Journal of Obesity & Metabolic Syndrome

Search

Article

J Obes Metab Syndr 2025; 34(1): 54-64

Published online January 30, 2025 https://doi.org/10.7570/jomes24027

Copyright © Korean Society for the Study of Obesity.

Ratio of Skeletal Muscle Mass to Visceral Fat Area Is a Useful Marker for Assessing Left Ventricular Diastolic Dysfunction among Koreans with Preserved Ejection Fraction: An Analysis of the Random Forest Model

Jin Kyung Oh1, Yuri Seo2, Wonmook Hwang1, Sami Lee2, Yong-Hoon Yoon1, Kyupil Kim2, Hyun Woong Park1, Jae-Hyung Roh1, Jae-Hwan Lee1, Minsu Kim1,*

1Division of Cardiology, Department of Internal Medicine, Chungnam National University Sejong Hospital, Chungnam National University College of Medicine, Sejong; 2Department of Family Medicine, Chungnam National University Sejong Hospital, Sejong, Korea

Correspondence to:
Minsu Kim
https://orcid.org/0000-0001-9230-3137
Division of Cardiology, Department of Internal Medicine, Chungnam National University Sejong Hospital, Chungnam National University College of Medicine, 20 Bodeum 7-ro, Sejong 30099, Korea
Tel: +82-44-995-5691
Fax: +82-44-995-5698
E-mail: mskimep80@gmail.com

The first two authors contributed equally to this study.

Received: September 13, 2024; Reviewed : December 6, 2024; Accepted: January 9, 2025

This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0/) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.

Background: Although the presence of both obesity and reduced muscle mass presents a dual metabolic burden and additively has a negative effect on a variety of cardiometabolic parameters, data regarding the associations between their combined effects and left ventricular diastolic function are limited. This study investigated the association between the ratio of skeletal muscle mass to visceral fat area (SVR) and left ventricular diastolic dysfunction (LVDD) in patients with preserved ejection fraction using random forest machine learning.
Methods: In total, 1,070 participants with preserved left ventricular ejection fraction who underwent comprehensive health examinations, including transthoracic echocardiography and bioimpedance body composition analysis, were enrolled. SVR was calculated as an index of sarcopenic obesity by dividing the appendicular skeletal muscle mass by the visceral fat area.
Results: In the random forest model, age and SVR were the most powerful predictors of LVDD. Multivariate logistic regression analysis demonstrated that older age (adjusted odds ratio [OR], 1.11; 95% confidence interval [CI], 1.07 to 1.15) and lower SVR (adjusted OR, 0.08; 95% CI, 0.01 to 0.57) were independent risk factors for LVDD. SVR showed a significant improvement in predictive performance and fair predictability for LVDD, with the highest area under the curve noted in both men and women, with statistical significance. In non-obese and metabolically healthy individuals, the lowest SVR tertile was associated with a greater risk of LVDD compared to the highest SVR tertile.
Conclusion: Decreased muscle mass and increased visceral fat were significantly associated with LVDD compared to obesity, body fat composition, and body muscle composition indices.

Keywords: Intra-abdominal fat, Muscle, Skeletal, Body composition, Ventricular dysfunction, Left, Random forest

The prevalence of overweightness and obesity has increased in modern communities, posing a significant health burden due to the heightened risk of cardiovascular disease.1,2 Obesity induces several changes in cardiac structure and function, including increased left ventricular (LV) mass and impaired LV systolic and diastolic function.3-5 Left ventricular diastolic dysfunction (LVDD) is characterized by impaired relaxation and deteriorated filling properties of the left ventricle and is commonly detected in otherwise healthy and asymptomatic patients.6-9 Its influence on diastolic function and hemodynamics seems to be multifactorial, stemming from elevated inflammation and oxidative stress associated with obesity10 as well as from related conditions like diabetes mellitus (DM), dyslipidemia, and hypertension (HTN).11

Body mass index (BMI) is a common metric for assessing obesity; however, its efficacy is limited as it fails to distinguish between fat mass and lean body mass and overlooks variations in body fat distribution.12 Sarcopenic obesity is the coexistence of sarcopenia and visceral obesity. As loss of skeletal muscle mass and increase in body fat are associated with increased morbidity and mortality, both obesity and reduced muscle mass may present a dual metabolic burden and collectively affect various cardiometabolic parameters.13-16 However, the association between these combined effects and LV diastolic function remains unclear. Therefore, this study aimed to evaluate the ratio of skeletal muscle mass to visceral fat area (SVR) as an index of skeletal muscle mass corrected by visceral fat area (VFA) and comprehensively assess LVDD in patients with preserved ejection fraction using random forest machine learning.

We screened 1,554 participants aged >18 years who underwent comprehensive health examinations, including transthoracic echocardiography and bioimpedance body composition analysis, at the Chungnam National University Sejong Hospital’s Healthcare Center, Republic of Korea, from July 1, 2020 to December 31, 2023. We excluded 281 patients with left ventricular ejection fraction (LVEF) of <50% and 152 patients with atrial fibrillation at baseline. Additionally, patients with significant valve disease, previous valve surgery, congenital heart disease, hypertrophic or restrictive cardiomyopathy, or renal insufficiency were excluded. A total of 1,070 participants (521 men and 549 women) were finally included in this study (Supplementary Fig. 1). The study protocol was approved by the Institutional Review Board and Ethics Committee of Chungnam National University Sejong Hospital (no. 2021-03-003), which waived the need for informed consent for this retrospective study. The study was conducted in accordance with the principles of the Declaration of Helsinki.

