Preamble

Correlation between the Albumin/Gamma-glutamyl Transferase Ratio and Metabolic-associated Fatty Liver Disease in Type 2 Diabetes Mellitus

YAN Ziwei¹,², NIE Jiahua¹, SHI Yan³, WEI Limin²*

¹Graduate School, Hebei North University, Zhangjiakou 075000, China
²Department of Endocrinology, Hebei General Hospital, Shijiazhuang 050000, China
³Graduate School, Hebei Medical University, Shijiazhuang 050000, China

*Corresponding author: WEI Limin, Chief physician; E-mail: 15133130672@163.com

Abstract

Background In recent years, the incidence and prevalence of metabolic-associated fatty liver disease (MAFLD) have increased annually, with over one-quarter of the global population affected. Type 2 diabetes mellitus (T2DM) is closely related to MAFLD, yet simple and accurate predictors for MAFLD in T2DM patients remain scarce. Objective To investigate the correlation between the albumin/γ-glutamyl transferase ratio (AGTR) and T2DM with MAFLD, and to construct a nomogram model for predicting the risk of T2DM combined with MAFLD. Methods A total of 1,050 adult T2DM patients hospitalized in the Department of Endocrinology at Hebei General Hospital from 2018 to 2023 were enrolled. After multiple rounds of rigorous screening, 723 eligible patients were included, comprising 430 in the T2DM with MAFLD group and 293 in the T2DM without MAFLD group. Patient baseline data were collected and analyzed, with MAFLD diagnosed via ultrasound. Spearman correlation analysis was used to examine associations between AGTR and MAFLD risk factors. Multivariate logistic regression analysis was performed to identify risk factors for T2DM with MAFLD, and an individualized nomogram for predicting this risk was constructed and validated. Results Compared with the T2DM without MAFLD group, the T2DM with MAFLD group showed significantly higher levels of BMI, ALT, AST, GGT, BA, FBG, TG, TC, LDL-C, VLDL-C, ApoB, and UA, while age, diabetes duration, ALB, and HDL-C were significantly lower (P<0.05). When patients were divided into T1–T3 groups based on AGTR tertiles, the T3 group exhibited lower BMI, ALT, GGT, FBG, TC, TG, LDL-C, VLDL-C, ApoB, and UA compared to T2 and T1 groups, lower HDL-C and AST compared to T1 group, and higher age and diabetes duration compared to T2 and T1 groups (P<0.05). The T2 group showed lower BMI, ALT, AST, GGT, TG, VLDL-C, and UA compared to T1 group (P<0.05). Spearman correlation analysis revealed that AGTR was positively correlated with age, diabetes duration, ALB, and HDL-C (P<0.05), and negatively correlated with VLDL-C, BMI, ALT, AST, GGT, BA, FBG, TC, TG, LDL-C, ApoB, and UA (P<0.05). Multivariate logistic regression showed that increased BMI [OR (95%CI) = 1.256 (1.187–1.330)], increased TG [OR (95%CI) = 1.272 (1.043–1.551)], and decreased AGTR [OR (95%CI) = 0.707 (0.562–0.890)] were influencing factors for T2DM with MAFLD. ROC curve analysis demonstrated that the combined model of these three factors predicted MAFLD in T2DM with an AUC of 0.827 (95%CI = 0.790–0.864). The calibration curve showed good agreement between predicted and observed values, and clinical decision curve analysis indicated good clinical predictive utility for T2DM with MAFLD. Conclusion Reduced AGTR is a protective factor for T2DM with MAFLD. The individualized nomogram model based on BMI, TG, and AGTR can effectively predict the risk of T2DM with MAFLD.

Keywords: Metabolic-associated fatty liver disease; Diabetes mellitus, type 2; Albumin; Gamma-glutamyl transferase; Albumin/gamma-glutamyl transferase; Root cause analysis

With rising living standards, the incidence and prevalence of non-alcoholic fatty liver disease (NAFLD) have gradually increased. Statistics indicate that in 2016, NAFLD patients accounted for 25% of the world population, with a global prevalence of 38% in 2019. It is projected that 20 million patients will eventually die from NAFLD-related liver disease, imposing a substantial economic burden on society. In 2020, NAFLD and non-alcoholic steatohepatitis were renamed and redefined as metabolic-associated fatty liver disease (MAFLD) and metabolic dysfunction-associated steatohepatitis (MASH). While liver biopsy remains the gold standard for diagnosing fatty liver disease, its invasive nature limits widespread clinical use. Consequently, investigating practical, direct, and reliable predictors of fatty liver disease holds significant clinical value.

