Ever since the significant role that a reduction in muscle mass and strength play in metabolic dysfunction was revealed,^1-3 sarcopenia has received attention as an important disease to treat. The co-existence of obesity and sarcopenia, two serious pathologic conditions, is called sarcopenic obesity (SO); the prevalence of SO increases with a sedentary lifestyle, a high calorie diet, and increased age.^4,5 SO is not only highly related to the prevalence of metabolic syndrome but also synergistically increases the risk of premature mortality compared with obesity or sarcopenia alone.^6-9 The presence of SO in patients with type 2 diabetes mellitus (T2DM) further increases the risk of metabolic disturbance. Previous studies have reported that T2DM patients with SO showed higher insulin resistance, a greater incidence of nonalcoholic fatty liver disease, and greater albuminuria than T2DM patients without SO.^10-12 This relationship emphasizes the importance of detecting and treating SO as early as possible and calls for the development of an adequate anthropometric index that can distinguish SO from simple obesity or sarcopenia, particularly in patients who already have metabolic risk factors such as T2DM.

Although body mass index (BMI) has been widely used because it is simple to calculate, it cannot distinguish abdominal/visceral fat and fat/muscle mass, which have different metabolic risks. Therefore, there has been a call for a new index to predict metabolic risk and mortality in elderly people, who often suffer from SO.^13,14 Waist circumference (WC), which accurately reflects abdominal obesity, has been verified to predict cardiovascular (CV) risk and CV mortality, but it shows limitations in reflecting low muscle mass.^13,15 Other indices have been suggested for measuring low muscle mass, including the circumference of the calf or mid upper arm,^16-18 but their low sensitivity and difficulty in measurement prevent them from being widely used.¹⁹

The weight-adjusted waist index (WWI) is a recently developed anthropometric index that standardizes the WC with the body weight, and it has shown effectiveness in predicting premature all-cause mortality, CV mortality, and T2DM.²⁰ Furthermore, in a previous Korean cohort study, the WWI demonstrated a remarkable association with high fat mass and low muscle mass.²¹ WC and WC-derived indices, the waist hip ratio (WHR) and waist height ratio (WHtR), have been analyzed for metabolic risk and mortality prediction as alternative body composition indices,^15,22 but they have never been evaluated as parameters of SO.

Therefore, we sought an appropriate anthropometric index to discriminate SO and predict SO-related metabolic risks in newly diagnosed T2DM patients by comparing several indices. We also compared the usefulness of each index across different age groups.

Study design and participants

We analyzed the Korea Guro Diabetes Program (KGDP) cohort, which contains data from newly diagnosed T2DM patients. T2DM was defined as a glycosylated hemoglobin (HbA1c) level of more than 6.5%. We recruited 515 participants from September 2014 to July 2021. All patients underwent blood and urine sample collection, anthropometric measurements, whole-body dual-energy X-ray absorptiometry (DEXA), and a diabetes mellitus complication study at their initial visit. Medical histories and lifestyle information were collected using a detailed questionnaire and personal interview. All participants submitted informed consent, and this study was approved by the Korea University Institutional Review Board (2014GR0140).

