Reference values of left and right atrial volumes and phasic function based on a large sample of healthy Chinese adults: A cardiovascular magnetic resonance study.

BACKGROUND
The left and right atrial (LA and RA) size and function are tightly linked to the morbidity and mortality of multiple cardiovascular diseases. We aimed to establish cardiovascular magnetic resonance (CMR) reference values for LA and RA volumes and phasic function based on a large sample of healthy Chinese adults.


METHODS
408 validated healthy Chinese adults (54% men; aged 21-70 years) were included. LA and RA maximum, minimum, and pre-atrial contraction volumes (Vmax, Vmin, and Vpac); total, passive, and booster emptying fractions (EF total, EF passive, and EF booster); and total, passive, and active emptying volumes (TEV, PEV, and AEV) were measured on cine CMR. Normal reference values were calculated and were stratified by sex and age decades.


RESULTS
Men demonstrated greater LAVmax, LAVmin, LAVpac, LAPEV, RAVmax, RAVmin, RAVpac, RATEV, and RAAEV, while women had higher LAEF total, LAEF booster, RAEF total, RAEF passive, and RAEF booster (all p < 0.05). Age was positively correlated with LAVpac and RAVpac in both sexes but was positively correlated with LAVmax, LAVmin, RAVmax, and RAVmin only in women (all p < 0.05). For both sexes, aging was associated with decreased LAEF total, LAEF passive, RAEF total, and RAEF passive, but increased LAEF booster (all p < 0.05).


CONCLUSION
We systematically provide age- and sex-specific CMR reference values for LA and RA volumes and phasic function based on a large sample of healthy Chinese adults with a wide age range. Both age and sex are closely associated with biatrial volumes and function.


Introduction
The left and right atrium (LA and RA) play an integral role in modulating ventricular lling with its reservoir, conduit, and booster pump function [1][2][3].It has been established that atrial volumes and function are closely related to the morbidity and mortality of multiple cardiovascular diseases, such as atrial brillation recurrence [4,5], hypertrophic cardiomyopathy [6], heart failure [7], acute myocardial infarction [8], pulmonary hypertension [9], and congenital heart diseases [10].Compared with traditional echocardiography, cardiovascular magnetic resonance (CMR) is well recognized as the "gold standard" method for assessing atrial structure and function, with higher accuracy and reproducibility [11][12][13].
Normal reference values are essential to differentiate between normal and pathological cardiovascular conditions.Evidence is available that there are signi cant racial differences in atrial size and function [14,15], emphasizing the necessity of the establishment of ethnic-speci c normal values for atrial structure and function.Although a few studies have reported normal reference values of LA and RA volumes and function for Chinese adults [16,17], their provided parameters are less comprehensive and are based on limited sample size.In addition, age-related normal values have not been well provided in the Chinese population.Therefore, the present study aimed to systematically establish age-and sex-speci c CMR reference values for LA, RA volumes and phasic function based on a large sample of healthy Chinese adults free of hypertension, diabetes, and obesity, and further investigate their associations with age and sex.

Study population
Our study population came from the local medical imaging and health screening institution, which aimed at the early detection of various diseases through systemic examinations, with cardiovascular disease screening being one of the key items.The main screening items included (1) baseline characteristics: age, sex, height, weight, exercise intensity, occupation, personal and family history, etc. (2) physical examination and electrocardiogram.(3) laboratory and biochemical tests: routine blood, urine, stool tests, and fasting blood glucose, insulin, lipid tests, and sex hormone tests, and thyroid, liver, kidney function tests, etc. (4) imaging examination: CMR, echocardiography, chest X-ray lm or chest CT, magnetic resonance examination of brain, spine, and abdomen.All the examinations for individuals were completed within the same day and the diagnosis was performed by physicians with more than ten years' experience.The authors were authorized full access to all the subject information.
A total of 1,164 consecutive Chinese adults completed the above full-body screening package from January 1, 2013, to June 1, 2020, in the local health screening institution.From this pool of population, subjects who met the following exclusion criteria were excluded: (1) subjects with chief complaint or cardiovascular symptoms, abnormal electrocardiogram, abnormal ndings of hypertrophic cardiomyopathy (de ned as end-diastolic left ventricular (LV) wall thickness ≥ 15 mm in any left ventricular segment [18]), congenital heart diseases, valvular heart diseases (de ned as observed dephasing jet), cardiac tumor, or known diseases may affect cardiac morphology and function, such as anemia, hyperthyroidism, gout, and major brain, lung, liver, and kidney diseases detected by comprehensive health screening; (2) subjects with hypertension (de ned as blood pressure ≥ 140/90 mmHg or use of antihypertensive medication according to the seventh Joint National Committee recommendation [19]), diabetes (de ned as fasting blood glucose (FBG) ≥ 126 mg/dl or history of hypoglycemic medication), or obesity (de ned as body mass index (BMI) ≥ 28 kg/m 2 for the Chinese populations [20]); (3) subjects with unquali ed images affecting the assessment of atrial volumes, such as artifacts and incomplete or suboptimal presentation of LA contours in the 2-or 4-chamber views.The present study was approved by the local institutional review board and the requirement for subject consent was waived by the Chinese Ethics Committee (reference number: ChiECRCT20190198).

