Diastolic dysfunction and sex-specific progression to HFpEF: current gaps in knowledge and future directions

The term left ventricular diastolic dysfunction (LVDD) refers to functional and mechanical problems during diastole, ultimately leading to inadequate filling of the left ventricle. LVDD is caused by a broad range of abnormalities such as altered myocardial relaxation, myocardial stiffness, and left atrial dysfunction [3]. LVDD is an imaging-based finding and does not necessarily cause symptoms. However, LVDD resulting in elevated left ventricular filling pressure, left atrial pressure, and increased pulmonary wedge pressure can cause exercise-induced shortness of breath and reduced exercise tolerance [4]. By the time these symptoms occur, HF is a common diagnosis in both women and men. Prevalence of LVDD ranges between 3.1 and 35% in the general community, these differences being highly dependent on age, and risk factors of the study population, and notably on the different definitions used [5‐9]. Multiple studies have shown that there are no important sex-differences in the prevalence of LVDD in community-based cohort studies [5‐7] (see Fig. 1). Nevertheless, these studies often fail to report the prevalence of LVDD by sex or by gender. LVDD by echocardiography is evaluated with similar cut-off values for women and men (see Table 1) [3], although for instance some differences in for example E/e′ ratio between women and men were found in healthy populations [10, 11]. Also, guidelines have changed their definition of LVDD over the years, but cut-offs do not differ between women and men. When applying the most recent 2016 guidelines [3] to a French population cohort, the prevalence of LVDD diastolic dysfunction was 0.2% in young individuals of 20 to 40 years of age compared to 1.1% and 3.1% in the age groups 40 to 60 and over 60 years of age [9]. Again, these prevalence numbers were not reported by sex. In addition, the prevalence was much lower compared to earlier guidelines. For example, the prevalence of LVDD was 12.9% in people over 60 years of age when applying the 2009 guidelines [12‐15].

To determine diastolic function, imaging is used, and the routine echocardiography report includes information on diastolic function of the heart classifying it as normal, indeterminate, or abnormal using four key parameters listed in Table 1 [3]. For each of these parameters, no sex-specific cut-offs exist and differences between sexes are reported to be small [10, 11, 26, 27]. Diastolic function parameters and all degrees of LVDD were associated with mortality in a large database of 436,360 women and men. Importantly, none of the reported diastolic function measures had a sex-specific association with all-cause mortality. Yet, all-cause mortality is the hardest of all clinical endpoints, and does not reflect sex differences in morbidity such as HFpEF [28]. Symptoms were not taken into account in this study, so it may be that diastolic function parameters have different prognostic consequences to clinically relevant endpoints in women and men.

Alternative echocardiographic parameters can be used to classify LVDD [29, 30]. Some of which differ by sex, e.g., left ventricular global longitudinal strain shows higher normal values in women compared to men [17, 31], and left ventricular mass index (LVMI) has a higher cut-off value for left ventricular hypertrophy in men compared to women (115 g/m² vs 95 g/m²) [16]. This reflects inherent sex-differences in cardiac structure and function (see Fig. 1). Men have higher left ventricular mass as compared to women. The difference in LV mass is attributed to the smaller hearts of women, even when indexed to body size, resulting in smaller left ventricular volumes and lower LV mass [32, 33]. To compensate for smaller cavity size, women have a slightly higher left ventricular ejection fraction [16] and higher global longitudinal strain [17, 31]. Still, smaller cavity of the left ventricle is associated with lower cardiac output after indexation to body surface area in healthy women at peak exercise, when compared to healthy men [34]. Furthermore, there is a greater and steeper increase in LV mass with aging women as compared to men [35]. Additionally, there is less cardiomyocyte loss in women during a lifespan [36], and it has been proposed that women are less susceptible to decreases in contractility when afterload increases, as compared to men [37]. Potentially, these dimorphisms in size and function of the heart form the female-specific substrate for a greater susceptibility to further concentric LV remodeling and evolving HFpEF.

HFpEF refers to a syndrome in which elevated left ventricular filling pressures and pulmonary pressures resulting from LVDD, cause symptoms and/or signs suggestive of HF, while left ventricular ejection fraction is preserved (≥ 50%) [38]. This might cause an increase in natriuretic peptide levels. The most reported symptom in both women and men with HFpEF is exercise-induced shortness of breath [39]. Heart failure with reduced ejection (HFrEF) fraction is considered the counterpart of HFpEF, since left ventricular ejection fraction is decreased. But, also in HFrEF, LVDD contributes to signs and symptoms through increased left ventricular filling pressures. However, in this review, we will mainly focus on HFpEF.

