Association of conventional and emerging adiposity indices with albuminuria in Indian type 2 diabetic adults

ABSTRACT

Background:

Central adiposity may contribute more strongly to renal injury than overall obesity in type 2 diabetes mellitus (T2DM), particularly in South Asian populations characterized by higher visceral fat at lower body mass index (BMI).

Objective:

To evaluate six adiposity indices—BMI, weight-adjusted waist index (WWI), body adiposity index (BAI), waist-to-height ratio (WHtR), a body shape index (ABSI), and body roundness index (BRI)—and identify the superior predictor of albuminuria in Indian adults with T2DM.

Methods:

In this hospital-based cross-sectional study, 364 adults with T2DM were enrolled consecutively. Albuminuria (normo-, micro-, or macroalbuminuria) was determined using the urinary albumin-to-creatinine ratio. Group differences were assessed using Welch’s ANOVA or Kruskal–Wallis tests. Associations were examined using Spearman’s correlation and binary and multinomial logistic regression adjusted for age, systolic blood pressure, HbA1c, smoking, and serum creatinine.

Results:

WWI, ABSI, BMI, and BAI differed across albuminuria categories (P < 0.05). WWI (r = 0.163, P = 0.002) and ABSI (r = 0.183, P < 0.001) correlated positively with UACR, whereas BAI correlated negatively (r = −0.137, P = 0.009). In adjusted models, WWI (OR = 1.32; 95% CI = 1.01–1.73) and ABSI (OR = 1.04 per 0.001-unit; 95% CI = 1.01–1.07) were independently associated with albuminuria, unlike BMI and BAI. The ABSI model showed the highest discrimination (AUC = 0.775). Stage-specific analysis demonstrated associations for WWI with macroalbuminuria (OR = 1.66; 95% CI: 1.09–2.53) and for ABSI with microalbuminuria (OR = 1.04; 95% CI: 1.00–1.07) and macroalbuminuria (OR = 1.07; 95% CI: 1.02–1.12).

Conclusion:

Central adiposity indices WWI and ABSI show modest, independent, and graded associations with albuminuria severity in Indian adults with T2DM.

Introduction

Type 2 diabetes mellitus (T2DM) remains a major global health challenge, with its prevalence increasing across both developed and developing countries.[1] India is among the countries with a rapidly increasing burden of diabetes.[2] Among these, diabetic kidney disease (DKD) is one of the most serious long-term complications of T2DM and a major contributor to chronic kidney disease and increased mortality.[3,4] Albuminuria is an early clinical marker of DKD and reflects glomerular injury, making it valuable for the early detection and monitoring of disease progression.[5] It is also associated with an increased risk of cardiovascular disease in individuals with T2DM.[6] Therefore, identifying factors associated with albuminuria may improve early risk assessment and preventive strategies.

Obesity has been associated with diabetic kidney disease and greater albuminuria severity in individuals with T2DM.[7] Beyond overall adiposity, body fat distribution appears to be a stronger determinant of cardiometabolic risk than body mass index (BMI) alone.[8] South Asians, including Indians, often exhibit greater body fat and central adiposity despite relatively low BMI, a pattern known as the “thin–fat” phenotype.[9,10] Consequently, BMI alone may underestimate obesity-related renal risk in this population.

Although BMI is widely used because of its simplicity, it does not adequately reflect body fat distribution. Alternative anthropometric indices have therefore been developed to better characterize central adiposity.[8] Waist-to-height ratio (WHtR) reflects abdominal fat relative to height,[11] body adiposity index (BAI) estimates body fat using hip circumference and height,[12] A Body Shape Index (ABSI) incorporates waist circumference adjusted for BMI and height,[13] body roundness index (BRI) estimates abdominal adiposity using a geometric model,[14] and weight-adjusted waist index (WWI) normalizes waist circumference to body weight and has emerged as a practical anthropometric index for cardiometabolic risk assessment.[15] All are derived from routine anthropometric measurements and are readily applicable in clinical practice.

Emerging evidence suggests that central adiposity indices may better identify cardiometabolic and renal risk than BMI alone. In support of this, ABSI has been associated with mortality independent of BMI,[16] while WWI has been linked to albuminuria and impaired kidney function in East Asian populations.[17] Despite these findings, evidence regarding the clinical utility of these newer indices in Indian adults with T2DM remains limited. In India, renal risk assessment has traditionally relied on conventional measures of kidney function, including estimated glomerular filtration rate (eGFR) and proteinuria.[18] However, comparative evaluation of both conventional and emerging adiposity indices within a single Indian T2DM cohort is still scarce, particularly for WWI, ABSI, and BRI.

