Abstract
Background: Age-related changes in left ventricular (LV) structure and function lower the threshold for the onset of heart failure with preserved ejection fraction (HFpEF). LV parameters change also with race; however, the racial differences in age-related changes in LV parameters with and without adjustment for body mass index (BMI), heart rate (HR), and blood pressure (BP) remain unclear.
Methods and Results: We performed a subanalysis of the World Alliance Society of Echocardiography Normal Values Study, an international cross-sectional study that examined normal echocardiographic values in 15 countries. The age-related changes in 2-dimensional echocardiographic derived parameters including LV size, systolic function, and mass, were compared between healthy Japanese (n=227) and healthy White (n=98) and Black (n=69) American participants. In men, age-related changes in all parameters did not differ significantly among races. However, compared with Japanese women, White American women had a smaller body surface area (BSA)-indexed LV volume, BSA-indexed LV internal dimension at end-systole, BSA-indexed LV stroke volume, and LV mass index to BSA, and a larger LV ejection fraction with age, even after adjusting for BMI, HR, and BP.
Conclusions: Age-related changes in LV structure and function, which are important for the pathophysiology of HFpEF, may differ by race. Therefore, future studies examining echocardiographic reference values for each age group in each race are needed.
With the aging population, the prevalence of heart failure (HF) with preserved ejection fraction (HFpEF) is increasing in many countries. Furthermore, HFpEF is more difficult to diagnose than HF with reduced ejection fraction, and treatment options for HFpEF are more limited, so there is a need to elucidate its complex and heterogeneous pathophysiology.1 Recently, the age-related changes in left ventricular (LV) structure and function have been suggested to lower the threshold for the onset of HFpEF in response to preload and afterload, such as weight gain and hypertension.2 An almost consistent finding of studies using echocardiography3–10 or cardiac magnetic resonance (CMR)11–14 in healthy adult participants is that LV volume and stroke volume decrease with age, associated with an increase in relative wall thickness (RWT) and LV ejection fraction (LVEF). There is also consistent evidence from small studies using catheters in healthy subjects or non-ischemic, non-hypertensive patients with normal LVEF that mean left atrial pressure does not change with healthy aging, whereas LV end-diastolic pressure increases with age within the normal range.15–18
LV structure and function differ not only in terms of age, but also with race. The World Alliance Societies of Echocardiography (WASE) Normal Values Study, a multicenter international observational prospective cross-sectional echocardiographic study to determine reference ranges for heart size and function from 15 countries that analyzed all images at core laboratories, revealed that White and Black people had larger LV dimensions, volumes, and stroke volumes than Asian and Mexican people.10,19 The reference ranges of the LV structure and function for the subjects overall, including all age groups, did not differ between White and Black people,10,19 but whether there are racial differences in age-related changes in these LV parameters remains unclear. In addition to age and race, body mass index (BMI)20 and vital signs, such as heart rate (HR)21 and blood pressure (BP),22 can influence LV structure and function. Moreover, the National Health and Nutrition Survey Japan 2019 conducted by the Ministry of Health, Labor and Welfare showed that the average BMI and BP of the Japanese population vary with age.23 Although the participants in the WASE study are healthy individuals, there are racial differences in age-related changes in factors such as BMI, HR, and BP, which affect LV size and function. Therefore, it is necessary to examine racial differences in age-related changes in LV structure and function after adjusting for these factors. The aim of the present study was to examine differences in age-related changes in each LV parameter on 2-dimensional (2D) echocardiography between Japanese participants and White and Black American participants, with and without adjustment for BMI, HR, and BP, in the WASE study.
Methods
Study Design and Population
This study was a subanalysis of the WASE study. The design of the WASE study has been described in detail elsewhere.24 Briefly, healthy individuals without histories of cardiac, lung, or kidney diseases were enrolled at each of 19 centers in 15 countries between September 2016 and January 2019. Of the 2,262 participants enrolled in the WASE study, 230, 98, and 74 were Japanese, White American, and Black American, respectively. In the present study, 2 Japanese men, 1 Japanese woman, and 1 Black American man with inadequate image quality or incomplete assessment of echocardiographic examinations, and 3 Black American men and 1 Black American woman who lacked information on demographic characteristics were excluded. Thus, the final study population comprised 227 Japanese, 98 White American, and 69 Black American individuals. Demographic information, including age, sex, systolic BP (SBP), diastolic BP (DBP), height, weight, BMI, body surface area (BSA), and self-reported race, was also collected. The study protocol adhered to the Declaration of Helsinki was approved by the local ethics committees; informed consent was obtained from all participants.
Echocardiographic Image Acquisition and Analysis Protocol
Comprehensive 2D transthoracic echocardiography was performed using ultrasound machines available at the enrollment sites (GE, Philips, and Siemens). Image acquisition was based on the WASE study-specific standard acquisition protocol. All images were stored in the Digital Imaging and Communications in Medicine format and were transferred to the core laboratory (MedStar Health Research Institute, Washington, DC, USA) for analysis. All measurements were performed using vendor-independent software (Image Arena; TomTec, Unterschleissheim, Germany) based on the American Society of Echocardiography/European Association of Cardiovascular Imaging guidelines.25
This study focused on 2D echocardiographic parameters: LV internal dimensions at end-diastole and end-systole (LVIDd and LVIDs, respectively) and LV wall dimensions (interventricular septal dimension in diastole [IVSd] and LV posterior wall dimension in diastole [LVPWd]); LV end-diastolic and end-systolic volumes (LVEDV and LVESV, respectively); LVEF; LV stroke volume (LVSV; calculated as LVEDV − LVESV); LV mass calculated by the cube formula; and RWT (calculated as [2 × LVPWd] / LVIDd). All LV internal dimensions, LV volumes, LVSV, and LV mass were indexed to BSA (indexed LVIDd, indexed LVIDs, LVEDVI, LVESVI, LVSVI, and LV mass index [LVMI], respectively).
