Warning: fopen(/home/virtual/enm-kes/journal/upload/ip_log/ip_log_2025-06.txt): failed to open stream: Permission denied in /home/virtual/lib/view_data.php on line 100 Warning: fwrite() expects parameter 1 to be resource, boolean given in /home/virtual/lib/view_data.php on line 101 Fatty Liver Index Dynamics as a Predictor of Hepatocellular Carcinoma in Patients with Type 2 Diabetes Mellitus and Non-Cirrhotic Livers
Skip Navigation
Skip to contents

Endocrinol Metab : Endocrinology and Metabolism

clarivate
OPEN ACCESS
SEARCH
Search
Advanced Search
mobile menu button

Articles

Page Path
HOME > Endocrinol Metab > Ahead-of print > Article
Original Article
Fatty Liver Index Dynamics as a Predictor of Hepatocellular Carcinoma in Patients with Type 2 Diabetes Mellitus and Non-Cirrhotic Livers
Eun-Hee Cho1*orcid, Min Gu Kang2*orcid, Chang Hun Lee3,4*orcid, Shinyoung Oh5,6, Chen Shen4,6, Ha Ram Oh4,6, Young Ran Park4,6, Hyun Lee7, Jong Seung Kim2,4,8, Ji Hyun Park4,6orcid

DOI: https://doi.org/10.3803/EnM.2024.2286
Published online: May 29, 2025

1Department of Internal Medicine, Kangwon National University School of Medicine, Chuncheon, Korea

2Department of Medical Informatics, Jeonbuk National University Medical School, Jeonju, Korea

3Division of Gastroenterology and Hepatology, Department of Internal Medicine, Jeonbuk National University Medical School, Jeonju, Korea

4Research Institute of Clinical Medicine, Jeonbuk National University, Biomedical Research Institute of Jeonbuk National University Hospital, Jeonju, Korea

5Department of Medicine, Jeonbuk National University Graduate School, Jeonju, Korea

6Division of Endocrinology and Metabolism, Department of Internal Medicine, Jeonbuk National University Medical School, Jeonju, Korea

7Department of Internal Medicine, Hanyang University College of Medicine, Seoul, Korea

8Department of Otorhinolaryngology, Jeonbuk National University Medical School, Jeonju, Korea

Corresponding author: Ji Hyun Park Division of Endocrinology and Metabolism, Department of Internal Medicine, Jeonbuk National University Medical School, 20 Geonji-ro, Deokjin-gu, Jeonju 54907, Korea Tel: +82-63-250-1780, Fax: +82-63-254-1609, E-mail: parkjh@jbnu.ac.kr
*These authors contributed equally to this work.
• Received: December 15, 2024   • Revised: February 26, 2025   • Accepted: March 13, 2025

Copyright © 2025 Korean Endocrine Society

This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0/) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.

