Hypertrophic cardiomyopathy (HCM) is a hereditary form of cardiomyopathy. Patients with HCM often have a range of clinical symptoms, including palpitations, dyspnea, and reduced exercise tolerance. An abnormal left ventricular (LV) diastolic function is an early pathophysiological change and an important cause of HCM progression. At present, echocardiography is the preferred non-invasive imaging method for the evaluation of LV diastolic function; the core indicator is E/e’. The American Society of Echocardiography/European Society 2016 Guidelines for Cardiovascular Imaging recommended an E/e’ $>$14 as the cut-off value for increased left ventricular filling pressure and an E/e’ 8 as the normal cut-off value [1]. Previous studies on LV diastolic function in HCM mainly focused on its causes and its impact on prognosis, while less attention was paid to exercise capacity and the tolerance of patients with HCM and an E/e’ of 8–14. Left atrial function plays an important role in the filling pressure of the left ventricle. Two dimensional Speckle tracking Technology (2DSTI) can sensitively and specifically evaluate LA function by analyzing LA strain (LAS). Over recent years, LAS has been shown to have the potential to independently assess LV diastolic function in a rapid and simple fashion [2]. However, LA function is very sensitive to stress [3]. Therefore, in this study, we used treadmill stress echocardiography combined with 2DSTI to evaluate the LA reservoir, conduits, contraction function, and its impact on metabolic equivalents (METS) in HCM patients with an E/e’ ranging from 8 to 14.

LV diastolic dysfunction includes insufficient or brady relaxation in the early stages of disease progression, and in advanced stages of disease, manifests mainly as reduced compliance and increased stiffness. These changes led to increased LV filling pressure and subsequently, may cause an increase in LA and pulmonary venous pressure. LA reservoir strain occurs during systole when the pulmonary veins fill the LA, thus causing the LA wall to stretch; this corresponds to LV isovolumic contraction and isovolumic relaxation. LA conduit strain occurs in the early diastolic period; when the mitral valve is opened, the LA blood content flows into the LV; this refers to the rapid filling period and the slow filling period which is regulated by LA compliance (LV relaxation and compliance). LA contraction strain is dependent on venous return and LV end-diastolic pressure [14, 15]. LAS alterations tend to progress between all stages of diastolic dysfunction. In particular, reservoir strain is significantly better than GLS and E/e’ and has been shown to predict cardiovascular events and represent a more accurate diagnostic tool to help classify diastolic dysfunction [16, 17, 18, 19, 20, 21, 22].

