We conducted prospective cross-sectional study on hypertensive patients who received medical care at the hospital from March 2022 to October 2023. The inclusion criteria for patients diagnosed with hypertension (HT) adhered strictly to the European Guidelines for the Prevention and Treatment of Hypertension (6), included 112 patients in HT group and 51 patients in HT-HFpEF group. The HT-HFpEF patients needed to satisfy the following criteria: duration of hypertension ≥5 years and according to the 2019 HFA–PEFF diagnostic algorithm: the consensus recommendation from the Heart Failure Association (HFA) of the European Society of Cardiology (ESC) (7), the score ≥ 5 points implies definite HFpEF; while mean score for the HT group was 2.9 ± 1.6. Patients with poor image quality, atrial fibrillation, moderate to severe valvular heart disease,congenital heart disease, hypertrophic and dilated cardiomyopathy,amyloidosis,coronary heart disease with stenosis more than 50%,constrictive pericarditis were excluded (Figure 1). We also included 62 normal subjects who underwent echocardiography examination during the same period with mean age of 56 ± 6 years.

M5S and 4V1 probes of GE-vivid 95 instrument (GE Ultrasound, Horten, Norway) were utilized for the image acquisition. Patients were instructed to hold their breath for 5–7 s and the 3D image frame rate was ensured to exceed 25 frames per second. Left ventricular end-diastolic diameter (LVEDd), right ventricular systolic pressure(RVSP), left ventricular mass index(LVMI), biplane-derived ejection fraction(EF), stroke volume (SV), cardiac index (CI) were obtained. From apical four-chamber views, E and A wave values were measured by transmitral flow velocity, early diastolic (e') and late diastolic (a') velocity were obtained using tissue doppler imaging (TDI),thus allowing to calculate E/e’ ratio. Auto function imaging (AFI) was applied to obtain global longitudinal strain (GLS),Global work index (GWI) represents the total workload under the pressure-strain curve, which is the product of strain and systolic pressure. Global constructive work (GCW) contributes to LV ejection and is the sum of work performed during LV systolic myocardial shortening and isovolumic diastolic myocardial lengthening.Global wasted work (GWW) is the work performed by myocardial lengthening during systole or shortening during isovolumic relaxation. Global work efficiency (GWE) is calculated as the ratio of GCW/(GCW + GWW) (8).

The treadmill exercise stress test was conducted according to the Bruce protocol and the test was stopped once the target heart rate (220-age) reached 85% (9). We recorded hemodynamic parameters as peak systolic arterial pressure (SBP-peak), peak heart rate(HR-peak), METs, and echocardiography performed immediately after exercise to measure E/e'(Ex-E/e'). Effective arterial elastance (Ea) was calculated as end-systolic pressure(ESP)/SV,and LV end systolic elastance (Ees) was derived as ESP {end-systolic volume (ESV) -V0}, with V0 considered negligible compared to ESV. The Ea/Ees ratio, equivalent to ESV/SV,was used to determine ventricular-arterial coupling (10). Additionally, the total systemic vascular resistance index (SVRI) was determined using the formula{ (mean BP- right atrial pressure(RAP)} × 80/cardiac index (CI) (11), which reflects cardiac afterload,the estimation of RAP is based on the inferior vena cava diameter and its collapsibility with respiration (12).

