Non-alcoholic fatty liver disease (NAFLD) is one of the most common chronic liver disorders worldwide and is closely linked to metabolic dysfunction, obesity, insulin resistance, type 2 diabetes mellitus, hypertension, and dyslipidemia [1]. The coexistence of multiple metabolic abnormalities may identify patients with a particularly high cardiometabolic risk profile. Insulin resistance is a central pathophysiological feature of NAFLD and contributes to abnormal glucose and lipid metabolism [2,3]. Direct measurement of insulin sensitivity, however, is generally impractical in routine clinical practice. Several surrogate markers based on routinely available biochemical and anthropometric variables have therefore been investigated [4,5]. The triglyceride-glucose (TyG) index, calculated from fasting triglyceride and glucose concentrations, has emerged as a simple surrogate marker of insulin resistance [4,5] Its combination with body mass index (BMI) and waist circumference produces TyG-BMI and TyG-WC, respectively, potentially incorporating information about overall and central adiposity [6,7]. The lipid accumulation product (LAP) is a combination of waist circumference and triglyceride concentration that has been proposed as an indirect marker of lipid over-accumulation and metabolic risk [8]. The atherogenic index of plasma (AIP) calculated from the logarithmic ratio of triglycerides to high-density lipoprotein cholesterol has been studied as a marker of atherogenic dyslipidemia [9]. Although these indices have individually been associated with insulin resistance and cardiometabolic risk comparative evidence regarding their ability to identify NAFLD patients with multiple metabolic comorbidities is limited [2,3]. The primary objective of this study was to compare TyG, TyG-BMI, TyG-WC, LAP, and AIP for identifying adults with NAFLD who had a high metabolic comorbidity burden, defined as the presence of at least two of three conditions: diabetes mellitus, hypertension, and dyslipidemia.

Study Design and Setting This observational cross-sectional study was conducted in the Department of Medicine at Nalanda Medical College and Hospital (NMCH), Patna, Bihar, India. The study included adults with NAFLD diagnosed on the basis of documented hepatic steatosis on abdominal ultrasonography between March 2018 to December 2019. The study was conducted in accordance with the principles of the Declaration of Helsinki and was approved by the Institutional Ethics Committee. The informed consent was obtained. Study Population A total of 112 adult patients with NAFLD were included. Patients were eligible when sufficient demographic, anthropometric, clinical, and biochemical information was available to calculate the metabolic indices. Definition of Metabolic Comorbidity Burden The primary outcome was the number of coexisting metabolic conditions among: 1. Diabetes mellitus [10] 2. Hypertension 3. Dyslipidemia Patients were categorized into: • Low metabolic comorbidity burden: 0-1 metabolic condition • High metabolic comorbidity burden: ≥2 metabolic conditions Of the 112 patients, 82 (73.2%) had low burden and 30 (26.8%) had high burden. The final clinical classification supplied by the investigators comprised 18 patients with diabetes plus hypertension, three with diabetes plus dyslipidemia, three with hypertension plus dyslipidemia, and six with all three conditions. The underlying spreadsheet currently contains a discrepancy in the individual combination coding; therefore, the patient-level statistical analyses were performed using the reproducible binary classification of 30 patients with ≥2 conditions versus 82 with 0-1 condition. Clinical and Biochemical Measurements The following variables were obtained from the medical records: age, sex, BMI, waist circumference, systolic blood pressure, diastolic blood pressure, fasting blood