Fatty liver index and risk of type 2 diabetes of adults with normoglycemia: Insights into insulin sensitivity and beta-cell function

Ji Hyun Bae; Min Jin Lee; Su Hyun Kim; Joo Yeon Kim; Ah Reum Khang; Yang Ho Kang; Dongwon Yi

doi:10.1371/journal.pone.0327058

Abstract

The fatty liver index (FLI) is a simple tool used to assess metabolic dysfunction-associated fatty liver disease (MAFLD). Previous studies have shown the utility of the FLI as an early predictor of diabetes for patients with prediabetes. We evaluated whether the FLI could predict the development of type 2 diabetes mellitus (T2DM) with normal glucose tolerance (NGT) by performing a retrospective assessment of a community-based cohort over the course of 18 years. We analyzed data of 6,083 adults with NGT available from the Korean Genome and Epidemiology Study database. Participants were stratified into the following three groups based on the FLI: low, FLI < 30; intermediate, FLI 30–59; and high, FLI ≥ 60. Cox proportional hazards regression models evaluated the T2DM risk differences. Insulin sensitivity and secretion markers were compared using multivariate linear regression and an analysis of covariance. The predictability of the FLI for T2DM was analyzed by comparing the area under the receiver-operator characteristic (ROC) curve (AUC) values from the ROC analysis. The cumulative incidence of T2DM was 31.9% for the high FLI group; however, it was 11.3% for the low FLI group (log-rank test, P < 0.0001). For individuals with NGT, a high FLI was associated with an increased T2DM risk (hazard ratio [HR], 3.42; 95% confidence interval [CI], 2.91–4.00). After adjusting for insulin sensitivity and secretion markers, FLI remained an independent predictor of T2DM (HR, 1.96, 95% CI, 1.54–2.50). The homeostasis model assessment of insulin resistance results and composite insulin sensitivity index of the high FLI group were higher than those of the other groups (P < 0.0001). However, the disposition index and insulin secretion-sensitivity index-2 of the high FLI group were lower than those of the intermediate FLI group (P = 0.027 and P = 0.011, respectively). The ROC analysis confirmed that the FLI had the highest predictive ability for T2DM (AUC, 0.654; P < 0.05) development in individuals with NGT compared to other insulin sensitivity and secretion markers. The FLI is an early predictor of T2DM that reflects underlying insulin sensitivity and β-cell function. These findings underscore the role of liver steatosis in the early T2DM pathogenesis and highlight the need for early preventive lifestyle interventions among individuals with normoglycemia and high FLI values.

Citation: Bae JH, Lee MJ, Kim SH, Kim JY, Khang AR, Kang YH, et al. (2025) Fatty liver index and risk of type 2 diabetes of adults with normoglycemia: Insights into insulin sensitivity and beta-cell function. PLoS One 20(6): e0327058. https://doi.org/10.1371/journal.pone.0327058

Editor: Guoying Wang, Johns Hopkins University Bloomberg School of Public Health, UNITED STATES OF AMERICA

Received: March 14, 2025; Accepted: June 8, 2025; Published: June 26, 2025

Copyright: © 2025 Bae 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: The data utilized in this study were provided by the Korean Genome and Epidemiology Study (KoGES) through the Cohort Data Access system (CODA), managed by the Korea Disease Control and Prevention Agency (KDCA). Because of ethical and legal restrictions imposed by the KDCA, public sharing of the dataset is not permitted, even in de-identified form. Although the data are anonymized, they contain detailed health and demographic information, which may carry a risk of re-identification when combined with other datasets. In accordance with institutional policy, access to the dataset is controlled and managed exclusively by KDCA. We are pleased to provide the official data access platform and the specific database ID corresponding to the dataset we were granted access to reference as follows: Data Access Platform: https://coda.nih.go.kr Database ID: CODA_D24032 (the database from which our study data were obtained) Contact Email: coda@korea.kr Phone: +82-43-249-3020 All data requests should be directed to the CODA administrative office. The authors do not hold any special privileges regarding data access.

Funding: This study was supported by a 2025 research grant from Pusan National University Yangsan Hospital.

Competing interests: The authors have declared that no competing interests exist.

Introduction

Type 2 diabetes mellitus (T2DM) is a leading cause of death associated with cardiovascular disease and a chronic progressive disease accompanied by various complications. Therefore, it creates a burden on individuals and society [1–3]. In 2021, the 10th edition of the International Diabetes Federation Diabetes Atlas reported that 537 million adults 20–79 years of age worldwide have diabetes, and this number is projected to increase to 783 million by 2045 [4]. According to the diabetic fact sheets in Korea, the prevalence of diabetes among individuals 30 years or older increased from 4.8 million (13.7%) in 2016 to 5.3 million (15.5%) in 2024 [5]. Hepatic regulation is crucial to glucose metabolism. Hepatic steatosis leads to hepatic insulin resistance, inappropriate endogenous glucose production, and abnormal lipid metabolism, resulting in T2DM [6,7]. The bidirectional association between nonalcoholic fatty liver disease (NAFLD) and T2DM is well-established, although its pathophysiology remains unclear. According to a large-scale meta-analysis, patients with NAFLD had a 2.22-times higher risk of developing T2DM compared to that of those without NAFLD. Furthermore, this risk can increase up to 4.74 times for individuals with severe NAFLD [8]. Recently, a shift from classifying liver steatosis as NAFLD based on its developmental pathway to classifying it as metabolic dysfunction-associated fatty liver disease (MAFLD) and steatotic liver disease has occurred. This new classification system focuses on the presence of liver steatosis accompanied by metabolic abnormalities, regardless of the underlying cause [9]. The MAFLD criteria include T2DM, obesity, and metabolic risk abnormalities (including the insulin resistance score), thus emphasizing a strong association with metabolic disorders [9]. Liver biopsy is the most reliable method of diagnosing steatotic liver disease. However, liver biopsy is not suitable for screening patients with mild fatty liver and is associated with a high cost and the risk of complications. Although abdominal ultrasonography is relatively accurate and useful, it is not suitable for repeated use because of its high cost. The fatty liver index (FLI) is as effective as ultrasonography for predicting fatty liver disease [10,11]. The FLI, which is based on a robust indicator of MAFLD pathophysiology, is a simple tool that can be used to assess MAFLD [12,13] and metabolic syndromes [14,15]. Previous studies have demonstrated the utility of the FLI as an early predictor of T2DM development in patients with prediabetes [16,17]. However, limited assessments of its effectiveness for individuals with normal glucose tolerance (NGT) have been conducted. Therefore, we aimed to assess whether the FLI could effectively reflect metabolic dysfunction and serve as an early predictor of T2DM, even for individuals with NGT, by performing a longitudinal assessment of a community-based cohort over the course of an 18-year period. Additionally, we compared the differences in insulin sensitivity and insulin secretion markers according to the FLI and investigated their correlation with the FLI. We also evaluated the predictive ability of the FLI relative to these markers for individuals with or without NGT.