  • 347 Views
  • 18 Download
Full Article

Download PDF Download PDF

  • Background
    Type 2 diabetes mellitus (T2DM) is a significant risk factor for hepatocellular carcinoma (HCC) in patients with nonalcoholic fatty liver disease; however, surveillance strategies for patients with T2DM, especially without cirrhosis, are inadequate. This study examined whether the fatty liver index (FLI) and its dynamic changes can effectively identify patients with T2DM at increased risk for HCC.
  • Methods
    Data from 92,761 individuals with T2DM aged 40 to 79 who underwent two health screenings (2012 to 2015) were analyzed. The FLI, calculated using waist circumference, body mass index, triglycerides, and gamma-glutamyl transferase, was used to stratify patients by baseline FLI and FLI changes between screenings. HCC cases were identified via International Classification of Diseases codes and reimbursement records (2016 to 2020).
  • Results
    Patients with baseline FLI of 30 to 59.9 had a 1.90-fold higher risk (P<0.01) and those with FLI ≥60 had a 2.94-fold higher risk (P<0.01) of developing HCC compared to those with FLI <30. An increase in FLI from <30 to ≥30 resulted in a 2.10-fold higher risk of HCC (P<0.01), while a reduction in FLI from ≥30 to <30 led to a 0.64-fold lower risk (P=0.03). Protective benefits of FLI reduction took approximately 3 years to manifest.
  • Conclusion
    Baseline and dynamic monitoring of FLI effectively identified HCC risk in T2DM patients with non-cirrhotic livers, supporting early detection and intervention.
  • Keywords: Diabetes mellitus, type 2; Carcinoma, hepatocellular; Risk; Non-alcoholic fatty liver disease
Hepatocellular carcinoma (HCC) is a serious complication of non-alcoholic fatty liver disease (NAFLD), currently referred to as metabolic dysfunction-associated steatotic liver disease, which is characterized by excessive hepatic fat without significant alcohol consumption [1]. NAFLD-related HCC typically presents in older individuals, often at an advanced stage, and can develop even without cirrhosis—distinguishing it from HCC due to viral hepatitis or other causes [2,3]. Although the overall incidence of HCC in patients with NAFLD is low, a significant proportion—approximately 50%—occurs in patients without cirrhosis, who are often not under active surveillance [4].
Diabetes is a major risk factor for HCC and advanced fibrosis in patients with NAFLD. A large European study involving 136,703 patients underscored diabetes as a critical risk factor for HCC, emphasizing the need for effective surveillance in this cohort [5-8]. Current guidelines recommend biannual abdominal ultrasounds for patients with cirrhosis when the annual HCC risk exceeds 1.5% [8]. However, expanding surveillance measures to include most patients with diabetes without cirrhosis could be a significant undertaking, considering the high prevalence and rapidly increasing incidence of diabetes worldwide. This gap highlights the urgent need to develop risk stratification models that accurately identify high-risk individuals with diabetes and non-cirrhotic livers. Consequently, simple, accessible predictive tools are vital for targeted HCC surveillance, improving early detection and intervention outcomes [9,10].
The fatty liver index (FLI) is a clinical scoring system comprising the body mass index (BMI), waist circumference, triglyceride levels, and gamma-glutamyl transferase (GGT) levels, quantified on a scale from 0 to 100. Developed by Bedogni et al. [11], FLI has proven to be a convenient and reliable predictor of NAFLD in the general Italian population. It effectively rules out NAFLD with an FLI cutoff <30 and confirms the disease’s presence with an FLI ≥60. Therefore, the FLI may aid physicians in stratifying and selecting high-risk patients for liver ultrasonography and intensified lifestyle modification [11]. Moreover, we previously [12] reported that an increase in FLI above 30 over a 2-year interval was associated with a 25% higher risk of developing HCC over 6 years, compared to individuals with stable FLI values. Conversely, a decrease in FLI is associated with a 32% reduced risk of HCC. These results confirm the utility of FLI as a critical clinical marker for assessing HCC risk in patients with NAFLD.
Given the complex interplay between diabetes, NAFLD, and HCC, this study aimed to evaluate the predictive value of dynamic changes in FLI for HCC development in a nationally representative cohort of patients with T2DM and non-cirrhotic livers.
This study was approved by the Institutional Review Board of Jeonbuk National University Hospital (CUH 2022-07-027). The need for informed consent was waived due to the retrospective nature of the study.
Data source
This study used data from the National Health Insurance Service (NHIS) database [13,14], a comprehensive nationwide population-based cohort in South Korea. All eligible Korean citizens (approximately 50 million) are mandatorily enrolled in the NHIS, which is managed by the government and serves as a single insurer. The NHIS database contains demographic characteristics, including age, sex, economic status, and residential area; medical records; and lifestyle data.