Statistical Analysis
Participants were stratified according to sex and then further stratified into 3 different racial groups. Continuous variables are expressed as the mean±SD. The Kolmogorov-Smirnov test was used to assess the normality of distribution for each continuous variable. Demographic characteristics and echocardiographic parameters were compared among races using one-way analysis of variance (ANOVA) or the Kruskal-Wallis analysis of variance, if assumptions of normality of distribution were not validated. Post hoc multiple comparisons after one-way ANOVA were performed using Tukey’s procedure or the Games-Howell test if homogeneity of variance was not verified. Pairwise comparisons using the Mann-Whitney U test were performed for any dependent variables for which the Kruskal-Wallis test was significant. Pearson’s product moment correlation coefficient (r) or Spearman’s rank correlation coefficient (ρ) if the assumption of normality of distribution was not validated was used to examine the association between age and baseline characteristics in men and women between each race. Then, to examine racial differences in linear associations between age and each LV parameter (indexed LVIDd, indexed LVIDs, LVEDVI, LVESVI, LVSVI, LVEF, LVMI, and RWT), the interaction was assessed via multiple regression analyses using a linear regression model including the following interaction terms:
LV parameter
= α + β1 × age + β2 × USAWT + β3 × USABK + β4 × age × USAWT + β5 × age × USABK
=(β1 + β4 × USAWT + β5 × USABK) × age + (α + β2 × USAWT + β3 × USABK)
where USAWT
and USABK
are the indicator variables that take ‘1’ if the participant is a White American participant and a Black American participant, respectively, and otherwise that take ‘0’. In these equations, β1, β1+β4, and β1+β5 correspond to increases in LV parameters per unit (years) increase in age in Japanese, White American, and Black American participants, respectively. We determined the interaction effect of White American vs. Japanese or Black American vs. Japanese on estimates of age-related changes in LV parameters based on the estimates of β4 or β5. After multiple regression analyses, the quantile-quantile plots and the Kolmogorov-Smirnov test were used to assess the normal distribution of residuals and the differences between the observed and predicted values. Finally, to examine the racial differences in linear associations between age and each LV parameter, after adjusting for BMI, HR, SBP and DBP, the interaction was assessed via multiple regression analysis using a linear regression model including the following interaction terms:
LV parameter
= α + β1 × age + β2 × USAWT + β3 × USABK + β4 × age × USAWT + β5 × age × USABK + β6 × BMI + β7 × HR + β8 × SBP + β9 × DBP
=(β1 + β4 × USAWT + β5 × USABK) × age + (α + β2 × USAWT + β3 × USABK + β6 × BMI + β7 × HR + β8 × SBP + β9 × DBP)
In these equations, β1, β1+β4, and β1+β5 correspond to increased LV parameters per unit (years) increase in age in Japanese, White American, and Black American participants, respectively, after adjusting for BMI, HR, SBP and DBP. We determined the interaction effect of age-related changes in LV parameters based on the estimates of β4 or β5 in White American vs. Japanese or Black American vs. Japanese. After multiple regression analyses, the quantile-quantile plots and the Kolmogorov-Smirnov test were used to assess the normal distribution of residuals and the differences between the observed and predicted values.
Data were analyzed using the SPSS version 29.0.1.0 (IBM Corp., Armonk, NY, USA). A two-tailed P value of less than 0.05 was considered statistically significant. The validity of these statistical analyses was reviewed using a statistical expert (T.K.).
Results
Comparisons of Demographic Characteristics Among Races
Demographic characteristics of the Japanese, White American, and Black American participants are presented in Table 1. There were no significant differences in terms of age and HR among the 3 racial groups in both sexes. However, the BMI, BSA, and SBP differed significantly among the 3 races in both sexes, and DBP differed among the 3 races in men. Further analysis with the post hoc test revealed that the BSA and BMI of Japanese participants were significantly lower than those of White and Black American participants for both sexes. Japanese men and women had a significantly higher SBP than White American men and women. Black American men also had a significantly higher SBP than White American men. Japanese men had a significantly higher DBP than White and Black American men.
Table 1.
Demographic Characteristics and Echocardiographic Parameters of the LV in Men and Women According to Race
| |
Men |
Women |
| JPN |
USAWT |
USABK |
P value |
JPN |
USAWT |
USABK |
P value |
| No. participants |
118 |
51 |
33 |
|
109 |
47 |
36 |
|
| Demographic characteristics |
| Age (years) |
|
|
|
0.092 |
|
|
|
0.333 |
| Mean±SD |
47±17 |
49±19 |
41±15 |
|
46±18 |
49±18 |
42±13 |
|
| Range |
22–82 |
20–86 |
20–68 |
|
21–86 |
21–87 |
24–73 |
|
| Heart rate (beats/min) |
62±9 |
62±9 |
67±12 |
0.055 |
66±10 |
65±9 |
71±13 |
0.093 |
| SBP (mmHg) |
128±11 |
121±14 |
126±11 |
<0.001*,†,§ |
124±12 |
118±13 |
122±11 |
0.022*,† |
| DBP (mmHg) |
75±8 |
72±8 |
71±9 |
0.004*,†,‡ |
73±10 |
72±8 |
72±9 |
0.773 |
| BSA (m2) |
1.76±0.13 |
2.04±0.20 |
2.04±0.24 |
<0.001*,†,‡ |
1.51±0.12 |
1.76±0.15 |
1.84±0.20 |
<0.001*,†,‡,§ |
| BMI (kg/m2) |
22.5±2.4 |
25.6±3.2 |
26.9±5.7 |
<0.001*,†,‡ |
21.4±2.7 |
24.6±3.3 |
28.2±5.8 |
<0.001*,†,‡,§ |
| LV parameters |
| IVSd (mm) |
8.2±1.4 |
8.3±1.4 |
8.4±1.5 |
0.855 |
7.0±1.5 |
7.4±1.4 |
7.6±1.5 |
0.068 |
| LVPWd (mm) |
8.7±1.2 |
8.7±1.1 |
9.4±1.1 |
0.011*,‡,§ |
7.7±1.4 |
7.8±1.0 |
8.0±1.1 |
0.571 |
| LVIDd (mm) |
45.1±5.1 |
48.0±5.3 |
47.2±4.4 |
0.001*,†,‡ |
40.8±3.9 |
43.5±4.4 |