In this context, the present study was designed to evaluate and compare six anthropometric indices—BMI, WWI, BAI, WHtR, ABSI, and BRI—in relation to albuminuria among Indian adults with T2DM. By examining both traditional and newer adiposity measures, this study aimed to determine which indices were most strongly associated with albuminuria as a marker of early renal involvement. These anthropometric indices may assist clinicians in renal risk stratification using measurements already obtained during routine diabetes care. For patients, improved assessment of obesity-related renal risk may facilitate more informed counseling and individualized management decisions, particularly in resource-constrained settings.

Methods

Study design and setting

This hospital-based analytical cross-sectional study was conducted from November 2023 to July 2025 at a tertiary care teaching hospital in Western India. Eligible adults with T2DM attending the General Medicine outpatient department were enrolled consecutively after providing written informed consent to minimize selection bias. Laboratory investigations were performed in the Department of Biochemistry following standardized procedures. The study evaluated the association between multiple adiposity indices and albuminuria at a single assessment, and the participant selection process is shown in Figure 1.

Figure 1.

Flow diagram showing participant screening, exclusions, and final inclusion for analysis

Study participants

A total of 364 adults (≥18 years) with confirmed T2DM were included. Diabetes was diagnosed according to the American Diabetes Association criteria.[19] Patients with type 1 or gestational diabetes, non-diabetic kidney disease, active urinary tract infection, hematuria, acute febrile illness, chronic liver disease, decompensated heart failure, chronic inflammatory or autoimmune disorders, active malignancy, pregnancy or lactation, or those receiving medications affecting renal hemodynamics or body composition (e.g. sodium–glucose cotransporter-2 inhibitors or systemic corticosteroids) were excluded. These criteria ensured clinical stability and minimized potential confounding of renal and adiposity measures.

Sample size estimation

The required sample size was calculated using the OpenEpi web-based calculator for estimation of a single population proportion.[20] Based on previously reported prevalence data, a 95% confidence level and 5.1% absolute precision were applied, yielding a minimum requirement of 364 participants. This sample size was considered sufficient to provide adequate statistical power (>80%) for correlation and multivariable regression analyses.

Clinical and anthropometric assessment

Data were collected using a structured, pre-tested questionnaire that recorded demographic characteristics, diabetes duration, treatment history, comorbidities, and smoking status. Anthropometric measurements were obtained by trained personnel using calibrated instruments. Height and weight were measured with participants barefoot and wearing light clothing using a wall-mounted stadiometer and calibrated digital weighing scale, respectively, and recorded to the nearest 0.1 cm and 0.1 kg. Waist circumference was measured midway between the lower rib and iliac crest, and hip circumference at the widest part of the buttocks using a non-stretchable measuring tape. All anthropometric measurements were performed in duplicate and averaged. Blood pressure was measured in the seated position after 5 minutes of rest using a validated automated oscillometric device with an appropriately sized cuff. Two readings were obtained 5 minutes apart and averaged for analysis.

Calculation of adiposity indices

Six adiposity indices were derived from standardized anthropometric measurements to characterize overall and central adiposity. BMI was calculated as weight (kg) divided by height squared (m²). WWI was computed as waist circumference (cm) divided by the square root of body weight (kg).[21] Similarly, BAI was calculated as hip circumference (cm) divided by height (m) raised to the power of 1.5, minus 18,[22] whereas WHtR was calculated by dividing waist circumference by height, with both measurements expressed in centimeters.[23] ABSI was calculated as waist circumference (m) divided by the product of BMI^(2/3) and height^(1/2).[24] For regression analyses, ABSI values were rescaled by multiplying by 1000 to improve interpretability. Finally, BRI was calculated using the geometric equation described in the original model, incorporating waist circumference and height to estimate body roundness and visceral adiposity.[14]

Biochemical and renal assessment

Following an overnight fast of 8–10 hours, venous blood samples were collected under aseptic conditions. Serum creatinine was measured using an enzymatic assay traceable to isotope-dilution mass spectrometry standards. Glycated hemoglobin (HbA1c) was analyzed by high-performance liquid chromatography aligned with National Glycohemoglobin Standardization Program criteria. Internal quality control procedures were conducted in accordance with laboratory standards throughout the study period.

Morning midstream urine samples were obtained to minimize diurnal variability. Urinary albumin was quantified using an immunoturbidimetric assay, and urinary creatinine was measured by the kinetic Jaffe method. The urine albumin-to-creatinine ratio (UACR) was calculated and categorized as normoalbuminuria (<30 mg/g), microalbuminuria (30–300 mg/g), or macroalbuminuria (>300 mg/g) according to established clinical criteria.[25] Estimated glomerular filtration rate (eGFR) was calculated using the CKD-EPI 2021 creatinine-based equation.[26]

Statistical analysis

Statistical analyses were performed using Jamovi (version 2.6).[27] Data normality was assessed using the Shapiro–Wilk test. Continuous variables are presented as mean ± standard deviation or median (interquartile range), as appropriate, and categorical variables as frequencies and percentages.

