https://orcid.org/0000-0002-8732-0191
Figures
Abstract
Background
TyG-BMI has been proposed as a marker of insulin resistance in metabolic-associated fatty liver disease, but its clinical utility remains uncertain. This study aims to evaluate the association between TyG-BMI and metabolic dysfunction-associated steatotic liver disease (MASLD) through a systematic review and meta-analysis, focusing on the diagnostic performance across different subgroups.
Methods
A comprehensive literature search was conducted in PubMed, Scopus, Embase, and Web of Science up to January 20, 2025. Studies evaluating the relationship between TyG-BMI and MASLD in adults were included. A random-effects model was employed to pool effect sizes, and subgroup analyses were conducted based on sex, disease definition, and population type.
Results
Thirty-five studies with 339,087 participants were included. The pooled mean difference for TyG-BMI between MASLD and non-MASLD groups was 42.72 (95% CI: 35.93–49.51; p < 0.0001). Subgroup analysis revealed higher mean differences in the metabolic-associated fatty liver disease (MAFLD) group (49.56, 95% CI: 39.38–59.74) compared to non-alcoholic fatty liver disease ase (NAFLD) (34.68, 95% CI: 28.45–40.91). The odds ratio per one-unit increment of the TyG-BMI was 1.05 (95% CI: 1.03–1.08). Sensitivity for TyG-BMI in diagnosing MASLD was 0.79 (95% CI: 0.73–0.84), and specificity was 0.76 (95% CI: 0.71–0.80). The pooled area under the curve (AUC) for TyG-BMI was 0.83 (95% CI: 0.81–0.86), with better performance in females (0.88) compared to males (0.83). Subgroup analysis by disease definition showed a higher AUC for MAFLD (0.87) compared to NAFLD (0.81).
Citation: Boushehri YG, Meymanatabadi Z, Tanha AE, Azami P, Alaei M, Alamdari AA, et al. (2025) Association of triglyceride glucose-body mass index (TyG-BMI) with metabolic dysfunction-associated steatotic liver disease: A systematic review and meta-analysis. PLoS One 20(8): e0324483. https://doi.org/10.1371/journal.pone.0324483
Editor: Anna Di Sessa, Universita degli Studi della Campania Luigi Vanvitelli Scuola di Medicina e Chirurgia, ITALY
Received: February 27, 2025; Accepted: April 25, 2025; Published: August 4, 2025
Copyright: © 2025 Boushehri et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the paper and its Supporting Information files.
Funding: The author(s) received no specific funding for this work.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Metabolic dysfunction-associated steatotic liver disease (MASLD), formerly known as non-alcoholic fatty liver disease (NAFLD), is one of the most common causes of chronic liver disease worldwide, with an increasing prevalence due to the rising global burden of metabolic disorders such as obesity, type 2 diabetes mellitus (T2DM), and hypertension [1]. In comparison to NAFLD, MASLD is characterized by excessive fat accumulation in the liver, independent of significant alcohol consumption, and is associated with various metabolic abnormalities, including insulin resistance, dyslipidemia, and hypertension. As MASLD progresses, it may lead to severe complications, such as liver cirrhosis, hepatocellular carcinoma, and cardiovascular disease, making early diagnosis and intervention essential [2,3].
The TyG index, calculated as the product of fasting triglycerides and fasting glucose, has been recognized as a reliable surrogate marker for insulin resistance. Research indicates that the TyG index alone is a useful tool for identifying individuals at risk for T2DM, cardiovascular diseases, and metabolic syndrome [4]. The TyG index was recently shown to be a useful tool for identifying patients with MASLD, especially in lean patients [5,6]. Recently, other TyG index-related surrogate markers, such as TyG-waist circumference (WC), and TyG-waist circumference-to-height ratio (WHtR) have emerged. The triglyceride glucose-body mass index (TyG-BMI) is an emerging, simple, and cost-effective biomarker that combines triglyceride levels, glucose levels, and body mass index (BMI). TyG-BMI has been proposed as a surrogate marker for insulin resistance and a predictor of various disorders, including MASLD, cardiovascular diseases, and T2DM [7,8]. When BMI is included in the calculation, the resulting TyG-BMI index provides a more nuanced understanding of metabolic health, particularly in populations with varying body compositions. Studies have shown that TyG-BMI outperforms the TyG index in predicting insulin resistance, as it accounts for body fat distribution and obesity’s role in metabolic health [9].
Several studies have investigated the association between TyG-BMI and MASLD, but the results remain inconsistent, particularly when considering differences by sex and across various study groups [10–12]. This inconsistency highlights the need for further research to address these discrepancies and provide a more comprehensive understanding of the relationship between TyG-BMI and MASLD. Given the clinical importance of identifying reliable biomarkers for early detection and risk stratification of MASLD, a comprehensive evaluation of the available evidence is needed.
This systematic review and meta-analysis aim to synthesize the existing literature on the association between TyG-BMI and MASLD. By pooling data from multiple studies, we seek to provide a clearer understanding of the potential role of TyG-BMI as a diagnostic and prognostic tool in MASLD and to explore the strength and consistency of this association across different populations and study designs. We have used the latest term, MASLD, throughout the manuscript. However, all fatty liver diseases defined by NAFLD, MAFLD, and MASLD were considered in this study. Also, due to the limited number of studies using MASLD definition, we merged MASLD and MAFLD patients for subgroup analysis by disease definition.
