ABSTRACT
-
Background/Aims
Previous studies suggest that hypothyroidism is associated with metabolic dysfunction-associated steatotic liver disease (MASLD) and its histological severity, but clinical outcome data are largely lacking. We aimed to study the impact of hypothyroidism on liver-related events (LREs).
-
Methods
Patients with MASLD were identified from a territory-wide registry in Hong Kong during 2000–2024. Thyroid status was determined using diagnosis codes and thyroid function tests. The primary outcome, LRE, was defined as a composite of hepatic decompensation, hepatocellular carcinoma, liver transplantation, and liver-related death.
-
Results
A total of 20,478 patients with MASLD were included in the final analysis (mean age 56.4±13.2 years; 43.9% male). At baseline, 18,178 (88.8%) patients were euthyroid, 598 (2.9%) were hyperthyroid, and 1,702 (8.3%) were hypothyroid. Compared with euthyroid patients, both hyperthyroidism and overt hypothyroidism were associated with cirrhosis. At a median follow-up of 4.8 years, 179 patients developed LREs, and 26 died from liver disease. Compared with patients with normal serum thyroid-stimulating hormone (TSH) levels of 0.4–4 mIU/L, those with subclinical (4–10 mIU/L; adjusted time-dependent cause-specific hazard ratio [aCSHR], 2.49; 95% CI, 1.51–4.13) and overt hypothyroidism (>10 mIU/L; aCSHR, 4.91; 95% CI, 1.56–15.47) had an increased risk of LREs. Time-dependent, but not baseline, TSH and thyroid status were associated with LRE risk.
-
Conclusions
Subclinical and overt hypothyroidism are associated with an increased risk of LREs in a dose-dependent manner. The association with time-dependent but not baseline thyroid status underscores the importance of thyroid monitoring and suggests that correction of hypothyroidism may mitigate LRE risk.
-
Keywords: Thyroid function test; Subclinical hypothyroidism; Fatty liver; Cirrhosis; Hepatocellular carcinoma
Study Highlights
• Hypothyroidism is associated with increased risk of LREs in patients with MASLD.
• Thyroid function can change over time due to treatment and fluctuating disease status.
• The findings suggest that dynamic thyroid function assessment may provide better prognostic value in predicting LRE risk and support timely intervention to reduce adverse liver outcomes.
Graphical Abstract
INTRODUCTION
Metabolic dysfunction-associated steatotic liver disease (MASLD) affects approximately 30% of people worldwide and has emerged as the most common cause of chronic liver disease [
1]. It represents a spectrum from simple steatosis to metabolic dysfunction-associated steatohepatitis (MASH), a more severe form with a higher risk of cirrhosis, hepatocellular carcinoma (HCC), and liver-related mortality [
2]. In the United States, MASH is currently the second leading indication for liver transplantation [
3]. The underlying pathophysiology of MASLD involves insulin resistance, excessive nutrient intake, and dysregulated glucose and lipotoxicity.
Thyroid hormones are key regulators in hepatic glucose and lipid metabolism by promoting fatty acid oxidation, cholesterol clearance, and mitochondrial activity [
4,
5]. In hepatocytes, deiodinase 1 (DIO1) converts thyroxine (T4) to the biologically active triiodothyronine (T3), which mediates through binding thyroid hormone receptor-beta (THR-β) to regulate transcription of metabolic target genes [
5]. Hepatic hypothyroidism may result from thyroid hormone deficiency, elevated thyroid-stimulating hormone (TSH), impaired T4-to-T3 conversion due to reduced DIO1 and increased deiodinase 3 (DIO3) expression, and downregulation of THR-β [
4,
5]. These alterations impair intrahepatic thyroid hormone signalling and contribute to oxidative stress, lipid accumulation and systematic inflammation; overlap with mechanisms driving MASLD progression and hepatocarcinogenesis [
4,
5]. Therapeutically, T4 supplementation may improve lipid metabolism [
6,
7], while liver-targeted THR-β agonist, resmetirom, the first approved drug for noncirrhotic MASH, shows promise in restoring intrahepatic thyroid signalling and reducing hepatic steatosis, offering a targeted approach to MASLD treatment [
5].
Furthermore, recent evidence has established a significant association between hypothyroidism and the presence of MASLD and its severity [
8-
10]. However, sparse data exist regarding its role in the development of liver-related events (LREs). Specifically, while some cross-sectional and case-control studies have explored the link between hypothyroidism and cirrhosis or HCC [
11-
14], longitudinal studies evaluating hypothyroidism and the incidence of LREs are lacking.