Comparisons across albuminuria categories were performed using Welch’s ANOVA or the Kruskal–Wallis test, followed by appropriate post hoc analyses, while categorical variables were compared using the Chi-square test. Spearman’s rank correlation assessed associations between adiposity indices and renal parameters.

To evaluate independent associations with albuminuria, binary logistic regression was performed after confirming model assumptions and the absence of multicollinearity. Adjusted odds ratios (95% confidence intervals) were reported. Multinomial logistic regression was subsequently used to assess stage-specific associations using normoalbuminuria as the reference category. Model performance was evaluated using the Akaike Information Criterion, Nagelkerke’s R², and the area under the receiver operating characteristic curve. A two-sided P < 0.05 was considered statistically significant.

Results

Participant characteristics

During the study period, 460 individuals were screened, of whom 364 adults with confirmed T2DM met eligibility criteria and were included in the final analysis [Figure 1]. Among these participants, 196 (53.8%) had normoalbuminuria, 120 (33.0%) had microalbuminuria, and 48 (13.2%) had macroalbuminuria.

Baseline characteristics stratified by albuminuria status are presented in Table 1. Age differed significantly across categories (P = 0.030, Welch ANOVA), with higher mean values observed in the macroalbuminuria group (58.63 ± 8.93 years) compared with normoalbuminuria (55.14 ± 9.76 years). However, sex distribution was comparable between groups (P = 0.445).

Table 1.

Baseline characteristics of the study cohort stratified by albuminuria status

Characteristic	Normoalbuminuria (n=196)	Microalbuminuria (n=120)	Macroalbuminuria (n=48)	P	Statistical Test	Post hoc Comparisons
Demographics
Age (years)	55.14±9.76	57.49±11.72	58.63±8.93	0.030	Welch ANOVA	Macro vs Micro: 0.778 Macro vs Normo: 0.051 Micro vs Normo: 0.158
Female sex, n (%)	93 (47.4%)	51 (42.5%)	22 (45.8%)	0.445	Chi-square	—
Anthropometrics
BMI (kg/m²)	26.6 [24.2–30.3]	25.7 [23.7–28.1]	25.6 [23.3–28.0]	0.013	Kruskal–Wallis	Macro vs Micro: 0.494 Macro vs Normo: 0.032 Micro vs Normo: 0.085
Waist circumference (cm)	89.0 [87.0–92.7]	90.1 [87.0–93.2]	90.4 [86.4–95.2]	0.229	Kruskal–Wallis	All P>0.34
Hip circumference (cm)	101.0 [95.5–103.0]	100.0 [90.5–103.0]	100.0 [86.5–104.0]	0.259	Kruskal–Wallis	All P>0.21
Adiposity Indices
Weight-Adjusted Waist Index (WWI)	10.71±0.87	10.89±1.05	11.14±0.84	0.006	Welch ANOVA	Macro vs Micro: 0.232 Macro vs Normo: 0.006 Micro vs Normo: 0.264
Body Adiposity Index (BAI) (%)	29.8 [26.7 – 33.8]	28.5 [24.6 – 32.3]	28.6 [25.0 – 32.6]	0.040	Kruskal–Wallis	Macro vs Micro: 0.891 Macro vs Normo: 0.363 Micro vs Normo: 0.039
WHtR	0.555±0.045	0.553±0.050	0.562±0.055	0.635	Welch ANOVA	All P>0.60
ABSI	0.079 [0.074–0.084]	0.081 [0.076–0.087]	0.082 [0.078–0.089]	0.007	Kruskal–Wallis	Macro vs Normo: 0.029 Micro vs Normo: 0.038 Macro vs Micro: 0.940
Body Roundness Index (BRI)	4.50±0.95	4.46±1.04	4.63±1.16	0.664	Welch ANOVA	All P>0.63
Metabolic Parameters
HbA1c (%)	9.93 [8.40–11.30]	9.91 [7.80–11.70]	9.86 [7.80–11.70]	0.974	Kruskal–Wallis	All P≈1.00
Creatinine (mg/dL)	0.74 [0.63–0.86]	0.98 [0.82–1.12]	1.42 [1.05–1.85]	<0.001	Kruskal–Wallis	All P<0.001
eGFR (mL/min/1.73 m²)	104.0 [95.0–119.0]	90.0 [70.0–116.0]	53.0 [26.0–82.0]	<0.001	Kruskal–Wallis	All P<0.001
UACR (mg/g)	14.6 [8.7–20.5]	80.4 [37.7–96.8]	1302.0 [447.0–1702.0]	<0.001	Kruskal–Wallis	All P<0.001
Clinical Variables
SBP (mmHg)	133.0 [120.0–144.0]	138.0 [122.0–152.0]	153.0 [140.0–166.0]	<0.001	Kruskal–Wallis	Macro vs Micro: 0.001 Macro vs Normo: <0.001 Micro vs Normo: 0.074
DBP (mmHg)	80.0 [74.0–86.5]	82.0 [74.0–88.0]	85.0 [78.0–92.5]	0.013	Kruskal–Wallis	Macro vs Normo: 0.011 Macro vs Micro: 0.246 Micro vs Normo: 0.292
Diabetes duration (years)	4.0 [3.0–8.0]	8.0 [5.9–11.3]	11.5 [8.0–15.0]	<0.001	Kruskal–Wallis	All P<0.001
Hypertension, n (%)	78 (39.8%)	74 (61.7%)	42 (87.5%)	<0.001	Chi-square	Macro vs Micro: 0.003 Macro vs Normo: <0.001 Micro vs Normo: <0.001

