Abstract
In the era of individualized medicine, gaps in knowledge remain about sex-specific risk factors, diagnostic and treatment options that might reduce mortality from cardiovascular disease (CVD) and improve outcomes for both women and men. In this review, contributions of biological mechanisms involving the sex chromosomes and the sex hormones on the cardiovascular system will be discussed in relationship to the female-specific risk factors for CVD: hypertensive disorders of pregnancy, menopause and use of hormonal therapies for contraception and menopausal symptoms. Additionally, sex-specific factors to consider in the differential diagnosis and treatment of four prevalent CVDs (hypertension, stroke, coronary artery disease and congestive heart failure) will be reviewed with emphasis on areas where additional research is needed.
Cardiovascular disease (CVD) as a major global health problem carried a total cost of US$863 billion dollars in 2010, which is expected to reach $1044 billion by 2030 [1]. In the USA, all-cause mortality for CVD (including congenital CVD) is greater in women than men among all ethnic groups [2]. This health disparity reflects complex interactions among cultural and socioeconomic issues, such as access to medical care, as well as basic biological differences between women and men. These biological differences often are not accounted for in basic science studies evaluating mechanisms of disease or in development of new therapies. The under-representation of women in clinical studies and the lack of attention to conditions specific to women such as pregnancy and menopause also contribute to health disparities. In 2001, the Institute of Medicine's report emphasized the need for more research into the basic cellular and integrated physiological processes and behaviors which contribute to sex differences in health and disease [3].
In this review, contributions of biological mechanisms involving the sex chromosomes and the sex hormones on the cardiovascular (CV) system will be discussed in relationship to the female-specific risk factors for CVD: disorders of pregnancy, menopause and use of hormonal therapies for contraception and menopausal symptoms. Additionally, sex-specific factors to consider in the differential diagnosis and treatment of four prevalent CVDs (hypertension, stroke, coronary artery disease and congestive heart failure) will be reviewed with emphasis on areas where additional research is needed.
Biological mechanisms of sex differences
Sex chromosomes & X-inactivation
The basic biological factor characterizing sex is the complement of sex chromosomes. Each human cell contains 23 pairs of chromosomes: 22 are autosomal chromosomes, and the remaining are sex chromosomes. In females, there are two X chromosomes, and males there is one X and one Y chromosome [4]. The Y chromosome is less than 40% of the size of the X chromosome and contains 421 genes, 47 of which are known to be protein coding and are mostly related to male sexual development and reproduction but some associated with innate immunity may increase risk for CVD [4–6]. In contrast, the X chromosome has 1965 genes, 825 of which code into protein [4] including genes important in brain development, blood clotting and visual pigmentation [6]. Because males lack a second X chromosome, they are at greater risk for developing X-linked inherited disorders, such as hemophilia and color blindness, and may be more susceptible to learning disabilities than women. In addition, the androgen receptor is on the X chromosome and genetic variations in the gene coding this receptor will affect androgen sensitivity in males [4] (Figure 1).

In females, inactivation of one X chromosome in each cell results in a ‘single-dose’ of X chromosomal gene product thus reducing potential sex differences [9]. Xist (X-inactive specific transcript), a non-coding RNA molecule, is responsible for this process which occurs in somatic cells during early embryogenesis. Maternal versus paternal X allele inactivation is usually random [9]; thus, female somatic tissues are a mosaic of cells expressing either a maternal or paternal X. However, skewing of the X inactivation ratio can occur [9] such that asymmetrical expression of the X chromosome may be greater than once thought, as visualized and quantified in mice through the use of cellular resolution maps in a variety of tissues, including cardiac muscle [10]. In the setting of an X-linked recessive disease, skewed X-activation can result in a very wide range of phenotypes in women due to the unequal skewing of the mutated X allele and the nonaffected X allele [11]. The X-inactivation phenomenon in female cells can serve a unique purpose in elucidating vascular physiological mechanisms. For example, evaluation of X-methylation in samples of human arteries demonstrated a nonrandom (skewed) X-inactivation pattern suggesting that the cells in the plaque arose from a single or similar group of cells [12,13].
Approximately 15% of X-linked genes ‘escape’ inactivation [14]. Many of these genes exist in the pseudoautosomal regions (PAR) of sex chromosomes and are expressed on both the X and Y chromosomes [14]. Therefore, both sexes receive a ‘double-dose’ of these gene products [9]. Women with Turner's syndrome, having only a single X chromosome and thus only a ‘single-dose’ of specific gene products, are characterized by short stature, infertility, lymphedema and often congenital heart disease [15]. An increased risk for bicuspid aortic valve and aortic coarctation in women affected by this condition suggests that full X chromosome expression is needed for proper cardiac development. The control of X-inactivation is an active area of research and much remains to be learned about how this process may contribute to sex-specific risk of CVD in women.
Sex steroid hormones
Genetic factors directed by the sex chromosomes interact and can be modulated by sex hormones which are produced by the chromosomal-directed sex organs, the ovaries and the testes. Effects of sex steroid hormones are either activational (reversible and dependent on the presence of a given hormone) or organizational (nonreversible and are relatively permanent following the removal of the hormones). Both effects are mediated either through genomic pathways requiring gene transcription, leading to either stimulation or inhibition of protein synthesis, or nongenomic pathways, which affect receptor and/or enzyme sensitivity/activity, causing rapid intracellular signaling cascades [16].
Estrogen
Estrogen is the major sex steroid in females. Two estrogen receptors (ER), ER-α and ER-β, are found in all cells including those of the CV system. However, their distribution seems to be tissue specific, and controversy remains regarding localization of ER-β to the mitochondria [17,18]. While ER-α is localized to the cell membrane and the nucleus [19,20], ER-β is observed in the nucleus and cytosol [19,20].
