Normal reference values and determinants of right heart strain derived from speckle tracking echocardiography: a systematic review and meta-analysis Authors: Ningbo Yin, Jingjing Shi, Wenqi Liu, Wanzhen Li, Jiaying Li, Fan Zhang, Yifei Chen, Yiyuan Gao, Jie Lin Published: 2026 ABSTRACT BackgroundNormal ranges for speckle tracking echocardiography (STE)-derived right atrial (RA) and right ventricular (RV) strain indices have not yet been fully standardized. This study aimed to quantitatively summarize available evidence regarding normal RV and RA deformation parameters in healthy adults and to investigate factors contributing to measurement variability.MethodsPubMed, Embase, and Web of Science were systematically searched for studies published until July 1, 2025 that reported STE-derived right-heart strain and strain-rate measurements in healthy adults. A random-effects model was applied to generate pooled estimates for RV global longitudinal strain (RV-GLS), RV peak systolic and diastolic longitudinal strain rates (RV-PSSRL and RV-PDSRL), RA reservoir, conduit, and booster strain (RA-εs, RA-εe, and RA-εa), as well as RA peak positive, early negative, and late negative strain rates (RA-SRs, RA-SRe, and RA-SRa). Subgroup and meta-regression analyses assessed the effects of race, sex, age, feature tracking (FT) software, STE vendor and STE imaging mode [two-dimensional (2D)-and three-dimensional (3D)-STE]. Heterogeneity was assessed using the I2 statistic, and publication bias was examined by Egger regression analysis.ResultsThe pooled means for RV-GLS, RV-PSSRL, and RV-PDSRL were −25.62%, −1.50 s−1, and 1.69 s−1, respectively. RA-εs, RA-εe, and RA-εa were 42.95%, 20.86%, and 19.30%, respectively; while RA-SRs, RA-SRe, and RA-SRa were 2.03 s−1, −2.14 s−1, and −1.80 s−1, respectively. Variability in right-heart deformation measurements was significantly associated with race, sex, age, FT software, and STE imaging mode (all P<0.05), whereas the influence of STE vendor was generally limited (most P>0.05).ConclusionsThis study established pooled reference values for STE-derived RV and RA strain and strain-rate parameters in healthy adult populations. STE demonstrates strong potential as a practical and reliable technique for the quantitative assessment of right heart function in routine clinical practice. Introduction In recent years, accumulating evidence has highlighted the critical role of the right atrium (RA) and right ventricle (RV) in the diagnosis and prognosis of a wide spectrum of cardiovascular diseases (1). These chambers are now recognized as distinct anatomical and functional entities, rather than mere counterparts of the left heart (2,3). The concept of strain, originating from continuum mechanics, describes the degree of structural deformation and is defined as the relative change in a material’s original length (4). In cardiovascular imaging, myocardial strain represents the relative change in myocardial length, while strain rate, defined as the time derivative of strain, describes the speed of myocardial deformation (5). Traditionally, the right heart has received less attention than the left in cardiac imaging (6-8). However, with advances in imaging technologies, particularly speckle tracking echocardiography (STE), the accuracy and reproducibility of RA and RV strain parameters have been validated across a range of conditions, including pulmonary hypertension, heart failure, and valvular heart disease (9-11). Echocardiography is the first-line imaging modality for assessing cardiac structure and function (12-14). STE quantifies myocardial deformation by tracking naturally occurring acoustic speckles within the myocardium and offers advantages over cardiac magnetic resonance-feature tracking (CMR-FT), including wider availability, lower cost, and easier bedside application (15-17). As a result, STE has become a key technique for myocardial strain assessment (18), and a practical tool for quantifying right heart function in routine clinical practice (19). However, STE remains sensitive to acoustic window quality and is operator dependent, which may affect accuracy and reproducibility and highlight the need for robust normal reference values to support standardized clinical use (20). To date, most meta-analyses of myocardial strain have focused on left heart indices (21) whereas data on RA and RV strain are relatively limited and often incomplete for key parameters (22-24). Existing reference intervals for right heart strain are mainly derived from single-center or registry-based studies and are influenced by multiple confounders, including imaging vendor, geographic region, analysis software, and demographic factors such as sex and age, leading to substantial interstudy heterogeneity (25). Systematic identification and adjustment of these sources of