Study population
We focused on two consecutive health screenings to measure changes in FLI over the two periods. The study population included patients aged 40 to 79 years with T2DM (International Classification of Diseases, 10th revision diagnosis codes [ICD-10] code E11.x) between January 1, 2009, and December 31, 2011, who underwent two consecutive health screenings (Fig. 1). Notably, 859 patients were excluded owing to a diagnosis of malignant neoplasm of the liver and intrahepatic bile ducts before 2016, and 1,781 were excluded owing to mortality before 2016. Additionally, we excluded 83 patients who underwent liver transplantation before the end of the follow-up period and 3,087 patients with alcoholic liver disease, chronic viral hepatitis, autoimmune hepatitis, hemochromatosis, Willson’s disease, or cirrhosis at baseline. Ultimately, 92,761 patients with T2DM were included in the final analysis.
Exposures: FLI changes
This study characterized exposure as an FLI change between the second and first health screenings in patients with T2DM. We utilized FLI threshold values of 30 and 60 to categorize changes [11]. Specifically, with a threshold value of 30, FLI <30 at the first health screening and FLI ≥30 at the second health screening was considered an increase, whereas FLI ≥30 at the first health screening and FLI <30 at the second health screening was considered a decrease [12]. FLI was calculated using the following formula [11]:
FLI=e0.953×loge (triglycerides)+0.139×(BMI)+0.718×loge (GGY)+0.053×(waist circumference)–15.745(1+e0.953×loge (triglycerides)+0.139×(BMI)+0.718×loge (GGY)+0.053×(waist circumference)–15.745)×100
Study outcome: incident HCC
The primary outcome was new-onset HCC according to FLI change in patients with T2DM. The secondary outcome was the development of a new-onset HCC, according to the baseline FLI in patients with T2DM.
HCC was defined as at least one diagnosis with special benefit codes V193 or V194, based on the ICD-10 diagnosis codes C22, C220, C227, or C229 [12]. A special benefit system was established to alleviate the financial burden on patients with serious, rare, and severe congenital diseases by reducing out-of-pocket medical expenses.
The study population was followed up from the index date (January 1, 2016) until the first date of new HCC diagnosis, death, or the end of the study period (December 31, 2020), whichever occurred first.
Other variables
Other variables considered in the study were age, sex, economic status, residential area, and lifestyle habits such as alcohol consumption, smoking, and physical activity. The NHIS in Korea evaluates lifestyle habits during health checkups using specific criteria. For smoking, non-smokers and former smokers are considered ‘good,’ while current smokers are ‘bad.’ Alcohol consumption is deemed ‘good’ if men consume ≤2 drinks per day and women ≤1 drink per day, while ‘bad’ or high-risk drinking is defined as ≥7 drinks per session or ≥14 drinks per week for men, and ≥5 drinks per session or ≥7 drinks per week for women. Exercise habits are categorized as ‘good’ when individuals engage in moderate-intensity aerobic exercise ≥5 days/week (≥30 minutes/day) or vigorous-intensity ≥3 days/week (≥20 minutes/day), along with strength training ≥2 times per week. Exercise is considered ‘bad’ if aerobic exercise is performed <3 days/week and strength training ≤1 time/week.
Comorbidities were defined using specific ICD-10 codes (Supplemental Table S1) [15].
Statistical analysis
The distribution of variables among different groups was evaluated using the standardized mean difference (SMD), with an SMD <0.25 indicating a balanced distribution between groups. The HCC incidence rate was calculated by dividing the number of occurrences by the total follow-up period and expressed per 100,000 person-years (PYs). Kaplan–Meier curves were employed to visualize the cumulative incidence of HCC based on changes in the FLI. The differences in the curves were statistically evaluated using the log-rank test. Cox proportional hazard regression analysis was used to investigate the association between new-onset HCC and FLI changes. Multivariable Cox proportional hazard regression analysis was fully adjusted for age, sex, economic status, residential area, alcohol consumption, smoking, physical activity, hypertension, and hyperlipidemia. Statistical analyses were performed using SAS version 9.4 (SAS Institute Inc., Cary, NC, USA) and R version 3.2.3 (R Foundation for Statistical Computing, Vienna, Austria), and significance was set at P<0.05.
Baseline characteristics
Table 1 presents the baseline characteristics of the patients categorized by the baseline FLI groups, as determined during the second health screening in 2014 to 2015. Among the patients with T2DM, 49,168 (53.0%) had FLI <30; 27,520 (29.7%) had FLI 30–59.9; and 16,073 (17.3%) had FLI ≥60. Notably, a higher FLI was associated with a greater proportion of males and higher likelihood of having unhealthy lifestyle habits (such as alcohol consumption and smoking) (SMD >0.25).
Primary outcome: new-onset HCC according to FLI change