Some evidence suggests that ER-α in the vascular endothelium may reduce the risk of atherosclerosis [21], whereas ER-β may be a negative regulator of ER-α [22]. For example, expression of ER-β in areas of coronary calcifcation and plaque in both women and men [23,24] suggests an association with development of coronary artery disease. The ratio of ER-α to ER-β expression in other tissues may also alter cardiovascular risk. In human hearts, the ratio of ER-α to ER-β is greater in the left ventricle (64:36) than in the right ventricle (52:48) [25]. However, neither the age nor sex of these individuals was reported, making it difficult to hypothesize any sex-dependent or hormonal effects as they might relate to expression of estrogen receptors and development of cardiac myopathies or heart failure. Additionally, genetic polymorphisms in estrogen receptors may alter cardiovascular risk in females [26,27], but these effects need further confirmation and study.
Alteration in function of the vascular endothelium (i.e., endothelial dysfunction) occurs early in the development of atherosclerosis. Estrogen affects endothelial function through activational effects on endothelial nitric oxide synthase (eNOS) and production of endothelial nitric oxide (NO). These activational effects on eNOS occur through increased gene expression and/or a change in mRNA stability and phosphorylation of eNOS [28]. NO mediated effects are important in both males and females; as in a man with genetic defects in ER-a, flow-mediated arterial dilatation was absent in the presence of accelerated atherosclerosis [29].
In postmenopausal women, exogenous estrogen increases plasma NO and enhances endothelium-dependent relaxations [30]. Thus, either the ligand for the receptor or changes in the receptors themselves, as may occur through methylation of the estrogen receptor gene with aging [31], may affect development of vascular disease in women.
Other endothelium-derived factors which have vasodilatory effects may also be modulated by sex steroids. For example, endothelial-derived hyperpolarizing factor(s) (EDHF) facilitates relaxation of vascular smooth muscle through activation of ion channels or gap junction-dependent electrical signals [32–34] and may act synergistically with NO or compensate for its absence [35]. Estrogen-dependent EDHF-mediated relaxation may require ER-β [36]. Effects of EDHF may depend upon the anatomical location of the blood vessel, having greater significance in small compared with large arteries [37]. Additional investigations are required to better understand hormonal regulation of EDHF in relationship to CVD in both women and men.
Estrogen also modulates cellular function via calcium signaling through a rapid nongenomic pathway involving the G protein-coupled estrogen receptor (GPER). This receptor, previously called GPR30, is localized to the endoplasmic reticulum; its function has been reviewed in detail recently [38]. In brief, GPER binds estrogen, eliciting calcium mobilization and synthesis of inositol trisphosphate [39]. Consequences of modulation of this receptor across the lifespan requires further study in relationship to risk for development of CVD in women and men.
Estrogen affects angiogenesis, vasculogenesis and remodeling in response to shear stress via proliferation and migration of both endothelial cells (ECs) and vascular smooth muscle cells (VSMCs). In ECs, estrogen promotes proliferation and migration of cells through activation of MAPK via receptor ER-α [40,41] which may reduce development of intimal hyperplasia following mechanical damage to the endothelium [42]. In contrast, VSMCs treated or cultured with estrogen show blunted proliferation and migration via several mechanisms, including inhibition of PDGF signaling [43,44] and reduced MAPK activity via ER-β stimulation [40,41]. Additionally, activation of GPER may inhibit VSMC proliferation via ERK and Akt phosphorylation pathways [45]. Other effects of estrogen may include upregulation of PGC-1α expression leading to inhibition of oleic acid driven VSMC proliferation and migration [46]. Overall, estrogen-regulated genes involved in cell proliferation and migration may include caveolin-1, two LIM proteins, and Id3a [47]. While several studies have elucidated the biological mechanisms responsible for estrogen-dependent blunting of VSMC proliferation and migration, novel approaches to harness these pathways after estrogen loss remain to be studied.
Effects of estrogen on other cells of the cardiovascular system are under investigation. For example, estrogen receptor and MAPK-dependent mechanisms appear to increase proliferation of fibroblasts derived from adult females [48]. However, the consequences of these activities related to development of cardiac disease in women remain to be defined. In cardiac tissue, 36 estrogen-dependent genes were regulated in a sex-specific manner. Some of these genes affected myosin activity which would be manifested in altered contractility in male tissue only; this has implications for differences in development of ventricular dysfunction between men and women [49].
Other sex hormones
Effects of progesterone on components of the cardiovascular system have not been studied extensively [50]. Briefly, evidence suggests that progesterone enhances proliferation, migration and apoptosis of VSMCs in the female rat aorta [51]. Additionally, progesterone may inhibit calcium flux through L-type calcium channels, therefore attenuating smooth muscle contraction. However, these experiments were conducted using tissue from male animals, and effects in female tissue remain to be validated [52].
In addition, effects of testosterone on components of the cardiovascular system in women and men are controversial. The controversy reflects, in part, that testosterone can be converted to estrogen by aromatase which is present in many tissues (nerves, cardiac tissue, adipose tissues) [53–56]. The relationship among these sex steroids warrants further investigation given the use of testosterone products by both women and men and the use of aromatase inhibitors for treatment of breast cancer.
Female-specific conditions
The concept that the net effect of endogenous sex steroids is decelerated progression of CVD in premenopausal women compared with age-matched men is not without challenge [57]. However, changes in circulating levels of endogenous sex hormones such as those that occur during pregnancy and menopause can affect current and future cardiovascular risk through direct effects on the vasculature and cardiac muscle as well as indirect effects mediated through changes in metabolism and coagulation (i.e., glucose, lipids, coagulation proteins). Exogenous hormones in the form of hormonal contraceptives in premenopausal women and menopausal hormone therapy in postmenopausal women modulate the hormonal environment, and subsequently sex-specific cardiovascular risk.