A previous systematic analysis and meta-evaluation of 2452 healthy subjects [23] showed that the normal reference range for LA reservoir strain was 39% while that of conduit strain was 23% and contraction strain was 17%. Our current analysis yielded similar outcomes: the LA reservoir, conduit and contraction strain were 44.39 $\pm$ 8.20%, –27.97 $\pm$ 6.45%, and –16.41 $\pm$ 3.73%, respectively. A previous study [24] showed that when the LA reservoir strain was 19%, there was a significant association between increased LV filling pressure and diastolic dysfunction. Another study [25] showed that a LA reservoir strain 18% and contraction strain 8% were more predictive of increased LV filling pressure than LAVI and conventional doppler parameters (p 0.05). Furthermore, the accuracy of diagnosed elevated LV filling pressure with only reservoir strain was 75%, the accuracy of diagnosed elevated LV filling pressure with only contraction strain was 72%, and the accuracy of LAS combined with conventional parameters in diagnosed elevated LV filling pressure was 82%. Gillebert et al. [26] reported that reduced LA reservoir function was a sensitive marker of early diastolic dysfunction and the coexistence of low reservoir and low contraction function indicated more severe heart failure, atrial fibrillation, thrombotic complications, acute heart failure syndrome and even death. When LV diastolic function decreases, end-diastolic volume increases, thus leading to a reduction in LA reservoir strain. Therefore, a reduction in LA reservoir strain may reflect a reduction in LV diastolic function at a time point much earlier. In the HCM-2 group, although LA and LAVI were still within the normal range, the resting E/e’ was also between 8 and 14; furthermore, the resting LA reservoir strain (25.05 $\pm$ 5.23% vs 44.39 $\pm$ 8.20%), conduit strain (–17.11 $\pm$ 8.27% vs –28 $\pm$ 6.45%) and contraction strain (–8.76 $\pm$ 0.91% vs –16.4 $\pm$ 3.73%) were all significantly lower than normal group (p 0.001), thus indicating that the reservoir, conduit, and contraction function were reduced (as shown in Figs. 6,7). The reduction of LA reservoir strain indicated that LA relaxation and compliance had been reduced, therefore suggesting that patients in the HCM-2 group may have entered the early stage of LV diastolic hypofunction. In this study, the LA conduit and contraction strain in the HCM-2 group were reduced, thus indicating that the compliance of the left atrium was reduced. The LA can reduce the function of the conduit by increasing the end-diastolic contraction compensation. In late diastole, the LA myocardial contraction was reduced, the ability to actively pump blood to the LV was reduced, and the contraction function of the LA was also reduced. We also found that the HCM-2 group had “lighter” impairment of the LA reservoir and conduit strain than the HCM-1 group; importantly, there was no difference between the HCM-1 and HCM-2 groups with regards to LA contraction strain (p $>$ 0.05). Thus, it is possible that LA contraction strain was not sensitive enough for the detection of reduced diastolic function. LA contraction function is a biphasic process in the course of the disease; when diastolic function is damaged in the early stage, the LA contraction function still has a certain reserve which can increase the contraction capacity. As the diastolic function further decreased, the LA pressure further increased, exceeding the reserve capacity of LA contraction; thus, LA contraction dysfunction and LA contraction strain decreased. Therefore, when the E/e’, as measured by conventional echocardiography, is between 8 and 14, if the LAS is measured in combination, the diastolic function can be quickly and accurately assessed, thus meeting the needs of clinical work for predicting cardiac function.

Fig. 6.

The LAS of rest and peak in Normal group. LAS, left atrial strain.

Fig. 7.

The LAS of rest and peak in HCM group. LAS, left atrial strain.

Patel et al. [27] showed that the lack of LA functional reserve and reduced LV diastolic function were associated with reduced exercise capacity. In the present study, we found that there was no statistical difference in the LA reservoir strain, conduit strain, and GLS in the HCM-1 group when compared between peak level and the rest stage. Moreover, there was no statistical difference in LA reservoir strain or conduit strain in the HCM-2 group when compared between peak level and the rest stage, thus indicating that the reserve capacity of the LA reservoir, conduit, contraction strain and GLS were reduced in both groups. It was also found that at rest and peak stages, the LA reservoir strain and conduit strain in the HCM-2 group was higher than in the HCM-1 group. In the HCM-2 and HCM-1 groups, the LA contraction strain reserve ($\mathrm{\Delta }$LASct%) and the $\mathrm{\Delta }$GLS% were significantly lower than in the normal group; furthermore, the $\mathrm{\Delta }$LASct% and $\mathrm{\Delta }$GLS% of the HCM-2 group were higher than that the HCM-1 group (p 0.05). The Rest-LASr, Rest-LAScd, Peak-LASr, and $\mathrm{\Delta }$LaSct% were all positively correlated with METS; these indices were significantly higher in the HCM-2 group than in the HCM-1 group (p 0.05). These findings are consistent with the fact that the METS in the HCM-2 group was higher than that in the HCM-1 group and that there was no statistical difference in METS when compared between the HCM-2 group and normal group. The better METS in the HCM-2 group can be attributed to the higher reservoir strain and conduit strain both in the rest and peak stages, and the Peak-GLS in the HCM-2 group in addition to the higher contraction strain reserve capacity and the higher $\mathrm{\Delta }$GLS%. The contraction of skeletal muscle after exercise promoted a greater amount of venous return to the right atrium. The contraction strain reserve of the HCM-2 group was still able to compensate for the left ventricle by increasing the atrial contraction and pumping blood to make up for the stroke volume with reduced diastolic function. Furthermore, there was a greater Peak-GLS in the HCM-2 group after exercise, thus indicating enhanced left ventricular contraction function. Although the METS in the HCM-2 group and normal group were not significantly different, the real hemodynamics and pathophysiological mechanisms underlying the METS were different; thus, the risk of cardiovascular events also differed.