CMR protocol
CMR studies were performed on 1.5T magnetic resonance scanners (Signa HDxt, General Electric Medical Systems, Waukesha, WI, and Magnetom Essenza, Siemens, Erlangen, Germany) with 16-channel phasedarray surface coil.Balanced steady-state free precession (SSFP) with breath-hold were used to acquire images in the supine position, comprising a stack of contiguous parallel short-axis slices covering the whole left and right ventricle from base to apex.Moreover, three long-axis slice (two-, three-and fourchamber views) images were also acquired.All the individuals were examined under a sinus rhythm.

Volumetric analysis
Volumetric analysis was performed using commercial software (cvi42® version 5.12.1, Circle Cardiovascular Imaging, Canada).All the basic information about subjects (including age, sex, etc.) was hidden during image analysis.LA endocardial contours were manually traced in the 2-and 4-chamber views excluding pulmonary veins and the LA appendage.RA endocardial contours were tracked in the 4chamber view, only.LA volumes and RA volumes were calculated using the previously validated biplane area-length and the single plane area-length method, respectively [21].The maximum volume (Vmax), preatrial contraction volume (Vpac), and minimum volume (Vmin) for both LA and RA were assessed at ventricular end-systole, at ventricular diastole before atrial contraction, and at late ventricular diastole after atrial contraction, respectively (Fig. 1).

Statistical analysis
All statistical analyses were performed using SPSS (version 26.0,IBM SPSS Inc, Chicago, IL, USA) and GraphPad Prism (version 8.0.2,La Jolla, CA).Normally distributed continuous variables are presented as mean ± standard deviation (SD) and non-normally distributed variables are expressed as median (interquartile range).Sex differences in demographic characteristics, atrial volumes, and phasic function parameters were compared using Student's t-test, or Mann-Whitney-U test where appropriate.The subjects were strati ed by sex (men and women) and age decade (21-30, 31-40, 41-50, 51-60, 61-70 years).The associations of atrial volumes and phasic function parameters with age were investigated using simple linear regression analysis.Intra-and inter-observer reproducibility were assessed in 30 randomly selected subjects using intraclass correlation coe cients (ICC) and Bland-Altman analyses.Differences were regarded as statistically signi cant at p < 0.05.All p values were two-sided.