Hormonal contraceptives
Oral contraceptive pills (OCPs) are used by up to 80% of women in the USA during their lifetime, and the introduction of exogenous sex hormones into the body in the presence of endogenously produced hormone is known to alter the cardiovascular ystem, which may place women at risk for increased adverse cardiovascular events [58,59]. OCPs are also used to treat irregular menses, menorrhagia, dysmenorrhea and acne. OCPs were first marketed in the 1960s, and since then multiple ‘generations’ of the drugs have been introduced. There are four generations of OCPs, all containing synthetic combinations of ethinyl estradiol and progestin which inhibit ovulation; first-generation OCPs have the greatest estrogen levels (150 μg) while fourth-generation OCPs contain the lowest concentrations (20 μg) [58,60]. The typical regimen of OCP use is for the individual to take active forms of the pills for 21 consecutive days, followed by 7 days of placebo pill (or no pills) [60]. In addition to pills, combined hormonal contraceptives can be delivered using vaginal hormonal rings, intrauterine devices, transdermal patches and subcutaneous implants. Progestin-only formulations also exist [58].
Chronic use of some formulations of OCPs may increase blood pressure [58,61]. While OCPs containing the highest levels of ethinyl estradiol have been associated with increased blood pressure [58], lower concentrations as found in newer generations of OCPs have also been linked to increases in 24-h ambulatory systolic blood pressure by up to 8 mmHg in healthy, normotensive women [62] as well as hypertensive women [63]. These increases may be due to increased production of angiotensinogen and subsequently angiotensin II [64]. Ethinyl estradiol, binds to estrogen receptors and has a long half-life, substantially enhances this process [65]. However, use of a combined OCP containing drospirenone, a synthetic progestin with anti-mineralocorticoid activity, for 6 months does not change blood pressure, heart rate or hemodynamic variables (cardiac output and total peripheral resistance) [66]. Drospirenone, being an analog of the diuretic spironolactone with anti-mineralocorticoid activity, is capable of counteracting estrogenic effects on the renin-angiotensin-aldosterone system. However, this type of OCP comes with a risk of its own: drospirenone-containing OCPs may increase the relative risk of venous thromboembolism [67]. In addition, one study has shown that drospirenone may alter the normal heart rhythm by prolonging the QT interval, placing individuals at risk for ventricular arrhythmias and sudden cardiac death [68]. This risk for a prolonged QT interval was greatest in women who do not take other QT-prolonging medications and do not have cardiovascular or metabolic comorbidities. While progestin-only formulations of hormonal contraceptives have not been associated with blood pressure elevation and hypertension risk [69], they have been linked with decreased endothelial function (endothelium-dependent vasodilation) [70]. Long-term use of these formulations is not recommended, as they cause hypoestrogenism and decreases in bone mineral density [71].
OCPs carry the risk for venous thrombosis, myocardial infarction and stroke [72]. However, the risk may be related to OCP formulation (dose) and an individual's underlying risk for CVD. In the Danish Cohort Study which followed women between 1995 and 2009, use of combined OCPs containing 20 μg of ethinyl estradiol increased the relative risk of both thrombotic stroke and myocardial infarction by 1.60 (95% CI: 1.37–1.86) and 1.40 (95% CI: 1.07–1.81), respectively, in comparison to non-OCP users. The progestin type in these pills did not vary the level of risk. In addition, progestin-only formulations did not increase the risk of either stroke or myocardial infarction [73]. While the absolute risk of stroke and myocardial infarction is fairly low with OCP use in healthy women, it remains important for clinicians to be aware of the factors which may modify this risk and which OCP formulations are safest to prescribe based on the patient's medical history.
Pregnancy
During pregnancy, a woman's cardiovascular system adapts to support the growing fetus. Increases in sex steroids (estradiol and progesterone) during pregnancy [74], in addition to embryonic production of substances, such as human chorionic gonadotropin (hCG), are responsible for promoting progesterone production by the maternal ovaries in order to maintain the pregnancy until progesterone generation can be accomplished by the placenta [75]. Throughout the duration of pregnancy, estrogen increases, as the placenta synthesizes this sex hormone from circulating androgens [75].
By the sixth week of gestation, several hemodynamic changes occur in the woman's body including 45% expansion of plasma volume and 20–30% increase in red blood cells mass [75]. Decreases in mean arterial pressure and systemic vascular resistance are seen early in pregnancy due to peripheral vasodilation likely mediated by increased production of progesterone, prostaglandins and nitric oxide and their actions on vascular smooth muscle [74–76]). There is also a 30–50% increase in cardiac output, mediated by increased stroke volume and heart rate [74,75]. In the renal system, kidney plasma flow and glomerular filtration rate increase by 50% or more while vascular resistance decreases in the first trimester of pregnancy [74] and plasma renin activity and aldosterone concentrations increase. During normal pregnancy, there is an increase in both low- and high-density lipoprotein cholesterols and triglycerides which provide energy to the growing fetus [75,77].
Vascular, metabolic and immunologic adaptions that occur to a women's body during pregnancy have been collectively viewed as a ‘stress test’ for the cardiovascular system. Complications, such as a hypertensive disorder or gestational diabetes, may arise and can place a woman at long-term risk for the development of CVD. It is possible that pregnancy may simply reveal preexistent cardiovascular dysfunction rather than cause it, as many risk factors (e.g., hypertension, insulin resistance or diabetes, obesity and dyslipidemia) are shared between CVD and pregnancy disorders [78] (Figure 2).

