Skip to main content
Download PDF

Developing a risk model for early diagnosis of metabolic syndrome in Chinese adults aged 40 years and above based on BMI/HDL-C: a cross-sectional study

BMC Endocrine Disorders volume 24, Article number: 223 (2024) Cite this article

Abstract

Background

This study aimed to compare the diagnostic accuracy of four indicators, including waist-to-height ratio (WHTR), vascular adiposity index (VAI), TG/HDL-C, and BMI/HDL-C for metabolic syndrome (MS) in Chinese adults aged 40 years and above. Additionally, the study aimed to develop an efficient diagnostic model displayed by a nomogram based on individual’s BMI and circulating HDL-C level.

Methods

A cross-sectional study was conducted on 699 participants aged 40 years and above. Quartiles of BMI/HDL-C, TG/HDL-C, VAI, and WHTR were used as independent variables, and metabolic syndrome was used as the dependent variable. Logistic regression was conducted to explore the impact of each parameter on the risk of MS. The areas under the receiver operating characteristics were compared to determine the accuracy of the indicators in diagnosing MS in the participants. Logistic regression was run to construct the nomograms, and the performance of the nomogram was assessed by a calibration curve.

Results

MS subjects had higher levels of BMI, BFM, PBF, VFA, AMC, WC, SCR, TG, and insulin, but lower LDH and HDL-C levels than the subjects without MS. The BMI/HDL-C ratio was positively correlated with the prevalence of MS and its components. The final diagnostic model included five variables: gender, BFM, WC, TG, and BMI/HDL-C. The model showed good calibration and discrimination power with an AUC of 0.780. The cut-off value for the nomogram was 0.623 for diagnosing MS.

Conclusions

BMI/HDL-C ratio was an independent risk factor for MS in Chinese adults. BMI/HDL-C was significantly correlated with MS and its components. BMI/HDL-C was the most powerful diagnostic indicator compared to other indicators, including TG/HDL-C, VAI and WHTR for diagnosing MS. The nomogram drawn based on the diagnostic model provided a practical tool for diagnosing MS in Chinese adults.

Peer Review reports

Background

Metabolic syndrome (MS) is composed of a series of common metabolic disorders, including obesity, hyperglycemia, hypertension, and disturbance of lipid metabolism [1]. Currently, it is estimated that a quarter of the global population has MS, which affects over a billion people [2,3,4]. In 2017, the reported prevalence of MS in China was approximately 31.3% [5]. MS causes varying degrees of damage to the cardiovascular, digestive, and endocrine systems of the body, leading to a higher mortality rate among patients compared to those without MS [6]. Because of the severe impact of MS on health, early diagnosis and prediction of MS have become especially crucial. The traditional diagnosis of MS mainly relies on clinical physical examination and biochemical testing, which makes it difficult to screen sensitively and accurately diagnose in the community residents [7]. Meanwhile, the clinical community has not widely adopted emerging early diagnostic methods such as metabolomics and mass spectrometry [8]. Therefore, early screening, detection, and health management to prevent the progression of MS is becoming an essential public health concern.

Many blood biochemical and anthropometry indicators have been used to predict MS for early diagnosis and prevention [9]. Besides, the emerging discipline developed in the last century, metabolomics, was also used for diagnosing MS [10, 11]. Abnormal glucose and lipid metabolism are the typical characteristics of MS [12]. Moreover, insulin resistance was closely related to the pathology of MS. Homeostasis model assessment, including insulin resistance (HOMA-IR) [13], triglyceride and glucose index (TyG) [14], was commonly used to predict MS [15]. The World Health Organization (WHO) also proposed that MS needs to focus on insulin resistance and hyperglycemia [16]. Additionally, the TG/HDL-C index was also considered a classic diagnostic of MS, which fully considered the lipid metabolites including triglycerides and high-density lipoprotein cholesterol (HDL-C) [17].

Anthropometry indicators were widely used to diagnose MS and assess the prevalence of risk factors for their convenient access, and body mass index (BMI) and waist circumference (WC) are extensively used for the diagnosis of MS. However, the inaccuracy of measurement reduced the prediction efficiency as individuals with higher muscle mass were also classified as overweight or obese according to the cutoff value of BMI and WC. Studies had confirmed a positive association existed between neck circumference and the risk of MS, and neck circumference was considered as a surrogate parameter for BMI and WC in diagnosing MS [18, 19]. Besides, increasing studies try to explore the indicators that combine multiple factors to enhance diagnostic stability, such as Waist-to-Height Ratio (WHTR) and Visceral Adiposity Index (VAI) [20], etc. WHTR was considered to have better diagnostic power on cardiovascular metabolism-related risk factors than waist circumference and BMI [21]. The VAI index, an effective indicator reflecting visceral fat accumulation, was reported to be associated with various components of MS [22]. Although BMI is a useful indicator for evaluating health status due to its simplicity, it cannot directly reflect the degree of metabolic disorder [23].

Due to the complex components of metabolic syndrome, it is necessary to construct a reliable diagnostic model using routine clinical indicators. To explore an indicator with better diagnostic performance for MS, our study proposed a new diagnostic indicator that combines body composition with blood biochemistry (BMI/HDL-C). Moreover, different from previous studies that only constructed indicators, our study also evaluated the diagnostic power with conventional diagnostic indicators including TG/HDL-C, WHTR, and VAI.

Materials and methods

Participants

The adults aged 40 years and above who accepted health examination were collected at Suzhou Science and Technology City Hospital and Dong Zhu Health Service Center in 2020. A total of 699 participants, including 215 males and 484 females, were recruited. Three participants were included in the baseline data analysis but were excluded from the subsequent studies for lacking blood biochemical indicators. The study protocol was approved by the Committee on Medical Ethics of Capital Medical University (No. 2012SY23), and the study procedures followed the ethical standards of the Helsinki Declaration of 1975. Informed consent was signed by all participants.