LA images were analyzed using EchoPAC 204 workstation software(4D-auto LAQ),a well-established and widely applied technique (13). Manual adjustments were made on apical four-chamber, three-chamber, and two-chamber views at end-systole and end-diastole to outline the LA border while excluding pulmonary veins and LA appendage regions, the process performed by experienced sonographers. This process generated time-volume curve for LA, and provided strain data, including longitudinal and circumferential reservoir phase strain (LASr,LASr-c), conduit phase strain (LAScd,LAScd-c), pump phase strain (LASct,LASct-c),maximum LA volume (LAVmax),minimum LA volume (LAVmin), pre-atrial contraction LA volume (LAVpreA), LA ejection volume (LAEV) and LA emptying fraction (LAEF) were automatically calculated by 3D LA analysis (14). The left atrial volume index (LAVI) was calculated by normalizing LAVmax according to body surface area (BSA),the left atrial stiffness index-LASI (E/e'/LASr, E/e'/LASr-c) was evaluated by computing the ratio of E/e' value to LA reservoir strain.We referred to previous study by snader et al. (15), who studied exercise tolerance METs to predict all-cause mortality in population of 3,400 patients; METs <6 were considered to be associated with reduced and poor long-term prognosis. Therefore, we evaluated the diagnostic value of LASI, Ex-E/e', LAVI, and RVSP in HT-HFpEF patients with METs <6,and conducted Delong test.

Statistical analysis was conducted using SPSS version 26.0.Continuous variables were compared using one-way analysis of variance (ANOVA) with Bonferroni correction or the non-parametric Kruskal-Wallis test. General linear regression analysis was conducted to evaluate the contributions of clinical variables, echocardiographic parameters, and exercise hemodynamic measures to exercise capacity. Variables identified as significant in univariable analysis were subsequently incorporated into multivariable models to determine independent predictors. An optimized model was generated, and partial r was defined as the association between the predicted variable and achieved METs. Key parameters were utilized for receiver operating characteristic (ROC) and area under the curve (AUC) analysis, to identify HT-HFpEF patients with reduced exercise capacity,and AUC values were compared using the DeLong test. Intra/inter-observer variability for LA function measurements was determined by estimating intraclass correlation coefficients (ICCs) for LA longitudinal strain indices including reservoir strain, conduit strain, and contraction strain, with ICCs of 0.92(95%CI: 0.82–0.97)/0.84(95%CI:0.68–0.93); 0.91(95%CI:0.81–0.94)/0.79(95%CI:0.61–0.92), and 0.89(95%CI:0.74–0.95)/0.81(95%CI:0.69–0.94), respectively.

Significant differences were identified in age, body mass index (BMI), and BSA among the three groups (P < 0.05). Additionally, variations in clinical medication were observed between different HT groups, as shown in Table 1. Significant differences were found among the three groups for LVEDd, LVMI, E/e' and RVSP (P < 0.05). Similarly, SV and CI showed significant differences across the groups (P < 0.05), except for EF (P > 0.05). Baseline GLS levels were notably lower in both the HT-HFpEF and HT groups compared to the control group (P < 0.05) (Table 2).

Myocardial work indices, which serve as load-independent parameters revealed decrease in GWI, GCW, and GWE in HT-HEpEF group (P < 0.05). Conversely, in the HT group, GWI/GCW exhibited a significant increase (P < 0.05), which is consistent with previous findings (16). Ea/Ees as metric indicative of ventricular-arterial coupling, did not vary significantly (P > 0.05); However, SVRI was significantly elevated in HT-HFpEF patients compared to those in HT group, with the lowest values in normal subjects. (P < 0.05). In terms of post-exercise variables, HR-peak did not differ significantly between the groups (P > 0.05). However, as expected, Ex-E/e' was significantly higher in HT-HFpEF group, and METs reflect exercise capacity was markedly lower compared to the other two groups (P < 0.05) (Table 2).

As depicted in Table 4, the values of LAVmin, LAVmax, LAVI, and LAEF were significantly elevated in HT group (P < 0.05). In the HT-HFpEF cohort, LA volume variables were increased at all time phases; however, for LAEF, longitudinal and circumferential strain were notable decreased (P < 0.05).LASr, LASr-c, LASct-c were differed significantly between the HT and HT-HFpEF groups (P < 0.05), thus indicating that these represent may serve as sensitive indicators of cardiac dysfunction in HT-HFpEF patients.

