Sex-specific differences in the prognostic value of METS-IR for long-term outcomes in patients with MASLD and advanced liver fibrosis: a nationwide study

Abstract

Objective

Insulin resistance (IR) plays a critical role in shaping long-term outcomes in patients with metabolic dysfunction-associated steatotic liver disease (MASLD). Recent findings suggest that biological sex may influence the onset and progression of MASLD, yet it remains unclear whether sex modifies the link between IR and mortality in those with MASLD and advanced liver fibrosis.

Methods

We analyzed data from 14,081 MASLD patients (7327 men and 6754 women) drawn from the 2001–2018 cycles of the National Health and Nutrition Examination Survey (NHANES). Participants were categorized based on sex-specific deciles of the Metabolic Score for Insulin Resistance (METS-IR). Kaplan–Meier survival analysis and Cox proportional hazards models were used to assess the association between METS-IR and all-cause mortality. Restricted cubic spline (RCS) modeling explored potential non-linear relationships.

Results

Marked sex-related disparities were identified in clinical and metabolic characteristics. Elevated METS-IR significantly predicted increased all-cause mortality in females with MASLD (log-rank p < 0.001), whereas this trend was not evident in males (p = 0.54). Multivariable Cox models showed that higher METS-IR independently correlated with mortality in women with MASLD and advanced fibrosis, but not in their male counterparts.

Conclusion

The prognostic significance of METS-IR differs by sex in MASLD. Elevated METS-IR independently increases long-term mortality risk in females, supporting the need for sex-specific risk evaluation in managing metabolic liver disease.

Supplementary Information

The online version contains supplementary material available at 10.1186/s40001-025-03783-x.

Introduction

In 2023, the term metabolic dysfunction-associated steatotic liver disease (MASLD) was introduced to replace non-alcoholic fatty liver disease (NAFLD), emphasizing the key role of metabolic disturbances in disease onset and progression [1]. MASLD now affects more than 30% of adults worldwide, with incidence rates rising steadily each year [2]. MASLD can progress to fibrosis and cirrhosis [3], especially in patients with combined type 2 diabetes [4]. In addition to liver-related issues, MASLD is strongly linked to a range of extrahepatic conditions, such as chronic kidney disease, cardiovascular complications, and certain malignancies, thereby exerting a substantial impact on both public health and medical resources [5–8].

Growing research highlights pronounced sex-related disparities in liver function and disease evolution. These variations stem from both inherent biological sex and sociocultural gender influences, which shape the initiation, progression, and clinical outcomes of hepatic disorders [9]. As a result, multiple investigations have uncovered sex-specific determinants that play roles in the development and advancement of MASLD [10–16]. Gaining insight into these sex-based distinctions is essential for refining risk assessment, halting disease advancement, and uncovering novel therapeutic strategies tailored by sex.

Insulin resistance (IR) is widely acknowledged as a pivotal driver in the initiation and progression of metabolic dysfunction-associated steatotic liver disease (MASLD) [17]. Persistent IR promotes hepatic fat accumulation by enhancing de novo lipogenesis while simultaneously reducing fatty acid oxidation efficiency [18]. This leads to an overload of triglycerides and free fatty acids in hepatocytes, which in turn worsens insulin resistance through pathways such as inflammation, oxidative damage, endoplasmic reticulum stress, and lipotoxic responses [19].

Recent studies have demonstrated sex-related variations in both the risk patterns and underlying mechanisms of type 1 diabetes, type 2 diabetes, and prediabetic states [20, 21]. Moreover, differences between males and females have been observed in how insulin resistance indices correlate with the likelihood of developing kidney failure [22]. Nevertheless, it is still uncertain whether biological sex influences the link between insulin resistance and long-term prognosis in individuals with MASLD and advanced stages of liver fibrosis.

To fill this knowledge gap, we utilized the Metabolic Score for Insulin Resistance (METS-IR), a recently developed surrogate index of insulin resistance that does not rely on insulin measurements [23]. Using a post hoc analysis of a large, nationally representative longitudinal dataset, our study aimed to explore whether the prognostic utility of METS-IR for predicting all-cause mortality differs between sexes in patients with MASLD and those with advanced liver fibrosis.