Echocardiographic evaluation and definition of diastolic dysfunction

Transthoracic echocardiography was performed using commercially available echocardio-graphic systems (Epiq CVx [Philips Healthcare] or Vivid E95 [GE Healthcare]). Subsequently, the stored images were reviewed and analyzed by a single experienced echocardiographic specialist, JKO, to evaluate diastolic function. LV and left atrium (LA) chamber quantification was performed according to the recommendations of the American Society of Echocardiography,17 and a modified biplanar method was used to calculate the LVEF. LV and LA volumes were assessed from apical four- and two-chamber views using Simpson’s biplane method and indexed by body surface area. LV diastolic function was analyzed using the mitral inflow pattern. From an apical four-chamber view, the transmitral flow was sampled using pulsed-wave Doppler at the level of the mitral valve leaflet tips. The early mitral inflow peak velocity (E) and late mitral inflow peak velocity (A) were measured, and their ratio (E/A) was calculated.18 Doppler tissue interrogation of the early diastolic mitral annular relaxation velocity (E′) and late diastolic velocity (A′) were recorded at the septal annulus, and the E/E′ ratio was calculated as an index of LV filling pressures. LVDD with normal LVEF was defined by an LVEF of >50% and present if three or four variables met the following cutoff values: (1) average E/E′ ratio >14; (2) septal E’ velocity <7 cm/sec or lateral E′ velocity <10 cm/sec; (3) tricuspid regurgitation velocity >2.8 m/sec; and (4) LA volume index >34 mL/m2.18

Body composition assessment using bioelectrical impedance measurements

All participants’ body compositions were analyzed using dual bioelectrical impedance measurements (Inbody 970; Biospace) on the same day as echocardiography. According to the instructions provided by the manufacturer, the participants stood upright and comfortably on the analyzer’s footplate barefoot, with their legs and arms apart. The lean body mass of the arms and legs, appendicular skeletal muscle mass (ASM), fat-free mass, total body fat percentage, and VFA were measured.19 The ASM was estimated as the sum of the muscle mass estimated individually for each arm and leg.20 Absolute ASM was converted to skeletal muscle mass index (SMI) by dividing the value by the square of the height in meters (kg/m2),21 and SVR (kg/cm2) was calculated as an index of sarcopenic obesity by dividing the ASM (kg) by VFA (cm2).

Development of a random forest prediction model for LVDD

Random forest, a classic machine learning algorithm,22 was applied to identify associations between body composition assessment data and LVDD (as defined above). For the initial random forest model development, we included 20 variables (nine demographic/comorbidities, three laboratory data, and eight body composition assessment data) (Supplementary Table 1). From the random forest model including 20 variables, the important predictors of LVDD were identified using the R package ‘Boruta,’23 which is a feature-selection method paired with random forest and selects features that have discriminating importance scores (Z-scores) greater than randomly permuted features (shadow matrixes). The proportion of missing values was minimal, and these were imputed using the ‘miss Forest’ algorithm before model development.

Statistical analysis

Continuous variables are presented as mean±standard deviation or median (interquartile range) values, whereas categorical variables are presented as numbers and percentages. Group comparisons were evaluated using the chi-squared or Fisher’s exact test for categorical variables and Student’s t-test or the Mann–Whitney U-test for continuous variables, as appropriate. Univariate logistic regression analyses were performed to evaluate the predictive values of each variable, and significant variables entered a multiple logistic regression model. Multivariable models included significant univariate variables and known clinical risk factors like age, sex, systolic blood pressure (SBP), HTN, DM, dyslipidemia, hemoglobin, estimated glomerular filtration rate (eGFR), fasting plasma glucose, total cholesterol, BMI, waist-hip ratio, SVR, and SMI. Odds ratios (ORs) and 95% confidence intervals (CIs) were reported. To evaluate the prognostic utility of SVR in predicting LVDD, the receiver operating characteristic (ROC) curve was plotted, and the area under the ROC curve (AUC) was calculated. The incremental value of SVR was assessed in a series of logistic models, in which the first model consisted of fitting the clinical data entered as a block (age, sex, HTN, DM, dyslipidemia, SBP, hemoglobin, glucose, and eGFR) in the test group. The χ2 values were compared across models. We also calculated the integrated discrimination improvement and continuous net reclassification improvement for the prediction of LVDD, following the methodology of Pencina et al.24,25 All reported P-values were two-sided, and statistical significance was set at P<0.05. All analyses were performed using SPSS version 21.0 (IBM Corporation) and R version 3.6.0 (R Foundation for Statistical Computing).

Study population

The baseline demographic and clinical characteristics of the patients according to the presence of LVDD are shown in Table 1. The patients were divided into a normal LV diastolic function group (n=997) and an LVDD group (n=73). When comparing the two groups, patients with LVDD were older and included greater proportions of women and individuals with HTN, DM, dyslipidemia, or a history of cerebrovascular disease. Among the laboratory findings, hemoglobin, eGFR, total cholesterol, and low-density lipoprotein cholesterol levels were lower, while fasting glucose was higher, in the group with LVDD. Following echocardiography, it was revealed that the LVDD group had a greater LA volume index and LV mass index and higher early and late mitral inflow peak velocities (E and A), early diastolic mitral annular relaxation velocity (E′), E/E′ ratio, and tricuspid regurgitation peak velocity, respectively, than the normal LV diastolic function group. LV dimensions and LVEF were not significantly different between the two groups.

