Abstract
Background
This study aims to compare the discriminative ability of the Visceral Obesity Index (VOI), body mass index (BMI) and waist circumference (WC) in identifying abnormal glucose metabolism (previously undiagnosed type 2 diabetes and prediabetes) in adults during a single screening visit.
Methods
This cross-sectional observational study included 4,250 individuals aged 25 to 60 years. All analyses were based on data obtained during a single screening visit. Participants were recruited consecutively from outpatient clinics and general medical practices in Moscow (Russian Federation) and Dubai (United Arab Emirates). The methods included anthropometric measurements (BMI, WC) and laboratory biochemical tests. Statistical analysis involved univariate and multivariate logistic regression to assess associations with type 2 diabetes, adjusted for age and gender; odds ratios (OR) with 95% confidence intervals (CI) were calculated. A receiver operating characteristic (ROC) curve analysis was performed to compare the diagnostic performance of VOI, BMI and WC. The area under the curve (AUC) was then calculated for each indicator. Type 2 diabetes was diagnosed in accordance with American Diabetes Association (ADA) guidelines.
Results and conclusions
VOI demonstrated a higher AUC compared with BMI and WC at identifying previously undiagnosed type 2 diabetes and prediabetes in adults. VOI may be a useful indicator for identifying individuals with abnormal glucose metabolism in screening settings.
Trial registration
Not applicable.
Clinical trial number
Not applicable.
Keywords: General obesity, Increased body mass, Metabolic syndrome, Obesity, Prediabetes
Introduction
The escalating predominance of type 2 diabetes constitutes a substantial global public health predicament, particularly within populations exhibiting a high prevalence of obesity and abdominal adiposity [1]. In Russia, elevated body mass was found in 67% of men and 62% of women, with 19% and 28%, respectively, classified as obese [2]. Similar trends are observed worldwide [3].
Standard anthropometric indicators such as BMI and WC have long been used to evaluate MetS risk; however, they have limitations in assessing visceral adiposity. More advanced indices—such as the Body Shape Index (BSI) and Body Roundness Index (BRI)—offer improved accuracy by integrating age, gender, height, and waist measures [4–7]. Recent studies have validated the superiority of BRI over BMI and waist-to-height ratio in identifying metabolic risk, particularly in women [4, 8–11].
The VOI, incorporating BMI, WC, triglycerides, and HDL cholesterol, VOI has been proposed as an index associated with metabolic risk and insulin resistance in several populations, including those from China, India, and South Korea [12–15]. Unlike BMI and WC, VOI better reflects internal fat distribution. Therefore, indices like VOI serve as practical screening alternatives. In this study, VOI may provide additional information on visceral adiposity-related risk [16].
Recent studies underscore the critical role of novel anthropometric indices in the identification of abnormal glucose metabolism [7, 8, 17, 18]. Unlike traditional measures such as Body Mass Index (BMI), emerging parameters—such as A Body Shape Index (ABSI), the triglyceride-glucose (TyG) index, fat distribution indices, and waist-to-height ratio—demonstrate superior predictive accuracy. According to data from population-based studies conducted in South Africa and Asia, the use of these indices increased the sensitivity of diabetes detection by 10–25% compared to reliance on BMI alone [15–17, 19]. This is particularly relevant in regions with a high prevalence of abdominal obesity and metabolic syndrome.
The aim of this study was to compare the discriminative ability of VOI, BMI, and WC in identifying abnormal glucose metabolism (previously undiagnosed type 2 diabetes and prediabetes) in adults during a single screening visit. This study also makes a contribution to the ongoing evaluation of anthropometric and metabolic indices in different populations, and addresses the need for further validation in a variety of clinical settings [20–22].
Materials and methods
Study design and population
This cross-sectional observational study aimed to evaluate the discriminative ability of the VOI, BMI and WC to identify abnormalities in glucose metabolism, including undiagnosed type 2 diabetes and prediabetes, in adults during a single screening visit.
As all analyses were performed using baseline data collected at a single time point, the study assessed the ability of the indices to identify existing metabolic abnormalities rather than predict future disease progression.
The study population consisted of adults aged 25 to 60 years, recruited from outpatient clinics and general practitioner offices in Moscow (Russian Federation) and Dubai (United Arab Emirates). After applying predefined inclusion and exclusion criteria, a total of 4,250 participants were included in the final analysis. Recruitment was based on self-consent and eligibility according to predefined criteria.
The primary objective was to compare the ability of VOI, BMI and WC to discriminate between individuals with and without glucose metabolism abnormalities (fasting hyperglycaemia, glucose intolerance and newly diagnosed type 2 diabetes). The secondary objective was to assess the strength of the association between the VOI and metabolic abnormalities, adjusting for age and sex.
Individuals with a previous diagnosis of type 2 diabetes were excluded prior to enrolment. All participants underwent a standardised baseline assessment, which included anthropometric measurements, fasting blood sampling and an oral glucose tolerance test (OGTT).
The final sample included 1,870 women and 2,380 men. The cohort consisted of an urban outpatient population rather than a random sample of the general population.
Ethnicity and population characteristics
The study population predominantly consisted of individuals of European and Middle Eastern origin, reflecting the demographic composition of urban outpatient populations in Moscow and Dubai. Information on specific ethnic subgroups (e.g. South Asian, Pacific Islander or indigenous populations) was not systematically collected during recruitment.