Methods

Data source and study population

This investigation was conducted as a secondary analysis based on data from the National Health and Nutrition Examination Survey (NHANES), which is administered by the U.S. National Center for Health Statistics (NCHS). NHANES adopts a sophisticated, multistage probability sampling approach to gather health and nutrition information from a nationally representative, non-institutionalized civilian population. The dataset is publicly available at https://wwwn.cdc.gov/nchs/nhanes/Default.aspx. The survey protocol was reviewed and approved by the NCHS Institutional Review Board, with all participants providing informed consent. Relevant ethics documentation is accessible online.

We analyzed data from the NHANES 2001–2018 survey cycles, which initially included 91,351 individuals. After implementing inclusion and exclusion criteria (illustrated in Fig. 1), a final sample of 14,081 MASLD cases was selected for analysis, consisting of 7327 male and 6754 female participants.

Fig. 1.

Flow diagram outlining the selection process of study participants

Definition of METS-IR

The Metabolic Score for Insulin Resistance (METS-IR) is a validated index that estimates insulin resistance without the need for direct insulin measurement. It is computed using the following formula:

METS-IR = ln {[(2 × fasting blood glucose (mg/dL) + triglyceride level (mg/dL)) × body mass index (kg/m²)] ÷ ln (high-density lipoprotein cholesterol (HDL-C) in mg/dL)}.

Definition of MASLD and advanced liver fibrosis

Due to the lack of imaging data for hepatic steatosis across most NHANES cycles, we employed the Fatty Liver Index (FLI) to identify cases of steatotic liver disease (SLD), following previously validated methodologies [24, 25]. Participants with an FLI score of 60 or above were categorized as having hepatic steatosis.

MASLD diagnosis was based on the 2023 Delphi consensus, which requires both hepatic steatosis (FLI ≥ 60) and the presence of at least one cardiometabolic risk factor [26, 27], Additionally, participants with alternative liver disease etiologies—such as viral hepatitis, autoimmune or genetic liver conditions, drug-induced liver injury, or excessive alcohol intake (≥ 30 g/day in men, ≥ 20 g/day in women)—were excluded [25].

Assessment of advanced liver fibrosis relied on non-invasive biomarkers, namely the NAFLD Fibrosis Score (NFS), Fibrosis-4 Index (FIB-4), and AST-to-Platelet Ratio Index (APRI), as described in earlier studies[PMID: 35334881]. Participants were considered at high risk for advanced fibrosis if they exceeded any of the following thresholds: APRI > 1, FIB-4 > 2.67, or NFS > 0.676 [26, 28, 29].

Outcome definition

The primary outcome of interest was all-cause mortality among individuals diagnosed with MASLD and those categorized as having a high risk of advanced liver fibrosis. Mortality information was obtained from the NHANES Linked Mortality File, with updates extending through December 31, 2019. This dataset was matched to the National Death Index (NDI) via a probabilistic linkage process. Causes of death were classified according to the 10th Revision of the International Classification of Diseases (ICD-10). The duration of follow-up was determined from the date of the participant’s initial NHANES interview until either the date of death or December 31, 2019, depending on which occurred first.

Statistical analysis

All statistical analyses were performed using R version 4.2.0 and EmpowerStats software (www.empowerstats.com). A two-sided p value < 0.05 was considered statistically significant. In accordance with NHANES analytical guidelines, all analyses incorporated appropriate sampling weights, clustering, and stratification to account for the complex survey design. For combined analysis across the nine NHANES cycles (2001–2018), adjusted sampling weights were calculated by dividing the 2-year cycle weights by nine [23], and survey design parameters were applied using the survey package in R.

Continuous variables were described as weighted means with 95% confidence intervals (CIs) and compared using weighted linear regression. Categorical variables were summarized as weighted proportions with 95% CIs and analyzed using weighted chi-square tests.

To evaluate the association between METS-IR and all-cause mortality in patients with MASLD and advanced liver fibrosis, multivariable Cox proportional hazards regression models were employed. Four models were constructed:

Model 1 (crude): No covariates adjusted.