The MS was defined according to the Clinical guidelines for prevention and treatment of type 2 diabetes mellitus in China (2022 version) [24]. The MS defined by Chinese Diabetes Society (CDS) guidelines consists of three or more of the following: (1) Abdominal Obesity: Abdominal obesity was defined as WC ≥ 90 cm for males, WC ≥ 85 cm for females. (2) Hyperglycemia: FPG ≥ 6.1 mmol/L or OGTT ≥ 7.8 mmol/L or those who have been diagnosed with diabetes and treated. (3) Hypertension: Blood pressure ≥ 130/85 mmHg (1 mmHg = 0.133 kPa) or those who have been diagnosed with hypertension and treated. (4) Hypertriglyceridemia: Triglycerides level of ≥ 1.70 mmol/L was used as the cut-off level. (5) Hypo-high-density lipoproteinemia (HHDL): HDL-Cholesterol level of < 1.04 mmol/L was used as the cut-off value.

Data collection

All data were collected by trained medical staff according to standard operating procedures. A questionnaire survey (see Additional file 1) was conducted at the beginning of the examination to collect basic demographic information, disease history (past medical history, medication history, and medication compliance), family disease history, and personal lifestyle. Physical examination and biochemical testing were conducted. All the information was input into a database, and strict quality control was carried out.

Physical examination and blood sample collection

Anthropometric measurement was conducted according to the guidelines of the WHO [25]. The height was measured by a standard height gauge (Seca). The weight-machine was required to calibrate the scale daily (Seca). The waist circumference was measured using a plastic tape measure, taken at the midpoint between the lower margin of the last palpable rib and the top of the iliac crest. The body composition, including body fat mass (BFM), body fat percentage (BFP), and visceral fat area (VFA), was measured by the Body Composition Analyzer (INBODY S10). The upper arm of the subject was raised at an angle of about 45 degrees, the palm of the hand was upward and clenched the fist and bent the elbow forcefully, then the measurer stood on the side and measured the muscle circumference of the upper arm by circling the tape around the thickest part of the biceps of the upper arm to obtain the value of arm muscle circumference (AMC). A sphygmomanometer was used to measure blood pressure twice in a sitting position.

Subjects were required to fast for at least 12 h. Venous blood samples (5 ml) were collected from 6:30 − 10:00 a.m. After centrifugation (480 g, 20 min), the plasma was separated. Plasma triglyceride (TG), creatinine (SCR) and lactate dehydrogenase (LDH) were measured by ILAB600 clinical chemistry analyzer (Instrumentation Laboratory, Lexington, WI, USA). Plasma HDL cholesterol (HDL-C) was measured using a commercially available assay from the Instrumentation Laboratory (Lexington, WI, United States). Low-density lipoprotein cholesterol (LDL-C) was calculated based on the Friedewald formula (FF) [26]. The patient took 75 g of glucose orally after fasting blood collection and drew venous blood two hours after meals to measure insulin levels. Plasma insulin concentrations were measured according to the manufacturer’s instructions (ELISA; Mercodia Ltd).

Body mass index (BMI) was classified as normal 18.5–23.9 kg/m2, overweight 24–27.9 kg/m2 and obese ≥ 28 kg/m2 [27]. Educational level was defined as illiteracy or primary school level and secondary school or above. Smoking status was categorized as non-smoker and smoker (including ex-smoker and current smoker).

Composite indicators are calculated according to the formula. TG/HDL-C: triglyceride to high-density lipoprotein cholesterol ratio. WHTR: waist-to-height ratio. VAI: Visceral Adiposity Index, was calculated as:

$$\:VAI\:\left(males\right)=\frac{WC}{39.68+\left(1.88\times\:BMI\right)}\times\:\frac{TG}{1.03}\times\:\frac{1.31}{HDL-c}$$
$$\:VAI\:\left(females\right)=\frac{WC}{36.58+(1.89\times\:BMI)}\times\:\frac{TG}{0.81}\times\:\frac{1.52}{HDL-c}$$

WC was measured in centimeters, BMI in kg/m2, TG and HDL-C in mmol/L.

Statistical analysis

SPSS 26.0 and R 4.2.2 were used for statistical analysis. Measurement data were expressed as M (SD). If the data conformed to the normal distribution, the t-test was used to compare the differences between groups; otherwise, the rank sum test was applied to compare the differences between groups. The categorical data were expressed in numbers and proportions, and the χ2 test was used to compare differences between groups. The Cramer’s V coefficient was used to evaluate the correlation strength when there was statistical significance in the χ2 test, and 0.1 < Cramer’s V < 0.3 indicates a weak intensity correlation; 0.3 ≤ Cramer’s V < 0.5 indicates a moderate intensity correlation; Cramer’s V ≥ 0.5 indicates a high intensity correlation. Cramer’s V was represented by the pound sign (#). Quartiles of BMI/HDL-C, TG/HDL-C, VAI, and WHTR were used as independent variables, and metabolic syndrome was used as the dependent variable. Logistic regression was conducted to explore the impact of each parameter on the risk of MS and to construct diagnostic models for MS. After adjusting for confounding factors, five variables were included in the final model, including gender, BFM, WC, TG, and BMI/HDL-C, and the OR value and 95% CI of MS risk were analyzed. The calibration curve and the receiver operating characteristic curve (ROC) of the prediction model were plotted to evaluate its diagnostic value. P < 0.05 was considered statistically significant.