42.6±4.2 |
0.001*,† |
| LVIDs (mm) |
28.9±3.8 |
31.1±4.1 |
30.5±3.5 |
0.001*,† |
25.7±2.6 |
28.1±3.8 |
27.1±3.3 |
<0.001*,† |
| Indexed LVIDd (mm/m2) |
25.6±2.5 |
23.7±2.8 |
23.5±3.6 |
<0.001*,†,‡ |
27.0±2.4 |
24.9±2.7 |
23.4±2.9 |
<0.001*,†,‡,§ |
| Indexed LVIDs (mm/m2) |
16.4±2.0 |
15.3±2.2 |
15.2±2.5 |
0.001*,†,‡ |
17.0±1.7 |
16.1±2.4 |
14.9±2.1 |
<0.001*,‡,§ |
| LVEDV (mL) |
99.2±19.9 |
125.9±29.0 |
120.2±22.6 |
<0.001*,†,‡ |
76.8±14.3 |
94.0±23.6 |
90.1±15.4 |
<0.001*,†,‡ |
| LVESV (mL) |
37.0±9.3 |
46.5±11.0 |
45.0±9.4 |
<0.001*,†,‡ |
27.3±5.5 |
34.0±9.8 |
33.2±6.1 |
<0.001*,†,‡ |
| LVEDVI (mL/m2) |
56.3±10.2 |
61.4±11.7 |
59.8±12.9 |
0.015*,† |
50.7±8.4 |
53.7±13.7 |
49.3±8.5 |
0.109 |
| LVESVI (mL/m2) |
21.0±4.9 |
22.7±4.5 |
22.4±5.2 |
0.040*,† |
18.0±3.3 |
19.4±5.6 |
18.2±3.4 |
0.131 |
| LVSV (mL) |
62.1±11.5 |
79.4±19.2 |
75.3±13.9 |
<0.001*,†,‡ |
49.5±9.4 |
60.0±14.5 |
56.9±9.9 |
<0.001*,†,‡ |
| LVSVI (mL/m2) |
35.3±5.8 |
38.7±7.7 |
37.4±8.0 |
0.025*,† |
32.7±5.6 |
34.3±8.4 |
31.2±5.4 |
0.088 |
| LVEF (%) |
62.9±3.0 |
63.1±2.8 |
62.7±2.5 |
0.740 |
64.4±2.6 |
64.1±2.9 |
63.2±2.3 |
0.059 |
| LV mass (g) |
123.8±30.3 |
138.5±32.6 |
141.7±26.1 |
0.001*,†,‡ |
88.0±23.1 |
101.4±21.2 |
101.7±25.6 |
<0.001*,†,‡ |
| LVMI (g/m2) |
70.2±15.7 |
67.5±13.6 |
69.8±11.2 |
0.539 |
57.9±13.9 |
57.6±10.3 |
55.2±11.4 |
0.515 |
| RWT |
0.39±0.08 |
0.37±0.07 |
0.40±0.07 |
0.060 |
0.38±0.08 |
0.36±0.07 |
0.38±0.06 |
0.451 |
Unless indicated otherwise, data are presented as the mean±SD. *P<0.05 among races; †P<0.05 between Japanese (JPN) and White American (USAWT) participants; ‡P<0.05) between JPN and Black American (USABK) participants; §P<0.05 between USAWT and USABK participants. BMI, body mass index; BSA, body surface area; DBP, diastolic blood pressure; indexed LVIDd, BSA-indexed left ventricular (LV) internal dimension at end-diastole; indexed LVIDs, BSA-indexed LV internal dimension at end-systole; IVSd, interventricular septal dimension in diastole; LVEDV, LV end-diastolic volume; LVEDVI, BSA-indexed LVEDV; LVEF, LV ejection fraction; LVESV, LV end-systolic volume; LVESVI, BSA-indexed LVESV; LVIDd, LV internal dimension at end-diastole; LVIDs, LV internal dimension at end-systole; LVMI, LV mass index; LVPWd, LV posterior wall dimension in diastole; LVSV, LV stroke volume; LVSVI, BSA-indexed LVSV; RWT, relative wall thickness; SBP, systolic blood pressure.
Comparison of Echocardiographic LV Parameters Among Races
Table 1 presents echocardiographic LV parameters among races. Japanese participants had significantly smaller LV internal dimensions (LVIDd, LVIDs) than White American participants for both sexes. The LVIDd of Japanese men was also smaller than that of Black American men. The Japanese participants had a small BSA. Thus, Japanese men had significantly larger BSA-indexed LV internal dimensions (indexed LVIDd, indexed LVIDs) than White and Black American men. Japanese women had the highest indexed LVIDd, followed by White and Black American women. Japanese and White American women had a larger indexed LVIDs than Black American women. Japanese men and women had lower LV volumes (LVEDV, LVESV) than White American men and women. Japanese men had significantly lower LV volumes indexed to BSA (LVEDVI, LVESVI) than White American men. In contrast, the LV volumes indexed to BSA did not differ in women between races. Japanese participants had a significantly lower LVSV than Black and White American participants in both sexes. Japanese men had a significantly lower LVSVI than White American men. However, there were no racial differences in women. Further, LVEF did not differ significantly across the races in both sexes. The LV mass of White and Black American participants was significantly larger than that of Japanese participants in both sexes. However, there was no significant difference in LVMI among the races in both sexes. RWT did not differ significantly among the races in both sexes.
Association Between Age and Demographic Characteristics
Table 2 shows the results of the correlation analysis between age and demographic characteristics for men and women of each race. HR was not significantly correlated with age in both sexes of all races. SBP increased significantly with age in both sexes of all races, except for Black American men. DBP increased significantly with age in White American women and Japanese men and women. BSA was not significantly correlated with age in both sexes of all races, except in Japanese men, whose BSA was significantly negatively correlated with age. BMI increased significantly with age in White American and Japanese women.
Table 2.
Associations Between Age and Demographic Characteristics in Men and Women of Each Race
| |
Men |
Women |
| JPN |
USAWT |
USABK |
JPN |
USAWT |
USABK |
| r or ρ |
P value |
r or ρ |
P value |
r or ρ |
P value |
r or ρ |
P value |
r or ρ |
P value |
r or ρ |
P value |
| Heart rate (beats/min) |
−0.160 |
0.083 |
−0.095 |
0.507 |
0.173 |
0.335 |
0.086 |
0.374 |
−0.128 |
0.391 |
0.323 |
0.055 |
| SBP (mmHg) |
0.252 |
0.006* |
0.340 |
0.015* |
−0.136 |
0.452 |
0.606 |
<0.001* |
0.336 |
0.021* |
0.348 |
0.038* |
| DBP (mmHg) |
0.183 |
0.047* |
−0.001 |
0.997 |
0.219 |
0.220 |
0.310 |
0.001* |
0.397 |
0.006* |
−0.204 |
0.233 |
| BSA (m2) |
−0.188 |
0.041* |
0.058 |
0.686 |
−0.026 |
0.888 |
0.076 |
0.435 |
0.180 |
0.227 |
0.103 |
0.549 |
| BMI (kg/m2) |
−0.001 |
0.995 |
0.119 |
0.406 |
0.099 |
0.582 |
0.423 |
<0.001* |
0.432 |
0.002* |
−0.047 |
0.786 |
*P<0.05 between age and demographic characteristics. ρ, Spearman’s rank correlation coefficient if the assumption of normality of distribution was not validated; JPN, Japanese participants; r, Pearson’s product rate correlation coefficient; USABK, Black American participants; USAWT, White American participants. Other abbreviations as in Table 1.