Notes: Data are presented as mean±standard deviation for normally distributed variables, median [interquartile range] for non-normally distributed variables, or frequency (percentage). Welch ANOVA with Games-Howell post hoc test was used for normally distributed variables with unequal variances. Kruskal-Wallistest with Dwass-Steel-Critchlow-Fligner post hoc comparisons was used for non-normally distributed variables. Chi-square test was used for categorical variables. Abbreviations: BMI, Body Mass Index; WHtR, Waist-to-Height Ratio; ABSI, A Body Shape Index; eGFR, estimated Glomerular Filtration Rate; UACR, Urine Albumin-to-Creatinine Ratio

Regarding anthropometric measures, median BMI varied significantly (P = 0.013, Kruskal–Wallis). Participants with macroalbuminuria had a lower BMI [25.6 kg/m² (23.3–28.0)] than those with normoalbuminuria [26.6 kg/m² (24.2–30.3); P = 0.032]. Waist and hip circumferences did not differ significantly.

Among the six adiposity indices, WWI (P = 0.006), BAI (P = 0.040), and ABSI (P = 0.007) showed significant differences across albuminuria categories. Mean WWI increased progressively from normoalbuminuria (10.71 ± 0.87) to macroalbuminuria (11.14 ± 0.84), with a significant difference between macro- and normoalbuminuria (P = 0.006). ABSI values were higher in both microalbuminuria [0.081 (0.076–0.087)] and macroalbuminuria [0.082 (0.078–0.089)] compared with normoalbuminuria [0.079 (0.074–0.084)]; P = 0.038 and P = 0.029, respectively. In contrast, WHtR and BRI did not demonstrate significant variation. These distributional differences across albuminuria categories are visually illustrated in Figure 2.

Figure 2.

Comparison of anthropometric and body composition indices across albuminuria status groups in T2DM

A clear gradient was observed in renal parameters. Median serum creatinine increased from 0.74 mg/dL (0.63–0.86) in normoalbuminuria to 1.42 mg/dL (1.05–1.85) in macroalbuminuria (P < 0.001), while median eGFR declined from 104.0 mL/min/1.73 m² (95.0–119.0) to 53.0 mL/min/1.73 m² (26.0–82.0, P < 0.001). Systolic and diastolic blood pressures increased progressively across stages (P < 0.001 and P = 0.013, respectively). The prevalence of hypertension rose from 39.8% (78/196) in normoalbuminuria to 87.5% (42/48) in macroalbuminuria (P < 0.001). Duration of diabetes was significantly longer in higher albuminuria categories (P < 0.001), whereas HbA1c did not differ significantly (P = 0.974).

Correlation analysis

Given the significant differences observed in WWI, BAI, and ABSI across albuminuria categories, these indices were further examined using Spearman’s correlation analysis [Table 2].

Table 2.