Recent studies have demonstrated that type 2 diabetes mellitus (T2DM) is a crucial risk factor for MAFLD development. T2DM patients drive MAFLD progression from simple hepatic steatosis to steatohepatitis and fibrosis, with T2DM patients exhibiting significantly higher MAFLD risk compared to non-diabetic individuals. Therefore, studying predictive factors for T2DM with MAFLD carries substantial clinical importance. The albumin/γ-glutamyl transferase ratio (AGTR) is a recently proposed inflammatory marker that has been confirmed as a protective factor for MAFLD. Research has shown that T2DM induces free radical production, leading to decreased serum albumin, while GGT is closely associated with T2DM and serves as an important predictor, with T2DM risk increasing alongside GGT elevation. Thus, AGTR is also closely related to T2DM, though its predictive value for T2DM with MAFLD remains unclear. This study retrospectively analyzed the relationship between T2DM with MAFLD and AGTR and established an individualized prediction model for T2DM with MAFLD risk.

1.1 Clinical Data

We retrospectively selected 1,050 adult T2DM patients hospitalized in the Department of Endocrinology at Hebei General Hospital from 2018 to 2023. After multiple rounds of rigorous screening, 723 patients met the inclusion criteria and were enrolled, including 430 with T2DM and MAFLD and 293 with T2DM alone [FIGURE:1]. T2DM diagnosis followed the 1999 WHO criteria, while MAFLD was defined as hepatic steatosis accompanied by overweight/obesity, T2DM, or two or more metabolic risk abnormalities. Exclusion criteria comprised: (1) type 1 diabetes, gestational diabetes, or pregnant women; (2) T2DM patients under 18 years; (3) acute diabetic complications such as ketoacidosis, hyperosmolar hyperglycemic syndrome, or hypoglycemic coma; (4) history of hepatitis B/C, cirrhosis, or liver surgery; and (5) recent severe systemic diseases (infectious diseases, kidney disease, severe cardiovascular/cerebrovascular disease, cancer, blood disorders). This study was approved by the Hebei General Hospital Ethics Committee (Approval No. 2025-LM-0072).

1.2 Research Methods

We collected patient baseline information (age, diabetes duration) and measured height and weight to calculate BMI. After an 8–10 hour fast, elbow venous blood samples were collected by professional medical staff the following morning. Fasting blood glucose (FBG), uric acid (UA), serum creatinine (Cr), total cholesterol (TC), triglycerides (TG), high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), very low-density lipoprotein cholesterol (VLDL-C), apolipoprotein A1 (ApoA1), apolipoprotein B (ApoB), aspartate aminotransferase (AST), alanine aminotransferase (ALT), γ-glutamyl transferase (GGT), bile acids (BA), and albumin (ALB) were measured using an automatic biochemical analyzer, and AGTR was calculated. Abdominal ultrasound was used to diagnose fatty liver disease.

1.3 Grouping

Based on ultrasound findings, patients were divided into a T2DM without MAFLD group (n=293) and a T2DM with MAFLD group (n=430).

1.4 Statistical Methods

Data analysis was performed using SPSS 25.0 and R software 4.2.2 (rms and caret packages). Continuous variables following normal distribution were expressed as (x̄±s) and compared between groups using independent t-tests; non-normally distributed data were expressed as M(P25, P75) and compared using Mann-Whitney U test for two groups or Kruskal-Wallis H test for multiple groups. Spearman rank correlation analysis explored relationships between AGTR and MAFLD-related indicators. Multivariate logistic regression analysis identified influencing factors for T2DM with MAFLD. Based on regression results, an individualized nomogram model was developed to predict T2DM with MAFLD risk, validated via Bootstrap resampling. Calibration curves assessed clinical consistency, receiver operating characteristic (ROC) curves evaluated predictive value for T2DM with MAFLD, and decision curves assessed clinical applicability. P<0.05 indicated statistical significance.