Model 2: Adjusted for age, race/ethnicity, PIR, and education level.

Model 3: Adjusted for all covariates listed in Table 1.

Table 1.

Cox regression analysis of METS-IR with all-cause mortality in patients of MASLD and advanced liver fibrosis by gender

Gender	Characteristic	Model 1 HR (95%CI)	Model 2 HR (95%CI)	Model 3 HR (95%CI)	Model 4 HR (95%CI)	P for interaction
						0.0059a
Male	MASLD	0.78 (0.61, 0.99)	1.72 (1.33, 2.22)	1.20 (0.92, 1.56)	1.16 (0.89, 1.51)
Female	MASLD	1.72 (1.27, 2.33)	2.79 (2.07, 3.77)	1.77 (1.28, 2.47)	1.80 (1.30, 2.49)
						0.0032b
Male	Liver fibrosis	0.56 (0.36, 0.86)	0.88 (0.55, 1.41)	0.64 (0.39, 1.06)	0.65 (0.39, 1.07)
Female	Liver fibrosis	1.53 (0.91, 2.57)	3.08 (1.79, 5.32)	2.53 (1.41, 4.55)	2.51 (1.42, 4.56)

Model 1 was adjusted for no covariates;

Model 2 was adjusted for age, race and education;

Model 3 was adjusted for age, race, education, smoking status, activity, TC, asthma, CVD, cancers, UA, diabetes, BP, PIR, total suger, total fat and total suger

Model 4 was adjusted for age, race, education, smoking status, activity, TC, TBIL, ALB, AST, ALT, GGT, asthma, CVD, cancers, UA, diabetes, BP, PIR, total suger, total fat and total suger

^arepresents MASLD patient p for interaction

^brepresents liver fibrosis patient p for interaction

Additionally, we constructed a sensitivity model (Model 4) further adjusting for serum total bilirubin (TBIL), albumin (ALB), aspartate aminotransferase (AST), alanine aminotransferase (ALT), and γ-glutamyl transferase (GGT), to account for liver function and endogenous antioxidant capacity.

Kaplan–Meier survival curves were constructed to visualize differences in survival across METS-IR levels, and log-rank tests were used to compare survival distributions.

To identify potential threshold effects of METS-IR on all-cause mortality, restricted cubic spline (RCS) regression was performed within the multivariable Cox proportional hazards framework. Knots were placed at predefined percentiles of the METS-IR distribution. When a significant non-linear association was detected, a two-piecewise Cox regression model was further applied. The optimal cutoff value (inflection point) was determined using a maximum likelihood–based approach, in which the likelihood of models with different candidate inflection points was compared, and the point yielding the highest log-likelihood was selected. This procedure was conducted separately for each sex and disease subgroup, resulting in four subgroup-specific cutoff values used in Fig. 3. In addition, a subgroup analysis was conducted to assess the potential effect modification by relevant clinical confounders. For age stratification, 55 years was selected as a threshold to approximate menopausal transition in women, as the majority of females undergo menopause around this age. This cutoff has been commonly applied in epidemiological studies examining sex-specific metabolic and cardiovascular risks [30–32]. Interaction analyses were conducted by including cross-product terms between sex and METS-IR in the multivariate Cox regression models to formally evaluate potential sex-specific differences in the associations.

Fig. 3.

Restricted cubic spline plots displaying the association between METS-IR and mortality: A MASLD in females; B Advanced liver fibrosis in females; C MASLD in males; D Advanced liver fibrosis in males. The red curves indicate the reference hazard ratio (HR), while shaded blue regions represent the 95% confidence intervals

Results

Baseline characteristics

The final study population included 14,081 individuals diagnosed with MASLD, consisting of 7327 men and 6754 women. Of these, 969 males and 846 females met the criteria for advanced liver fibrosis. Throughout the follow-up period, all-cause mortality was recorded in 1294 men and 842 women.

