Transthoracic echocardiography plays an important role as an imaging tool for the evaluation of heart failure with preserved ejection fraction (HFpEF). Among the imaging findings, the E/e' ratio, a surrogate for left atrial (LA) filling pressure, is a robust indicator that supports the diagnosis of HFpEF. Other findings such as left ventricular (LV) early diastolic tissue velocity (e'), LA volume, LA strain and LV global longitudinal strain are also related to LA filling pressure, and these parameters are useful for the diagnosis of HFpEF. Although some patients with HFpEF do not have abnormalities in these indices at rest, they may develop abnormalities in LA filling pressure exclusively during exercise.1 Therefore, HFpEF is difficult to diagnose in symptomatic elderly outpatients with limited exercise capacity (Figure 1). There are two types of probability tests in which HFpEF is present in patients with dyspnoea based on the findings obtained in clinical settings: the H2FPEF score2 and the HFA-PEFF score.3 Although both scores included echocardiographic indices, the H2FPEF score focuses on comorbidities, whereas the HFA-PEFF score includes natriuretic peptide levels. Most patients with HFpEF have a history of hypertension, and their blood pressure levels may no longer be elevated by medications at the time of evaluation.4 Furthermore, most patients with HFpEF show a high BMI, which is not the case in older patients.5 Because natriuretic peptide levels are occasionally within normal limits in patients with HFpEF, they may not be necessary for the diagnosis of HFpEF.2 Haemodynamic exercise tests play a key role when one cannot make a definitive decision of developing HFpEF. Noninvasive exercise testing does not discriminate between patients with cardiac and non-cardiac causes of dyspnoea.6 Although exercise echocardiography is a useful tool in place of invasive stress testing, there are contradictory results regarding the significance of the E/e' ratio for LA filling pressure.7, 8 Since LV diastolic function is affected by afterload, we recently evaluated LV diastolic function as a vascular resistance-integrated index: the ratio of LV diastolic elastance (Ed) to arterial elastance (Ea) = (E/e')/(0.9 × systolic blood pressure).9 This noninvasive index shows the ratio of LA filling pressure to LV end-systolic pressure and is a significant index of all-cause mortality in a multivariate Cox proportional hazards regression analysis performed by adjusting for age, comorbidities, natriuretic peptide levels and echocardiographic indices in elderly patients with HFpEF.10, 11 Stress tests cause an elevation of systolic blood pressure, and the extent of the change in blood pressure is different in each patient. Using an index such as Ed/Ea, the conflicting consequences described above may disappear. For the diagnosis of heart failure (HF), patients are aware of certain symptoms such as exertional dyspnoea, chest discomfort and fatigue. However, the symptoms are sometimes obscured in older patients. In clinical settings, it is important to identify patients with latent HFpEF who will likely be admitted for HF treatment in the near future. When populations with asymptomatic hypertension undergo invasive exercise stress or exercise echocardiography, some portions may exhibit an elevated LA filling pressure during exercise. We may call these populations ‘pre-HFpEF’ patients in terms of clinical entity.12 Asymptomatic LV hypertrophy is a potent risk factor for HFpEF.13 The main pathologies of HFpEF include cardiomyocyte remodelling and interstitial collagen deposition resulting from endothelial dysfunction and microvascular inflammatory changes.14-16 These latent populations might have treatment as ‘pre-HFpEF’ to avoid admission for HF. The diagnosis and treatment of asymptomatic elderly outpatients without a history of admission for HF, who possibly show exercise-induced elevation of LA filling pressure, remain to be defined. Recently, emerging differences in the phenotyping of HFpEF have been shown.17, 18 Asymptomatic patients with certain phenotypes must undergo stress echocardiography to diagnose pre-HFpEF. Under these conditions, the measurement of the Ed/Ea ratio may play a key role in the discrimination of several phenotypes in a multivariate model. As the medications and prognosis for HFpEF may be different from its phenotype,19, 20 it is urgent to elucidate the type of HFpEF or pre-HFpEF that should be medicated according to treatment type to reduce the incidence of admission for HF and mortality in the era of a super-ageing society. Among elderly clinic patients, we must detect symptomatic outpatients with possible HFpEF as well as asymptomatic ‘pre-HFpEF’ outpatients to reduce admission for HF in the near future and to conserve resources in the era of a super-ageing society. A vascular resistance-integrated LV diastolic index, (E/e')/(0.9 × systolic blood pressure), may play a role in revealing these patients in addition to possible comorbidities. None. No funding was received in relation to this article.