As ethnicity is a recognised factor in the risk of type 2 diabetes and obesity-related metabolic complications, the generalisability of the results to populations with different ethnic compositions should be interpreted with caution. This limitation is discussed in the ‘Discussion’ section.
Consideration of medication use
To eliminate the potential confounding effects of pharmacological therapy on study outcomes, a detailed medical history regarding current medication use was collected during the questionnaire and screening stages. Participants receiving hormonal treatments or undergoing active therapy for chronic illnesses were excluded from the study. However, some enrolled participants may have been taking basic vitamins, dietary supplements, or symptomatic medications (such as analgesics or antihistamines) that do not directly affect carbohydrate metabolism or overall metabolic status. These data were taken into account during the analysis and interpretation of results, and stratification by the presence or absence of ancillary therapy was performed when necessary.
Type 2 diabetes was diagnosed in accordance with American Diabetes Association (ADA) guidelines. Due to the absence of detailed ethnic subgroup stratification, the findings should be interpreted with caution regarding generalizability to other ethnic populations.
Sample and inclusion/exclusion criteria
A survey protocol was developed and approved, outlining the selection criteria for the study. The sample size was set at a minimum of 5,000 individuals, both men and women, aged 25 to 60 years. Participants were recruited through advertisements in clinics and medical centers. In accordance with the Helsinki Declaration, all participants provided informed consent.
Inclusion criteria
Ages 25–60.
Permanent residence in the Russian Federation or the UAE.
No previous diagnosis of type 2 diabetes.
Written informed consent.
Exclusion criteria
Pregnancy or lactation.
Known endocrine disorders affecting glucose metabolism.
Active oncological disease.
Severe chronic systemic illness.
Use of systemic glucocorticoids or hormonal therapy.
Participants with pre-VOI undiagnosed diabetes detected during the baseline OGTT were included in the analysis as newly diagnosed cases.
The primary outcome was abnormal glucose metabolism identified at baseline screening. This included:
newly diagnosed type 2 diabetes (according to ADA criteria, this is defined as a fasting plasma glucose level of ≥ 7.0 mmol/L or a 2-hour OGTT level of ≥ 11.1 mmol/L);
impaired fasting glucose (defined as a level of 5.6–6.9 mmol/L);
impaired glucose tolerance (defined as a 2-hour OGTT level of 7.8–11.0 mmol/L).
The study evaluated the ability of VOI, BMI and WC to identify existing, but previously undiagnosed, metabolic abnormalities, rather than predicting incident diabetes.
Data collection and blinding
Demographic data (age and sex) and anthropometric measurements (weight, height and waist circumference) were collected during a standardised baseline visit, as were biochemical parameters (fasting glucose, triglycerides, HDL cholesterol and OGTT results).
All measurements were performed by trained personnel who were unaware of the study objectives. Laboratory and anthropometric data were anonymised prior to analysis. The data analysis phase was not blinded.
Anthropometric measurements
Body weight and height were measured using calibrated instruments, and BMI was calculated as weight (kg)/height (m²). Waist circumference was measured at the midpoint between the lower rib margin and iliac crest at the end of normal expiration using a non-elastic tape measure. BMI was categorized according to WHO criteria (≥ 25 kg/m² overweight, ≥ 30 kg/m² obesity).
Biochemical analysis
Venous blood samples were collected following an overnight fast of at least 12 h. Fasting plasma glucose, triglyceride and HDL cholesterol levels were measured using standardised laboratory methods. All participants underwent a 75 g oral glucose tolerance test, with plasma glucose measured at fasting and two hours post-load.
Insulin measurements were only available for a subset of participants and were therefore not included in the primary analysis. Insulin resistance indices, such as HOMA-IR, could not be calculated due to incomplete data availability across the full cohort.
Due to logistical constraints, HbA1c was not measured. Additionally, blood pressure and family history of type 2 diabetes were not systematically collected and should be considered when interpreting the findings.
Risk factor interpretation
The VOI was calculated using previously published formulas that are specific to sex:
For males:
 |
For females:
 |
VOI values were categorised into risk groups according to previously published reference thresholds [23].
Statistical analysis
All statistical analyses were performed using SPSS version 25.0.
Continuous variables are presented as the mean ± standard deviation, while categorical variables are presented as percentages.
Three separate regression models were constructed to avoid mathematical collinearity between VOI and its components:
Model 1: BMI + age + sex; Model 2: waist circumference + age + sex; and Model 3: VOI + age + sex.
Variance inflation factors (VIF) were calculated to confirm the absence of multicollinearity.
Receiver operating characteristic (ROC) curve analysis was performed to compare the discriminative ability of the three anthropometric indicators. The area under the curve (AUC) was calculated for VOI, BMI and WC, and differences between the values were assessed using standard statistical methods.
Age-stratified analyses were additionally performed using the following age groups: <40 years, 40–50 years, and > 50 years. Multivariable logistic regression models were used to estimate adjusted odds ratios (ORs) with 95% confidence intervals.
A p-value of less than 0.05 was considered statistically significant.
Ethical considerations
The research protocol was approved by the Ethics Committee of the Sechenov First Moscow State Medical University (Protocol No. 4565 of 16.04.2022). All procedures were performed in accordance with the Declaration of Helsinki. The participants gave informed consent to both participate in the study and for the use of their data for scientific purposes.
Results
Screening results
Based on the application of the above criteria, the following participants were excluded:
Individuals with hereditary disorders: 25;
Individuals with chronic diseases: 78;
Individuals with psychiatric or neurological conditions: 45;
Pregnant women: 35;
Individuals using hormonal medications: 112;
Individuals with oncological conditions: 37.
