Cardiovascular risk in women: the cardiologist's perspective

The incidence of coronary heart disease (CHD) is much lower in younger women than in age‐matched men, and this has led to the popular misconception that cardiovascular disease is a disease of men, and is relatively rare in women. This, however, is not the case. The incidence of CHD may be much lower in young women than in men of the same age, up to the age of 65, with the risk of dying from CHD being 3.5‐fold higher in men than in women. After age 65, however, the risk equalizes for both sexes.1 Overall, CHD is a very significant cause of morbidity and mortality amongst women. The reasons for the lower incidence of CHD in younger women are unclear. Because of the increased incidence with age, it has been proposed that oestrogen may be responsible for the protective effects seen amongst younger women. Loss of this protection after menopause may therefore be responsible for the increased CHD incidence at older age. However, Heller et al.2 examined the mortality data of the Registrar General of England and Wales over a 5‐year period (1970–4) and suggested that there was no real acceleration in the increase of CHD death in women aged over 50 years (that is, postmenopausal women) but that the rate of increase for men over the age of 50 was in fact lower than in younger men. They therefore proposed that around the age of 50 years it was the men who may have lost some ‘protective’ factor, thus putting them at higher risk of CHD, rather than the increased CHD risk amongst women being due to their loss of a ‘protective’ factor! This viewpoint is contrary to popular belief, but highlights the uncertainty over the problem of cardiovascular risk in women, ranging from what the actual risk is, to the role of hormone replacement therapy (HRT) in postmenopausal women. Indeed, the replacement of oestrogen following menopause has been proposed as ‘cardiovascular protection’ for the older woman. Much of the evidence for the use of HRT has come from observational population studies, which have been criticized for using healthy cohorts and to be unrepresentative of the general population. It is only very recently, after decades of HRT use that any well‐conducted prospective randomized trials have been performed to confirm or refute the role of HRT. The misconception that women are at low risk of CHD has been reinforced by many large clinical trials of cardiovascular disease in which women are under‐represented, with the majority of trials being conducted in White, middle‐aged men.3–5 In addition to, or as a result of, the treatment bias, women with CHD also have a worse outcome than men. The precise reasons for the poorer outcomes are difficult to ascertain, but women tend to present at an older age and have more complicating factors such as diabetes, hypertension and heart failure than men.6,7 In an age when evidence‐based medicine is promoted, we find that the evidence needed for good clinical practice is often not available for women but must be extrapolated from the findings from male‐dominated studies. For example, the West of Scotland Coronary Prevention Study (WOSCOPS), a large study examining the use of cholesterol‐lowering using a statin in the primary prevention of CHD, was conducted solely in men.8 This trial showed a dramatic reduction in cardiovascular end points in subjects receiving the statin, but whether this data can be confidently extrapolated to women remains unknown. Other lipid‐lowering studies have excluded women9,10 while other studies have recruited far greater numbers of men than women. This bias against women in clinical trials may perhaps influence the clinical care of women by physicians. For example, it has been found that amongst patients with a diagnosis of angina in a London area, men were more likely to receive surgical intervention than women.11 Furthermore, Clarke et al.12 reported that women with myocardial infarction were less likely than men to be admitted to coronary care units, were less likely to receive thrombolytic therapy than were men, and in addition, were less likely to be discharged on secondary prophylaxis. Whether the arrival of the menopause alters a woman's risk of CHD remains controversial. Many, but not all, studies have demonstrated an increase in rates of CHD and atherosclerosis following the menopause but, since natural menopause is highly correlated with age, as is the incidence of CHD, this association may well be spurious.13 Natural menopause has also been found to occur earlier in smokers than in non‐smokers, which is not taken into account in population studies and further confuses the association between CHD and menopause.14 Also the number of subjects undergoing an early, natural or surgical menopause in these studies is small and thus many studies analyse the data for natural and surgical menopause together. Nevertheless, an abrupt cessation of ovarian function and oestrogen production following a surgical menopause may have an entirely different effect to a naturally occurring menopause in which ovarian function declines gradually. The evidence for the association between CHD and menopause comes from many different types of study. Population studies have given an epidemiologist's perspective, but even these have provided conflicting results. For example in the Framingham study, an analysis of a cohort of 2873 women followed up for 24 years reported that the incidence of cardiovascular disease in men under 60 years was twice that of women aged under 60.13,14 This study also reported that the incidence of cardiovascular disease amongst postmenopausal women up to the age of 55 years was twice that found in premenopausal women, with 20 cardiovascular events occurring in premenopausal women compared to 70 events among postmenopausal women. Their definition of ‘cardiovascular events’ has however been criticized, as it included patients with unspecified chest pain and leg pain. When the harder endpoints of stroke, heart attack and congestive cardiac failure are used in the analysis, the number of events was much smaller, with only 9 and 39 events in the two groups, respectively, with the difference being