Introduction

The global health burden of fatty liver disease has seen a significant increase. Trend analyses from 1991 to 2019 show that its worldwide prevalence rose from 21.9% to 37.3%, representing an annual increase of 0.7% [1]. It is estimated that non-alcoholic fatty liver disease (NAFLD) affects about a quarter of the global adult population, with a global prevalence that positions it as the most common chronic liver condition [2]. The condition, which encompasses hepatic steatosis in over 5% of hepatocytes in the absence of excessive alcohol use and other chronic liver diseases, is closely linked to metabolic dysfunction [2]. NAFLD demonstrates strong clinical associations with type 2 diabetes, hypertension, obesity, and dyslipidemia [3]. Statistics indicate that NAFLD affects 47.3-63.7% of individuals with type 2 diabetes and over 80% of individuals with obesity [4]. Given its strong association with metabolic dysfunction, this condition has recently been redefined as metabolic dysfunction-associated steatotic liver disease (MASLD). Diagnosis typically depends on imaging or histopathology, where liver biopsy continues to be regarded as the diagnostic gold standard [5]. Nevertheless, the practical application of biopsy is constrained by several drawbacks, such as sampling error, physician interpretation subjectivity, and its invasive character, potentially leading to negative impacts on patients [6]. Additionally, NAFLD/MASLD is a multisystem disease that may evolve into cirrhosis and liver cancer over the long term, while also increasing the risk for extrahepatic complications [3]. As a result, there is a pressing need to create and implement new diagnostic indicators for the early detection and monitoring of NAFLD/MASLD, allowing for effective interventions to prevent further disease advancement.

Despite the identification of numerous biomarkers for early NAFLD/MASLD diagnosis – including the fatty liver index, NashTest, and APRI – their utility in routine practice is limited by complexity and expense [7]. As the hepatic manifestation of metabolic syndrome, NAFLD/MASLD is primarily influenced by elevated triglycerides (TGs) and fasting plasma glucose (FPG), two factors that are important for managing the condition [8]. The triglyceride-glucose (TyG) index, integrating these two parameters, offers a simple and low-cost alternative for evaluating insulin resistance (IR). IR is a state of impaired peripheral glucose uptake and oxidation [9], recognized as a core pathophysiological mechanism underlying fatty liver disease [10]. As the TyG index is a simple and reliable surrogate marker of IR, it is reasonable to hypothesize that the TyG index may be associated with NAFLD/MASLD [3]. This interplay suggests a potential association between the TyG index and NAFLD/MASLD. Moreover, owing to the central pathophysiological role of triglycerides in steatosis and the high prevalence of NAFLD/MASLD [11], the TyG index may be a sensitive indicator for detecting steatosis.

The growing prevalence of fatty liver in China, now at 29.2% in the general population [12], is fueled by sedentary lifestyles and high-calorie diets. This problem is exacerbated in high-stress occupations characterized by extended working hours and sleep deprivation, making healthcare workers a particularly vulnerable group [13]. Despite their professional understanding of the disease and potential for greater preventive efforts, the prevalence of fatty liver among healthcare workers remains notably high at 24% [12], and statistics show a concerning upward trend [14], posing a substantial threat to their health and career longevity. Moreover, healthcare workers are routinely exposed to high occupational stress, which may predispose them to more significant IR – a core mechanism of fatty liver. In this context, the TyG index, as a simple and low-cost surrogate marker of IR, is particularly well-suited for routine self-monitoring among healthcare workers who undergo regular health examinations. This paradox and practical consideration underscore a critical gap: despite the severe health challenges this group faces, the potential association between the TyG index and fatty liver has been scarcely investigated. This study aims to address this gap by leveraging large-scale health examination data from medical staff to construct statistical models. Our objective is to quantify the relationship between the TyG index and the prevalence of fatty liver and to evaluate the effectiveness of the TyG index as a potential predictive biomarker for fatty liver in this population.