Demographic characteristics of subjects
In the total population of 1,164 subjects, 756 subjects were excluded, leaving 408 validated healthy Chinese adults (220 men, 44.4 ± 12.1 years).The baseline characteristics of the study population are shown in Table 1.Effect of gender on LA, RA volumes and phasic function The reference values of LA volumes and phasic function for both sexes are shown in Table 2. LA Vmax, LA Vmin, and LA Vpac in men were all greater than those in women (all p < 0.01).After indexing by BSA, LA Vmax and LA Vpac were greater for women (both p < 0.05), while no signi cant sex difference was observed in LA Vmin (p = 0.618).LAEF total and LAEF booster in women were higher than those in men (both p < 0.001), but LAEF passive showed no signi cant gender difference (p = 0.404).LATEV, LAPEV, and LAAEV were comparable between both sexes (p = 0.157, p = 0.055, and p = 0.861, respectively), while women had a signi cantly higher LAEI (p < 0.001).The normal parameters of RA volumes and phasic function for men and women are presented in Table 2.
The absolute and indexed RA Vmax, RA Vmin, and RA Vpac were all greater in men (all p < 0.05), while RAEFs (including RAEF total, RAEF passive, and RAEF booster) were higher in women (all p ≤ 0.001).Men had signi cantly greater RATEV and RAAEV compared with women, but a lower RAEI (all p < 0.01).There was no signi cant sex difference in RAPEV (p = 0.462).
Age-related differences in atrial volumes and phasic function LA, RA volumes and phasic function parameters by age decades for both sexes are shown in Tables 3  and 4. Simple linear regression analysis showed that age was positively correlated with the absolute and indexed LA Vmin, LA Vpac, RA Vmax, RA Vmin, and RA Vpac (all p < 0.05), while there was no signi cant correlation between the absolute and indexed LA Vmax and age (p = 0.463 and p = 0.165).There was a signi cantly negative correlation between age and LA and RA reservoir function (r = -0.19,r = -0.19,r = -0.16,and r = -0.17for LAEF total, LAEI, RAEF total, and RAEI, respectively; all p < 0.001), except for LATEV and RATEV (p = 0.335 and p = 0.747).Moreover, age was inversely related to LA and RA conduit function (r = -0.57,r = -0.40,r = -0.42,and r = -0.28 for LAEF passive, LAPEV, RAEF passive, and RAPEV, respectively; all p ≤ 0.001), while was positively correlated with biatrial booster function (r = 0.34, r = 0.43, r = 0.14, and r = 0.28 for LAEF booster, LAAEV, RAEF booster, and RAAEV, respectively; all p < 0.01) (Table 5).

Intra-observer and inter-observer reproducibility
Intra-observer and reproducibility are presented in Table S1.and as Bland-Altman plots in Fig. 2 and Fig. 3.There were excellent intra-and inter-observer reproducibility with ICC of more than 0.85 in both LA and RA volumes.

Discussion
In Previous multi-ethnic studies have con rmed that there are signi cant racial differences in atrial volumes and function, as described by the smaller atrial sizes and higher atrial function parameters for Asians compared with those for whites [14,15].Compared with several studies in western populations [11,13,21,25], LA and RA volumetric parameters (including Vmax, Vmin, and Vpac) in the present study were smaller, while the phasic function parameters (including EF, EV, and EI) were slightly higher, which may con rm the racial difference in arial volumes and function as well.

Gender differences LA and RA volumes phasic function
In general, men presented with higher absolute LA Vmax, LA Vmin, and LA Vpac than women, which was consistent with a few previous studies [26,27].Interestingly, we found that BSA-indexed LA Vmax and LA Vpac in women were signi cantly greater than those in men.Besides, we con rmed that LA phasic function was also affected by sex.Speci cally, women showed higher LAEF total and LAEF booster than men, which was in accordance with the ndings of Vasconcellos et al. [14].In contrast, some other studies, such as those by Maceira et al. [13] and Truong et al. [26], did not observe these sex differences.Currently, there are still few studies on gender difference in LAEV and LAEI.Maceira et al. [13] reported that women presented with signi cantly smaller LAPEV and higher LAEI than men.Different from their study, we demonstrated that there was no signi cant sex difference in LATEV, LAPEV, and LAAEV, while women had a signi cantly higher LAEI.
For the right atrium, in line with several previous studies [15,28], we demonstrated that the absolute and indexed RA Vmax, RA Vmin, and RA Vpac for men were all signi cantly greater than those for women.
However, there are still considerable controversies about gender differences in RAEFs.For example, several studies showed that RAEF booster was not associated with gender [13,15,16].In comparison, our study demonstrated that female sex was associated with greater RAEFs including RAEF total, RAEF passive, and RAEF booster, which was consistent with the ndings from Peluso et al. [28].We that the differences in sample size, ethnicity, and baseline characteristics of subjects may together contribute to these inconsistencies among studies.Furthermore, we con rmed that men had signi cantly greater RATEV and RAAEV and a lower RAEI compared with women, while there was no signi cant sex difference in RAPEV, which was generally agree with the study by Maceira et al. [13].