Increase in FLI based on the threshold of 30

Among 48,046 patients with FLI <30 at the first health screening, 40,602 and 7,444 maintained an FLI <30 and experienced an increase in FLI to ≥30, respectively, at the second health screening (Fig. 1). During follow-up, 62 (incidence rate, 31.2/100,000 PYs) and 28 (incidence rate, 77.2/100,000 PYs) new-onset HCC cases occurred in these groups, respectively. As depicted in Fig. 2A, the cumulative incidence of HCC significantly differed between patients who maintained FLI <30 and those who experienced an increase to FLI ≥30 (log-rank test, P<0.01). The risk of new-onset HCC was 2.1-fold higher in patients whose FLI increased to ≥30 in the second screening compared with those who maintained FLI <30 in the second screening (adjusted hazard ratio [aHR], 2.10; 95% confidence interval [CI], 1.34 to 3.30) (Supplemental Table S2).

Decrease in FLI based on the threshold of 30

Among 44,715 patients with FLI ≥30 at the first health screening, 36,149 and 8,566 maintained FLI ≥30 and experienced a decrease in FLI to <30, respectively, at the second health screening (Fig. 1). During follow-up, 181 (incidence rate, 102.63 per 100,000 PYs) and 27 (incidence rate, and 64.72 per 100,000 PYs) new-onset HCC cases occurred in these groups, respectively. As depicted in Fig. 2B, the cumulative incidence of HCC significantly differed between patients who maintained FLI ≥30 and those who experienced a decrease to FLI <30 (log-rank test, P=0.02), although no difference was observed during the first 2.5 years. The risk of new-onset HCC was 0.64-fold lower in patients whose FLI decreased to <30 in the second round compared with that of patients who maintained an FLI ≥30 in the second round (aHR, 0.64; 95% CI, 0.42 to 0.96) (Supplemental Table S3).

FLI change based on the threshold of 60

The increase and decrease in FLI were not statistically significant when the threshold was set at 60 with regards to the risk of new-onset HCC (increase [aHR, 1.58; 95% CI, 1.00 to 2.50; P>0.05]; decrease [aHR, 0.70; 95% CI, 0.46 to 1.08; P=0.11]) (Figs. 2C, D, and 3, Supplemental Tables S4, S5).

