,164. Oxytocin — a peptide hormone principally produced in the hypothalamus — is best characterized for its roles in uterine contraction and milk production165. However, oxytocin has also been shown to cause vasodilation and reduce BP in male rats, but this has not been investigated in females165,166. Oxytocin also may have roles in the developmental programming of BP, as infants who do not receive breast milk are at increased risk of developing CVD later in life167. By contrast, elevated prolactin levels have been associated with endothelial dysfunction and increased vascular stiffness in postmenopausal women168. Vasopressin (also known as anti-diuretic hormone) plays a major part in water homeostasis and is a potent vasoconstrictor169. Levels of plasma vasopressin are higher in male rats than in female rats, resulting in greater decreases in urine flow and free water clearance and an increase in urine osmolality in males170,171. This finding is of particular interest given that recurrent dehydration is a risk factor for hypertension, kidney failure and CVD and suggests that the lower levels of vasopressin are cardioprotective164,172. Further investigation into the cardiovascular properties of these hormones is warranted, especially in the context of sexual dimorphism.
Sex chromosomal gene expression
In addition to dimorphic responses to sex hormones, sex differences in hypertension and CVD risk are also dependent on the differential expression of sex chromosomal genes173. Certain factors, including the testis determinant factor (also known as SRY; encoded by Sry) on the Y chromosome, initiate sex-specific gene expression, which regulates processes such as gonad differentiation during early development. Differences in the sex chromosome complement lead to a genetic imbalance in males and females, such that genes unique to the Y chromosome are expressed in males only and genes that escape X chromosome inactivation may be more highly expressed in females. Although the number of X chromosome genes that are capable of escaping inactivation — 12–20% in humans and 3–7% in mice174,175 — is small, some of these genes (KDM5C and KDM6A, which encode histone demethylases, and DDX3X, which encodes an RNA helicase involved in transcription) are important regulators of downstream gene expression; as such, escape from X inactivation has widespread consequences176. In addition, as only females inherit a paternal X chromosome, any paternally imprinted genes on the X chromosome will be expressed in females only, which, for example, may alter growth patterns in offspring177. Sex chromosomes have important roles in BP control and CVD, as demonstrated in the four core genotype model in which Sry is translocated from the Y chromosome to one allele of chromosome 3, leading to four core genotypes (XX-male, XY-male, XX-female and XY-female). Consequently, generation of the male phenotype is not dependent on Sry being located on the Y chromosome. Importantly, concentrations of serum testosterone (between XX-male and XY-male) and oestradiol (between XX-female and XY-female) do not differ in this model178. Comparison of gonad-intact animals and gonadectomized animals of this model makes it possible to differentiate between factors regulated by sex hormones and/or the sex chromosome complement. Studies using this model have shown that sex chromosomes contribute to the angiotensin II-mediated pressor response179, the development and progression of abdominal aneurysm180, ischaemia–reperfusion injury in the heart181 and atherosclerosis182. Thus, the sex chromosomes can cause sex differences in non-gonadal tissues independent of the effects of gonadal hormones in mice. The effects of sex chromosomes and hormones can move in opposite directions, driving both cardio-renal protection and augmented vascular, cardiac and renal injury. For example, although oestrogen exerts cardioprotective effects (see above), the XX sex chromosome complement is associated with worse outcomes and larger myocardial infarct area following ischaemia–reperfusion injury regardless of gonadal sex in the four core genotype mouse model181. Therefore, outcomes are determined by the balance of these sex-dependent mechanisms, which need to be considered in concert. In addition, mice with two X chromosomes with or without the Sry gene (XX-female and XX-male with Sry on chromosome 3) gain greater body mass on a high-fat diet than mice with a single X chromosome (XY-female and XY-male with Sry on chromosome 3)183. Gonadectomy limited this increase in body mass in response to a high-fat diet but did not ablate the relative difference in body mass between the XY (male or female gonads) and XX (male or female gonads) mice183. Together with evidence suggesting that sex hormone balance (hypoestrogenaemia and hyperandrogenaemia) plays a part in loss of cardiorotection in women with endothelial dysfunction184, these data indicate that both sex hormones and X chromosome number greatly influence obesity outcomes. Indeed, in humans, sex chromosome abnormalities are also associated with increased cardiometabolic risk. For example, women with Turner syndrome (XO sex chromosome complement) have an increased body mass index (BMI), a fourfold increased risk of type 2 diabetes mellitus (T2DM) and a threefold increased risk of mortality related to CVD and cerebrovascular disease, which has been linked to perturbations in glucose metabolism, lipid abnormalities and hypertension185. Similarly, men with Klinefelter syndrome (XXY sex chromosome complement) have a twofold increased risk of cerebrovascular disease, obesity and T2DM, which were also linked to disturbances in metabolic pathways186.