Results

Demographic characteristics of the participants

The demographic characteristics and plasma parameters of the participants were shown in Tables 1 and 2. The percentage of females was higher in the MS group than that in the non-MS group (P < 0.05). The subjects with MS had higher levels of BMI, BFM, PBF, VFA, AMC, WC, SCR, TG, and insulin, but lower LDH and HDL-C levels than the subjects without MS (P < 0.05). The subjects with MS had a lower education level than the subjects without MS (P < 0.05). The BMI/HDL-C, TG/HDL-C, WHTR, and VAI levels in the MS group are higher than those in the non-MS group (P < 0.001).

Table 1 Demographic characteristics and body composition parameters of the participants
Full size table
Table 2 Comparison of plasma parameters between MS and non-MS subjects
Full size table

Correlation between MS and quartiles of BMI/HDL-C, TG/HDL-C, WHTR, and VAI

The prevalence of MS and its components were compared according to the quartiles of BMI/HDL-C, TG/HDL-C, WHTR, and VAI, and the Cramer’s V index was used to evaluate the correlation (Table 3; Fig. 1).

After grouping the participants according to the quartile of BMI/HDL-C ratio, the results revealed significant differences in the prevalence of MS, abdominal obesity, hyperglycemia, hypertension, hypertriglyceridemia, and HHDL (P < 0.001). The prevalence of MS, abdominal obesity, hyperglycemia, hypertension, hypertriglyceridemia, and HHDL increases in proportion to the BMI/HDL-C ratio. The quartile of BMI/HDL-C is weakly correlated with the prevalence of abdominal obesity, hyperglycemia, and hypertension, and moderately correlated with the prevalence of MS and hypertriglyceridemia, and strongly correlated with the prevalence of HHDL (P < 0.05, Cramer’s V MS = 0.414, Cramer’s V abdominal obesity = 0.270, Cramer’s V hyperglycemia = 0.134, Cramer’s V hypertension = 0.160, Cramer’s V hypertriglyceridemia = 0.301, Cramer’s V HHDL = 0.630) (Table 3; Fig. 1).

The prevalence of MS, hypertriglyceridemia, and HHDL varies based on TG/HDL-C ratio quartiles (P < 0.001). The quartile of TG/HDL-C is moderately correlated with the prevalence of MS and HHDL, but strongly linked with the prevalence of hypertriglyceridemia, and the relationship is statistically significant (P < 0.05, Cramer’s V MS = 0.358, Cramer’s V hypertriglyceridemia = 0.789, Cramer’s V HHDL = 0.452) (Table 3; Fig. 1).

The prevalence of MS, abdominal obesity, and hyperglycemia varies according to the quartiles of WHTR, and a positive correlation between WHTR and the prevalence of MS, abdominal obesity, hyperglycemia, hypertension, hypertriglyceridemia, and HHDL was observed (P < 0.05). The WHTR quartile is weakly correlated with the prevalence of hyperglycemia, moderately correlated with the prevalence of MS, and strongly correlated with the prevalence of abdominal obesity (P < 0.05, Cramer’s V MS = 0.328, Cramer’s V abdominal obesity = 0.811, Cramer’s V hyperglycemia = 0.108) (Table 3; Fig. 1).

Based on the quartile of VAI, there are variations in the prevalence of MS, hypertriglyceridemia, and HHDL (P < 0.001). The prevalence of MS and its components increases correspondingly with the increase of VAI. The quartile of VAI is moderately connected with the prevalence of MS and HHDL but strongly correlated with the prevalence of hypertriglyceridemia, and the relationship between parameters was statistically significant (P < 0.05, Cramer’s V MS = 0.395, Cramer’s V hypertriglyceridemia = 0.739, Cramer’s V HHDL = 0.431) (Table 3; Fig. 1).

Table 3 Prevalence of MS and its components according to the quartiles of BMI/HDL-C, TG/HDL-C, WHTR and VAI
Full size table
Fig. 1
figure 1

Prevalence of MS and its components according to the quartiles of BMI/HDL-C, TG/HDL-C, WHTR, and VAI. (A) Comparison of the prevalence of MS in the quartile of four key diagnostic indicators. (B) Comparison of the prevalence of abdominal obesity in the quartile of four key diagnostic indicators. (C) Comparison of the prevalence of hyperglycemia in the quartile of four key diagnostic indicators. (D) Comparison of the prevalence of hypertension in the quartile of four key diagnostic indicators. (E) Comparison of the prevalence of hypertriglyceridemia in the quartile of four key diagnostic indicators. (F) Comparison of the prevalence of HHDL in the quartile of four key diagnostic indicators. BMI, Body Mass Index; HDL-C, high-density lipoprotein cholesterol; TG, triglyceride; WHTR: waist-to-height ratio. VAI: visceral adiposity index. (*P < 0.05, **P < 0.01, ***P < 0.001. a: P < 0.05 vs. Q1 of the same indicator, b: P < 0.05 vs. Q2 of the same indicator, c: P < 0.05 vs. Q3 of the same indicator. #: 0.1 < Cramer’s V < 0.3, indicates a weak intensity correlation; ##: 0.3 ≤ Cramer’s V < 0.5, indicates a moderate intensity correlation; ###: Cramer’s V ≥ 0.5, indicates a high intensity correlation)