Hypertensive disorders of pregnancy affect 6–8% of pregnancies in the USA and include the conditions of gestational hypertension, chronic hypertension and pre-eclampsia [78]. Gestational hypertension is defined as new onset of hypertension (≥140/90 mmHg) after 20 weeks' gestation in a woman who was originally normotensive [79]. Women who develop hypertension prior to 20 weeks of gestation are diagnosed with chronic hypertension [79]. Women who suffer from severe gestational hypertension (≥160/110 mmHg) are at a greater risk of progressing to the development of pre-eclampsia and may be at increased risk for developing CVD later in life [80].
Three percent of pregnancies are affected by pre-eclampsia, classified by the new onset of hypertension (≥140/90 mmHg) after 20 weeks of gestation and either proteinuria (0.3 g/24 h) [78] or end-organ dysfunction (i.e., renal insufficiency, thrombocytopenia, impaired liver function, pulmonary edema or cerebral/visual symptoms) [79]. Causes of pre-eclampsia are multifactorial and may involve poor placental perfusion and hypoxia, accompanied by endothelial dysfunction in the setting of hypoxia, increased oxidative stress and an imbalance in vasoactive factors [81–84].
The course of pre-eclampsia may be complicated by an atherosclerosis-like phenotype with endothelial dysfunction, vasoconstriction and activation of the coagulation cascade [85]. The condition is also associated with either normal or increased triglyceride levels and decreased levels of high-density lipoprotein cholesterol, which may further promote endothelial dysfunction in this condition [77]. Left untreated, preeclampsia progresses to eclampsia, which is characterized by generalized seizures [79].
Historically, it was believed that the adverse cardiovascular effects of pre-eclampsia and other hypertensive pregnancy disorders were limited to the time of pregnancy; however, there is growing consensus that CVD risks persist to later in life far beyond the affected pregnancy [86]. A meta-analysis conducted by Bellamy et al. showed that in comparison to women with normal pregnancies, women who suffered from pre-eclampsia had a 3.70 (95% CI: 2.70–5.05) relative risk for developing hypertension 14 years after pregnancy, a 2.16 (95% CI: 1.86–2.52) relative risk for ischemic heart disease after 12 years, a 1.81 (95% CI: 1.45–2.27) relative risk for stroke after 10 years and a 1.79 (95% CI: 1.37–2.33) relative risk for venous thromboembolism after 5 years [87]. Earlier occurrence of pre-eclampsia in pregnancy is associated with poorer outcomes; in addition, the severity of pre-eclampsia is correlated with the severity of CVD later in life [78].
Gestational diabetes mellitus may also increase future risk of CVD, as suggested by retrospective and cross-sectional studies. However, longitudinal studies are necessary to better confirm and this relationship [88,89].
Because disease during pregnancy is associated with both short- and long-term CVD risks, it is important that clinicians provide follow-up care and counseling to postpartum women, as well as for primary care providers to obtain a pregnancy history of their patients during the course of their care [86]. Such information may allow for the early identification and interventions to slow or limit development of severe CVD. Challenges to understanding causes and effects of pregnancy on the cardiovascular system of the mother include the absence of characterization of the cardiovascular health of women prior to pregnancy, longitudinal follow-up of cardiovascular function of women after the pregnancy and uniform information about pregnancy disorders during pregnancy (uniform recording in the medical record and changes in coding for various conditions). While some experimental models of pregnancy-associated hypertension exist [90], they were developed in animals of uniform genetic background and absence of pre-existing cardiovascular risk factors which may limit the translation of the results to humans. Additional studies are needed to characterize the etiology of hypertension in women and female animals as most basic science studies of hypertension have been derived from studies of male spontaneous hypertensive rats or salt-sensitive rats.
Menopause
Given the constellation of hormonal effects on the cardiovascular system discussed above (Figure 3), it might be predicted that following menopause, the risk for hypertension, stroke, coronary artery disease and heart failure would increase. In one study of rural Chinese women, the percentage of postmenopausal women with hypertension was twice that of premenopausal women (49.3 vs 21.9%) even when these data were corrected for confounders including age, hyperlipidemia, body mass index and exercise [91]. In addition, it might be expected that mortality risk and risk for CVD would accompany early ovarian insufficiency or surgical unilateral or bilateral oophorectomy [92,93].

In addition to the loss of direct effects of estrogen on the cardiovascular system, indirect effects resulting from loss of estrogenic modulation of immunological and prothrombotic proteins [94] by the liver, altered lipid and glucose metabolism [95,96] and autonomic function [94] contribute to the increased risk for women. Although not a specific topic in this review, interactions between the autonomic nervous system and perhaps the renin angiotensin system in women warrant further investigation.
Menopausal hormone therapy
Because risk for development of CVD increases at menopause, it was hypothesized that use of exogenous hormones might reduce this risk. Multiple observational, epidemiological and case-control studies of women taking menopausal hormone treatments (MHT) for relief of menopausal symptoms supported this hypothesis [97]. However, the conclusions of these studies were criticized for representing women who were proactive about their medical care (i.e., ‘healthy user bias’). A randomized, prospective clinical trial, the Women's Health Initiative (WHI), was designed to test the hypothesis that MHT would reduce CVD. However, the WHI enrolled women about 8 years past menopause without menopausal symptoms, who in clinical practice would not have been prescribed MHT. Thus, it should not be surprising that the outcomes of the WHI did not support the original hypothesis [98]. In the WHI, women randomized to continuous oral conjugated equine estrogen (0.625 mg/day) plus medroxyprogesterone acetate had a higher relative risk of developing coronary heart disease, pulmonary embolism and stroke compared with women randomized to placebo treatment. Subsequent, subanalyses showed that women less than 60 years of age were not adversely affected by the treatment as compared with older women who were many years past menopause [99]. Interpretation and conclusions of randomized studies to investigate the hypothesis that MHT is protective against CVD are challenged by several parameters that differ among studies: age and health of participants, type of MHT (oral, transdermal, conjugated equine estrogen, 17-β estradiol or esterified estrogens with or without continuous or pulsed synthetic or natural progestogens) and duration of treatment. These divergent findings among studies could be due to pre-existing CVD in older women [99], a conclusion supporting the hypothesis developed from experimental studies suggesting that there is an ideal time during which MHT is most beneficial in preventing progression of atherosclerosis [99]. For example, in another prospective, randomized, but open-label trial in which women who were within 3 to 24 months of menopause were given 17-β estradiol with and without norethisterone acetate or no treatment [100], women randomized to MHT had a lower rate of death, myocardial infarction and heart failure (assessed as a composite endpoint), and did not have increases in pulmonary embolism or deep venous thrombosis across 10 years of treatment. This study was criticized for having an open-label design with no placebo [101,102]. However, MHT to women with oophorectomy prior to the natural age of menopause also reduced mortality from CVD compared with untreated women with either unilateral or bilateral oophorectomy [92,93].