Sex-specific autosomal gene expression
Hundreds of genes are differentially expressed in males and females, and these dimorphic effects can be tissue-specific187,188. For example, 841 genes, including renal sodium transporters, are differentially expressed between the sexes in an age-dependent manner in rats189, which may contribute to sex differences in pressure natriuresis and long-term regulation of BP48,190. Similarly, a substantial number of liver genes are differentially expressed in males and females, but as these are different from those identified in the kidney191, a gene may be upregulated in one tissue and downregulated in another tissue in females compared with males188. In regard to hypertension and CVD, both the SNS and the RAAS are differentially regulated by sex chromosomes and hormones. For example, the sympathetic neurotransmitter tyrosine hydroxylase is upregulated by Sry192, which has been linked to hypertension in male spontaneously hypertensive rats, a genetic model of essential hypertension193. Interestingly, genes encoding AT2R and ACE2 are located on the X chromosome; thus, increased dosage of these genes may contribute to the relative cardioprotection in females20. Moreover, both oestrogen (see above) and the X chromosome, which has been demonstrated to modulate the BP response to angiotensin II infusion in the four core genotype model179, drive the RAAS towards the cardioprotective depressor pathways in females156,179,194. By contrast, both the Y chromosome and testosterone shift the balance of the RAAS towards the pro-hypertensive pressor arm194,195,196. Polymorphisms in the gene encoding AT2R are associated with preeclampsia197 and the development of hypertension in men198. Taken together, differences in genotype frequencies of X chromosome-linked polymorphic alleles between males and females are likely to contribute to sex differences in physiology and pathophysiology. However, few studies to date have compared gene expression and the effect on the proteome between the sexes in humans, and further studies are required.
Inflammation contributes to the pathogenesis of hypertension and CVD. Importantly, several genes encoding proteins involved in both the innate and adaptive immune responses are located on the X chromosome199, including forkhead box protein P3 (FOXP3), which controls the development and function of regulatory T (Treg) cells, and the CD40 ligand, which has an important role in T cell activation. In Rag1−/− mice, which lack both T and B cells owing to the absence of V(D)J recombination-activating protein 1, the pressor response to angiotensin II is blunted, which can be rescued with adoptive T cell transfer but not with B cells, highlighting a key role for T cells in mediating the pressor effects of this factor200,201,202. Importantly, both the sex of the recipient animal and the sex of the T cell donor determine the hypertensive response in this model: when T cells from wild-type female mice are transferred to male Rag1−/− mice, the pressor response to angiotensin II is blunted as compared with that of male Rag1−/− mice receiving T cells from wild-type male mice201. By contrast, adoptive T cell transfer from wild-type males results in a greater pressor response to angiotensin II in male Rag1−/− mice than in female Rag1−/− mice, and this finding has been associated with greater renal T cell infiltration202. Sex differences in renal T cell subpopulations have also been reported in spontaneously hypertensive rats: whereas females have more anti-inflammatory Treg cells, males have more pro-inflammatory T helper 17 (TH17) cells203,204, which may contribute to the sex difference in BP in this model. However, additional studies in other species and strains are required to confirm the contribution of different T cell populations in the development of hypertension. Sex differences in T cells are also observed in humans, with the proportion of circulating CD4+ T cells being greater in women than in men205. Moreover, CD4+ T cells from women produce more IFNγ, a marker of an anti-inflammatory phenotype206, whereas CD4+ T cells from men produce more IL-17, a pro-inflammatory marker206, suggesting that sex differences exist in the CD4+ T cell subpopulations. Interestingly, a sex-specific role for T cells in the development of hypertension has also been suggested: compared with men, women have a ninefold increased risk of developing the autoimmune disease systemic lupus erythematosus (SLE)207, which is characterized by an increase in circulating TH17 cells208 and a reduction in Treg cells209, and women with SLE have a tenfold increased risk of developing hypertension compared with that of healthy women210. Moreover, preeclampsia — the most severe form of hypertension during pregnancy — is associated with an increase in circulating CD4+ T cells211 and TH17 cells212 as well as a reduction in Treg cells compared with T cell populations in normotensive pregnant women213. Importantly, this association may have important long-term consequences for cardiovascular health given that preeclampsia is a risk factor for developing CVD in later life. A study examining the gene expression profile in macrophages following an oestrogen surge in mice showed that oestrogen promoted anti-inflammatory and injury resolution pathways in macrophages214. Although these data support the hypothesis that dysregulation of macrophage and T cell populations contributes to the development of hypertension in women, the underlying mechanisms are yet to be elucidated and warrant further study.
