Preamble

Correlation Analysis Between Chinese Visceral Adiposity Index and the Risk of Diabetes Mellitus in Residents Aged 18 and Above

Abstract

Objective: To investigate the correlation between the Chinese Visceral Adiposity Index (CVAI) and the risk of diabetes mellitus (DM) among residents aged 18 and above in China, and to evaluate the predictive value of CVAI for DM risk.

Methods: Data were derived from the 2015 China Adult Chronic Disease and Nutrition Surveillance (CACDNS). A total of 164,188 participants aged 18 and above with complete data were selected. CVAI was calculated based on age, body mass index (BMI), waist circumference (WC), triglycerides (TG), and high-density lipoprotein cholesterol (HDL-C). Participants were divided into four groups (Q1–Q4) based on CVAI quartiles. Multivariate logistic regression models were used to analyze the association between CVAI and DM risk. Restricted cubic spline (RCS) analysis was employed to characterize the dose-response relationship, and subgroup analyses were performed to assess consistency across different populations.

Results: Among the 164,188 participants, the prevalence of DM was 11.9%. After adjusting for potential confounders, multivariate logistic regression analysis showed that compared with the lowest quartile (Q1), the odds ratios (OR) and 95% confidence intervals (CI) for DM in the Q2, Q3, and Q4 groups were 2.14 (1.95–2.35), 3.85 (3.52–4.21), and 7.82 (7.16–8.54), respectively (P for trend < 0.001). RCS analysis revealed a non-linear positive correlation between CVAI and DM risk (P for non-linearity < 0.001). Subgroup analyses indicated that the positive association between CVAI and DM remained significant across different age groups, genders, and regions.

Conclusion: CVAI is significantly and positively correlated with the risk of diabetes in Chinese adults. As a comprehensive indicator of visceral fat distribution and metabolic health, CVAI may serve as a valuable tool for identifying individuals at high risk of diabetes in clinical and public health settings.

Introduction

Diabetes mellitus (DM) has become a major global public health challenge, with its prevalence increasing

1.110005 辽宁省沈阳市，中国医科大学预防医学研究院

Health Liaoning Action Monitoring and Evaluation Office, Liaoning Provincial Center for Disease Control and Prevention, Shenyang, Liaoning Province; Department of Non-communicable Disease Prevention and Control, Dalian Municipal Center for Disease Control and Prevention, Dalian, Liaoning Province; Department of Non-communicable Disease Prevention and Control, Benxi Municipal Center for Disease Control and Prevention, Benxi, Liaoning Province; Medical Records Department, The First Hospital of China Medical University, Shenyang, Liaoning Province; Department of Ultrasound, The Fourth Hospital of China Medical University, Shenyang, Liaoning Province.

背景

The Chinese Visceral Adiposity Index (CVAI) is a novel and simplified metric for evaluating human visceral fat, which has been shown to be closely associated with cardiovascular and metabolic diseases. Currently, research regarding the correlation between CVAI and diabetes remains limited. This study aims to explore the association between CVAI and diabetes to provide a scientific basis for diabetes screening and prediction, while offering new insights and recommendations for the prevention, treatment, and management of the disease. From January to December 2023, a cross-sectional survey was conducted among urban and rural residents aged 40 and older in Liaoning Province using stratified multi-stage cluster sampling. The investigation included questionnaires, physical examinations, and laboratory tests. Measurements of height, body mass, waist circumference (WC), blood pressure, total cholesterol (TC), high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), glycated hemoglobin ($HbA_{1c}$), and fasting plasma glucose (FPG) were collected to calculate the CVAI. Multivariable binary logistic regression analysis and restricted cubic spline (RCS) plots were employed to evaluate the correlation between the CVAI and diabetes. Receiver operating characteristic (ROC) curves were used to assess the predictive value of CVAI, BMI, and WC for diabetes risk in the total population as well as in male and female subgroups. The area under the curve (AUC), sensitivity, specificity, and optimal cutoff values were also calculated. A total of 32,813 participants were included in this study, of whom 8,421 had diabetes, representing a prevalence of 25.7%. The mean CVAI was $119.37 \pm 37.01$. When categorized by CVAI quartiles, the prevalence of diabetes in the Q1, Q2, Q3, and Q4 groups was 13.9% (1,138/8,204), 22.7% (1,862/8,203), 29.4% (2,415/8,203), and 36.6% (3,006/8,203), respectively. Multivariable binary logistic regression results indicated that after adjusting for confounding factors, compared with the Q1 group, the Q2 group ($OR = 1.74, 95\% CI = 1.60\text{--}1.89$), Q3 group ($OR = 2.38, 95\% CI = 2.20\text{--}2.58$), and Q4 group ($OR = 3.18, 95\% CI = 2.94\text{--}3.44$) were all significantly associated with a higher risk of diabetes ($P < 0.05$).

Restricted cubic spline analysis revealed a significant non-linear relationship between CVAI and the risk of diabetes ($P < 0.01$). As CVAI levels increased, the risk of diabetes rose significantly ($P < 0.05$). ROC curve analysis showed that the AUCs for CVAI, BMI, and WC in predicting diabetes risk were 0.635 ($95\% CI = 0.628\text{--}0.642, P < 0.001$), 0.611 ($95\% CI = 0.604\text{--}0.608, P < 0.001$), and 0.579 ($95\% CI = 0.572\text{--}0.586, P < 0.001$), respectively. In conclusion, there is a significant positive correlation between CVAI and the prevalence of diabetes. This index can effectively identify high-risk populations at an early stage, facilitating timely health management and intervention measures to address the onset and progression of diabetes.