glucose, glycated hemoglobin (HbA1c), triglycerides, and high-density lipoprotein (HDL) cholesterol. BMI was calculated as weight in kilograms divided by height in meters squared. Calculation of Metabolic Indices TyG was calculated as [4,5]: TyG = ln [fasting triglycerides (mg/dL) × fasting glucose (mg/dL) / 2]. TyG-BMI was calculated as: TyG-BMI = TyG × BMI (kg/m²). TyG-WC was calculated as: TyG-WC = TyG × waist circumference (cm). LAP was calculated using sex-specific formulas [8]. Men: LAP = (waist circumference − 65) × triglycerides (mmol/L) Women: LAP = (waist circumference − 58) × triglycerides (mmol/L) AIP was calculated as [9]: AIP = log10 [triglycerides (mmol/L) / HDL cholesterol (mmol/L)]. Statistical Analysis Continuous variables are presented as mean ± standard deviation (SD), and categorical variables as frequency and percentage. Participants with low and high metabolic comorbidity burden were compared using appropriate statistical tests according to the distribution and type of variable. Spearman rank correlation was used to assess associations between each metabolic index and the number of metabolic comorbidities. Binary logistic regression was performed with high metabolic comorbidity burden (≥2 vs 0-1) as the dependent variable. Odds ratios (ORs) were expressed per one-SD increase in each metabolic index. Multivariable models were adjusted for age and sex. ROC curves were constructed to evaluate the discriminatory performance of the indices for identifying high metabolic comorbidity burden. AUCs with 95% confidence intervals (CIs), optimal thresholds based on the Youden index, sensitivity, and specificity were calculated. A two-sided p<0.05 was considered statistically significant. Statistical analyses were performed using SPSS 31.0.

Study Population A total of 112 adults with NAFLD were included. Thirty participants (26.8%) had two or more metabolic comorbidities, whereas 82 (73.2%) had zero or one. The clinical distribution of patients with multiple metabolic conditions was 18 with diabetes plus hypertension, three with diabetes plus dyslipidemia, three with hypertension plus dyslipidemia, and six with diabetes, hypertension, and dyslipidemia. Baseline Characteristics Patients with high metabolic comorbidity burden were significantly older than those with low burden (62.50 ± 8.39 vs 45.87 ± 13.74 years, p<0.001). The proportion of males was similar between the groups (50.0% vs 54.9%, p=0.807). BMI was significantly higher in the high-burden group (29.29 ± 2.54 vs 26.67 ± 3.98 kg/m², p=0.002), as was waist circumference (97.63 ± 7.60 vs 91.22 ± 8.60 cm, p<0.001). Fasting glucose was significantly higher among patients with high burden (142.23 ± 26.74 vs 101.63 ± 20.33 mg/dL, p<0.001), as was HbA1c (7.95 ± 1.44% vs 5.94 ± 1.31%, p<0.001). Both systolic and diastolic blood pressure were significantly higher in the high-burden group. Mean systolic blood pressure was 142.73 ± 7.85 versus 128.88 ± 8.70 mmHg (p<0.001), while mean diastolic blood pressure was 89.67 ± 5.68 versus 82.27 ± 4.95 mmHg (p<0.001). Triglyceride concentration was modestly higher in the high-burden group (154.27 ± 23.84 vs 144.38 ± 24.97 mg/dL, p=0.049). HDL cholesterol was 44.93 ± 9.87 versus 40.02 ± 10.31 mg/dL (p=0.025). (Table 1). Table 1. Baseline Characteristics According to Metabolic Comorbidity Burden Variable 0-1 Comorbidities (n=82) ≥2 Comorbidities (n=30) P Value Age, years 45.87 ± 13.74 62.50 ± 8.39 <0.001 Male sex, n (%) 45 (54.9) 15 (50.0) 0.807 BMI, kg/m² 26.67 ± 3.98 29.29 ± 2.54 0.002 Fasting glucose, mg/dL 101.63 ± 20.33 142.23 ± 26.74 <0.001 HbA1c, % 5.94 ± 1.31 7.95 ± 1.44 <0.001 Triglycerides, mg/dL 144.38 ± 24.97 154.27 ± 23.84 0.049 HDL cholesterol, mg/dL 40.02 ± 10.31 44.93 ± 9.87 0.025 Waist circumference, cm 91.22 ± 8.60 97.63 ± 7.60 <0.001 Systolic BP, mmHg 128.88 ± 8.70 142.73 ± 7.85 <0.001 Diastolic