Thus, 4,250 individuals remained in the study: 44% (1,870) were females and 56% (2,380) were males. The youngest participant was 25 years old, while the oldest was 60 years old. Standardized methods of anthropometric parameter measurement were employed to ensure the accuracy of the data. The units of measurement for triglyceride (mmol/L), high-density cholesterol (mmol/L), and fasting glucose (mmol/L) measurements are provided in the data table.
Association between anthropometric indices and abnormal glucose metabolism
Multivariable logistic regression analysis was performed to assess the association between anthropometric indicators and abnormal glucose metabolism (newly diagnosed type 2 diabetes, impaired fasting glucose, and impaired glucose tolerance).
To avoid mathematical collinearity between the Visceral Obesity Index (VOI) and its constituent elements, three distinct models were meticulously constructed.
In Model 1 (which included BMI, age, and sex as independent variables), overweight and obesity were significantly associated with abnormal glucose metabolism (OR = 1.10; 95% CI: 1.05–1.15; p = 0.002).
In Model 2 (waist circumference + age + sex), elevated waist circumference was also significantly associated with abnormal glucose metabolism (OR = 1.12; 95% CI: 1.07–1.17; p < 0.001).
In Model 3 (VOI + age + sex), VOI demonstrated the strongest association among the three anthropometric indicators (OR = 1.25; 95% CI: 1.15–1.35; p < 0.001).
The findings demonstrated that age remained a robust independent predictor, with an odds ratio (OR) of 1.30 for every 10-year increase (95% confidence interval [CI]: 1.22–1.39; p < 0.001). Additionally, male sex exhibited a statistically significant yet comparatively weaker association (OR = 1.08; 95% CI: 1.01–1.15; p = 0.030).
Anthropometric measurements
The BMI results are presented in Table 1.
Table 1.
BMI and categorization of participants, according to the increasing risk of type 2 diabetes occurrence
| Weight Category | BMI (kg/m²) | Total Number of Participants (4250) | Men (2380) | Women (1870) | Fasting Glucose (Median, IQR) | Triglycerides (Median, IQR) | HDL- Cholesterol (Median, IQR) | Insulin (Median, IQR) | Correlation with VOI (r) |
|---|---|---|---|---|---|---|---|---|---|
| Underweight | < 18.5 | 176 (4%) | 119 (5%) | 57 (3%) | 4.2 (3.8–4.6) | 1.1 (0.9–1.3) | 1.7 (1.5–1.9) | 8.5 (7.8–9.2) | 0.34 |
| Healthy | 18.5–24.9 | 2944 (69%) | 1523 (64%) | 1421 (76%) | 5.1 (4.9–5.4) | 1.4 (1.2–1.6) | 1.5 (1.4–1.7) | 9.2 (8.7–9.8) | 0.52 |
| Overweight | 25.0-29.9 | 700 (16%) | 476 (20%) | 224 (12%) | 5.8 (5.5–6.2) | 1.9 (1.7–2.2) | 1.3 (1.2–1.5) | 10.0 (9.5–10.5) | 0.62 |
| Obese | 30.0–40.0 | 430 (10%) | 262 (11%) | 168 (9%) | 6.5 (6.2–6.8) | 2.4 (2.1–2.7) | 1.1 (1.0-1.3) | 11.0 (10.5–11.5) | 0.70 |
The analysis showed that 26% of participants (1130 individuals) had excess weight and obesity, with 31% being men (738 individuals) and 21% being women (392 individuals).
The results of WC measurements were categorized and presented in Table 2.
Table 2.
Waist circumference (WC) and categorization of participants, according to increasing risk of type 2 diabetes
| Risk of Diabetes According to WC | WC Parameters for Men and Women (cm) |
Total Number of Participants 4250 |
|---|---|---|
| Low Risk – 63% (2690 individuals) | Men ≤ 94 | 66% (1568) |
| Men 94–101 cm | 15% (360) | |
| Women ≤ 80 | 60% (1122) | |
| Women 80–87 cm | 18% (365) | |
| Elevated and High Risk – 37% (1560 individuals) | Men ≥ 102 | 34% (812) |
| Women ≥ 88 | 40% (748) |
The WO measurements revealed that 37% of the participants (1560 individuals) were at an elevated or high risk of developing type 2 diabetes. This is 10% higher than the risk identified by the BMI.
Biochemical analyses
Table 3 presents biochemical test results, including triglyceride levels and HDL cholesterol levels.
Table 3.
Triglyceride levels* and categorization of participants according to increasing risk of type 2 diabetes
| Blood Triglyceride Level (mmol/L) | Risk of Type 2 Diabetes | Total Number of Participants (4250) | Men (2380) | Women (1870) |
|---|---|---|---|---|
| Normal (0.2–1.7) | Low | 545 (13%) | 430 (10%) | 115 (3%) |
| Borderline High (1.70–2.25) | Elevated | 630 (15%) | 270 (6%) | 360 (8%) |
| High (2.26–5.65) | High | 2020 (48%) | 1120 (26%) | 900 (21%) |
| Very High (> 5.65) | High | 1055 (25%) | 560 (13%) | 495 (12%) |
Note: *Reference values of triglyceride levels and variations in values during pathological conditions were obtained from the source titled “2019 ESC/EAS Guidelines for the management of dyslipidaemias: Lipid modification to reduce cardiovascular risk.”