Keywords: Diabetes; Visceral Adiposity Index; Cross-sectional study; Liaoning Province
CLC Number: R 587.1
Document Code: A

Chinese General Practice https Prevention, Shenyang 110005, China for the occurrence and development of diabetes in a timely manner.

Diabetes has become a severe challenge to global public health. Statistics indicate that in 2021, the prevalence of diabetes among adults aged 20 to 79 worldwide was 10.5% (536.6 million people), a figure projected to rise to 12.2% (783.2 million people) by 2045. In China specifically, the incidence of diabetes has been increasing annually due to shifting lifestyles and rising obesity rates, with an estimated 113.9 million Chinese adults expected to be affected. The onset of diabetes is closely linked to visceral fat. Recently, research into the correlation between visceral adiposity and diabetes has garnered significant attention. The Chinese Visceral Adiposity Index (CVAI) is regarded as a novel metric for assessing human visceral fat, offering substantial clinical value for the early diagnosis and intervention of diabetes.

The CVAI incorporates indicators tailored to the physiological characteristics of the Chinese population, allowing for the assessment of visceral fat distribution and metabolic status. Existing studies have demonstrated a strong association between CVAI and the occurrence of various chronic diseases, particularly diabetes \cite{3-4}. Based on data from the 2023 screening of populations at high risk for stroke in Liaoning Province, this study analyzes the prevalence of diabetes and explores the association between CVAI and the disease in depth. Furthermore, it evaluates the efficacy of body obesity-related indicators in screening and predicting diabetes risk among middle-aged and elderly Chinese adults, aiming to provide new perspectives and a theoretical foundation for the prevention and treatment of diabetes.

1.1 研究对象

From January to December 2023, a multi-stage stratified cluster random sampling method was employed to select participants from eight cities in Liaoning Province: Shenyang, Dalian, Benxi, Dandong, Chaoyang, Yingkou, Jinzhou, and Liaoyang.

XING Liying, Chief Physician, Doctoral Supervisor.

Background

The Chinese visceral adiposity index (CVAI) is a new simple index to assess visceral fat in the human body, which is closely related to cardiovascular diseases and metabolic diseases. At present, there is limited research on the correlation between CVAI and diabetes.

Objective The purpose of this study was to explore the association between CVAI nd diabetes, to provide a scientific basis for the screening and prediction of diabetes, and to provide new enlightenment and suggestions for the clinical practice of diabetes prevention and treatment and diabetes management.

Methods

From January to December 2023, a stratified multi-stage cluster sampling method was adopted to conduct questionnaire surveys, physical examinations, and laboratory tests among urban and rural residents aged 40 years and above in Liaoning Province. Indicators including height, body weight, waist circumference (WC), blood pressure, total cholesterol (TC), high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), glycosylated hemoglobin (HbA ), and fasting plasma glucose (FPG) of the participants were determined to calculate the Cardiovascular Age Index (CVAI). Binary Logistic multivariate regression analysis and restricted cubic spline plots were employed to evaluate the association between the CVAI index and diabetes mellitus (DM).

Receiver operating characteristic (ROC)curves were utilized to assess the predictive value of CVAI, body mass index (BMI), and WC for the risk of DM in the total population, as well as in male and female subgroups separately. The area under the ROC curve (AUC), sensitivity, specificity, and optimal cut-off values were calculated accordingly.

Results

A total of 32 813 participants were included in this study, among whom 8 421 were diabetic patients, with a prevalence rate of 25.7%. The mean CVAI was (119.37±37.01). When CVAI was grouped by quartiles, the prevalence rates of diabetes in the Q1, Q2, Q3, and Q4 groups were 13.9% (1 138/8, 204), 22.7% (1 862/8, 203), 29.4% (2 415/8, 203), and 36.6% (3 006/8, 203), respectively. Result s of the binary Logistic multivariate regression analysis showed that after adjusting for confounding factors, compared with the Q1 group, the Q2 group ( 1.74, 95% =1.60-1.89), Q3 group ( 2.38, 95% =2.20-2.58), and Q4 group ( 3.18, 95% =2.94-3.44) of CVAI were all associated with a higher risk of diabetes ( 0.05). Result s of the restricted cubic restricted cubic spline analysis showed a significant non-linear relationship between CVAI and diabetes risk ( 0.01). The risk of diabetes increased significantly with the elevation of CVAI levels ( 0.05). Result s of the ROC curve analysis indicated that the AUCs of CVAI, BMI, and WC for predicting the risk of diabetes were 0.635 (95% =0.628-0.642, 0.001), 0.611 (95% =0.604-0.608, 0.001), and 0.579 =0.572-0.586, 0.001), respectively.

Conclusion

CVAI is positively correlated with diabetes, and it is recommended to promote CVAI to identify potential high-risk groups of diabetes at an early stage and carry out health management and intervention Key words Diabetes; Chinese visceral adiposity index; Cross-sectional study; Liaoning province

Permanent residents were included in the study according to the following criteria: (1) permanent residency in the local area (residing in the region for no less than 6 months per year); (2) clear verbal expression and communication skills; and (3) consent to the collection of biological samples and cooperation in completing the research requirements.

This study was approved by the Ethics Committee of the Liaoning Provincial Center for Disease Control and Prevention (Ethics Approval No. 2023-004), and all participants provided written informed consent. A total of 34,384 cases were initially included in the investigation.

After excluding subjects with missing information, logical errors, or outliers, a total of 32,813 cases were included in the final analysis.