Baseline characteristics, including demographic, clinical, and laboratory parameters, stratified by sex, are detailed in Supplementary Tables 1 and 2. Age distribution, metabolic profiles, and behavioral factors showed notable variation across METS-IR deciles for both sexes. Among female MASLD patients, higher METS-IR levels were significantly associated with increased all-cause mortality (p < 0.01), whereas this association was not statistically significant in males (p = 0.326). Furthermore, smoking prevalence in women rose consistently with increasing METS-IR deciles (p < 0.01), a pattern not observed in men (p = 0.79). In both sexes, elevated METS-IR levels were linked to higher serum uric acid and increased rates of hypertension, diabetes, asthma, and advanced liver fibrosis.

Prognostic role of METS-IR by sex

The Kaplan–Meier survival analysis indicated a strong association between higher METS-IR levels and elevated all-cause mortality among women with MASLD (Log-rank p < 0.001; Fig. 2A). Conversely, no statistically significant difference in survival was observed in men across METS-IR levels (Log-rank p = 0.54; Fig. 2B). In patients with advanced liver fibrosis, METS-IR appeared to have limited ability to discriminate mortality risk in either sex (Figs. 2C–D).

Fig. 2.

Kaplan–Meier survival analysis of all-cause mortality by METS-IR levels: A Female MASLD patients; B Male MASLD patients; C Female patients with advanced liver fibrosis; D Male patients with advanced liver fibrosis

In fully adjusted multivariate Cox regression models (Model 3), increased METS-IR was independently linked to higher all-cause mortality risk in females with MASLD (HR: 1.77; 95% CI 1.28–2.47). However, this association was not statistically significant in male participants (HR: 1.20; 95% CI 0.92–1.56). A comparable sex-related trend was seen in those with advanced liver fibrosis: elevated METS-IR was significantly predictive of mortality in females (HR: 2.53; 95% CI 1.41–4.55), but not in males (HR: 0.64; 95% CI 0.39–1.06) (Table 1). In sensitivity analyses incorporating bilirubin, albumin, AST, ALT, and GGT (Model 4), the results were consistent with those of Model 3. Elevated METS-IR remained significantly associated with increased mortality risk in women with MASLD (HR: 1.80, 95% CI 1.30–2.49) and in those with advanced fibrosis (HR: 2.51, 95% CI 1.42–4.56), but no significant association was observed in men (Table 3). Importantly, formal tests for interaction confirmed significant sex differences in the prognostic effect of METS-IR on mortality. The p for interaction was 0.0059 in patients with MASLD and 0.0032 in those with advanced liver fibrosis, supporting the robustness and statistical significance of the observed sex-specific associations.

Table 3.

Subgroup analysis of METS-IR with all-cause mortality in patients of MASLD and advanced liver fibrosis in men

Characteristic		Model 1 HR (95%CI)	Model 2 HR (95%CI)	Model 3 HR (95%CI)
MASLD	Age < 55	2.19 (1.16, 4.13)	2.19 (1.15, 4.15)	1.22 (0.62, 2.42)
	Age ≥ 55	1.19 (0.92, 1.55)	1.06 (0.81, 1.39)	0.70 (0.53, 0.93)
	Mexican American	1.02 (0.55, 1.89)	1.87 (0.96, 3.64)	1.28 (0.64, 2.56)
	White	0.71 (0.53, 0.95)	1.74 (1.26, 2.38)	1.15 (0.82, 1.61)
	Black	0.88 (0.47, 1.67)	1.58 (0.84, 2.98)	1.11 (0.57, 2.16)
	Other Race	2.23 (0.56, 8.97)	3.24 (0.79, 13.38)	2.08 (0.41, 10.61)
Liver fibrosis	Age < 55	0.68 (0.10, 4.85)	0.34 (0.03, 3.58)	0.00 (0.00, inf.)
	Age ≥ 55	0.62 (0.39, 0.97)	0.55 (0.35, 0.88)	0.44 (0.26, 0.72)
	Mexican American	0.59 (0.19, 1.82)	1.15 (0.34, 3.81)	1.18 (0.26, 5.49)
	White	0.39 (0.22, 0.68)	0.71 (0.39, 1.29)	0.44 (0.23, 0.84)
	Black	1.08 (0.38, 3.06)	1.68 (0.54, 5.18)	0.96 (0.27, 3.38)
	Other race	0.46 (0.03, 6.03)	0.97 (0.05, 17.78)	inf. (0.00, Inf)