An analysis of triglyceride levels showed that 72% of the participants were at risk of developing type 2 diabetes, with 48% having elevated levels and 25% having high levels. The HDL cholesterol levels showed that 69% of individuals were at risk for type 2 diabetes, with a notably higher percentage of men (40%) compared to women (29%, Table 4).
Table 4.
HDL cholesterol level* and categorization of participants according to increasing risk of type 2 diabetes
| HDL Cholesterol Level (mmol/L) | Risk of Type 2 Diabetes | Total Number of Participants (4250) | Men (2380) | Women (1870) |
|---|---|---|---|---|
| Normal | Low | 688 (16%) | 542 (14%) | 146 (8%) |
| Elevated | Elevated | 642 (15%) | 380 (16%) | 262 (14%) |
| Low | High | 1692 (40%) | 1120 (47%) | 572 (31%) |
| Very Low | High | 1228 (29%) | 338 (14%) | 890 (48%) |
Note: *The reference values for high-density lipoprotein (HDL) cholesterol levels and variations in values during pathological conditions were obtained from the source titled “2019 ESC/EAS Guidelines for the management of dyslipidaemias: Lipid modification to reduce cardiovascular risk.”
Oral glucose tolerance test results
Table 5 shows the OGTT results. The OGTT revealed 60% of the participants (2561 individuals) had impaired glucose tolerance, and 17% were diagnosed with type 2 diabetes. IGT was defined as a 2-hour plasma glucose level between 7.8 and 11.0 mmol/L during the OGTT, in accordance with ADA criteria.
Table 5.
OGTT results and categorization of participants according to increasing risk of type 2 diabetes
| Reference Values for Glucose, Venous Blood, mmol/L: | Total Number of Participants 4250 | Men 2380 (56%) |
Women 1870 (44%) |
|---|---|---|---|
|
Impaired fasting glycemia: Fasting: ≥ 6.1 and < 7.0 2 h after: < 7.8 |
258 (6%) | 143 (6%) | 115 (7%) |
|
Normal values: Fasting: < 6.1 2 h after: < 7.8 |
1431 (34%) | 1071 (45%) | 360 (50%) |
|
Impaired glucose tolerance: Fasting: < 7.0 2 h after: ≥ 7.8 and < 11.1 |
1828 (43%) | 928 (39%) | 900 (31%) |
|
Diabetes mellitus: Fasting: ≥ 7.0 2 h after: ≥ 11.1 |
733 (17%) | 238 (10%) | 495 (12%) |
VOI
Table 6 shows data on the VOI among the participants. According to the VOI, 48% (2020 individuals) of the participants of the total study population were at high risk of developing type 2 diabetes. Among them, there were 1120 men and 900 women.
Table 6.
Index (VOI) results and the level of risk for developing type 2 diabetes
| Risk of Diabetes According to VOI | Total Number of Participants 4250 | Men 2380 (56%) |
Women 1870 (44%) |
|---|---|---|---|
|
Normal = 1.0-2.5 |
545 (13%) | 430 (10%) | 115 (3%) |
|
Elevated 2.6–3.5 |
1055 (25%) | 560 (13%) | 495 (12%) |
|
High > 3.5 |
2020 (48%) | 1120 (26%) | 900 (21%) |
| Not classified (missing VOI data) | 630 (15%) | 270 (11%) | 360 (19%) |
VOI categories were defined according to previously published reference values [23].
According to the VOI, 48% (2020 individuals) of the participants were at a high risk of developing type 2 diabetes. Among them, there were 1120 men and 900 women.
Distribution of participants classified as high risk
Table 7 shows the proportion of participants categorised as high risk based on the assessed anthropometric and biochemical indicators at the start of the study. These descriptive findings illustrate the distribution of risk classifications within the study population, rather than providing information on discriminatory or predictive performance.
Table 7.
Proportion of participants classified as high risk by different screening methods*
| Diagnostic Parameter | Number of Individuals with High Risk | Number of Men | Number of Women |
|---|---|---|---|
| BMI | 1130 (26%) | 738 (31%) | 329 (18%) |
| Waist Circumference | 1560 (37%) | 812 (40%) | 748 (40%) |
| Triglycerides | 3075 (73%) | 1680 (39%) | 1395 (33%) |
| HDL Cholesterol | 2933 (69%) | 1692 (40%) | 1228 (29%) |
| VOI (high category) | 2020 (48%) | 1120 (47%) | 900 (48%) |
| Glucose Tolerance Test (Impaired Glucose Tolerance + Diabetes) | 2561 (60%) | 1166 (49%) | 1395 (43%) |
*Table 7 presents the number and proportion of participants classified as high risk according to each screening method at baseline. These results reflect risk classification and should not be interpreted as incident diabetes outcomes, as follow-up data on incident diabetes were not collected in this study
Figures 1 and 2 show how participants were distributed across the high-risk categories according to the indicators evaluated. Within this study population, VOI classified a greater proportion of participants as high risk than BMI and waist circumference did.
Fig. 1.
Comparison of diagnostic anthropometric indices and combined VOI (Visceral Obesity Index)
Fig. 2.