Racial Differences in Associations Between Age and LV Parameters
Table 3 shows the slopes of each LV parameter plotted against age in each race, including the interaction effect of White American participants vs. Japanese participants and the interaction effect of Black American participants vs. Japanese participants. Supplementary Table 1 shows the results of a multiple regression analysis including an interaction term to examine whether the linear regression slopes of various echocardiographic parameters with respect to age differ among races. Figures 1 and 2 show the linear regression equations between the various LV parameters and age in each race. Table 4, Supplementary Table 2, and Figures 3 and 4 show the linear regression model adjusted for BMI, HR, SBP, and DBP.
Table 3.
Linear Regression Estimates of Change in Each Echocardiographic Parameter of the LV Based on Age Unit (Years) Grouped According to Race for Men and Women
| |
Men |
Women |
| Indexed LVIDd (mm/m2) |
Indexed LVIDs (mm/m2) |
Indexed LVIDd (mm/m2) |
Indexed LVIDs (mm/m2) |
| Estimate |
SE |
P value |
Estimate |
SE |
P value |
Estimate |
SE |
P value |
Estimate |
SE |
P value |
| JPN |
−0.043 |
0.014 |
0.003 |
−0.031 |
0.011 |
0.005 |
−0.040 |
0.013 |
0.002 |
−0.031 |
0.009 |
0.001 |
| USAWT |
−0.057 |
0.020 |
0.005 |
−0.047 |
0.016 |
0.003 |
−0.062 |
0.020 |
0.002 |
−0.076 |
0.014 |
<0.001 |
| USABK |
−0.034 |
0.032 |
0.289 |
−0.016 |
0.025 |
0.520 |
−0.065 |
0.032 |
0.045 |
−0.028 |
0.024 |
0.241 |
| Interaction USAWT vs. JPN |
−0.014 |
0.025 |
0.577 |
−0.016 |
0.019 |
0.405 |
−0.022 |
0.023 |
0.354 |
−0.046 |
0.017 |
0.008* |
| Interaction USABK vs. JPN |
0.009 |
0.035 |
0.791 |
0.015 |
0.027 |
0.572 |
−0.025 |
0.035 |
0.467 |
0.003 |
0.025 |
0.901 |
| |
LVEDVI (mL/m2) |
LVESVI (mL/m2) |
LVEDVI (mL/m2) |
LVESVI (mL/m2) |
| Estimate |
SE |
P value |
Estimate |
SE |
P value |
Estimate |
SE |
P value |
Estimate |
SE |
P value |
| JPN |
−0.263 |
0.053 |
<0.001 |
−0.120 |
0.023 |
<0.001 |
−0.190 |
0.044 |
<0.001 |
−0.076 |
0.017 |
<0.001 |
| USAWT |
−0.332 |
0.074 |
<0.001 |
−0.133 |
0.033 |
<0.001 |
−0.509 |
0.068 |
<0.001 |
−0.221 |
0.027 |
<0.001 |
| USABK |
−0.327 |
0.118 |
0.006 |
−0.132 |
0.052 |
0.012 |
−0.199 |
0.111 |
0.076 |
−0.081 |
0.044 |
0.064 |
| Interaction USAWT vs. JPN |
−0.068 |
0.091 |
0.451 |
−0.013 |
0.040 |
0.747 |
−0.318 |
0.081 |
<0.001* |
−0.145 |
0.032 |
<0.001* |
| Interaction USABK vs. JPN |
−0.063 |
0.129 |
0.623 |
−0.012 |
0.057 |
0.832 |
−0.009 |
0.120 |
0.941 |
−0.006 |
0.047 |
0.907 |
| |
LVSVI (mL/m2) |
LVEF (%) |
LVSVI (mL/m2) |
LVEF (%) |
| Estimate |
SE |
P value |
Estimate |
SE |
P value |
Estimate |
SE |
P value |
Estimate |
SE |
P value |
| JPN |
−0.144 |
0.032 |
<0.001 |
0.044 |
0.015 |
0.005 |
−0.114 |
0.029 |
<0.001 |
0.017 |
0.013 |
0.205 |
| USAWT |
−0.199 |
0.046 |
<0.001 |
0.018 |
0.021 |
0.409 |
−0.287 |
0.045 |
<0.001 |
0.073 |
0.020 |
<0.001 |
| USABK |
−0.195 |
0.073 |
0.008 |
0.014 |
0.034 |
0.682 |
−0.117 |
0.074 |
0.113 |
0.005 |
0.034 |
0.875 |
| Interaction USAWT vs. JPN |
−0.055 |
0.056 |
0.323 |
−0.026 |
0.026 |
0.324 |
−0.173 |
0.054 |
0.001* |
0.057 |
0.024 |
0.022* |
| Interaction USABK vs. JPN |
−0.051 |
0.080 |
0.520 |
−0.030 |
0.037 |
0.427 |
−0.003 |
0.079 |
0.966 |
−0.012 |
0.036 |
0.749 |
| |
LVMI (g/m2) |
RWT |
LVMI (g/m2) |
RWT |
| Estimate |
SE |
P value |
Estimate |
SE |
P value |
Estimate |
SE |
P value |
Estimate |
SE |
P value |
| JPN |
0.100 |
0.078 |
0.200 |
0.002 |
0.000 |
<0.001 |
0.361 |
0.061 |
<0.001 |
0.003 |
0.000 |
<0.001 |
| USAWT |
−0.105 |
0.109 |
0.339 |
0.002 |
0.000 |
<0.001 |
0.118 |
0.094 |
0.208 |
0.002 |
0.001 |
0.001 |
| USABK |
−0.102 |
0.174 |
0.557 |
0.001 |
0.001 |
0.385 |
0.182 |
0.154 |
0.240 |
0.002 |
0.001 |
0.007 |
| Interaction USAWT vs. JPN |
−0.205 |
0.134 |
0.128 |
−0.001 |
0.001 |
0.393 |
−0.242 |
0.112 |
0.031* |
−0.001 |
0.001 |
0.078 |
| Interaction USABK vs. JPN |
−0.202 |
0.190 |
0.289 |
−0.001 |
0.001 |
0.074 |
−0.179 |
0.166 |
0.281 |
−0.000 |
0.001 |
0.606 |
*P<0.05 between White American (USAWT) and Japanese (JPN) participants, or between Black American (USABK) and JPN participants. Abbreviations as in Table 1.