The LASI, derived from left atrial strain, showed a consistent upward trend, being significantly higher in the HT-HFpEF group (P < 0.05),specifically, the ratio of E/e' to LASr was 0.58 (0.52, 0.66) in the HT-HFpEF group vs. 0.46 (0.33, 0.59) in the HT group, similarly, the ratio of E/e' to LASr-c was 0.5 (0.38, 0.6) vs. 0.36 (0.27, 0.43). These findings emphasize LASI as key index for evaluating impaired atrioventricular compliance in HT-HFpEF patients, underscoring its critical role in detecting atrioventricular dysfunction in this population (Table 3). Conversely, no significant differences were observed in LAScd and LAScd-c between the HT and HT-HFpEF groups, indicating that conduction function remains preserved (P > 0.05). Figure 2 presents an example of different time-phase 3D LA volume and strain parameters for HT-HFpEF and control patient. Figure 3 depicts the points utilized for determining circumferential strain lines.

A general linear regression analysis was conducted to assess the relative contribution of each variable to exercise capacity through a heatmap,and lambda plots was used to facilitate analysis. Significant statistical correlations (P < 0.05) were observed between LA function (inclusive of volume and strain), GLS, GCW, SVRI and METs; indicated that these parameters could be considered as candidate variables in multiple regression models to determine the independent predictors of METs. Subsequently, Age, SBP-rest, and BMI emerged as independent general predictors of METs. And an analysis combining E/e'/LASr, E/e'/LASr-c with LVAI, GLS, E/e', RVSP and SVRI was conducted (referred as echocardiographic Model 2). Ultimately, a comprehensive model that integrates both general and echocardiographic predictors (Model 3) and found that E/e'/LASr still independently related, in Table 4. To assess the clinical value of E/e'/LASr, we utilized unstandardized coefficients from Model 3 to generate overlay scatter plots, comparing METs achieved during exercise with those predicted by the model, the results demonstrated strong positive linear correlation. Partial r represented the strength of association between the model variable and METs achieved (Figure 4). Notably, LASI represents the elastic modulus during diastole, and serves as a marker of myocardial stiffness or flexibility. The correlation diagram between LASI (E/e'/LASr, E/e'/LASr-c),candidate variables and METs are displayed (P < 0.05) (Figure 5). The strongest correlations were observed between Ex-E/e’ and METs (r = −0.521, P < 0.001), followed by E/e’/LASr and METs (r = −0.462, P < 0.001).

The AUC and 95% CI values for the variables are presented in Figure 6. The following cut-off values and AUCs were identified: E/e'/LASr (cut-off: 0.46; AUC: 0.836; 95% CI: 0.805–0.892), and E/e'/LASr-c (cut-off: 0.33; AUC: 0.709; 95% CI: 0.660–0.797). Additionally, the diagnostic performance for LAVI (AUC: 0.789; 95% CI: 0.753–0.865), Ex-E/e' (AUC: 0.867 95% CI: 0.812–0.913), and RVSP (AUC: 0.825; 95% CI: 0.789–0.881) were evaluated. No significant differences were found between E/e'/LASr and Ex-E/e' (Delong test, P = 0.239) (Graphical Abstract).

In this study, we mainly focused on the hypertensive subtype of HFpEF, However,the sample size was relatively small and needs to be increased in future work. For the heterogeneity of HFpEF, including patients both with hypertension and renal dysfunction, did not entirely excluded. Despite the use of high-quality spatial and temporal resolution in our three-dimensional speckle tracking analysis, there was still inter-observer difference,and intervendor variability, image quality on loading conditions that could have theoretically impacted upon our results. In addition, long-term follow-up and large-scale prospective studies are now needed to determine the clinical predictive value of early changes in left atrial function in patients with HT.Furthermore, more in-depth analysis of pathological changes in the left atrium is required across different cardiovascular diseases to enhance our understanding of the relevant mechanisms involved. In addition, LVEDP, E/e',and LASr may be affected by volume overload and may change with the use of diuretics. We lacked objective data relating to the assessment of pulmonary capillary wedge pressure (PCWP) based on invasive hemodynamics. Further studies are needed to investigate whethe increased LASI can represent an independent determinant for predicting HF hospitalization or mortality.