described as being ‘almost statistically significant’. The analysis from the Framingham study used 5‐year age intervals and this has been shown in other studies to be far too long.15 Smoking was also not included in the analysis, which is another factor which may have affected the results. In the Nurses’ Health Study,15 121 700 female nurses aged 30–55 were contacted by a mailed questionnaire every 2 years for 6 years. They were asked to supply information about their age, smoking habit, height, weight, menopausal status and history of CHD, diabetes, hypertension and hyperlipidaemia. After adjusting for age in 5‐year categories, the risk of CHD among women with a natural menopause was increased when compared with the risk amongst premenopausal women, but when this analysis was repeated, controlling for age in 1‐year intervals, there was a reduction in the risk ratio from 1.7 to 1.2. When cigarette smoking was also included in the analysis, the risk was reduced to 1.0. There may be differences between risk of CHD in patients undergoing a surgical menopause (by ‘castration’ or bilateral oophorectomy) compared to those undergoing a natural menopause. In one early series of autopsy reports,16 there was an increase in atherosclerosis amongst women who had surgically‐induced menopause (bilateral oophorectomy). In an angiographic study, however, no association was observed between the presence of CHD and the presence or absence of ovaries.17 A cohort study of 146 women in Malmo who had bilateral oophorectomy at an early age (between 15 and 30 years) reported that the difference in myocardial infarction (MI) and angina occurring under the age of 70 in the surgical menopause group was significant when compared to age‐matched controls, although the numbers involved were small (16 compared with 1).18 In the Nurses’ Health Study,15 women who had surgical menopause with bilateral oophorectomy and did not use oestrogen replacement therapy were found to be at an increased risk of CHD when compared to age‐matched controls (risk ratio 2.2); even when other risk factors were controlled for, this risk was only slightly reduced (risk ratio 1.7). These findings are in contrast to the results found with the natural menopause. However, some studies suggest that a long ‘incubation period’ precedes the adverse effects of an early menopause. For example, Parrish et al.19 studied the autopsy records of 80 patients who had a bilateral oophorectomy before the age of 50 and compared them with 80 age‐matched controls: those undergoing surgical menopause 10 years before the age of 50 had significantly more severe coronary atherosclerosis than did those with intact ovaries. By contrast, this was not found to be the case in women who had a surgical menopause 0 to 9 years before the age of 50. A follow up study of women who had X‐irradiation treatment for metropathia haemorrhagica20 found a small excess of deaths from CHD 10 years following treatment. The study by Rosenberg et al.21 compared the incidence of MI among women under 56 years of age with that in a control group, and found that the relative risk of developing an MI increased with decreasing age at bilateral oophorectomy; this risk was 7.2 in women who had bilateral oophorectomy before the age of 35 compared to pre‐menopausal women. This increased risk seemed to occur in spite of cigarette smoking at the time of MI. No such increase in risk was seen in women experiencing a natural menopause, although the numbers were small. In the Renfrew and Paisley survey of the relationship between coronary risk and coronary mortality, there was a startlingly lower incidence of CHD in women compared to men, which was difficult to explain.22 Women tended to smoke less than men, but non‐smoking women had only half the coronary mortality of non‐smoking men. Women were also more likely to have higher serum cholesterol than men, yet their CHD mortality was lower. The authors therefore concluded that in women, any given level of risk marker was associated with lower CHD rates than men and it was suggested that this reduction in risk may be due to the protection conferred by female sex hormones; this protection may also continue for many years following the menopause. Hypertension is one of the major risk factors for the development of CHD. The incidence of hypertension increases with age and is higher in men than in women up to the age of about 50. Beyond middle age, however, blood pressure in women exceeds that in men. It has been suggested that menopause may potentiate the age‐related increase in systolic pressure, perhaps as a result of reduced arterial compliance. Staessen et al.23 measured the blood pressures of 315 women and followed them up for a median of 5.2 years. Those women who were post‐menopausal had a 4–5 mmHg higher systolic blood pressure than their pre‐ and perimenopausal counterparts, and also, while there was no change in systolic blood pressure in premenopausal women during follow‐up, the systolic blood pressure increased by 4 mmHg in 5 years in the postmenopausal and perimenopausal women. In particular, isolated systolic hypertension is estimated to affect 30% of women over 65 years of age and was linked to an elevated risk of death from stroke and coronary heart disease. However, recent data, such as the Systolic Hypertension in the Elderly Program (SHEP)24 showed that a 36% reduction in the incidence of stroke and a 25% reduction in the incidence of CHD can be achieved by treating isolated systolic hypertension. Several cross‐sectional studies examining blood pressure in premenopausal and postmenopausal women have given conflicting results. For example, Taylor et al.25 found that the prevalence of hypertension was not higher amongst a group of postmenopausal women compared to that in control et found no relationship between blood pressure and menopause in a cohort of women. In one study of women in a small in systolic and blood pressure with age was but there was no significant difference in blood pressure when premenopausal women was there any increase in the use of when status In the Framingham no relationship was between a change in menopausal status and systolic or blood