Comparison of body composition assessment using bioelectrical impedance measurements

Table 1 presents the comparison of body composition assessment using bioelectrical impedance measurements of the patients according to the presence of LVDD. In the LVDD group, the mean values of BMI, percentage body fat, VFA, and SVR were 25.6 kg/m², 33.4%, 165.3 cm2, and 0.19 kg/cm2, respectively. The mean waist circumference and waist-hip ratio were 71.8 cm and 1.02, respectively. Significant differences were observed in body fat and muscle composition between the two groups. The body surface area, waist circumference, waist-hip ratio, and VFA were greater and the body fat percentage and SVR were higher in the LVDD group than in the normal diastolic function group. However, SMI and appendicular muscle mass were significantly lower in the LVDD group. BMI and fat-free mass index were not statistically different between the two groups.

Feature selection and ranking by random forest

Fig. 1 displays the important variables contributing to the risk of LVDD, as identified by random forest, with the 20 variables ranked by relative importance. In the random forest model, age was the most important predictor of LVDD. SBP, HTN, diastolic blood pressure, history of cardiovascular disease, and eGFR emerged as important clinical risk factors. However, other comorbidities, such as DM, dyslipidemia, and a history of coronary artery disease, had relatively low importance, indicating little predictive value. Regarding the body composition assessment data, SVR was the second most powerful predictor among all variables. VFA, body fat percentage, waist-hip ratio, waist circumference, and SMI were also predictors with greater importance. However, BMI, a more commonly recognized obesity index, offered comparatively little predictive importance.

Predictors of LVDD

Table 2 shows the results of the logistic regression analysis performed to reveal the relationship between the variables and LVDD prediction. On univariate logistic regression analysis, body fat percentage, waist circumference, waist-hip ratio, VFA, SVR, and SMI showed significant associations with LVDD. In multivariate analyses, age (adjusted OR, 1.11; 95% CI, 1.07 to 1.15; P<0.001) and SVR (adjusted OR, 0.08; 95% CI, 0.01 to 0.57; P=0.027) showed a significant association with LVDD among the parameters significant in the univariate analysis. The results indicated that older age and lower SVR are independent risk factors for LVDD after adjusting for age, sex, SBP, HTN, DM, dyslipidemia, smoking, alcohol consumption, hemoglobin level, eGFR, fasting plasma glucose level, cholesterol level, BMI, waist-hip ratio, SVR, and SMI (Supplementary Table 2). To reduce multicollinearity between the data, body fat percentage, VFA, and arm muscle circumference were not simultaneously entered into the multivariate logistic regression model.

Incremental prognostic value of SVR for LVDD

The prognostic performances of SVR, waist-hip ratio, BMI, and SMI are shown in Fig. 2. Model 1, including clinical variables (age, sex, HTN, DM, dyslipidemia, SBP, hemoglobin, fasting plasma glucose, cholesterol, and eGFR), yielded an overall χ2 value of 100.6. Model 2 adjustments included the factors in model 1 and SMI, whereas model 3 considered the adjusted factors in model 1 and BMI. In model 4, which incorporated the waist-hip ratio into model 1, the χ2 value increased from 100.6 to 109.4 (P=0.003). When SVR was added to model 1, the value increased from 100.6 to 117.6 in model 5 (P<0.001) (Fig. 2). The waist-hip ratio and SVR provided incremental prognostic values; however, the addition of SVR increased the prediction performance more than the combination of clinical data and SMI, BMI, and waist-hip ratio. Adding SVR showed a significant improvement in predictive performance (C-statistic=0.853 [95% CI, 0.814 to 0.892; P=0.014]; net classification index=0.411 [95% CI, 0.188 to 0.634; P<0.001]; and integrated discrimination improvement=0.016 [95% CI, 0.005 to 0.028; P=0.006]).

ROC curves for the prediction of LVDD in men and women

ROC analysis revealed the strength of the univariate relationship between different indices of body composition assessment (SVR, waist-hip ratio, BMI, and SMI) and LVDD in men and women (Fig. 3). The AUC for waist-hip ratio, BMI, and SMI were 0.57 (95% CI, 0.47 to 0.67) in men and 0.67 (95% CI, 0.59 to 0.75) in women, 0.52 (95% CI, 0.39 to 0.64) in men and 0.64 (95% CI, 0.55 to 0.72) in women, and 0.64 (95% CI, 0.55 to 0.72) in men and 0.52 (95% CI, 0.44 to 0.61) in women, respectively. SVR showed fair predictability for LVDD, with the highest AUC noted for men (AUC=0.69; 95% CI, 0.61 to 0.77) and women (AUC=0.71; 95% CI, 0.64 to 0.78) with statistical significance. The whole patient population was divided into sex-specific SVR tertiles (Q), as follows: Q1, <0.29 kg/cm2; Q2, 0.29–0.41 kg/cm2; Q3, >0.41 kg/cm2 in men and Q1, <0.17 kg/cm2; Q2, 0.17–0.26 kg/cm2; and Q3, >0.26 kg/cm2 in women. According to study findings, the prevalence of LVDD decreased with increasing SVR. In subgroup analysis, women had a greater prevalence of LVDD than men. Among men, the prevalence rates of LVDD in the Q1, Q2, and Q3 tertiles were 7.8%, 2.0%, and 0%, respectively. Among women, the prevalence rates of LVDD in the Q1, Q2, and Q3 levels were 18.7%, 6.35%, and 3.8%, respectively (P<0.05). In non-obese (adjusted OR, 3.5) and metabolically healthy patients (OR, 17.8), the lowest SVR tertile had a higher OR compared to the highest SVR tertile (Fig. 4). All individuals were considered metabolically healthy,26 as characterized by the absence of DM, HTN, or dyslipidemia diagnoses.