Materials and methods

Study population

During this study, we analyzed data from the Ansan and Ansung cohort, which is a subset of the Korean Genome and Epidemiology Study (KoGES), which is a longitudinal, prospective cohort study conducted by the Korean Centers for Disease Control and Prevention (KCDC), to investigate the genetic and environmental determinants of chronic diseases. The KoGES is a large-scale prospective study of 10,030 Korean individuals 40–69 years of age that was initiated in 2001; participants were recruited based on age, sex, and residential location from the urban area of Ansan using a random sampling method and from the rural area of Ansung using a cluster sampling method. Baseline data collection was conducted from 2001 to 2002. Thereafter, continuous biennial follow-up assessments were performed. Participants underwent standardized assessments every 2 years that included anthropometric measurements, laboratory tests, and structured questionnaires [18]. A 75-g oral glucose tolerance test was also conducted biennially to reassess the glycemic status until a clinical diagnosis of diabetes was determined. We obtained and retrospectively analyzed the baseline survey data (2001–2002) and follow-up data (2019–2020) from the National Institute of Health of Korea and Korea Disease Control and Prevention Agency on November 23, 2023. The KoGES provides only de-identified data in compliance with the Personal Information Protection Act and the Statistics Act, thus ensuring that individuals cannot be identified from the survey data. Therefore, we only had access to anonymized data that did not contain any personally identifiable information during or after data collection.

From the 10,030 participants at baseline, we excluded individuals with preexisting diabetes, those undergoing treatment for malignancy or viral hepatitis (hepatitis B or C), those who were treated with corticosteroids in the past 3 months, and those with missing essential data regarding fasting plasma glucose (FPG) and the oral glucose tolerance test (OGTT). Additionally, participants who were newly diagnosed with T2DM at baseline, those with impaired fasting glucose (IFG) and/or impaired glucose tolerance (IGT), and those with other missing data were excluded. After applying these criteria, 6,083 participants were included in the final analysis (Fig 1). Ethical approval of the study protocol was obtained from the Korean Centers for Disease Control and Prevention Ethics Committee, and written informed consent was obtained from all participants. This study adhered to the principles of the Declaration of Helsinki and was approved by the Institutional Review Board of Pusan National University Yangsan Hospital (IRB no. 55-2023-028).

Download:

Fig 1. Flowchart of the study population.

Abbreviations: FPG, fasting plasma glucose; IFG, impaired fasting glucose; IGT, impaired glucose tolerance; OGTT, oral glucose tolerance test; T2DM, type 2 diabetes mellitus.

https://doi.org/10.1371/journal.pone.0327058.g001

Assessment of anthropometric and biochemical parameters

The participants visited designated community clinics where trained healthcare professionals conducted anthropometric measurements and biochemical tests. The body mass index (BMI) was calculated by dividing the weight (kg) by the square of the height (m²). Waist circumference (WC) was measured midway between the lowest rib margin and the iliac crest during exhalation. Blood pressure (BP) was measured three times on the right upper arm using a mercury sphygmomanometer, and the average of the second and third readings was recorded. Smoking history was categorized as never smoker (fewer than 100 cigarettes during the lifetime of the individuals), ex-smoker (at least 100 cigarettes during the lifetime of the individual but not currently smoking), or current smoker (at least 100 cigarettes during the lifetime of the individual and currently smoking). Alcohol consumption data were gathered during the interviews and quantified as grams per week.

Hypertension was defined as systolic BP (SBP) ≥140 mmHg, diastolic BP (DBP) ≥90 mmHg, or the use of antihypertensive medications. A family history of diabetes was noted if first-degree relatives were involved. Biochemical testing required at least 8 h of fasting and included FPG, insulin, total cholesterol (TC), high-density lipoprotein cholesterol (HDL-C), and triglyceride (TG) levels. This cohort did not include direct measurements of low-density lipoprotein cholesterol (LDL-C). Therefore, LDL-C levels were calculated using the Friedewald formula [19]: LDL-C = TC − (TG/5) − HDL-C, excluding cases with TG > 400 mg/dL (112 individuals). Plasma samples for glucose and insulin were collected at 0, 60, and 120 min during a 2-h 75-g OGTT. Plasma glucose levels were measured using the hexokinase method, whereas insulin levels were determined using a radioimmunoassay.

Definition and outcomes

NGT was defined as an FPG level <100 mg/dL and 2-h plasma glucose level <140 mg/dL during the OGTT; this definition excluded individuals diagnosed with T2DM. T2DM was diagnosed based on any of the following criteria: FPG ≥ 126 mg/dL; 2-h plasma glucose ≥200 mg/dL during the OGTT; or the use of antidiabetic medications, including insulin. HbA1c (glycated hemoglobin) levels ≥ 6.5% were not included in the diagnostic criteria because the study was conducted before the American Diabetes Association introduced HbA1c as a standard diagnostic tool in 2010. The primary outcome of this study was the incidence of T2DM during the 18-year follow-up period.

For individuals with NGT, we calculated the FLI based on initial data from 2001. The variables included in the FLI calculation were the TG level, BMI, gamma-glutamyl transferase (GGT) level, and WC. The FLI ranged between 0 and 100.

The FLI values were categorized as follows and as previously described [20]: FLI < 30, low; FLI 30–59, intermediate; and FLI ≥ 60, high. A low FLI indicated a very low likelihood of fatty liver (negative likelihood ratio = 0.2), whereas a high FLI was strongly associated with a high probability of fatty liver (positive likelihood ratio = 4.3) [20]. Insulin resistance (homeostasis model assessment [HOMA] of insulin resistance [HOMA-IR]) was calculated using the following formula: HOMA-IR = fasting plasma insulin (FPI) (mmol/L) × FPG (μU/mL)/ 22.5. Pancreatic β-cell function (HOMA of pancreatic β-cell function) was calculated using the following formula: HOMA of pancreatic β-cell function = [FPI (μU/mL) × 20]/ [FPG (mmol/L) − 3.5] [21]. Markers derived from the OGTT to evaluate insulin sensitivity and secretion were calculated as follows: composite or Matsuda insulin sensitivity index (ISI): ISI = 10,000/ √(FPG × FPI × mean plasma glucose × mean plasma insulin) [22]; 60-min insulinogenic index (IGI₆₀): IGI₆₀ = (plasma insulin at 60 min – FPI)/ (plasma glucose at 60 min − FPG) [23]; and oral disposition index₆₀: disposition index = IGI₆₀ × composite ISI [24]. The area under the receiver-operator characteristic (ROC) curve (AUC)_{insulin/glucose} was calculated as (AUC_insulin)/ (AUC_glucose) using the trapezoidal rule during the 75-g OGTT [25]. The insulin secretion-sensitivity index-2 (ISSI-2) was calculated as follows: ISSI-2 = AUC_{insulin/glucose}× composite ISI [24].

Statistical analysis

We compared the baseline characteristics of the study population across the FLI groups (low, intermediate, and high) using a one-way analysis of variance for continuous variables and a chi-squared test for categorical variables. Logarithmic or modified Box-Cox transformations were applied to normalize the skewed variables. Continuous data are expressed as means ± standard deviations or medians (25th and 75th percentiles), whereas categorical data are summarized as frequencies and percentages. A multivariate linear regression analysis was conducted to assess the independent association between the FLI as a continuous variable and insulin sensitivity and secretion markers in the baseline population. Model 1 comprised the correlation between the FLI and markers of insulin sensitivity and secretion. Model 2 comprised the beta coefficient of the FLI and P value after adjusting for age, SBP, DBP, TC, and HDL-C. Model 3 was adjusted for BMI to examine the association with each marker. Additionally, we used an analysis of covariance to evaluate the differences in insulin sensitivity and secretion markers across the three FLI groups. Covariates, including age, sex, SBP, BMI, TC, and HDL-C, which are established risk factors for T2DM, were adjusted [26].