FLI change based on the threshold at baseline and follow-up

The graph in Fig. 4 represents a cumulative event curve of HCC incidence over a 5-year period, categorized by different groups based on FLI at baseline and follow-up. The red group (persistently high FLI) has the highest cumulative event rate of HCC, while the purple group (reference group with persistently low FLI) has the lowest. The green and blue groups fall in between. The cumulative incidence of HCC significantly differed between the groups (log-rank test, P<0.01).
Secondary outcome: new-onset HCC according to baseline FLI
Based on follow-up observations, 89, 112, and 97 new-onset HCC cases occurred among patients with a baseline FLI <30, FLI 30–59.9, and FLI ≥60, respectively. The corresponding incidence rates per 100,000 PYs were 37.02, 83.48, and 123.62, respectively. Our data showed an HCC incidence of 98.29 per 100,000 PYs in patients with non-cirrhotic NAFLD (FLI ≥30) and T2DM. The cumulative incidence of HCC differed significantly among the FLI groups (log-rank test, P<0.01) (Fig. 5). Patients with an FLI of 30–59.9 exhibited a 1.90-fold increased risk of developing new-onset HCC compared with those with an FLI <30 (aHR, 1.90; 95% CI, 1.44 to 2.52). Additionally, patients with an FLI ≥60 demonstrated a 2.94-fold higher risk of developing new-onset HCC compared with those with an FLI <30 (aHR, 2.94; 95% CI, 2.18 to 3.97) (Supplemental Table S6).
Our study aimed to evaluate the predictive value of dynamic changes in FLI for HCC development in a nationally representative cohort of patients with T2DM and non-cirrhotic livers. Our results reveal the significant impact of FLI dynamics on HCC risk in this patient cohort. Specifically, an increase in FLI from <30 to ≥30 over a 2-year period correlated with more than a doubling of HCC risk. In contrast, a decrease in FLI >30 significantly reduced HCC risk. These observations support the critical role of FLI as a dynamic and reliable marker for HCC risk stratification in this specific patient group.
Non-cirrhotic HCC typically presents at more advanced stages compared to its cirrhotic counterpart, often characterized by larger tumor sizes, higher tumor grades, and more extensive vascular invasion [16,17]. These characteristics underscore the critical importance of early detection and prevention of non-cirrhotic HCC which generally emerges de novo, in contrast to the stepwise carcinogenesis observed in cirrhotic HCC. It progresses from regenerative to dysplastic nodules and frequently involves mutations in tumor suppressor genes, such as p53. Conversely, p53 mutations are less common non-cirrhotic HCC; however, there is a higher prevalence of β-catenin mutations [18]. The pathogenesis of non-cirrhotic NAFLD-HCC encompasses systemic inflammation, insulin resistance, oxidative stress, immune dysregulation, and genetic polymorphisms, driving its distinct and aggressive pathology [19].
Changes in FLI did not show a statistically significant impact on HCC risk at the higher threshold of 60. This suggests that a threshold of 30 is more effective for early detection and intervention in our study population with T2DM, emphasizing the importance of identifying an optimal FLI cutoff for clinical use. Similar to how low-density lipoprotein cholesterol targets vary based on cardiovascular risk, it is crucial to determine the most appropriate FLI cutoff tailored to the patient’s metabolic risk factors to prevent HCC effectively. Customizing these thresholds can enhance early detection and preventive strategies, ultimately improving patient outcomes.
Moreover, it is imperative to note that the effectiveness of FLI in diagnosing NAFLD varies by ethnicity and sex due to differences in body composition and metabolic risk factors [11,20]. For instance, the optimal FLI cutoff for diagnosing NAFLD in the Korean population is 29, with sex-specific values of 31 for males and 18 for females [21]. Similarly, magnetic resonance imaging suggests an optimal FLI cutoff of ≥37.64 for diagnosing NAFLD in the Chinese Han population [22]. These variations underscore the need for customized diagnostic thresholds to enhance the predictive accuracy of HCC risk assessment, separate from the diagnosis of NAFLD.
Furthermore, our findings support the concept of the ‘metabolic memory’ seen in diabetes management, where the benefits of improved metabolic control may take time to manifest [23]. This delayed response could be due to the time required to reverse underlying pathophysiological changes after the fatty liver condition improves, highlighting the importance of early intervention in reducing FLI and warranting further investigation.
In our study population of Korean patients with T2DM and non-cirrhotic livers, the incidence of HCC was 77.2 per 100,000 PYs among those with an FLI ≥30, corresponding to an aHR of 2.10 compared to those with an FLI <30. This correlation is supported by recent meta-analysis data, emphasizing the necessity for HCC surveillance [24]. Additionally, El-Serag et al. [25] reported a duration-response relationship, where longer diabetes duration was associated with higher HCC risk, reinforcing the need for ongoing monitoring in these patients. This is because hepatocarcinogenesis can be promoted by diabetes at the molecular level through mechanisms that include chronic inflammation and oxidative stress, which lead to cellular proliferation and genomic instability [26].
Recent data demonstrated that the rates of NAFLD progression to HCC varied from study to study according to different ethnicities and fibrosis stages. For example, in a nationwide real world United States study, the incidence rate of HCC was rare, 5 per 100,000 PYs in 751,603 patients with NAFLD without documented cirrhosis [27]. In a prospective study with 1,773 adults covering the histologic spectrum of NAFLD, the incidence of HCC was 40 per 100,000 PYs in stages F0 to F2 (no, mild, or moderate fibrosis), increasing to 340 per 100,000 PYs in stage F3 with bridging fibrosis [28]. Similarly, a Korean nationwide cohort indicated a higher HCC incidence rate in patients with non-cirrhotic NAFLD compared with those with non-NAFLD (21 per 100,000 PYs vs. 15 per 100,000 PYs) [29]. Moreover, recent publications highlight that the risk of incident HCC is 8.36 times higher in individuals with T2DM than in those without T2DM (3.68% [95% CI, 2.18 to 5.77] vs. 0.44% [95% CI, 0.11 to 1.33]) over 5 years, based on data from cohorts in the USA, Japan, and Türkiye [24]. Thus, our findings show an HCC incidence rate of 98.29 per 100,000 PYs in patients with T2DM and non-cirrhotic livers having an FLI ≥30 underscore the notably higher incidence of non-cirrhotic HCC in this population compared with those in patients without T2DM.
Moreover, current clinical guidelines do not differentiate HCC screening based on etiology, emphasizing the importance of timely detection of advanced fibrosis in patients with T2DM and non-cirrhotic livers to effectively assess HCC risk. International guidelines [8] recommend biannual ultrasonography and optional alpha-fetoprotein testing for all patients with cirrhosis and selected patients without cirrhosis but with chronic hepatitis B. However, surveillance protocols for non-cirrhotic livers are less established due to limited data, highlighting the need for practical surveillance strategies tailored to patients with T2DM at higher risk for HCC. Consequently, our study suggests that the FLI should be incorporated into the assessment protocols for this patient cohort. Unlike traditional fibrosis markers such as the NAFLD fibrosis score, fibrosis-4 (FIB-4) index, BARD score, and aspartate aminotransferase to platelet ratio index [28,29], which primarily evaluate advanced stages of liver fibrosis or non-alcoholic steatohepatitis, the FLI can be used effectively at earlier stages of NAFLD. This allows for earlier intervention and targeted surveillance that could prevent the development of HCC. While our study highlights the potential role of FLI in HCC risk stratification, a direct comparison with FIB-4 was not feasible due to the lack of platelet count data in our dataset. Future studies incorporating FLI dynamics and FIB-4 dynamics could provide further insight into their respective clinical utility in risk prediction.
Nonetheless, this study had some limitations. First, it only included Korean patients with T2DM, therefore, the results cannot be generalized to other ethnicities. Second, we defined HCC as having at least one diagnosis with a special benefit code of V193 or V194 based on the ICD-10 diagnosis codes C22, C220, C227, or C229 and cirrhosis with ICD-10 codes such as K740.x, K741. x, K742.x, K746.x, K743.x, K744.x, K745.x, K761, P788, or K717. Therefore, there is a possibility that some patients with HCC or cirrhosis were missing because of the absence of ICD-10 codes in the hospitals. However, we believe that combining the ICD-10 code for HCC with the special benefit code V193 or V194 is a stringent and reliable method for confirming the diagnosis.
In conclusion, this study demonstrates that initial assessment and dynamic monitoring of FLI in patients with T2DM and non-cirrhotic livers is feasible and clinically reliable for detecting and managing HCC risk. Given that the impact of FLI varies by demographic factors, further research is required to refine customized FLI values to improve predictive accuracy in high-risk groups, such as patients with diabetes. Establishing tailored surveillance strategies that incorporate FLI could significantly enhance early detection and improve outcomes in patients at increased risk for HCC. Additionally, interventions aimed at lowering FLI appear to have delayed protective effects, which may reflect a metabolic memory-like phenomenon. This finding underscores the importance of initiating timely intervention to improve FLI and achieve earlier protective effects against HCC.

