Objective To investigate the association and diagnostic value of the triglyceride-total cholesterol-body weight index (TCBI) with the risk of metabolic-associated fatty liver disease (MAFLD), and compare its performance with the triglyceride-glucose (TyG) index.

Methods A retrospective analysis was conducted on patients admitted to the Department of General Medicine at the Second Affiliated Hospital of Bengbu Medical University from November 2022 to June 2025. The TCBI was transformed using natural logarithm to obtain TCBI-LN, which, along with the TyG index, was divided into four quartile groups (Q1, Q2, Q3, Q4). Patient demographic data, medical history, complete blood count, liver and kidney function, blood glucose and lipid profiles were collected from hospital records. Differences among groups were analyzed using analysis of variance, Kruskal-Wallis H test and χ² test. Multivariate logistic regression models and restricted cubic spline models were used to assess the relationship between TCBI-LN, TyG index and MAFLD risk. The predictive value of TCBI-LN and TyG index for MAFLD was evaluated using receiver operating characteristic curves and DeLong tests.

Results A total of 408 patients were included, among whom 237 were diagnosed with MAFLD, accounting for 58.09%. With the increase of TCBI-LN and TyG indices, the proportion of MAFLD patients and risk of MAFLD occurrence in each group increased, and the differences were statistically significant (P ＜ 0.01). The results of multivariate logistic regression analysis showed that after adjusting for covariates such as hypertension, diabetes, and hyperlipidemia, with the lowest quard group Q1 of TCBI-LN as the reference, the risk of MAFLD in group Q4 was 6.753 times (OR = 6.753, 95%CI: 2.451–18.576, P ＜ 0.01); The risk of TyG index in group Q4 was 3.386 times higher than that in group Q1 (OR = 3.386, 95%CI: 1.342–8.591, P = 0.01). Trend test and restricted cubic spline analysis showed that The TCBI-LN (Ppopulation ＜ 0.01, P nonlinearity = 0.26) and TyG index (Ppopulation = 0.01, P nonlinearity = 0.19) were both linearly and dose-response associated with the risk of MAFLD. Receiver operating characteristic curve analysis showed that the area under the curve of TCBI-LN for predicting MAFLD was 0.728 (95%CI: 0.678–0.778), with a sensitivity of 75.1% and a specificity of 63.2%. The area under the TyG index curve was 0.714 (95%CI: 0.663–0.765), with a sensitivity of 54.4% and a specificity of 80.7%. There was no statistically significant difference in the area under the curve between the two (P = 0.43).

Conclusions Both TCBI and TyG indices are independently associated with the risk of MAFLD, and there is a significant dose-response relationship. As a simple and readily available new indicator, TCBI can be used for the preliminary assessment of the risk of MAFLD in hospitalized populations.