Effects of MHT on blood pressure are mixed and may depend on age of initiation of therapy, type, dose and duration of treatment as well as initial health status such as existing conventional risk factors (e.g., body mass index, blood pressure, lipid profile, fasting glucose and smoking status) [98,103–105]. Current treatment guidelines do not recommend the use of MHT for stroke prevention or recurrence [97,106–107].
Although there is a need to better understand these effects of MHT on development of peripheral and cerebral CVD, it is unlikely that additional large prospective trials will be performed regarding dose and formulation given the existing controversy and fear of potential harm regarding use of MHT and expense associated with longitudinal follow-up. While there appears to be a consensus among multiple studies that MHT is most appropriately used in women closely following menopause and for short-term duration for menopausal symptom management, controversy still exists. In addition, the question remains how early during the menopausal transition should MHT be initiated. Clinicians must consider the risk factors which may modulate MHT effects and discuss the potential risks and benefits with their patients including, but not limited to, age at the onset of menopause, time since menopause, personal health history and the severity of symptoms and the degree to which they affect quality of life [108]. Thus, consensus is that MHT should be used close to the onset of menopause (within 6–10 years), including women with oophorectomy prior to the age of natural menopause, in women under the age of 60 years old and for a short duration of time [99,107–110].
Traditional risk factors
While there are numerous other risk factors for the development of CVD which may display sex differences, such as physical activity, cholesterol levels and obesity, these will not be discussed here and have been reviewed elsewhere [111,112]. US guidelines have recently been published regarding proper assessment and treatment of atherosclerotic CVD (ASCVD), including modification of lifestyle, blood cholesterol and obesity. Specific to women, statin treatment decisions should now be based on ASCVD instead of LDL-cholesterol levels, as was done previously [113].
Female-specific diagnosis & treatment
There are multiple variables that influence the development of treatment guidelines for CVD (Figure 3) [114]. In general, and unfortunately, sex as a biological variable is not often taken into account in the development of guidelines for diagnosis and treatment of CVDs [115]. The American Heart Association's guidelines for stroke prevention in women indicate that a women's pregnancy history is a potential factor for guiding CV monitoring and treatment [106], but it is unclear how these suggestions would be implemented in general clinical practice. The American College of Cardiology and American Heart Association's Guidelines on treatment of blood cholesterol to reduce atherosclerotic cardiovascular risk in adults suggests that treatment decisions for women be based on the ASCVD risk [113]. However, further improvements to these guidelines may be necessary, as the ASCVD risk was based on literature from preclinical studies, the majority of which used male animals and a male pattern of atherosclerotic disease. In addition, sex- and age-specific laboratory values for lipids were absent, and the clinical trials included more men than women and often reported data that were not dichotomized by sex [116]. With new policies instituted by the US National Institutes of Health [117], as well as those by other granting agencies around the world to ensure that preclinical studies consider sex differences in experimental design, the evidence basis to build sex-specific recommendations should improve in the future. The following evaluates female-specific diagnosis and treatment approaches for four major categories of CVD.
Hypertension
The American Heart Association includes hypertension in its definition of total CVD [118]. Proper diagnosis and treatment of hypertension early and effectively can prevent cardiovascular complications such as stroke, heart failure, myocardial infarction and chronic kidney disease [119].
Hypertensive women have a higher prevalence of isolated systolic hypertension than men [120], and 44.8% of women versus 51.1% of men have satisfactory management of their hypertension, defined as a systolic blood pressure less than 140 mmHg and diastolic blood less than 90 mmHg pressure for most individuals. However, this disparity may be explained by a higher prevalence of isolated office (‘white-coat’) hypertension in women than in men [121] and that management of hypertension in women is better assessed by 24-h ambulatory blood pressure monitoring [122]. Daugherty et al. suggested that management of blood pressure is also influenced by age [122]. For example until the age of 65 years, women have better blood pressure management than age-matched men [121].
The 2014 Evidence-based Guidelines for the Management of High Blood Pressure in Adults provide recommended dosing targets for the general classes of drugs affecting blood pressure: angiotensin converting enzymes, angiotensin receptor blockers, β-adrenergic receptor antagonists, calcium channel antagonists and thiazide-type diuretics [123]. However, these guidelines allow discretion on the part of the physician for selection and dosing but do not provide information relative to first-line approaches based on potential differences in efficacy or dosing for women to reach target blood pressure control. There are conflicting reports on hypertension outcomes and the need for sex-specific guidelines for hypertension. Results from a retrospective cohort study have suggested that outcomes for hypertension differ based on the sex of an individual [121]. This study reported that women with hypertension were more likely to have a lower glomerular filtration rate than men, as well as more likely to develop chronic kidney disease. Also, hypertensive women had a lower risk of stroke and myocardial infarction than men but had a higher risk of developing heart failure with preserved ejection fraction [119]. However, Peters et al. show that systolic blood pressure is not associated with higher risks of stroke or ischemic heart disease in one sex versus the other [124]. A review of randomized clinical trials by Turnbull et al. concluded that differential hypertension treatment for men and women is unnecessary [125] because the efficacy of the various classes of pharmacological therapies (see ‘2014 Evidence-based Guidelines’ above) to prevent adverse cardiovascular events is similar in both populations [125].