variation are therefore essential to establish more reliable and generalizable reference intervals for right heart strain and to improve consistency across studies. Therefore, this study aimed to perform a systematic meta-analysis to establish normal reference intervals for the principal STE-derived strain parameters of the right ventricle and right atrium in healthy adults under different conditions, and to investigate potential sources of variation across these conditions, thereby providing broadly applicable reference standards for clinical practice and research. We present this article in accordance with the PRISMA reporting checklist (available at https://qims.amegroups.com/article/view/10.21037/qims-2026-1-0013/rc) (26). Methods This meta-analysis has been officially registered in the International Prospective Register of Systematic Reviews (PROSPERO, ID: CRD420251238094). Literature retrieval Employing a systematic methodology, two separate reviewers (N.Y. and J.S.) conducted extensive literature searches across various databases, such as PubMed, Embase, and Web of Science, with a focus on literature published until July 1, 2025. The search strategy was framed around the following keywords: echocardiography, STE, speckle tracking echocardiography, myocardial strain, right atrium, right ventricles, strain, among others. The comprehensive search formula, detailing the terms utilized and Boolean operators, can be found in Appendix 1. Both reviewers independently examined the titles and abstracts of all identified publications to evaluate their eligibility based on the study goals. Articles that did not fulfill the inclusion criteria were eliminated according to predetermined exclusion criteria, which included case reports, non-original articles (such as editorials, letters, and reviews), and studies unrelated to the research theme. Any inconsistencies observed during the selection process were addressed through consensus discussions to uphold methodological integrity and scientific reliability. The inclusion and exclusion criteria The criteria for inclusion in this research were established as follows: (I) studies that reported on RV or RA strain alongside strain rate parameters obtained from STE; (II) populations studied consisting of healthy adults over 18 years of age, having no known cardiovascular or metabolic disorders (such as diabetes or structural heart disease), absence of clinical indicators of cardiac dysfunction, and no identified abnormalities in MRI or STE; (III) studies that involved more than 30 healthy adult subjects who participated in STE; (IV) publications appearing in peer-reviewed journals written in English. The criteria for exclusion included: conference abstracts, review papers, animal research, studies involving competitive athletes, investigations using stress echocardiography or cardiac MRI methods, studies lacking a healthy control group, studies with fewer than 30 healthy participants undergoing STE, and articles published in non-peer-reviewed or non-English journals. Any differences arising during the study selection process were addressed through discussion and mutual agreement among the co-authors. Literature screening and data extraction All retrieved studies were imported into EndNote (version 20.0) for initial screening by two independent reviewers (N.Y. and J.S.). Titles and abstracts were carefully evaluated to exclude studies not aligned with the study objectives, while potentially eligible studies were identified according to the predefined inclusion and exclusion criteria. For studies with incomplete datasets, attempts were made to contact the corresponding authors to obtain additional information when necessary. Study screening and data collection were conducted in accordance with predefined criteria to ensure a rigorous and comprehensive assessment of the available literature. Prior to data extraction, a standardized electronic form was designed to uniformly collect relevant study information, including article title, author, year of publication, country of origin, sample size, STE vendor, imaging mode, feature tracking (FT) software, number of healthy participants, and other relevant details. The main objective of this meta-analysis was to determine reference ranges for STE-derived right-heart strain parameters, including RV global longitudinal strain (RV-GLS), RV peak longitudinal systolic and diastolic strain rates (RV-PSSRL and RV-PDSRL), RA reservoir, conduit, and booster strain (RA-εs, RA-εe, and RA-εa), together with RA peak positive, early negative, and late negative strain rates (RA-SRs, RA-SRe, and RA-SRa). Longitudinal Lagrangian strain refers to the percentage change in myocardial length from end-diastole (L0) to a given time point (L), calculated as: ε = (L – L0)/L0 × 100%. Strain rate is the temporal