Model 1 was adjusted for no covariates;

Model 2 was adjusted for age, race and education;

Model 3 was adjusted for age, race, education, smoking status, activity, TC, asthma, CVD, cancers, UA, diabetes, BP, PIR, total suger, total fat and total suger

Restricted cubic spline and threshold effect analysis

In women with MASLD, restricted cubic spline (RCS) modeling revealed a non-linear association between METS-IR and all-cause mortality (p for non-linearity = 0.038). Mortality risk escalated notably once METS-IR surpassed the threshold of 2.308 (Fig. 3A). For females within the advanced liver fibrosis subgroup, a linear positive trend between METS-IR and mortality was detected (p for non-linearity = 0.146), suggesting a gradual, dose-dependent increase in risk (Fig. 3B).

In contrast, no statistically significant linear or non-linear relationships were found between METS-IR and mortality among male patients with MASLD or advanced liver fibrosis (all p-values > 0.05).

Subgroup analysis

Subgroup analyses demonstrated marked sex-based heterogeneity in the relationship between METS-IR and all-cause mortality among patients with MASLD and advanced liver fibrosis. In female MASLD patients, elevated METS-IR was significantly linked to higher mortality risk in those aged 55 years or older (HR = 1.29; 95% CI 1.01–1.84), as well as in White (HR = 1.77; 95% CI 1.16–2.71) and African American populations (HR = 2.37; 95% CI 1.04–5.37). A similar trend was observed among females with advanced fibrosis, where higher METS-IR levels were associated with increased mortality in those aged ≥ 55 (HR = 1.98; 95% CI: 1.07–3.65), Whites (HR = 1.77; 95% CI 1.16–2.71), and African Americans (HR = 2.37; 95% CI 1.04–5.37) (Table 2).

Table 2.

Subgroup analysis of METS-IR with all-cause mortality in patients of MASLD and advanced liver fibrosis in women

Characteristic		Model 1 HR(95%CI)	Model 2 HR(95%CI)	Model 3 HR(95%CI)
MASLD	Age < 55	2.46 (1.04, 5.80)	2.68 (1.10, 6.49)	0.76 (0.28, 2.10)
	Age ≥ 55	2.18 (1.58, 3.01)	2.15 (1.55, 2.98)	1.29 (1.01, 1.84)
	Mexican American	1.32 (0.63, 2.77)	2.05 (0.96, 4.39)	1.34 (0.60, 2.98)
	White	1.68 (1.14, 2.49)	2.58 (1.76, 3.77)	1.77 (1.16, 2.71)
	Black	2.25 (1.09, 4.67)	5.05 (2.45, 10.41)	2.37 (1.04, 5.37)
	Other race	1.76 (0.30, 10.12)	2.86 (0.45, 18.29)	2.33 (0.19, 28.46)
Liver fibrosis	Age < 55	0.70 (0.09, 5.23)	0.81 (0.10, 6.65)	0.45 (0.02, 8.87)
	Age ≥ 55	2.06 (1.19, 3.57)	2.24 (1.27, 3.95)	1.98 (1.07, 3.65)
	Mexican American	0.48 (0.13, 1.71)	1.11 (0.30, 4.18)	1.07 (0.20, 5.58)
	White	1.65 (0.79, 3.44)	3.23 (1.47, 7.08)	2.77 (1.18, 6.47)
	Black	3.20 (1.08, 9.52)	7.08 (2.35, 21.36)	7.53 (1.89, 30.09)
	Other race	3.43 (0.49, 23.79)	11.70 (0.43, 320.44)	inf. (0.00, Inf)

Model 1 was adjusted for no covariates;

Model 2 was adjusted for age, race and education;

Model 3 was adjusted for age, race, education, smoking status, activity, TC, asthma, CVD, cancers, UA, diabetes, BP, PIR, total suger, total fat and total suger

Conversely, in male MASLD patients, higher METS-IR was unexpectedly associated with lower mortality in individuals aged ≥ 55 (HR = 0.70; 95% CI 0.53–0.93). Among males with advanced liver fibrosis, elevated METS-IR was also significantly linked to decreased all-cause mortality in both the ≥ 55 age group (HR = 0.44; 95% CI 0.26–0.72) and in White participants (HR = 0.44; 95% CI 0.23–0.84) (Table 3).