Spearman’s correlation of significant adiposity indices with study variables in patients with T2DM

Variable	Weight-Adjusted Waist Index (WWI)		Body Adiposity Index (BAI)		A Body Shape Index (ABSI)
	r	P	r	P	r	P
Body Mass Index (BMI)	−0.405***	<0.001	0.468***	<0.001	−0.603***	<0.001
HbA1c (%)	−0.043	0.414	−0.192***	<0.001	−0.002	0.964
Creatinine (mg/dL)	−0.013	0.802	−0.156**	0.003	0.064	0.225
eGFR (mL/min/1.73 m²)	−0.110*	0.035	0.022	0.670	−0.109*	0.038
UACR (mg/g)	0.163**	0.002	−0.137**	0.009	0.183***	<0.001
SBP (mmHg)	0.023	0.659	−0.011	0.837	−0.007	0.899
DBP (mmHg)	−0.073	0.166	−0.106*	0.043	−0.021	0.688

Note. This table presents Spearman’s rank correlation coefficients (r) for the three adiposity indices found to be significant in univariate analysis. The P-values correspond to the null hypothesis that the correlation coefficient is zero. Abbreviations: eGFR, estimated glomerular filtration rate; UACR, urine albumin-to-creatinine ratio; SBP, systolic blood pressure; DBP, diastolic blood pressure. Significance levels are *P<0.05, **P<0.01, and ***P<0.001

WWI and ABSI were negatively correlated with BMI (r = −0.405 and r = −0.603; both P < 0.001), whereas BAI correlated positively with BMI (r = 0.468; P < 0.001). With respect to renal measures, WWI (r = 0.163, P = 0.002) and ABSI (r = 0.183, P < 0.001) demonstrated positive correlations with UACR, while BAI showed an inverse correlation (r = −0.137, P = 0.009). Both WWI and ABSI were weakly but significantly inversely correlated with eGFR (r = −0.110 and r = −0.109; both P < 0.05), whereas BAI showed no significant association. Additionally, BAI was negatively correlated with HbA1c (r = −0.192, P < 0.001) and diastolic blood pressure (r = −0.106, P = 0.043). None of the indices were significantly associated with systolic blood pressure.

Multivariable binary logistic regression

To assess whether the observed associations were independent of established clinical factors, multivariable logistic regression models were constructed with albuminuria (micro- or macroalbuminuria) as the outcome [Table 3]. Each model included one adiposity index and was adjusted for systolic blood pressure, serum creatinine, HbA1c, age, sex, and smoking status. BMI (OR = 0.95, 95% CI: 0.90–1.00) and BAI (OR = 0.95, 95% CI: 0.91–1.00) showed inverse associations with albuminuria, whereas WWI (OR = 1.32, 95% CI: 1.01–1.73) and ABSI (per 0.001-unit increase; OR = 1.04, 95% CI: 1.01–1.07) were positively associated (all P < 0.05). Serum creatinine and systolic blood pressure remained significant predictors across all models (P < 0.001). Model discrimination was comparable (AUC 0.769–0.775), while the ABSI model demonstrated the best fit, with the lowest AIC (427.0). The adjusted associations are shown in Figure 3.

Table 3.

Multivariable binary logistic regression models for the association of adiposity indices with albuminuria in T2DM

Predictor	Model 1: BMI OR (95% CI)	Model 2: WWI OR (95% CI)	Model 3: BAI OR (95% CI)	Model 4: ABSI OR (95% CI)
Anthropometric Index
Body Mass Index (BMI)	0.95 (0.90 - 1.00)*	—	—	—
Weight-Adjusted Waist Index (WWI)	—	1.32 (1.01 - 1.73)*	—	—
Body Adiposity Index (BAI)	—	—	0.95 (0.91 - 1.00)*	—
A Body Shape Index (ABSI) (per 0.001 unit)	—	—	—	1.04 (1.01 - 1.07)*
Clinical Covariates
Systolic BP (per mmHg)	1.02 (1.00 - 1.03)*	1.02 (1.00 - 1.03)*	1.02 (1.00 - 1.03)*	1.02 (1.00 - 1.03)*
Creatinine (per mg/dL)	12.18 (4.95 - 29.96)**	13.21 (5.28 - 33.06)**	11.95 (4.92 - 29.03)**	12.77 (5.11 - 31.90)**
HbA1c (per %)	1.07 (0.96 - 1.21)	1.10 (0.98 - 1.23)	1.07 (0.95 - 1.20)	1.09 (0.97 - 1.22)
Age (per year)	1.00 (0.97 - 1.02)	1.00 (0.97 - 1.02)	1.00 (0.98 - 1.03)	1.00 (0.97 - 1.02)
Sex (Male vs. Female)	0.77 (0.47 - 1.27)	0.89 (0.53 - 1.49)	0.65 (0.38 - 1.11)	0.83 (0.51 - 1.36)
Smoking (Yes vs. No)	0.58 (0.31 - 1.08)	0.56 (0.30 - 1.05)	0.63 (0.33 - 1.20)	0.57 (0.30 - 1.07)
Model Fit
Akaike Information Criterion (AIC)	429	429	429	427
Nagelkerke’s R²	0.291	0.292	0.291	0.298
Area Under the Curve (AUC)	0.771	0.774	0.769	0.775
Sensitivity (%)	72.6	70.2	68.5	70.8
Specificity (%)	71.4	70.4	66.8	69.9