Materials and methods
Study design
This systematic review and meta-analysis aimed to investigate the association between the TyG-BMI index and MASLD. The review was conducted following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines (Registration number: CRD42024554538).
Literature search
We conducted a comprehensive literature search in the following electronic databases: PubMed, Scopus, Embase, and Web of Science. The search strategy was developed using a combination of MeSH terms and keywords related to “TyG-BMI,” “NAFLD,” “MASLD, “ and “MAFLD.” Studies were investigated from inception up to January 20, 2025. We also reviewed the reference lists of included studies and relevant reviews to identify additional articles. The full version of the search strategy is available in S1 Table.
Eligibility criteria
We included observational studies that assessed the relationship between TyG-BMI MASLD in adult populations. Studies meeting the following criteria with no time and language restrictions were eligible: (1) adults aged 18 years or older; (2) studies that measured TyG-BMI (formula: TyG-BMI = (ln(fasting triglycerides (mg/dL)×fasting glucose (mg/dL))/2)×BMI) and evaluated its association and predictability power in diagnosing MASLD; (3) observational studies, including cross-sectional, cohort, and case-control studies. Studies were excluded if they: focused on animal models or children, used non-validated diagnostic criteria for MASLD, and did not report sufficient data to calculate effect sizes or outcomes related to TyG-BMI. Letters, commentaries, reviews, case reports, and case series were also deemed ineligible and, therefore, excluded from the systematic review.
After removing duplicates, two authors (YGB and ZM), independently, assessed the titles and abstracts of all identified papers to determine eligibility according to the predefined inclusion and exclusion criteria. Subsequently, both authors independently conducted a thorough review of the full texts of studies that met these criteria. Any disagreements during the review process were resolved through consensus.
Data extraction
Two independent reviewers (YGB and AET) extracted the following data from each eligible study including, study characteristics (first author, year of publication, study design, sample size, country of origin, follow-up duration), population characteristics (age, sex, and comorbid conditions), and outcome data (MASLD diagnosis criteria, effect estimates (odds ratios [OR], hazard ratios [HR], mean/median TyG-BMI in each group, sensitivity, specificity, area under the curve [AUC]), and their 95% confidence intervals [CIs] and standard deviation [SD]). Discrepancies between the two reviewers were resolved through discussion or by consulting a third reviewer (MA).
Risk of bias assessment
The methodological quality and risk of bias of included studies were assessed using the Newcastle-Ottawa Scale (NOS) for cohort, cross-sectional, and case-control studies. Two independent authors (AH and PA) assessed the qualities and in case of disagreement, a third author (AAA) resolved the issue. Studies scoring 7–9 stars were considered high quality, indicating a low risk of bias. Those scoring 4–6 stars were deemed to be of fair quality, suggesting a moderate risk of bias. Studies scoring 0–3 stars were classified as poor quality, indicating a high risk of bias.
Statistical analysis
The statistical analysis for this meta-analysis was performed using R version 4.4.2 with the ‘meta’ and ‘metafor’ packages [13]. The primary outcomes of interest included the AUC, sensitivity, specificity, ORs, and mean differences of the TyG-BMI in association with MASLD. For each included study, we extracted AUCs, ORs, sensitivity, specificity, and TyG-BMI mean in each study group, along with their respective CIs or SDs. Among studies using the same database (e.g., NHANES), we included the most general population with the highest sample size. The methods proposed by Wan et al. [14] and Luo et al. [15] were used to convert median and interquartile range values into means and SDs. In cases where the 95% CI for the AUC was not provided, we applied the Hanley and McNeil (1982) [16] method to calculate it. A random-effects model was employed to pool the data, accounting for both within-study and between-study variability. Specifically, we used the restricted maximum likelihood (REML) method to estimate the between-study variance (τ²), which is particularly suitable given the expected heterogeneity among the studies. Heterogeneity was assessed using the I² statistic and Cochran’s Q test. A p-value of < 0.05 was considered statistically significant. An I² value greater than 50% or a significant Q test (p < 0.10) was considered indicative of substantial heterogeneity. Subgroup analyses were performed based on sex, disease definition, and population type where applicable. Also, due to the limited number of studies using MASLD definition, we merged MASLD and MAFLD patients for subgroup analysis by disease definition. Publication bias was assessed using Begg’s [17] and Egger’s tests [18] and visual inspection of funnel plots. Asymmetry in the funnel plots or a significant Egger’s test (p < 0.05) was considered indicative of potential publication bias.
Result
Study selection
The initial search identified a total of 204 results: 57 from Embase, 47 from ISI, 50 from PubMed, and 50 from Scopus (S2 Table). After merging the results and removing duplicates, 59 unique studies remained. These 59 studies were screened for eligibility based on titles and abstracts, resulting in the selection of 43 studies for full-text review. Following a thorough review of the full texts, eight studies were excluded (3 due to an undesired measure of effect, 2 as conference abstracts, and 3 due to undesired outcomes) (S3 Table). As a result, 35 studies were deemed eligible and included in the qualitative synthesis for the meta-analysis. Fig 1 presents a flow diagram of the study selection process.
Basic characteristics
Overall, 35 studies with a total number of 339,087 individuals from countries including China (n = 23), Korea (n = 4), Iran (n = 3), Japan (n = 2), Tanzania (n = 1), Mexico (n = 1), Ghana (n = 1). Of the included studies, 33 had a cross-sectional design, one was a cohort study, and one was a case-control study. Four studies used the NHANES database, and two studies used the ‘DATADRYAD’ database. Among the included studies, 24 focused on the general population, 5 on individuals with T2DM, 4 on normal BMI individuals, and 2 on overweight individuals and patients with obesity. Ultrasound, transient elastography, and CT-scan were among the imaging techniques to diagnose fatty liver disease. The basic characteristics of the included studies and effect sizes are summarized in Tables 1 and 2.