1.2.1 问卷调查

This study utilized a standardized questionnaire to collect basic demographic data from the participants. Data collection was conducted by investigators who had undergone formal training provided by the Liaoning Provincial Center for Disease Control and Prevention. Information was gathered through face-to-face interviews and included variables such as age, sex, urban or rural residency, physical inactivity, smoking status, alcohol consumption, and average annual household income.

1.2.2 体格检查

Standardized physical examinations were conducted by professionally trained investigators, encompassing measurements of height, body weight, waist circumference (WC), and blood pressure. During the measurement process, participants were required to remove their shoes, bags, and hats. Height and body weight were recorded with a precision of 0.1 cm and 0.1 kg, respectively. Body Mass Index (BMI) was subsequently calculated using the formula: $BMI = \text{weight (kg)} / \text{height (m)}^2$.

Waist circumference was measured using a tape measure positioned horizontally at the level of the umbilicus, while hip circumference was measured at the widest point of the buttocks, both with a precision of 0.1 cm. Blood pressure measurements were performed in a relatively quiet environment using an Omron J30 electronic sphygmomanometer (Omron Corporation, Japan). Measurements were taken from the participant's left upper arm a total of three times, with an interval of 1–2 minutes between each reading; the average of these three measurements was then recorded for analysis.

1.2.3 实验室检查

Nurses at the project site hospitals collected 10 mL venous blood samples from the participants (following a fasting period of $\ge 8$ hours). Laboratory technicians utilized the Abbott 800i automated analyzer (Abbott Laboratories, USA) to process the blood samples through standing and centrifugation in accordance with relevant testing requirements.

Biochemical indicators measured included total cholesterol (TC), high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), glycated hemoglobin ($HbA_{1c}$), and fasting plasma glucose (FPG). The diagnostic criteria were defined as follows: (1) Diabetes: Based on the ADA Standards of Care in Diabetes (2024), diabetes was defined as $FPG \ge 7.0$ mmol/L and/or $HbA_{1c} \ge 6.5\%$ ($\ge 48$ mmol/mol), or a previous medical history of diabetes. (2) Hypertension: According to the Chinese Guidelines for Prevention and Treatment of Hypertension (2018 Revision), hypertension was defined as an office systolic blood pressure (SBP) $\ge 140$ mmHg (1 mmHg = 0.133 kPa) and/or a diastolic blood pressure (DBP) $\ge 90$ mmHg, and/or the current use of antihypertensive medication. (3) Dyslipidemia: Based on the Guideline for the Prevention and Treatment of Dyslipidemia in Chinese Adults (2016 Revision), dyslipidemia was defined as $TC \ge 6.2$ mmol/L, or triglycerides ($TG$) $\ge 2.3$ mmol/L, or $LDL\text{-}C \ge 4.1$ mmol/L, or $HDL\text{-}C < 1.0$ mmol/L, or the recent use of lipid-lowering medications. (4) Smoking and Alcohol Consumption: Smoking was defined as self-reported smoking of at least one cigarette per day for more than one year. Drinking was defined as self-reported alcohol consumption occurring at least once per week for at least one year.

The Chinese Visceral Adiposity Index (CVAI) was calculated as follows \cite{9}:
For males: $CVAI = -267.93 + 0.68 \times \text{Age} + 0.03 \times BMI + 4.00 \times WC + 22.00 \times \lg TG - 16.32 \times HDL\text{-}C$

(5) The CVAI calculation formulas are:

Equation (1)
For females: $CVAI = -187.32 + 1.71 \times \text{Age} + 4.23 \times BMI + 1.12 \times WC + 39.76 \times \lg TG - 11.66 \times HDL\text{-}C$

Quality Control

A quality control group was jointly established by personnel from the provincial, municipal, and district Centers for Disease Control and Prevention (CDC). Members of the on-site quality control team strictly monitored data quality and the standardization of all field operations. Any identified issues were promptly reported to the relevant staff for re-investigation and correction. All investigators and operators underwent standardized training and were required to pass an assessment before participating in the field survey. The quality control group regularly supervised and evaluated the quality of questionnaires submitted by the project sites as well as the reproducibility of the data.

Statistical Methods

Data analysis was performed using SPSS 26.0 software. Quantitative data following a normal distribution were described as mean $\pm$ standard deviation ($\bar{x} \pm s$), and comparisons between groups were conducted using independent samples analysis of variance (ANOVA). Categorical data were described using relative numbers (percentages), and comparisons between groups were performed using the $\chi^2$ test.

2 检验。采用

Binary logistic multivariate regression analysis was employed to investigate the association between the Chinese Visceral Adiposity Index (CVAI) and the prevalence of hyperglycemia, with the significance level set at $\alpha = 0.05$. The variable assignments and coding details are presented in [TABLE:1].

Subsequently, R software (version 4.4.0) was utilized to treat CVAI as a continuous variable, and restricted cubic spline (RCS) analysis was performed to explore the dose-response relationship between CVAI and the risk of diabetes. To evaluate the predictive value of CVAI, Body Mass Index (BMI), and Waist Circumference (WC) for diabetes risk, Receiver Operating Characteristic (ROC) curve analysis was conducted. The area under the ROC curve (AUC), sensitivity, specificity, and the optimal cutoff values were calculated for each metric. A $P$-value of less than 0.05 was considered to indicate a statistically significant difference.

2.1 基本情况

A total of 32,813 participants were included in this study, with a mean age of $63.38 \pm 9.27$ years. The cohort consisted of 11,934 males (36.4%) and 20,882 females (63.6%).