In recently menopausal women enrolled in the Kronos Early Estrogen Prevention Study (KEEPS), blood pressure was the single variable that associated with carotid intimal medial thickening [126] prior to randomization to hormonal treatments. This finding is significant in that both systolic and diastolic blood pressure were within current normative ranges and women were at low risk for developing CVD based on age, body mass index and metabolic parameters of glucose and lipids. It is important to remember that blood pressure is a continuous variable and to consider both short-term and long-term consequences of elevated blood pressure on end organ function (damage).
There are sex disparities in the economic aspects of hypertension, particularly in the areas of prescription medication costs, inpatient and outpatient expenditures, and office-related care costs [127]. For example, the average inpatient expenditure for hypertension in women is $61 to $159 greater per individual than in men until the age of 70 years when these costs become slightly higher in men. Furthermore, expenditures for medication, outpatient care and office care are greater in women until the sixth decade of life when this disparity is reversed [127].
As mentioned above, specific episodes of hypertension, such as hypertensive pregnancy disorders, place a woman at life-long risk for other adverse cardiovascular events. It is imperative to better understand sex differences in underlying etiology of hypertension to better direct treatment and reduce associated adverse events associated with chronic hypertension in women.
Stroke
The prevalence of stroke in the USA is 3%, with approximately 42,500 women being affected each year [2]. Eighty-seven percent of strokes are classified as ischemic and the remaining are hemorrhagic (intra-cerebral or subarachnoid) [106]. Men suffer from ischemic stroke more than women, with the exception of women over the age of 85 years old; following menopause, women have a higher incidence of subarachnoid hemorrhagic stroke [106]. Women and men who suffer from an acute stroke do not always present similarly. While the rank of specific signs and symptoms of a stroke may be the same between women and men, their actual prevalence may differ. For example, in a population-based assessment, the most frequent symptoms reported in both men and women with stroke onset are limb and facial weakness, followed by dysarthria [128]. The fourth and fifth most common symptoms of stroke are paresthesia and dysphasia, respectively; however, paresthesia is seen more frequently in men, and dysphasia is more common in women. Women are also more likely to present with loss of consciousness and incontinence and less likely to have ataxia and dysarthria than men. There are conflicting reports on whether men or women have higher rates of visual impairment with stroke [128,129].
Currently, guidelines for the treatment of stroke in men and women are the same [106]. Thrombolytic therapy, in the form of tissue-type plasminogen activator (tPA), is widely used following the onset of ischemic stroke and has been shown to improve neurological function when given within 3.0–4.5 h of initial symptoms [130]. Women may be less likely to receive intravenous tPA than men after presenting to the emergency department [131]; however, recent data from the Promoting Acute Thrombolysis for Ischaemic Stroke (PRACTISE) study suggests that the number of women and men treated with this therapy is similar [132]. Any differences in treatment may be due to the longer amount of time it takes for women to arrive at an emergency department following stroke onset, varied symptom prevalence between women and men, the tendency for women to present with stroke at an older age (tPA should be used with caution in individuals over the age of 80 years old [130]), or contraindications to the medication [132,133]. Men are more likely to receive cerebrovascular and cardiac revascularization/reperfusion therapies, which may also be related to the amount of time it takes for an individual to reach the hospital, in addition to physician bias and assumptions on how a female patient may or may not want to be treated [132]. There is some evidence that newly prescribed medications, such as antiplatelet and anticoagulant drugs, are prescribed at the same rate upon discharge in women versus men [128].
Sex differences in stroke outcomes can often be explained by confounding variables. For example, women tend to be older during incident stroke, which might explain why women subsequently have a higher rate of handicap and show less independence in their activities of daily living [2]. In addition, women often have greater stroke severity, morbidity and chronic conditions (e.g., arthritis and dementia) prior to stroke onset, which can influence poststroke status [128,132,134]. Functional outcomes in women and men are similar 3 months after thrombolytic therapy [132,135]. However, men have a higher mortality with this treatment and are at a greater risk of suffering from intracranial hemorrhage as a side effect [135].
Stroke mortality is age related, with women having a lower age-adjusted mortality from stroke than men, with the exception of women over the age of 85 years [106]. In contrast, in-hospital mortality may be greater in elderly men over the age of 85 [136].
Many studies report sex differences in stroke diagnosis, treatment and outcomes; however, they do not always take into account confounding factors or covariates which may explain these findings [134]. In addition, more studies need to designate the presence or absence of sex differences as a primary outcome variable to better assess how sex and gender disparities are related to stroke outcomes.