The cumulative incidence of T2DM was illustrated using Kaplan–Meier curves and log-rank tests. Differences in the cumulative incidence across the FLI groups were assessed using log-rank tests. To determine the risk of T2DM while controlling for potential confounding factors, we used multivariate Cox proportional hazard regression models. Hazard ratios (HRs) for the T2DM incidence were reported with corresponding 95% confidence intervals (CIs). Model 1 evaluated the crude association between the T2DM risk and FLI. Model 2 adjusted for common T2DM risk factors, including age, sex, BMI, BP, lipid profile, smoking, alcohol consumption, and family history of diabetes [26]. Model 3 included additional adjustments for baseline insulin sensitivity and secretion markers to assess the independent effects of the FLI. The Schoenfeld residuals test showed that the proportional hazards assumption was satisfied in these models.

The ROC analysis was performed to calculate the AUC to evaluate the accuracy of the FLI, insulin sensitivity, and insulin secretion markers when predicting the T2DM incidence. The Youden index was calculated to improve diagnostic precision. Differences between the AUCs were examined using the DeLong test, and statistical significance was set at a two-tailed P < 0.05. All statistical analyses were performed using R software (version 4.2.3; R Foundation for Statistical Computing, Vienna, Austria).

Results

Baseline characteristics

The baseline characteristics of the 6,083 participants with NGT were categorized into three FLI groups (Table 1). In our study, we observed a distinct sex-based difference in the distribution of FLI categories. Among the 2,891 male participants, 530 (18.3%) were classified as the high FLI group (≥60), whereas only 200 (6.3%) of the 3,192 female participants were classified as the high FLI group. Participants in the high FLI group had higher mean BMI, WC, SBP, DBP, FPG, 1-h glucose, and 2-h glucose levels during the OGTT as well as higher mean HbA1c, TC, TG, AST, ALT, and GGT levels than those of participants in the low FLI group. Conversely, HDL-C levels were lower in the high FLI group. Although smoking status did not differ significantly among the groups, alcohol consumption (measured in grams of alcohol) was higher in the high FLI group. No significant difference in the family history of diabetes was observed. The HOMA-IR values increased and the composite ISI values decreased in the high FLI group. IGI₆₀ and AUC_{insulin/glucose}, which are insulin secretion markers, were increased in the high FLI group. The ISSI-2 decreased in the high FLI group, although the 60-min disposition index (DI₆₀) showed no significant differences across the FLI groups.

Download:

Table 1. Baseline characteristics of the study population according to the fatty liver index (FLI).

https://doi.org/10.1371/journal.pone.0327058.t001

Association among the FLI and insulin sensitivity and secretion markers

We used a multiple linear regression analysis to evaluate the association between the FLI, insulin sensitivity, and secretion markers in the NGT state. The standardized β coefficients for the FLI were calculated for each marker, and the results are summarized in Table 2. The FLI was not significantly associated with the oral DI₆₀. After adjusting for age, SBP, DBP, TC, and HDL-C, the FLI was positively correlated with the HOMA-IR and negatively correlated with the composite ISI and ISSI-2. However, after further adjustment for BMI, only the composite ISI retained a statistically significant correlation with the FLI (β= −0.024, P = 0.036, adjusted R² = 0.105). Additionally, we conducted an analysis of covariance with covariate adjustments to examine the differences across the categorized FLI groups (Fig 2). We selected covariates, including age, sex, SBP, BMI, TC, and HDL-C. The HOMA-IR and composite ISI showed significant differences between the FLI groups (P < 0.0001). The oral DI₆₀ and ISSI-2 demonstrated statistically significant differences across the FLI groups, even after covariate adjustment (P = 0.027 and P = 0.011, respectively). A further post hoc analysis indicated that the oral DI₆₀ and ISSI-2 were statistically significantly lower in the high FLI group than those of the intermediate FLI group.

Download:

Table 2. Standardized beta coefficient of the fatty liver index (FLI) for insulin sensitivity or secretion markers according to the multivariate linear regression.

https://doi.org/10.1371/journal.pone.0327058.t002

Download:

Fig 2. OGTT-derived insulin sensitivity and secretion markers across FLI groups adjusted for covariates.

The P values at the top of the figures were calculated using an ANCOVA adjusted for age, sex, SBP, BMI, TC, and the HDL-C level. ^aThe logarithmic transformation was performed to calculate the P values. ^bModified Box-Cox transformation was applied to calculate P values. ^cSignificant difference (P < 0.05, after Bonferroni’s correction for post hoc analysis). Abbreviations: ANCOVA, analysis of covariance; BMI, body mass index; DI, disposition index; FLI, fatty liver index; HDL-C, high-density lipoprotein cholesterol; HOMA-IR, homeostasis model assessment of insulin resistance; ISI, insulin sensitivity index; ISSI-2, insulin secretion-sensitivity index-2; OGTT, oral glucose tolerance test; SBP, systolic blood pressure; TC, total cholesterol.

https://doi.org/10.1371/journal.pone.0327058.g002

Risk of the incidence of T2DM across the FLI groups

We analyzed the cumulative incidence of T2DM during the 18-year follow-up period. The incidence was calculated biennially. Of the 6,083 participants, 1,006 (16.5%) individuals developed T2DM. In the low FLI group, the cumulative incidence of T2DM was 11.3% (n = 430). The high FLI group comprised 31.9% of the patients (n = 233). The median follow-up duration was 12.4 years (interquartile range, 9.8–17.8 years). The cumulative incidence of T2DM over the course of 18 years was illustrated using a cumulative incidence curve (Fig 3). The high FLI group exhibited a significantly higher incidence rate, as indicated by the curve (log-rank test; P < 0.0001).

Download:

Fig 3. Cumulative incidence of T2DM over the 18-year follow-up period across the FLI groups. Abbreviations: FLI, fatty liver index; T2DM, type 2 diabetes mellitus.

https://doi.org/10.1371/journal.pone.0327058.g003

We conducted a multivariate Cox proportional hazards regression analysis to determine whether the FLI was an independent predictor of incident T2DM (Table 3). In model 1, the HRs for incident T2DM were 2.08 (95% CI, 1.80–2.40) for the intermediate FLI group and 3.41 (95% CI, 2.91–4.00) for the high FLI group. In model 2, we adjusted for factors known to influence the occurrence of T2DM, such as age, sex, BMI, TC, HDL-C, SBP, current smoking status, alcohol intake, and family history of diabetes [27]. The HRs were 1.71 (95% CI, 1.43–2.05) and 2.57 (95% CI, 2.02–3.26) for the intermediate and high FLI groups, respectively. In model 3, which was further adjusted for insulin sensitivity and secretion markers, including the HOMA-IR, composite ISI, oral DI₆₀, and ISSI-2, the HRs were 1.57 (95% CI, 1.31–1.88) for the intermediate FLI group and 1.96 (95% CI, 1.54–2.50) for the high FLI group.

Download:

Table 3. Hazard ratios and 95% confidence intervals of T2DM according to the FLI.

https://doi.org/10.1371/journal.pone.0327058.t003

Additionally, to investigate sex-specific differences in the incidence of T2DM in our study, we conducted a sex-stratified multivariate Cox proportional hazards regression analysis (Table S1). Interaction terms (sex × FLI groups) were not statistically significant in any of the three models (all P for interaction >0.05).