The mean difference of TyG-BMI in MASLD vs. non-MASLD groups
The random effects model for the mean difference analysis included 24 studies and yielded a pooled mean difference of 42.7214 (95% CI: 35.9331–49.5096; z = 12.33; p < 0.0001). Substantial heterogeneity was observed (I² = 99.3%, Q = 3185.45, df = 23, p < 0.0001). Patients with MASLD have a significantly higher TyG-BMI compared to patients without MASLD (Fig 2).
In the subgroup analysis by disease definition, 11 studies focused on NAFLD with a mean difference of 34.6765 (95% CI: 28.4465–40.9065; τ² = 97.6531; τ = 9.8820), while 13 studies on MAFLD reported a mean difference of 49.5614 (95% CI: 39.3799–59.7428; τ² = 342.3150; τ = 18.5018). The test for subgroup differences revealed significant heterogeneity between the two groups (Q = 5.97, df = 1, p = 0.0145). TyG-BMI mean difference was significantly higher in the MAFLD group compared to the NAFLD group (S1 Fig).
In the subgroup analysis by population type, 17 studies focusing on the general population reported a mean difference of 49.9946 (95% CI: 43.6121–56.3770; τ² = 172.4196; τ = 13.1309). For the T2DM subgroup (4 studies), the mean difference was 19.4570 (95% CI: 17.7224–21.1917; τ² < 0.0001; τ = 0.0015). The subgroup of overweight participants and patients with obesity (2 studies) yielded a mean difference of 25.3400 (95% CI: 13.2598–37.4202; τ² = 42.3958; τ = 6.5112), while the non-obese subgroup (2 studies) had a mean difference of 31.0773 (95% CI: 25.8420–36.3126; τ² = 11.4452; τ = 3.3831). The test for subgroup differences indicated significant heterogeneity between subgroups (Q = 93.63, df = 3, p < 0.0001) (S2 Fig).
The odds ratio for the association of TyG-BMI with MASLD
The random effects model for the odds ratio (OR) of TyG-BMI fourth quartile vs. first quartile meta-analysis, based on 13 studies, yielded an odds ratio of 29.3217 (95% CI: 15.8361–54.2914; z = 10.75; p < 0.0001). The test of heterogeneity was significant (I² = 97.5%, Q = 483.48, df = 12, p < 0.0001) (Fig 3). For a 1-unit increment, the random effects model included 11 studies and reported an OR of 1.0517 (95% CI: 1.0252–1.0788; z = 3.87; p = 0.0001, I² = 98.0%, Q = 495.27, df = 10, p < 0.0001). (Fig 4A). For a 1-SD increment, based on 5 studies, the random effects model reported an OR of 3.7306 (95% CI: 3.1255–4.4529; z = 14.58; p < 0.0001). The test of heterogeneity was also significant (I² = 94.3%, Q = 70.19, df = 4, p < 0.0001). (Fig 4B).
Sensitivity and specificity of TyG-BMI for diagnosing MASLD
The sensitivity meta-analysis, based on 23 studies, revealed a pooled sensitivity of 0.7876 (95% CI: 0.7274–0.8375) under the random effects model (I² = 99.3%) (Fig 5). Subgroup analysis by disease definition indicated a sensitivity of 0.7548 (95% CI: 0.6493–0.8365) for NAFLD and 0.8229 (95% CI: 0.7722–0.8643) for MAFLD. However, the test for subgroup differences was not significant (Q = 1.82, p = 0.1776) (S3 Fig).
The specificity meta-analysis, based on the same 23 studies reported a pooled specificity of 0.7557 (95% CI: 0.7074–0.7983) under the random effects model (I² = 99.7%) (Fig 6). Subgroup analysis by disease definition showed specificities of 0.7492 (95% CI: 0.6726–0.8128) for NAFLD and 0.7654 (95% CI: 0.7050–0.8167) for MAFLD, with no significant subgroup differences (Q = 0.13, p = 0.7224) (S4 Fig).
Subgroup analysis by sex indicated a sensitivity of 0.8117 (95% CI: 0.7772–0.8419) for males and 0.8677 (95% CI: 0.8222–0.9029) for females under the random effects model, with significant subgroup differences (Q = 4.05, p = 0.0441). Specificity results were 0.7487 (95% CI: 0.6941–0.7963) for males and 0.7498 (95% CI: 0.6801–0.8086) for females, with no significant subgroup differences (Q = 0.00, p = 0.9787) (S5 and S6 Figs).
Subgroup analysis by population type showed a sensitivity of 0.8034 (95% CI: 0.7340–0.8583) for the general population, 0.7488 (95% CI: 0.5325–0.8864) for individuals with T2DM, 0.6984 (95% CI: 0.5568–0.8102) for overweight participants and patients with obesity, and 0.7824 (95% CI: 0.7464–0.8145) for non-obese population under the random effects model, with no significant differences between subgroups (Q = 2.51, p = 0.4733). Specificity results were 0.7709 (95% CI: 0.7111–0.8215) for the general population, 0.7463 (95% CI: 0.7173–0.7733) for T2DM, 0.6416 (95% CI: 0.4596–0.7903) for overweight participants and patients with obesity, and 0.7346 (95% CI: 0.7038–0.7633) for non-obese populations, with no significant subgroup differences (Q = 2.86, p = 0.4142) (S7 and S8 Figs).