In a previous study, Badran et al. [28] found that in a normal group, the rest and peak HR were 82 $\pm$ 8 and 165 $\pm$ 1, the rest and peak SBP were 118 $\pm$ 3 and 155 $\pm$ 7, and the rest and peak DBP were 79 $\pm$ 4 and 97 $\pm$ 4l. However, in the HCM group, the rest and peak HR were 70 $\pm$ 11 and 142 $\pm$ 21, the rest and peak SBP were 128 $\pm$ 22 and 145 $\pm$ 32, and the rest and peak DBP were 81 $\pm$ 14 and 81 $\pm$ 12. Analysis showed that both the rest and peak SBP differed significantly between the normal and HCM groups. In the present study, we obtained similar results; in the normal group, the rest and peak HR were 71 $\pm$ 10 and 173 $\pm$ 6, the rest and peak SBP were 120 $\pm$ 8 and 168 $\pm$ 15, and the rest and peak DBP were 78 $\pm$ 8 and 77 $\pm$ 11. However, in the HCM group, the rest and peak HR were 79 $\pm$ 11 and 172 $\pm$ 15, the rest and peak SBP were 126 $\pm$ 24 and 172 $\pm$ 27, and the rest and peak DBP were 77 $\pm$ 12 and 75 $\pm$ 17. Only rest SBP differed significantly between the normal and HCM groups (Table 1). For the HCM-1, HCM-2, and normal groups, we observed significant differences for SBP before and after exercise (Table 4). During exercise, the sympathetic nerve is excited, the HR increased, the heart pumps more blood, the pressure of blood vessel wall increases, which can lead to increased blood pressure. The peak SBP does not exceed 190–210 mmHg at normal, while the DBP basically does not change or slightly decreases. Because SBP is the pressure of blood against the walls of blood vessels when the heart contracts. During exercise, the circulation speeds up and the volume of heart strokes increases. The more blood the heart pumps out, the more pressure is created against the walls of blood vessels, so the SBP rises. DBP mainly reflects vascular elasticity and peripheral circulation resistance. In order to obtain more oxygen during exercise, the body reduces the resistance generated by peripheral arterioles and increases the blood supply to the exercise system through blood pressure regulation mechanism, thus making DBP stable or slightly decreased.