In this study, we explored the association between SVR and LVDD in a Korean population with preserved LVEF. SVR emerged as an independent factor associated with impaired LV diastolic function, with a lower SVR significantly elevating the risk of LVDD, even after adjusting for clinical risk factors and other body composition indices. Notably, SVR exhibited a greater predictive capacity for LVDD compared to BMI, VFA, and SMI. Moreover, lower SVR was significantly associated with LVDD in non-obese and metabolically healthy individuals.

As aging is related to a gradual loss of skeletal muscle mass and an increased incidence of obesity, several studies have shown that a reduction in muscle mass and an increase in fat mass contribute independently to mobility impairment in the elderly.5,13,14 Additionally, as individuals age, intra-abdominal fat tends to increase progressively due to the redistribution of body fat, even among healthy individuals who maintain a relatively stable BMI.27,28 Although BMI is the most widely used index for defining obesity, it does not provide information on the distribution of fat, such as visceral (around organs) versus subcutaneous (under the skin) fat, which contributes more significantly to an increased risk of cardiovascular disease.29 BMI tends to reflect overall body weight, which includes both lean body mass (muscles, bones, and organs) and fat mass. This is because the formula for BMI does not differentiate between these two components. Individuals with similar BMIs may have different body compositions, leading to variations in muscle mass and fat mass. Thus, to obtain a more comprehensive assessment of an individual’s body composition related to excess body fat, additional measures such as waist circumference, body fat percentage, and imaging techniques are needed.

Obesity can have diverse impacts on the structure and function of the LV. Obesity is associated with an overall increase in body mass, leading to an increase in LV mass.30 Interestingly, recent studies have investigated the association between adipose tissue distribution and cardiac function and have consistently demonstrated a robust association between visceral adiposity and LVDD. The visceral fat is metabolically active and releases inflammatory mediators. These bioactive substances can directly affect the myocardium, leading to alterations in LV diastolic function. Additionally, metabolic, hormonal, and hemodynamic factors are involved in these mechanisms and participate in complex interplay.31,32

Some studies have suggested that an appropriate ratio of healthy body fat to muscle mass may positively impact LV mass and diastolic function. Optimal lean body mass can reduce the load on the heart and enhance diastolic function. Additionally, it can improve insulin sensitivity and contribute to inflammatory suppression, thereby reducing the likelihood of cardiovascular and vascular diseases.33-35 Thus, a combined consideration of skeletal muscle mass and visceral fat provides a more comprehensive understanding of cardiac function. VFA calculation by computed tomography (CT) imaging is considered the gold-standard measure for assessing visceral adiposity owing to its accurate and direct quantification of visceral fat. However, its use is limited by factors, such as exposure to ionizing radiation, cost considerations, and requirements for specialized equipment.36,37 Thus, this study explored the association between SVR and LVDD in a large Korean population, using bioelectrical impedance measurements to assess skeletal muscle mass and VFA. The findings corroborate previous research, highlighting SVR as a more robust independent predictor of LVDD compared to BMI, SMI, or VFA, even among individuals who are metabolically healthy and not obese. As an index of sarcopenic obesity, SVR may be associated with LVDD through several mechanisms, including metabolic influences, inflammatory modulation, hormonal regulation, insulin sensitivity, and potential hemodynamic effects. The skeletal muscle is a metabolically active tissue that plays a crucial role in glucose metabolism and insulin sensitivity. Increased skeletal muscle mass may improve glucose metabolism and insulin sensitivity; however, visceral fat accumulation is associated with insulin resistance. Skeletal muscles secrete myokines, which are known to have anti-inflammatory effects. Conversely, visceral fat is associated with the release of pro-inflammatory adipokines. The balance achieved through SVR, with higher skeletal muscle and lower visceral fat, may contribute to an anti-inflammatory environment that potentially mitigates inflammation-related cardiac dysfunction.

The strengths of our study include its large sample size and accurate assessment of LV diastolic function according to the current guidelines. However, this study had several limitations. First, our study was a single-center retrospective analysis, and the limited data may have affected the statistical validity of BMI, VFA, SMI, and SVR. Second, a dual bioelectrical impedance analyzer was used to measure the VFA. Abdominal CT imaging is the gold standard for visceral adiposity assessment; however, a previous study revealed an obvious correlation between VFA measurements obtained through CT imaging and a dual bioelectrical impedance analyzer.38 Finally, given potential ethnic disparities in body composition, uncertainty remains about whether our findings in the Korean population can be extrapolated to other ethnic groups. Therefore, multicenter prospective studies are needed to obtain more significant results on this issue.