derivative of the strain curve (dε/dt), reported in s–1 (27). In all included studies, RV-GLS was determined as the mean strain value derived from three segments of the RV free wall: basal, midventricular, and apical. The interventricular septum was explicitly excluded from analysis, as its measurement is additionally influenced by left ventricular systolic function, which consequently lowers the absolute strain value of RV-GLS (28). Contour delineation algorithms identify cardiac borders, endocardial/epicardial tracking measures myocardial displacement, and temporal smoothing reduces noise to generate reliable deformation curves. These are critical steps for conducting effective strain analysis, and methodological differences among various FT software have led to significant heterogeneity in the studies. When overlapping populations were identified among different publications originating from the same dataset, priority was given to the study with the largest sample size in order to avoid data duplication and minimize potential bias in the pooled analysis. Data quality assessment Bias risk in the included studies was assessed using the Quality Assessment of Diagnostic Accuracy Studies-2 (QUADAS-2) tool implemented in Review Manager 5.3 (29). This instrument evaluates potential sources of bias related to study design and conduct, including participant selection, allocation procedures, blinding, and outcome assessment, thereby addressing selection, performance, and detection bias (30). Each study was examined across four domains: patient selection, index test, reference standard, flow, and timing, with the risk level categorized as low, high, or unclear. The results of the quality assessment were incorporated into the study characteristics and summarized in the RevMan risk-of-bias tables to ensure a standardized and thorough evaluation process. Statistical analysis Mean values and corresponding standard errors (SEs) were extracted from all included studies. Continuous outcomes were synthesized using a random-effects model, and pooled estimates were presented with 95% confidence interval (CI). Between-study and subgroup heterogeneity were assessed using the I2 statistic, with I2 values >50% considered indicative of substantial heterogeneity. Publication bias was evaluated by Egger’s test. In addition, meta-regression analyses were performed to investigate the potential effects of study-level covariates on heterogeneity. All statistical tests were two-sided, and a P value <0.05 was considered statistically significant. Statistical analyses were conducted using STATA software (version 17, StataCorp). Results Study selection A comprehensive literature search across databases yielded a total of 4,856 records. After duplicate removal, 3,602 records were retained for further evaluation. Screening of titles and abstracts led to the exclusion of 3,447 studies because of irrelevance to the study topic (n=1,443), inclusion of non-healthy populations (n=901), studies involving animals, athletes, pregnant women, or children (n=379), review articles (n=546), case reports (n=142), conference abstracts (n=27), and book chapters (n=9). Then 155 articles were included for further screening. Following full-text evaluation, 102 articles were excluded due to invalid strain data (n=37), sample size of healthy population <30 (n=25), publications in non-peer-reviewed English-language journals (n=16), and missing data (n=24). In total, 53 studies satisfied the eligibility criteria and were subsequently included in the final meta-analysis. The detailed procedure for study identification and selection is illustrated in the PRISMA flowchart (Figure 1). PRISMA flowchart. Study characteristics The characteristics of the included studies are presented in Table 1. Overall, 53 eligible studies involving 5,693 healthy adults were included in the present meta-analysis. All selected studies met the predefined search and eligibility criteria. Among them, 43 studies (n=3,010) reported RV strain data (43 for RV-GLS); 9 studies (n=1,132) reported RV strain rate data (7 for RV-PSSRL, 5 for RV-PDSRL); 18 studies (n=3,045) reported RA strain data (18 for RA-εs, 13 for RA-εe, and 11 for RA-εa); and 10 studies (n=423) reported RA strain rate data (9 for RA-SRs, 8 for RA-SRe, and 10 for RA-SRa). Table 1Characteristics of included studiesFirst author (ref.)YearSample sizeStudy designAge (years), mean ± SDMale [%]CountryImaging modeVendorSoftwareChamberAndreas P. Kalogeropoulos et al. (31)200839Prospective43±1813 [33]USA2DGEEchoPACRVMichael Becker et al. (32)201031Prospective23.1±2.614 [45]Germany2DGEEchoPACRVNowell M. Fine et al. (33)2013186Prospective44±1672 [39]USA2DPhilips/Siemens/GESyngo VVIRVMasoumeh Lotfi-Tokaldany et al. (34)201369Prospective30.03±5.3529 [42]Germany2DGENARVYuman Li et al. (35)201331Prospective37±1412 [39]China2DGEEchoPACRVDiletta Peluso et al. (36)2013210Prospective43±1587 [41]Italy2D/3DGEEchoPACRAMarijana Tadic et al. (37)201345Prospective38±70Serbia2D/3DGEEchoPACRA, RVMarcia M. Barbosa et al. (38)201438Prospective44±9.222 [58]Brazil2DGEEchoPACRVMarijana Tadic et al. (39)201455Prospective45±829 [53]Serbia2D/3DGEEchoPACRA, RVBenjamin C. F. Smith et al. (40)201440Prospective40.6±1224 [60]UK3DToshibaToshibaRVE. Durmus et al. (41)201540Prospective45.9±7.619 [48]Turkey2DGEEchoPACRA, RVDaniel A et al. (42)2015187Prospective37.2±1392 [49]Germany2DGEEchoPACRVMarijana Tadic et al. (43)201554Prospective51±825 [47]Italy2D/3DGEEchoPACRA, RVAntonello D’Andrea et al. (44)201645Prospective65.4±8.323 [51]Italy2DEsaoteX-strainRVAntonio Vitarelli MD et al. (45)201530Prospective46.2±13.411 [37]Italy3DGEEchoPACRVAntonio Vitarelli MD et al. (46)201530Prospective54±1513 [43]Italy2D/3DGEEchoPACRVLan Zhang et al. (47)201640Prospective60 ± 718 [45]China2D/3DPhilipsQLABRVMonica Roşca et al. (48)201530Prospective48±912 [40]Romania 2DGEEchoPACRVDenisa Muraru et al. (49)2016276Prospective44±17.88123 [45]Romania 2DGEEchoPACRVYonghuai Wang et al. (50)201648Prospective55.9±8.322 [46]China2DGEEchoPACRATor Skibsted Clemmensen et al. (51)201641Prospective51±1224 [59]Denmark2D/3DGEEchoPACRVLiYing Cai et al. (52)201755Retrospective62±1323 [42]USA2DPhilipsQLABRVMarijana Tadic et al. (53)201745Prospective49±1.324 [53]Serbia2DGEEchoPAC RVHala Badran et al. (54)201733Prospective32.9±12.0425 [76]Egypt2DEsaoteNARA, RVJae-Hyeong Park et al. (55)2018493Prospective47±15232 [47]South Korea2DGEEchoPAC RVAnna Brand et al. (56)2018123Prospective50.9±13.80Germany2DGEEchoPACRAAi-Li Li et al. (57)201830Retrospective54 ± 1414 [47]China2DGEEchoPACRVJunfang Li et al. (58)201830Prospective46.82±5.4517 [57]China2DSiemensVVIRAMihaela Berceanu et al. (59)201990Prospective30±859 [66]Romania2DGEEchoPACRVMaria Sanz-de la Garza et al. (60)201980Prospective37.1±4.540 [50]Spain2DGEEchoPACRVRoberto M. Saraiva et al. (61)201977Prospective40.4±10.436 [46]Brazil2DGEEchoPACRVYiu-Fai Cheung et al. (62)201943Prospective30.3±12.618 [42]China2DGEEchoPACRASánchez-Hernández JE et al. (63)201930Prospective27.3±9.35 [17]Mexico2DPhilipsCMQRAEsther Scheirlynck et al. (64)202082Prospective44±1142 [51]Belgium 2DGEEchoPACRVOzge Can Bostan et al. (65)202070Prospective33.9±9.544 [63]Turkey2DPhilipsNARA, RVVien T. Truong et al. (66)202061Prospective45±1330 [49]USA2DPhilipsEchoPACRAJun Huang et al. (67)202034Prospective46±1219 [56]China2DGEEchoPACRALaurie Soulat-Dufour et al. (68)20212,008Prospective47±171033 [51]NA2D/3DNANARAFatih Akkaya et al. (69)2021105Retrospective43.4±12.765 [62]Turkey2DPhilipsQLABRVAltug Cincin et al. (70)2021100Retrospective42.1±9.483 [83]Turkey2DPhilipsQLABRVBlanca Domenech-Ximenos et al. (71)202172Prospective33.58±3.4642 [58]Spain2DGEEchoPACRVRosa Lillo et al. (72)202149Retrospective48.3±1724 [49]Italy2DToshibaTomTecRVÜnlü et al. (73)202140Prospective22.9±3.421 [53]Turkey2D/3DGEEchoPACRVMohammad Iqbal Janhangeer et al. (74)202135Prospective30.9±7.9NAEgypt2DPhilipsQLABRVYanting Zhang et al. (75)202131Retrospective51.5±8.418 [58]China3DPhilipsTomTecRVLaura Houard et al. (76)202130Prospective35±1120 [67]Belgium2DPhilipsTomTecRVShehab Anwer et al. (77)202250Retrospective43.83±19.3135 [70]Switzerland2DGETomTecRA, RVThuy Thu Pham et al. (78)202230Prospective44±148 [27]Vietnam2DPhilipsEPIQ RVTing Wu et al. (79)202349Prospective52.29±11.1522 [45]China2DGEEchoPACRVEmre Gürel et al. (80)202332Prospective26.8±5.414 [44]Turkey2DPhilipsQLABRA, RVFlavia Fusco et al. (81)202449Retrospective38.48±14.2925 [51]Italy2D/3DGEEchoPACRVZsuzsanna Ladányi et al. (82)202439Prospective55±97 [18]Hungary3DGEEchoPACRVJie Hu et al. (83)202433Retrospective38.9±13.17 [21]China2DPhilipsQLABRA, RV2D, two-dimensional; 3D, three-dimensional; NA, not available; RA, right atrial; RV, right ventricular; SD, standard deviation. Characteristics of included studies 2D, two-dimensional; 3D, three-dimensional; NA, not available; RA, right atrial; RV, right ventricular; SD, standard deviation. Of the included studies, nine (16.98%) enrolled more than 100 participants, with the largest single cohort comprising 2,008 individuals. Nine studies exclusively recruited healthy or normal subjects (n=3,307; 58.09% of the total population), whereas the remaining 44 studies included a healthy or normal control group (n=2,386). Bias risks included in studies Figure 2 summarizes the assessment of bias risk and applicability concerns across the included studies. Most studies were judged to have a low risk of bias with minimal applicability concerns, suggesting overall good methodological quality and reliable applicability of the findings. QUADAS-2 scores for 53 studies. QUADAS-2, Quality Assessment of Diagnostic Accuracy Studies-2. Pooled analysis of normal