The mean Chinese Visceral Adiposity Index (CVAI) of the study subjects was $119.4 \pm 37.0$ cm, the mean waist circumference (WC) was $88.3 \pm 10.1$ cm, and the mean body mass index (BMI) was $25.5 \pm 3.5$ kg/m². CVAI values were divided into quartiles: Q1 ($\leq 94.97$), Q2 ($94.97$–$119.30$), Q3 ($119.30$–$143.63$), and Q4 ($\geq 143.63$). Statistically significant differences ($P < 0.05$) were observed across the four groups (Q1, Q2, Q3, and Q4) regarding age, sex, urban/rural residence, diabetes status, physical inactivity, smoking, alcohol consumption, average annual household income, BMI, WC, systolic blood pressure (SBP), diastolic blood pressure (DBP), high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), total cholesterol (TC), triglycerides (TG), glycated hemoglobin ($HbA_{1c}$), and fasting plasma glucose (FPG), as shown in [TABLE:2].

Residents of Liaoning Province categorized by gender and urban/rural status...

Chinese General Practice https Covariate assignment in binary Logistic multivariate regression

analysis

Average annual household income was categorized as: 0 = <5,000 RMB; 1 = 5,000–9,999 RMB.

Note: WC = waist circumference; SBP = systolic blood pressure; DBP = diastolic blood pressure; TC = total cholesterol; TG = triglycerides; HDL-C = high-density lipoprotein cholesterol; LDL-C = low-density lipoprotein cholesterol; HbA1c = glycated hemoglobin; FPG = fasting plasma glucose.

Prevalence of Diabetes

In this study, a total of 8,421 participants (25.7%) were diagnosed with diabetes. Among these, 3,253 cases were male (27.3%) and 5,168 were female (24.7%). Regarding residency, 4,737 cases (27.7%) were urban residents, while 3,684 cases (23.5%) were from rural areas. The prevalence of diabetes showed a significant upward trend as Chinese Visceral Adiposity Index (CVAI) levels increased across the total population, as well as when stratified by sex and area of residence. These differences were all statistically significant ($P < 0.01$), as shown in [TABLE:3].

Multivariate Logistic Regression Analysis of the Relationship Between CVAI and Diabetes Risk

A binary logistic regression analysis was conducted with diabetes status as the dependent variable (0 = No, 1 = Yes) and CVAI quartiles as the independent variable (Q1, Q2, Q3, and Q4). The results indicated that after adjusting for confounding factors—including age, sex, average annual household income, physical inactivity, smoking, alcohol consumption, SBP, and DBP—in Model 3, the risk of diabetes increased significantly with higher CVAI. Compared to the Q1 group, the risk of diabetes in the Q2, Q3, and Q4 groups increased to 1.74 times ($OR = 1.74$, 95% $CI: 1.60\text{--}3.89$), 2.38 times ($OR = 2.38$, 95% $CI: 2.20\text{--}2.58$), and 3.18 times ($OR = 3.18$, 95% $CI: 2.94\text{--}3.44$), respectively. Detailed results are presented in [TABLE:5].

The results of the restricted cubic spline (RCS) analysis revealed a significant non-linear relationship between CVAI and the risk of diabetes in the overall study population ($P < 0.01$). As CVAI levels increased, the risk of diabetes rose markedly, as illustrated in [FIGURE:1]A. Further sex-stratified analysis demonstrated that the significant non-linear relationship between CVAI and diabetes risk persisted for both males and females ($P < 0.05$). Notably, this dose-response relationship was more pronounced among female participants, as shown in [FIGURE:1]B and [FIGURE:1]C.

Predictive Value of CVAI for Diabetes Risk

分析

The ROC curve results demonstrate that the Area Under the Curve (AUC) values for CVAI, BMI, and WC in predicting the risk of diabetes were 0.635 (95% CI = 0.628–0.642, $P < 0.001$), 0.611 (95% CI = 0.604–0.608, $P < 0.001$), and 0.579 (95% CI = 0.572–0.586, $P < 0.001$), respectively; further details are provided in [FIGURE:2]. Based on the calculation of the Youden index from the ROC curves, the optimal cutoff value for CVAI to predict diabetes in the total population was 116.40 (132.35 for males and 117.10 for females). For WC, the optimal cutoff value in the total population was 87.45 cm (91.45 cm for males and 87.45 cm for females). Regarding BMI, the optimal cutoff value in the total population was

24.71 (24.74 for males and 25.05 for females), as shown in

3 讨论

This study evaluated the association between the Chinese Visceral Adiposity Index (CVAI) and the risk of diabetes among residents aged $\ge 40$ years in Liaoning Province. The findings indicate that CVAI is closely associated with the prevalence of diabetes, and higher CVAI levels serve as a significant risk factor for the disease.

Research by JIANG et al. \cite{...} demonstrates that the visceral adiposity coefficient, as an indicator for assessing visceral fat content, reflects the body's metabolic status and degree of insulin resistance. Excessive visceral fat accumulation tends to induce insulin resistance and cause metabolic disorders, thereby increasing the risk of developing diabetes.

The results of the multivariate binary logistic regression analysis in this study show that after adjusting for confounding factors, the risk of diabetes in the highest CVAI quartile (Q4) was 3.18 times that of the lowest quartile (Q1) ($OR = 3.18$, $95\% CI = 2.94\text{--}3.44$). CVAI levels were significantly and positively correlated with diabetes risk. These findings are consistent with the research results of XIA \cite{...}, ANOOP \cite{...}, and WEI et al. \cite{...}, further validating CVAI as a potential independent risk factor for diabetes. Furthermore, this study utilized restricted cubic splines (RCS) to illustrate the non-linear relationship between CVAI and diabetes. The analysis revealed that the risk of diabetes begins to increase when CVAI exceeds 119.00; as CVAI levels continue to rise, the risk of diabetes increases significantly. These results refine the understanding of the relationship between visceral adiposity indices and diabetes in middle-aged and elderly Chinese populations. By establishing a dose-response relationship and identifying a critical threshold for CVAI, this study provides an important quantitative indicator for assessing diabetes risk, which helps optimize personalized prevention and management strategies.