In conclusion, our study demonstrated that decreased muscle mass and increased visceral fat are significantly associated with LVDD in patients with preserved LVEF. Lower SVR level is an independent predictor of LVDD compared to other obesity, body fat composition, and body muscle composition indices. The combined consideration of skeletal muscle mass and visceral fat provided a more effective means of predicting LVDD, even in non-obese and metabolically healthy individuals.

Study concept and design: JKO, YS, and MK; acquisition of data: WH, SL, YHY, and KK; analysis and interpretation of data: JKO and YS; drafting of the manuscript: JKO; critical revision of the manuscript: HWP, JHR, and JHL; obtained funding: JKO; and study supervision: MK.

Fig. 1. Important feature selection by random forest. The plot displays the features leading up to the left ventricular diastolic dysfunction in order of importance of each variable. Important variables are in green boxes, rejected variables are in blue boxes, and tentative variables are in pink boxes. SVR, ratio of skeletal muscle mass to visceral fat area; VFA, visceral fat area; BMI, body mass index; SMI, skeletal muscle mass index; DBP, diastolic blood pressure; SBP, systolic blood pressure; eGFR, estimated glomerular filtration rate; CAD, coronary artery disease; CVD, cerebrovascular disease.
Fig. 2. Incremental prognostic value of the ratio of skeletal muscle mass to visceral fat area (SVR) in the prediction of left ventricular diastolic dysfunction. The addition of SVR to the model including established risk factors resulted in significant improvement in the prediction. P-value for model 1 vs. each of the other models. HTN, hypertension; DM, diabetes mellitus; SBP, systolic blood pressure; eGFR, estimated glomerular filtration rate; SMI, skeletal muscle mass index; BMI, body mass index; cNRI, continuous net reclassification index; IDI, integrated discrimination improvement; CI, confidence interval.
Fig. 3. The receiver operating characteristic curves for predicting left ventricular diastolic dysfunction according to sex. Receiver operating characteristic curves in (A) men and (B) women. *P<0.05 for the comparison between the SVR’s area under the curve and other models’ areas under the curve are indicated by the symbol. SVR, ratio of skeletal muscle mass to visceral fat area; CI, confidence interval; WHR, waist-hip ratio; SMI, skeletal muscle mass index; BMI, body mass index.
Fig. 4. Prevalence of left ventricular diastolic dysfunction (LVDD) across tertiles of the ratio of skeletal muscle mass to visceral fat area (SVR). Patients were divided into tertiles based on sex-specific SVR values. Prevalence of LVDD according to (A) sex, (B) presence of obesity, and (C) metabolically health patients. BMI, body mass index.

Baseline characteristics of the study population according to the presence of left ventricular diastolic dysfunction