values for RV strain and strain rate parameters The pooled mean reference value for RV-GLS across all studies was −25.62% (95% CI: −26.60%, −24.63%), with significant heterogeneity observed (Figure 3). Egger’s test revealed no significant publication bias (P=0.05). Subgroup and meta-regression analyses identified race, age, sex, FT software, STE vendor and STE imaging mode as significant contributors to heterogeneity in RV-GLS (all P<0.05). The forest plot of pooled normal value for RV-GLS. CI, confidence interval; ES, effect size; GLS, global longitudinal strain; RV, right ventricular. For RV strain rates, the pooled mean reference values for RV-PSSRL and RV-PDSRL were −1.50 s–1 (95% CI: −1.64, −1.36) and 1.69 s–1 (95% CI: 1.33, 2.05), respectively, both showing significant heterogeneity (Figures S1,S2). Egger’s test again indicated no significant publication bias (P=0.774 and 0.15 for RV-PSSRL and RV-PDSRL, respectively). Subgroup and meta-regression analyses demonstrated that race, age, and sex were significant sources of heterogeneity in RV-PSSRL (all P<0.05), whereas no variables significantly contributed to heterogeneity in RV-PDSRL (all P>0.05). Detailed results are summarized in Tables 2,3. Table 2Subgroup and meta-regression analyses for right ventricular and right atrial strain and strain rateVariablesPublication yearRaceSexAgeSoftwareVendorImaging modeCoefficient (95% CI)P valueCoefficient (95% CI)P valueCoefficient (95% CI)P valueCoefficient (95% CI)P valueCoefficient (95% CI)P valueCoefficient (95% CI)P valueCoefficient (95% CI)P valueRV strain RV-GLS−0.37 (−2.17, 1.43)0.675−2.60 (−5.04, −0.15)0.03*3.73 (1.46, 6.012)0.003*−2.33 (−4.37, −0.29)0.026*0.71 (0.25, 1.16)0.004*1.81 (0.56, 3.07)0.007*−3.09 (−5.99, −0.19)0.037*RV strain rate RV-PSSRLNANA−0.29 (−0.57, −0.01)0.043*−0.44 (−0.87, −0.01)0.046*−0.84 (−1.39, 0.29)0.011*NANA−0.08 (−0.27, 0.11)0.274NANA RV-PDSRLNANA−0.62 (−1.92, 0.68)0.1780.35 (−0.69, 1.38)0.285−0.33 (−1.74, 1.07)0.414NANA0.22 (−0.92, 1.36)0.584NANARA strain RA-εs−2.42 (−10.08, 5.24)0.516−0.06 (−11.99, 11.86)0.044*11.90 (−2.68, 26.49)0.03*−12.28 (−22.29, −2.28)0.02*−1.80 (−3.30, −0.30)0.024*2.68 (−3.26, 8.61)0.352NANA RA-εe2.21 (−6.85, 11.26)0.595−8.18 (−15.66, −0.70)0.036*−6.94 (−12.24, −1.65)0.022*6.14 (−1.00, 11.29)0.029*6.16 (0.71, 11.60)0.035*0.95 (−8.12, 10.03)0.818NANA RA-εa0.85 (−59.82, 61.52)0.888−4.21 (−7.69, −0.74)0.028*7.94 (4.21, 11.67)0.004*−4.22 (−8.01, −0.42)0.037*−6.41 (−11.91, −0.91)0.03*−0.88 (−79.73, 77.97)0.91NANARA strain rate RA-SRs0.26 (−0.50, 1.02)0.3990.49 (0.01, 0.98)0.047*1.07 (0.21, 1.93)0.021*−0.70 (−1.46, 0.07)0.064NANA0.41 (0.04, 0.79)0.037*NANA RA-SRe0.39 (−1.93, 2.72)0.694−1.11 (−2.15, −0.07)0.041*−2.11 (−3.21, −1.01)0.003*−0.71 (−1.72, −0.30)0.122NANA−0.49 (−1.08, 0.10)0.084NANA RA-SRa−0.14 (−0.91, 0.63)0.6850.33 (0.06, 0.61)0.026*0.45 (−0.16, 1.07)0.111*−0.53 (−0.98, −0.08)0.029*NANA−0.12 (−0.50, 0.27)0.495NANA*, P<0.05. CI, confidence interval; NA, not available; RA, right atrial; RA-SRa, right atrial peak late negative strain rates; RA-SRe, right atrial peak early negative strain rates; RA-SRs, right atrial peak positive strain rates; RA-εa, right atrial booster strain; RA-εe, right atrial conduit strain; RA-εs, right atrial reservoir strain; RV, right ventricular; RV-GLS, right ventricular global longitudinal strain; RV-PDSRL, right ventricular peak longitudinal diastolic strain rates; RV-PSSRL, right ventricular peak longitudinal systolic strain rates. Subgroup and meta-regression analyses for right ventricular and right atrial strain and strain rate *, P<0.05. CI, confidence interval; NA, not available; RA, right atrial; RA-SRa, right atrial peak late negative strain rates; RA-SRe, right atrial peak early negative strain rates; RA-SRs, right atrial peak positive strain rates; RA-εa, right atrial booster strain; RA-εe, right atrial conduit strain; RA-εs, right atrial reservoir strain; RV, right ventricular; RV-GLS, right ventricular global longitudinal strain; RV-PDSRL, right ventricular peak longitudinal diastolic strain rates; RV-PSSRL, right ventricular peak longitudinal systolic strain rates. Table 3Pooled values of right ventricular strain parametersSubgroupRV-GLSRV-PSSRLRV-PDSRLNMean (95% CI)I2 (%)NMean (95% CI)I2 (%)NMean (95% CI)I2 (%)Overall43−25.62 (−26.60, −24.63)98.57−1.50 (−1.64, −1.36)98.651.69 (1.33, 2.05)99.1Software EchoPAC26−26.30 (−27.34, −25.26)983−1.69 (−1.91, −1.48)92.841.69 (1.26, 2.12)99.3 TomTec4−26.93 (−28.20, −25.65)97.3NANANANANANA Q-lab6−23.13 (−25.08, −21.19)95.5NANANANANANA EPIQ1−30.02 (−31.56, −28.48)NANANANANANANA Toshiba1−17.90 (−19.26, −16.54)NANANANANANANA CMQNANANANANANANANANASTE vendor Philips10−24.50 (−26.21, −22.78)96.61−1.20 (−1.23, −1.17)NANANANA GE28−26.68 (−27.61, −25.79)97.76−1.60 (−1.81, −1.39)98.741.69 (1.26, 2.12)99.3 SiemensNANANA1−1.20 (−1.24, −1.16)NANANANA