Full size image

Comparison of diagnostic models

Logistic regression models were established based on the quartiles of BMI/HDL-C, TG/HDL-C, VAI, and WHTR. Then, the odds ratio (OR) values under different adjustment conditions were compared. After adjusting for confounding factors, the subjects with Q3 (PBMI/HDL−C= 0.001, ORBMI/HDL−C = 3.077, 95% CIBMI/HDL−C: 1.575–6.010; PWHTR = 0.006, ORWHTR = 4.171, 95%CIWHTR: 1.505–11.556) level of BMI/HDL-C and WHTR showed a higher risk of MS than the subjects in the Q1 group, and Q4 (PBMI/HDL−C < 0.001, ORBMI/HDL−C = 4.629, 95% CIBMI/HDL−C: 1.982–10.811) level of BMI/HDL-C continuously showed a higher risk of MS than the subjects in the Q1 group, while the subjects with Q2 to Q4 (Q2, PTG/HDL−C < 0.001, ORTG/HDL−C = 3.063, 95% CITG/HDL−C: 1.662–5.643, PVAI = 0.010, ORVAI = 2.189, 95% CIVAI: 1.202–3.987; Q3, PTG/HDL−C < 0.001, ORTG/HDL−C = 11.884, 95% CITG/HDL−C: 5.968–23.665, PVAI < 0.001, ORVAI = 15.914, 95% CIVAI: 8.097–31.280; Q4, PTG/HDL−C < 0.001, ORTG/HDL−C = 75.468, 95% CITG/HDL−C: 28.850–197.413, PVAI< 0.001, ORVAI = 80.318, 95% CIVAI: 29.840–216.182) level of TG/HDL-C and VAI consistently have a higher risk of MS than the subjects in the Q1 group (Fig. 2. A-D, Table S1).

The ROC curves of models under different adjustment conditions were depicted (Fig. 2. E - G) and the diagnostic value of the models was compared. In model 1, no covariates were corrected (Fig. 2. E). In model 2, we further adjusted for gender, education level, and smoking (Fig. 2. F). In model 3, BMI, BFM, PBF, VFA, AMC, WC, SCR, LDH, TG, HDL-C, and insulin were adjusted (Fig. 2. G). We also found that the AUC of the ROC curves of model 3 (AUC = 0.780, 95% CI: 0.746–0.814) constructed based on BMI/HDL-C was higher than the AUC of the ROC curves of model 1and 2 (AUC Model 1 = 0.731, 95% CI: 0.693–0.768; AUC Model 2 = 0.752, 95% CI: 0.716–0.788) (Table 4).

Fig. 2
figure 2

Forest plot and ROC curves of the models constructed based on BMI/HDL-C, TG/HDL-C, VAI, and WHTR. Correlation between quartile of BMI/HDL-C (A), TG/HDL-C (B), VAI (C), and WHTR (D) and MS with adjustment of different confounding factors. Different background of forest plot represents different confounding factor correction: Blue - without adjustment of confounding factors; red - adjustment of confounding factors including gender, education level, and smoking; yellow - adjustment of confounding factors including BMI, BFM, PBF, VFA, AMC, WC, SCR, LDH, TG, HDL-C, and insulin. Comparison of ROC curves (E) without adjustment of confounding factors; (F) with adjustment of confounding factors including gender, education level, and smoking; (G) with adjustment of confounding factors including BMI, BFM, PBF, VFA, AMC, WC, SCR, LDH, TG, HDL-C, and insulin. BMI, Body Mass Index; HDL-C, high-density lipoprotein cholesterol; TG, triglyceride; WHTR: waist-to-height ratio. VAI: visceral adiposity index

Full size image
Table 4 The sensitivity, specificity, AUC and cutoff value of different models
Full size table

The sensitivity, specificity, AUC, and the cutoff values of the prediction models were shown in Table 4. The sensitivity and specificity of the model are both higher than 60%. The model 3 based on BMI/HDL-C and Model 3 based on VAI have the highest AUC values compared to the AUC values of other models, which are 0.780 (95% CI: 0.746, 0.814) and 0.780 (95% CI: 0.745, 0.814).

The optimal model

Considering both simplicity (easy to calculate) and accuracy (the AUC value of the ROC curve), and its correlation with MS, model 3 based on BMI/HDL-C was selected as the optimal model to predict metabolic syndrome.

Beginning with the risk factors of MS (Table 1), results of logistic regression analysis showed that gender, BFM, WC, TG, and BMI/HDL-C remained important predictors after adjusting for confounding factors. Then a nomogram (Fig. 3A) was built to predict the risk of metabolic syndrome. To use the nomogram, clinicians could position on the variable axis based on the patients’ situation, and extend a line upwards to determine the points of each variable based on the topmost axis. Then add the points obtained from each variable, extend a line downwards from the total point axis to the risk axis to determine the risk of metabolic syndrome at the lower line of the nomogram. The cut-off value of BMI/HDL-C for the nomogram was 0.623 for diagnosing metabolic syndrome.

Fig. 3
figure 3

Nomogram and calibration curve of the optimal model. (A) Nomogram of the optimal prediction model of metabolic syndrome. (B) Calibration curve of the optimal prediction model of metabolic syndrome. BFM, Body Fat Mass; WC, Waist Circumference; TG, triglyceride; Q1, Q3, Q4, quartile of BMI/HDL-C

Full size image

We plot the calibration curve (Fig. 3B) and ROC curve (Fig. 2G) to evaluate the diagnostic efficiency of the models. Sensitivity, specificity, the AUC, and cut-off value of different models were used to analyze the ROC curve, and we found that the AUC value of the model is 0.780 (0.746, 0.814), indicating a powerful prediction ability.