Characteristic LVDD P
No (n=997) Yes (n=73)
Demographics
Age (yr) 54.9 ± 12.5 67.8 ± 8.8 < 0.001
Male sex 499 (50.1) 22 (30.1) 0.001
Hemodynamics
SBP (mmHg) 128.2 ± 16.1 134.8 ± 15.8 0.001
DBP (mmHg) 77.2 ± 11.8 74.4 ± 10.4 0.036
Heart rate (bpm) 77.2 ± 12.6 74.5 ± 13.6 0.109
History of disease data
Hypertension 347 (34.8) 45 (61.6) < 0.001
Diabetes mellitus 132 (13.2) 22 (30.1) < 0.001
Dyslipidemia 322 (32.3) 35 (47.9) 0.010
History of coronary heart disease 66 (6.6) 8 (11.0) 0.154
History of cerebrovascular disease 23 (2.3) 5 (6.8) 0.037
Smoking 0.916
Never 722 (74.0) 53 (72.6)
Past 174 (17.8) 13 (17.8)
Current 80 (8.2) 7 (9.6)
Alcohol consumption 0.621
None 444 (46.1) 36 (52.2)
Moderate 440 (45.7) 28 (40.6)
Heavy 79 (8.2) 5 (7.2)
Biochemical markers
Hemoglobin (mmol/L) 14.2 ± 1.4 13.4 ± 1.5 < 0.001
Serum creatinine (µmol/L) 0.8 ± 0.3 0.9 ± 1.3 0.324
eGFR (mL/min) 94.2 ± 19.2 89.1 ± 23.5 0.032
Fasting plasma glucose (mmol/L) 104.2 ± 31.3 115.2 ± 36.4 0.005
Total cholesterol (mmol/L) 190.7 ± 44.6 169.8 ± 44.7 < 0.001
HDL cholesterol (mmol/L) 53.4 ± 12.6 50.3 ± 11.4 0.057
LDL cholesterol (mmol/L) 116.5 ± 40.6 96.1 ± 39.0 < 0.001
Triglycerides (mmol/L) 145.3 ± 109.7 142.3 ± 111.3 0.825
Echocardiographic data
LV end-systolic diameter (mm) 38.3 ± 10.2 39.1 ± 10.4 0.539
LV end-diastolic diameter (mm) 37.8 ± 13.5 37.3 ± 9.8 0.737
LA volume index (mL/m2) 24.2 ± 4.8 32.3 ± 9.0 < 0.001
LV mass index (g/m2) 72.1 ± 15.2 84.0 ± 16.5 < 0.001
LV ejection fraction (%) 60.9 ± 2.4 60.5 ± 2.4 0.171
Mitral E velocity (cm/sec) 60.3 ± 16.4 66.2 ± 16.6 0.003
Mitral A velocity (cm/sec) 65.4 ± 16.8 66.2 ± 16.6 < 0.001
E/A ratio 1.0 ± 0.4 0.9 ± 0.7 0.013
Mitral E/e’ ratio 7.7 ± 2.0 13.6 ± 3.3 < 0.001
Septal E’ (cm/sec) 8.1 ± 2.3 5.3 ± 1.3 < 0.001
TR peak velocity (m/sec) 2.3 ± 0.2 2.6 ± 0.3 < 0.001
Body composition assessment
Body height (m) 164.0 ± 8.9 157.6 ± 9.5 < 0.001
Body weight (kg) 67.1 ± 13.1 63.7 ± 11.4 0.016
Body mass index (kg/m2) 24.8 ± 3.5 25.6 ± 3.6 0.091
Body surface area (m2) 1.7 ± 0.3 1.7 ± 0.2 0.002
Body fat composition
Body fat mass (kg) 20.3 ± 3.4 21.2 ± 5.8 0.194
Percentage body fat (%) 30.2 ± 6.9 33.4 ± 6.8 < 0.001
Fat-free mass (kg) 46.8 ± 10.1 42.4 ± 6.9 < 0.001
Fat-free mass index 17.2 ± 2.3 16.9 ± 2.1 0.171
Waist circumference (cm) 60.0 ± 42.9 71.8 ± 37.7 0.022
Waist-hip ratio 0.91 ± 0.20 1.02 ± 0.22 < 0.001
Visceral fat area (cm2) 114.7 ± 94.8 165.3 ± 106.4 < 0.001
SVR (kg/cm2) 0.4 ± 0.5 0.2 ± 0.1 < 0.001
Body muscle composition
Skeletal muscle mass (kg) 25.8 ± 6.1 22.9 ± 5.3 < 0.001
Skeletal muscle mass index 7.1 ± 1.1 6.7 ± 1.0 0.003
Segmental lean analysis (kg)
Right arm muscle mass 2.5 ± 0.7 2.3 ± 0.6 0.001
Left arm muscle mass 2.5 ± 0.7 2.2 ± 0.6 0.003
Trunk muscle mass 21.3 ± 4.5 19.4 ± 3.9 < 0.001
Right leg muscle mass 7.2 ± 1.7 6.2 ± 1.6 < 0.001
Left leg muscle mass 7.1 ± 1.7 6.2 ± 1.6 < 0.001
Arm muscle circumference (cm) 26.9 ± 2.8 26.3 ± 2.5 0.059

Values are presented as mean±standard deviation or number (%).

LVDD, left ventricular diastolic dysfunction; SBP, systolic blood pressure; DBP, diastolic blood pressure; eGFR, estimated glomerular filtration rate; HDL, high-density lipoprotein; LDL, low-density lipoprotein; LV, left ventricle; LA, left atrium; TR, tricuspid regurgitation; SVR, ratio of skeletal muscle mass to visceral fat area.

Univariate and multivariable logistic regression analyses for left ventricular diastolic dysfunction

Variable Univariate analysis Multivariable analysis
OR (95% CI) P Adjusted OR (95% CI) P
Age (yr) 1.13 (1.09–1.16) < 0.001 1.11 (1.07–1.15) < 0.001
Male sex 0.43 (0.26–0.72) 0.001 0.35 (0.11–1.40) 0.081
SBP (mmHg) 1.03 (1.01–1.04) 0.001 1.01 (0.99–1.02) 0.523
Hypertension 3.01 (1.85–4.91) < 0.001 1.34 (0.72–2.50) 0.351
Diabetes mellitus 2.83 (1.66–4.81) < 0.001 1.26 (0.60–2.66) 0.539
Dyslipidemia 1.93 (1.20–3.11) 0.007 0.98 (0.55–1.76) 0.946
Smoking 1.19 (0.52–2.71) 0.675 2.25 (0.77–6.62) 0.140
Alcohol consumption 0.78 (0.30–2.05) 0.615 2.69 (0.80–9.08) 0.111
Hemoglobin (mmol/L) 0.70 (0.59–0.82) < 0.001 0.81 (0.64–1.03) 0.090
eGFR (mL/min) 0.99 (0.97–0.99) 0.031 1.00 (0.99–1.02) 0.754
Fasting plasma glucose (mmol/L) 1.01 (1.00–1.01) 0.007 1.00 (0.99–1.01) 0.469
Total cholesterol (mmol/L) 0.99 (0.98–0.99) < 0.001 0.99 (0.99–1.00) 0.115
Body mass index (kg/m2) 1.06 (0.99–1.13) 0.084 0.97 (0.84–1.12) 0.659
Waist-hip ratio 8.20 (3.02–22.28) < 0.001 0.72 (0.09–5.85) 0.757
SVR (kg/cm2) 0.01 (0.01–0.04) < 0.001 0.08 (0.01–0.57) 0.027
Skeletal muscle mass index 0.72 (0.57–0.90) 0.004 1.36 (0.70–2.63) 0.364

Percentage body fat, visceral fat area, and arm muscle circumference were not included in the multivariable model due to multicollinearity with body mass index and skeletal muscle mass index.