Supplemental Table S1.

ICD-10 Codes for Comorbidities and Exclusion Criteria
enm-2024-2286-Supplemental-Table-S1.pdf

Supplemental Table S2.

Association between FLI Increase (Threshold: 30) and Risk of New-Onset HCC among Patients Diagnosed with T2DM
enm-2024-2286-Supplemental-Table-S2.pdf

Supplemental Table S3.

Association between FLI Decrease (Threshold: 30) and Risk of New-Onset HCC among Patients Diagnosed with T2DM
enm-2024-2286-Supplemental-Table-S3.pdf

Supplemental Table S4.

Association between FLI Increase (Threshold: 60) and Risk of New-Onset HCC among Patients Diagnosed with T2DM
enm-2024-2286-Supplemental-Table-S4.pdf

Supplemental Table S5.

Association between FLI Decrease (Threshold: 60) and Risk of New-Onset HCC among Patients Diagnosed with T2DM
enm-2024-2286-Supplemental-Table-S5.pdf

Supplemental Table S6.

Association between Baseline FLI and Risk of New-Onset HCC among Patients Diagnosed with T2DM
enm-2024-2286-Supplemental-Table-S6.pdf

CONFLICTS OF INTEREST

No potential conflict of interest relevant to this article was reported.

ACKNOWLEDGMENTS

This work was supported by the Basic Science Research Program through the National Research Foundation of Korea, funded by the Ministry of Education (grant numbers 2017R1D1A3B04028873 and 2022R1I1A1A01069493).

AUTHOR CONTRIBUTIONS

Conception or design: E.H.C., M.G.K., C.H.L., J.H.P. Acquisition, analysis, or interpretation of data: E.H.C., M.G.K., C.H.L., S.O., H.L., J.S.K., J.H.P. Drafting the work or revising: E.H.C., M.G.K., C.H.L., S.O., C.S., H.R.O., Y.R.P., H.L., J.S.K., J.H.P. Final approval of the manuscript: E.H.C., M.G.K., C.H.L., S.O., C.S., H.R.O., Y.R.P., H.L., J.S.K., J.H.P.

Fig. 1.
Study population flow and design. T2DM, type 2 diabetes mellitus; ALD, alcoholic liver disease; CVH, chronic viral hepatitis; AH, autoimmune hepatitis, WD, Willson disease; FLI, fatty liver index; HCC, hepatocellular carcinoma.
enm-2024-2286f1.jpg
Fig. 2.
Changes in fatty liver index (FLI) and cumulative incidence of hepatocellular carcinoma (HCC) in patients with type 2 diabetes mellitus (T2DM) and non-cirrhotic liver. (A) Cumulative incidence of HCC by an FLI threshold of 30. The graph shows the cumulative incidence of HCC over 5 years among patients with T2DM and non-cirrhotic livers. The blue line represents individuals who maintained a FLI <30, while the red line represents individuals whose FLI increased from <30 to ≥30. A statistically significant increase in HCC incidence (log-rank test, P<0.01) was observed in patients with rising FLI levels. (B) Cumulative incidence of HCC by FLI reduction from ≥30 to <30. This graph depicts the cumulative incidence of HCC among patients who initially had an FLI ≥30 and experienced a decrease to <30 (red line) compared to those who maintained an FLI ≥30 (blue line) over a 5-year period. A significant reduction in HCC risk (log-rank test, P=0.02) was noted with a decreased FLI. (C) Cumulative incidence of HCC by an FLI threshold of 60. The graph illustrates the cumulative incidence of HCC based on changes in FLI with a threshold of 60 over 5 years. The red line indicates individuals whose FLI increased from <60 to ≥60, and the blue line shows those who maintained an FLI <60. The difference was significant (log-rank test, P<0.01), but did not remain significant after multivariable adjustment (Fig. 3). (D) Cumulative incidence of HCC by FLI reduction from ≥60 to <60 (log-rank test, P=0.30).
enm-2024-2286f2.jpg
Fig. 3.
Association between fatty liver index changes and new-onset hepatocellular carcinoma in multivariable proportional hazard regression analysis. CI, confidence interval.
enm-2024-2286f2.jpg
Fig. 4.
The cumulative incidence of hepatocellular carcinoma over a 5-year period, categorized based on the fatty liver index (FLI) at baseline and follow-up (FU).
enm-2024-2286f4.jpg
Fig. 5.
Association of hepatocellular carcinoma occurrence with baseline fatty liver index (FLI).
enm-2024-2286f5.jpg
Table 1.
Baseline Characteristics of Study Population
Total Baseline FLI
 SMD
<30 30–59.9 ≥60
Total 92,761 (100.0) 49,168 (53.0) 27,520 (29.7) 16,073 (17.3)
Age 60–79 yr 60,765 (65.5) 33,132 (67.4) 18,981 (69.0) 8,652 (53.8) 0.21
Male sex 45,246 (48.8) 19,103 (38.9) 15,184 (55.2) 10,959 (68.2) 0.41
Economic status 0.05
 Middle 46,175 (49.8) 24,124 (49.1) 13,800 (50.1) 8,251 (51.3)
 Low 16,527 (17.8) 8,837 (18.0) 4,711 (17.1) 2,979 (18.5)
 High 30,059 (32.4) 16,207 (33.0) 9,009 (32.7) 4,843 (30.1)
Residential area 0.04
 Mid-size and small cities 23,805 (25.7) 12,372 (25.2) 7,195 (26.1) 4,238 (26.4)
 Rural areas 10,849 (11.7) 5,581 (11.4) 3,450 (12.5) 1,818 (11.3)
 Metropolitan cities 58,107 (62.6) 31,215 (63.5) 16,875 (61.3) 10,017 (62.3)
Drinkinga, badb 19,082 (20.6) 6,602 (13.4) 6,362 (23.1) 6,118 (38.1) 0.39
Smoking, bad 13,072 (14.1) 4,865 (9.9) 4,354 (15.8) 3,853 (24.0) 0.26
PA, bad 33,776 (36.4) 16,642 (33.8) 10,545 (38.3) 6,589 (41.0) 0.10
HTN, yes 36,773 (39.6) 17,012 (34.6) 12,110 (44.0) 7,651 (47.6) 0.18
Hyperlipidemia, yes 11,229 (12.1) 5,927 (12.1) 3,299 (12.0) 2,003 (12.5) 0.01
Metformin, yes 48,066 (51.8) 22,683 (46.1) 15,487 (56.3) 9,896 (61.6) 0.21
Statin, yes 50,715 (54.7) 24,775 (50.4) 16,179 (58.8) 9,761 (60.7) 0.14
SGLT2 inhibitor, yes 1,320 (1.4) 496 (1.0) 407 (1.5) 417 (2.6) 0.08