Area under the curve for TyG-BMI in MASLD prediction
The pooled analysis of 30 studies assessing the association between the TyG-BMI and MASLD demonstrated high diagnostic performance, with an overall AUC of 0.8343 (95% CI: 0.8101, 0.8584; p < 0.0001). However, significant heterogeneity was observed across the studies, with an I² value of 99.4%, τ² of 0.0042, and a heterogeneity test p-value < 0.0001 (Fig 7).
In the sex-specific subgroup analysis, the pooled AUC for the male population, based on 17 studies, was 0.8323 (95% CI: 0.8053, 0.8592; p < 0.0001), with substantial heterogeneity (I² = 98.8%, τ² = 0.0029). Similarly, for the female population, also based on 17 studies, the pooled AUC was higher at 0.8770 (95% CI: 0.8445, 0.9095; p < 0.0001), with significant heterogeneity (I² = 98.0%, τ² = 0.0043). A comparison of AUCs between males and females revealed a statistically significant difference (Q = 4.31, p = 0.0378). TyG-BMI had a better diagnostic performance in females compared to males for MASLD (S9 Fig).
The analysis also evaluated differences based on disease definition. For studies investigating MAFLD, the pooled AUC from 12 studies was 0.8650 (95% CI: 0.8266, 0.9034; p < 0.0001), with high heterogeneity (I² = 98.7%, τ² = 0.0043). In contrast, for studies focusing on NAFLD, the pooled AUC from 18 studies was slightly lower at 0.8147 (95% CI: 0.7868, 0.8425; p < 0.0001), with similarly high heterogeneity (I² = 98.3%, τ² = 0.0033). The test for subgroup differences between MAFLD and NAFLD showed statistically significant results (Q = 4.32, p = 0.0377). TyG-BMI had a better diagnostic performance for patients with MAFLD compared to NAFLD (S10 Fig).
Subgroup analysis by population type further revealed variations in the diagnostic performance of the TyG-BMI index. For the general population, comprising 21 studies, the pooled AUC was 0.8604 (95% CI: 0.8376, 0.8831; p < 0.0001), with high heterogeneity (I² = 99.5%, τ² = 0.0027). Among patients with T2DM, the pooled AUC from five studies was lower at 0.7564 (95% CI: 0.7302, 0.7825; p < 0.0001), with minimal heterogeneity (I² = 31.6%, τ² = 0.0003). For patients with obesity or overweight populations, based on two studies, the pooled AUC was 0.6947 (95% CI: 0.6429, 0.7465; p < 0.0001), with no heterogeneity (I² = 0%). In the non-obese population, based on four studies, the pooled AUC was 0.8305 (95% CI: 0.8053, 0.8557; p < 0.0001, I² = 89.1%,). The test for subgroup differences by population type was statistically significant (Q = 56.49, p < 0.0001) (S11 Fig).
Risk of bias assessment
Independent investigators evaluated the quality of the studies included in the review using the modified Newcastle-Ottawa Scale (NOS), specific to cohort, case-control, and cross-sectional studies. The potential for bias in these studies was minimal. Quality assessments are presented in Table 3. Most studies received high scores, indicating good quality, while only one scored as fair. None were classified as having a “poor” score.
The funnel plot for publication bias is available in Fig 8. Egger’s regression test was conducted to evaluate funnel plot asymmetry. The test result indicated a t-value of −2.11 and a p-value of 0.0440, suggesting statistically significant evidence of publication bias (p < 0.05). The bias estimate was −6.3135, with a standard error (SE) of 2.9934. The multiplicative residual heterogeneity variance (tau^2) was 139.4794. Additionally, we used the rank correlation test by Begg and Mazumdar (1993) to further examine funnel plot asymmetry. The test result yielded a z-value of 0.38 with a p-value of 0.7076, indicating no statistically significant evidence of publication bias (p > 0.05). The bias estimate was 21.0000, with a standard error (SE) of 55.9911. The discrepancy between the two tests highlights the importance of using multiple methods to assess publication bias and interpreting the results with caution.
Discussion
This systematic review and meta-analysis demonstrated a strong association between the TyG-BMI index and the fatty liver disease defined by NAFLD, MAFLD and MASLD. Individuals with these conditions had significantly higher TyG-BMI values compared to those without. The association was more pronounced in MAFLD than in NAFLD. Additionally, an increase in the TyG-BMI index was linked to higher odds of having MASLD. The diagnostic performance of TyG-BMI was notable, showing a relatively high sensitivity and specificity. Overall, the index demonstrated strong diagnostic accuracy, with better performance in females compared to males, in MAFLD compared to NAFLD, with the highest AUC value in the general population and the lowest value in patients with T2DM.
The TyG index was indeed first introduced by Simental-Mendía et al. in 2008 [50]. In their study, they explored the use of fasting triglyceride and glucose levels as a marker for insulin resistance, which is a key factor in metabolic disorders. The TyG index was shown to be a valuable surrogate for assessing insulin resistance in clinical practice, providing a more accessible alternative to more complex and expensive methods [50]. TyG-BMI is an emerging composite indicator that combines the TyG index, a reliable surrogate marker of insulin resistance, with BMI, a widely used measure of obesity. By integrating metabolic and anthropometric parameters, TyG-BMI provides a more comprehensive assessment of metabolic dysfunction compared to either marker alone [51].