Anthropometric, laboratory, and body composition measurements

BMI was calculated as weight (kg)/height (m)², and WC was measured at the midpoint between the lower edge of the rib and the top of the iliac crest. The hip circumference (HC) was obtained at the widest portion of the buttocks when the patient was in the standing position, following World Health Organization guidelines.²³ All blood samples were obtained in the morning after a 12-hour overnight fast and were immediately stored at –80 °C for subsequent assays. The detailed method used to assess the laboratory parameters, including high-sensitivity C-reactive protein (hs-CRP), HbA1c, and albuminuria, was previously described in the KGDP cohort study.¹¹ The estimated glomerular filtration rate (eGFR) was calculated using the Modification of Diet in Renal Disease equation.²⁴ Each patient underwent whole-body DEXA using fan-beam technology (Hologic Discovery A; Hologic) to measure muscle and fat mass. The appendicular skeletal muscle (ASM, kg) mass is the sum of the lean muscle mass in a patient’s arms and legs, and ASM/height (m)² was calculated by dividing ASM with square of height to diagnose SO. The truncal fat amount (TFA, g) was measured by DEXA to quantify the fat mass. The WHR was defined as the WC (cm)/HC (cm), and the WHtR was calculated as the WC (cm) divided by the height (cm). The WWI was defined as the WC (cm)/√ weight (√kg).²⁰ As complication studies of T2DM, the carotid intima-media thickness (CIMT) was measured using high-resolution B-mode ultrasonography (EnVisor; Philips Healthcare) with a 5 to 12 MHz transducer 1 to 3 cm proximal to the carotid bifurcation.²⁵ The brachial-ankle pulse wave velocity (baPWV) was measured after a 5-minute rest using a BP-203RPE II volume plethysmographic apparatus (Colin) with the patient in a supine position. The means of the left and right baPWV and CIMT values were defined as the mean baPWV and CIMT, respectively.

Definition of sarcopenic obesity

In this study, the components of sarcopenia were measured and defined according to the 2019 consensus update of the Asian Working Group for Sarcopenia (AWGS)²⁶: low muscle mass plus low muscle strength and/or reduced physical performance. The cutoff value for low muscle mass was a DEXA-measured ASM/height² <7.0 kg/m² for men and <5.4 kg/m² for women.²⁶ Muscle strength was measured by using a dynamometer (Takei) to assess the handgrip strength of both hands with full elbow extension; the maximum value was obtained during two or three trials. Low muscle strength was defined as a handgrip <28.0 kg for men and <18.0 kg for women.²⁶ Physical performance was determined by the usual gait speed and regarded as low when the speed was <1.0 m/sec.²⁶ Obesity was defined as a WC ≥90 cm in men and ≥85 cm in women.²⁷ SO was diagnosed when both sarcopenia and obesity were present.

Statistical analysis

Each variable was assessed for a normal distribution. The data are expressed as the mean±standard deviation. Comparisons of subjects with and without SO were performed using Student t-tests and Mann-Whitney U-tests. Associations between the baseline characteristics of each participant and the body composition indices (WC, WHR, WHtR, and WWI) were estimated using a Spearman correlation analysis. Those correlations were subsequently analyzed by age group (age ≥65 years, age <65 years). We use a scatterplot to show the relationships between variables. Multivariate logistic regression analyses were performed to calculate the odds ratios (ORs) of SO based on tertiles of WC, WHR, WHtR, and WWI. Model 1 was adjusted for age and sex, and model 2 was further adjusted for smoking, alcohol use, and physical activity. Model 3 was further adjusted for the systolic blood pressure, HbA1c, and eGFR. Receiver operating characteristic (ROC) curve analyses were done to assess how well each index predicted SO, and pairwise comparisons of the area under the curve (AUC) for each index were performed using the DeLong test method. We used the Youden J-index to determine optimal cutoff values that maximize sensitivity and specificity. All statistical analyses were performed using SPSS software version 20.0 (SPSS Inc.), and statistical significance was regarded as a P<0.05.

The baseline characteristics of newly diagnosed T2DM patients with or without SO are shown in Table 1. Among the participants, 46 subjects were diagnosed with SO. The population with SO was older than that without SO (59.70 years vs. 53.63 years) and had higher WHR and WWI values than those without SO (WHR: 0.95 vs. 0.92; WWI: 11.01 vs. 10.73). In addition, SO patients had higher baPWV than subjects without SO (1,625.68 vs. 1,515.27).