Lifestyle and environmental factors
Many lifestyle and environmental factors influence BP and CVD risk, and these have been extensively reviewed elsewhere215,216,217 (Fig. 1). However, as discussed below, some risk factors seem to have a greater effect in women, whereas others have a greater effect in men. Although evidently some lifestyle and environmental factors disproportionately affect one sex, the underlying mechanisms are not well understood and warrant further investigation.
Low-level alcohol consumption (1–2 alcoholic drinks per day) has been suggested to provide cardioprotection in both men and women, but moderate to high levels (3–4 or more alcoholic drinks per day) have been associated with increased risk of CVD218. However, although alcohol consumption is in general lower in women than in men, those women with moderate to high alcohol consumption have an increased CVD risk compared with men218,219. However, these data are controversial, as many potentially confounding factors, such as self-reporting of alcohol intake, existence of comorbidities or level of physical activity, were not taken into account. A population-based cohort study comprising 1,937,360 participants and another meta-analysis of 19 trials reported no difference in CVD risk between the sexes related to alcohol intake once several other risk factors were taken into account219,220.
A meta-analysis of 26 studies comprising 3,912,809 participants clearly demonstrated that smoking is associated with a greater risk of coronary heart disease in women than in men220, and another meta-analysis comprising 82 prospective cohort studies with 3,980,359 participants demonstrated a greater risk of stroke in women than in men221,222. Interestingly, smoking has been associated with an increase in testosterone levels in men and women223 and has been shown to cause oestrogen deficiency in women and to bring forward the onset of menopause224. Therefore, alterations in sex hormone levels may be one mechanistic pathway by which smoking causes a greater risk of CVD in women.
In the USA, more women (∼36%) than men (32%) are obese225. The cardioprotection normally observed in women <55 years of age is lost in the context of obesity and metabolic syndrome226,227,228, which is of particular concern given the dramatic rise in the prevalence of obesity worldwide229. Obese premenopausal women have a threefold higher risk of hypertension than lean women217,230. Furthermore, a comparable increase in BMI causes a greater increase in systolic BP in women than in men231. However, why obesity has a greater adverse effect on BP in females has not be explored in depth. Women are generally smaller than men, with a greater proportion of body fat and less muscle mass232. Interestingly, obese women also develop more obesity-related conditions — such as hyperlipidaemia, insulin resistance and T2DM — than obese men, and these conditions are associated with more severe outcomes in women227.
The mechanisms by which obesity blunts cardiovascular protection in females are not well understood. However, several studies have provided evidence that obesity-induced hypertension develops via distinct pathways in the sexes. For example, adipocytes from women produce ∼30% more leptin than adipocytes from men, which contributes to higher circulating leptin levels in obese women than in BMI-matched men233. Leptin is an adipokine that drives hypertension via central activation of the SNS in males but not in females, suggesting that other pathways are involved in females234. Leptin has been shown to stimulate the synthesis and release of aldosterone from the adrenal gland in female mice235, resulting in endothelial dysfunction and hypertension236. Consistent with these findings, obese women have disproportionately high levels of plasma aldosterone compared with the levels in obese men237. The hypothesis that sex differences in aldosterone pathways can contribute to the deleterious effects of obesity in females is supported by evidence from studies in humans and various animal models showing that mineralocorticoid pathways are differentially regulated in a sex-specific, tissue-specific and age-specific manner238.
Interestingly, studies have shown that in response to a high-fat diet, RAAS depressor pathways are activated and prevent a BP increase in female mice239,240,241. However, with longer exposure to the high-fat diet, these cardioprotective effects may wane, potentially owing to changes in the secretory phenotype of adipocytes with long-standing obesity, thereby creating a pro-inflammatory pro-hypertensive environment242. Indeed, further long-term studies in animal models of obesity are needed, as weight gain occurs inexorably over many decades in association with the development of insulin resistance and T2DM in humans.
Disparities in body composition and energy storage are also likely to contribute to greater weight gain in women in response to calorie excess. Furthermore, although women generally have a healthier diet (greater portions of fruit and vegetables) than men, men are typically more physically active243,244. In line with the 'fat-but-fit' hypothesis245, the greater muscle mass and greater level of physical activity may confer a higher relative degree of protection from obesity in men than in women. Taken together, the effects of sex-specific differences in body composition, lifestyle and environmental factors on hypertension and cardiovascular risk warrant further investigation, as greater insights into these effects might enable the development of sex-specific therapies for the treatment of obesity-induced hypertension.