BP, mmHg 82.27 ± 4.95 89.67 ± 5.68 <0.001 Values are mean ± SD unless otherwise indicated. BMI, body mass index; HbA1c, glycated hemoglobin; HDL, high-density lipoprotein; BP, blood pressure. Comparison of Metabolic Indices TyG was significantly higher in patients with high metabolic comorbidity burden (9.27 ± 0.25 vs 8.87 ± 0.23, p<0.001). Similarly, TyG-BMI was higher in the high-burden group (271.35 ± 21.77 vs 236.68 ± 36.62, p<0.001), as was TyG-WC (905.45 ± 76.99 vs 809.07 ± 79.61, p<0.001). LAP was also significantly higher among patients with ≥2 comorbidities (63.19 ± 16.29 vs 47.89 ± 14.88, p<0.001). AIP did not differ significantly between groups (0.181 ± 0.076 vs 0.206 ± 0.086, p=0.133). (Table 2). Table 2. Metabolic Indices According to Metabolic Comorbidity Burden Index 0-1 Comorbidities (n=82) ≥2 Comorbidities (n=30) P Value TyG 8.87 ± 0.23 9.27 ± 0.25 <0.001 TyG-BMI 236.68 ± 36.62 271.35 ± 21.77 <0.001 TyG-WC 809.07 ± 79.61 905.45 ± 76.99 <0.001 LAP 47.89 ± 14.88 63.19 ± 16.29 <0.001 AIP 0.206 ± 0.086 0.181 ± 0.076 0.133 TyG, triglyceride-glucose index; BMI, body mass index; WC, waist circumference; LAP, lipid accumulation product; AIP, atherogenic index of plasma. Correlation Analysis Spearman correlation analysis demonstrated a significant positive association between the number of metabolic comorbidities and TyG (ρ=0.631, p<0.001), TyG-BMI (ρ=0.560, p<0.001), TyG-WC (ρ=0.524, p<0.001), and LAP (ρ=0.390, p<0.001). AIP showed a weak inverse association with the number of metabolic comorbidities (ρ=-0.218, p=0.021). (Table 3). Table 3. Spearman Correlations Between Metabolic Indices and Metabolic Comorbidity Count Index Spearman ρ P Value TyG 0.631 <0.001 TyG-BMI 0.560 <0.001 TyG-WC 0.524 <0.001 LAP 0.390 <0.001 AIP -0.218 0.021 Logistic Regression Analysis In unadjusted logistic regression, each one-SD increase in TyG, TyG-BMI, TyG-WC, and LAP was associated with significantly greater odds of high metabolic comorbidity burden. TyG showed the strongest unadjusted association (OR, 6.97; 95% CI, 3.24-15.01; p<0.001). After adjustment for age and sex, the association remained significant (OR, 7.92; 95% CI, 3.07-20.47; p<0.001). TyG-WC was independently associated with high metabolic comorbidity burden after adjustment (OR, 5.91; 95% CI, 2.35-14.86; p<0.001). Significant adjusted associations were also observed for TyG-BMI (OR, 3.95; 95% CI, 1.61-9.68; p=0.003) and LAP (OR, 2.80; 95% CI, 1.47-5.36; p=0.002). AIP was not independently associated with high metabolic comorbidity burden after adjustment (OR, 0.78; 95% CI, 0.46-1.30; p=0.336). (Table 4). Table 4. Logistic Regression Analysis Index Unadjusted OR (95% CI) P Value Adjusted OR* (95% CI) P Value TyG 6.97 (3.24-15.01) <0.001 7.92 (3.07-20.47) <0.001 TyG-BMI 4.04 (2.04-8.02) <0.001 3.95 (1.61-9.68) 0.003 TyG-WC 3.89 (2.14-7.08) <0.001 5.91 (2.35-14.86) <0.001 LAP 2.74 (1.66-4.52) <0.001 2.80 (1.47-5.36) 0.002 AIP 0.74 (0.48-1.14) 0.170 0.78 (0.46-1.30) 0.336 *Adjusted for age and sex. ORs represent the odds associated with a one-SD increase in each index. ROC Analysis ROC analysis demonstrated that TyG had the highest discriminatory performance for identifying patients with two or more metabolic comorbidities, with an AUC of 0.887. TyG-WC demonstrated an AUC of 0.806, followed by TyG-BMI (AUC, 0.771) and LAP (AUC, 0.758). AIP showed poor discrimination, with an AUC of 0.407. Using the Youden index, the optimal TyG cutoff was approximately 8.96, with sensitivity of 96.7% and specificity of 72.0%.(Table 5) Fig 1. Table 5. ROC Performance of Metabolic Indices Index AUC 95% CI Optimal Cutoff Sensitivity Specificity TyG 0.887 0.820-0.953 8.96 96.7% 72.0% TyG-WC 0.806 0.715-0.897 870.60 76.7% 76.8% TyG-BMI 0.771 0.684-0.857 251.64 83.3% 65.9% LAP 0.758 0.661-0.854 45.44 90.0% 52.4% AIP 0.407 0.291-0.523 0.062 100.0% 3.7% ROC, receiver operating characteristic; AUC, area under the curve; CI, confidence interval.