The diagnosis of HFpEF is complex, also because of the multiple cardiac and non-cardiac comorbidities associated with the disease, such as atrial fibrillation, diabetes, chronic obstructive pulmonary disease (COPD), and renal dysfunction. These comorbidities may be a cause, contributing factor for developing HFpEF, or an alternative diagnosis (“mimic”) for patients presenting with shortness of breath or reduced exercise tolerance. Most of HFpEF comorbidities are hypothesized to contribute to a systemic pro-inflammatory state [40] that can lead to endothelial dysfunction and impaired coronary flow reserve and coronary microvascular dysfunction. The latter were proven to have greater impact on the incidence of major outcomes in women compared to men referred for coronary angiography [41].

We know that aging women from the general population report more exercise-induced complaints, e.g., more severe breathlessness, compared to men [42]. HF is often underdiagnosed in primary care possibly due to limited diagnostic tools such as electrocardiography and measurement of natriuretic peptide plasma levels. On the other hand, spirometry is more readily available upon presentation with shortness of breath. Pulmonary fluid overload may cause pulmonary obstruction and makes it easy to misclassify HF as chronic obstructive pulmonary disease. Indeed, 20% of primary care patients labeled with COPD were diagnosed with HF after undergoing an extensive pulmonary and cardiac assessment, and about half of these HF patients are classified as HFpEF [43]. In men and women aged 65 years or older, who visited their general practitioner for reasons of exertional shortness of breath, resting echocardiography diagnosed 16.5% of men and 15% of women with HF [44]. Interestingly, 76% of these newly detected HF cases were HFpEF cases. Undetected HF was even more prevalent in individuals with diabetes with a prevalence of 27.7%, and again most had HFpEF (83%), with female sex being a predictor of HF [21]. Screening studies like this are scarce and show that HFpEF is frequently underdiagnosed in the elderly. Without a doubt, underdiagnosis or a hampering diagnosis results in lower quality of life and increased health care consumption [45]. Underdiagnosis seems to affect women more often than men, also for myocardial infarction: 30% of electrocardiogram detected myocardial infarction remained unrecognized in women, compared to 16% in men [46]. The more chronic nature of cardiovascular disease in women may go hand in hand with a higher burden of symptoms, or symptoms that are perceived as being atypical or non-cardiac, as shown by a recent meta-analysis of studies in women and men with coronary syndromes [47]. Whether disease presentation is exactly the same in women and men with HFpEF is still unclear.

The prevalence of established HF worldwide is around 1 to 2.5%, depending on the diagnostic criteria used, and this percentage is equal for women and men [66]. In Western populations, the lifelong risk of HF at the age of 40 years is 21% for women and 20% for men [67] and, at the age of 55 years, 29% and 31% for women and men, respectively [68].

In the period 2000–2010, the incidence of HF in the USA decreased by ~ 5% per year [69], most likely due to better treatment of myocardial infarction [66]. The incidence of HFpEF also decreased, with similar overall rate changes for women and men over 10 years (− 27%; − 2.7% per year), probably due to better treatment of comorbidities [69]. However, mortality and hospitalization rates in HF patients did not decrease over time and remain high [69, 70]. Mortality rates in HF patients are 20% in the first year and reach 50% over 5 years [69]. While total HF prevalence is similar in men and women, women outnumber men with respect to men HFpEF. In community-based studies, women with HF had HFpEF in 67% of the cases, compared to 42% of men with HF having HFpEF [71]. Women account for 55–66% of all HFpEF hospitalizations and only 29–42% of all HFrEF hospitalizations [55, 72, 73]. The proportion of HFpEF cases with respect to overall HF hospitalization is also increasing. In 2010, 39% of hospitalized HF patients had HFpEF, whereas this was 33% in 2005. Unfortunately, this was not reported for women and men separately [74].