Values are expressed as number (%).

FLI, fatty liver index; SMD, standardized mean difference; Bad, lifestyle habits that need improvement; PA, physical activity; HTN, hypertension; SGLT2, sodium-glucose cotransporter 2.

a Alcohol consumption;

b Life habits were categorized based on the physician’s judgment into whether improvement was needed (bad) for drinking, smoking, and PA, or if the individual’s habits were considered normal (good) (normal life habits).

  • 1. Huang DQ, El-Serag HB, Loomba R. Global epidemiology of NAFLD-related HCC: trends, predictions, risk factors and prevention. Nat Rev Gastroenterol Hepatol 2021;18:223–38.ArticlePubMedPMCPDF
  • 2. Younossi ZM, Otgonsuren M, Henry L, Venkatesan C, Mishra A, Erario M, et al. Association of nonalcoholic fatty liver disease (NAFLD) with hepatocellular carcinoma (HCC) in the United States from 2004 to 2009. Hepatology 2015;62:1723–30.ArticlePubMedPDF
  • 3. Stine JG, Wentworth BJ, Zimmet A, Rinella ME, Loomba R, Caldwell SH, et al. Systematic review with meta-analysis: risk of hepatocellular carcinoma in non-alcoholic steatohepatitis without cirrhosis compared to other liver diseases. Aliment Pharmacol Ther 2018;48:696–703.ArticlePubMedPMCPDF
  • 4. Foerster F, Gairing SJ, Muller L, Galle PR. NAFLD-driven HCC: safety and efficacy of current and emerging treatment options. J Hepatol 2022;76:446–57.ArticlePubMed
  • 5. Younossi ZM, Golabi P, de Avila L, Paik JM, Srishord M, Fukui N, et al. The global epidemiology of NAFLD and NASH in patients with type 2 diabetes: a systematic review and meta-analysis. J Hepatol 2019;71:793–801.ArticlePubMed
  • 6. Yang JD, Ahmed F, Mara KC, Addissie BD, Allen AM, Gores GJ, et al. Diabetes is associated with increased risk of hepatocellular carcinoma in patients with cirrhosis from nonalcoholic fatty liver disease. Hepatology 2020;71:907–16.ArticlePubMedPDF
  • 7. El-Serag HB, Hampel H, Javadi F. The association between diabetes and hepatocellular carcinoma: a systematic review of epidemiologic evidence. Clin Gastroenterol Hepatol 2006;4:369–80.ArticlePubMed
  • 8. Loomba R, Lim JK, Patton H, El-Serag HB. AGA clinical practice update on screening and surveillance for hepatocellular carcinoma in patients with nonalcoholic fatty liver disease: expert review. Gastroenterology 2020;158:1822–30.ArticlePubMed
  • 9. Reeves HL, Zaki MY, Day CP. Hepatocellular carcinoma in obesity, type 2 diabetes, and NAFLD. Dig Dis Sci 2016;61:1234–45.ArticlePubMedPDF
  • 10. Otgonsuren M, Estep MJ, Hossain N, Younossi E, Frost S, Henry L, et al. Single non-invasive model to diagnose non-alcoholic fatty liver disease (NAFLD) and non-alcoholic steatohepatitis (NASH). J Gastroenterol Hepatol 2014;29:2006–13.PubMed
  • 11. 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.ArticlePubMedPMCPDF
  • 12. Kang MG, Lee CH, Shen C, Kim JS, Park JH. Longitudinal changes in fatty liver index are associated with risk of hepatocellular carcinoma: a nationwide cohort study in Korea. J Hepatol 2024;80:e216–8.ArticlePubMed
  • 13. Shin DW, Cho J, Park JH, Cho B. National General Health Screening Program in Korea: history, current status, and future direction. Precis Future Med 2022;6:9–31.ArticlePDF
  • 14. Kim MK, Han K, Lee SH. Current trends of big data research using the Korean National Health Information Database. Diabetes Metab J 2022;46:552–63.ArticlePubMedPMCPDF
  • 15. Benhammou JN, Leng M, Shah SC, Cholankeril G, Dong TS, Patel AA, et al. Exposure to agent orange and hepatocellular carcinoma among US military personnel. JAMA Netw Open 2023;6:e2346380.ArticlePubMedPMC
  • 16. Jain A, Mazer B, Deng Y, Ciarleglio M, Jain D, Taddei T, et al. Hepatocellular carcinoma: does the background liver with or without cirrhosis matter? Am J Clin Pathol 2022;157:305–13.PubMed
  • 17. Liu Y, Li H, Ye N, Luo CJ, Hu YY, Wu H, et al. Non-cirrhotic liver is associated with poor prognosis of hepatocellular carcinoma: a literature review. Med Sci Monit 2019;25:6615–23.ArticlePubMedPMC
  • 18. Gaddikeri S, McNeeley MF, Wang CL, Bhargava P, Dighe MK, Yeh MM, et al. Hepatocellular carcinoma in the noncirrhotic liver. AJR Am J Roentgenol 2014;203:W34–47.ArticlePubMed
  • 19. Cespiati A, Cinque F, Meroni M, Lombardi R, Dongiovanni P, Fracanzani AL. An overview of hepatocellular carcinoma surveillance focusing on non-cirrhotic NAFLD patients: a challenge for physicians. Biomedicines 2023;11:586.ArticlePubMedPMC
  • 20. Lajeunesse-Trempe F, Boit MK, Kaduka LU, De Lucia-Rolfe E, Baass A, Paquette M, et al. Validation of the fatty liver index for identifying non-alcoholic fatty liver disease in a Kenyan population. Trop Med Int Health 2023;28:830–8.PubMed
  • 21. Cho EJ, Jung GC, Kwak MS, Yang JI, Yim JY, Yu SJ, et al. Fatty liver index for predicting nonalcoholic fatty liver disease in an asymptomatic Korean population. Diagnostics (Basel) 2021;11:2233.ArticlePubMedPMC
  • 22. Zhang Q, Zhu Y, Yu W, Xu Z, Zhao Z, Liu S, et al. Diagnostic accuracy assessment of molecular prediction model for the risk of NAFLD based on MRI-PDFF diagnosed Chinese Han population. BMC Gastroenterol 2021;21:88.ArticlePubMedPMCPDF
  • 23. Kato M, Natarajan R. Epigenetics and epigenomics in diabetic kidney disease and metabolic memory. Nat Rev Nephrol 2019;15:327–45.ArticlePubMedPMCPDF
  • 24. Huang DQ, Noureddin N, Ajmera V, Amangurbanova M, Bettencourt R, Truong E, et al. Type 2 diabetes, hepatic decompensation, and hepatocellular carcinoma in patients with non-alcoholic fatty liver disease: an individual participant-level data meta-analysis. Lancet Gastroenterol Hepatol 2023;8:829–36.ArticlePubMedPMC
  • 25. El-Serag HB, Tran T, Everhart JE. Diabetes increases the risk of chronic liver disease and hepatocellular carcinoma. Gastroenterology 2004;126:460–8.ArticlePubMed
  • 26. Teng PC, Huang DQ, Lin TY, Noureddin M, Yang JD. Diabetes and risk of hepatocellular carcinoma in cirrhosis patients with nonalcoholic fatty liver disease. Gut Liver 2023;17:24–33.ArticlePubMed
  • 27. Huang DQ, Tran S, Barnett S, Zou B, Yeo YH, Cheung R, et al. Incidence and predictors of hepatocellular carcinoma in NAFLD without diagnosed cirrhosis: a nationwide real-world U.S. study. Hepatol Int 2024;18:540–9.ArticlePubMedPDF
  • 28. Stal P. Liver fibrosis in non-alcoholic fatty liver disease: diagnostic challenge with prognostic significance. World J Gastroenterol 2015;21:11077–87.ArticlePubMedPMC
  • 29. Contreras D, Gonzalez-Rocha A, Clark P, Barquera S, Denova-Gutierrez E. Diagnostic accuracy of blood biomarkers and non-invasive scores for the diagnosis of NAFLD and NASH: systematic review and meta-analysis. Ann Hepatol 2023;28:100873.ArticlePubMed