OR, odds ratio; CI, confidence interval; SBP, systolic blood pressure; eGFR, estimated glomerular filtration rate; SVR, ratio of skeletal muscle mass to visceral fat area.

  1. Powell-Wiley TM, Poirier P, Burke LE, Despres JP, Gordon-Larsen P, Lavie CJ, et al. Obesity and cardiovascular disease: a scientific statement from the American Heart Association. Circulation 2021;143:e984-1010.
    Pubmed KoreaMed CrossRef
  2. Cercato C, Fonseca FA. Cardiovascular risk and obesity. Diabetol Metab Syndr 2019;11:74.
    Pubmed KoreaMed CrossRef
  3. Avelar E, Cloward TV, Walker JM, Farney RJ, Strong M, Pendleton RC, et al. Left ventricular hypertrophy in severe obesity: interactions among blood pressure, nocturnal hypoxemia, and body mass. Hypertension 2007;49:34-9.
    Pubmed CrossRef
  4. Iacobellis G, Ribaudo MC, Leto G, Zappaterreno A, Vecci E, Di Mario U, et al. Influence of excess fat on cardiac morphology and function: study in uncomplicated obesity. Obes Res 2002;10:767-73.
    Pubmed CrossRef
  5. Russo C, Jin Z, Homma S, Rundek T, Elkind MS, Sacco RL, et al. Effect of obesity and overweight on left ventricular diastolic function: a community-based study in an elderly cohort. J Am Coll Cardiol 2011;57:1368-74.
    Pubmed KoreaMed CrossRef
  6. Kloch-Badelek M, Kuznetsova T, Sakiewicz W, Tikhonoff V, Ryabikov A, González A, et al. Prevalence of left ventricular diastolic dysfunction in European populations based on cross-validated diagnostic thresholds. Cardiovasc Ultrasound 2012;10:10.
    Pubmed KoreaMed CrossRef
  7. Chadha DS, Gupta N, Goel K, Pandey RM, Kondal D, Ganjoo RK, et al. Impact of obesity on the left ventricular functions and morphology of healthy Asian Indians. Metab Syndr Relat Disord 2009;7:151-8.
    Pubmed CrossRef
  8. Wang YC, Liang CS, Gopal DM, Ayalon N, Donohue C, Santhanakrishnan R, et al. Preclinical systolic and diastolic dysfunctions in metabolically healthy and unhealthy obese individuals. Circ Heart Fail 2015;8:897-904.
    Pubmed KoreaMed CrossRef
  9. Kossaify A, Nicolas N. Impact of overweight and obesity on left ventricular diastolic function and value of tissue Doppler echocardiography. Clin Med Insights Cardiol 2013;7:43-50.
    Pubmed KoreaMed CrossRef
  10. van Heerebeek L, Franssen CP, Hamdani N, Verheugt FW, Somsen GA, Paulus WJ. Molecular and cellular basis for diastolic dysfunction. Curr Heart Fail Rep 2012;9:293-302.
    Pubmed CrossRef
  11. Saeedi Borujeni MJ, Esfandiary E, Taheripak G, Codoñer-Franch P, Alonso-Iglesias E, Mirzaei H. Molecular aspects of diabetes mellitus: resistin, microRNA, and exosome. J Cell Biochem 2018;119:1257-72.
    Pubmed CrossRef
  12. Heymsfield SB, Cefalu WT. Does body mass index adequately convey a patient's mortality risk?. JAMA 2013;309:87-8.
    Pubmed CrossRef
  13. Zamboni M, Mazzali G, Zoico E, Harris TB, Meigs JB, Di Francesco V, et al. Health consequences of obesity in the elderly: a review of four unresolved questions. Int J Obes (Lond) 2005;29:1011-29.
    Pubmed CrossRef
  14. Schrager MA, Metter EJ, Simonsick E, Ble A, Bandinelli S, Lauretani F, et al. Sarcopenic obesity and inflammation in the InCHIANTI study. J Appl Physiol (1985) 2007;102:919-25.
    Pubmed KoreaMed CrossRef
  15. Kohara K. Sarcopenic obesity in aging population: current status and future directions for research. Endocrine 2014;45:15-25.
    Pubmed CrossRef
  16. Kim TN, Park MS, Lim KI, Yang SJ, Yoo HJ, Kang HJ, et al. Skeletal muscle mass to visceral fat area ratio is associated with metabolic syndrome and arterial stiffness: the Korean Sarcopenic Obesity Study (KSOS). Diabetes Res Clin Pract 2011;93:285-91.
    Pubmed CrossRef
  17. 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.
    Pubmed CrossRef
  18. Nagueh SF, Smiseth OA, Appleton CP, Byrd BF 3rd, Dokainish H, Edvardsen T, et al. Recommendations for the evaluation of left ventricular diastolic function by echocardiography: an update from the American Society of Echocardiography and the European Association of Cardiovascular Imaging. J Am Soc Echocardiogr 2016;29:277-314.
    Pubmed CrossRef
  19. Kyle UG, Bosaeus I, De Lorenzo AD, Deurenberg P, Elia M, Gómez JM, et al. Bioelectrical impedance analysis: part I: review of principles and methods. Clin Nutr 2004;23:1226-43.
    Pubmed CrossRef