The high proportion of women with HFpEF could be accounted for by higher life expectancy in women. Also, a higher prevalence of comorbidities such as chronic kidney disease, hypertension, and valve and lung disease in women with HFpEF explains the female predominance (see Fig. 1) [55]. A study combining data from 4 large population-based cohort studies concluded that women and men have an equal risk to develop HFpEF, after correction for comorbidities and age [75], but that the risk to develop HFrEF is lower in women, as compared to men. Hence, female sex was not an independent risk factor for HF (HR = 0.86 (95% CI: 0.71, 1.04) [76] or HFpEF [25], while male sex was an independent risk factor for HFrEF (HR = 1.84 (95% CI: 1.55, 2.19). However, in a community study among people aged 60 years or over with type 2 diabetes, female sex was an important predictor of previously undetected HF (> 80% HFpEF), but more importantly, in this group of people with type 2 diabetes, the age-stratified prevalence of HFpEF among women was significantly higher than in men [21]. Altogether, despite the finding that sex or gender may not be an independent risk factor for HF development, there is a higher prevalence of HFpEF in women. Therefore, it is important to better understand the role of risk factors contributing to the progression from LVDD to HFpEF that may be associated to female sex.

While the mechanistic role of LVDD in the development of HFpEF is evident, longitudinal data on how LVDD deteriorates towards HFpEF is relatively sparse [77]. As HFpEF is difficult to treat and carries a poor prognosis, preventing HFpEF and limiting disease progression are critical. Therefore, predicting progression from LVDD to HFpEF is key to guide aggressive risk factor management and earlier intervention. Eleven longitudinal studies described the progression of LVDD towards HF (Table 2). The percentage of women participating ranged from 19 to 61%. The proportion of participants with mild to severe diastolic dysfunction that developed HF ranged from 0.8 to 37% during a follow-up time of 1.2 to 11 years. Out of these 11 studies, only one distinguished between HFpEF and HFrEF when investigating the progression of LVDD towards HF [8]. In this study, with a median follow-up of 11 years, LVDD was present in 36% of the participants at baseline. These participants had a high risk of developing HFpEF (HR = 1.88, 95% CI: 1.13, 3.13) even after correction for age, sex, body mass index, systolic blood pressure, hypertension treatment, cholesterol levels, diabetes mellitus, prior myocardial infarction, and valvular heart disease [8]. The main risk factor for progression in this study was airflow limitation which could be a manifestation of sub-clinical pulmonary disease, leading to low-grade inflammation [8]. Further risk factors for the progression of LVDD or pre-clinical HF to overt HF were older age [78‐80], hypertension [78, 81], peripheral vascular disease [81], diabetes [78], coronary artery disease [78], (subclinical) renal impairment [8, 79], anemia [8], and the Charlson comorbidity score [80]. These risk factors are exemplary for the multi-organ involvement of the HFpEF syndrome. Given the higher prevalence of HFpEF in women, it may be that this comorbidity-driven progression of LVDD towards HFpEF is sex-specific (see Fig. 1). On the other hand, the observation that female sex was not unequivocally an independent risk factor for HF(pEF) questions this idea. This is indeed also confirmed by three studies that reported that sex was not significantly influencing the progression from LVDD towards HF [80, 82, 83], suggesting that the risk of progression from LVDD to HFpEF is similar in women and men. Nevertheless, most studies do not test for effect modification by sex, do not perform sex-stratified analyses, or study female-specific associations, as was previously also shown in a systematic review on LVDD/HFpEF [84]. This is important because stronger associations of comorbid conditions for one of the sexes may lead to an absent relation of sex itself in multivariable analyses correcting for comorbidities. We therefore highlight several areas in HFpEF research in which the incorporation of sex and gender analyses are likely to enable advancements in the field.

There is a significant knowledge gap on the exact mechanisms that are implicated in the progression from LVDD to HFpEF. We hereby review the risk factors associated with HFpEF, the knowledge on the mechanisms, and whether influences of sex are reported (see Fig. 1 and Table 3).

Apart from differences in the magnitude of the associations between risk factors and HFpEF in women and men, female-specific factors are often not studied but important to consider. We describe several female-specific and female-prevalent factors or disorders that might influence progression to HFpEF (see Fig. 1 and Table 3).

Women and men with HF have equal mortality rates [72, 130], but the probability of re-hospitalization for HF is higher in women (34% re-admissions in women compared to 27% in men) [72]. Data on mortality and hospitalization, however, are not consistent. Three studies reported significantly better outcomes in women with HFpEF compared to men with HFpEF [55, 131, 132]. Also, women with HF were more frequently admitted for non-cardiovascular causes [130], and women hospitalized with HFpEF were at higher risk of poor post-discharge outcomes (adjusted HR = 1.54 (95% CI: 1.14, 2.07) than men [133], which may be due to high comorbidity burden in women. This high comorbidity burden together with a higher prevalence of obesity and worse diastolic and vascular function and greater exercise limitations might reflect different HFpEF etiologies and can partly explain the inconsistencies in prognostic studies [39, 131, 134, 135]. Additionally, women with HFpEF have a worse quality of life compared to men with HFpEF, and this is also consistently observed in the general community [134]. A lower quality of life in women is potentially attributed to a higher symptom burden, less social support, or more depression [134]. Additionally, women may perceive impairment as more severe compared to men [134]. Two community studies showed that a lower quality of life or lower self-rated health, respectively, are associated with asymptomatic LVDD [5, 136], and counter-intuitively, the age-adjusted association of self-rated health with LVDD was only significant in men (OR = 3.49 (95% CI: 1.0, 11.9)) [136].