Based on our findings, in almost all the included studies investigating the association of TyG-related indices, TyG-BMI was superior to TyG index in predicting MASLD; however, results regarding the superiority of TyG-BMI over TyG-WC and TyG-WHtR were inconsistent. The measured AUC in our study for the diagnostic performance of TyG-BMI in MASLD was higher than the diagnostic performance of the TyG index (0.834 vs. 0.75) in a study conducted by Wang et. al [52]. The TyG-BMI includes BMI, which reflects overall adiposity and body weight status, but the TyG index alone does not account for body composition or fat distribution, which are crucial in metabolic dysfunction. This makes TyG-BMI a more comprehensive indicator of metabolic health, as obesity or excess weight plays a critical role in metabolic disorders like MASLD [25]. Moreover, TyG-BMI has been shown in studies to correlate better with insulin resistance compared to the TyG index. Insulin resistance is a major contributor to conditions like MASLD, making TyG-BMI a stronger marker for identifying at-risk populations [53,54]. While our analysis focused on TyG-BMI, future research should directly compare the predictive performance of TyG-BMI against central obesity-adjusted indices (e.g., TyG-WC, TyG-WHtR) in well-characterized cohorts with standardized adiposity measurements. Pooled analyses of such studies may help resolve current controversies regarding optimal index selection.
Our subgroup analyses showed that TyG-BMI performed as a better predictor in females compared to males. Overall, females have a lower risk of MASLD compared to males [55]. However, in our systematic review, the mean/median age in most of the studies included was above 45 years. At this age, most women are undergoing menopause, characterized by declining estrogen levels. Research has shown that postmenopausal women have a prevalence of MASLD comparable to that of men, likely due to increased weight gain, fat redistribution, and dyslipidemia, all of which contribute to a higher risk of MASLD [55]. Furthermore, our findings underscore the differences in MASLD when the same increase in TyG-BMI is applied across different sexes. This may be attributed to sex-based variations in carbohydrate and lipid metabolism, as well as menopausal changes in body fat distribution and heightened susceptibility to metabolic complications [56].
The subgroup analyses by disease definition demonstrated that TyG-BMI performed as a better predictor in the MAFLD group compared to the NAFLD group. The diagnosis of MAFLD incorporates metabolic dysfunction as a primary criterion (e.g., insulin resistance, obesity, and metabolic syndrome), in addition to hepatic steatosis. TyG-BMI, which reflects both insulin resistance and obesity, aligns closely with these metabolic parameters [57]. Nevertheless, NAFLD criteria do not require the presence of metabolic dysfunction [58]. Thus, TyG-BMI may be less predictive in NAFLD because it primarily focuses on metabolic markers. Also, obesity is a core diagnostic criterion for MAFLD and is a strong driver of hepatic fat accumulation and systemic metabolic dysfunction [57]. Since TyG-BMI incorporates BMI, it directly captures the role of obesity in MAFLD.
The findings of this meta-analysis support the use of TyG-BMI as a non-invasive, cost-effective, and accessible marker for identifying individuals at risk of MASLD. Its strong association with insulin resistance and superior diagnostic performance, particularly in specific subgroups such as females and patients with MAFLD, highlights its potential for use in clinical settings. However, clinicians should consider population-specific factors, including age, sex, and metabolic profiles, when interpreting TyG-BMI values.
Our study was the first to determine the relationship between the TyG-BMI and MASLD. Moreover, the high number of included studies and subgroup analyses can provide good evidence and clues for further research. While this meta-analysis provides robust evidence for the association and diagnostic utility of TyG-BMI in MASLD, several limitations should be acknowledged. First, the included studies were heterogeneous in terms of population characteristics, study design, and diagnostic criteria for MASLD. Although the subgroup analyses partially addressed this issue, residual heterogeneity may affect the generalizability of the findings. Second, most studies were cross-sectional, limiting the ability to infer causality or assess the predictive value of TyG-BMI over time. Third, most studies were conducted on the Chinese population and few studies included diverse populations, such as those from low- and middle-income countries, where the burden of MASLD is rapidly increasing. Fourth, the cutoff values for the TyG-BMI varied among the included studies, which could result in differences in classifying individuals into high or low TyG-BMI groups. This variability may impact the observed associations with MASLD outcomes (for Q4 vs. Q1 OR and AUC). Fifth, BMI differences in different study populations were not investigated due to limitation in data. Lastly, differences in the adjustment for confounding factors, as well as inadequate adjustment, can result in biased estimates of the association. These limitations highlight the need for future research on the association of TyG-BMI with MASLD.
Supporting information
S2 Table. List of all studies identified in literature search.
https://doi.org/10.1371/journal.pone.0324483.s002
(DOCX)
S3 Table. Excluded studies after full text review.
https://doi.org/10.1371/journal.pone.0324483.s003
(DOCX)
S1 Fig. Mean difference of TyG-BMI in MASLD vs. non-MASLD patients in subgroups categorized by disease definition.
https://doi.org/10.1371/journal.pone.0324483.s005
(TIF)
S2 Fig. Mean difference of TyG-BMI in MASLD vs. non-MASLD patients in subgroups categorized by population type.
https://doi.org/10.1371/journal.pone.0324483.s006
(TIF)
S3 Fig. Sensitivity of TyG-BMI for diagnosing MASLD in subgroups categorized by disease definition.
https://doi.org/10.1371/journal.pone.0324483.s007
(TIF)
S4 Fig. Specificity of TyG-BMI for diagnosing MASLD in subgroups categorized by disease definition.