Although MRI and CT provide the most accurate measurements of body fat, these methods are impractical for large-scale cohort studies and routine screenings due to their complexity, time-consuming nature, and high cost. While traditional metrics such as Body Mass Index (BMI) and Waist Circumference (WC) can measure visceral obesity, they fail to comprehensively account for metabolic indicators, thus limiting their effectiveness in identifying metabolic diseases. In contrast, CVAI is tailored to the physique of the Chinese population and integrates critical factors such as lipid levels, age, and gender, providing a more robust tool for assessing obesity and identifying metabolic disorders.

[TABLE:1] (Note: Data fragments indicate age distributions across groups: $58.5 \pm 9.4$, $62.5 \pm 8.4$, $65.1 \pm 8.2$, $67.3 \pm 8.7$, and a total mean of $63.4 \pm 9.3$, with $P < 0.001$).

a <0.001

a <0.001

a <0.001

[TABLE:1]

The baseline characteristics of the study population across different Chinese Visceral Adiposity Index (CVAI) quartiles are presented above. Significant differences were observed across all physiological and biochemical parameters ($P < 0.001$). Specifically, mean waist circumference increased progressively from $78.2 \pm 7.1$ kg/m in the lowest quartile to $99.3 \pm 7.3$ kg/m in the highest quartile. Similarly, both systolic blood pressure (SBP) and diastolic blood pressure (DBP) showed an upward trend across quartiles, with SBP rising from $133 \pm 21$ mmHg to $150 \pm 21$ mmHg. Regarding lipid profiles, high-density lipoprotein cholesterol (HDL-C) levels decreased significantly as CVAI increased ($1.59 \pm 0.40$ mmol/L to $1.20 \pm 0.31$ mmol/L), while triglyceride (TG) levels nearly doubled from the first to the fourth quartile ($1.30 \pm 0.75$ mmol/L vs. $2.34 \pm 1.31$ mmol/L).

[FIGURE:1]

The prevalence of diabetes mellitus was analyzed across CVAI quartiles, stratified by both sex and geographic region. The results indicate a consistent positive correlation between CVAI levels and the prevalence of diabetes in all subgroups.

Association Between CVAI and Diabetes Risk

To further evaluate the relationship between CVAI and the risk of diabetes, a multifactorial Logistic regression analysis was conducted. The results are summarized across three progressive models:

• Model 1: Unadjusted crude analysis.
• Model 2: Adjusted for sex, age, and average annual household income.
• Model 3: Further adjusted for sex, age, alcohol consumption, smoking status, physical inactivity, average annual household income, SBP, and DBP.

The analysis demonstrates that CVAI remains a significant independent risk factor for diabetes even after adjusting for multiple confounding variables. (Note: "—" indicates that the value was not applicable or not calculated for that specific category).

Chinese General Practice https Predictive value of CVAI，BMI，and WC for diabetes risk in the overall population，men，and women

2 ） 0.643 0.450 <0.001

2 ） 0.605 0.512 <0.001

2 ） 0.640 0.466 <0.001

Note: A represents the total population, B represents males, and C represents females; CVAI = Chinese Visceral Adiposity Index.

Restricted cubic spline analysis of the relationship between CVAI and diabetes.

1 - Specificity, 1 - Specificity, 1 - Specificity.

Note: A represents the total population, B represents males, and C represents females; ROC = Receiver Operating Characteristic; CVAI = Chinese Visceral Adiposity Index; BMI = Body Mass Index; WC = Waist Circumference.

ROC curves of CVAI, BMI, and WC for predicting diabetes risk in the overall population, men, and women provide more comprehensive and accurate evaluation metrics.

The AUC (Area Under the Curve), specificity, and sensitivity of the CVAI demonstrate superior performance compared to both Body Mass Index (BMI) and Waist Circumference (WC). Research indicates that, within Asian populations, CVAI outperforms conventional anthropometric measurements and visceral fat indices as a marker for abdominal obesity.

Based on the results of the ROC curve analysis in this study, the AUC for CVAI in the total population was 0.635, while the AUC for BMI was 0.539 and the AUC for WC was 0.611. These findings indicate that CVAI demonstrates superior performance in the prediction of diabetes compared to traditional anthropometric indices.

Abstract

In recent years, the rapid development of computer vision and artificial intelligence (CVAI) has significantly impacted various fields, including medical imaging, biometric identification, and social governance. However, the performance and fairness of these algorithms across different demographic groups—particularly among Chinese populations and other ethnic groups—remain critical areas of research. This paper examines the current state of CVAI applications, focusing on the technical challenges and ethical considerations associated with ethnic diversity. We analyze how data representation affects model accuracy and discuss strategies for mitigating algorithmic bias to ensure equitable outcomes across diverse populations.

Introduction

Computer vision and artificial intelligence (CVAI) technologies have transitioned from theoretical research to large-scale practical deployment. In the context of a globalized society, the ability of these systems to accurately recognize, categorize, and serve individuals from different ethnic backgrounds is paramount. For the Chinese population, which encompasses a vast array of regional and ethnic variations, as well as for other global ethnic groups, the "one-size-fits-all" approach to model training often leads to performance degradation or systemic bias.