2.1 Baseline Data Comparison Between T2DM and T2DM with MAFLD Groups

Compared with the T2DM without MAFLD group, the T2DM with MAFLD group showed significantly increased BMI, ALT, AST, GGT, BA, FBG, TC, TG, LDL-C, VLDL-C, ApoB, and UA levels, while age, diabetes duration, ALB, and HDL-C were significantly decreased (P<0.05) [TABLE:1].

2.2 Comparison of General Data Across Different AGTR Levels

Patients were further divided into three groups based on AGTR tertiles: T1 group (AGTR: 0.036–1.116), T2 group (AGTR: 1.117–1.918), and T3 group (AGTR: 1.919–5.458). Significant differences were observed among the three groups in age, diabetes duration, BMI, ALB, ALT, AST, GGT, FBG, TC, TG, HDL-C, LDL-C, VLDL-C, ApoB, and UA (P<0.05). The T3 group exhibited lower BMI, ALT, GGT, FBG, TC, TG, LDL-C, VLDL-C, ApoB, and UA compared to T2 and T1 groups; lower HDL-C and AST compared to T1 group; and higher age and diabetes duration compared to T2 and T1 groups (P<0.05). The T2 group showed lower BMI, ALT, AST, GGT, TG, VLDL-C, and UA compared to T1 group (P<0.05). No significant differences were found in BA, ApoA1, or Cr among the three groups (P>0.05) [TABLE:2].

2.3 Correlation Analysis Between AGTR and General Data

Spearman rank correlation analysis demonstrated that AGTR levels were positively correlated with age, diabetes duration, ALB, and HDL-C (rs=0.267, 0.146, 0.258, 0.152; P<0.05), and negatively correlated with BMI, ALT, AST, GGT, BA, FBG, TG, TC, VLDL-C, LDL-C, ApoB, and UA (rs=-0.310, -0.266, -0.319, -0.390, -0.094, -0.216, -0.308, -0.169, -0.184, -0.143, -0.148, -0.222; P<0.05) [TABLE:3].

2.4 Multivariate Logistic Regression Analysis of Factors Influencing T2DM with MAFLD

Using MAFLD presence as the dependent variable (yes=1, no=0) and indicators with P<0.05 from Table 1 (age, BMI, duration, AST, ALT, BA, UA, FBG, TC, TG, HDL-C, LDL-C, VLDL-C, ApoB, AGTR) as independent variables (actual values), multivariate logistic regression analysis revealed that BMI, TG, and AGTR were influencing factors for T2DM with MAFLD (P<0.05) [TABLE:4].

2.5 Construction of a Nomogram Prediction Model for T2DM with MAFLD Risk

Based on multivariate logistic regression results, an individualized nomogram model was constructed to predict T2DM with MAFLD risk, incorporating three predictors: BMI, TG, and AGTR [FIGURE:2].

2.6 Validation of the Nomogram Risk Prediction Model

Using Bootstrap resampling for internal validation, the calibration curve approximated the Y=X line. Hosmer goodness-of-fit test for the calibration curve yielded P=0.606 (P>0.05), indicating good model fit [FIGURE:3]. The ROC curve showed an AUC of 0.827 (95%CI=0.790–0.864, P<0.001), with sensitivity of 68.8% and specificity of 83.8%, suggesting the nomogram based on BMI, TG, and AGTR has predictive value [FIGURE:4]. The decision curve indicated that when threshold probability ranged from 20% to 83%, the model offered ideal clinical utility for predicting T2DM with MAFLD risk [FIGURE:5].

As MAFLD prevalence and incidence continue to rise, it has become the most common cause of chronic liver disease, imposing a substantial societal burden. MAFLD fibrosis progression is closely related to T2DM, with over half of T2DM patients also having MAFLD. The detection rate of T2DM with MAFLD in this study was 59%, making early identification and treatment a shared clinical concern. While imaging modalities like ultrasound, CT, and MRI can effectively diagnose fatty liver by detecting hepatic fat accumulation, no specific laboratory markers exist. MAFLD lacks early clinical symptoms, and diagnostic methods are relatively expensive with limited equipment and technical expertise in remote areas, often resulting in detection only after severe complications develop. Therefore, identifying convenient, inexpensive screening markers to assess T2DM patients' MAFLD risk for early diagnosis and intervention holds important clinical significance.