Toshiba2−22.36 (−31.08, −13.64)991−1.70 (−1.86, −1.54)NANANANA Esaote2−21.49 (−27.47, −15.52)96.9NANANA11.70 (1.56, 1.84)NAImaging mode 2D31−25.91 (−26.94, −24.89)98.67−1.50 (−1.64, −1.35)98.651.69 (1.33, 2.05)99.1 3D4−23.35 (−26.55, −20.14)95.7NANANANANANARace Asian7−25.55 (−26.68, −24.42)98.82−1.55 (−1.65, −1.45)91.541.79 (1.43, 2.16)98.1 Non-Asian36−25.82 (−27.16, −24.47)87.67−1.43 (−1.54, −1.32)95.611.30 (1.27, 1.33)NACI, confidence interval; I2, inconsistency index; N, number of studies; NA, not available; RV-GLS, right ventricular global longitudinal strain; RV-PDSRL, right ventricular peak longitudinal diastolic strain rates; RV-PSSRL, right ventricular peak longitudinal systolic strain rates. Pooled values of right ventricular strain parameters CI, confidence interval; I2, inconsistency index; N, number of studies; NA, not available; RV-GLS, right ventricular global longitudinal strain; RV-PDSRL, right ventricular peak longitudinal diastolic strain rates; RV-PSSRL, right ventricular peak longitudinal systolic strain rates. Pooled analysis of normal values for RA strain and strain rate parameters The pooled mean reference values for RA-εs, RA-εe, and RA-εa were 42.95% (95% CI: 40.69%, 45.21%), 20.86% (95% CI: 18.94%, 22.78%), and 19.30% (95% CI: 15.75%, 22.84%), respectively. The corresponding forest plots are shown in Figures 4-6. All three RA strain parameters demonstrated significant heterogeneity across studies. Egger’s test revealed no evidence of publication bias for any of the RA strain parameters (P=0.612, 0.420 and 0.091 for RA-εs, RA-εe, and RA-εa, respectively). Subgroup and meta‑regression analyses indicated that race, age, sex, and strain‑tracking software were significant contributors to heterogeneity in RA-εe and RA-εa (all P<0.05), while race, sex, and software were significant contributors to heterogeneity in RA-εs (all P<0.05). The forest plot of pooled normal value for RA-εs. εs, reservoir strain; CI, confidence interval; ES, effect size; RA, right atrial. The forest plot of pooled normal value for RA-εe. εe, conduit strain; CI, confidence interval; ES, effect size; RA, right atrial. The forest plot of pooled normal value for RA-εa. εa, booster strain; CI, confidence interval; ES, effect size; RA, right atrial. Regarding RA strain‑rate parameters, the pooled mean reference values for RA-SRs, RA-SRe, and RA-SRa were 2.03 s–1 (95% CI: 1.78, 2.28), −2.14 s–1 (95% CI: −2.55, −1.72), and −1.80 s–1 (95% CI: −2.04, −1.56), respectively. The forest plots are presented in Figures S3-S5. All three strain‑rate parameters also showed significant heterogeneity. Egger’s test revealed no significant publication bias for RA-SRs and RA-SRe (P=0.104 and 0.342, for RA-SRs and RA-SRe, respectively), whereas slight publication bias was detected for RA-SRa (P=0.04). Subgroup and meta-regression analyses identified race, sex, and STE vendor as significant contributors to heterogeneity in RA-SRs, while race and sex were associated with heterogeneity in RA-SRe, and race and age with heterogeneity in RA-SRa. Detailed results are summarized in Tables 2,4. Table 4Pooled values of right atrial strain parametersSubgroupRA-εsRA-εeRA-εaRA-SRsRA-SReRA-SRaNMean (95% CI)I2 (%)NMean (95% CI)I2 (%)NMean (95% CI)I2 (%)NMean (95% CI)I2 (%)NMean (95% CI)I2 (%)NMean (95% CI)I2 (%)Overall1842.95 (40.69, 45.21)97.71320.86 (18.93, 22.78)98.71119.30 (15.75, 22.84)99.592.03 (1.78, 2.28)97.48−2.14 (−2.55, −1.72)98.910−1.80 (−2.04, −1.56)97.6Software EchoPAC941.14 (39.13, 43.15)86723.41 (20.98, 25.84)90715.30 (13.58, 17.02)89.971.90 (1.64, 2.15)97.56−1.87 (−2.22, −1.51)98.510−1.74 (−2.00, −1.47)98.1 TomTec136.00 (34.41, 37.59)NA114.46 (13.51, 15.41)NA121.70 (20.72, 22.68)NANANANANANANANANANA Q-lab344.84 (36.16, 53.52)95.9227.77 (12.38, 43.15)98.4213.73 (4.42, 23.04)98.9NANANANANANANANANA EPIQNANANANANANANANANANANANANANANANANANA ToshibaNANANANANANANANANANANANANANANANANANA CMQ145.00 (41.85, 48.15)NA115.00 (13.39, 16.61)NANANANANANANANANANANANANASTE vendor Philips642.09 (37.22, 46.95)93.5522.02 (17.28, 26.77)96.7415.01 (9.80, 20.21)97.911.70 (1.53, 1.87)NA1−1.80 (−2.01, −1.59)NA2−2.04 (−3.18, −0.89)79 GE941.02 (38.66, 43.39)91.6722.21 (17.92, 26.50)97.8616.83 (13.87, 19.79)97.261.93 (1.64, 2.22)97.96−1.32 (−2.94, 0.30)98.95−1.77 (−2.06, −1.47)98.4 Siemens141.60 (38.85, 44.36)NANANANANANANA12.21 (1.99, 2.44)NA1−1.97 (−2.19, −1.75)NA1−2.18 (−2.38, −1.98)NA ToshibaNANANANANANANANANANANANANANANANANANA Esaote177.80 (67.12, 88.48)NANANANANANANA13.00 (2.59, 3.41)97.31−3.99 (−4.24, −3.74)NA1−1.57 (−1.87, −1.27)NAImaging mode 2D1542.32 (39.81, 44.83)93.01021.61 (18.28, 24.94)97.2816.52 (13.21, 19.84)97.992.03 (1.78, 2.28)97.48−2.14 (−2.55, −1.72)98.910−1.80 (−2.04, −1.56)97.6 3DNANANANANANANANANANANANANANANANANANARace Asian443.17 (35.25, 51.09)95.5327.01 (18.99, 35.03)95.8317.23 (15.08, 