[FIGURE:1]

The core of these challenges lies in the diversity of physiological features, cultural contexts, and the inherent biases present in training datasets. When CVAI systems are trained predominantly on datasets lacking sufficient representation of specific groups, the resulting models may exhibit higher error rates when applied to those underrepresented populations. This phenomenon, often referred to as "algorithmic bias," poses significant risks in sensitive applications such as facial recognition, healthcare diagnostics, and automated surveillance.

Methodology and Data Representation

To address these disparities, researchers have begun to emphasize the importance of inclusive data collection and robust evaluation metrics. The performance of a model $\mathcal{M}$ on a population $\mathcal{P}$ can be defined by its loss function $L(\mathcal{M}, \mathcal{P})$. If $\mathcal{P}{Chinese}$ and $\mathcal{P}$ represent different ethnic cohorts, a fair system aims to minimize the discrepancy:

$$\Delta = |L(\mathcal{M}, \mathcal{P}{Chinese}) - L(\mathcal{M}, \mathcal{P})|$$

Data Diversity in Chinese Populations

The Chinese population presents a unique case study due to its internal diversity. Beyond the Han majority, there are 55 recognized ethnic minorities, each with distinct facial characteristics and cultural markers. Standard datasets often fail to capture this granularity, leading to models that perform well in urban centers but struggle in border regions or

Studies have successfully predicted diabetic events within the Chinese general population \cite{9, 16}. Research indicates that various anthropometric indices, including the Visceral Adiposity Index (VAI), Body Mass Index (BMI), and Waist Circumference (WC), are closely associated with the risk of diabetes. Notably, the Chinese Visceral Adiposity Index (CVAI) has demonstrated superior diagnostic performance for overall diabetes and prediabetes among Chinese adults compared to traditional metrics such as BMI, WC, and waist-to-hip ratio \cite{12-13, 17–20}. This superiority primarily stems from the fact that CVAI more accurately reflects visceral adiposity. In contrast, BMI and WC fail to effectively distinguish between visceral and subcutaneous obesity, and they struggle to identify the specific issues of excessive central obesity prevalent among Chinese adults.

[21]. Furthermore, the ROC curve analysis demonstrated that the Area Under the Curve (AUC) for CVAI was 0.653 for females and 0.606 for males. These findings are consistent with the results reported by Wu et al.

These findings are consistent with the research results reported by LIN et al. This phenomenon may be related to gender-specific differences in the mechanisms by which visceral obesity leads to diabetes, as well as variations in visceral fat deposition patterns and regional adipose tissue distribution. In summary, CVAI not only demonstrates superior diagnostic efficacy compared to traditional metrics such as BMI and WC, but also provides a more precise risk assessment by integrating biochemical indicators and age. Therefore, CVAI should be considered an important evaluation tool in the screening and management of diabetes. Identifying high-risk individuals early and implementing effective intervention measures will help reduce the incidence of the disease and slow its progression.

The large sample size of this study effectively reflects the association between CVAI and diabetes prevalence among residents aged 40 and above in Liaoning Province, thereby enhancing the generalizability and validity of the findings. However, this study also has several limitations.

First, this study employed a cross-sectional design to explore the relationship between CVAI levels and the prevalence of diabetes, which inherently limits the ability to establish causal inferences. Second, the diagnosis of diabetes in this study lacked 2-hour postprandial glucose test data, which may have led to the underdiagnosis of some individuals and an overall incomplete diagnostic assessment. Future research should consider adopting a prospective cohort design to further validate the longitudinal correlation between CVAI and diabetes.

4 小结

CVAI serves as an effective predictor for diabetes, as elevated CVAI levels are closely associated with increased diabetes risk. This study offers new perspectives and methodologies for the early prevention and control of diabetes; specifically, CVAI enables the earlier identification of at-risk individuals, facilitating timely intervention and reducing the overall burden of the disease. Consequently, the assessment and management of CVAI should be prioritized within diabetes prevention and healthcare strategies.

Furthermore, future research should build upon these findings to explore the relationship between CVAI and other chronic diseases, as well as to evaluate the effectiveness and feasibility of various intervention strategies and clinical management approaches targeting CVAI.

Author Contributions: Xiaohe Wang proposed the research concept, designed the study protocol, formulated the research propositions (including specific viewpoints or methodologies), and drafted the manuscript. Han Yan, Li Jing, and Yuanmeng Tian were responsible for conducting the experiments and implementing the research process, including performing trials or surveys, selecting subjects, collecting samples, and performing laboratory testing and detection. Yiheng Zhou and Xiaoxia An were responsible for data collection, cleaning, statistical analysis, and the creation of figures and tables. Guangxiao Li and Shuang Liu were responsible for the initial drafting of the paper. Liying Xing was responsible for the final revision and serves as the guarantor for the manuscript.

The authors declare no conflicts of interest.

参考文献

SUN H，SAEEDI P, KARURANGA S, et al. IDF Diabetes Atlas:

Global, regional and country-level diabetes prevalence estimates for 2021 and projections for 2045[J]. Diabetes Res Clin Pract, 2022, 183:

XU Y, WANG L M, HE J, et al. Prevalence and control of diabetes in Chinese adults[J]. JAMA, 2013, 310(9): 948-959. DOI: 10.1001/ jama.2013.168118.

YU Junmei, XIONG Huali, SHU Qiang, et al. Prevalence and influencing factors of prediabetes among residents aged 40–79 years in Rongchang District, Chongqing, 2022 [J]. Practical Preventive Medicine, 2024, 31(2): 133-137.