Our data demonstrated that AGTR was significantly lower in the T2DM with MAFLD group compared to the T2DM alone group. Further analysis by AGTR tertiles revealed that lower AGTR was associated with higher MAFLD incidence, showing a negative correlation between AGTR and T2DM with MAFLD. Multivariate logistic regression confirmed AGTR as a protective factor for T2DM with MAFLD. AGTR, derived from ALB/GGT, has previously served as an independent predictor of coronary artery disease and been validated as a MAFLD protective factor, though its protective role in T2DM with MAFLD was previously unconfirmed. ALB is a crucial protein synthesized and metabolized by the liver; hepatic lipid accumulation promotes MAFLD development, impairs liver function, and reduces ALB synthesis. Conversely, decreased ALB promotes oxidative stress, further participating in MAFLD progression. GGT is a major enzyme localized on cell membranes; increased hepatic lipid deposition and decreased albumin stimulate GGT synthesis and release. Elevated GGT correlates with increased risks of cerebrovascular disease, diabetes, metabolic syndrome, and all-cause mortality, and enhances insulin resistance risk, which is intimately involved in MAFLD pathogenesis. Higher GGT levels thus promote MAFLD development, while GGT also serves as an important indicator of intrahepatic cholestasis, which is associated with MAFLD occurrence. Collectively, these mechanisms underscore AGTR's close relationship with T2DM and MAFLD.

This study also identified BMI and TG as independent risk factors for T2DM with MAFLD, consistent with previous research. Obesity increases MAFLD prevalence, and both obesity and elevated TG can cause insulin resistance, thereby promoting MAFLD progression. Elevated TG represents a characteristic feature of metabolic syndrome, which is the strongest risk factor for MAFLD, making TG elevation an independent risk factor for T2DM with MAFLD.

Using multivariate logistic regression, we identified risk factors for T2DM with MAFLD and, for the first time, constructed a risk prediction model with good internal validation and clinical predictive value. Our findings indicate this nomogram can effectively predict T2DM with MAFLD risk to enable early detection and intervention.

This study has several limitations. First, as a cross-sectional study, causal relationships cannot be established. Second, while liver biopsy is the gold standard for diagnosing fatty liver disease, this study used ultrasound, which may underestimate MAFLD incidence. Finally, data were collected only from Hebei General Hospital, limiting sample size and potentially introducing bias.

In conclusion, T2DM patients with MAFLD exhibited significantly lower AGTR levels than those without MAFLD, indicating AGTR is a protective factor for T2DM with MAFLD. The constructed risk prediction model effectively predicts T2DM with MAFLD risk.

Author Contributions: WEI Limin was responsible for feasibility analysis, conceptualization, and quality control. YAN Ziwei contributed to conceptualization, statistical analysis, interpretation, and manuscript writing. NIE Jiahua and SHI Yan collected, organized, and entered data. YAN Ziwei and WEI Limin took overall responsibility for the article.

Conflict of Interest: None declared.

References

[1] YOUNOSSI ZM, GOLABI P, PAIK JM, et al. The global epidemiology of nonalcoholic fatty liver disease (NAFLD) and nonalcoholic steatohepatitis (NASH): a systematic review. Hepatology. 2023;77(4):1335-1347. DOI: 10.1097/HEP.0000000000000004.

[2] RINELLA M, CHARLTON M. The globalization of nonalcoholic fatty liver disease: Prevalence and impact on world health. Hepatology. 2016;64(1):19-22. DOI: 10.1002/hep.28524.

[3] ESLAM M, SANYAL AJ, GEORGE J, et al. MAFLD: a consensus-driven proposed nomenclature for metabolic associated fatty liver disease. Gastroenterology. 2020;158(7):1999-2014.e1. DOI: 10.1053/j.gastro.2019.11.312.

[4] DECRAECKER M, DUTARTRE D, HIRIART JB, et al. Long-term prognosis of patients with metabolic (dysfunction)-associated fatty liver disease by non-invasive methods. Aliment Pharmacol Ther. 2022;55(5):580-592. DOI: 10.1111/apt.16760.