Figure & Data

References

    Citations

    Citations to this article as recorded by  

      • PubReader PubReader
      • ePub LinkePub Link
      • Cite
        Cite
        export Copy Download
        Close
        Download Citation
        Download a citation file in RIS format that can be imported by all major citation management software, including EndNote, ProCite, RefWorks, and Reference Manager.
        Format:
        • RIS — For EndNote, ProCite, RefWorks, and most other reference management software
        • BibTeX — For JabRef, BibDesk, and other BibTeX-specific software
        Include:
        • Citation for the content below
        Fatty Liver Index Dynamics as a Predictor of Hepatocellular Carcinoma in Patients with Type 2 Diabetes Mellitus and Non-Cirrhotic Livers
        Close
      • XML DownloadXML Download
      Figure
      • 0
      • 1
      • 2
      • 3
      • 4
      Related articles
      Fatty Liver Index Dynamics as a Predictor of Hepatocellular Carcinoma in Patients with Type 2 Diabetes Mellitus and Non-Cirrhotic Livers
      Image Image Image Image Image
      Fig. 1. Study population flow and design. T2DM, type 2 diabetes mellitus; ALD, alcoholic liver disease; CVH, chronic viral hepatitis; AH, autoimmune hepatitis, WD, Willson disease; FLI, fatty liver index; HCC, hepatocellular carcinoma.
      Fig. 2. Changes in fatty liver index (FLI) and cumulative incidence of hepatocellular carcinoma (HCC) in patients with type 2 diabetes mellitus (T2DM) and non-cirrhotic liver. (A) Cumulative incidence of HCC by an FLI threshold of 30. The graph shows the cumulative incidence of HCC over 5 years among patients with T2DM and non-cirrhotic livers. The blue line represents individuals who maintained a FLI <30, while the red line represents individuals whose FLI increased from <30 to ≥30. A statistically significant increase in HCC incidence (log-rank test, P<0.01) was observed in patients with rising FLI levels. (B) Cumulative incidence of HCC by FLI reduction from ≥30 to <30. This graph depicts the cumulative incidence of HCC among patients who initially had an FLI ≥30 and experienced a decrease to <30 (red line) compared to those who maintained an FLI ≥30 (blue line) over a 5-year period. A significant reduction in HCC risk (log-rank test, P=0.02) was noted with a decreased FLI. (C) Cumulative incidence of HCC by an FLI threshold of 60. The graph illustrates the cumulative incidence of HCC based on changes in FLI with a threshold of 60 over 5 years. The red line indicates individuals whose FLI increased from <60 to ≥60, and the blue line shows those who maintained an FLI <60. The difference was significant (log-rank test, P<0.01), but did not remain significant after multivariable adjustment (Fig. 3). (D) Cumulative incidence of HCC by FLI reduction from ≥60 to <60 (log-rank test, P=0.30).
      Fig. 3. Association between fatty liver index changes and new-onset hepatocellular carcinoma in multivariable proportional hazard regression analysis. CI, confidence interval.
      Fig. 4. The cumulative incidence of hepatocellular carcinoma over a 5-year period, categorized based on the fatty liver index (FLI) at baseline and follow-up (FU).
      Fig. 5. Association of hepatocellular carcinoma occurrence with baseline fatty liver index (FLI).

      Fig. 1.

      Fig. 2.

      Fig. 3.

      Fig. 4.

      Fig. 5.

      Fatty Liver Index Dynamics as a Predictor of Hepatocellular Carcinoma in Patients with Type 2 Diabetes Mellitus and Non-Cirrhotic Livers
      Total Baseline FLI
      		 SMD
      <30 30–59.9 ≥60
      Total 92,761 (100.0) 49,168 (53.0) 27,520 (29.7) 16,073 (17.3)
      Age 60–79 yr 60,765 (65.5) 33,132 (67.4) 18,981 (69.0) 8,652 (53.8) 0.21
      Male sex 45,246 (48.8) 19,103 (38.9) 15,184 (55.2) 10,959 (68.2) 0.41
      Economic status 0.05
       Middle 46,175 (49.8) 24,124 (49.1) 13,800 (50.1) 8,251 (51.3)
       Low 16,527 (17.8) 8,837 (18.0) 4,711 (17.1) 2,979 (18.5)
       High 30,059 (32.4) 16,207 (33.0) 9,009 (32.7) 4,843 (30.1)
      Residential area 0.04
       Mid-size and small cities 23,805 (25.7) 12,372 (25.2) 7,195 (26.1) 4,238 (26.4)
       Rural areas 10,849 (11.7) 5,581 (11.4) 3,450 (12.5) 1,818 (11.3)
       Metropolitan cities 58,107 (62.6) 31,215 (63.5) 16,875 (61.3) 10,017 (62.3)
      Drinkinga, badb 19,082 (20.6) 6,602 (13.4) 6,362 (23.1) 6,118 (38.1) 0.39
      Smoking, bad 13,072 (14.1) 4,865 (9.9) 4,354 (15.8) 3,853 (24.0) 0.26
      PA, bad 33,776 (36.4) 16,642 (33.8) 10,545 (38.3) 6,589 (41.0) 0.10
      HTN, yes 36,773 (39.6) 17,012 (34.6) 12,110 (44.0) 7,651 (47.6) 0.18
      Hyperlipidemia, yes 11,229 (12.1) 5,927 (12.1) 3,299 (12.0) 2,003 (12.5) 0.01
      Metformin, yes 48,066 (51.8) 22,683 (46.1) 15,487 (56.3) 9,896 (61.6) 0.21
      Statin, yes 50,715 (54.7) 24,775 (50.4) 16,179 (58.8) 9,761 (60.7) 0.14
      SGLT2 inhibitor, yes 1,320 (1.4) 496 (1.0) 407 (1.5) 417 (2.6) 0.08
      Table 1. Baseline Characteristics of Study Population

      Values are expressed as number (%).

      FLI, fatty liver index; SMD, standardized mean difference; Bad, lifestyle habits that need improvement; PA, physical activity; HTN, hypertension; SGLT2, sodium-glucose cotransporter 2.

      Alcohol consumption;

      Life habits were categorized based on the physician’s judgment into whether improvement was needed (bad) for drinking, smoking, and PA, or if the individual’s habits were considered normal (good) (normal life habits).

      Table 1.

      Endocrinol Metab : Endocrinology and Metabolism
      © Korean Endocrine Society.
      TOP