Abstract

Objective To investigate the prevalence of prediabetes and its associated influencing factors among residents aged 40–79 years in Rongchang District, Chongqing, in 2022, providing a scientific basis for the prevention and control of diabetes.

Methods Using a multi-stage stratified cluster random sampling method, a total of 5,851 permanent residents aged 40–79 in Rongchang District were selected for the study. Data were collected through questionnaire surveys, physical examinations, and laboratory testing. The prevalence of prediabetes was calculated, and the influencing factors were analyzed using both univariate analysis and multivariate logistic regression.

Results The prevalence of prediabetes among the study population was 24.13% (standardized prevalence: 23.45%). Multivariate logistic regression analysis indicated that older age (OR=1.026, 95% CI: 1.017–1.035), hypertension (OR=1.428, 95% CI: 1.258–1.621), dyslipidemia (OR=1.352, 95% CI: 1.193–1.531), and central obesity (OR=1.233, 95% CI: 1.082–1.405) were independent risk factors for prediabetes. Conversely, a high level of physical activity (OR=0.793, 95% CI: 0.672–0.936) was identified as a protective factor.

Conclusion The prevalence of prediabetes among residents aged 40–79 in Rongchang District is relatively high. Targeted interventions focusing on weight management, blood pressure and lipid control, and increasing physical activity should be implemented to reduce the risk of progression from prediabetes to clinical diabetes.

Introduction

Diabetes mellitus has become a major global public health challenge, with its prevalence increasing annually. Prediabetes, a state

YU J M, XIONG H L, SHU Q, et al. Prevalence of pre-diabetes mellitus and its related influencing factors among people aged 40- 79 years in Rongchang District of Chongqing Municipality, 2022[J].

Practical Preventive Medicine, 2024, 31(2): 133-137.

Correlation Analysis Between Chinese Visceral Adiposity Index and the Risk of Diabetes in Elderly Chinese Based on CHARLS Data

Authors: Zhan Bowen, Yang Hongguang, Deng Guifang, et al.
Journal: Modern Preventive Medicine (2023)

Abstract

Objective: To investigate the correlation between the Chinese Visceral Adiposity Index (CVAI) and the risk of developing diabetes among the elderly population in China, providing a scientific basis for the prevention and control of diabetes in this demographic.

Methods: This study utilized data from the China Health and Retirement Longitudinal Study (CHARLS) conducted in 2011 and 2015. A total of 3,117 participants aged 60 and older who did not have diabetes at baseline were included. The CVAI was calculated based on age, body mass index (BMI), waist circumference (WC), triglycerides (TG), and high-density lipoprotein cholesterol (HDL-C). Participants were divided into four groups (Q1–Q4) based on CVAI quartiles. Cox proportional hazards regression models were employed to analyze the relationship between CVAI and the risk of diabetes onset.

Results: During the 4-year follow-up period, 432 new cases of diabetes were identified, resulting in a cumulative incidence rate of 13.86%. After adjusting for potential confounders such as gender, age, education level, smoking status, and alcohol consumption, the Cox regression analysis revealed that the risk of diabetes increased with higher CVAI levels. Compared to the lowest quartile (Q1), the hazard ratios (HR) and 95% confidence intervals (CI) for groups Q2, Q3, and Q4 were 1.42 (1.02–1.98), 1.85 (1.34–2.56), and 2.48 (1.81–3.40), respectively ($P_{trend} < 0.001$). Subgroup analyses indicated that the positive correlation between CVAI and diabetes risk remained consistent across different genders and age groups.

Conclusion: A high CVAI is significantly associated with an increased risk of diabetes in the elderly Chinese population. CVAI may serve as a valuable clinical indicator for identifying individuals at high risk of diabetes.

Introduction

Diabetes mellitus has become a major global public health challenge

ZHAN B W, YANG H G, DENG G F, et al. Correlation analyses between Chinese visceral adipose index and the risk of diabetes mellitus in the elderly base on China Health and Retirement Longitudinal Study[J]. Modern Preventive Medicine, 2024, 51(2): 216- American Diabetes Association Professional Practice Committee. 2. diagnosis and classification of diabetes: standards of care in diabetes- 2024[J]. Diabetes Care, 2024, 47(Suppl 1): S20-S42. DOI: 10.2337/ dc24-S002.

Chinese Diabetes Society. Guideline for the prevention and treatment of type 2 diabetes mellitus in China (2020 edition) (Part 1) [J]. Chinese Journal of Practical Internal Medicine, 2021, 41(8): 668-695. DOI: 10.19538/j.issn.1005-2216.2021.08.006. Joint Committee for the Revision of Guidelines for the Prevention and Treatment of Hypertension in China, Hypertension League (China), Chinese Society of Cardiology, Hypertension Professional Committee of the Chinese Medical Doctor Association, et al. Guidelines for the prevention and treatment of hypertension in China (2018 revision) [J]. Chinese Journal of Cardiovascular Medicine, 2019, 24(1): 24-56.

Joint Committee for the Revision of Guidelines for the Prevention and Treatment of Dyslipidemia in Chinese Adults, ZHU Junren, GAO Runlin, et al. Guidelines for the prevention and treatment of dyslipidemia in Chinese adults (2016 revision) [J].

Chinese Journal of General Practitioners, 2017, 16(1): 15-35. DOI: 10.3760/cma.j.issn.1671-7368.2017.01.006. XIA M F, CHEN Y, LIN H D, et al. An indicator of visceral adipose dysfunction to evaluate metabolic health in adult Chinese [J]. Sci Rep, 2016, 6: 38214. DOI: 10.1038/srep38214.