Predictive ability of the FLI for T2DM

We evaluated the predictive ability of the FLI for T2DM and compared it with that of insulin sensitivity and secretion markers. We conducted a ROC analysis that revealed that the FLI had the highest AUC value at 0.654 (95% CI, 0.636–0.673); ISSI-2, composite ISI, oral DI₆₀, and HOMA-IR had the next highest AUC values (Table 4 and Fig 4). Differences between the AUCs were tested using the DeLong test, which showed that the FLI was significantly different from other insulin markers (P < 0.05).

Download:

Table 4. Predictive ability of the FLI for T2DM.

https://doi.org/10.1371/journal.pone.0327058.t004

Download:

Fig 4. ROC curve of the FLI and insulin sensitivity and secretion markers.

Abbreviations: FLI, fatty liver index; ROC, receiver-operating characteristic.

https://doi.org/10.1371/journal.pone.0327058.g004

Discussion

Our large-scale, community-based, retrospective cohort study showed that a high FLI, even for individuals with NGT, could serve as an early predictor of T2DM. High FLI levels are associated with reduced insulin sensitivity and secretion. The FLI exhibited a slightly better predictive ability than other insulin sensitivity and secretion markers. Furthermore, the FLI remained independently associated with the incidence of T2DM, even after adjusting for common metabolic factors related to T2DM development, insulin sensitivity, and secretion markers.

The Diabetes Prevention Program focused on individuals with prediabetes to study the outcomes related to progression to T2DM. Among those who experienced progression to T2DM, clear differences were noted when they were compared with individuals who did not experience progression to T2DM. Those who experienced progression to T2DM exhibited low liver attenuation values (<40 Hounsfield units), indicating a fatty liver, high level of insulin resistance, high WC, and high TG level [28]. These factors correspond to the components included in the FLI. However, the absence of initial data regarding fatty liver from the start of the Diabetes Prevention Program study made it difficult to definitively establish causality. We compared the initial data and data obtained during 18 years of follow-up to determine the prevalence of T2DM among the FLI groups.

Previous FLI studies have primarily focused on populations with prediabetes at high risk for progression to new-onset T2DM. The results indicated that the FLI is a simple and practical tool that can be used to predict progression to T2DM [16,17,29]. Additionally, several studies have investigated the role of the FLI as a predictive factor for the general population, including individuals without diabetes [30–32]. However, those studies only included individuals with prediabetes. Because glucose intolerance increases the risk of T2DM, studies that specifically focused on populations with normal glucose tolerance have been relatively limited. A study in Japan that used real-world data analyzed the incidence of T2DM according to FLI groups in populations with and without IFG. However, individuals with IGT within the prediabetes range were not excluded because OGTT data were not used. Therefore, this population cannot be truly defined as having NGT [33]. Additionally, it is important to note that the study performed in Japan did not compare insulin resistance and was not adjusted for it, which could have influenced the results. The lower incidence of T2DM observed in individuals without IFG in that study compared to that of our study may be attributable to the small sample size and lack of OGTT results, which prevented the complete identification of true diabetes cases. To the best of our knowledge, only one study has included populations with NGT as determined by the OGTT. A study conducted in Spain focused on individuals with NGT and compared populations with NGT and prediabetes separately. Among the 2,260 participants, 641 had prediabetes and 1,619 had OGTT-confirmed NGT. The Cox regression model included the HOMA-IR as an adjustment variable. In the ROC curve analysis, the AUC values were calculated for each group; the addition of the FLI to conventional T2DM risk factors resulted in an improvement in the AUC values. This improvement was particularly evident in models that did not include the HOMA-IR [34]. In our study, among individuals with NGT, the correlation between the FLI and composite ISI was stronger than that between the FLI and HOMA-IR. Additionally, in terms of the predictive value, the FLI had the highest AUC; the ISSI-2 and composite ISI had the next highest AUC. Therefore, as a predictor of T2DM before the onset of glucose intolerance, the FLI appears to be superior to the HOMA-IR and other insulin markers. Notably, our study included a large population and excluded individuals with IFG and IGT based on the OGTT results. Moreover, we compared insulin markers of individuals with OGTT-confirmed NGT and identified their differences. After adjusting for insulin sensitivity markers, we also demonstrated that the FLI differed significantly even among individuals with NGT.

The mechanisms that link NAFLD and T2DM remain unclear. However, lipid accumulation in the liver is linked to both hepatic insulin resistance and hepatic inflammation. Moreover, currently, increasing evidence indicates that lipid accumulation in the liver is also closely associated with insulin resistance in adipose tissue and muscle [35–37]. In our study, individuals with normal glucose levels in the high FLI group exhibited low insulin sensitivity and secretion. A dose-response meta-analysis revealed a direct relationship between the FLI and incidence of T2DM [38]. Therefore, we further aimed to investigate the association between the FLI and established insulin resistance markers of individuals with NGT. As shown in Table 2, the multivariate linear regression analysis of model 1 demonstrated a linear relation among the FLI and HOMA-IR, composite ISI, and ISSI-2. However, the oral DI₆₀ did not exhibit a significant linear association. In model 3, the composite ISI remained significantly associated after adjusting for BMI, whereas the associations with HOMA-IR and ISSI-2 disappeared. This suggests that these markers are highly associated with BMI. In contrast, a group-based analysis revealed that the differences between the FLI groups remained statistically significant even after adjusting for BMI (Fig 2). Participants in the high FLI group exhibited more pronounced impairments in insulin secretion. Both the oral DI₆₀, which is associated with early insulin secretion, and the ISSI-2, which is an insulin secretion index adjusted for insulin sensitivity, of the high FLI group were lower than those of the intermediate FLI group.

After performing multivariate Cox proportional hazards regression, we demonstrated that the FLI remained independently associated with the incidence of T2DM even after adjusting for established metabolic factors, insulin sensitivity, and secretion markers (Table 3). Additionally, compared with insulin resistance and secretion markers, the FLI had a better predictive rate for T2DM. However, the AUC value was approximately 0.654, which was not particularly high. However, this low predictive rate may have occurred because our study focused on individuals with NGT and included approximately half of the participants in the low FLI group, who likely had a low probability of NAFLD.

This study had several strengths. First, we analyzed data of a large, community-based cohort who underwent extensive follow-up for 18 years and provided robust insights into disease progression. Second, to our knowledge, this is the first study to evaluate the role of the FLI as an early predictor of T2DM in individuals with NGT as determined by OGTT results in Asia. Our findings underscore the potential of the FLI as a noninvasive tool that can be used to assess the early T2DM risk, thus suggesting that liver steatosis may predict T2DM development earlier than previously anticipated. Finally, this is the first study to compare the predictive ability of the FLI with OGTT-derived insulin sensitivity and secretion markers; the results of this comparison demonstrated that the FLI outperformed these markers. Our study had several limitations. First, we excluded data regarding lifestyle factors, particularly physical activity, which is an important factor in the progression of NGT to prediabetes and T2DM. because of the nature of the data, it was challenging to accurately account for physical activity; therefore, it was excluded from the analysis. Second, this cohort study measured data at 0, 60, and 120 min during the OGTT, thus limiting the ability to use blood glucose and insulin values at 30 min. Therefore, OGTT-derived insulin sensitivity and secretion markers based on these measurements may not fully reflect insulin secretion and sensitivity of adults with NGT [39]. However, the oral DI₆₀ calculated from the IGI₆₀ demonstrated a strong correlation with the 30-min IGI (IGI₃₀), making it a valid indicator of early insulin secretion. Additionally, the ISSI-2, as derived from the OGTT, exhibits a moderate correlation with the DI, supporting its reliability [24]. Third, misclassification of the NGT status may occur because of the limited reproducibility of a single OGTT measurement because individual glucose tolerance may exhibit some degree of variability [40]. However, our study included a large-scale cohort and the OGTT was conducted at baseline, which may help mitigate this limitation.