After years of disappointing clinical trials, the first evidence-based HFpEF treatment has been found. Two trials on sodium-glucose co-transporter 2 (SGLT2) inhibition, studying empagliflozin and dapagliflozin, respectively, in HFpEF patients, were able to meet their primary endpoint of reducing cardiovascular mortality and HF hospitalization, in both sexes [65, 137]. At the moment, SGLT2 inhibition is recommended in the American HF guidelines (level of evidence 2A), and it is expected that European guidelines will follow soon [138]. Now that these pharmacological treatments for HFpEF become available, aggressive management of pre-clinical LVDD with the same drugs should be investigated, to prevent deterioration to HFpEF. Further current guideline recommendations include treatment with diuretics in congested HFpEF patients (level 1A of evidence) [38, 138], and the American guidelines also have a 2B level of evidence recommendation for treating selected HFpEF patients with sacubitril-valsartan, angiotensin receptor blockers, or mineralocorticoid receptor antagonists. Interestingly, although sacubitril-valsartan did not convincingly reduce the composite outcome of HF hospitalization and cardiovascular death in patients with HFpEF from the PARAGON-HF trial, sex appeared to modify the effect of treatment on the outcome. A benefit was indeed seen in women, in which the rate ratio for the primary outcome for sacubitril-valsartan versus valsartan was 0.73 (95% CI, 0.59–0.90), while in men, no benefit was reported (rate ratio = 1.03 (95% CI, 0.84–1.25)) [139]. Since the average ejection fraction is higher in women, it was hypothesized that a proportion of women in the trial had mild systolic dysfunction. This could represent a plausible explanation for the observed benefit of sacubitril-valsartan in women, considering that the drug has been clearly demonstrated to be effective in the presence of LV systolic dysfunction [139]. Another example of sex-specific treatment response to HF drugs comes from an exploratory post hoc analysis of the TOPCAT trial, showing a reduced risk of in all-cause mortality in women treated with spironolactone (HR = 0.66 (95% CI: 0.48, 0.90), while no effect was observed in men (HR = 1.06 (95% CI: 0.81, 1.39) [140]. A more pronounced protective effect on cardiac remodeling has been hypothesized as one of the contributing factors of the response to spironolactone in women. Sex-differences in pharmacokinetics and pharmacodynamics underpin these differences in treatment responses and have also been demonstrated for other HF drugs such as ACE-inhibitors, ARBs, and beta-blockers. Observational and routine health care data studies showed that women with HF are better off with lower doses of these drugs, bringing into question whether or not optimal medical treatment should rather be defined sex-specifically [141, 142]. Additionally, it should be noted that women were underrepresented in HFpEF trials testing drug therapies, and although post hoc analyses did not show effect modification by sex, those sub-analyses were underpowered and thus unlikely to detect sex differences.

Since there is a broad understanding that HFpEF is a multifactorial, multi-organ, multi-comorbidity syndrome, numerous pre-clinical models have been developed to understand disease mechanisms and to identify therapeutic targets. Over time, there has been a transition from simple single-hit models to multi-hit models involving age, a Western high fat/high sugar diet, diabetes, hypertension, hypercholesterolemia, and kidney dysfunction as stressors and/or comorbidities [148]. These models enable sex-specific and phenotype-specific research [148, 149]. However, a major drawback is the HFpEF definition. Many studies define disease outcomes based on structural and functional parameters, and the models actually represent extended LVDD models [149]. To overcome this, signs of congestion, such as lung weight, natriuretic peptide levels, and, ultimately, symptoms, should be taken into account. In our opinion, pre-clinical models are not fully suitable to study the natural progression of LVDD towards HFpEF, but especially aging and hypertension/kidney disease models provide opportunities to investigate the pre-clinical stage of HFpEF in a sex-specific way.