[5] MUTHIAH MD, HAN NC, SANYAL AJ. A clinical overview of non-alcoholic fatty liver disease: a guide to diagnosis, the clinical features, and complications-What the non-specialist needs to know. Diabetes Obes Metab. 2022;24(Suppl 2):3-14. DOI: 10.1111/dom.14521.

[6] BUTT AS, HAMID S, HAIDER Z, et al. Nonalcoholic fatty liver diseases among recently diagnosed patients with diabetes mellitus and risk factors. Euroasian J Hepatogastroenterol. 2019;9(1):9-13. DOI: 10.5005/jp-journals-10018-1288.

[7] LI XM, LIU SL, HE YJ, et al. Using new indices to predict metabolism dysfunction-associated fatty liver disease (MAFLD): analysis of the national health and nutrition examination survey database. BMC Gastroenterol. 2024;24(1):109. DOI: 10.1186/s12876-024-03190-2.

[8] YARIBEYGI H, ATKIN SL, SAHEBKAR A. A review of the molecular mechanisms of hyperglycemia-induced free radical generation leading to oxidative stress. J Cell Physiol. 2019;234(2):1300-1312. DOI: 10.1002/jcp.27164.

[9] ZHAO W, TONG JJ, LIU J, et al. The dose-response relationship between gamma-glutamyl transferase and risk of diabetes mellitus using publicly available data: a longitudinal study in Japan. Int J Endocrinol. 2020;2020:5356498. DOI: 10.1155/2020/5356498.

[10] LEE YS, CHO Y, BURGESS S, et al. Serum gamma-glutamyl transferase and risk of type 2 diabetes in the general Korean population: a Mendelian randomization study. Hum Mol Genet. 2016;25(17):3877-3886. DOI: 10.1093/hmg/ddw226.

[11] GABIR MM, HANSON RL, DABELEA D, et al. The 1997 American Diabetes Association and 1999 World Health Organization criteria for hyperglycemia in the diagnosis and prediction of diabetes. Diabetes Care. 2000;23(8):1108-1112. DOI: 10.2337/diacare.23.8.1108.

[12] YOUNOSSI ZM, KOENIG AB, ABDELATIF D, et al. Global epidemiology of nonalcoholic fatty liver disease-Meta-analytic assessment of prevalence, incidence, and outcomes. Hepatology. 2016;64(1):73-84. DOI: 10.1002/hep.28431.

[13] GOLABI P, PAIK JM, ALQAHTANI S, et al. Burden of non-alcoholic fatty liver disease in Asia, the Middle East and North Africa: Data from Global Burden of Disease 2009-2019. J Hepatol. 2021;75(4):795-809. DOI: 10.1016/j.jhep.2021.05.022.

[14] YOUNOSSI ZM, STEPANOVA M, ONG J, et al. Nonalcoholic steatohepatitis is the most rapidly increasing indication for liver transplantation in the United States. Clin Gastroenterol Hepatol. 2021;19(3):580-589.e5. DOI: 10.1016/j.cgh.2020.05.064.

[15] MCPHERSON S, HARDY T, HENDERSON E, et al. Evidence of NAFLD progression from steatosis to fibrosing-steatohepatitis using paired biopsies: implications for prognosis and clinical management. J Hepatol. 2015;62(5):1148-1155. DOI: 10.1016/j.jhep.2014.11.034.

[16] ARRESE M, BARRERA F, TRIANTAFILO N, et al. Concurrent nonalcoholic fatty liver disease and type 2 diabetes: diagnostic and therapeutic considerations. Expert Rev Gastroenterol Hepatol. 2019;13(9):849-866. DOI: 10.1080/17474124.2019.1649981.

[17] BESSONE F, RAZORI MV, ROMA MG. Molecular pathways of nonalcoholic fatty liver disease development and progression. Cell Mol Life Sci. 2019;76(1):99-128. DOI: 10.1007/s00018-018-2947-0.

[18] TOPAL S, KIZILTUNC E, SEZENOZ B, et al. Gamma-glutamyl transferase to albumin ratio can predict severity of coronary artery disease detected by coronary computed tomography angiography. Anatol J Cardiol. 2021;25(9):653-660. DOI: 10.5152/AnatolJCardiol.2021.36330.