Nevertheless, data from the Health of found that systolic blood pressure was not to the menopausal status of women, although blood pressure was significantly higher amongst postmenopausal women. Furthermore, Staessen et the relationship between the menopause and blood pressure in a of pre‐ and post‐menopausal women, and found that hypertension was more observed in postmenopausal women, even after for age and of blood pressure occurring with the menopause have not been and thus are to account for the increase in the incidence of CHD which in older women. It has been suggested that oestrogen the premenopausal from hypertension and thus to the lower risk of CHD in women compared to men. However, this remains as the menopause is also associated with and of which influence blood pressure as In addition to oestrogen given in the has been found to increase blood and in some may although this is by of the on cholesterol and cardiovascular risk has been performed on men and the data extrapolated to women. This has been since it is that oestrogen studies have found a sex difference in These differences may in the difference in the incidence of CHD between the Indeed, the found a association between low and cardiovascular mortality in women, which was only to age as a of cardiovascular death in women. women, however, have lower cholesterol and higher compared to In contrast, women have higher cholesterol and younger men have lower In a cross‐sectional study, et found that postmenopausal women had significantly higher serum cholesterol compared to pre‐menopausal women. This was in with cross‐sectional and In the Framingham women between the of and years who were in a study for years demonstrated a significant in serum cholesterol between premenopausal and postmenopausal with the a time of the of the menopause, thus a Nevertheless, not studies have this For example, in a study of women, no association was found between the menopause and serum cholesterol It has been suggested that this is due to the of which is lower in cholesterol than that of and factors may an and role in the increase in cholesterol seen in older women. is a which has been to and data, from cross‐sectional studies, have shown that this can be associated with an increased risk of CHD and In a study by et of were significantly increased compared to following and bilateral In the Framingham however, no significant association was found between the menopause and In addition to the information available from studies of post‐menopausal women some studies of of oestrogen have shown a in and an increase in in has been as a of the development of is a major of cardiovascular but whether it significantly by the menopause remains Indeed, in with the menopause have been demonstrated in some but not cross‐sectional The study by et reported a significantly higher level of factor in postmenopausal women than premenopausal women, but no significant differences in of factors or or et found higher of factors and after the menopause. The of these are as yet but it has been suggested that higher of factors may to increased risk of CHD. increases the risk of CHD in women, and them at the same risk of CHD as men of the same age. Much of this excess risk is due to the excess in other coronary risk factors which occur in There not to be an association between menopause and after the of age, and smoking are taken into However, the presence of to any that a may In the Nurses’ Health was associated with death due to CHD, with the risk of CHD over higher among women with a of or Much of this increased risk can be to on blood pressure, and This increase in risk of CHD with increased has not been a in other studies, although a from the Framingham study that the prevalence of risk factors for CHD at of over 20 No relationship has however been found between the menopause and smoking is the cause of death in both men and women, and it has been estimated that the risk of coronary heart disease is between two and greater in women who with a increase in Women tend to smoke less than and this may to the lower incidence of coronary heart disease seen in women. smoking has been linked to early menopause and may to the increase in cardiovascular risk seen in postmenopausal women. Indeed, the risk of myocardial infarction significantly in women who therapy has effects on the in an increase in cholesterol by 10 to and a cholesterol by The increase in is in due to of the which The addition of to oestrogen in HRT the increase in seen with oestrogen The reported a trial in post‐menopausal women, found that therapy increased and although the effect on was greater with oestrogen In the a randomized trial of oestrogen or oestrogen in women with cholesterol in both groups, cholesterol increased in women oestrogen but not in those However, women had higher than women on in the this being more in women receiving oestrogen is an risk factor for and there is some evidence that increases with age in In a small study of 30 et found that there was a reduction in in the 10 women receiving HRT. In addition, et found that HRT the in seen after surgically‐induced menopause. In postmenopausal hormone replacement the risk of cardiovascular disease the effects oestrogen has on the have a with blood the it has been that there is a risk of developing hypertension with the Indeed, oestrogen increases and increases on the other with at the and treatment also to cause There has been about the use of HRT in postmenopausal women with and studies have been to the effect of HRT on blood pressure in women. For example, et studied postmenopausal women with HRT for menopausal After a of in this study, no significant change in blood pressure was found in the women receiving HRT as or oestrogen only et conducted a study in postmenopausal women with to hypertension and found no increase in blood pressure in this group after treatment with oestrogen or although a in and in women was Other studies on the effect of HRT on blood pressure in women have shown a significant in blood pressure and no change in systolic blood et in a prospective study of and women showed a in systolic blood pressure in the group after 6 treatment with HRT. However, in