Notes: Results are presented as odds ratios (OR) with 95% confidence intervals (CI) from binary logistic regression models, each adjusted for all listed covariates. The A Body Shape Index (ABSI) was rescaled (multiplied by 1000) for clinical interpretability; its aOR represents the change in odds per 0.001-unit increase in raw ABSI. Model fit was assessed using the Akaike Information Criterion (AIC) and Nagelkerke’s R², while discriminative performance was evaluated by the Area Under the Curve (AUC), with sensitivity and specificity calculated at a predicted probability cut-off of 0.5. Significance levels are *P<0.05 and **P<0.001

Figure 3.

Adjusted odds ratios for albuminuria according to adiposity indices in multivariable logistic regression models. Note: Models were adjusted for age, sex, systolic blood pressure, serum creatinine, HbA1c, and smoking status

Multinomial logistic regression

To evaluate stage-specific associations, multinomial logistic regression analyses were performed using normoalbuminuria as the reference category [Table 4]. BMI was not associated with microalbuminuria and showed only a marginal inverse association with macroalbuminuria (OR = 0.91, 95% CI: 0.82–1.00; P = 0.054). WWI was independently associated with macroalbuminuria (OR = 1.66, 95% CI: 1.09–2.53; P = 0.019), with a near-significant association for microalbuminuria (P = 0.072). HbA1c was independently associated with macroalbuminuria (OR = 1.21, 95% CI: 1.01–1.45; P = 0.035), whereas BAI showed no significant associations. In the ABSI model, each 0.001-unit increase was independently associated with both microalbuminuria (OR = 1.04, 95% CI: 1.00–1.07; P = 0.032) and macroalbuminuria (OR = 1.07, 95% CI: 1.02–1.12; P = 0.011). Model fit was slightly better for the WWI and ABSI models (Nagelkerke R² =0.23–0.24). The significant stage-specific associations are illustrated in Figure 4.

Table 4.

Multinomial logistic regression analysis for the association of adiposity indices with severity of albuminuria in T2DM