[19] PEISELER M, SCHWABE R, HAMPE J, et al. Immune mechanisms linking metabolic injury to inflammation and fibrosis in fatty liver disease - novel insights into cellular communication circuits. J Hepatol. 2022;77(4):1136-1160. DOI: 10.1016/j.jhep.2022.06.012.

[20] NEGRE-SALVAYRE A, SALVAYRE R, AUGE N, et al. Hyperglycemia and glycation in diabetic complications. Antioxid Redox Signal. 2009;11(12):3071-3109. DOI: 10.1089/ars.2009.2484.

[21] ORTEGA E, KOSKA J, SALBE AD, et al. Serum gamma-glutamyl transpeptidase is a determinant of insulin resistance independently of adiposity in Pima Indian children. J Clin Endocrinol Metab. 2006;91(4):1419-1422. DOI: 10.1210/jc.2005-1783.

[22] FIGUEIRA TR, VERCESI AE, OLIVEIRA HCF. Lack of plasma albumin impairs intravascular lipolysis and explains the associated free fatty acids deficiency and hypertriglyceridemia. Lipids Health Dis. 2010;9:146. DOI: 10.1186/1476-511X-9-146.

[23] NDREPEPA G, COLLERAN R, KASTRATI A. Gamma-glutamyl transferase and the risk of atherosclerosis and coronary heart disease. Clin Chim Acta. 2018;476:130-138. DOI: 10.1016/j.cca.2017.11.026.

[24] YI SW, LEE SH, HWANG HJ, et al. Gamma-glutamyltransferase and cardiovascular mortality in Korean adults: a cohort study. Atherosclerosis. 2017;265:102-109. DOI: 10.1016/j.atherosclerosis.2017.08.028.

[25] HOSSAIN IA, RAHMAN SHAH MM, RAHMAN MK, et al. Gamma glutamyl transferase is an independent determinant for the association of insulin resistance with nonalcoholic fatty liver disease in Bangladeshi adults: Association of GGT and HOMA-IR with NAFLD. Diabetes Metab Syndr. 2016;10(1 Suppl 1):S25-S29. DOI: 10.1016/j.dsx.2015.09.005.

[26] LIU D, ZHOU L, YANG M, et al. Oxidative stress mediates the association between dietary fat intake and cognition in US older adults. Am J Geriatr Psychiatry. 2022;30(7):761-773. DOI: 10.1016/j.jagp.2022.01.001.

[27] SHI ZJ, CHEN G, CAO Z, et al. Gut microbiota and its metabolite deoxycholic acid contribute to sucralose consumption-induced nonalcoholic fatty liver disease. J Agric Food Chem. 2021;69(13):3982-3991. DOI: 10.1021/acs.jafc.0c07467.

[28] LIU JY, AYADA I, ZHANG XF, et al. Estimating global prevalence of metabolic dysfunction-associated fatty liver disease in overweight or obese adults. Clin Gastroenterol Hepatol. 2022;20(3):e573-e582. DOI: 10.1016/j.cgh.2021.02.030.

[29] MISHRA SR, WALLER M, CHUNG HF, et al. Epidemiological studies of the association between reproductive lifespan characteristics and risk of Type 2 diabetes and hypertension: a systematic review. Maturitas. 2022;155:14-23. DOI: 10.1016/j.maturitas.2021.09.009.

[30] GASEVIC D, FROHLICH J, JOHN MANCINI GB, et al. The association between triglyceride to high-density-lipoprotein cholesterol ratio and insulin resistance in a multiethnic primary prevention cohort. Metabolism. 2012;61(4):583-589. DOI: 10.1016/j.metabol.2011.09.009.

[31] HUANG PL. A comprehensive definition for metabolic syndrome. Dis Model Mech. 2009;2(5/6):231-237. DOI: 10.1242/dmm.001180.

[32] FRIEDMAN SL, NEUSCHWANDER-TETRI BA, RINELLA M, et al. Mechanisms of NAFLD development and therapeutic strategies. Nat Med. 2018;24(7):908-922. DOI: 10.1038/s41591-018-0104-9.

(Received: January 10, 2025; Revised: May 15, 2025)
(Editor: ZHAO Yuecui)