the no effect on blood pressure was seen with HRT. In addition to the in seen with oestrogen oestrogen is have other effects on the The may in patients on HRT. studies, oestrogen to arterial and to serum and and have The recent and small of lower in women with HRT. surgical menopause, HRT use may also result in in of function and the The effects of HRT may therefore to on the blood For example, may increase cardiac and coronary blood Furthermore, may et found in a study that there was a significant reduction in the of in women with HRT. This was also found in a prospective study by et with of the women receiving HRT developing diabetes, compared to in the control The small numbers however, not one to any about the effects of HRT on the development of Nevertheless, in the and were slightly reduced in women that is by HRT. hormone replacement therapy (HRT) has been reported as the risk of CHD in postmenopausal women. However, it is perhaps an of the with which CHD in women is that only have prospective clinical trials which the and of HRT have been very recently, the studies of the effects of HRT on cardiovascular disease had been studies. These on the concluded that HRT had a effect and the use of HRT in post‐menopausal women to to the development of heart disease. of these studies suggested that protection was for secondary prevention in women with CHD. Nevertheless, these trials can be criticized for many of the studies were observational studies, not into account for HRT. For example, may be in HRT to the and it from higher risk HRT may be an a in a The in some of the studies were in that included the development of chest pain and leg pain of only the end points such as cardiovascular myocardial infarction or et a of the effects of HRT on a of outcomes coronary heart disease. They studies between and which examined this and that women who had used HRT had a relative risk for CHD of compared to The relative risk of CHD death amongst of oestrogen was 30 studies have the relationship between CHD and HRT. The majority of these studies have demonstrated a 30 to reduction in cardiovascular and mortality in of In the Nurses’ Health of the 121 700 women studied over deaths were in postmenopausal women, with due to CHD, due to and due to When the incidence of death was to HRT those who were of HRT had a relative risk of death from CHD of compared to due to stroke was reduced risk as was mortality due to risk Women who had one or more risk factors for CHD and used HRT had a in mortality compared to women who were at low risk for CHD had less risk any from HRT was by an increased risk of In another prospective population study, the women were for an of deaths due to cardiovascular disease were in the of oestrogen and in the oestrogen This was to a relative risk of for oestrogen after a analysis, adjusting for age, smoking and blood In the Framingham an analysis of postmenopausal women found no difference in mortality between oestrogen and this study found an increase in risk of cardiovascular death among oestrogen This increase was in spite of a more cardiovascular risk amongst oestrogen It be that much of the data available from the population studies was for oestrogen are used in with oestrogen in to the risk of due to in women with an intact Nevertheless, there is less data available for the cardiovascular effects of In an analysis of the Nurses’ Health Study population to whether the addition of the effects seen with et found that the addition of did not to the effects of The trial was a trial although it did not clinical studied the effects of different HRT on cardiovascular risk In this study, cholesterol was increased by in the group with oestrogen and to a in those women The age‐related increase in was by There was no significant change in blood pressure, although there was a increase in in women with were also They found a large and early increase in the of and a in in women with but no change in factor or factor They concluded that this increase in result in adverse cardiovascular trial of oestrogen oestrogen in women with found cholesterol and cholesterol in groups, with the effect on being more in the in and large well‐conducted randomized trials have been the Health which is a primary prevention study and has not yet been and the and Study a secondary prevention The trial was a trial of oestrogen in postmenopausal women 80 years of age, who were to have CHD. They were followed up every 4 for an of years. The primary outcomes in this study were myocardial infarction or CHD outcomes included coronary congestive heart cardiac stroke or attack and arterial disease. The study some results. There were no significant differences in end points between the treatment and women in the treatment group and women in the group had a primary outcome in spite of lower cholesterol and higher cholesterol in the treatment There was however a significant time with to CHD with women in the treatment group experiencing more events in the and in the and years. The authors concluded that not HRT for secondary prevention in CHD, but a women were it be for to The of whether premenopausal women are from disease by of their status remains The same risk factors for cardiovascular disease to in women as in men, although the of women in the large studies that far less information about the of these risk factors is available to Indeed, hypertension and are significant risk factors for the development of CHD in women and be The of women from clinical trials is not as it the misconception that cardiovascular disease is a and half the population may receive treatment. of the of to CHD in women is the fact it has taken decades for randomized trials of HRT to be many population studies have suggested a effect with HRT in reduced risk of cardiovascular it is to that these studies can be in that there may well be significant which effect the The studies are the which to In of the recent the use of HRT in postmenopausal women must more information is available from these studies. to of studies of hormone replacement therapy studies of hormone replacement therapy trials of hormone replacement therapy trials of hormone replacement therapy