https://doi.org/10.1371/journal.pone.0324483.s008
(TIF)
S5 Fig. Sensitivity of TyG-BMI for diagnosing MASLD in subgroups categorized by sex.
https://doi.org/10.1371/journal.pone.0324483.s009
(TIF)
S6 Fig. Specificity of TyG-BMI for diagnosing MASLD in subgroups categorized by sex.
https://doi.org/10.1371/journal.pone.0324483.s010
(TIF)
S7 Fig. Sensitivity of TyG-BMI for diagnosing MASLD in subgroups categorized by population type.
https://doi.org/10.1371/journal.pone.0324483.s011
(TIF)
S8 Fig. Specificity of TyG-BMI for diagnosing MASLD in subgroups categorized by population type.
https://doi.org/10.1371/journal.pone.0324483.s012
(TIF)
S9 Fig. Area under the curve for TyG-BMI in MASLD prediction in subgroups categorized by sex.
https://doi.org/10.1371/journal.pone.0324483.s013
(TIF)
S10 Fig. Area under the curve for TyG-BMI in MASLD prediction in subgroups categorized by disease definition.
https://doi.org/10.1371/journal.pone.0324483.s014
(TIF)
S11 Fig. Area under the curve for TyG-BMI in MASLD prediction in subgroups categorized by population type.
https://doi.org/10.1371/journal.pone.0324483.s015
(TIF)
References
- 1. Khanmohammadi S, Ramos-Molina B, Kuchay MS. NOD-like receptors in the pathogenesis of metabolic (dysfunction)-associated fatty liver disease: Therapeutic agents targeting NOD-like receptors. Diabetes Metab Syndr. 2023;17(7):102788. pmid:37302383
- 2. Khanmohammadi S, Kuchay MS. Toll-like receptors and metabolic (dysfunction)-associated fatty liver disease. Pharmacol Res. 2022;185:106507. pmid:36252773
- 3. Khanmohammadi S, Masrour M, Fallahtafti P, Habibzadeh A, Schuermans A, Kuchay MS. The relationship between nonalcoholic fatty liver disease and frailty: A systematic review and meta-analysis. Diabetes Metab Syndr. 2025;19(2):103187. pmid:39798236
- 4. Sun Y, Ji H, Sun W, An X, Lian F. Triglyceride glucose (TyG) index: A promising biomarker for diagnosis and treatment of different diseases. Eur J Intern Med. 2025;131:3–14. pmid:39510865
- 5. Ying Y, Ji Y, Ju R, Chen J, Chen M. Association between the triglyceride-glucose index and liver fibrosis in adults with metabolism-related fatty liver disease in the United States: a cross-sectional study of NHANES 2017-2020. BMC Gastroenterol. 2025;25(1):3. pmid:39748306
- 6. Zhu R, Xu C, Jiang S, Xia J, Wu B, Zhang S, et al. Risk factor analysis and predictive model construction of lean MAFLD: a cross-sectional study of a health check-up population in China. Eur J Med Res. 2025;30(1):137. pmid:40001266
- 7. Rao X, Xin Z, Yu Q, Feng L, Shi Y, Tang T, et al. Triglyceride-glucose-body mass index and the incidence of cardiovascular diseases: a meta-analysis of cohort studies. Cardiovasc Diabetol. 2025;24(1):34. pmid:39844258
- 8. Hu M, et al. Prediction of MASLD using different screening indexes in Chinese type 2 diabetes mellitus. Diabetol Metab Syndr. 2025;17(1):10.
- 9. Lim J, Kim J, Koo SH, Kwon GC. Comparison of triglyceride glucose index, and related parameters to predict insulin resistance in Korean adults: An analysis of the 2007-2010 Korean National Health and Nutrition Examination Survey. PLoS One. 2019;14(3):e0212963. pmid:30845237
- 10. Bockarie AS, Nartey YA, Nsiah P, Edzie EKM, Tuoyire D, Acquah S, et al. Fatty liver biomarkers and insulin resistance indices in the prediction of non-alcoholic fatty liver disease in Ghanaian patients. Endocrinol Diabetes Metab. 2023;6(6):e456. pmid:37814510
- 11. Chang M, Shao Z, Shen G. Association between triglyceride glucose-related markers and the risk of metabolic-associated fatty liver disease: a cross-sectional study in healthy Chinese participants. BMJ Open. 2023;13(5):e070189. pmid:37130686
- 12. Kilonzo SB, Kamala E, Jaka H, Ngoya P. Non-alcoholic fatty liver disease in Tanzania: prevalence, determinants, and diagnostic performance of triglycerides-glucose index and triglycerides-glucose index -body mass index compared to the hepatic ultrasound in overweight and obese individuals. BMC Gastroenterol. 2024;24(1):96. pmid:38438930
- 13. Balduzzi S, Rücker G, Schwarzer G. How to perform a meta-analysis with R: a practical tutorial. Evid Based Ment Health. 2019;22(4):153–60. pmid:31563865
- 14. Wan X, Wang W, Liu J, Tong T. Estimating the sample mean and standard deviation from the sample size, median, range and/or interquartile range. BMC Med Res Methodol. 2014;14:135. pmid:25524443
- 15. Luo D, Wan X, Liu J, Tong T. Optimally estimating the sample mean from the sample size, median, mid-range, and/or mid-quartile range. Stat Methods Med Res. 2018;27(6):1785–805. pmid:27683581
- 16. Hanley JA, McNeil BJ. The meaning and use of the area under a receiver operating characteristic (ROC) curve. Radiology. 1982;143(1):29–36.