We found that the EDV in the HCM group at rest and peak stages was higher than in the normal group. Furthermore, the LV-GLS and LAS in the HCM group at the peak stage were smaller than in the rest stage. Thus, from a mechanical point of view, the peak myocardial contractility was reduced, therefore the EDV increased. In a previous study, Wu et al. [29] study found that in a normal group, the rest and peak LV-GLS were –20.8 $\pm$ 2.1 and –27.6 $\pm$ 1.9, $\mathrm{\Delta }$GLS was –4.75 $\pm$ 1.78, while in the HCM group, the rest and peak LV-GLS were –17.1 $\pm$ 2.8 and –20.0 $\pm$ 3.3, $\mathrm{\Delta }$GLS was –2.93 $\pm$ 1.58, the age of the subject was 52 $\pm$ 12, and the LV-GLS increased after exercise in both the normal and HCM groups. Badran et al. [28] found that in a normal group, the rest and peak LV-GLS were –18.5 $\pm$ 2 and –23.1 $\pm$ 2.7, $\mathrm{\Delta }$GLS was –3.4 $\pm$ 1.13, furthermore, in the HCM group, the rest and peak LV-GLS were –13.5 $\pm$ 5.8 and –11.8 $\pm$ 4.9, $\mathrm{\Delta }$GLS was –3.23 $\pm$ 5.0, the age of the subjects was 42 $\pm$ 12, and the LV-GLS increased after exercise in the normal group but decreased after exercise in the HCM group. Mahfouz et al. [30] found that in a normal group, the LA reservoir, conduit, and contraction strain were 52.1 $\pm$ 8.2%, –27.5 $\pm$ 4.4%, and –21.5 $\pm$ 4.0%, respectively. Similar results were obtained in the present study: in the normal group, the rest and peak LV-GLS were –25.25 $\pm$ 2.27 and –35.67 $\pm$ 2.5, while in the HCM group, the rest and peak LV-GLS were –20.07 $\pm$ 2.95 and –18.91 $\pm$ 6.09, respectively. The LV-GLS increased after exercise in the normal group but decreased after exercise in the HCM group (as shown in Figs. 8,9). LA reservoir, conduit and contraction strain were 44.39 $\pm$ 8.20%, –27.97 $\pm$ 6.45%, and –16.41 $\pm$ 3.73%, respectively. A previous study found that $\mathrm{\Delta }$GLS = GLS_peak-GLS_rest, but we consider that the percentage of the increase is more reasonable and accurate; therefore, in our study, we used the $\mathrm{\Delta }$GLS%, $\mathrm{\Delta }$LASr%, $\mathrm{\Delta }$LAScd%, $\mathrm{\Delta }$LASct%; these values were 40%, 51%, 40%, and 74%, respectively. Although it seems that the LV-GLS data from the normal group in our study was larger than that in the two previous studies, it is also important to consider that LV-GLS may be affected by age, race, gender; furthermore, the sample size of our normal group was 30 and may have had an impact on our findings. The blood flow of the left atrium increased sharply in the supine position immediately after the exercise stopped, and the deformation of the left atrium reached its maximum state, the following factors were taken into consideration, which would also lead to increased the LA strain after exercise cessation: (1) with the increase of intensity during exercise, the contractile force of the heart is enhanced, and the venous return to right atrium is increased; (2) the contraction and diastole of skeletal muscle during exercise can lead to the increase of venous return, (3) with the increase of intensity of exercise, the intrapleural pressure is negative pressure, the transmural pressure of the large vein in the thoracic cavity is larger, during inspiration, the thoracic volume is increased, and the negative pressure value of the pleural cavity is further increased. The large vein in the chest cavity and the right atrium are more dilated, the pressure is further reduced, the blood flow in the peripheral vein returns to the right atrium, and the blood flow from the pulmonary vein into the left atrium increases, (4) after the peak stage, the patient immediately shifts to the left decubitus position, the venous return flow increases. Genovese et al. [2] found that: in the normal group, LA strain is load dependent but to a lesser degree than LA volume, LA strain had relative advantage over LV volume in diagnostic paradigm. The dynamic images we collected were clear and were carried out in strict accordance with the analytical process; when heart rate reached approximately 170 bpm at the peak stage of exercise, we analyzed the images three times and took the mean value. All data are true and effective. At present, there are very few studies on left atrial strain after exercise. Therefore, we need to further expand the sample size, acquire the reference range of left atrial strain in healthy subjects after exercise, and facilitate the improvement of our follow-up research.

Fig. 8.

The GLS of rest and peak in normal group. GLS, LV global longitudinal strain.

Fig. 9.

The GLS of rest and peak in HCM group. GLS, LV global longitudinal strain; HCM, hypertrophic cardiomyopathy.