Coronary artery disease
There is a positive correlation between coronary atherosclerosis and myocardial infarction [137], and while the death rate due to coronary artery disease (CAD) has fallen overall from 1999 to 2010 [118], there are various sex-dependent differences in development, diagnosis and treatment outcomes for CAD. The Women's Ischemia Syndrome Evaluation (WISE) Study identified patterns of coronary artery plaque of women which differed from that observed in men [138,139]. In women, plaque was diffuse compared with the more defined, restrictive narrowing of arteries of men. The slow erosion of plaque in women leads to more microvascular disease compared with what has been defined as an ‘explosion’ or abrupt rupture of an obstruction in men. Factors contributing to differences in the development of plaque in men and women remain to be elucidated. Accumulation of lipids within the vascular wall accompanied by inflammation is considered a fundamental characteristic. However, most of the studies of these mechanisms were developed in male experimental animals and it is unclear the extent to which these same processes are present in women. With the focus on this pattern of disease development, a now common front line treatment to prevent CAD is the use of statins. Guidelines for statin therapy are not based on sex-specific normative values for lipids and treatment efficacy is not evaluated by sex. A meta-analysis of statin studies found that the risk of Type II diabetes with statin use was associated with the number of women enrolled in the trial [116]. As Type II diabetes carries risk for cardiovascular complications on its own, additional research is needed to evaluate the long-term use of statins and to resolve the controversy of the use of this therapy for primary prevention of CAD in women.
Symptoms of CAD vary by sex. Chest pain is less predictive of CAD in women than in men [140,141]. Women often experience prodromal symptoms before or during acute myocardial infarction, including unusual fatigue, sleep disturbance and shortness of breath [140]. One study found that black women are less likely to experience chest pain but instead report stomach discomfort more often [142]. Additionally, women with CAD undergoing percutaneous coronary intervention may be more likely than men to report nonchest pain during 60-s balloon inflation [143]. These variable symptoms must be fully recognized in order to accurately diagnose CAD in women, especially because many women may undergo an invasive angiography procedure (clinically indicated due to chest pain or suspected myocardial ischemia) when they in fact do not have CAD [144]. A noninvasive computed tomographic technique may diagnose women with chest pain with at least as much sensitivity as it does in men [145].
Exercise stress testing is the most popular method of CAD diagnosis in women [146]. While exercise ECG may be less successful in identifying those with CAD in younger (<65 years) women [147], adding electrocardiography to exercise testing might improve sensitivity and specificity overall [148,149], perhaps even above that of the male population [150]. Additionally, adding high-frequency components of the QRS complexes (HFQRS) analysis to the standard exercise ECG may also improve specificity and accuracy [151]. The fact that exercise ECG was considered of lower value until more studies involving female subjects were performed highlights the importance of the need to include women in clinical studies related to CVD. Other methods that have shown promise in diagnosing females with CAD include single-photon emission computed tomographic (SPECT) imaging, which may be able to diagnose CAD in low-risk women, and perhaps even more effectively in high-risk women than in high-risk men [152,153]. A meta-analysis determined that 26 SPECT studies had a mean sensitivity of 84.2% and a mean specificity of 78.7% in women [154]. Recently, cardiovascular magnetic resonance (CMR) offers a similar specificity but higher sensitivity than SPECT in the diagnosis of CAD [155]. These studies have shown clear sex differences in the usefulness of various CAD diagnosis strategies, and this knowledge can lead to better treatment and outcomes for women with CAD as there are sex disparities in the treatment of CAD.
Women admitted for acute coronary syndromes appear more likely than men to die in the hospital [156], and this risk may be even greater in black women [157]. While reasons are unclear, a coronary artery bypass grafting procedure carries a higher mortality in women than men [158]. A percutaneous coronary intervention procedure may also lead to higher 1-year mortality in women, but women also present at older age with more comorbidities [159]. Drug-eluting stents appear to be more effective in comparison to bare metal stents, especially in women [160,161]. Regarding cardiovascular drugs, women may be prescribed lipid-modifying agents and ACE inhibitors less than men [162], and sex differences in drug prescriptions persist throughout the lifetime [163]. Additional research and attention to sex in treatment modalities will improve efficiencies and outcomes for both men and women, along with reducing costs.
Heart failure
Evidence is constantly emerging that the diagnosis and treatment of heart failure should incorporate sex-specific considerations [164]. Women are more likely to develop heart failure with preserved ejection fraction (HFpEF) [165], but effective treatments for this form of heart failure do not yet exist [166]. While the prevalence of heart failure is greater in men, women carry a higher mortality [167]. Therefore, better diagnosis and treatment of heart failure specific to women is essential.
Methods to diagnose heart failure have sex-specific variables which must be considered. For example, men and women may have different predictors of incident heart failure during an electrocardiographic exam [168]. Also, cardiac catheterization for CAD, as well as baseline echocardiography of patients with HFpEF, elucidate differences in arterial and left ventricular characteristics between the sexes including increased arterial stiffness and greater diastolic dysfunction in women, both of which are hypothesized to contribute to the development of HFpEF in this sex [169,170]. These functional differences are consistent with anatomical differences in number of cardiomyocytes, and intracellular contractile mechanisms between males and females [49]. Additionally, peak oxygen consumption (VO2max) testing for heart transplant candidacy uses VO2 values based on males, but it is well known that females have a lower VO2max, such that “a woman with [heart failure] and peak VO2 of 9 ml/kg/min has the same 2-year outcome as a man with peak VO2 of 14 ml/kg/min” [171]. These sex-dependent differences are crucial in effective diagnosis, as well as future treatment, of women with heart failure.
Although men and women currently undergo similar treatment for heart failure, the outcome of a given therapy is sex-specific [172]. A recent review identifies the need for comparative effectiveness research (CER) for these treatment regimens as characteristics of study participants often do not mirror heart failure patients in clinical practice (i.e., sex, age and comorbidities) [173]. For example, 60% of heart failure patients are female, yet only 20–30% of participants in clinical trials are female [173]. This may represent, in part, the absence of differentiation between heart failure with reduced compared with preserved ejection fraction. Because cardiac anatomy and function reflect sex-specific characteristics, it is imperative that future trials use a CER approach in which study participants are comparable to those patients seen in the clinic.