- 17. Begg CB, Mazumdar M. Operating characteristics of a rank correlation test for publication bias. Biometrics. 1994;50(4):1088–101. pmid:7786990
- 18. Egger M, Davey Smith G, Schneider M, Minder C. Bias in meta-analysis detected by a simple, graphical test. BMJ. 1997;315(7109):629–34. pmid:9310563
- 19. Chen B, Peng X-E. Association between serum uric acid to creatinine ratio and metabolic-associated fatty liver disease in Southeast China (TyG-BMI as a potential mediator). Diabetes Metab Syndr Obes. 2024;17:4711–20. pmid:39678226
- 20. Gao Q, Feng L, Zhou W, Li X, Yin L, Wang Y. Non-traditional blood lipid indices for metabolism dysfunction-associated fatty liver disease prediction in non-obese type 2 diabetes mellitus. Diabetes Metab Syndr Obes. 2023;16:2345–54. pmid:37577041
- 21. Han AL, Lee HK, Shin SR. Diagnostic performance of insulin resistance indices for identifying metabolic dysfunction-associated fatty liver disease. Metab Syndr Relat Disord. 2024;22(5):402–9. pmid:38574322
- 22. Hu H, Han Y, Cao C, He Y. The triglyceride glucose-body mass index: a non-invasive index that identifies non-alcoholic fatty liver disease in the general Japanese population. J Transl Med. 2022;20(1):398. pmid:36064712
- 23. Khamseh ME, Malek M, Abbasi R, Taheri H, Lahouti M, Alaei-Shahmiri F. Triglyceride glucose index and related parameters (triglyceride glucose-body mass index and triglyceride glucose-waist circumference) identify nonalcoholic fatty liver and liver fibrosis in individuals with overweight/obesity. Metab Syndr Relat Disord. 2021;19(3):167–73. pmid:33259744
- 24. Khamseh ME, Malek M, Jahangiri S, Nobarani S, Hekmatdoost A, Salavatizadeh M, et al. Insulin resistance/sensitivity measures as screening indicators of metabolic-associated fatty liver disease and liver fibrosis. Dig Dis Sci. 2024;69(4):1430–43. pmid:38438774
- 25. Kim AH, Son DH, Lee YJ. Modified triglyceride-glucose index indices are reliable markers for predicting risk of metabolic dysfunction-associated fatty liver disease: a cross-sectional study. Front Endocrinol. 2023;14.
- 26. Kim HS, et al. Triglyceride glucose-waist circumference is superior to the homeostasis model assessment of insulin resistance in identifying nonalcoholic fatty liver disease in healthy subjects. J Clin Med. 2021;11(1).
- 27. Lee J, et al. Performance of noninvasive indices for discrimination of metabolic dysfunction-associated steatotic liver disease in young adults. Gut and liver. 2024.
- 28. Li D, Wu S, Zhang M, Xie C, Tan H, Li N. Correlation between triglyceride-glucose index and related parameters and nonalcoholic fatty liver disease in northwest China. Int J Diabetes Dev Ctries. 2023;44(3):496–502.
- 29. Li H, Shi Z, Chen X, Wang J, Ding J, Geng S, et al. Relationship between six insulin resistance surrogates and nonalcoholic fatty liver disease among older adults: A cross-sectional study. Diabetes Metab Syndr Obes. 2023;16:1685–96. pmid:37309507
- 30. Li N, Tan H, Xie A, Li C, Fu X, Xang W, et al. Value of the triglyceride glucose index combined with body mass index in identifying non-alcoholic fatty liver disease in patients with type 2 diabetes. BMC Endocr Disord. 2022;22(1):101. pmid:35428219
- 31. Li S, Feng L, Ding J, Zhou W, Yuan T, Mao J. Triglyceride glucose-waist circumference: the optimum index to screen nonalcoholic fatty liver disease in non-obese adults. BMC Gastroenterol. 2023;23(1):376. pmid:37919650
- 32. Li Y, Zheng R, Li J, Feng S, Wang L, Huang Z. Association between triglyceride glucose-body mass index and non-alcoholic fatty liver disease in the non-obese Chinese population with normal blood lipid levels: a secondary analysis based on a prospective cohort study. Lipids Health Dis. 2020;19(1):229. pmid:33109219
- 33. Liu Z, He H, Dai Y, Yang L, Liao S, An Z, et al. Comparison of the diagnostic value between triglyceride-glucose index and triglyceride to high-density lipoprotein cholesterol ratio in metabolic-associated fatty liver disease patients: a retrospective cross-sectional study. Lipids Health Dis. 2022;21(1):55. pmid:35752830
- 34. Malek M, Khamseh ME, Chehrehgosha H, Nobarani S, Alaei-Shahmiri F. Triglyceride glucose-waist to height ratio: a novel and effective marker for identifying hepatic steatosis in individuals with type 2 diabetes mellitus. Endocrine. 2021;74(3):538–45. pmid:34355342
- 35. Otsubo N, Fukuda T, Cho G, Ishibashi F, Yamada T, Monzen K. Utility of indices obtained during medical checkups for predicting fatty liver disease in non-obese people. Intern Med. 2023;62(16):2307–19. pmid:36517035
- 36. Peng H, et al. Prediction of MAFLD and NAFLD using different screening indexes: A cross-sectional study in U.S. adults. Front Endocrinol. 2023;14.