Comparison of biochemical parameters
Diagnostic performance of anthropometric indicators
ROC curve analysis revealed that VOI exhibited the greatest discriminatory capacity (AUC = 0.84), surpassing waist circumference (AUC = 0.78) and BMI (AUC = 0.76). Although the magnitude of improvement was modest, the differences between VOI and the other indices were statistically significant. A statistical comparison of the area under the curve (AUC) values revealed that the VOI exhibited a significantly superior performance in comparison to BMI (p < 0.05). The discrepancy between VOI and waist circumference decreased, yet it remained statistically significant (p < 0.05). No statistically significant difference was observed between BMI and waist circumference (p = 0.12).
Age-stratified analysis
Subsequent analyses were conducted in three distinct age groups: under 40 years, 40–50 years, and over 50 years.
In all age groups, higher VOI values were associated with a higher prevalence of abnormal glucose metabolism. However, the strength of the association increased with age, reaching its most pronounced levels among participants over 50 years of age.
Logistic regression model results
In addition to the ROC analysis, a logistic regression was performed to quantify the contribution of individual predictors to the risk of developing type 2 diabetes mellitus. The results are summarized in Table 8.
Table 8.
Results of the logistic regression model predicting the risk of type 2 diabetes mellitus
| Variable | OR adjusted | 95% CI | p-value |
|---|---|---|---|
| VOI (high level) | 1.25 | 1.15–1.35 | < 0.001 |
| BMI (≥ 25 kg/m²) | 1.10 | 1.05–1.15 | 0.002 |
| WC (elevated category) | 1.12 | 1.07–1.17 | < 0.001 |
| Triglycerides (> 1.7 mmol/L) | 1.18 | 1.09–1.27 | < 0.001 |
| HDL (reduced < 1.0 in men, < 1.2 in women) | 1.20 | 1.12–1.28 | < 0.001 |
| Age (per 10-year increment) | 1.30 | 1.22–1.39 | < 0.001 |
| Male sex | 1.08 | 1.01–1.15 | 0.030 |
The model demonstrated good calibration (Hosmer–Lemeshow test: p = 0.42) and high discriminative performance (AUC = 0.87). All anthropometric and biochemical variables were significantly associated with abnormal glucose metabolism. Among the anthropometric indicators, VOI demonstrated the strongest association; however, the magnitude of the differences between the indicators was moderate.
The odds ratio (OR) for VOI in predicting type 2 diabetes was 1.25 (95% CI: 1.15–1.35); for BMI, it was 1.10 (95% CI: 1.05–1.15), and for WC, it was 1.12 (95% CI: 1.07–1.17). These findings indicate a significant association between all three indicators and the risk of developing type 2 diabetes, with VOI having the greatest impact. These findings suggest a strong link between all three indicators and abnormal glucose metabolism, with VOI showing the most significant association among the anthropometric indices. The observed differences in discriminatory performance were modest yet statistically significant. Therefore, VOI may offer supplementary discriminatory information compared to BMI and WC in cross-sectional screening settings. The present study design does not allow for causal inference or prediction of future diabetes.
Discussion
VOI showed the highest discriminatory performance in a cross-sectional setting among the evaluated anthropometric indices in this study. The present study found that 48% of participants were classified as high-risk according to VOI, compared to 26% by BMI and 37% by waist circumference. These results underscore the added value of integrative indices combining anthropometric and metabolic parameters.
The novelty of this study lies in the comprehensive comparative assessment of the associative strength of the Visceral Obesity Index (VOI), Body Mass Index (BMI), and Waist Circumference (WC) for the early identification of type 2 diabetes risk. In contrast to most existing studies, this research was conducted using a design involving two geographically and ethnically distinct cohorts (Russia and the United Arab Emirates), thereby allowing comparison between two geographically distinct urban populations. Furthermore, the study employs a standardised approach to calculating VOI, which has not previously been applied to such a large sample (n = 4,250) in the context of type 2 diabetes risk assessment, incorporating both biochemical parameters and oral glucose tolerance test (OGTT) results. The findings offer a more precise evaluation of VOI as a screening tool and provide a strong rationale for its integration into routine clinical practice.
The present study provides a comprehensive assessment of various anthropometric and biochemical indicators used to evaluate the risk of type 2 diabetes among a large adult population. The main findings demonstrate that traditional indicators such as BMI and WC identify a substantial proportion of individuals at elevated risk (26% and 37%, respectively), while more sensitive and integrative indices, particularly the VOI, identify an even higher proportion—48% of participants.
In ROC analysis, VOI demonstrated the highest discriminatory performance among anthropometric indices (AUC = 0.84), compared with waist circumference (AUC = 0.78) and BMI (AUC = 0.76). However, the differences between the indices were modest yet statistically significant. These findings suggest that VOI offers more discriminatory information than conventional anthropometric measures when assessing abnormal glucose metabolism in cross-sectional studies. These results reflect cross-sectional associations between anthropometric indices and abnormal glucose metabolism and should not be interpreted as evidence of future risk prediction.
Additionally, the OGTT revealed a high prevalence of carbohydrate metabolism disorders: impaired glucose tolerance (defined as 2-hour OGTT glucose 7.8–11.0 mmol/L) was observed in 60% of participants, and 17% were newly diagnosed with type 2 diabetes. This underscores the importance of early screening, especially given that many of these individuals may remain undiagnosed if relying solely on conventional metrics such as BMI. IGT was included as a diagnostic category in accordance with ADA criteria, being defined as an OGTT glucose level of between 7.8 and 11.0 mmol/L over a period of two hours.