19.39)78.441.80 (1.55, 2.06)94.33−1.54 (−1.80, −1.28)93.34−1.83 (−2.34, −1.32)73.1 Non-Asian1242.17 (39.63, 44.71)92.6920.60 (17.36, 23.83)97.3715.59 (12.09, 19.09)98.452.19 (1.94, 2.43)91.35−2.49 (−3.03, −1.96)98.26−1.75 (−1.88, −1.62)98.5CI, confidence interval; I2, inconsistency index; N, number of studies; NA, not available; RA-SR, right atrial strain rate; RA-SRa, right atrial peak late negative strain rates; RA-SRe, right atrial peak early negative strain rates; RA-SRs, right atrial peak positive strain rates; RA-εa, right atrial booster strain; RA-εe, right atrial conduit strain; RA-εs, right atrial reservoir strain. Pooled values of right atrial strain parameters CI, confidence interval; I2, inconsistency index; N, number of studies; NA, not available; RA-SR, right atrial strain rate; RA-SRa, right atrial peak late negative strain rates; RA-SRe, right atrial peak early negative strain rates; RA-SRs, right atrial peak positive strain rates; RA-εa, right atrial booster strain; RA-εe, right atrial conduit strain; RA-εs, right atrial reservoir strain. Discussion The present meta-analysis integrated evidence from 53 studies that evaluated STE-derived RV and RA strain as well as strain-rate parameters in healthy adult populations under various conditions. Study identification, screening, and data extraction were conducted according to predefined methodological standards to ensure rigorous and comprehensive assessment of the available literature. The results indicated that demographic characteristics, particularly race, sex, and age, together with differences in FT software, were major sources of variation in right-heart deformation measurements. In contrast, the influence of STE vendors and imaging mode appeared to be comparatively modest. This meta-analysis provides several advantages over previous reports of pooled means for right heart strain parameters derived from STE. First, previous studies only evaluated either RV (22-24) or RA (84) strain individually, with no normal reference values established for strain rate. In contrast, our study analyzed a comprehensive range of right heart strain parameters, including RV-GLS, RA-εs, RA-εe, RA-εa, and their corresponding strain rates. Second, the sample size was substantially greater, including 43 studies for RV and 18 studies for RA evaluation, comprising 3,010 participants for RV assessment and 3,045 for RA evaluation. Moreover, the present analysis established pooled reference values for right heart strain and strain rates across multiple demographic and technical conditions, including race, sex, age, FT software, STE vendor, and imaging mode. Potential determinants contributing to interstudy variability were also explored, thereby providing useful reference data for future research and clinical interpretation. Race, sex, and age have long been recognized as important factors influencing left ventricular morphology and function (85-89). Although findings from the left heart cannot be directly extrapolated to the right heart (2), the present study revealed that most pooled reference values of RV and RA deformation measured by STE also demonstrated significant variation according to race, sex, and age. The significant racial differences were observed between Asian and non-Asian (European, American) populations in parameters such as RV-GLS, RA-εe, and RA-εa. Specifically, Asians exhibited higher values of RA-SRs and RA-SRa (coefficients >0), but lower values of RV-GLS, RV-PSSRL, RA-εe, RA-εa, and RA-SRe (coefficients <0); Europeans exhibited higher values of RA-εs, RA-εa, RA-SRs and RA-SRa (coefficients >0), but lower values of RV-GLS, RV-PSSRL, RA-εe, and RA-SRe (coefficients <0) than Americans. Such disparities may be attributable to inherent differences in RV and RA myocardial characteristics, variations in cardiovascular adaptation, environmental influences such as altitude, and lifestyle-related factors (90). Likewise, sex significantly influenced both ventricular (RV-GLS, RV-PSSRL) and atrial (RA-εe, RA-εa, RA-SRe, RA-SRs) strain parameters, consistent with prior reports (89,91,92). These differences may be partly attributable to sex differences in mitochondrial function, myocardial mass, and ventricular cavity size, potentially leading to differences in cardiac diastolic function (93-95). Moreover, participants aged >50 years exhibited lower values of RV-GLS, RV-PSSRL, RV-PDSRL, RA-εs, RA-εa, and RA-SRa (coefficients <0), but higher RA-εe value (coefficient >0). These age-related alterations may be associated with physiological changes including myocardial cell loss, progressive fibrosis, and impaired RV diastolic function (96). Taken together, race, sex, and age exert measurable effects on right heart strain parameters derived from STE. The findings of the present study also indicated that the choice