Discussion

Using data from a nationally representative cohort of U.S. adults, this study is the first to explore the sex-specific association between the Metabolic Score for Insulin Resistance (METS-IR) and long-term all-cause mortality in individuals diagnosed with metabolic dysfunction-associated steatotic liver disease (MASLD) and advanced fibrosis. We observed that elevated METS-IR was significantly linked to increased mortality in women but not in men, implying that insulin resistance may exert sex-dependent prognostic effects in these conditions. The sex-specific divergence was statistically supported, as evidenced by significant p values for the interaction between sex and METS-IR (p for interaction = 0.0059 for MASLD; p = 0.0032 for advanced fibrosis). These results reinforce the notion that insulin resistance exerts a stronger adverse prognostic impact in females than in males, highlighting the importance of considering sex-specific differences in clinical risk assessment and management strategies.To the best of our knowledge, this is the most comprehensive analysis to date investigating sex-related disparities in this context using a large, longitudinal dataset.

The pathogenesis and progression of MASLD are influenced by a multifaceted combination of genetic factors, nutritional patterns, physical activity, and alterations in gut microbiota composition [33, 34]. Insulin resistance serves as a critical link connecting these variables to the accumulation of hepatic fat and subsequent disease progression [35, 36]. Previous studies have indicated that elevated insulin resistance markers, such as HOMA-IR, are associated with a higher likelihood of developing non-alcoholic steatohepatitis (NASH) and advancing to more severe fibrosis [37]. METS-IR, as an insulin-independent surrogate index, has likewise demonstrated notable diagnostic performance for NAFLD, with an area under the receiver operating characteristic curve (AUROC) of 0.824 [38].

In recent years, the prognostic implications of various insulin resistance–related indices have been the focus of growing research. Analyses based on NHANES data have shown that the triglyceride-glucose (TyG) index and its related parameters—such as TyG-WHtR, TyG-BMI, and TyG-WC—are useful predictors of both cardiovascular and overall mortality in individuals with MASLD [25, 39]. Findings from the Dalian Health Management Cohort indicated that TyG-BMI showed the strongest correlation with MASLD presence and progression to hepatic fibrosis [40]. Additional studies have corroborated these associations, linking TyG-based metrics with histologically confirmed liver fibrosis [41, 42]. Nonetheless, some contradictory evidence exists; for instance, data from a Korean population indicated that lower METS-IR scores were unexpectedly linked to increased risks of hepatocellular carcinoma and liver decompensation [43]. These inconsistencies highlight the importance of conducting analyses tailored to specific populations and ethnic backgrounds.

This study contributes novel evidence by demonstrating a significant and independent link between high METS-IR levels and increased long-term mortality risk among female patients with MASLD and advanced fibrosis. Several biological and behavioral mechanisms may underlie this sex-specific relationship:

Hormonal and postmenopausal factors

A substantial proportion of female participants were postmenopausal (with nearly 68% aged over 50), a demographic marked by estrogen deficiency. Consistent with this, we stratified our analyses at 55 years of age to approximate menopausal transition, recognizing that estrogen decline after menopause is closely linked to increased visceral adiposity, dyslipidemia, and heightened insulin resistance, thereby amplifying the prognostic role of METS-IR in women. Estrogen plays a protective role in maintaining insulin sensitivity, and its reduction post-menopause leads to visceral fat gain, lipid profile disturbances, and heightened insulin resistance [31, 32, 44]. Visceral adiposity is more metabolically active and more strongly linked to IR than subcutaneous fat [45].