Model equation

The equation for the developed model is provided below:

$$\eqalign{& Logit = - 16.994 - 1.780 \times Gender\left( {male} \right) \cr & - 0.103 \times BFM + 0.195 \times WC\left( {cm} \right) + 1.763 \cr & \times TG{\rm{(}}mmol/L{\rm{)}} + 1.124 \times Q3 + 1.532 \times Q4 \cr}$$

Discussion

In this cross-sectional study, we investigated the correlation between TG/HDL-C, WHTR, VAI, and BMI/HDL-C ratios and the prevalence of MS. Our data indicated that higher level of TG/HDL-C, WHTR, VAI, and BMI/HDL-C ratios were positively associated with MS, abdominal obesity, hyperglycemia, hypertension, hypertriglyceridemia, and HHDL in adults aged 40 to 85. BMI, BFM, PBF, VFA, AMC, WC, SCR, TG, insulin, LDH, and HDL-C varied between subjects with MS and subjects without MS. The prevalence of MS, abdominal obesity, hyperglycemia, hypertension, hypertriglyceridemia, and HHDL was compared according to the quartiles of BMI/HDL-C, TG/HDL-C, VAI, and WHTR. We found that, in comparison with classical indicators including TG/HDL-C, VAI, and WHTR, BMI/HDL-C showed a significantly stronger correlation with MS. Based on the BMI/HDL-C quartile, we proposed a new model to predict MS, which includes five factors: gender, BFM, WC, TG, and BMI/HDL-C ratio. The nomogram visually displayed the model.

Previous prospective and cross-sectional studies have reported that patients with metabolic syndrome (MS) exhibited an increased level of BMI, BFM, PBF, VFA, AMC, WC, SCR, TG, and insulin as comparing with non-MS subjects (P < 0.05) [28, 29]. These findings align with our results. Body fat mass commonly reflects the content of visceral fat and is a widely accepted parameter for evaluating visceral obesity. A study found that the increase in BFM leads to an increased risk of MS in adults [30]. Subjects who had accumulated visceral fat were at high risk for MS, which had a cumulative effect on the development of obesity [31]. Serum creatinine level in Chinese adults showed a positive correlation with the risk of MS. Even for individuals with SCR levels within the normal range, a high SCR level implies a high risk of MS [32]. A previous study found that the levels of total LDH, LDH1,2,4, and 5 in obese and diabetes patients were lower than in normal control subjects [33]. In our study, we also found that the level of total LDH in the MS patients was lower than in the non-MS subjects (mean LDH 186.3 [SD 36.1] vs. 194.3 [50.2]; P = 0.043). However, other studies reported inconsistent results. Vizir OO et al. found that the subjects with severe diabetes were characterized by a significant change in LDH activity [34]. This was demonstrated by an increase in the activity of total LDH, LDH4, and LDH5, but a decrease in LDH1 and LDH2 activity when compared to the healthy subjects [34]. These data indicated the LDH subtype-dependent relation between LDH activity and metabolic diseases. Gender differences in the prevalence of MS were demonstrated by a previous study, which reported that females showed a lower risk for MS than the males [22]. Consistent with previous results, our research also found that the proportion of MS in males (40.0%) was lower than in females (50.2%). The gender difference in the prevalence of MS may be attributed to the discrepant distribution of fat in females and males, and the study has reported that females have more body fat than males. In addition, unhealthy lifestyles such as smoking and drinking in the male population are more common than in females, which are also risk factors for the development of MS [35].

For elderly people in China, there is a stable association between obesity and lipid-related indicators and MS, which were considered effective indicators for diagnosing chronic diseases [36]. As one of the most extensively used indices for evaluating body shape, abnormal BMI was strongly associated with the risk of obesity and metabolic syndrome [37]. Additionally, numerous studies have confirmed the correlation between BMI and MS. A cohort study analyzed the correlation between BMI and the risk of MS, and recommended a BMI level of 27 as the ideal cut-off value for identifying metabolic syndrome [38]. Another longitudinal study divided the participants into fast trajectory (slope = 1.447) and slow trajectory (slope = 1.433) according to the growth rate of BMI with age, which confirmed that the population with a fast growth rate of BMI had a higher incidence of MS than the population with a slow growth rate (OR = 3.40) [39]. The prevalence of MS increased with the rise of BMI and age [40]. As blood pressure, blood glucose, and abdominal obesity increased with age, people at the age of 60 years old and above have more metabolic complications [35]. As a reverse marker of MS and cardiovascular disease, the subclasses of HDL-C, especially the level of HDL3, were reported to have a negative relationship with MS, which could provide more information on diagnosing MS than HDL-C [41].

Studies consistently report that MS patients have a higher TG/HDL-C ratio, WHTR, and VAI than non-MS subjects [42, 43]. In our study, we found a significant association between BMI/HDL-C and MS, abdominal obesity, hyperglycemia, hypertension, hypertriglyceridemia, and HHDL (P < 0.05, Cramer’s V > 0.1). The correlations of TG/HDL-C, WHTR, and VAI with MS were also significant (P < 0.05, Cramer’s V > 0.1), but the correlations of these parameters with hyperglycemia and hypertension were undetected (P > 0.05), implying a better diagnostic efficacy of BMI/HDL-C than the other three indicators. It is reported that the TG/HDL-C ratio is a reliable predictor of MS, particularly because it showed a significant correlation with MS in the elderly Chinese population [44]. Furthermore, studies have indicated that the TG/HDL-C ratio is an efficient biomarker for abnormal lipid metabolism and cardiovascular disease [45]. In our study, we found that the diagnostic potency of TG/HDL-C in diagnosing central obesity, hyperglycemia, and hypertension was poor. Consistently, studies have also reported that WC was not related to the TG/HDL-C ratio. Therefore, we speculate that WC was not an effective predictor of central obesity in adults [46]. As our results showed, there was no significant correlation between TG/HDL-C and abdominal obesity (P = 0.406). Due to comprehensively consideration of waist circumference and height, the WHTR value was suggested as the best predictor of central obesity. However, in our study, the diagnostic effect of WHTR on blood lipids and blood pressure abnormalities was not significant (P > 0.05). The same drawback appeared in the VAI value, which failed to reflect the blood pressure status, and the calculation of VAI was complex [47]. Comparing the models constructed with BMI/HDL-C, TG/HDL-C, VAI, and WHTR, our data demonstrated that model based on BMI/HDL-C was the most comprehensive and reliable model with the greatest potential for diagnosing MS in Chinese adults aged 40–85.