  20. Heymsfield SB, Smith R, Aulet M, Bensen B, Lichtman S, Wang J, et al. Appendicular skeletal muscle mass: measurement by dual-photon absorptiometry. Am J Clin Nutr 1990;52:214-8.
    Pubmed CrossRef
  21. Cruz-Jentoft AJ, Baeyens JP, Bauer JM, Boirie Y, Cederholm T, Landi F, et al. Sarcopenia: European consensus on definition and diagnosis: report of the European Working Group on Sarcopenia in Older People. Age Ageing 2010;39:412-23.
    Pubmed KoreaMed CrossRef
  22. Breiman L. Random forests. Mach Learn 2001;45:5-32.
    CrossRef
  23. Kursa MB, Rudnicki WR. Feature selection with the Boruta package. J Stat Softw 2010;36:1-13.
    CrossRef
  24. Pencina MJ, D'Agostino RB Sr, D'Agostino RB Jr, Vasan RS. Evaluating the added predictive ability of a new marker: from area under the ROC curve to reclassification and beyond. Stat Med 2008;27:157-72.
    Pubmed CrossRef
  25. Pencina MJ, D'Agostino RB Sr, Steyerberg EW. Extensions of net reclassification improvement calculations to measure usefulness of new biomarkers. Stat Med 2011;30:11-21.
    Pubmed KoreaMed CrossRef
  26. Caleyachetty R, Thomas GN, Toulis KA, Mohammed N, Gokhale KM, Balachandran K, et al. Metabolically healthy obese and incident cardiovascular disease events among 3.5 million men and women. J Am Coll Cardiol 2017;70:1429-37.
    Pubmed CrossRef
  27. Farajian P, Renti E, Manios Y. Obesity indices in relation to cardiovascular disease risk factors among young adult female students. Br J Nutr 2008;99:918-24.
    Pubmed CrossRef
  28. Neeland IJ, Ross R, Després JP, Matsuzawa Y, Yamashita S, Shai I, et al. Visceral and ectopic fat, atherosclerosis, and cardiometabolic disease: a position statement. Lancet Diabetes Endocrinol 2019;7:715-25.
    Pubmed CrossRef
  29. Kouli GM, Panagiotakos DB, Kyrou I, Georgousopoulou EN, Chrysohoou C, Tsigos C, et al. Visceral adiposity index and 10-year cardiovascular disease incidence: the ATTICA study. Nutr Metab Cardiovasc Dis 2017;27:881-9.
    Pubmed CrossRef
  30. Turkbey EB, McClelland RL, Kronmal RA, Burke GL, Bild DE, Tracy RP, et al. The impact of obesity on the left ventricle: the Multi-Ethnic Study of Atherosclerosis (MESA). JACC Cardiovasc Imaging 2010;3:266-74.
    Pubmed KoreaMed CrossRef
  31. Sawada N, Daimon M, Kawata T, Nakao T, Kimura K, Nakanishi K, et al. The significance of the effect of visceral adiposity on left ventricular diastolic function in the general population. Sci Rep 2019;9:4435.
    Pubmed KoreaMed CrossRef
  32. Canepa M, Strait JB, Milaneschi Y, AlGhatrif M, Ramachandran R, Makrogiannis S, et al. The relationship between visceral adiposity and left ventricular diastolic function: results from the Baltimore Longitudinal Study of Aging. Nutr Metab Cardiovasc Dis 2013;23:1263-70.
    Pubmed KoreaMed CrossRef
  33. Daniels SR, Kimball TR, Morrison JA, Khoury P, Witt S, Meyer RA. Effect of lean body mass, fat mass, blood pressure, and sexual maturation on left ventricular mass in children and adolescents: statistical, biological, and clinical significance. Circulation 1995;92:3249-54.
    Pubmed CrossRef
  34. Bella JN, Devereux RB, Roman MJ, O'Grady MJ, Welty TK, Lee ET, et al. Relations of left ventricular mass to fat-free and adipose body mass: the strong heart study. The Strong Heart Study Investigators. Circulation 1998;98:2538-44.
    Pubmed CrossRef
  35. de Simone G, Pasanisi F, Ferrara AL, Roman MJ, Lee ET, Contaldo F, et al. Relative fat-free mass deficiency and left ventricular adaptation to obesity: the Strong Heart Study. Int J Cardiol 2013;168:729-33.
    Pubmed KoreaMed CrossRef
  36. Ryo M, Kishida K, Nakamura T, Yoshizumi T, Funahashi T, Shimomura I. Clinical significance of visceral adiposity assessed by computed tomography: a Japanese perspective. World J Radiol 2014;6:409-16.
    Pubmed KoreaMed CrossRef
  37. Shuster A, Patlas M, Pinthus JH, Mourtzakis M. The clinical importance of visceral adiposity: a critical review of methods for visceral adipose tissue analysis. Br J Radiol 2012;85:1-10.
    Pubmed KoreaMed CrossRef
  38. Yamashina A, Tomiyama H, Takeda K, Tsuda H, Arai T, Hirose K, et al. Validity, reproducibility, and clinical significance of noninvasive brachial-ankle pulse wave velocity measurement. Hypertens Res 2002;25:359-64.
    Pubmed CrossRef