Table 4.
BMI-, Heart Rate-, SBP and DBP-Adjusted Liner Regression Estimates of Change in Each Echocardiographic Parameter of the LV According to Age Unit (Years) According to Race for Men and Women
| |
Men |
Women |
| Indexed LVIDd (mm/m2) |
Indexed LVIDs (mm/m2) |
Indexed LVIDd (mm/m2) |
Indexed LVIDs (mm/m2) |
| Estimate |
SE |
P value |
Estimate |
SE |
P value |
Estimate |
SE |
P value |
Estimate |
SE |
P value |
| JPN |
−0.047 |
0.014 |
<0.001 |
−0.036 |
0.010 |
<0.001 |
−0.016 |
0.013 |
0.224 |
−0.015 |
0.010 |
0.137 |
| USAWT |
−0.052 |
0.019 |
0.006 |
−0.043 |
0.015 |
0.004 |
−0.035 |
0.019 |
0.061 |
−0.060 |
0.014 |
<0.001 |
| USABK |
−0.017 |
0.030 |
0.576 |
−0.004 |
0.023 |
0.849 |
−0.064 |
0.031 |
0.041 |
−0.023 |
0.023 |
0.323 |
| Interaction USAWT vs. JPN |
−0.005 |
0.023 |
0.817 |
−0.007 |
0.018 |
0.706 |
−0.019 |
0.022 |
0.383 |
−0.045 |
0.016 |
0.006* |
| Interaction USABK vs. JPN |
0.030 |
0.033 |
0.357 |
0.032 |
0.025 |
0.211 |
−0.048 |
0.033 |
0.150 |
−0.008 |
0.025 |
0.740 |
| |
LVEDVI (mL/m2) |
LVESVI (mL/m2) |
LVEDVI (mL/m2) |
LVESVI (mL/m2) |
| Estimate |
SE |
P value |
Estimate |
SE |
P value |
Estimate |
SE |
P value |
Estimate |
SE |
P value |
| JPN |
−0.324 |
0.050 |
<0.001 |
−0.145 |
0.022 |
<0.001 |
−0.209 |
0.047 |
<0.001 |
−0.073 |
0.019 |
<0.001 |
| USAWT |
−0.362 |
0.069 |
<0.001 |
−0.140 |
0.031 |
<0.001 |
−0.516 |
0.066 |
<0.001 |
−0.218 |
0.027 |
<0.001 |
| USABK |
−0.263 |
0.109 |
0.017 |
−0.112 |
0.049 |
0.023 |
−0.161 |
0.110 |
0.145 |
−0.062 |
0.044 |
0.166 |
| Interaction USAWT vs. JPN |
−0.038 |
0.084 |
0.650 |
0.005 |
0.038 |
0.889 |
−0.307 |
0.077 |
<0.001* |
−0.145 |
0.031 |
<0.001* |
| Interaction USABK vs. JPN |
0.061 |
0.120 |
0.612 |
0.033 |
0.054 |
0.547 |
0.048 |
0.117 |
0.679 |
0.012 |
0.047 |
0.802 |
| |
LVSVI (mL/m2) |
LVEF (%) |
LVSVI (mL/m2) |
LVEF (%) |
| Estimate |
SE |
P value |
Estimate |
SE |
P value |
Estimate |
SE |
P value |
Estimate |
SE |
P value |
| JPN |
−0.179 |
0.031 |
<0.001 |
0.047 |
0.016 |
0.003 |
−0.136 |
0.031 |
<0.001 |
0.000 |
0.015 |
0.992 |
| USAWT |
−0.222 |
0.043 |
<0.001 |
0.009 |
0.022 |
0.693 |
−0.298 |
0.043 |
<0.001 |
0.064 |
0.021 |
0.003 |
| USABK |
−0.150 |
0.068 |
0.028 |
0.024 |
0.034 |
0.485 |
−0.099 |
0.072 |
0.171 |
−0.008 |
0.035 |
0.820 |
| Interaction USAWT vs. JPN |
−0.043 |
0.052 |
0.407 |
−0.039 |
0.026 |
0.141 |
−0.162 |
0.050 |
0.002* |
0.064 |
0.024 |
0.010* |
| Interaction USABK vs. JPN |
0.028 |
0.075 |
0.705 |
−0.023 |
0.038 |
0.533 |
0.037 |
0.077 |
0.634 |
−0.008 |
0.037 |
0.827 |
| |
LVMI (g/m2) |
RWT |
LVMI (g/m2) |
RWT |
| Estimate |
SE |
P value |
Estimate |
SE |
P value |
Estimate |
SE |
P value |
Estimate |
SE |
P value |
| JPN |
0.049 |
0.077 |
0.530 |
0.002 |
0.000 |
<0.001 |
0.289 |
0.069 |
<0.001 |
0.002 |
0.000 |
<0.001 |
| USAWT |
−0.191 |
0.108 |
0.078 |
0.002 |
0.000 |
0.002 |
0.055 |
0.096 |
0.572 |
0.001 |
0.001 |
0.011 |
| USABK |
−0.086 |
0.170 |
0.612 |
0.001 |
0.001 |
0.404 |
0.156 |
0.160 |
0.333 |
0.002 |
0.001 |
0.031 |
| Interaction USAWT vs. JPN |
−0.240 |
0.130 |
0.068 |
−0.001 |
0.001 |
0.278 |
−0.235 |
0.112 |
0.037* |
−0.001 |
0.001 |
0.130 |
| Interaction USABK vs. JPN |
−0.135 |
0.187 |
0.473 |
−0.002 |
0.001 |
0.070 |
−0.133 |
0.170 |
0.434 |
−0.000 |
0.001 |
0.684 |
*P<0.05 between White American (USAWT) and Japanese (JPN) participants, or between Black American (USABK) and JPN participants. Abbreviations as in Table 1.