The advantage of this study is that we identified a new key indicator and developed a new model to predict MS in patients clinically. BMI/HDL-C was easy to obtain and had better accuracy and comprehensiveness than other indicators (TG/HDL-C, VAI, WHTR), which provided a new tool for early prediction of MS. Figure 3 showed a nomogram of the developed MS diagnostic model, which provided help for the clinician to assess the risk of MS according to clinical indicators.

There are some limitations in this study. Firstly, the cross-sectional study design makes us fail to draw a causal conclusion. Secondly, the sample size of this study is small and age-related analysis of metabolic syndrome is lacking, thus the conclusions derived from this study need to be further verified in large-scale population studies. Thirdly, the study was conducted in Chinese; therefore, extrapolation of our results to other populations should be done with caution.

Conclusions

This study established a new model based on BMI/HDL-C to predict MS in Chinese adults aged 40–85, which compensated for the lack of existing research on the prediction of MS based on BMI/HDL-C quartile. The model established in our study might provide a sensitive and effective tool for diagnosing MS.

Data availability

Availability of data and materialsThe datasets generated during and analyzed during the current study are available from the corresponding author upon reasonable request.

References

  1. Lemieux I, Despres JP. Metabolic syndrome: past, Present and Future. Nutrients. 2020;12(11).

  2. Saklayen MG. The global epidemic of the metabolic syndrome. Curr Hypertens Rep. 2018;20(2):12.

    Article  PubMed  PubMed Central  Google Scholar 

  3. Pan WH, Yeh WT, Weng LC. Epidemiology of metabolic syndrome in Asia. Asia Pac J Clin Nutr. 2008;17(Suppl 1):37–42.

    CAS  PubMed  Google Scholar 

  4. Bener A, Zirie M, Musallam M, et al. Prevalence of metabolic syndrome according to adult treatment panel III and international diabetes federation criteria: a population-based study. Metab Syndr Relat Disord. 2009;7(3):221–9.

    Article  PubMed  Google Scholar 

  5. Yao F, Bo Y, Zhao L et al. Prevalence and influencing factors of metabolic syndrome among adults in China from 2015 to 2017. Nutrients. 2021;13(12).

  6. Lu J, Wang L, Li M, et al. Metabolic syndrome among adults in China: the 2010 China noncommunicable disease surveillance. J Clin Endocrinol Metab. 2017;102(2):507–15.

    PubMed  Google Scholar 

  7. Li R, Li W, Lun Z, et al. Prevalence of metabolic syndrome in Mainland China: a meta-analysis of published studies. BMC Public Health. 2016;16:296.

    Article  PubMed  PubMed Central  Google Scholar 

  8. Chen Y, Xu W, Zhang W, et al. Plasma metabolic fingerprints for large-scale screening and personalized risk stratification of metabolic syndrome. Cell Rep Med. 2023;4(7):101109.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  9. Cho Y, Lee SY. Useful biomarkers of metabolic syndrome. Int J Environ Res Public Health. 2022;19(22).

  10. Pujos-Guillot E, Brandolini M, Petera M, et al. Systems metabolomics for prediction of metabolic syndrome. J Proteome Res. 2017;16(6):2262–72.

    Article  CAS  PubMed  Google Scholar 

  11. Christakoudi S, Tsilidis KK, Muller DC, et al. A body shape index (ABSI) achieves better mortality risk stratification than alternative indices of abdominal obesity: results from a large European cohort. Sci Rep. 2020;10(1):14541.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  12. Silva VM, Vinagre CG, Dallan LA, et al. Plasma lipids, lipoprotein metabolism and HDL lipid transfers are equally altered in metabolic syndrome and in type 2 diabetes. Lipids. 2014;49(7):677–84.

    Article  CAS  PubMed  Google Scholar 

  13. Son DH, Lee HS, Lee YJ, et al. Comparison of triglyceride-glucose index and HOMA-IR for predicting prevalence and incidence of metabolic syndrome. Nutr Metab Cardiovasc Dis. 2022;32(3):596–604.

    Article  CAS  PubMed  Google Scholar 

  14. Raimi TH, Dele-Ojo BF, Dada SA, et al. Triglyceride-glucose index and related parameters predicted metabolic syndrome in nigerians. Metab Syndr Relat Disord. 2021;19(2):76–82.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  15. Kang SW, Kim SK, Kim YS, et al. Risk prediction of the metabolic syndrome using TyG index and SNPs: a 10-year longitudinal prospective cohort study. Mol Cell Biochem. 2023;478(1):39–45.

    Article  CAS  PubMed  Google Scholar 

  16. Alberti KG, Zimmet PZ. Definition, diagnosis and classification of diabetes mellitus and its complications. Part 1: diagnosis and classification of diabetes mellitus provisional report of a WHO consultation. Diabet Med. 1998;15(7):539–53.