Material and methods

Study design and population

This retrospective, cross-sectional analysis was conducted using data from 2,137 healthcare workers who received health check-ups at The Third Xiangya Hospital from January to September 2022. Inclusion criteria: 1) healthcare workers aged above 18 years; 2) complete baseline clinical data, specifically including triglyceride and fasting plasma glucose measurements. Exclusion criteria: 1) unavailable triglyceride data; 2) unavailable fasting plasma glucose data; 3) missing data for covariates including age, body mass index (BMI), smoking, or alcohol consumption. The enrollment process is detailed in Figure 1. This study had been approved by the Ethics Committee of the Third Xiangya Hospital of Central South University (Ethics No. 25065). All participants provided their written informed consent.

Data collection

Demographic and clinical information, including age, sex, ethnicity, BMI, smoking history, alcohol consumption, total cholesterol (TC), high-density lipoprotein cholesterol (HDL-C), self-reported hypertension, fasting plasma glucose (FPG), and triglyceride (TG) levels, was collected through a review of electronic medical records and questionnaires.

TyG index calculation

Triglyceride levels were determined using an enzymatic colorimetric method on a Hitachi 7600 fully automated biochemical analyzer. The TyG index was calculated using the following formula and was evaluated as both a continuous variable and a categorical variable based on quartiles.

TyG index = Ln [TG (mg/dl) × FPG (mg/dl)/2]

Definition of fatty liver

Fatty liver was diagnosed based on abdominal ultrasonography, performed by two specialized sonographers using standardized procedures. No further assessment was conducted to differentiate the etiology of fatty liver, such as excessive alcohol consumption, history of viral hepatitis, or use of hepatotoxic drugs.

Covariates

Detailed information was collected on age, sex, ethnicity, smoking status, alcohol consumption, and medical history. BMI was calculated as weight in kilograms divided by height in meters squared. Participants were categorized into normal weight (< 25 kg/m²) and overweight (≥ 25 kg/m²) groups based on their BMI.

Blood pressure was measured according to a standard protocol [15] using a mercury sphygmomanometer. Participants were excluded from measurement if either arm presented with conditions such as skin rash, bandaging, cast, edema, paralysis, open wounds, muscular atrophy, arteriovenous fistula, or post-mastectomy status. Measurements were routinely taken on the right arm unless it was unavailable or the participant objected. Each participant had 1 to 4 blood pressure readings taken during the study; those with no valid readings were excluded. For participants with a single reading, that value was recorded. For those with multiple readings, the first was discarded, and the average of the subsequent readings was used. Hypertension was defined as current use of antihypertensive medication, a prior clinical diagnosis of hypertension, or an average systolic blood pressure ≥ 140 mmHg and/or diastolic blood pressure ≥ 90 mmHg based on consecutive measurements. Clinical biomarkers, including total cholesterol and high-density lipoprotein cholesterol, were analyzed in the Central Laboratory of the Third Xiangya Hospital.

Statistical analysis

All statistical analyses were performed using R version 4.4.1. Continuous variables are presented as mean ± standard deviation, while categorical variables are expressed as percentages. A multivariable weighted logistic regression model was employed to examine the association between the TyG index and fatty liver.

Model 1: Unadjusted.

Model 2: Adjusted for age, sex, and ethnicity.

Model 3: Adjusted for age, sex, ethnicity, BMI, smoking status, alcohol consumption, TC, HDL-C, and hypertension.

To assess the robustness of our findings, the TyG index was categorized into quartiles to evaluate the odds of fatty liver across different levels. The TyG index was analyzed both as a continuous variable and in quartiles, with the presence of fatty liver among healthcare workers as the dependent variable (yes = 1, no = 0).

A restricted cubic spline (RCS) model was used to explore potential non-linear relationships between the TyG index and the odds of fatty liver. Three knots were placed at specified percentiles based on the distribution of the continuous variable. The independent variable was transformed using the RCS function, the model was fitted with an appropriate link function, and the non-linear relationship was visualized graphically. A p-value for non-linearity was calculated, with p > 0.05 indicating that the non-linear association was not statistically significant.