Conclusion and outlook
Overwhelming evidence indicates that sex differences exist in all aspects of cardiovascular function examined to date, but the extent of these differences is currently poorly understood, with many questions remaining to be answered (Box 1). For example, the causes and consequences of sex differences in non-gonadal tissues warrant further investigation. Importantly, studies using isolated cells to investigate mechanistic pathways should take into account the sex of cells, and experiments should be conducted in cells from both males and females to obtain new insights into sexual dimorphism in the pathophysiology of disease246. Moreover, differential patterns of gene expression between the sexes are also affected by age and environmental exposures. The effects of these interactions are likely complex, and elucidating the effect of environmental factors on differential gene activation will therefore likely require sophisticated computational tools to analyse gene networks. Improved understanding of the environmental conditions that modify different mechanistic pathways could aid the development of strategies to prevent hypertension and CVD.
In addition, sex differences in mitochondrial function could potentially contribute to longevity in females and confer protection against the causes and consequences of hypertension. However, the finding that obesity and T2DM potentially switch off cardioprotective mechanisms in young women is of great concern. To enable improved treatment of hypertension and CVD, further research is required to understand sex differences in cardiovascular physiology and pathophysiology.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Nature Reviews Nephrology thanks D. Pollock, J. Sullivan and the other anonymous reviewer(s) for their contribution to the peer review of this work.
K.M.M.C is supported by an Australian National Health and Medical Research Council (NHMRC) CJ Martin Research Fellowship (1112125), and K.M.D is supported by an NHMRC Senior Research Fellowship (1041844).
- Hypertensive disorders of pregnancy
Hypertension that develops during pregnancy and is resolved following delivery of the offspring.
- Polycystic ovary syndrome
(PCOS). A hormonal condition in which testosterone levels are elevated that is associated with irregular menstrual cycles, excessive facial and body hair, obesity, reduced fertility and increased risk of diabetes.
- Hypothalamic hypoestrogenaemia
A condition in which plasma oestrogen levels are low owing to abnormal pituitary regulation of oestrogen.
- Senescence-associated secretory phenotype
(SASP). A phenotype of senescent cells in which there is a high production of cytokines (IL-β1, tumour necrosis factor), chemokines (monocyte chemoattractant proteins), reactive oxygen species (superoxide) and remodelling factors (transforming growth factor–β), which results in recruitment of immune cells to the site of injury.
- Cellular senescence
A process that results in permanent cellular proliferative arrest.
NAD+-dependent deacetylases that act on forkhead homeobox transcription factors, peroxisome proliferator-activated receptor-α and nuclear factor-κB.
- Telomere uncapping
Telomere ends form a loop structure, known as a cap, which prevents the chromosome ends from being recognized as double-stranded DNA breaks and initiating a DNA damage response. Telomere uncapping increases with advancing age and is associated with hypertension.
- Oestrogenic compound
A compound with oestrogen-like activity. In women, there are three major endogenous oestrogenic compounds: estrone, oestradiol and estriol. In rodents, 17β-oestradiol is the major oestrogenic compound.
The use of radio-waves to transmit information from a device to a remote receiver. This is the gold-standard method used to measure blood pressure (BP) in rodents, as BP can be measured continuously in unrestrained animals.
- Tail-cuff plethysmography
A non-invasive method that uses plethysmography (volume displacement) to measure blood pressure (BP) in rodents via a cuff placed around the tail. This method is associated with stress and results in underestimation or overestimation of BP, but it is an inexpensive method that lends itself to long-term tracking of BP.
- Sympathetic nervous system
(SNS). The part of the autonomic nervous system that serves to increase heart rate, constrict blood vessels and raise blood pressure when activated. Overactivity of the sympathetic nerves, particularly in the kidney, drives the development of hypertension.
A state characterized by androgen excess in females, for example, as seen in polycystic ovary syndrome.
- Sex chromosome complement
The number and type of sex chromosomes present in an organism, with XY being usual in males and XX in females.
- Four core genotype model
In the four core genotype (FCG) model, Sry, the testis-determining gene, is deleted from the Y chromosome and inserted into chromosome 3. In XY-male males (gonadal and chromosomal male), the autosome becomes male-determining instead of the Y chromosome. The FCGs are generated by crossing this XY-male (Sry on chromosome 3) with a wild-type XX-female, which gives XX-males (gonadal males with XX chromosomal complement), XY-males (gonadal males with XY chromosomal complement), XX-females (gonadal females with XX chromosomal complement) and XY-females (gonadal females with XY chromosomal complement).
- Adoptive T cell transfer
The transfer of T cells into an individual; the cells may originate from the same subject or from another individual.
- 'Fat-but-fit' hypothesis
The hypothesis that cardiovascular fitness (and muscle mass) ameliorates the adverse impact of obesity on cardiometabolic health.