    Article  CAS  PubMed  Google Scholar 

  17. Ho CI, Chen JY, Chen SY, et al. Relationship between TG/HDL-C ratio and metabolic syndrome risk factors with chronic kidney disease in healthy adult population. Clin Nutr. 2015;34(5):874–80.

    Article  CAS  PubMed  Google Scholar 

  18. Zanuncio VV, Sediyama C, Dias MM, et al. Neck circumference and the burden of metabolic syndrome disease: a population-based sample. J Public Health (Oxf). 2022;44(4):753–60.

    Article  CAS  PubMed  Google Scholar 

  19. Laohabut I, Udol K, Phisalprapa P, et al. Neck circumference as a predictor of metabolic syndrome: a cross-sectional study. Prim Care Diabetes. 2020;14(3):265–73.

    Article  PubMed  Google Scholar 

  20. Al-Shami I, Alkhalidy H, Alnaser K, et al. Assessing metabolic syndrome prediction quality using seven anthropometric indices among Jordanian adults: a cross-sectional study. Sci Rep. 2022;12(1):21043.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  21. Ashwell M, Gunn P, Gibson S. Waist-to-height ratio is a better screening tool than waist circumference and BMI for adult cardiometabolic risk factors: systematic review and meta-analysis. Obes Rev. 2012;13(3):275–86.

    Article  CAS  PubMed  Google Scholar 

  22. Amato MC, Giordano C, Galia M, et al. Visceral adiposity index: a reliable indicator of visceral fat function associated with cardiometabolic risk. Diabetes Care. 2010;33(4):920–2.

    Article  PubMed  PubMed Central  Google Scholar 

  23. Radetti G, Fanolla A, Grugni G, et al. Indexes of adiposity and body composition in the prediction of metabolic syndrome in obese children and adolescents: which is the best? Nutr Metab Cardiovasc Dis. 2019;29(11):1189–96.

    Article  PubMed  Google Scholar 

  24. [National guidelines for the prevention and control of diabetes in primary care. (2022)]. Zhonghua Nei Ke Za Zhi, 2022,61(3):249–262.

  25. World Health Organization. The WHO STEPwise approach to noncommunicable disease risk factor surveillance. Geneva 2017.[Z].

  26. Friedewald WT, Levy RI, Fredrickson DS. Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge. Clin Chem. 1972;18(6):499–502.

    Article  CAS  PubMed  Google Scholar 

  27. Chen C, Lu FC. The guidelines for prevention and control of overweight and obesity in Chinese adults. Biomed Environ Sci. 2004;17(Suppl):1–36.

    PubMed  Google Scholar 

  28. Cho SA, Joo HJ, Cho JY, et al. Visceral fat area and serum adiponectin level predict the development of metabolic syndrome in a community-based asymptomatic population. PLoS ONE. 2017;12(1):e169289.

    Article  Google Scholar 

  29. Ying X, Jiang Y, Qin G, et al. Association of body mass index, waist circumference, and metabolic syndrome with serum cystatin C in a Chinese population. Med (Baltim). 2017;96(10):e6289.

    Article  Google Scholar 

  30. Oh YH, Choi S, Lee G et al. Changes in body composition are associated with metabolic changes and the risk of metabolic syndrome. J Clin Med. 2021;10(4).

  31. Kim SH, Kang HW, Jeong JB, et al. Association of obesity, visceral adiposity, and Sarcopenia with an increased risk of metabolic syndrome: a retrospective study. PLoS ONE. 2021;16(8):e256083.

    Article  Google Scholar 

  32. Wang J, Li X, Han X, et al. Serum creatinine levels and risk of metabolic syndrome in a middle-aged and older Chinese population. Clin Chim Acta. 2015;440:177–82.

    Article  CAS  PubMed  Google Scholar 

  33. Johari TY, Ghoneim MA, Moselhy SS. Thyroid profile and LDH isoenzymes as prognostic biomarkers for diabetic and/or obese subjects. Afr Health Sci. 2018;18(3):697–706.

    Article  PubMed  PubMed Central  Google Scholar 

  34. Vizir OO. [Activity of blood serum lactate dehydrogenase in diabetes mellitus]. Probl Endokrinol (Mosk). 1977;23(3):15–7.

    CAS  PubMed  Google Scholar 

  35. Slagter SN, van Waateringe RP, van Beek AP, et al. Sex, BMI and age differences in metabolic syndrome: the dutch lifelines cohort study. Endocr Connect. 2017;6(4):278–88.

    Article  PubMed  PubMed Central  Google Scholar 

  36. Gui J, Li Y, Liu H, et al. Obesity- and lipid-related indices as a predictor of obesity metabolic syndrome in a national cohort study. Front Public Health. 2023;11:1073824.

    Article  PubMed  PubMed Central  Google Scholar 

  37. Bramante CT, Palzer EF, Rudser KD, et al. BMI metrics and their association with adiposity, cardiometabolic risk factors, and biomarkers in children and adolescents. Int J Obes (Lond). 2022;46(2):359–65.

    Article  CAS  PubMed  Google Scholar 

  38. Kobo O, Leiba R, Avizohar O, et al. Normal body mass index (BMI) can rule out metabolic syndrome: an Israeli cohort study. Med (Baltim). 2019;98(9):e14712.

    Article  Google Scholar 

  39. Ying M, Hu X, Li Q, et al. Long-term trajectories of BMI and cumulative incident metabolic syndrome: a cohort study. Front Endocrinol (Lausanne). 2022;13:915394.

    Article  PubMed  Google Scholar 

  40. Razzouk L, Muntner P. Ethnic, gender, and age-related differences in patients with the metabolic syndrome. Curr Hypertens Rep. 2009;11(2):127–32.