Subgroup analyses were conducted to examine the relationship between the TyG index and the likelihood of fatty liver across different strata. Data were stratified by sex (male/female), BMI (< 25 kg/m²/≥ 25 kg/m²), hypertension (yes/no), alcohol consumption (yes/no), and smoking status (yes/no) to explore potential effect modification. Interaction p-values were calculated by including product terms of the TyG index and each stratification factor in the logistic regression equation. A p-value < 0.05 was considered statistically significant for interaction.

Results

Baseline characteristics of study participants

The overall prevalence of fatty liver was 23.49%, which increased progressively across ascending TyG index quartiles. Participants in higher TyG quartiles were generally older, more likely to be male, and had higher BMI, total cholesterol, and prevalence of hypertension, smoking, and alcohol consumption, along with lower HDL-C levels. Significant differences across quartiles were observed for all baseline characteristics (all p < 0.05).

Association between the TyG index and fatty liver

The association between the TyG index and the odds of fatty liver is presented in Table 2. Our findings indicate that a higher TyG index is associated with an increased odds of fatty liver. This association was significant in both the unadjusted model (OR = 10.14, 95% CI: 8.02-12.81, p < 0.001) and the minimally adjusted model (OR = 6.40, 95% CI: 5.01-8.19, p < 0.001). After full adjustment, the TyG index remained positively correlated with odds of fatty liver. Specifically, each one-unit increase in the TyG index was associated with a 2.78-fold higher odds of fatty liver. When the TyG index was analyzed as quartiles in the fully adjusted model (Model 3), participants in the highest quartile had 3.35-fold higher odds of fatty liver (OR = 3.35, 95% CI: 1.94-5.79, p < 0.001) compared to those in the lowest quartile.

RCS analysis

In regression analysis, applying a linear regression model directly may lead to inaccurate results if the relationship between the independent and dependent variables is non-linear [16]. Therefore, we used an RCS model to assess whether the observed positive association between the TyG index and the odds of having fatty liver had an underlying non-linear component. As shown in Figure 2, no statistically significant non‑linear association was observed between the TyG index and the odds of fatty liver (p for overall effect < 0.001; p for non-linearity = 0.288).

Subgroup analysis

To determine whether the relationship between the TyG index and fatty liver differed significantly across populations with different characteristics, we conducted subgroup analyses stratified by baseline features (Fig. 3). The results indicated no significant interaction effects between any of the stratification factors and the TyG index on the increased odds of fatty liver (all p for interaction > 0.05). A significant association between the baseline TyG index and an increased odds of fatty liver was observed in the male subgroup (OR = 2.47, 95% CI: 0.99-6.16), as well as among smokers (OR = 2.31, 95% CI: 0.50-10.79), alcohol drinkers (OR = 3.29, 95% CI: 0.88-12.31), and participants with hypertension (OR = 7.23, 95% CI: 0.56-93.27).

Discussion

Within diverse occupational groups, healthcare professionals in China face particularly demanding work environments due to the nature of their profession. They are routinely exposed to heavy clinical workloads, extended working hours, and exceptional levels of responsibility coupled with limited autonomy. These conditions contribute to frequent interpersonal conflicts and associated health issues such as professional burnout, physical fatigue, obesity, and sleep disturbances [17]. Given that fatty liver disease has become a serious public health concern, its prevalence among healthcare workers warrants focused attention.

In this study of 2,137 healthcare professionals, we found that a higher TyG index was associated with an increased odds of fatty liver, demonstrating an approximately linear relationship. The cut‑off point observed in this study was identified at 8.36, beyond which the odds of fatty liver increased significantly. No significant interactions were observed between the baseline TyG index and the stratification variables. Subgroup analysis further indicated that this approximately linear positive association remained consistent across different demographic and clinical subgroups of healthcare workers. In summary, our findings demonstrate that the TyG index may serve as an effective predictive marker for early fatty liver assessment in healthcare professionals.