Predictor	Macroalbuminuria (n=48)			Microalbuminuria (n=120)
	β (SE)	OR [95% CI]	P	β (SE)	OR [95% CI]	P
4a. BMI model
Intercept	−8.08 (2.61)	-	0.002	−2.50 (1.61)	-	0.122
Age (years)	−0.00 (0.02)	1.00 [0.96–1.04]	0.997	−0.00 (0.01)	1.00 [0.97–1.02]	0.911
BMI (kg/m²)	−0.10 (0.05)	0.91 [0.82–1.00]	0.054	−0.05 (0.03)	0.95 [0.90–1.01]	0.118
SBP (mmHg)	0.04 (0.01)	1.04 [1.02–1.06]	<0.001	0.01 (0.01)	1.01 [1.00–1.02]	0.075
Creatinine (mg/dL)	3.07 (0.48)	21.60 [8.40–55.80]	<0.001	2.15 (0.44)	8.60 [3.60–20.40]	<0.001
Smoking (Yes vs No)	−0.62 (0.54)	0.54 [0.19–1.55]	0.250	−0.59 (0.33)	0.56 [0.29–1.05]	0.071
HbA1c (%)	0.15 (0.09)	1.16 [0.97–1.40]	0.098	0.05 (0.06)	1.05 [0.93–1.19]	0.394
Model fit	χ²=126, df=12, P<0.001, Pseudo-R²: McFadden=0.18, Nagelkerke=0.23, Information criteria: AIC=605, BIC=660
4b. WWI model
Intercept	−16.43 (3.32)	-	<0.001	−6.57 (1.98)	-	<0.001
Age (years)	−0.00 (0.02)	1.00 [0.96–1.04]	0.966	−0.00 (0.01)	1.00 [0.97–1.02]	0.826
WWI	0.51 (0.22)	1.66 [1.09–2.53]	0.019	0.25 (0.14)	1.28 [0.98–1.67]	0.072
SBP (mmHg)	0.04 (0.01)	1.04 [1.02–1.06]	<0.001	0.01 (0.01)	1.01 [1.00–1.02]	0.086
Creatinine (mg/dL)	3.23 (0.50)	25.40 [9.50–67.60]	<0.001	2.31 (0.46)	10.00 [4.10–24.50]	<0.001
Smoking (Yes vs No)	−0.61 (0.54)	0.54 [0.19–1.57]	0.259	−0.59 (0.33)	0.55 [0.29–1.04]	0.068
HbA1c (%)	0.19 (0.09)	1.21 [1.01–1.45]	0.035	0.07 (0.06)	1.08 [0.95–1.21]	0.233
Model fit	χ²=128, df=12, P<0.001, Pseudo-R²: McFadden=0.18, Nagelkerke=0.23, Information criteria: AIC=603, BIC=658
4c. BAI model
Intercept	−9.36 (2.49)	-	<0.001	−2.77 (1.55)	-	0.074
Age (years)	0.01 (0.02)	1.00 [0.96–1.04]	0.817	0.00 (0.01)	1.00 [0.97–1.02]	0.829
BAI (%)	−0.05 (0.04)	0.95 [0.89–1.03]	0.208	−0.04 (0.02)	0.97 [0.92–1.01]	0.149
SBP (mmHg)	0.04 (0.01)	1.04 [1.02–1.06]	<0.001	0.01 (0.01)	1.01 [1.00–1.04]	0.106
Creatinine (mg/dL)	3.03 (0.48)	20.70 [8.00–53.40]	<0.001	2.08 (0.44)	8.00 [3.40–19.00]	<0.001
Smoking (Yes vs No)	−0.56 (0.54)	0.57 [0.20–1.65]	0.299	−0.56 (0.33)	0.57 [0.30–1.09]	0.090
HbA1c (%)	0.15 (0.09)	1.17 [0.97–1.40]	0.101	0.05 (0.06)	1.05 [0.93–1.19]	0.397
Model fit	χ²=124, df=12, P<0.001, Pseudo-R²: McFadden=0.18, Nagelkerke=0.23, Information criteria: AIC=607, BIC=662
4d. ABSI model
Intercept	−16.01 (3.10)	-	<0.001	−6.56 (1.82)	-	<0.001
Age (years)	−0.00 (0.02)	1.00 [0.96–1.04]	0.969	−0.00 (0.01)	1.00 [0.97–1.02]	0.796
ABSI (per 0.001 unit)	0.06 (0.03)	1.07 [1.02–1.12]	0.011	0.04 (0.02)	1.04 [1.00–1.07]	0.032
SBP (mmHg)	0.04 (0.01)	1.04 [1.02–1.06]	<0.001	0.01 (0.01)	1.01 [0.99–1.03]	0.072
Creatinine (mg/dL)	3.17 (0.49)	23.70 [9.00–62.40]	<0.001	2.24 (0.45)	9.30 [3.90–22.60]	<0.001
Smoking (Yes vs No)	−0.63 (0.54)	0.53 [0.18–1.54]	0.244	−0.60 (0.33)	0.55 [0.29–1.04]	0.068
HbA1c (%)	0.18 (0.09)	1.20 [1.00–1.43]	0.050	0.06 (0.06)	1.06 [0.94–1.20]	0.310
Model fit	χ²=130, df=12, P<0.001, Pseudo-R²: McFadden=0.18, Nagelkerke=0.24, Information criteria: AIC=602, BIC=656

Note: This table presents detailed results from four separate multinomial logistic regression models. The reference category for the outcome is Normoalbuminuria. All models are adjusted for all variables listed. The A Body Shape Index (ABSI) was rescaled (multiplied by 1000) for clinical interpretability; its odds ratio represents the change in odds per 0.001-unit increase in raw ABSI. Abbreviations: β, regression coefficient; SE, standard error; OR, odds ratio; CI, confidence interval

Figure 4.

Adjusted odds ratios for stage-specific associations between adiposity indices and albuminuria. Note: Models were adjusted for age, sex, systolic blood pressure, serum creatinine, HbA1c, and smoking status

Discussion

This study demonstrates that central adiposity indices—WWI and ABSI—are independently associated with albuminuria among Indian adults with T2DM, whereas BMI and BAI showed neutral or inverse associations. Both indices exhibited graded associations across albuminuria stages, suggesting that body fat distribution is more closely linked to early renal injury than total body mass.