Differences in body composition and activity levels

Compared to men, women typically exhibit greater total fat mass and engage in less physical activity [46], both of which impair insulin signaling. Exercise enhances insulin action via mechanisms such as improved capillary perfusion and GLUT4 translocation [47], suggesting that lower activity levels in women may amplify the adverse effects of IR.

Sex-specific inflammatory responses

Emerging evidence indicates that higher METS-IR is associated with elevated levels of inflammatory cytokines—including EGF, Eotaxin, and MCP-1—in women but not in men [48]. These cytokines contribute to systemic inflammation and IR [49], which may partially explain the stronger association between METS-IR and mortality in females.

Notably, while the prevalence of MASLD, metabolic-associated steatohepatitis (MASH), and advanced fibrosis tends to be higher in males [50], our results indicate that the adverse prognostic influence of insulin resistance—measured via METS-IR—appears more pronounced in females. These findings emphasize the necessity of incorporating sex as a key biological determinant in the risk assessment and clinical management of MASLD.

Importantly, in sensitivity analyses that included bilirubin, albumin, AST, ALT, and GGT as additional covariates, the association between METS-IR and mortality in females remained robust. Bilirubin and albumin are well-recognized markers of liver function and endogenous antioxidant capacity. Recent studies have reported sex-specific associations of bilirubin with cancer risk and interactions with lifestyle factors, suggesting potential biological mechanisms underlying sex-related heterogeneity [51–53]. These findings further support that the prognostic relevance of METS-IR in women is independent of traditional liver function parameters, thereby strengthening the robustness and clinical interpretability of our results.

This study has several strengths, including a large, nationally representative sample, a long follow-up period, and comprehensive adjustment for confounding variables. The use of METS-IR, a validated and easily accessible IR surrogate, adds practical value for clinical screening and risk assessment.

However, several limitations must be acknowledged. First, due to the observational design, causal inferences cannot be made. Second, although we adjusted for many potential confounders, residual confounding from unmeasured variables (e.g., medication use, unrecorded comorbidities) cannot be excluded. Third, METS-IR was assessed at a single time point, limiting our ability to evaluate longitudinal changes or treatment effects. Fourth, our findings are based on a U.S. adult population and may not be generalizable to other ethnic or age groups. Further validation in diverse cohorts is warranted.Finally, both the Fatty Liver Index (FLI) and the Metabolic Score for Insulin Resistance (METS-IR) are surrogate, non-invasive indices rather than definitive diagnostic tools. FLI does not directly confirm hepatic steatosis by imaging or histology, and METS-IR reflects insulin resistance indirectly rather than through direct insulin-based measurements. Therefore, misclassification cannot be completely exclude.

Conclusion

In conclusion, this study highlights a clear sex-specific relationship between METS-IR and all-cause mortality among individuals with MASLD and advanced liver fibrosis. Elevated METS-IR was independently associated with increased mortality risk in women but showed no significant prognostic relevance in men. These findings underscore the importance of integrating sex-specific considerations into MASLD risk stratification and clinical decision-making. Future research, particularly prospective and mechanistic investigations, is warranted to uncover the underlying biological pathways and to confirm the prognostic value of METS-IR across varied populations.

Author contributions

Binbin Song (BS): Writing-original draft, Conceptualization, Data curation and Investigation. Yuan Zhou (YZ): Writing-original draft, Methodology and Software Rui Su (RS): Writing-original draft, Visualization. Yan Wang (YW): Writing-original draft,Formal Analysis. ·Song Chen (SC): Writing-review & editing and editing, Methodology. Wen jin Chen (WC): Writing-review & editing, Project administration.

Funding

Not applicable.

Data availability

The datasets analyzed during the current study are publicly available from the National Health and Nutrition Examination Survey (NHANES) at: https://wwwn.cdc.gov/nchs/nhanes/Default.aspx.

Declarations

Ethics approval and consent to participate

Ethical approval for this study was granted by the Research Ethics Review Board of the National Center for Health Statistics (NCHS). All participants provided informed consent prior to participation.

Competing interests

The authors declare no competing interests.