In conclusion, we demonstrated that the FLI can serve as an early predictor of T2DM, even among individuals with NGT. By comparing the FLI with established markers of insulin resistance and insulin secretion, we further evaluated its ability to reflect underlying insulin sensitivity and β-cell function. These findings emphasize the critical role of liver steatosis, as indicated by the FLI, in the early pathogenesis of T2DM, even among individuals with NGT. Therefore, individuals with a high FLI and NGT may be at high risk for metabolic dysfunction with liver steatosis, thus emphasizing the need for early preventive lifestyle modifications to mitigate the development of T2DM.

Supporting information

S1 Table. Hazard ratios and 95% confidence intervals of T2DM according to the FLI, stratified by sex.

https://doi.org/10.1371/journal.pone.0327058.s001

(DOCX)

Acknowledgments

Data in this study were from the Korean Genome and Epidemiology Study (KoGES; 6635−302), National Institute of Health, and Korea Disease Control and Prevention Agency, Republic of Korea.

References

1. Ali MK, Bullard KM, Saydah S, Imperatore G, Gregg EW. Cardiovascular and renal burdens of prediabetes in the USA: analysis of data from serial cross-sectional surveys, 1988–2014. Lancet Diabetes Endocrinol. 2018;6(5):392–403.
- View Article
- Google Scholar
2. Rawshani A, Rawshani A, Franzén S, Eliasson B, Svensson A-M, Miftaraj M, et al. Mortality and cardiovascular disease in type 1 and type 2 diabetes. N Engl J Med. 2017;376(15):1407–18. pmid:28402770
- View Article
- PubMed/NCBI
- Google Scholar
3. Rawshani A, Rawshani A, Franzén S, Sattar N, Eliasson B, Svensson A-M, et al. Risk factors, mortality, and cardiovascular outcomes in patients with type 2 diabetes. N Engl J Med. 2018;379(7):633–44. pmid:30110583
- View Article
- PubMed/NCBI
- Google Scholar
4. Cho NH, Shaw JE, Karuranga S, Huang Y, da Rocha Fernandes JD, Ohlrogge AW, et al. IDF Diabetes Atlas: Global estimates of diabetes prevalence for 2017 and projections for 2045. Diabetes Res Clin Pract. 2018;138:271–81. pmid:29496507
- View Article
- PubMed/NCBI
- Google Scholar
5. Park SE, Ko SH, Kim JY, Kim K, Moon JH, Kim NH, et al. Diabetes fact sheets in Korea 2024. Diabetes & Metabolism J. 2025;49(1):24.
- View Article
- Google Scholar
6. Targher G, Corey KE, Byrne CD, Roden M. The complex link between NAFLD and type 2 diabetes mellitus - mechanisms and treatments. Nat Rev Gastroenterol Hepatol. 2021;18(9):599–612. pmid:33972770
- View Article
- PubMed/NCBI
- Google Scholar
7. Marušić M, Paić M, Knobloch M, Liberati Pršo A-M. NAFLD, insulin resistance, and diabetes mellitus type 2. Can J Gastroenterol Hepatol. 2021;2021:6613827. pmid:33681089
- View Article
- PubMed/NCBI
- Google Scholar
8. Mantovani A, Byrne CD, Bonora E, Targher G. Nonalcoholic fatty liver disease and risk of incident type 2 diabetes: a meta-analysis. Diabetes Care. 2018;41(2):372–82. pmid:29358469
- View Article
- PubMed/NCBI
- Google Scholar
9. Gofton C, Upendran Y, Zheng MH, George J. MAFLD: how is it different from NAFLD?. Clinical and Molecular Hepatology. 2023;29(Suppl):S17.
- View Article
- Google Scholar
10. Carvalhana S, Leitão J, Alves AC, Bourbon M, Cortez-Pinto H. How good is controlled attenuation parameter and fatty liver index for assessing liver steatosis in general population: correlation with ultrasound. Liver Int. 2014;34(6):e111-7. pmid:24034415
- View Article
- PubMed/NCBI
- Google Scholar
11. Yang BL, Wu WC, Fang KC, Wang YC, Huo TI, Huang YH, et al. External validation of fatty liver index for identifying ultrasonographic fatty liver in a large-scale cross-sectional study in Taiwan. PLoS One. 2015;10(3):e0120443.
- View Article
- Google Scholar
12. Kosmalski M, Śliwińska A, Drzewoski J. Non-alcoholic fatty liver disease or type 2 diabetes mellitus-the chicken or the egg dilemma. Biomedicines. 2023;11(4):1097. pmid:37189715
- View Article
- PubMed/NCBI
- Google Scholar
13. Han AL. Validation of fatty liver index as a marker for metabolic dysfunction-associated fatty liver disease. Diabetol Metab Syndr. 2022;14(1):44. pmid:35317824
- View Article
- PubMed/NCBI
- Google Scholar
14. Khang AR, Lee HW, Yi D, Kang YH, Son SM. The fatty liver index, a simple and useful predictor of metabolic syndrome: analysis of the Korea National Health and Nutrition Examination Survey 2010–2011. Diabetes, Metabolic Syndrome and Obesity: Targets and Therapy. 2019;:181–90.
- View Article
- Google Scholar
15. Rogulj D, Konjevoda P, Milić M, Mladinić M, Domijan A-M. Fatty liver index as an indicator of metabolic syndrome. Clin Biochem. 2012;45(1–2):68–71. pmid:22056738
- View Article
- PubMed/NCBI
- Google Scholar
16. Kitazawa A, Maeda S, Fukuda Y. Fatty liver index as a predictive marker for the development of diabetes: A retrospective cohort study using Japanese health check-up data. PLoS One. 2021;16(9):e0257352. pmid:34543321
- View Article
- PubMed/NCBI
- Google Scholar
17. Busquets-Cortés C, Bennasar-Veny M, López-González A-A, Fresneda S, Aguiló A, Yanez A. Fatty liver index and progression to type 2 diabetes: a 5-year longitudinal study in Spanish workers with pre-diabetes. BMJ Open. 2021;11(8):e045498. pmid:34433590
- View Article
- PubMed/NCBI
- Google Scholar
18. Kim Y, Han B-G, KoGES group. Cohort Profile: the Korean Genome and Epidemiology Study (KoGES) consortium. Int J Epidemiol. 2017;46(2):e20. pmid:27085081
- View Article
- PubMed/NCBI
- Google Scholar
19. Friedewald WT, Levy RI, Fredrickson DS. Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge. Clin Chem. 1972;18(6):499–502. pmid:4337382
- View Article
- PubMed/NCBI
- Google Scholar
20. Bedogni G, Bellentani S, Miglioli L, Masutti F, Passalacqua M, Castiglione A, et al. The Fatty Liver Index: a simple and accurate predictor of hepatic steatosis in the general population. BMC Gastroenterol. 2006;6:33. pmid:17081293