These findings may partly reflect the limitations of BMI as a surrogate of adiposity, particularly its inability to distinguish lean mass from metabolically active visceral fat. The inverse association between BMI and albuminuria is consistent with observations related to the “obesity paradox” described in chronic kidney disease, where higher BMI has been associated with apparently favorable outcomes despite its limited ability to reflect body composition.[28] In South Asians, this may be further influenced by the well-described “thin–fat” phenotype, characterized by greater central adiposity and metabolic risk despite relatively low BMI.[9,29] Collectively, these observations suggest that BMI may underestimate metabolically adverse fat accumulation, whereas WWI and ABSI have demonstrated stronger associations with renal outcomes than BMI in previous studies.[30,31]

The observed correlation patterns further support this interpretation. WWI and ABSI correlated positively with UACR and inversely with eGFR, whereas BAI showed opposite or neutral associations. These findings are biologically plausible, as visceral adipose tissue contributes to glomerular injury through inflammatory and metabolic pathways that promote proteinuria and progressive renal dysfunction.[32] Excess visceral fat induces oxidative stress, endothelial dysfunction, and activation of the renin–angiotensin–aldosterone system, thereby increasing intraglomerular pressure and permeability.[33] Moreover, lipid overflow and lipotoxicity contribute to podocyte injury and renal fibrosis through intracellular lipid accumulation and organelle dysfunction.[34] Together, these mechanisms support the observed association between central adiposity and renal injury.

Consistent with these mechanisms, WWI and ABSI remained independently associated with albuminuria after adjustment for blood pressure, glycemia, and renal function. Similar findings have been reported in East Asian populations, where ABSI has been independently associated with albuminuria and chronic kidney disease.[35,36] Furthermore, the stage-dependent associations observed in multinomial models support a dose–response relationship between central adiposity and albuminuria severity.

In contrast, the lack of an independent association for BAI suggests that peripheral adiposity may confer less renal risk than central adiposity. Gluteofemoral fat has been associated with a more favorable cardiometabolic profile than visceral fat,[37] highlighting the importance of fat distribution over absolute adiposity in renal vulnerability.

Our results are also consistent with the growing body of evidence supporting the clinical relevance of central adiposity indices. East Asian studies have linked WWI with albuminuria and chronic kidney disease,[17,30] while ABSI has been associated with impaired kidney function and elevated urinary albumin excretion.[35,36] However, most previous studies evaluated individual indices or non-Indian populations. These observations may be particularly important in South Asian populations, who characteristically exhibit greater central adiposity despite lower BMI.[29] Accordingly, these findings suggest that central body shape indices may complement routine diabetes assessment for renal risk stratification. By comparing six conventional and emerging adiposity indices in a single cohort of Indian adults with T2DM, our study provides population-specific evidence supporting the potential utility of WWI and ABSI for renal risk assessment.

From a clinical perspective, WWI and ABSI are derived from routine anthropometric measurements and may complement renal risk stratification, particularly where advanced imaging is unavailable. Lifestyle modification and GLP-1 receptor agonists improve body weight and cardiometabolic risk factors, which may indirectly benefit renal health.[38]

The present study has several strengths. These include evaluation of six adiposity indices, use of albuminuria as an early renal marker, and consistent findings across regression models. Standardized measurements and multivariable adjustment strengthen internal validity. Nevertheless, several limitations should be acknowledged. The cross-sectional design precludes causal inference, imaging-based assessment of visceral fat was not performed, and the single-center setting may limit generalizability. Residual confounding from diet and physical activity also cannot be excluded. Future prospective studies should determine whether longitudinal changes in WWI or ABSI predict incident albuminuria or eGFR decline and whether combining anthropometric indices with inflammatory or adipokine biomarkers improves risk prediction.

Conclusion

In Indian adults with T2DM, central adiposity indices—WWI and ABSI—were independently and progressively associated with albuminuria, whereas BMI and BAI were not. These findings suggest that central fat distribution, rather than overall body mass, appears to better reflect renal risk in this population. As WWI and ABSI are derived from routine anthropometry, they may serve as practical tools for improved risk stratification.

Ethical considerations

The study was approved by the Institutional Ethics Committee of the study center. All procedures complied with the Declaration of Helsinki and Good Clinical Practice guidelines. Written informed consent was obtained from all participants before enrollment. Participant confidentiality was maintained through anonymized coding and secure data storage with restricted access.

Data availability statement

The data supporting this study are not publicly available to protect participant privacy. De-identified datasets are available from the corresponding author upon reasonable request and institutional approval.

Author contributions

• Study Design: Mehul Kaliya, Sagar Dholariya, Ragini Singh

• Data Collection and Drafting: Mehul Kaliya, Sagar Dholariya

• Statistical Analysis and Interpretation: Sagar Dholariya, Krupal Joshi

• Critical Review, Revision, and Supervision: Ragini Singh, Garima Anandani, Krupal Joshi, Parth Goswami

• Final Approval: All authors

All authors have reviewed the manuscript and approved its final version for submission.

Conflicts of interest

There are no conflicts of interest.

Funding Statement

This study is a secondary analysis of an intramural research project (Project No. IM/F/05/2023-24) supported by the Institute. No external funding was received.