Gender and Survival in Patients With Heart Failure
Gender and Survival in Patients With Heart Failure
This large-scale meta-analysis, based upon patient-level data from almost 42 000 individuals, represents the largest assessment of the association between sex and survival for patients with HF. The main finding of our study is that compared with men, women with HF have lower all-cause mortality over 3 years, irrespective of EF. Our analysis confirms that women with HF are on average older, are more likely to have a history of hypertension and diabetes, but are less likely to have HF of ischaemic aetiology. We also found that women had more severe functional limitation than men as reflected by NYHA class. Importantly, women were less likely than men to be prescribed evidence-based therapies, particularly among those patients with HF with reduced EF, for which there is unequivocal evidence of benefit from these agents.
Previous studies assessing potential differences in survival among men and women with HF have presented conflicting results, with some suggesting better survival for women and others failing to identify such an association. Many of these studies have been limited by relatively small numbers of patients and have presented mortality risks for men and women with wide and overlapping confidence intervals, preventing definitive conclusions from being drawn. Moreover, consideration of specific interactions of sex with aetiology of HF or with reduced/preserved EF has been limited. Our study, using a large individual patient data set, is appropriately powered to ascertain the prognostic significance of sex in patients with HF.
The current analysis suggests that while crude unadjusted mortality rates in men and women were very similar, when adjusted for age the risk of death was higher in men than in women with HF. Further, the influence of age on survival was similar in men and women (P for gender ×age interaction = 0.63), suggesting that better survival in women is associated with factors other than age. While women have a higher prevalence of HF with preserved EF which was associated with a better prognosis in this study, we observed a higher risk of death in men, irrespective of whether they had HF with reduced or preserved EF (Figure 2).
There are a number of alternative potential explanations for the better outcomes in women with HF. The female heart appears to respond to injury differently from the male heart. For example, women have been reported to have less ventricular remodelling, preservation of right ventricular function, and protection against ventricular arrhythmias, neurohormonal activation, genetic mutations, myocyte necrosis, and apoptosis. Some of these advantages could be related to pregnancy and to sex-specific differences in gene expression.
In accordance with previous studies, we found lower prescription of ACE inhibitors or ARBs in women than in men. Perhaps surprisingly, this was particularly evident in patients with reduced EF, where the evidence for these therapies is strongest. In fact, for all age groups with reduced EF, women received ACE inhibitors or ARBs less frequently than men (data not shown). Given this pattern of prescribing, the lower risk of death for women compared with men is all the more remarkable.
The reported prevalence of diabetes in patients with HF is highly variable, with figures between 13% and 29%, probably due to the definitions of both diabetes and HF used and the heterogeneous nature of the populations studied. Our data are consistent with previous reports of higher risk of death among patients with HF with co-existing diabetes. In the current analysis we have extended these previous observations to a large population that included patients with reduced or preserved EF. In both groups we observed diabetes to be a strong independent risk factor for mortality in patients with HF. This association was particularly evident among women, especially those with reduced EF, where the presence of diabetes attenuates the apparently protective 'effect' of female sex on prognosis. A similar interaction has been described for patients with ischaemic heart disease, where diabetes also attenuates the gender gap in mortality. In contrast, female gender has been suggested to be associated with greater likelihood of pre-clinical diabetic cardiomyopathy.
Our finding that the association between female sex and better survival appears to be stronger in patients with non-ischaemic HF is in agreement with several previous studies. Importantly, we have shown clearly that this sex-related difference in prognosis is seen regardless of whether EF is reduced or preserved. A single previous report, from the second Cardiac Insufficiency Bisoprolol Study (CIBIS II), suggested no sex-related difference in mortality in patients with non-ischaemic aetiology. However, in CIBIS II, the aetiology of HF was undefined in 36% of men and 47% of women, limiting markedly the ability of this trial to compare prognosis between ischaemic and non-ischaemic HF.
Our analysis is constrained by the underlying limitations of the original individual studies. However, by incorporating large amounts of data from both randomized trials and observational studies, resulting in a wide range of patients, with long follow-up and a large number of clinical events, the results are likely to be an accurate reflection of patients with the syndrome of HF seen in routine clinical practice. The interaction between diabetes status and sex-related outcomes is robust, whereas the interaction with aetiology is less certain.