In this retrospective cross-sectional study of 112 adults with NAFLD, 26.8% had 2 or more metabolic comorbidities. The main finding was that TyG was most strongly associated with increasing metabolic comorbidity burden and had the highest discriminatory performance in identifying patients with multiple metabolic abnormalities. The TyG index combines fasting glucose and triglyceride concentrations, both of which reflect important components of insulin resistance and metabolic dysfunction [4,5]. The strong correlation observed between TyG and the number of metabolic comorbidities and the high ROC AUC suggest that TyG may be useful for identifying NAFLD patients who warrant closer cardiometabolic evaluation. TyG-WC demonstrated the second-highest discriminatory performance [7]. Waist circumference provides information regarding central adiposity, which may complement the metabolic information captured by fasting glucose and triglycerides. The significant adjusted association between TyG-WC and high metabolic comorbidity burden supports this interpretation. TyG-BMI also demonstrated significant association with high metabolic comorbidity burden. BMI reflects overall adiposity but does not distinguish visceral from subcutaneous fat [6]. This may partly explain why TyG-WC performed better than TyG-BMI in the present cohort. LAP was significantly higher in patients with multiple metabolic comorbidities and remained independently associated with high burden after adjustment [8] LAP incorporates both waist circumference and triglyceride concentration and may therefore capture an adverse metabolic phenotype through complementary anthropometric and lipid information. AIP was unlike the other indices. There was no significant difference between groups and the ROC analysis showed poor discrimination [9]. AIP is a function of the triglyceride and HDL concentrations but does not take into account glucose or direct measurements of adiposity. This may have limited the ability to identify a composite burden including diabetes and hypertension. These results could have practical implications. TyG only needs fasting glucose and triglycerides which are routinely available in most clinical settings. TyG-BMI, TyG-WC and LAP are also based on simple anthropometric measurements. These indices could be used as low cost tools to identify NAFLD patients who require more in-depth assessment of their metabolic risk. LIMITATIONS Several limitations should be acknowledged. First, the retrospective cross-sectional design does not allow causal inferences. Second, the study was performed at a single tertiary care center with a relatively small sample size, especially in the high comorbidity group. Third, NAFLD was diagnosed by ultrasonography and not by liver biopsy or advanced quantitative imaging. Fourth, we did not have direct measures of insulin resistance. Fifth, possible confounding factors such as dietary intake, physical activity, medication use and socioeconomic factors were not adequately examined. Sixth, the evolving nomenclature for NAFLD, including metabolic dysfunction-associated fatty liver disease (MAFLD), was not applied in the present cohort [10]. In addition, the present analysis was based on a composite count of diabetes mellitus, hypertension, and dyslipidemia rather than a validated global cardiometabolic risk score. The ROC thresholds should therefore be considered exploratory rather than definitive clinical cutoffs. Future prospective multicenter studies with larger cohorts should validate these findings and determine whether these indices provide incremental prognostic information beyond conventional risk factors. Their ability to predict cardiovascular events, progression of hepatic fibrosis, and other clinically meaningful outcomes should also be evaluated.

