Outcomes in women vs men with heart failure and preserved ejection fraction

Cardiology Review® Online, March 2007, Volume 24, Issue 3

The epidemiologic finding that women with heart failure have better overall survival than men may be because of the higher prevalence of diastolic heart failure or heart failure with preserved ejection fraction (HF-PEF) among women.

The epidemiologic finding that women with heart failure have better overall survival than men may be because of the higher prevalence of diastolic heart failure or heart failure with preserved ejection fraction (HF-PEF) among women.1-3 In patients with systolic heart failure, higher survival rates have also been shown for women.4,5 Few data exist regarding the role of sex in disease severity, symptoms, or clinical presentation of HF-PEF or about the contribution of these differences to sex differences in morbidity and mortality in patients with HF-PEF.6 To shed light on this issue, we performed a retrospective analysis of sex differences in clinical presentation, morbidity, and mortality among subjects with HF-PEF enrolled in clinical trial.

Subjects and methods


Among 719 subjects with HF-PEF and a normal left ventricular ejection fraction (≥ 50%) who were enrolled in the ancillary arm of the Digitalis Investigation Group (DIG) trial, we assessed the differences between men and women in baseline clinical characteristics, all-cause mortality, cardiovascular hospitalization, and outcomes of hospitalization for worsening heart failure.7 To compare the risk of the outcome between women and men, we performed multivariate and univariate Cox proportional hazards modeling. Covariates in multivariate models included variables that correlated with hospitalization or death in bivariate analysis ( < .10) or that were found to be important variables from prior studies. Models for hospitalization were censored for death. We also examined the hazard for mortality by sex in subgroups that were defined by New York Heart Association (NYHA) functional class, diabetes, hypertension, history of myocardial infarction (MI), and ischemic cause of heart failure. To determine whether the effects of sex varied in the subgroups, we performed tests of interaction.


shows the baseline characteristics of the subjects by sex. A total of 341 subjects (47.4%) were women and 378 (52.6%) were men. Women with HF-PEF more often had a history of hypertension or diabetes, less often had a prior MI, and were older compared with men. Although a nonischemic cause of heart failure was more prevalent in women than in men, ischemia was identified as the primary cause of heart failure in 46% of the women. At baseline, women also had greater impairment of renal function (as shown by lower estimated glomerular filtration rate) and a higher heart rate and systolic blood pressure than men.

Table 1

The duration of heart failure was similar in both groups. However, women were more symptomatic and appeared to have greater clinical severity of heart failure. This was shown by a greater proportion of women with a history of orthopnea and dyspnea at rest, fewer women in NYHA functional class I, and more women in NYHA functional class III-IV. Compared with men, women also more frequently had a higher cardiothoracic index and vascular congestion on chest x-ray, as well as peripheral edema and rales on physical examination. Furthermore, more women were taking diuretics. Similar numbers of men and women were taking angiotensin-converting enzyme inhibitors, nitrates, and vasodilators at baseline and were randomly assigned to either digoxin (Lanoxin) or placebo.






A total of 176 deaths occurred over a follow-up period of 39 months. There were no significant differences between women and men in crude rates of heart failure mortality (8.2% vs 6.9%, respectively; = .50), cardiovascular mortality (17.6% vs 18.0%, respectively; = .89), and all-cause mortality (25% vs 24%, respectively; = .93). Hospitalizations for heart failure (27% for women vs 16% for men; < .001) and for any cardiovascular cause (53% for women vs 46% for men; = .05), however, occurred more often in women.

shows the hazard ratios by sex for the various outcomes. Female sex was shown to be an independent predictor of lower mortality but not of cardiovascular hospitalization or heart failure on multivariable analysis that was adjusted for differences in baseline characteristics, including severity of heart failure. As shown in the


, the trend for the favorable effect of female sex on death rates occurred in all subgroups that were examined. Subjects with ischemic causes appeared to have a greater benefit than those with nonischemic causes, but the interaction for this difference was not significant ( = .18). Furthermore, no statistical significance was shown for any of the 2-way interactions tested for the other subgroup variables.

Table 2Figure


Our results show that women with HF-PEF have worse heart failure than men and are also older. Almost half the women in our study were classified as having ischemic heart disease as the etiologic factor for heart failure. In addition, women had better risk-adjusted survival but similar risk of hospitalization compared with men.

Prior studies have shown differences between men and women in the clinical severity of heart failure in patients with systolic heart failure. Women with heart failure have been shown to have more signs and symptoms of heart failure along with a lower exercise tolerance than men. Our study was the first to show similar differences in patients with HF-PEF. The observed sex differences in clinical presentation of HF-PEF may be the result of several biologic factors. Compared with men, older women have a greater hypertrophic response to pressure and volume overload.8,9 In hypertensive women, extreme cardiac hypertrophy is typified by noticeable impairment of diastolic function, with episodes of pulmonary edema.10,11 Only women have been reported to have a distinct diabetic cardiomyopathy with increased wall thickness and left ventricular mass.12 Furthermore, an increased sensitivity to fluid shifts may result from the higher ventricular and vascular elastance that has been noted in women,13 as may the increased occurrence of renal dysfunction in women shown in our study. At a less advanced stage of heart failure, women may have more signs, symptoms, and morbidity than men, which may result from an interaction of these factors, including higher systolic blood pressure and greater diastolic dysfunction associated with stiffer hearts and vasculature.

This study provided interesting findings regarding the etiology of HF-PEF in women. Hypertension, which is usually thought to be the chief cause of HF-PEF in women,14 was shown to be the etiologic factor in less than a third of the women in the study cohort, and a history of hypertension was absent in another third of the women. However, myocardial ischemia was noted as the primary cause of heart failure in almost half of the women in the study cohort. As shown by these results, it is crucial to assess women with HF-PEF for coronary heart disease.

Despite the fact that this study did not provide insight into the reasons for the independent survival benefit in female patients with HF-PEF, several possible mechanisms may exist. Differences in myocardial adaptation to aging, injury, or stress between men and women may contribute to the sex differences in mortality.2,9,15 Compared with the “adaptive” remodeling with concentric hypertrophy and small cavity size in women with systolic hypertension, for example, men may show maladaptive remodeling with left ventricular dilatation and eccentric left ventricular hypertrophy.9 In addition, women with aortic stenosis have greater left ventricular hypertrophy with smaller cavities, lower end-systolic wall stress, and better preservation of left ventricular ejection performance compared with men with aortic stenosis.15 As a result, the higher mortality observed in men in our study may be a consequence of maladaptive remodeling, that is, the progression of HF-PEF to left ventricular dilatation with left ventricular systolic dysfunction may be more frequent in men. Furthermore, the increased burden of ischemic heart disease in men may result in a higher incidence of MI or of larger infarcts at follow-up, with a subsequent higher prevalence of systolic dysfunction and thus higher mortality in men compared with women.


In subjects with HF-PEF enrolled in the ancillary arm of the DIG trial, we found that women had greater clinical severity of heart failure at enrollment compared with men. During follow-up, women were more often hospitalized for heart failure but had similar mortality compared with men. However, after adjusting for the differences in clinical severity and other baseline characteristics between men and women, female sex was an independent predictor of better survival but not of heart failure hospitalization. Although sex differences in myocardial adaptation to pressure overload, aging, or injury may be invoked as explanations for our findings, further studies in different populations are needed to confirm these sex differences in patients with HF-PEF and to examine factors contributing to the observed differences, including the role of etiology of heart failure, severity of coronary heart disease, and echocardiographic measures of left ventricular hypertrophy, dimensions, and mass.