As shown in Figs. 2,3 and Tables 6,7, the AUC for LaSr was 0.759 (sensitivity 0.833, specificity 0.6, cutoff value –16.91%) and had the greatest discriminatory value. However, the specificity of LaSr was low, although the sensitivity and AUC were high enough to reduce the chances of missed diagnosis. In this study, when the AUC of E/e’ was 0.757, the cutoff value was 17.06, at this time, the index E/e’ and the AUC cannot be used in patients with E/e’ between 8 and 14. Therefore, we need to add LASr into the model to assist clinical judgment. When Rest-LASr+E/e’-rest was used as the prediction model, the AUC was 0.800. Although E/e’ and LASr+E/e’ had similar AUC values (0.757 vs 0.800, p = 0.4386) (as shown in Fig. 10) , the technology and principle for obtaining E/e’ and LASr are different. E/e’ was obtained through pulsed wave-tissue Doppler imaging, this technique is angle dependent. In addition, E/e’ can be affected by valve and left ventricular outflow tract obstruction, and was limited in terms of clinical application [23]. LA strain is based on speckle tracking technology which can quantitatively evaluate mechanical function of the atrial myocardium by analyzing myocardial deformation and can analyze the working characteristics of the atrial myocardium in different cardiovascular diseases; furthermore, this parameter has been shown to have predictive value for the risk of adverse cardiovascular events. LASr can be accurately measured even in combination with valve disease, heart failure and arrhythmia; these conditions are more common in clinical diagnosis and treatment. Furthermore, speckle tracking echocardiography has become an increasingly standard imaging method in routine clinical practice [31]. Moreover, LASr is independent of LAV, LV-GLS, age, LVEF, and E/e’ [32]. Thus, when E/e’ lies between 8 and 14, the routine assessment of left atrial function can be an important strategy to supplement the current prediction models.

Fig. 10.

Comparison the AUC of E/e’ and LASr+E/e’. LASr, LA reservoir strain; E/e’, the ratio of early diastolic forward mitral flow velocity and early-diastolic mitral annular velocity.

In conclusion, analysis showed that when the E/e’ was between 8 and 14, the LAS and reserve capacity of HCM patients were significantly reduced. Our findings suggest that the routine assessment of LAS +E/e’ can be a strategy with which to supplement current predictive models and facilitate clinical management strategies.

This study was limited to a single center sample, and many simultaneous factors affected METS, such as systemic inflammation, endothelial dysfunction, changes in intracellular and extracellular structures of cardiomyocytes, skeletal muscle bioenergetics, pulmonary functional status, mitral and tricuspid regurgitation, left ventricular outflow tract obstruction,etc, the pathophysiological changes of HCM may result in the above problems, which we did not refine analysis. At the same time, due to the different algorithms of the analysis software of different ultrasonic diagnostic instruments, the measurement results of LAS will be affected. It may be necessary to further expand the sample size and expand the ultrasonic diagnostic instrument model and analysis software to analyze LAS and METS in HCM patients, and follow-up to observe the occurrence of cardiovascular events in HCM patients. Another factor to consider is that the assessment of exercise tolerance was only based on METS. Other indices, such as the maximal oxygen consumption and the minute ventilation/carbon dioxide production slope, were not used. The original purpose of this study was to focus on the “grey uncertain population” with an E/e of 8–14 in order to quickly and accurately assist in the judgement of cardiac diastolic function and to predict exercise tolerance; thus, we considered METS as the short-term result. In clinical work, if it is necessary to predict the exercise tolerance but the patient cannot perform exercise tests (due to disability or other reasons), we can also consider using the LASr+E/e’ model to estimate the exercise tolerance in an approximate manner and provide a basis for clinical treatment decisions. Of course, we are also continuing to follow-up these patients and continue to observe long-term cardiovascular events for further detailed research reports.

The data used to support the findings of this study are available from the author upon request.

Contributions: (I) Conception and design—YS, LY, CL; (II) Administrative support—LY, CL; (III) Provision of study materials or patients—YS; (IV) Collection and assembly of data—YS; (V) Data analysis and interpretation—YS; (VI) Manuscript writing—All authors; (VII) Final approval of manuscript—All authors.

The retrospective research was approved by the Medical Ethics Committee of Sichuan Academy of Medical Sciences and Sichuan Provincial People’s Hospital (approval No.185). All procedures performed in this study involving human participants were in accordance with the Declaration of Helsinki (as revised in 2013). All procedures strictly followed the protection of patient privacy and all data were anonymized in the study. All patients signed informed consent forms.

Not applicable.

This work was supported by the Fundamental Research Funds for the Central Universities [Grant Number: ZYGX2020ZB038].

The authors declare no conflict of interest.