Among adults with NAFLD, TyG demonstrated the strongest association with metabolic comorbidity burden and the highest discriminatory ability for identifying patients with two or more metabolic comorbidities. TyG-WC, TyG-BMI, and LAP also demonstrated significant associations and useful discriminatory performance, whereas AIP performed poorly. These findings suggest that simple insulin resistance and adiposity indices, particularly TyG, may have potential utility as inexpensive tools for cardiometabolic risk stratification in patients with NAFLD. Prospective multicenter validation is required before these indices can be incorporated into routine clinical decision-making.

1. Younossi ZM, Koenig AB, Abdelatif D, Fazel Y, Henry L, Wymer M: Global epidemiology of nonalcoholic fatty liver disease—meta-analytic assessment of prevalence, incidence, and outcomes. Hepatology. 2016, 64:73-84. 10.1002/hep.28431 2. Gastaldelli A, Cusi K: From NASH to diabetes and from diabetes to NASH: mechanisms and treatment options. JHEP Rep. 2019, 1:312-328. 10.1016/j.jhep.2019.07.002 3. Mantovani A, Byrne CD, Targher G: Diabetes, nonalcoholic fatty liver disease and cardiovascular risk. Semin Immunopathol. 2021, 43:31-47. 10.1007/s00281-020-00808-1 4. Simental-Mendía LE, Rodríguez-Morán M, Guerrero-Romero F: The product of fasting glucose and triglycerides as surrogate for identifying insulin resistance in apparently healthy subjects. Metab Syndr Relat Disord. 2008, 6:299-304. 10.1089/met.2008.0034 5. Navarro-González D, Sánchez-Íñigo L, Pastrana-Delgado J, Fernández-Montero A, Martínez JA: Triglyceride-glucose index in comparison with fasting plasma glucose improved diabetes prediction in patients with normal fasting glucose. Prev Med. 2016, 86:99-105. 10.1016/j.ypmed.2016.01.022 6. Er LK, Wu S, Chou HH, Hsu LA, Teng MS, Sun YC, Ko YL: Triglyceride glucose-body mass index is a simple and clinically useful surrogate marker for insulin resistance in nondiabetic individuals. PLoS One. 2016, 11:e0149731. 10.1371/journal.pone.0149731 7. Song S, Son DH, Baik SJ, Cho WJ, Lee YJ: Triglyceride glucose-waist circumference (TyG-WC) is a reliable marker to predict non-alcoholic fatty liver disease. Biomedicines. 2022, 10:2251. 10.3390/biomedicines10092251 8. Kahn HS: The “lipid accumulation product” performs better than the body mass index for recognizing cardiovascular risk: a population-based comparison. Am J Clin Nutr. 2005, 82:1213-1219. 10.1093/ajcn/82.6.1213 9. Dobiásová M, Frohlich J: The plasma parameter log(TG/HDL-C) as an atherogenic index: correlation with lipoprotein particle size and esterification rate in apoB-lipoprotein-depleted plasma. Clin Biochem. 2001, 34:583-588. 10.1016/S0009-9120(01)00263-6 10. Eslam M, Sanyal AJ, George J: MAFLD: a consensus-driven proposed nomenclature for metabolic associated fatty liver disease. Gastroenterology. 2020, 158:1999-2014.e1. 10.1053/j.gastro.2019.11.312.