of FT software has a substantial impact on the measurement of right-heart strain parameters. Among the included studies, feature-tracking analysis software mainly comprised EchoPAC, TomTec, QLAB, and others, with EchoPAC being used most frequently (32 of 53 studies). The choice of software significantly affected measurements of RV-GLS, RA-εe, and RA-εa, which is consistent with earlier findings (84). Such inter-software variation likely arises from differences in contour delineation algorithms, endocardial/epicardial tracking, and temporal smoothing techniques (28). Consequently, when measuring RV and RA strain through STE, it is essential to consider the choice of FT software and, where feasible, establish software-specific normative ranges to minimize heterogeneity attributable to analytical methods. Meanwhile, STE acquisition mode exerts a substantial impact on RV strain heterogeneity, consistent with previous findings (19). This discrepancy primarily arises from differences in the ability of each technique to capture the three-dimensional (3D) motion of the myocardium. 3D strain analysis acquires data from the entire left ventricular myocardial volume, enabling a more comprehensive and accurate assessment of global and regional myocardial function. In contrast, two-dimensional (2D) strain analysis is restricted to a single imaging plane and fails to fully account for the complex 3D myocardial motion, which increases the likelihood of measurement bias (14,97). These modality-related effects should be carefully considered when establishing future reference values. Additionally, STE vendor (e.g., Philips, GE, Siemens) significantly influenced certain parameters, including RV-GLS and RA-SRs. This underscores the need for further investigation into inter-vendor variability in right heart strain measurements, which will be crucial for standardization in both clinical and research settings. Given inherent technical variability, these pooled means should serve as directional references rather than rigid diagnostic thresholds. Clinically, their primary value lies in screening for major deviations and monitoring longitudinal intra-individual trends, where relative changes—particularly those assessed using consistent vendors/software—are more reliable than absolute values. Additionally, they provide a foundational baseline for research, provided the acoustic window is adequate. Several limitations should be acknowledged. First, inherent heterogeneity across the included studies may have influenced the pooled findings, particularly owing to variations in participant characteristics, reporting methods, and the possibility of publication bias. Second, because only a limited number of eligible studies were available, pooled reference ranges for RV global circumferential and radial strain rates could not be established, and further exploration of factors contributing to variability in RV peak longitudinal diastolic strain rate was not feasible. Third, the accuracy and reproducibility of STE-derived parameters are influenced by multiple technical, physiological, and anatomical factors that are not fully accounted for in current reference ranges. As a highly operator-dependent technique, STE is sensitive to image quality and acquisition settings, such as acoustic window and frame rate, which can substantially influence tracking performance and measurement precision (98,99); in addition, strain measurements are load dependent and may vary with changes in preload and afterload even in healthy individuals, while anatomical or extrinsic factors, including anterior chest wall deformities, may further alter myocardial deformation patterns and introduce measurement bias (100). Fourth, sex-specific reference values were not available for most right-heart strain parameters, which may limit the clinical applicability of the current pooled estimates. Finally, differences in right-heart deformation parameters between 2D- and 3D-STE remain insufficiently defined because only a small number of 3D systems were included, and most right-heart indices still lack separate 2D- and 3D-specific reference values. These limitations underscore the need for further high-quality, multicenter studies to refine normative standards for right-heart deformation assessed by STE. Conclusions This meta-analysis established pooled reference values for RV and RA strain and strain-rate parameters derived from STE in healthy adults. Race, sex, age, FT software and imaging mode were major contributors to heterogeneity in most right-heart strain parameters, whereas STE vendor had only limited influence. Overall, STE demonstrates strong potential as a practical and reliable technique for the quantitative assessment of right heart function in routine clinical practice. Supplementary The article’s supplementary files as