    Article  PubMed  Google Scholar 

  41. Yang HS, Hur M, Kim H, et al. HDL subclass analysis in predicting metabolic syndrome in koreans with high HDL cholesterol levels. Ann Lab Med. 2020;40(4):297–305.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  42. Jialal I, Adams-Huet B, Remaley AT. A comparison of the ratios of C-reactive protein and triglycerides to high-density lipoprotein-cholesterol as biomarkers of metabolic syndrome in African americans and non-hispanic whites. J Diabetes Complications. 2022;36(7):108231.

    Article  CAS  PubMed  Google Scholar 

  43. Ma A, Fang K, Dong J, et al. Prevalence and related factors of metabolic syndrome in Beijing, China (Year 2017). Obes Facts. 2020;13(6):538–47.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  44. Nie G, Hou S, Zhang M, et al. High TG/HDL ratio suggests a higher risk of metabolic syndrome among an elderly Chinese population: a cross-sectional study. BMJ Open. 2021;11(3):e41519.

    Article  Google Scholar 

  45. Kosmas CE, Rodriguez PS, Bousvarou MD et al. The triglyceride/high-density lipoprotein cholesterol (TG/HDL-C) ratio as a risk marker for metabolic syndrome and cardiovascular disease. Diagnostics (Basel). 2023;13(5).

  46. Ren X, Chen ZA, Zheng S, et al. Association between triglyceride to HDL-C ratio (TG/HDL-C) and insulin resistance in Chinese patients with newly diagnosed type 2 diabetes mellitus. PLoS ONE. 2016;11(4):e154345.

    Article  Google Scholar 

  47. Lazzer S, D’Alleva M, Isola M et al. Cardiometabolic index (CMI) and visceral adiposity index (VAI) highlight a higher risk of metabolic syndrome in women with severe obesity. J Clin Med. 2023;12(9).

Download references

Acknowledgements

Not applicable.

Funding

This work was supported by the National Natural Science Foundation of China (82173508; 81973027), and the Beijing High-level Public Health Technical Personnel Training Program (2022–3 − 032); and the Suzhou Science and Technology City Hospital Talent Introduction Project (Yj202208).

Author information

Author notes

  1. Yu Liu and Xixiang Wang contributed equally to this work.

Authors and Affiliations

  1. School of Public Health, Capital Medical University, Beijing, 100069, P.R. China

    Yu Liu, Xixiang Wang, Xiaojun Ma, Jingjing Xu, Lu Liu, Xiuwen Ren & Linhong Yuan

  2. Suzhou Research Center of Medical School, Suzhou Hospital, Affiliated Hospital of Medical School, Nanjing University, Suzhou, China

    Jie Mu, Yiyao Gu, Zhi Duan & Ying Wang

  3. School of Science, Faculty of Engineering and Science, University of Greenwich, Central Avenue, Chatham, ME4 4TB, UK

    Shaobo Zhou

Authors
  1. Yu Liu
    View author publications

    You can also search for this author inPubMed Google Scholar

  2. Xixiang Wang
    View author publications

    You can also search for this author inPubMed Google Scholar

  3. Jie Mu
    View author publications

    You can also search for this author inPubMed Google Scholar

  4. Yiyao Gu
    View author publications

    You can also search for this author inPubMed Google Scholar

  5. Shaobo Zhou
    View author publications

    You can also search for this author inPubMed Google Scholar

  6. Xiaojun Ma
    View author publications

    You can also search for this author inPubMed Google Scholar

  7. Jingjing Xu
    View author publications

    You can also search for this author inPubMed Google Scholar

  8. Lu Liu
    View author publications

    You can also search for this author inPubMed Google Scholar

  9. Xiuwen Ren
    View author publications

    You can also search for this author inPubMed Google Scholar

  10. Zhi Duan
    View author publications

    You can also search for this author inPubMed Google Scholar

  11. Linhong Yuan
    View author publications

    You can also search for this author inPubMed Google Scholar

  12. Ying Wang
    View author publications

    You can also search for this author inPubMed Google Scholar

Contributions

Yu Liu: Conceptualization, Methodology, Software, Writing - Original Draft. Xixiang Wang: Methodology, Formal analysis, Visualization. Jie Mu: Investigation. Yiyao Gu: Investigation. Shaobo Zhou: Writing – Review & Editing. Xiaojun Ma: Investigation. Jingjing Xu: Investigation. Lu Liu: Investigation. Xiuwen Ren: Investigation. Zhi Duan: Investigation. Linhong Yuan: Conceptualization, Supervision, Writing - Review & Editing, Funding acquisition. Ying Wang: Conceptualization, Supervision, Writing - Review & Editing, Funding acquisition.

Corresponding authors

Correspondence to Linhong Yuan or Ying Wang.

Ethics declarations

Ethics approval

The study was approved by the Committee on Medical Ethics of Capital Medical University. The study protocol was approved by the Committee on Medical Ethics of Capital Medical University (No. 2012SY23), and the study procedures followed the ethical standards of the Helsinki Declaration of 1975.

Consent to participate

Written informed consent was obtained from individual or guardian participants.

Consent for publication

Not applicable.

Competing interests

The authors have no relevant financial or non-financial interests to disclose.

Additional information

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Supplementary Material 1

Supplementary Material 2

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Liu, Y., Wang, X., Mu, J. et al. Developing a risk model for early diagnosis of metabolic syndrome in Chinese adults aged 40 years and above based on BMI/HDL-C: a cross-sectional study. BMC Endocr Disord 24, 223 (2024). https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s12902-024-01752-9

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s12902-024-01752-9

Keywords

BMC Endocrine Disorders

ISSN: 1472-6823

Contact us