- 37. Priego-Parra BA, Reyes-Diaz SA, Ordaz-Alvarez HR, Bernal-Reyes R, Icaza-Chávez ME, Martínez-Vázquez SE, et al. Diagnostic performance of sixteen biomarkers for MASLD: A study in a Mexican cohort. Clin Res Hepatol Gastroenterol. 2024;48(7):102400. pmid:38901566
- 38. Sheng G, Lu S, Xie Q, Peng N, Kuang M, Zou Y. The usefulness of obesity and lipid-related indices to predict the presence of Non-alcoholic fatty liver disease. Lipids Health Dis. 2021;20(1):134. pmid:34629059
- 39. Song S, et al. Triglyceride glucose-waist circumference (TyG-WC) is a reliable marker to predict non-alcoholic fatty liver disease. Biomedicines. 2022;10(9).
- 40. Tian J, Cao Y, Zhang W, Wang A, Yang X, Dong Y, et al. The potential of insulin resistance indices to predict non-alcoholic fatty liver disease in patients with type 2 diabetes. BMC Endocr Disord. 2024;24(1):261. pmid:39633354
- 41. Wang M, Chang M, Shen P, Wei W, Li H, Shen G. Application value of triglyceride-glucose index and triglyceride-glucose body mass index in evaluating the degree of hepatic steatosis in non-alcoholic fatty liver disease. Lipids Health Dis. 2023;22(1):186. pmid:37924128
- 42. Wang R, Dai L, Zhong Y, Xie G. Usefulness of the triglyceride glucose-body mass index in evaluating nonalcoholic fatty liver disease: insights from a general population. Lipids Health Dis. 2021;20(1):77. pmid:34321005
- 43. Wang X, Wang Z, Du W, Ma X, Ma J, Chen Z, et al. Predictive value of TyG and TyG-BMI indices for non-alcoholic fatty liver disease in high-altitude regions of China: A cross-sectional study. J Clin Med. 2024;13(23):7423. pmid:39685881
- 44. Xue Y, Xu J, Li M, Gao Y. Potential screening indicators for early diagnosis of NAFLD/MAFLD and liver fibrosis: Triglyceride glucose index-related parameters. Front Endocrinol (Lausanne). 2022;13:951689. pmid:36120429
- 45. Yang X, Rao H, Yuan Y, Hu N, Zhang X, Zeng Y, et al. Correlation analysis of the triglyceride-glucose index and related parameters in metabolic dysfunction-associated fatty liver disease. Sci Rep. 2025;15(1):23. pmid:39748005
- 46. Yang Z, Yu B, Wang Z, Li Z, Yang B, Zeng H, et al. Comparison of the prognostic value of a comprehensive set of predictors in identifying risk of metabolic-associated fatty liver disease among employed adults. BMC Public Health. 2023;23(1):584. pmid:36991357
- 47. Zeng P, Cai X, Yu X, Gong L. Markers of insulin resistance associated with non-alcoholic fatty liver disease in non-diabetic population. Sci Rep. 2023;13(1):20470. pmid:37993481
- 48. Zhang S, et al. Triglyceride glucose-body mass index is effective in identifying nonalcoholic fatty liver disease in nonobese subjects. Medicine (United States). 2017;96(22).
- 49. Zou H, Ma X, Zhang F, Xie Y. Comparison of the diagnostic performance of twelve noninvasive scores of metabolic dysfunction-associated fatty liver disease. Lipids Health Dis. 2023;22(1):145. pmid:37674196
- 50. 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(4):299–304. pmid:19067533
- 51. Er LK. et al. Triglyceride glucose-body mass index is a simple and clinically useful surrogate marker for insulin resistance in nondiabetic individuals. PLOS ONE. 2016;11(3):e0149731.
- 52. Wang J, et al. The diagnostic and prognostic value of the triglyceride-glucose index in metabolic dysfunction-associated fatty liver disease (MAFLD): A systematic review and meta-analysis. Nutrients. 2022;14(23).
- 53. Huang N, Lu B, Zhu Z-Z, Zhu X-Y, Chen S, Shu Z-Y, et al. The association between triglyceride glucose-body mass index and kidney impairment in patients with type 2 diabetes mellitus. Diabetes Metab Syndr Obes. 2024;17:3447–53. pmid:39309307
- 54. Khanmohammadi S, Kuchay MS. Effects of metabolic dysfunction-associated steatotic liver disease on bone density and fragility fractures: associations and mechanisms. J Obes Metab Syndr. 2024;33(2):108–20. pmid:38740429
- 55. Lonardo A. et al. Sex differences in nonalcoholic fatty liver disease: State of the art and identification of research gaps. Hepatology. 2019;70(4): 1457–69.
- 56. Ling Q, Chen J, Liu X, Xu Y, Ma J, Yu P, et al. The triglyceride and glucose index and risk of nonalcoholic fatty liver disease: A dose-response meta-analysis. Front Endocrinol (Lausanne). 2023;13:1043169. pmid:36743937
- 57. Huang J, Ou W, Wang M, Singh M, Liu Y, Liu S, et al. MAFLD criteria guide the subtyping of patients with fatty liver disease. Risk Manag Healthc Policy. 2021;14:491–501. pmid:33603515
- 58. Chalasani N, Younossi Z, Lavine JE, Charlton M, Cusi K, Rinella M, et al. The diagnosis and management of nonalcoholic fatty liver disease: Practice guidance from the American Association for the study of liver diseases. Hepatology. 2018;67(1):328–57. pmid:28714183