Cardiac rehabilitation is a successful treatment for both sexes, but may be more beneficial for women than men [174]. Cardiac resynchronization therapy is also more successful in women, including the ability to decrease all-cause mortality [175–178]. While the reasons for these differences are not completely understood [179], the presence of nonischemic cardiomyopathy may contribute to the enhanced outcome seen in women [177,178]. These studies point to the importance of personalized treatment based on the type of heart failure and sex of the patient.
A study of trends in cardiac medications from 1995 to 2004 found that heart failure medications are utilized differently by men and women. There are sex-specific effects of various medications, including ACE inhibitors, β-blockers, angiotensin receptor blockers and aldosterone antagonists (Table 1) [180]. While evidence suggests that digoxin therapy leads to higher mortality in women this conclusion is controversial [180] and needs further study. Additionally, effectiveness of β-adrenergic receptor antagonists may depend upon the type and age of the patient with HFpEF [181,182]. Interestingly, testosterone therapy may be a promising treatment for women with heart failure [183] but as mentioned in earlier sections, effects of testosterone therapies, in general, need rigorous testing in regard to direct effects as well as indirect effects related to aromatization to estrogen.
Post hoc analyses of pharmacological therapies for cardiovascular disease by sex.
African–Americans.
All analyses are post hoc, as there are no intentional studies of sex differences in the effects of pharmacological therapies.
Adapted with permission from [180] © Elsevier (2013).
Future perspective
Awareness and research into sex differences in physiology and pathophysiology will facilitate translation of knowledge obtained from basic science to clinical practice and improved health outcomes for women and men. Genetic studies will include evaluation of X- and Y-linked genes in the phenotypic expression and inheritability of disease. Reproductive history will become a standard part of the medical record and cardiovascular risk factor assessment for women that can be used to target early intervention strategies to limit progression of CVD. Pharmacological interventions developed and tested with attention to age, sex and hormonal status will reduce treatment side effects and maximize benefit.
Executive summary
Women have often been under-represented in cardiovascular research.
Sex-specific biological factors such as pregnancy-related disorders and menopause are important to consider in the risk assessment, diagnosis and treatment of CVD in women.
Sex chromosomes characterize individuals as female (XX) and males (XY). In female somatic cells, one of the two X chromosomes randomly undergoes X-inactivation and can result in differential expression of phenotypes in health and disease.
Sex steroid hormonal effeclis
– Estrogen exerts activational and organizational effects on the female cardiovascular system via nuclear and membrane estrogen receptors and endoplasmic reticulum localized G protein-coupled estrogen receptor 1 (GPER).
– Estrogen promotes proliferation and migration of vascular endothelial cells and reduces contractility of vascular smooth muscle.
– High-density lipoproteins (HDL) aid in cholesterol transport away from arteries, and postmenopausal women show lower levels of HDL than premenopausal women, perhaps contributing to atherosclerosis progression in postmenopausal women. Additionally, HDL stimulates eNOS activity.
– Progesterone also has activational effects on the cardiovascular system, but more research is necessary to understand them fully.
Increased risk of hypertension, venous thrombosis, myocardial infarction, stroke and arrhythmia with use of oral contraceptives depends on the type, and clinicians should prescribe these medications in the context of an individual's medical history.
Hypertensive pregnancy disorders (gestational hypertension, chronic hypertension and pre-eclampsia) place women at risk for cardiovascular complications not only during pregnancy but also later in life.
Women who undergo premature ovarian failure of oophorectomy prior to the age of natural menopause have increased risk for developing CVD.
Controversy remains regarding the long-term cardiovascular benefit of menopausal hormone treatments. Current practice guidelines recommend use of these products for relief of menopausal symptoms and not for prevention of CVD.
Hypertensilin
– Women have a higher incidence of isolated systolic hypertension and have poorer management of hypertension (when based on blood pressure alone) than men.
– While some studies suggest the need for sex-specific treatment guidelines for hypertension, there is strong evidence against this concept, as review of the literature has not shown differences in outcomes or treatment efficacy based on sex.
Strolie
– The prevalence of specific stroke symptoms differs between men and women, and clinicians should be cognizant of such differences when evaluating patients with a stroke-like presentation.
– Guidelines for stroke treatment are the same for men and women. Any variations seen in stroke treatment and outcomes in men versus women are likely due to confounding factors, such as time from stroke onset to presentation to hospital, age at the time of stroke and comorbidities.
Coronary artery disealie
– Coronary artery disease symptoms vary between women and men, as women are more likely to experience prodromal symptoms prior to a heart attack. Effective diagnosis methods also differ for women; exercise stress testing with electrocardiography or single-photon emission computed tomographic (SPECT) imaging show promising sensitivity and specificity in the female population.
– Successful treatment of coronary artery disease varies between men and women, including differential benefits of coronary artery bypass grafting, percutaneous coronary intervention, stents and cardiovascular drugs.
Heart failulie
– Women are more likely to develop heart failure with preserved ejection fraction (HFpEF) and experience a higher mortality due to heart failure. Diagnostic strategies specific to HFpEF need to be developed and the disparity in outcome must be recognized by clinicians.
– Women undergoing treatment for heart failure are more likely to benefit from cardiac rehabilitation or cardiac resynchronization therapy.
Financial & competing interests disclosure
Support for this article was derived from research grants to VM Miller from the National Institutes of Health Ag 44170 and HD65987 and the Mayo Foundation, to RE Harvey from the American Heart Association 14PRE18040000 and the Mayo Clinic Medical Scientist Training Program and to KE Coffman from the Mayo Clinic College of Medicine, Mayo Graduate School. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.
No writing assistance was utilized in the production of this manuscript.
Ethical conduct
The authors state that they have obtained appropriate institutional review board approval or have followed the principles outlined in the Declaration of Helsinki for all human or anima experimental investigations. In addition, for investigations involving human subjects, informed consent has been obtained from the participants involved.