Correlation between atrial volumes, phasic function, and age
It is controversial regarding the correlation between atrial volumes and age.For example, Petersen et al. [21] 21,27].In contrast, we demonstrated that age was positively correlated with all the RA volumetric parameters including RA Vmax, RA Vmin, and RA Vpac, even after normalized by BSA.These differences may be related to the narrow age range as well as the limited sample size in previous studies.
Our study demonstrated that age was a major determinant of biatrial phasic function, as found that the measurements of LA and RA reservoir and conduit function were decreased, whereas booter function was increased with advancing age.These ndings further reinforced the previously described associations between age and atrial function [13,15].The observed age-related variations in atrial function were in accordance with physiologic knowledge.Normal aging is associated with loss of cardiomyocytes, myocardial brosis, and impaired ventricular relaxation, which may together contribute to the increased atrial booster function, decreased atrial reservoir and conduit function [30][31][32].

Limitations
were several limitations in the present study.Firstly, normal reference values of elderly over 70 years old were not covered due to the limited healthy elderly population.In addition, this study is a singleethnic, cross-sectional study, and the relationship between age, sex, and atrial parameters needs to be further veri ed by multi-ethnic, longitudinal studies.

Figure 1
Figures

Table 1
Baseline characteristics of all the subjects and strati ed according to gender p-values were obtained from the student' t text, or Wilcoxon rank sum test.BMI, body mass index; BSA, body surface area; SBP, systolic blood pressure; DBP, diastolic blood pressure; FBG, fasting blood glucose; TC, total cholesterol; TG, triglyceride; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol.

Table 2
LA and RA volumes and phasic function parameters for all the subjects and strati ed according to gender Data are presented as means ± SD. p-values were obtained from the student' t text.LA, left atrial; RA, right atrial; Vmax, maximum volume; Vmin, minimum volume; Vpac, pre-atrial contraction volume; BSA, body surface area; EF, emptying fraction; TEV, total emptying volume; PEV, passive emptying volume; AEV, active emptying volume; EI, expansion index.Data are presented as means ± SD. p-values were obtained from the student' t text.LA, left atrial; RA, right atrial; Vmax, maximum volume; Vmin, minimum volume; Vpac, pre-atrial contraction volume; BSA, body surface area; EF, emptying fraction; TEV, total emptying volume; PEV, passive emptying volume; AEV, active emptying volume; EI, expansion index.

Table 5
Correlation between age and atrial parameters for all the subjects [23,24]R study, we systematically established age-and sex-speci c reference values for LA, RA volumes and phasic function based on a large sample of validated healthy Chinese adults free of hypertension, diabetes, and obesity.Moreover, we further con rmed that both sex and age had a considerable impact on atrial volumes and phasic function.Comparison with previous studiesAt present, although there are a few CMR studies on normal biatrial reference values for Chinese adults, this study still presents certain advantages by providing more comprehensive atrial volume and function parameters based on a signi cantly larger sample size.Speci cally, compared with the present study, a study on 200 healthy Chinese volunteers by Zhuang et al.[17]reported comparable normal values of LA Vmax, Vmin, and RA Vmax, but signi cantly greater RA Vmin (29.7 ± 10.9 mL versus 53.52 ± 16.12 mL).However, they did not provide atrial phasic function parameters including EF, EV, and EI.In addition, another CMR study byLi et al. [16]reported similar normal reference values of biatrial Vmax, Vmin, Vpac, EF total, EF passive, and EF booster based on 135 healthy Chinese adults.However, their provided parameters were not further strati ed by age.Finally, as de nitive indicators of adverse atrial remodeling in the clinical scenario such as acute coronary syndrome[23,24], information regarding normal values of atrial EV and EI has scarcely been reported.To the best of our knowledge, this is the rst CMR study to establish age-and sex-speci c normal values of atrial EV and EI in Chinese adults.
[29]ed that aging was associated with decreased LA Vmax based on 804 Caucasian adults aged 45 to 74 years.A CMR study byLi et al. [16]in 135 healthy Chinese volunteers aged 49.9 ± 17.1 years showed that age was positively correlated with LA Vmax, LA Vmin, and LA Vpac.In line with the study by Van Grootel et al.[29], we found that age showed no signi cant correlation with LA Vmax, while was associated with increased LA Vmin and LA Vpac.Besides, with respect to the right atrium, most previous studies reported aging was associated with no change in or decreased RA Vmax [16,