Previous studies across diverse populations have consistently supported an association between the TyG index and fatty liver disease. In U.S. adults, the TyG index was significantly associated with NAFLD/MAFLD and liver fibrosis [7]. Similar findings were reported in Japanese cohorts, where the TyG index showed strong predictive performance for metabolic-associated fatty liver disease [18], as well as in Chinese populations, where higher TyG levels were associated with lean NAFLD [19] and effective NAFLD prediction models incorporating the TyG index have been developed [20]. Additionally, the TyG index has shown significant correlations with liver fat content in specific patient groups such as those with acromegaly [18]. These previous findings are consistent with our results, providing further evidence for the observed association.

The relationship between the TyG index and odds of fatty liver appears to exhibit a threshold effect [21]. A cross-sectional study of U.S. adults identified an optimal TyG index cut-off value of 8.535 for predicting metabolic-associated fatty liver disease based on ROC curve analysis [22], which aligns closely with the cut-off point of 8.36 identified in our healthcare worker population. Our study further revealed significant associations between the TyG index and higher odds of fatty liver in males, smokers, alcohol drinkers, and individuals with hypertension. This underscores the importance for healthcare professionals, particularly male staff, to adopt healthier lifestyles by reducing smoking and alcohol consumption as key preventive strategies against fatty liver disease, thereby safeguarding both their own health and the quality of care they provide to patients.

The underlying mechanisms linking the TyG index to fatty liver remain incompletely understood but may involve the following factors. As a robust indicator of IR, the TyG index reflects the close and complex relationship between fatty liver and IR [23]. On one hand, the development of fatty liver is primarily driven by excessive fat accumulation within the liver, which impairs normal hepatic function. The liver plays a vital role in glucose metabolism by converting blood glucose into glycogen for storage and releasing it as needed [24]. However, excessive hepatic fat accumulation can impair hepatic glucose metabolism and contribute to dysregulated glucose homeostasis, leading to elevated blood glucose levels. Compensatory increases in insulin secretion may occur to maintain glucose homeostasis, while reduced hepatic responsiveness to insulin contributes to IR [25]. On the other hand, IR not only promotes fat deposition in the liver but also facilitates ectopic lipid accumulation in other tissues, such as within muscle, around blood vessels, in the pericardial region, and in visceral areas [26]. As insulin sensitivity declines, the processes of fat synthesis and breakdown are dysregulated. This alters the distribution and utilization of fats in the body, favoring their accumulation in the liver, muscle, and visceral tissues [23, 27]. These pathophysiological mechanisms provide a logical basis for the relationship between the TyG index and fatty liver disease.

This study has several limitations. First, the participants were recruited exclusively from healthcare workers at the Third Xiangya Hospital, which may limit the generalizability of our findings to the broader population. Second, due to the cross-sectional design, we cannot establish a causal relationship between the TyG index and fatty liver, and residual confounding may persist despite adjustments. Finally, although multivariable weighted logistic regression and RCS models were employed to analyze the association, statistical bias or model error could still exist due to factors such as sample size. In future research, we plan to expand the scope by collaborating with multiple medical institutions to establish a large, diverse health examination database for healthcare professionals. This could facilitate more comprehensive evaluation of health status and investigation of factors associated with fatty liver and other metabolic diseases, with the aim of improving the identification and management of metabolic health risks among healthcare workers.

Conclusions

This study found that the TyG index was positively associated with the prevalence of fatty liver among healthcare professionals. The TyG index may have potential as a simple and accessible marker for identifying individuals with a higher likelihood of fatty liver, although its clinical utility requires further validation. Prospective, multicenter studies are needed to determine whether the TyG index can improve the identification of individuals who subsequently develop fatty liver and to establish its potential role in health screening and management.

Disclosures

This research was supported by grants from Joint Funding Project of Hunan Provincial Natural Science Foundation and Hunan Xiangya Boai Rehabilitation Hospital Co Ltd. (Grant No.2024JJ9116).

This study had been approved by the Ethics Committee of the Third Xiangya Hospital of Central South University (Ethics No. 25065).

The authors declare no conflict of interest.

References