- View Article
- PubMed/NCBI
- Google Scholar
21. Matthews DR, Hosker JP, Rudenski AS, Naylor BA, Treacher DF, Turner RC. Homeostasis model assessment: insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia. 1985;28(7):412–9. pmid:3899825
- View Article
- PubMed/NCBI
- Google Scholar
22. Matsuda M, DeFronzo RA. Insulin sensitivity indices obtained from oral glucose tolerance testing: comparison with the euglycemic insulin clamp. Diabetes Care. 1999;22(9):1462–70.
- View Article
- Google Scholar
23. Seltzer HS, Allen EW, Herron AL Jr, Brennan MT. Insulin secretion in response to glycemic stimulus: relation of delayed initial release to carbohydrate intolerance in mild diabetes mellitus. J Clin Invest. 1967;46(3):323–35. pmid:6023769
- View Article
- PubMed/NCBI
- Google Scholar
24. Retnakaran R, Qi Y, Goran MI, Hamilton JK. Evaluation of proposed oral disposition index measures in relation to the actual disposition index. Diabet Med. 2009;26(12):1198–203. pmid:20002470
- View Article
- PubMed/NCBI
- Google Scholar
25. Oh TJ, Min SH, Ahn CH, Kim EK, Kwak SH, Jung HS, et al. Normal glucose tolerance with a high 1-hour postload plasma glucose level exhibits decreased β-cell function similar to impaired glucose tolerance. Diabetes Metab J. 2015;39(2):147–53. pmid:25922809
- View Article
- PubMed/NCBI
- Google Scholar
26. American Diabetes Association Professional Practice Committee. 2. diagnosis and classification of diabetes: standards of care in diabetes-2024. Diabetes Care. 2024;47(Suppl 1):S20–42. pmid:38078589
- View Article
- PubMed/NCBI
- Google Scholar
27. Care D. Standards of care in diabetes—2023. Diabetes care. 2023;46:S1–267.
- View Article
- Google Scholar
28. Zhu Y, Hu H, Wu Y, Rao Y, Li Q, Duan X, et al. The association between fatty liver index and onset of diabetes: secondary analysis of a population-based cohort study. BMC Public Health. 2023;23(1):679. pmid:37041534
- View Article
- PubMed/NCBI
- Google Scholar
29. Franch-Nadal J, Caballeria L, Mata-Cases M, Mauricio D, Giraldez-García C, Mancera J, et al. Fatty liver index is a predictor of incident diabetes in patients with prediabetes: The PREDAPS study. PLoS One. 2018;13(6):e0198327. pmid:29856820
- View Article
- PubMed/NCBI
- Google Scholar
30. Seo I-H, Lee HS, Lee Y-J. Fatty liver index as a predictor for incident type 2 diabetes in community-dwelling adults: longitudinal findings over 12 years. Cardiovascular Diabetology. 2022;21(1):209.
- View Article
- Google Scholar
31. Higashiura Y, Furuhashi M, Tanaka M, Takahashi S, Koyama M, Ohnishi H, et al. High level of fatty liver index predicts new onset of diabetes mellitus during a 10-year period in healthy subjects. Sci Rep. 2021;11(1):12830. pmid:34145341
- View Article
- PubMed/NCBI
- Google Scholar
32. Jäger S, Jacobs S, Kröger J, Stefan N, Fritsche A, Weikert C, et al. Association between the fatty liver index and risk of type 2 diabetes in the EPIC-Potsdam Study. PLoS One. 2015;10(4):e0124749. pmid:25902304
- View Article
- PubMed/NCBI
- Google Scholar
33. Hirata A, Sugiyama D, Kuwabara K, Hirata T, Tsutatani H, Funamoto M, et al. Fatty liver index predicts incident diabetes in a Japanese general population with and without impaired fasting glucose. Hepatol Res. 2018;48(9):708–16. pmid:29341419
- View Article
- PubMed/NCBI
- Google Scholar
34. García-Escobar E, Valdés S, Soriguer F, Vendrell J, Urrutia-Etxebarria IM, Maldonado-Araque C, et al. Fatty liver index as a predictor for type 2 diabetes in subjects with normoglycemia in a nationwide cohort study. Sci Rep. 2021;11(1):16453. pmid:34385479
- View Article
- PubMed/NCBI
- Google Scholar
35. Sung K-C, Jeong W-S, Wild SH, Byrne CD. Combined influence of insulin resistance, overweight/obesity, and fatty liver as risk factors for type 2 diabetes. Diabetes Care. 2012;35(4):717–22. pmid:22338098
- View Article
- PubMed/NCBI
- Google Scholar
36. Yki-Järvinen H. Ceramides: a cause of insulin resistance in nonalcoholic fatty liver disease in both murine models and humans. Hepatology. 2020;71(4):1499–501. pmid:31899812
- View Article
- PubMed/NCBI
- Google Scholar
37. Byrne CD, Targher G. NAFLD: a multisystem disease. J Hepatol. 2015;62(1 Suppl):S47–64. pmid:25920090
- View Article
- PubMed/NCBI
- Google Scholar
38. Movahedian M, Rahmani J, Hashemi Nazari SS, Mohamadi S, Naik G, Hekmatdoost A. Fatty liver index and risk of diabetes incidence: A systematic review and dose-response meta-analysis of cohort studies. Prim Care Diabetes. 2020;14(6):577–83. pmid:32694004
- View Article
- PubMed/NCBI
- Google Scholar
39. Hayashi T, Boyko EJ, Sato KK, McNeely MJ, Leonetti DL, Kahn SE, et al. Patterns of insulin concentration during the OGTT predict the risk of type 2 diabetes in Japanese Americans. Diabetes Care. 2013;36(5):1229–35. pmid:23275353
- View Article
- PubMed/NCBI
- Google Scholar
40. Ko GT, Chan JC, Woo J, Lau E, Yeung VT, Chow CC, et al. The reproducibility and usefulness of the oral glucose tolerance test in screening for diabetes and other cardiovascular risk factors. Ann Clin Biochem. 1998;35 ( Pt 1):62–7. pmid:9463740
- View Article
- PubMed/NCBI
- Google Scholar

[ref1] 1. Ali MK, Bullard KM, Saydah S, Imperatore G, Gregg EW. Cardiovascular and renal burdens of prediabetes in the USA: analysis of data from serial cross-sectional surveys, 1988–2014. Lancet Diabetes Endocrinol. 2018;6(5):392–403.
View Article
Google Scholar

[2] View Article

[3] Google Scholar

[ref2] 2. Rawshani A, Rawshani A, Franzén S, Eliasson B, Svensson A-M, Miftaraj M, et al. Mortality and cardiovascular disease in type 1 and type 2 diabetes. N Engl J Med. 2017;376(15):1407–18. pmid:28402770
View Article
PubMed/NCBI
Google Scholar