Discussion
This large-scale meta-analysis, based upon patient-level data from almost 42 000 individuals, represents the largest assessment of the association between sex and survival for patients with HF. The main finding of our study is that compared with men, women with HF have lower all-cause mortality over 3 years, irrespective of EF. Our analysis confirms that women with HF are on average older, are more likely to have a history of hypertension and diabetes, but are less likely to have HF of ischaemic aetiology. We also found that women had more severe functional limitation than men as reflected by NYHA class. Importantly, women were less likely than men to be prescribed evidence-based therapies, particularly among those patients with HF with reduced EF, for which there is unequivocal evidence of benefit from these agents.
Previous studies assessing potential differences in survival among men and women with HF have presented conflicting results, with some suggesting better survival for women and others failing to identify such an association. Many of these studies have been limited by relatively small numbers of patients and have presented mortality risks for men and women with wide and overlapping confidence intervals, preventing definitive conclusions from being drawn. Moreover, consideration of specific interactions of sex with aetiology of HF or with reduced/preserved EF has been limited. Our study, using a large individual patient data set, is appropriately powered to ascertain the prognostic significance of sex in patients with HF.
The current analysis suggests that while crude unadjusted mortality rates in men and women were very similar, when adjusted for age the risk of death was higher in men than in women with HF. Further, the influence of age on survival was similar in men and women (P for gender ×age interaction = 0.63), suggesting that better survival in women is associated with factors other than age. While women have a higher prevalence of HF with preserved EF which was associated with a better prognosis in this study, we observed a higher risk of death in men, irrespective of whether they had HF with reduced or preserved EF (Figure 2).
There are a number of alternative potential explanations for the better outcomes in women with HF. The female heart appears to respond to injury differently from the male heart. For example, women have been reported to have less ventricular remodelling, preservation of right ventricular function, and protection against ventricular arrhythmias, neurohormonal activation, genetic mutations, myocyte necrosis, and apoptosis. Some of these advantages could be related to pregnancy and to sex-specific differences in gene expression.
In accordance with previous studies, we found lower prescription of ACE inhibitors or ARBs in women than in men. Perhaps surprisingly, this was particularly evident in patients with reduced EF, where the evidence for these therapies is strongest. In fact, for all age groups with reduced EF, women received ACE inhibitors or ARBs less frequently than men (data not shown). Given this pattern of prescribing, the lower risk of death for women compared with men is all the more remarkable.
The reported prevalence of diabetes in patients with HF is highly variable, with figures between 13% and 29%, probably due to the definitions of both diabetes and HF used and the heterogeneous nature of the populations studied. Our data are consistent with previous reports of higher risk of death among patients with HF with co-existing diabetes. In the current analysis we have extended these previous observations to a large population that included patients with reduced or preserved EF. In both groups we observed diabetes to be a strong independent risk factor for mortality in patients with HF. This association was particularly evident among women, especially those with reduced EF, where the presence of diabetes attenuates the apparently protective 'effect' of female sex on prognosis. A similar interaction has been described for patients with ischaemic heart disease, where diabetes also attenuates the gender gap in mortality. In contrast, female gender has been suggested to be associated with greater likelihood of pre-clinical diabetic cardiomyopathy.
Our finding that the association between female sex and better survival appears to be stronger in patients with non-ischaemic HF is in agreement with several previous studies. Importantly, we have shown clearly that this sex-related difference in prognosis is seen regardless of whether EF is reduced or preserved. A single previous report, from the second Cardiac Insufficiency Bisoprolol Study (CIBIS II), suggested no sex-related difference in mortality in patients with non-ischaemic aetiology. However, in CIBIS II, the aetiology of HF was undefined in 36% of men and 47% of women, limiting markedly the ability of this trial to compare prognosis between ischaemic and non-ischaemic HF.
Our analysis is constrained by the underlying limitations of the original individual studies. However, by incorporating large amounts of data from both randomized trials and observational studies, resulting in a wide range of patients, with long follow-up and a large number of clinical events, the results are likely to be an accurate reflection of patients with the syndrome of HF seen in routine clinical practice. The interaction between diabetes status and sex-related outcomes is robust, whereas the interaction with aetiology is less certain.