JIANG K, LUAN H, PU X L, et al. Association between visceral adiposity index and insulin resistance: a cross-sectional study based on US adults[J]. Front Endocrinol(Lausanne), 2022, 13: 921067. DOI: 10.3389/fendo.2022.921067.

ANOOP S S, DASGUPTA R, REBEKAH G, et al. Lipid accumulation product(LAP)as a potential index to predict risk of insulin resistance

Chinese General Practice https in young, non-obese Asian Indian males from Southern India: observations from hyperinsulinemic-euglycemic clamp studies[J].

BMJ Open Diabetes Res Care, 2021, 9(1): e002414. DOI: 10.1136/ bmjdrc-2021-002414.

XIA M F, LIN H D, CHEN L Y, et al. Association of visceral adiposity and its longitudinal increase with the risk of diabetes in Chinese adults: a prospective cohort study[J]. Diabetes Metab Res Rev, 2018, 34(7): e3048. DOI: 10.1002/dmrr.3048.

WEI J X, LIU X, SHI Z M. Relationships of visceral adiposity index and body shape index with diabetes mellitus risks in Chinese adults(P21-037-19)[J]. Curr Dev Nutr, 2019, 3: nzz041.P21- nzz0437-19. DOI: 10.1093/cdn/nzz041.P21-037-19.

MOZAFAR SAADATI H, SABOUR S, MANSOURNIA M A, et al.

Effect modification of general and central obesity by sex and age on cardiovascular outcomes: Targeted maximum likelihood estimation in the atherosclerosis risk in communities study[J]. Diabetes Metab JAYEDI A, SOLTANI S, MOTLAGH S Z, et al. Anthropometric and adiposity indicators and risk of type 2 diabetes: systematic review and dose-response meta-analysis of cohort studies[J]. BMJ, 2022, 376: e067516. DOI: 10.1136/bmj-2021-067516.

［19］ NUSRIANTO R, AYUNDINI G, KRISTANTI M, et al. Visceral

adiposity index and lipid accumulation product as a predictor of type 2 diabetes mellitus: The Bogor cohort study of non-communicable diseases risk factors[J]. Diabetes Res Clin Pract, 2019, 155: 107798.

FENG X Y, WANG J Y, WANG S P, et al. Correlation analysis of anthropometric indices and type 2 diabetes mellitus in residents aged years and older[J]. Front Public Health, 2023, 11: 1122509. DOI: 10.3389/fpubh.2023.1122509.

WEI J X, LIU X, XUE H, et al. Comparisons of visceral adiposity index, body shape index, body mass index and waist circumference and their associations with diabetes mellitus in adults[J]. Nutrients, 2019, 11(7): 1580. DOI: 10.3390/nu11071580.

WU J S, GONG L L, LI Q F, et al. A novel visceral adiposity index for prediction of type 2 diabetes and pre-diabetes in Chinese adults: a 5-year prospective study[J]. Sci Rep, 2017, 7(1): 13784. DOI: 10.1038/s41598-017-14251-w.

HAN M H, QIN P, LI Q M, et al. Chinese visceral adiposity index: a reliable indicator of visceral fat function associated with risk of type 2 diabetes[J]. Diabetes Metab Res Rev, 2021, 37(2): e3370. DOI: 10.1002/dmrr.3370.

WANG L M, PENG W, ZHAO Z P, et al. Prevalence and treatment of diabetes in China, 2013-2018[J]. JAMA, 2021, 326(24): 2498-2506.

LIN M Y, XIE Y D, HONG L K, et al. Abnormal adiposity indices are associated with an increased risk of diabetes in a non-obese Asian population[J]. Public Health, 2024, 226: 144-151. DOI: 10.1016/ (Received: 2024-08-21; Revised: 2025-11-19) (Editor: Wang Shiyue)

Abstract

Objective To investigate the association between various adiposity indices and the risk of developing diabetes in a non-obese Asian population, and to evaluate the predictive value of these indices for diabetes onset.

Methods This study utilized a longitudinal cohort design. Data were collected from a non-obese Asian population (Body Mass Index, BMI < 25 $kg/m^2$). Adiposity indices, including Body Roundness Index (BRI), Visceral Adiposity Index (VAI), Lipid Accumulation Product (LAP), and others, were calculated based on anthropometric measurements and biochemical markers. Cox proportional hazards models were employed to estimate the hazard ratios (HRs) and 95% confidence intervals (CIs) for the association between these indices and diabetes incidence. Area under the receiver operating characteristic curve (AUC) was used to compare the predictive performance of different indices.

Results During the follow-up period, a significant number of participants developed diabetes. After adjusting for potential confounders, elevated levels of BRI, VAI, and LAP were significantly associated with an increased risk of diabetes. Specifically, participants in the highest quartile of these indices exhibited a substantially higher risk compared to those in the lowest quartile. Subgroup analyses indicated that these associations remained robust across different age groups and genders. Furthermore, the predictive models incorporating these adiposity indices showed superior performance compared to models using BMI alone.

Conclusion Abnormal adiposity indices are independent risk factors for diabetes in non-obese Asian individuals. These indices may serve as valuable tools for early screening and risk assessment of diabetes in populations where traditional BMI measurements may underestimate metabolic risk.

Introduction

The global prevalence of diabetes mellitus continues to rise, posing a significant challenge to public health systems worldwide. While obesity, typically defined by a high Body Mass Index (BMI), is a well-established risk factor for type 2 diabetes, a substantial proportion of the Asian population develops