[5] View Article

[6] PubMed/NCBI

[7] Google Scholar

[ref3] 3. Rawshani A, Rawshani A, Franzén S, Sattar N, Eliasson B, Svensson A-M, et al. Risk factors, mortality, and cardiovascular outcomes in patients with type 2 diabetes. N Engl J Med. 2018;379(7):633–44. pmid:30110583
View Article
PubMed/NCBI
Google Scholar

[9] View Article

[10] PubMed/NCBI

[11] Google Scholar

[ref4] 4. Cho NH, Shaw JE, Karuranga S, Huang Y, da Rocha Fernandes JD, Ohlrogge AW, et al. IDF Diabetes Atlas: Global estimates of diabetes prevalence for 2017 and projections for 2045. Diabetes Res Clin Pract. 2018;138:271–81. pmid:29496507
View Article
PubMed/NCBI
Google Scholar

[13] View Article

[14] PubMed/NCBI

[15] Google Scholar

[ref5] 5. Park SE, Ko SH, Kim JY, Kim K, Moon JH, Kim NH, et al. Diabetes fact sheets in Korea 2024. Diabetes & Metabolism J. 2025;49(1):24.
View Article
Google Scholar

[17] View Article

[18] Google Scholar

[ref6] 6. Targher G, Corey KE, Byrne CD, Roden M. The complex link between NAFLD and type 2 diabetes mellitus - mechanisms and treatments. Nat Rev Gastroenterol Hepatol. 2021;18(9):599–612. pmid:33972770
View Article
PubMed/NCBI
Google Scholar

[20] View Article

[21] PubMed/NCBI

[22] Google Scholar

[ref7] 7. Marušić M, Paić M, Knobloch M, Liberati Pršo A-M. NAFLD, insulin resistance, and diabetes mellitus type 2. Can J Gastroenterol Hepatol. 2021;2021:6613827. pmid:33681089
View Article
PubMed/NCBI
Google Scholar

[24] View Article

[25] PubMed/NCBI

[26] Google Scholar

[ref8] 8. Mantovani A, Byrne CD, Bonora E, Targher G. Nonalcoholic fatty liver disease and risk of incident type 2 diabetes: a meta-analysis. Diabetes Care. 2018;41(2):372–82. pmid:29358469
View Article
PubMed/NCBI
Google Scholar

[28] View Article

[29] PubMed/NCBI

[30] Google Scholar

[ref9] 9. Gofton C, Upendran Y, Zheng MH, George J. MAFLD: how is it different from NAFLD?. Clinical and Molecular Hepatology. 2023;29(Suppl):S17.
View Article
Google Scholar

[32] View Article

[33] Google Scholar

[ref10] 10. Carvalhana S, Leitão J, Alves AC, Bourbon M, Cortez-Pinto H. How good is controlled attenuation parameter and fatty liver index for assessing liver steatosis in general population: correlation with ultrasound. Liver Int. 2014;34(6):e111-7. pmid:24034415
View Article
PubMed/NCBI
Google Scholar

[35] View Article

[36] PubMed/NCBI

[37] Google Scholar

[ref11] 11. Yang BL, Wu WC, Fang KC, Wang YC, Huo TI, Huang YH, et al. External validation of fatty liver index for identifying ultrasonographic fatty liver in a large-scale cross-sectional study in Taiwan. PLoS One. 2015;10(3):e0120443.
View Article
Google Scholar

[39] View Article

[40] Google Scholar

[ref12] 12. Kosmalski M, Śliwińska A, Drzewoski J. Non-alcoholic fatty liver disease or type 2 diabetes mellitus-the chicken or the egg dilemma. Biomedicines. 2023;11(4):1097. pmid:37189715
View Article
PubMed/NCBI
Google Scholar

[42] View Article

[43] PubMed/NCBI

[44] Google Scholar

[ref13] 13. Han AL. Validation of fatty liver index as a marker for metabolic dysfunction-associated fatty liver disease. Diabetol Metab Syndr. 2022;14(1):44. pmid:35317824
View Article
PubMed/NCBI
Google Scholar

[46] View Article

[47] PubMed/NCBI

[48] Google Scholar

[ref14] 14. Khang AR, Lee HW, Yi D, Kang YH, Son SM. The fatty liver index, a simple and useful predictor of metabolic syndrome: analysis of the Korea National Health and Nutrition Examination Survey 2010–2011. Diabetes, Metabolic Syndrome and Obesity: Targets and Therapy. 2019;:181–90.
View Article
Google Scholar

[50] View Article

[51] Google Scholar

[ref15] 15. Rogulj D, Konjevoda P, Milić M, Mladinić M, Domijan A-M. Fatty liver index as an indicator of metabolic syndrome. Clin Biochem. 2012;45(1–2):68–71. pmid:22056738
View Article
PubMed/NCBI
Google Scholar

[53] View Article

[54] PubMed/NCBI

[55] Google Scholar

[ref16] 16. Kitazawa A, Maeda S, Fukuda Y. Fatty liver index as a predictive marker for the development of diabetes: A retrospective cohort study using Japanese health check-up data. PLoS One. 2021;16(9):e0257352. pmid:34543321
View Article
PubMed/NCBI
Google Scholar

[57] View Article

[58] PubMed/NCBI

[59] Google Scholar

[ref17] 17. Busquets-Cortés C, Bennasar-Veny M, López-González A-A, Fresneda S, Aguiló A, Yanez A. Fatty liver index and progression to type 2 diabetes: a 5-year longitudinal study in Spanish workers with pre-diabetes. BMJ Open. 2021;11(8):e045498. pmid:34433590
View Article
PubMed/NCBI
Google Scholar

[61] View Article

[62] PubMed/NCBI

[63] Google Scholar

[ref18] 18. Kim Y, Han B-G, KoGES group. Cohort Profile: the Korean Genome and Epidemiology Study (KoGES) consortium. Int J Epidemiol. 2017;46(2):e20. pmid:27085081
View Article
PubMed/NCBI
Google Scholar

[65] View Article

[66] PubMed/NCBI

[67] Google Scholar

[ref19] 19. Friedewald WT, Levy RI, Fredrickson DS. Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge. Clin Chem. 1972;18(6):499–502. pmid:4337382
View Article
PubMed/NCBI
Google Scholar

[69] View Article

[70] PubMed/NCBI

[71] Google Scholar

[ref20] 20. Bedogni G, Bellentani S, Miglioli L, Masutti F, Passalacqua M, Castiglione A, et al. The Fatty Liver Index: a simple and accurate predictor of hepatic steatosis in the general population. BMC Gastroenterol. 2006;6:33. pmid:17081293
View Article
PubMed/NCBI
Google Scholar

[73] View Article

[74] PubMed/NCBI

[75] Google Scholar

[ref21] 21. Matthews DR, Hosker JP, Rudenski AS, Naylor BA, Treacher DF, Turner RC. Homeostasis model assessment: insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia. 1985;28(7):412–9. pmid:3899825
View Article
PubMed/NCBI
Google Scholar

[77] View Article

[78] PubMed/NCBI

[79] Google Scholar

[ref22] 22. Matsuda M, DeFronzo RA. Insulin sensitivity indices obtained from oral glucose tolerance testing: comparison with the euglycemic insulin clamp. Diabetes Care. 1999;22(9):1462–70.
View Article
Google Scholar