Comparison of the correlations between WC, WHR, WHtR, and WWI and the body composition measurements

Correlation analyses between WC, WHR, WHtR, and WWI and the baseline characteristics of the study population are shown in Table 2. The WWI correlated positively with age (r=0.150, P=0.001), whereas WC, WHR, and WHtR were negatively associated with age. An increase in the WWI, WC, WHR, and WHtR correlated with increases in aspartate aminotransferase, hs-CRP, and homeostatic model assessment for insulin resistance levels. The relationships between the anthropometric parameters and the mass of fat and muscle are presented in Fig. 1. WC, WHR and WHtR showed significant positive correlations with both the TFA (WC: r=0.756, P<0.001; WHR: r=0.584, P<0.001; WHtR: r=0.797, P<0.001) and the ASM (WC: r=0.473, P<0.001; WHR: r=0.307, P<0.001; WHtR: r=0.094, P=0.034), suggesting that those indicators cannot appropriately reflect reduced muscle mass in obese patients. However, the WWI showed a remarkable positive correlation with the TFA (r=0.501, P<0.001) and a negative correlation with the ASM (r=–0.272, P<0.001). Therefore, this novel anthropometric index might be useful for detecting SO because it correlates with both a reduction of muscle mass and a growth of fat mass.

ORs of SO according to WC, WHR, WHtR, and WWI tertiles, the ROC curve analysis, and comparison of the AUC of each index for SO

Supplementary Table 1 shows the ORs of SO based on the increment of tertiles of WC, WHR, WHtR, and WWI. In the fully adjusted model, WWI showed the highest OR of SO among four indices (WWI [OR, 5.70; 95% confidence interval (CI), 2.04 to 15.94], WHR [OR, 4.40; 95% CI, 1.64 to 11.79], and WHtR [OR, 3.46; 95% CI, 1.20 to 9.98]). WC did not show statistical significance (OR, 1.46; 95% CI, 0.61 to 3.50; P=0.40). The ROC curves for WC, WHR, WHtR, and WWI in predicting SO are presented in Fig. 2. The AUC of the WWI was 0.684, which was the highest value among the indices; the AUC values of WC, WHR, and WHtR were 0.563, 0.656, and 0.584, respectively. The optimal cutoff value of the WWI for SO was 10.87 (sensitivity: 63%, specificity: 63%), whereas those of WC, WHtR, and WHR were 90.15, 0.55, and 0.93, respectively (sensitivity [WC, 61%; WHtR, 57%; WHR, 61%] and specificity [WC, 57%; WHtR, 58%; WHR, 62%]). The differences in the AUC values for each index in predicting SO are presented in Table 3. The WWI and WC showed a noteworthy difference of area (0.121; 95% CI, 0.037 to 0.204), as did the WWI and WHtR (0.100; 95% CI, 0.040 to 0.160), and WC and WHR (0.093; 95% CI, 0.044 to 0.143).

Comparison of WC, WHR, WHtR, and WWI in their correlation with metabolic parameters stratified by age group (<65 and ≥65 years)

Body composition changes were stratified according to age, so we performed a correlation analysis stratified by age in patients <65 and ≥65 years; the results are presented in Supplementary Table 2. Among them, the association between WC, WHR, WHtR, and WWI and albuminuria and mean baPWV are illustrated in Fig. 3. The WWI correlated significantly with albuminuria in patients 65 years and older (r=0.215, P=0.031), but no association was observed in the group aged <65 years (r=0.031, P=0.543). Meanwhile, WC showed no relationship with albuminuria in either age group, and the WHR showed a positive relationship with albuminuria in patients younger than 65 years (r=0.106, P=0.037) but no association in patients aged ≥65 years (r=0.142, P=0.158). Furthermore, although the WHR and WHtR had no significant correlation with the mean baPWV in either age group and WC was associated with the mean baPWV only in the younger group, the WWI showed a significant association with mean baPWV in the elderly group (r=0.260, P=0.009) but not in the younger group (r=0.042, P=0.409). The WWI also showed an age-dependent differential correlation with low density lipoprotein cholesterol (age ≥65 years: r=0.203, P=0.042; age <65 years: r=0.026, P=0.600) and triglycerides (age ≥65 years: r=0.281, P=0.004; age <65 years: r=–0.077, P=0.126), whereas WC, WHR, and WHtR revealed no differential relationship with those factors according to the age groups . Furthermore, the WWI showed an inverse relationship with the mean and max handgrip strength in the group aged ≥65 years. These results indicate the potential of the WWI as a proper index to reflect cardiometabolic risk and loss of muscle function, especially in elderly people.