The biochemical analyses supported these findings. High levels of triglycerides and low levels of HDL cholesterol were observed in the majority of participants, with 72% and 69%, respectively, at increased risk of developing diabetes. These metabolic disturbances were more pronounced in men, who consistently demonstrated higher levels of triglycerides and lower HDL values, highlighting the gender-specific nuances in diabetes risk profiles.
Finally, logistic regression analysis further strengthened the role of VOI as a robust predictor. Adjusted odds ratios showed that VOI had the highest associative strength (OR = 1.25; 95% CI: 1.15–1.35), compared to BMI (OR = 1.10) and WC (OR = 1.12). These findings reinforce the need to adopt VOI and similar complex indices in clinical and public health settings for large-scale screening and early prevention strategies.
Our findings corroborate earlier studies demonstrating the superior diagnostic utility of novel anthropometric indices compared with conventional BMI. For instance, in European and African cohorts, measures such as the A Body Shape Index (ABSI), the waist-to-height ratio, and waist circumference have been shown to correlate strongly with insulin resistance and hyperglycemia (p < 0.01), rendering them effective screening tools [18, 19, 24–26]. Moreover, the application of data-mining techniques based on anthropometric parameters increased the accuracy of diabetes prediction to 85–90%, opening avenues for the development of individualized risk-stratification and prevention algorithms [25].
This study compares the efficacy of various anthropometric measures in identifying the risk of developing type 2 diabetes. The BMI, a parameter developed by Adolphe Quetelet in 1835, has been the most commonly used measure for many years. Despite its frequent applicability, an increase in BMI does not always indicate excess weight. In some ethnic groups, a BMI of < 25 kg/m² may be unhealthy, although it is defined as normal by WHO guidelines. Thus, for Finnish residents, the specified BMI is not considered normal due to the redistribution of adipose tissue in the body. Additionally, BMI measurement does not account for gender and age [22, 27].
According to a study involving 4250 individuals in the UAE, overweight and obesity based on BMI criteria were observed in 26% of participants, with 31% being male and 21% female, consistent with global trends. While BMI is widely used, its limitations are increasingly recognized. For example, a Canadian guideline published in 2021 recommended abandoning BMI due to its low sensitivity in predicting obesity-related complications [28].
WC is considered a more reliable indicator of type 2 diabetes risk by many researchers, as it provides additional information beyond BMI [29]. However, some argue that WC does not directly measure subcutaneous or visceral fat [28]. Despite this, WC measurement demonstrated high sensitivity in identifying diabetes risk, with 37% of participants classified as having increased or high risk based on WC. Combining BMI and WC improved risk detection by approximately 10%.
Recent studies also emphasize the value of VOI in diagnosing diabetes risk due to their strong correlation with insulin resistance [19, 26, 30, 31]. VOI has shown comparable effectiveness to insulin resistance tests, which are often impractical for routine screening [24, 25]. For instance, a 2021 study by [29] found a direct correlation between VOI and arterial elasticity in heart failure patients, highlighting the cardiovascular implications of visceral fat accumulation [29].
The oral glucose tolerance test remains a simple and effective method for detecting diabetes risk, unaffected by factors such as athleticism, unlike BMI or insulin levels. This test, combined with anthropometric indices such as WC and VOI, may improve early identification and prevention strategies for type 2 diabetes [32, 33].
The current study compared the results of BMI, WC, and VOI. The findings show that VOI demonstrated the highest discriminatory performance among anthropometric indices. The VOI method identified 6% more people at risk than with waist circumference measurement and 16% more than with BMI. Therefore, among individuals aged 25 to 60 years, both men and women, VOI was the most effective in detecting the risk of developing type 2 diabetes. The high effectiveness of VOI is attributed to its calculation formula, which includes various parameters such as gender, anthropometric indicators, and biochemical markers.
For a long time, BMI has been considered the most efficient tool. Waist circumference indirectly measures visceral fat, making the combination of the two methods even more effective. However, the VOI exhibited the highest level of ability to detect the risk of developing type 2 diabetes.
The absence of categorization of participants by age groups is a limitation of this study. This omission may potentially impact the reliability of the findings, as the likelihood of developing type 2 diabetes can vary substantially across different age brackets. In the future, it is planned to expand this study by dividing the study groups based on age criteria. A more granular division into age cohorts would allow for a better understanding of age-related characteristics and potential variations in results. This, in turn, could enhance the interpretation of data and increase its practical significance.
A further limitation of the study is the use of a convenience sample, recruited through advertisements. Individuals who respond to such invitations may differ systematically from non-responders in terms of health awareness, lifestyle behaviours, or metabolic risk profiles. This has the potential to lead to an over- or underestimation of risk prevalence, which may, in turn, limit the generalizability of the findings to the broader adult population. Future studies should consider implementing population-based sampling techniques to enhance the external validity of the research.
Due to the cross-sectional design of this study, the observed associations reflect the current relationship between anthropometric indices and abnormal glucose metabolism. Therefore, no conclusions can be drawn regarding temporal prediction or incident diabetes.
Conclusion
The present study demonstrated that, among the evaluated anthropometric indices, VOI showed the highest discriminatory performance in a cross-sectional setting in identifying abnormal glucose metabolism in a cross-sectional setting. When combined with biochemical parameters, VOI provided additional information on metabolic risk compared with BMI and waist circumference. The observed differences were modest, and the present study design does not allow for causal inference or predictive performance to be established. These findings should be interpreted in light of several limitations, including the lack of detailed ethnic stratification, the cross-sectional design and the absence of longitudinal follow-up. VOI may be considered a useful adjunct marker for identifying individuals with abnormal glucose metabolism in similar urban clinical populations, although generalisability to other ethnic groups remains limited due to the lack of detailed ethnic stratification. However, the lack of detailed stratification by ethnic subgroup limits the generalisability of the findings to broader, more ethnically diverse populations.