The BSA-indexed LV internal dimensions (LVIDd, LVIDs) had a significant negative slope with age in Japanese and White American men, but not in Black American men. The BSA-indexed LV volumes and stroke volume (LVEDVI, LVESVI, LVSVI) had a significant negative slope with age in all 3 races. LVEF had a significantly positive slope with age only in Japanese men. LVMI did not change significantly with age in all races. Meanwhile, RWT increased significantly with age in Japanese and White American men. Further, there were no significant racial interactions in age-related changes for all echocardiographic parameters. The same analysis was performed after adjusting for BMI, HR, SBP, and DBP, and the results were similar.
In women, the indexed LVIDd had a significantly negative change with age in all 3 races, and there was no clear racial interaction. However, after adjusting for BMI, HR, SBP and DBP, the indexed LVIDd did not change significantly with age in all races. The slope of the indexed LVIDs against age was significantly negative in Japanese and White American women, but was significantly negative only in White American women after adjustment for BMI, HR, SBP, and DBP. The 1-year decrease in the indexed LVIDs of White American women was significantly greater than that of Japanese women, even after adjustment for BMI, HR, SBP, and DBP, but the difference between the 2 races was small (0.046 mm/m2). In contrast, LVEDVI, LVESVI, and LVSVI presented with significant negative age-related changes in Japanese and White American women, with and without adjustment for BMI, HR, SBP, and DBP. The 1-year decreases in LVEDVI, LVESVI, and LVSVI of White American women was significantly greater than those of Japanese women by 0.307, 0.145, and 0.162 mL/m2, respectively, after adjustment for BMI, HR, SBP, and DBP. That is, the difference in the decrease in LVEDVI, LVESVI, and LVSVI over 10 years is 3.07, 1.45, and 1.62 mL/m2, respectively, which is not negligible because these values correspond to 5.7%, 7.4%, and 4.7% of mean LVEDVI (53.7 mL/m2), LVESVI (19.4 mL/m2), and LVSVI (34.3 mL/m2) in White American women of all ages, respectively. The age-related change in LVEF was significantly positive only in White American women. The 1-year increase in LVEF of White American women was significantly greater than that of Japanese women, even after adjustment for BMI, HR, SBP, and DBP, but the difference between the 2 races was small (0.064%). The change in the LVMI with age was significantly positive only in Japanese women, and the 1-year increase in LVMI of White American was significantly lower than that of Japanese women by 0.235 g/m2, even after adjusting for BMI, HR, SBP, and DBP. The age-related change in RWT was significantly positive in all races, and there was no clear racial interaction, with and without adjustment for BMI, HR, SBP and DBP. There was no clear interaction effect on the association between age and all echocardiographic parameters, with and without adjustment for BMI, HR, SBP, and DBP, between Black American women and Japanese women.
Discussion
This study examined the racial differences in age-related changes in LV structure and LV systolic function using data collected from healthy Japanese, White American, and Black American participants among the WASE study participants. In men, age-related changes in all echocardiographic parameters did not differ significantly between Japanese and White American participants and between Japanese and Black American participants, with and without adjustment for factors, such as BMI, HR, and BP, which are related to LV structure and systolic function. In contrast, White American women had smaller BSA-indexed LV volumes, BSA-indexed LV internal dimension at end-systole, BSA-indexed LVSV, and LVMI, and a higher LVEF with age compared with Japanese women, even after adjustment for BMI, HR, and BP. There was no clear interaction effect on the association between age and all echocardiographic parameters, with and without adjustment for BMI, HR, and BP, between Black American and Japanese women. To the best of our knowledge, this study is the first to examine racial differences between ages and LV structure and function after adjusting for BMI, HR, and BP.
There are almost consistent data from previous cross-sectional studies using 2D or 3-dimensional (3D) echocardiographic studies showing an age-related decline in the LVEDVI and LVESVI in both sexes.3–8 The present study also showed that the LVEDVI and LVESVI were negatively correlated with age in both sexes in all 3 races, except for Black American women. A new finding in this study was that, although there were no racial differences in the slope of LVEDVI and LVESVI with age in men, in women the negative slope was significantly greater in White American participants than in Japanese participants, with or without adjustment for BMI, HR, and BP. This indicates that White American women are more likely to have smaller LV volumes with age than Japanese women.
In contrast, the association between LV internal dimensions and age was not consistent across studies.2 The present study also showed that the association between the LV internal dimensions and age differs based on sex, race, and after adjustments are made for BMI, HR, and BP. A novel finding of this study was the absence of statistically significant racial differences in the age-related change in the indexed LVIDd in both sexes, with and without adjustment for BMI, HR, and BP. Although the slope of the age-related decline in the indexed LVIDs was significantly greater in White American women than in Japanese women, the absolute value of the annual change was small.
There was a significant difference in the age-related slope of LV volume between White American and Japanese women. However, no significant difference was observed in the age-related slope of end-diastolic LV internal dimension and a significant but small difference in the end-systolic LV internal dimension. One study using echocardiography26 and another using CMR27 investigated age-related changes in the ratio of the LV long-axis diameter (LV end-diastolic length [LVEDL]) to the LV short-axis diameter (LVEDL/LVIDd). These studies revealed that the LVEDL/LVIDd ratio decreases with age, and the sphericity of the LV increases with age. Therefore, our results indicate a greater age-related shortening of the LV long axis length in White American women than in Japanese women. In the future, analysis using 3D echocardiography may be required.
LVSV indexed to the BSA (LVSVI) has almost consistently shown to decrease with age based on previous cross-sectional studies using 3D echocardiography4,8,10 or CMR.11–14 Regarding the association between age and LVEF, findings using 2D and 3D echocardiography vary.2 However, previous studies using CMR have reported that LVEF increases with age.11,12,14 In the present study, the LVSVI had a significant negative slope with age in both sexes of all races, except for Black American women. Moreover, the LVEF increased significantly with age only in Japanese men and White American women. A novel finding of this study was that there were no significant differences in age-related changes in LVSVI and LVEF in men based on race. However, compared with Japanese women, White American women presented with a significant decrease in LVSVI and an increase in LVEF with age.