In this study, we have examined the suitability of the WWI as an index to discriminate SO and other accompanying metabolic disturbances in newly diagnosed T2DM patients. Elderly patients who are undergoing age-related body composition changes (including high visceral fat and low muscle mass) are more desirable group than younger patients for applying the WWI to evaluate body composition and metabolic risks. The WWI showed a negative correlation with muscle mass, in contrast to WC, WHR, and WHtR, which exhibited a positive relationship with both fat mass and muscle mass. In addition, the WWI predicts SO better than the other three indicators, with an approximately five times higher OR in the highest tertile of the WWI compared with the lowest. The highest AUC in the ROC analysis for SO was shown in the WWI, which shows its statistical superiority to the AUC of the WHtR and WC. Furthermore, the WWI exhibited an obvious relationship with various cardiometabolic risk factors, such as arterial stiffness and albuminuria, especially in elderly people, a vulnerable group at higher risk than younger people for the progression of SO.

Previous studies have sought an index to reflect a metabolically hazardous fat distribution, which includes mostly visceral fat, and overcome the limitations of the BMI. WC and various WC-derived measures (such as the WHR and WHtR) were previously inspected for their ability to predict metabolic disease and mortality.^15,28-30 As the role of muscle composition has been increasingly emphasized, demand has also risen for an index that can distinguish between muscle and fat mass. Although computed tomography (CT), magnetic resonance imaging, and DEXA are the most accurate methods for measuring muscle mass, their high cost and radiation exposure limit their usage, which leads to an underdiagnosis of sarcopenia. A simple, cost-effective, and reliable anthropometric index is needed to screen for and diagnose sarcopenia. In 2019, the AWGS suggested calf circumference (CC) or the Simple Five-Item (SARC-F) questionnaire as screening tools for sarcopenia in community health care settings.^26,31,32 Despite the decent correlation between CC and low muscle mass, that index cannot reflect the increments in visceral fat percentage that are usually present in elderly patients with sarcopenia. In addition, the SARC-F questionnaire depends on self-reporting, which reduces its objectivity and introduces potential bias.

The WWI was developed by Park et al.²⁰ and is distinguished by its association with both low muscle mass and high fat, in contrast to other indices, which reflect only one or the other. In a previous study based on 602 subjects from the Ansan Geriatric cohort, the WWI also correlated positively with the total and visceral fat areas measured by CT, and it had a negative relationship with the ASM and ASM/height².²¹ However, prior studies did not test the power of the WWI to discriminate low muscle function (muscle strength and physical performance) due to limited data. Therefore, the ability of the WWI to identify sarcopenia, which involves the loss of both muscle mass and function, had not been verified. As Stenholm et al.⁴ underlined, a decline in muscle strength and function is more important to the pathologic outcomes of sarcopenia than the simple reduction in muscle mass. In this research, we evaluated the strength of various indices to identify SO, which defined using muscle mass, strength, and physical performance data according to the 2019 AWGS definition. We defined obesity by WC instead of BMI because the predictability of BMI for body composition and obesity-related mortality is reduced in elderly participants, who comprise most SO patients.^6,33 In addition, WC is closely related to metabolic risk in Asian people who accumulate intra-abdominal fat predominantly but have a relatively low BMI.^27,34 In this study, we also verified the relationship between the WWI and metabolic disturbances, represented by albuminuria and the mean baPWV, in addition to finding that the WWI correlates with T2DM, CV-related risk, and mortality. Albuminuria is an indicator of systemic inflammation and insulin resistance and a risk factor for chronic kidney disease (CKD), hypertension (HTN), and CV disease.^35,36 Arterial stiffness, as measured by the pulse wave velocity, reflects an increased risk of coronary heart disease, CKD, and HTN.^37,38 The WWI revealed a noticeably better relationship with albuminuria and mean baPWV than WC, the WHR, and the WHtR, especially in people ≥65 years of age, which suggests that the WWI is suitable for predicting metabolic risk in elderly people.