Acknowledgements
Not applicable.
Author contributions
All authors contributed to the study conception and design. Material preparation, data collection and analysis were performed by EK, AS, DM. The first draft of the manuscript was written by ES. All authors read and approved the final manuscript.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
Data availability
All data generated or analysed during this study are included in this published article.
Declarations
Ethical approval
All procedures were performed in accordance with the Declaration of Helsinki. The research protocol was approved by the Ethics Committee of the Sechenov First Moscow State Medical University (Protocol No. 4565 of 16.04.2022).
Consent to participate
The participants gave written informed consent to both participate in the study and for the use of their data for scientific purposes.
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Vera-Ponce VJ, Zuzunaga-Montoya FE, Vásquez-Romero LEM, Loayza-Castro JA, Paucar CRI, De Carrillo CIG, et al. Anthropometric measures of obesity as risk indicators for prediabetes. A systematic review and meta-analysis. Diabet Epidemiol Manag. 2024;16:100230. 10.1016/j.deman.2024.100230. [Google Scholar]
- 2.Martinchik AN, Laikam KE, Kozyreva NA, Keshabyants EE, Mikhailov NA, Baturin AK, et al. The prevalence of obesity in various socio-demographic groups of the population of Russia. Vopr Pitan. 2021;90:535. 10.33029/0042-8833-2021-90-3-67-76. [DOI] [PubMed] [Google Scholar]
- 3.Sadeghi E, Khodadadiyan A, Hosseini SA, Hosseini SM, Aminorroaya A, Amini M, et al. Novel anthropometric indices for predicting type 2 diabetes mellitus. BMC Public Health. 2024;24:1033. 10.1186/s12889-024-18541-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Amiri P, Javid AZ, Moradi L, Haghighat N, Moradi R, Behbahani HB, et al. Associations between new and old anthropometric indices with type 2 diabetes mellitus and risk of metabolic complications: a cross-sectional analytical study. J Vasc Bras. 2021;20:e20200236. 10.1590/1677-5449.200236. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Golabi S, Ajloo S, Maghsoudi F, Adelipour M, Naghashpour M. Associations between traditional and non-traditional anthropometric indices and cardiometabolic risk factors among inpatients with type 2 diabetes mellitus: a cross-sectional study. J Int Med Res. 2021;49:03000605211049960. 10.1177/03000605211049960. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Klisić A, Radoman Vujačić I, Kostadinović J, Patoulias D, Ninić A. Novel anthropometric parameters in the adult population with prediabetes. Eur Rev Med Pharmacol Sci. 2023;27:11063–72. 10.26355/eurrev_202311_34475. [DOI] [PubMed] [Google Scholar]
- 7.Konstantinova ED, Maslakova TA, Ogorodnikova SY. The predictive capability of several anthropometric indices for identifying the risk of metabolic syndrome and its components among industrial workers. Sci Rep. 2024;14:15327. 10.1038/s41598-024-66262-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Gesteiro E, Megía A, Guadalupe-Grau A, Fernandez-Veledo S, Vendrell J, González-Gross M. Early identification of metabolic syndrome risk: a review of reviews and proposal for defining pre-metabolic syndrome status. Nutr Metab Cardiovasc Dis. 2021;31:2557–74. 10.1016/j.numecd.2021.05.022. [DOI] [PubMed] [Google Scholar]
- 9.Guo X, Ding Q, Liang M. Evaluation of eight anthropometric indices for identification of metabolic syndrome in adults with diabetes. Diabetes Metab Syndr Obes. 2021;14:1431–43. 10.2147/DMSO.S294244. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Xu S, Ren R, Li W, Liang Y, Ma J, Zheng Y, et al. The association between obesity indicators and metabolic risk factors in type-2 diabetic patients. Heliyon. 2023;9:e20013. 10.1016/j.heliyon.2023.e20013. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Zalzala MH, Al-khfajy WS, Khaleel RA. Cytotoxic effect of 6-ethyl-chenodeoxycholic acid and cabazitaxel on PC-3 cells. Drug Dev Regist. 2023;12:52–8. 10.33380/2305-2066-2023-12-1-52-58. [Google Scholar]
- 12.Bai K, Chen X, Song R, Shi W, Shi S. Association of body mass index and waist circumference with type 2 diabetes mellitus in older adults: a cross-sectional study. BMC Geriatr. 2022;22:489. 10.1186/s12877-022-03145-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Bazyar H, Zare Javid A, Masoudi MR, Haidari F, Heidari Z, Hajializadeh S, et al. Assessing the predictive value of insulin resistance indices for metabolic syndrome risk in type 2 diabetes mellitus patients. Sci Rep. 2024;14:8917. 10.1038/s41598-024-59659-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Feng X, Wang J, Wang S, Wu S, Wang Z, Wei Y, et al. Correlation analysis of anthropometric indices and type 2 diabetes mellitus in residents aged 60 years and older. Front Public Health. 2023;11:1122509. 10.3389/fpubh.2023.1122509. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Montes YD, Vergara TA, Molina RT, Guerrero GM, Arrieta LAA, Aschner P, et al. The association between sociodemographic characteristics, clinical indicators and body mass index in a population at risk of type 2 diabetes: a cross-sectional study in two Colombian cities. Prim Care Diabetes. 2024;18:458–65. 10.1016/j.pcd.2024.06.001. [DOI] [PubMed] [Google Scholar]