LVMI did not have a consistent association with age in previous studies using echocardiography or CMR.2 In contrast, RWT, determined based on the ratio of wall thickness to end-diastolic LV internal dimension, is used as an indicator of concentric remodeling and has been shown to increase with age in healthy participants based on previous echocardiographic studies.9,26 In the present study, no statistically significant age-related changes in LVMI were observed in both sexes of all races, except in Japanese women. Even after adjusting for BMI, HR, and BP, Japanese women had a significantly greater age-related changes in LVMI compared with White American women. However, the magnitude of the change is small. The RWT increased with age in both sexes of all races, and there were no racial differences in age-related increases in RWT in both sexes, with and without adjustment for BMI, HR, and BP. The equation for estimating LV mass and RWT on echocardiography uses LVIDd and LVPWd measured at the basal part of the LV; therefore, it does not consider the entire LV wall, including the mid and apical part of the LV, and the age-related change in LV shape. In elderly patients with a shortened LV long axis diameter, the LV mass may be incorrectly measured. Therefore, it is necessary to assess racial differences in age-related changes in LV mass or concentric LV remodeling via CMR examinations in the future.
This study has some limitations. First, the number of participants was small because this study was a subanalysis of the WASE study, potentially losing some statistical power to detect differences. The small number of participants also led us to analyze racial differences in age-related changes in LV echocardiographic parameters under the assumptions of linear regression models rather than epidemiological breakdown tables. In linear regression analysis, generally 10 or more participants per explanatory variable can be included.28 In this study, we assumed that adjustment variables (BMI, HR, SBP, DBP) have a common slope for all races, and further assumed that the slopes of each echocardiographic index depending on age vary by race. Therefore, when race was combined for each sex, there were enough participants for each sex, which exceeded the standard of 90 participants (9 explanatory variables multiplied by 10 participants), so it was possible to analyze this. However, it should be noted that estimates of LV parameters are variable, because there are certainly errors in the estimated coefficients. Furthermore, there were few participants over 70 years of age, so it is unclear whether the results of this study apply to people over the age of 70 years. To confirm the results of this cross-sectional study, longitudinal studies, which evaluate serial echocardiograms of the same person over time, are needed. Second, because only data from Japanese and American participants in the WASE study were used, extrapolation to countries other than Japan and the US may not be possible. Hence, future studies should evaluate differences in age-related changes in LV structure and function between races other than Japanese and American. Third, the reason for racial differences in age-related changes in LV structure and function in women but not men could not be examined. LV geometry may be influenced by various factors, including lifestyle habits such as exercise routines, and race is not the sole determinant. Unfortunately, in this study we did not collect information regarding participants’ lifestyles, and therefore were unable to examine the effects of these lifestyle differences. Further research is necessary to understand the effects of lifestyle differences among races on LV geometry and function.
Conclusions
Compared with Japanese women, White American women had a smaller BSA-indexed LV volume, BSA-indexed LV internal dimension at end-systole, BSA-indexed LVSV, and LVMI, and a larger LVEF with age even after adjusting for BMI, HR, and BP. The age-related changes in these echocardiographic parameters did not differ significantly between Black American and Japanese women. Moreover, there were no racial differences in men. The results of this study suggest that age-related changes in LV structure and function may differ by race, which is important for a deeper understanding of the pathogenesis of HFpEF. Therefore, future studies examining echocardiographic reference values for each age group in each race are needed.
Sources of Funding
This work was partly supported by a Grant-in-Aid for Scientific Research C (22K12860) (T.N.) and (21K12701) (M.D.) from the Japan Society for the Promotion of Science.
Disclosures
M.D. is a member of Circulation Journal’s Editorial Team.
IRB Information
The Institutional Ethics Committee of The University of Tokyo approved this WASE study (Reference no. 11345).
Data Availability
The deidentified participant data will not be shared.
Supplementary Files
Please find supplementary file(s);
https://doi.org/10.1253/circj.CJ-24-0031
References
- 1.
Obokata M, Sorimachi H, Harada T, Kagami K, Saito Y, Ishii H. Epidemiology, pathophysiology, diagnosis, and therapy of heart failure with preserved ejection fraction in Japan. J Card Fail 2023; 29: 375–388, doi:10.1016/j.cardfail.2022.09.018.
- 2.
Peverill RE. Changes in left ventricular size, geometry, pump function and left heart pressures during healthy aging. Rev Cardiovasc Med 2021; 22: 717–729, doi:10.31083/j.rcm2203079.
- 3.
Daimon M, Watanabe H, Abe Y, Hirata K, Hozumi T, Ishii K, et al. Normal values of echocardiographic parameters in relation to age in a healthy Japanese population: The JAMP study. Circ J 2008; 72: 1859–1866, doi:10.1253/circj.CJ-08-0171.
- 4.
Kaku K, Takeuchi M, Otani K, Sugeng L, Nakai H, Haruki N, et al. Age- and gender-dependency of left ventricular geometry assessed with real-time three-dimensional transthoracic echocardiography. J Am Soc Echocardiogr 2011; 24: 541–547, doi:10.1016/j.echo.2011.01.011.
- 5.
Chahal NS, Lim TK, Jain P, Chambers JC, Kooner JS, Senior R. Population-based reference values for 3D echocardiographic LV volumes and ejection fraction. JACC Cardiovasc Imaging 2012; 5: 1191–1197, doi:10.1016/j.jcmg.2012.07.014.
- 6.
Fukuda S, Watanabe H, Daimon M, Abe Y, Hirashiki A, Hirata K, et al. Normal values of real-time 3-dimensional echocardiographic parameters in a healthy Japanese population: The JAMP-3D Study. Circ J 2012; 76: 1177–1181, doi:10.1253/circj.CJ-11-1256.
- 7.
Kou S, Caballero L, Dulgheru R, Voilliot D, De Sousa C, Kacharava G, et al. Echocardiographic reference ranges for normal cardiac chamber size: Results from the NORRE study. Eur Heart J Cardiovasc Imaging 2014; 15: 680–690, doi:10.1093/ehjci/jet284.
- 8.
Bernard A, Addetia K, Dulgheru R, Caballero L, Sugimoto T, Akhaladze N, et al. 3D echocardiographic reference ranges for normal left ventricular volumes and strain: Results from the EACVI NORRE study. Eur Heart J Cardiovasc Imaging 2017; 18: 475–483, doi:10.1093/ehjci/jew284.