[81] View Article

[82] Google Scholar

[ref23] 23. Seltzer HS, Allen EW, Herron AL Jr, Brennan MT. Insulin secretion in response to glycemic stimulus: relation of delayed initial release to carbohydrate intolerance in mild diabetes mellitus. J Clin Invest. 1967;46(3):323–35. pmid:6023769
View Article
PubMed/NCBI
Google Scholar

[84] View Article

[85] PubMed/NCBI

[86] Google Scholar

[ref24] 24. Retnakaran R, Qi Y, Goran MI, Hamilton JK. Evaluation of proposed oral disposition index measures in relation to the actual disposition index. Diabet Med. 2009;26(12):1198–203. pmid:20002470
View Article
PubMed/NCBI
Google Scholar

[88] View Article

[89] PubMed/NCBI

[90] Google Scholar

[ref25] 25. Oh TJ, Min SH, Ahn CH, Kim EK, Kwak SH, Jung HS, et al. Normal glucose tolerance with a high 1-hour postload plasma glucose level exhibits decreased β-cell function similar to impaired glucose tolerance. Diabetes Metab J. 2015;39(2):147–53. pmid:25922809
View Article
PubMed/NCBI
Google Scholar

[92] View Article

[93] PubMed/NCBI

[94] Google Scholar

[ref26] 26. American Diabetes Association Professional Practice Committee. 2. diagnosis and classification of diabetes: standards of care in diabetes-2024. Diabetes Care. 2024;47(Suppl 1):S20–42. pmid:38078589
View Article
PubMed/NCBI
Google Scholar

[96] View Article

[97] PubMed/NCBI

[98] Google Scholar

[ref27] 27. Care D. Standards of care in diabetes—2023. Diabetes care. 2023;46:S1–267.
View Article
Google Scholar

[100] View Article

[101] Google Scholar

[ref28] 28. Zhu Y, Hu H, Wu Y, Rao Y, Li Q, Duan X, et al. The association between fatty liver index and onset of diabetes: secondary analysis of a population-based cohort study. BMC Public Health. 2023;23(1):679. pmid:37041534
View Article
PubMed/NCBI
Google Scholar

[103] View Article

[104] PubMed/NCBI

[105] Google Scholar

[ref29] 29. Franch-Nadal J, Caballeria L, Mata-Cases M, Mauricio D, Giraldez-García C, Mancera J, et al. Fatty liver index is a predictor of incident diabetes in patients with prediabetes: The PREDAPS study. PLoS One. 2018;13(6):e0198327. pmid:29856820
View Article
PubMed/NCBI
Google Scholar

[107] View Article

[108] PubMed/NCBI

[109] Google Scholar

[ref30] 30. Seo I-H, Lee HS, Lee Y-J. Fatty liver index as a predictor for incident type 2 diabetes in community-dwelling adults: longitudinal findings over 12 years. Cardiovascular Diabetology. 2022;21(1):209.
View Article
Google Scholar

[111] View Article

[112] Google Scholar

[ref31] 31. Higashiura Y, Furuhashi M, Tanaka M, Takahashi S, Koyama M, Ohnishi H, et al. High level of fatty liver index predicts new onset of diabetes mellitus during a 10-year period in healthy subjects. Sci Rep. 2021;11(1):12830. pmid:34145341
View Article
PubMed/NCBI
Google Scholar

[114] View Article

[115] PubMed/NCBI

[116] Google Scholar

[ref32] 32. Jäger S, Jacobs S, Kröger J, Stefan N, Fritsche A, Weikert C, et al. Association between the fatty liver index and risk of type 2 diabetes in the EPIC-Potsdam Study. PLoS One. 2015;10(4):e0124749. pmid:25902304
View Article
PubMed/NCBI
Google Scholar

[118] View Article

[119] PubMed/NCBI

[120] Google Scholar

[ref33] 33. Hirata A, Sugiyama D, Kuwabara K, Hirata T, Tsutatani H, Funamoto M, et al. Fatty liver index predicts incident diabetes in a Japanese general population with and without impaired fasting glucose. Hepatol Res. 2018;48(9):708–16. pmid:29341419
View Article
PubMed/NCBI
Google Scholar

[122] View Article

[123] PubMed/NCBI

[124] Google Scholar

[ref34] 34. García-Escobar E, Valdés S, Soriguer F, Vendrell J, Urrutia-Etxebarria IM, Maldonado-Araque C, et al. Fatty liver index as a predictor for type 2 diabetes in subjects with normoglycemia in a nationwide cohort study. Sci Rep. 2021;11(1):16453. pmid:34385479
View Article
PubMed/NCBI
Google Scholar

[126] View Article

[127] PubMed/NCBI

[128] Google Scholar

[ref35] 35. Sung K-C, Jeong W-S, Wild SH, Byrne CD. Combined influence of insulin resistance, overweight/obesity, and fatty liver as risk factors for type 2 diabetes. Diabetes Care. 2012;35(4):717–22. pmid:22338098
View Article
PubMed/NCBI
Google Scholar

[130] View Article

[131] PubMed/NCBI

[132] Google Scholar

[ref36] 36. Yki-Järvinen H. Ceramides: a cause of insulin resistance in nonalcoholic fatty liver disease in both murine models and humans. Hepatology. 2020;71(4):1499–501. pmid:31899812
View Article
PubMed/NCBI
Google Scholar

[134] View Article

[135] PubMed/NCBI

[136] Google Scholar

[ref37] 37. Byrne CD, Targher G. NAFLD: a multisystem disease. J Hepatol. 2015;62(1 Suppl):S47–64. pmid:25920090
View Article
PubMed/NCBI
Google Scholar

[138] View Article

[139] PubMed/NCBI

[140] Google Scholar

[ref38] 38. Movahedian M, Rahmani J, Hashemi Nazari SS, Mohamadi S, Naik G, Hekmatdoost A. Fatty liver index and risk of diabetes incidence: A systematic review and dose-response meta-analysis of cohort studies. Prim Care Diabetes. 2020;14(6):577–83. pmid:32694004
View Article
PubMed/NCBI
Google Scholar

[142] View Article

[143] PubMed/NCBI

[144] Google Scholar

[ref39] 39. Hayashi T, Boyko EJ, Sato KK, McNeely MJ, Leonetti DL, Kahn SE, et al. Patterns of insulin concentration during the OGTT predict the risk of type 2 diabetes in Japanese Americans. Diabetes Care. 2013;36(5):1229–35. pmid:23275353
View Article
PubMed/NCBI
Google Scholar

[146] View Article

[147] PubMed/NCBI

[148] Google Scholar

[ref40] 40. Ko GT, Chan JC, Woo J, Lau E, Yeung VT, Chow CC, et al. The reproducibility and usefulness of the oral glucose tolerance test in screening for diabetes and other cardiovascular risk factors. Ann Clin Biochem. 1998;35 ( Pt 1):62–7. pmid:9463740
View Article
PubMed/NCBI
Google Scholar

[150] View Article

[151] PubMed/NCBI

[152] Google Scholar

Figures

Abstract

Introduction

Materials and methods

Study population

Assessment of anthropometric and biochemical parameters

Definition and outcomes

Statistical analysis

Results

Baseline characteristics

Association among the FLI and insulin sensitivity and secretion markers

Risk of the incidence of T2DM across the FLI groups

Predictive ability of the FLI for T2DM

Discussion

Supporting information

S1 Table. Hazard ratios and 95% confidence intervals of T2DM according to the FLI, stratified by sex.

Acknowledgments

References