- 16.Lee BJ, Yim MH. Comparison of anthropometric and body composition indices in the identification of metabolic risk factors. Sci Rep. 2021;11:9931. 10.1038/s41598-021-89422-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Mohammad A, Ziyab AH, Mohammad T. Anthropometric and DXA-derived measures of body composition in relation to pre-diabetes among adults. BMJ Open Diabetes Res Care. 2023;11:e003412. 10.1136/bmjdrc-2023-003412. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Sun J, Liu Z, Zhang Z, Zeng Z, Kang W. The correlation of prediabetes and type 2 diabetes with adiposity in adults. Front Nutr. 2022;9:818263. 10.3389/fnut.2022.818263. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Park MJ, Hwang SY, Kim NH, Kim SG, Choi KM, Baik SH, et al. A novel anthropometric parameter, weight-adjusted waist index represents sarcopenic obesity in newly diagnosed type 2 diabetes mellitus. J Obes Metab Syndr. 2023;32:130–40. 10.7570/jomes23005. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Baveicy K, Mostafaei S, Darbandi M, Hamzeh B, Najafi F, Pasdar Y. Predicting metabolic syndrome by visceral adiposity index, body roundness index and a body shape index in adults: a cross-sectional study from the Iranian RaNCD cohort data. Diabetes Metab Syndr Obes. 2020;13:879–87. 10.2147/DMETSO.S238153. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Jiang K, Luan H, Pu X, Wang M, Yin J, Gong R. Association between visceral adiposity index and insulin resistance: a cross-sectional study based on US adults. Front Endocrinol. 2022;13:921067. 10.3389/fendo.2022.921067. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Khorlampenko AA, Karetnikova VN, Kochergina AM, Ignatova YS, Belik EV, Gruzdeva OV, et al. Visceral adiposity index in patients with coronary artery disease, obesity and type 2 diabetes. Cardiovasc Ther Prev. 2020;19:2311. 10.15829/1728-8800-2020-2311. [Google Scholar]
- 23.Amato MC, Giordano C, Galia M, Criscimanna A, Vitabile S, Midiri M, et al. Visceral Adiposity Index: a reliable indicator of visceral fat function associated with cardiometabolic risk. Diabetes Care. 2010;33(4):920–2. 10.2337/dc09-1825. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Jiao M, Chen J, Wang X, Tao W, Feng Y, Yang H, et al. Anthropometric and metabolic parameters associated with visceral fat in non-obese type 2 diabetes individuals. Diabetol Metab Syndr. 2025;17:28. 10.1186/s13098-025-01583-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Liu AB, Lin YX, Meng TT, Tian P, Chen JL, Zhang XH, et al. Associations of the cardiometabolic index with insulin resistance, prediabetes, and diabetes in US adults: a cross-sectional study. BMC Endocr Disord. 2024;24:217. 10.1186/s12902-024-01676-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Lushchak VI, Covasa M, Abrat OB, Mykytyn TV, Tverdokhlib IZ, Storey KB, et al. Risks of obesity and diabetes development in the population of the Ivano-Frankivsk region in Ukraine. EXCLI J. 2023;22:1047. 10.17179/excli2023-6296. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Dyakova NA. Regression analysis in development of method for isolation and quantitative determination of water-soluble polysaccharides from sunflower roots of one-year-old. Drug Dev Regist. 2022;11:113–21. 10.33380/2305-2066-2022-11-1-113-121. [Google Scholar]
- 28.Stefanescu A, Revilla L, Lopez T, Sanchez SE, WilliaMetS MA, Gelaye B, Using. A Body Shape Index (ABSI) and Body Roundness Index (BRI) to predict risk of metabolic syndrome in Peruvian adults. J Int Med Res. 2020;48:0300060519848854. 10.1177/0300060519848854. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Statsenko ME, Fabritskaya SV, Ryndina YA, Shilina NN. The effect of visceral obesity on the condition of the main arteries in patients with chronic heart failure. J Volgograd State Med Univ. 2021;18:88–95. 10.19163/1994-9480-2021-4(80). [Google Scholar]
- 30.Ghassan F, Mohammed A, Kazim Z. The effect of the keto diet on lipid profile with type two diabetic patients. Drug Dev Regist. 2024;13(4):202–11. 10.33380/2305-2066-2024-13-4-1820. [Google Scholar]
- 31.Matveeva VA, Kozlova YS, Razumova SN, Plokhikh AI, Baikulova MD, Ruda O, et al. Model of tooth enamel for tribological study. J Int Den Med Res. 2023;16(2):619–23. [Google Scholar]
- 32.Bruess E. 3 reasons why BMI is not an accurate measure of your health or body weight — and what to use instead. Business Insider. 2021. https://www.businessinsider.com/guides/health/is-bmi-accurate. Accessed 8 Jan 2026.
- 33.Wharton S, Lau DC, Vallis M, Sharma AM, Biertho L, Campbell-Scherer D, et al. Obesity in adults: a clinical practice guideline. Can Med Assoc J. 2020;192:E875–91. 10.1503/cmaj.191707. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
All data generated or analysed during this study are included in this published article.