Recent studies have reported that T2DM patients, who already have a high metabolic risk, face further metabolic hazard when they are also diagnosed with SO. Kim et al.¹¹ demonstrated that insulin resistance is increased in T2DM patients with SO compared to those with simple obesity or low muscle mass. In previous studies, high incidences of albuminuria or CV disease were reported in T2DM patients with SO.^10,39 Therefore, applying the WWI to screen for SO in T2DM patients is important for preventing severe metabolic dysfunction and diabetic complications.

The predominant mechanism of sarcopenia is understood to be age-related sex hormone decline, and it is also characterized by mitochondrial dysfunction and an increase in protein proteolysis.³⁸ Leptin and proinflammatory cytokines secreted from adipocytes also play important roles that lead to muscle catabolism.⁶ A reduction in muscle mass and function promotes physical inactivity, and the disuse of muscles sequentially causes motor neuron loss. Therefore, identifying sarcopenia alongside obesity using a simple and practical method and offering early interventions could prevent a vicious cycle, particularly in the elderly population. The WWI is easy to calculate and can predict sarcopenia and obesity together simply by measuring a person’s WC and weight. We expect this marker to encourage clinicians to assess their patients for SO. A possible mechanism for the SO-predicting power of the WWI can be found in the process of formula development. Park et al.²⁰, who developed the WWI formula, removed the effect of weight on WC with a regression method, in the same way that the BMI was developed by removing the effect of height on weight. The BMI estimates that a person with a specific weight and short stature has higher risks than a person with the same weight and tall stature, and the WWI similarly assesses that a person with the certain WC and lower weight has more risks than a person with the same WC and more weight. For this reason, the WWI might well reflect the risk of abdominal obesity with low weight, which represents SO.

This study has several limitations. First, it was a cross-sectional study, so the temporal effects of the body composition measures in each index on metabolic disturbance were difficult to evaluate. Second, it was conducted in a relatively small cohort of Asian people from a single center. Larger longitudinal studies of people from various ethnicities are needed to estimate the causal relationships and effects across age and ethnic groups. Nonetheless, to the best of our knowledge, this is the first study to compare the use of anthropometric measures to identify SO, defined by muscle mass, strength, and physical performance. In addition, we estimated the correlation of each index with metabolic risks and then analyzed those findings by age group. The BMI has limitations, particularly in elderly patients who are diagnosed with both SO and SO-related metabolic disturbances; therefore, the WWI’s close association with the metabolic risk of elderly patients suggests that it can probably serve as a marker of SO and overcome the limitations of the BMI.

In conclusion, we have demonstrated the strength of the WWI as a parameter for SO and SO-related metabolic disturbances in newly diagnosed T2DM patients. The WWI might well predict metabolic dysfunction associated with SO, especially in an elderly population. Large-scale prospective studies are needed to verify the WWI as a discriminator of SO and to determine its clinical implications for fine risk stratification that could be used to predict cardiometabolic diseases in different age groups.

Supplementary materials can be found online at https://doi.org/10.7570/jomes23005.

This research was supported by a Korea Medical Device Development Fund grant funded by the Korean government (Ministry of Science and ICT, Ministry of Trade, Industry and Energy, Ministry of Health & Welfare, Ministry of Food and Drug Safety) (Project Number: 9991007469, KMDF_PR_20200901_0233) and by the National Research Foundation of Korea funded by the Ministry of Education of Korea (2021R1A2C2008792).

Study concept and design: MJP, NHK, SGK, and HJY; acquisition of data: KMC and HJY; analysis and interpretation of data: MJP, SYH, and HJY; drafting of the manuscript: MJP, SYH, and HJY; critical revision of the manuscript: KMC and SHB; statistical analysis: MJP and SYH; obtained funding: HJY; administrative, technical, or material support: NHK, SGK, KMC, SHB, and HJY; and study supervision: KMC, SHB, and HJY.