- 9.
Vriz O, Pirisi M, Habib E, Galzerano D, Fadel B, Antonini-Canterin F, et al. Age related structural and functional changes in left ventricular performance in healthy subjects: A 2D echocardiographic study. Int J Cardiovasc Imaging 2019; 35: 2037–2047, doi:10.1007/s10554-019-01665-y.
- 10.
Patel HN, Miyoshi T, Addetia K, Henry MP, Citro R, Daimon M, et al. Normal values of cardiac output and stroke volume according to measurement technique, age, sex, and ethnicity: Results of the World Alliance of Societies of Echocardiography study. J Am Soc Echocardiogr 2021; 34: 1077–1085.e1, doi:10.1016/j.echo.2021.05.012.
- 11.
Chuang ML, Gona P, Hautvast GL, Salton CJ, Breeuwer M, O’Donnell CJ, et al. CMR reference values for left ventricular volumes, mass, and ejection fraction using computer-aided analysis: The Framingham Heart Study. J Magn Reson Imaging 2014; 39: 895–900, doi:10.1002/jmri.24239.
- 12.
Yeon SB, Salton CJ, Gona P, Chuang ML, Blease SJ, Han Y, et al. Impact of age, sex, and indexation method on MR left ventricular reference values in the Framingham Heart Study offspring cohort. J Magn Reson Imaging 2015; 41: 1038–1045, doi:10.1002/jmri.24649.
- 13.
Eng J, McClelland RL, Gomes AS, Hundley WG, Cheng S, Wu CO, et al. Adverse left ventricular remodeling and age assessed with cardiac MR imaging: The Multi-Ethnic Study of Atherosclerosis. Radiology 2016; 278: 714–722, doi:10.1148/radiol.2015150982.
- 14.
Bülow R, Ittermann T, Dörr M, Poesch A, Langner S, Völzke H, et al. Reference ranges of left ventricular structure and function assessed by contrast-enhanced cardiac MR and changes related to ageing and hypertension in a population-based study. Eur Radiol 2018; 28: 3996–4005, doi:10.1007/s00330-018-5345-y.
- 15.
Merillon JP, Motte G, Masquet C, Azancot I, Guiomard A, Gourgon R. Relationship between physical properties of the arterial system and left ventricular performance in the course of aging and arterial hypertension. Eur Heart J 1982; 3 Suppl A: 95–102, doi:10.1093/eurheartj/3.suppl_a.95.
- 16.
Downes TR, Nomeir AM, Smith KM, Stewart KP, Little WC. Mechanism of altered pattern of left ventricular filling with aging in subjects without cardiac disease. Am J Cardiol 1989; 64: 523–527, doi:10.1016/0002-9149(89)90433-5.
- 17.
Carrick-Ranson G, Hastings JL, Bhella PS, Shibata S, Fujimoto N, Palmer MD, et al. Effect of healthy aging on left ventricular relaxation and diastolic suction. Am J Physiol Heart Circ Physiol 2012; 303: H315–H322, doi:10.1152/ajpheart.00142.2012.
- 18.
Wolsk E, Bakkestrøm R, Thomsen JH, Balling L, Andersen MJ, Dahl JS, et al. The Influence of Age on Hemodynamic Parameters During Rest and Exercise in Healthy Individuals. JACC Heart Fail 2017; 5: 337–346, doi:10.1016/j.jchf.2016.10.012.
- 19.
Asch FM, Miyoshi T, Addetia K, Citro R, Daimon M, Desale S, et al. Similarities and differences in left ventricular size and function among races and nationalities: Results of the World Alliance Societies of Echocardiography Normal Values Study. J Am Soc Echocardiogr 2019; 32: 1396–1406.e2, doi:10.1016/j.echo.2019.08.012.
- 20.
Alpert MA, Omran J, Mehra A, Ardhanari S. Impact of obesity and weight loss on cardiac performance and morphology in adults. Prog Cardiovasc Dis 2014; 56: 391–400, doi:10.1016/j.pcad.2013.09.003.
- 21.
Saba MM, Ibrahim MM, Rizk HH. Gender and the relationship between resting heart rate and left ventricular geometry. J Hypertens 2001; 19: 367–373, doi:10.1097/00004872-200103000-00003.
- 22.
Cuspidi C, Sala C, Tadic M, Gherbesi E, Facchetti R, Grassi G, et al. High-normal blood pressure and abnormal left ventricular geometric patterns: A meta-analysis. J Hypertens 2019; 37: 1312–1319, doi:10.1097/HJH.0000000000002063.
- 23.
Ministry of Health, Labour and Welfare (Japan). The national health and nutrition survey Japan, 2019 [in Japanese]. https://www.mhlw.go.jp/content/001066903.pdf (accessed April 9, 2024).
- 24.
Asch FM, Banchs J, Price R, Rigolin V, Thomas JD, Weissman NJ, et al. Need for a global definition of normative echo values: Rationale and design of the World Alliance of Societies of Echocardiography Normal Values Study (WASE). J Am Soc Echocardiogr 2019; 32: 157–162.e2, doi:10.1016/j.echo.2018.10.006.
- 25.
Lang RM, Badano LP, Mor-Avi V, Afilalo J, Armstrong A, Ernande L, et al. Recommendations for cardiac chamber quantification by echocardiography in adults: An update from the American Society of Echocardiography and the European Association of Cardiovascular Imaging. J Am Soc Echocardiogr 2015; 28: 1–39.e14, doi:10.1016/j.echo.2014.10.003.
- 26.
Støylen A, Mølmen HE, Dalen H. Importance of length and external diameter in left ventricular geometry: Normal values from the HUNT Study. Open Heart 2016; 3: e000465, doi:10.1136/openhrt-2016-000465.
- 27.
Hees PS, Fleg JL, Lakatta EG, Shapiro EP. Left ventricular remodeling with age in normal men versus women: Novel insights using three-dimensional magnetic resonance imaging. Am J Cardiol 2002; 90: 1231–1236, doi:10.1016/s0002-9149(02)02840-0.
- 28.
Schmidt AF, Finan C. Linear regression and the normality assumption. J Clin Epidemiol 2018; 98: 146–151, doi:10.1016/j.jclinepi.2017.12.006.