In this study, we aimed to investigate the association between hypothyroidism and the risk of HCC and cirrhotic complications in a retrospective, territory-wide cohort of patients with MASLD.
MATERIALS AND METHODS
Study design and data source
This is a territory-wide retrospective cohort study based on data retrieved from the Clinical Data Analysis and Reporting System (CDARS) under the Hospital Authority, Hong Kong, China. CDARS is an electronic healthcare database that covers patients’ demographics, death, diagnoses, procedures, drug prescription and dispensing history, and laboratory results from all public hospitals and clinics in Hong Kong, representing around 80% of the local population [
15]. Diagnoses are coded using the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) in CDARS. When validated against a review of consultation notes and radiological and laboratory reports, the positive predictive value for nonalcoholic fatty liver disease (NAFLD) by ICD-9-CM diagnosis coding was 98.4%, and cirrhotic complications had a positive predictive value of 85.7% and a negative predictive value of 95.9% [
16]. Furthermore, all patients with NAFLD confirmed by chart review had at least one cardiometabolic risk factor, suggesting that they meet the updated MASLD diagnostic criteria.
All patients aged 18 years or older with MASLD (ICD-9-CM diagnosis code 571.8) [
15], first diagnosed between January 1, 2000, and September 22, 2024, in Hong Kong, were identified. Patients were excluded if they were younger than 18 years old at baseline; lack of TSH and/or T4 measurements; were infected with hepatitis B virus (HBV) and/or hepatitis C virus (HCV) based on ICD-9-CM diagnosis codes, viral and serological markers, and/or use of antiviral therapy for HBV and/or HCV at any time during follow-up; did not perform HBV serology tests during follow-up; were infected with human immunodeficiency virus (HIV) based on ICD-9-CM diagnosis codes and/or use of antiviral therapy for HIV at any time during follow-up; had excessive use of alcohol based on the nursing assessment form or ICD-9-CM diagnosis codes; or had other hepatobiliary diseases based on ICD-9-CM diagnosis codes; had HCC and/or cirrhotic complications before baseline or within 3 months from baseline; had thyroid cancer or other malignancy before baseline or within 3 months from baseline; or had autoimmune hepatitis and primary biliary cholangitis based on ICD-9-CM diagnosis code (
Fig. 1,
Supplementary Table 1).
Patients with hypothyroidism were defined by exposure to T4 replacement treatment, thyroid function tests, or ICD-9-CM diagnosis codes (
Supplementary Table 1). Patients with hyperthyroidism were defined by thyroid function tests or ICD-9-CM diagnosis code. To avoid immortal time bias, the baseline date was defined as the second chronological event among the first diagnosis of MASLD or the first thyroid function test. Patients were followed until the diagnosis of HCC or cirrhotic complications, death, or last follow-up date (September 22, 2024), whichever came first. The study protocol conformed to the ethical guidelines of the 1975 Declaration of Helsinki as reflected in a priori approval by the Joint Chinese University of Hong Kong-New Territories East Cluster Clinical Research Ethics Committee (Ref. no.: 2020.359). Informed consent was waived because it was a registry study using deidentified data.
At baseline and during follow-up, patients’ demographic, biochemical, relevant diagnosis and procedures and medications were recorded. The calculation formula of body mass index (BMI) and definitions of hypertension and diabetes mellitus (DM) are shown in
Supplementary Methods.
Thyroid function was evaluated at baseline and throughout the follow-up period using serum levels of TSH, free T4, and free T3. Additionally, thyroglobulin, antithyroglobulin antibody (anti-TG), and anti-thyroid peroxidase antibody (anti-TPO) results were captured. The reference range for TSH is 0.4–4.0 mIU/L [
17]. A TSH level between 4.0 and 10.0 mIU/L is considered mildly elevated, whereas a level exceeding 10.0 mIU/L is classified as severely increased [
18]. The normal range of free T4 and free T3 levels was based on laboratory normal ranges. Thyroid status at baseline was determined using a combination of thyroid function test and relevant ICD-9-CM diagnosis codes (
Supplementary Tables 1,
2). During the follow-up period, thyroid status was assessed through subsequent thyroid function tests. When thyroid status was defined based on thyroid function tests, euthyroidism was defined as TSH and free T4 values within the reference range. Hyperthyroidism was identified by a decreased TSH level (<0.4 mIU/L) and/or elevated free T4 levels. Hypothyroidism was defined as an increased TSH (>4.0 mIU/L) level with a normal or decreased free T4 level. It was further classified as subclinical or overt hypothyroidism. Subclinical hypothyroidism was defined as a TSH level between 4.0 mIU/L and 10.0 mIU/L with a normal free T4 level. The diagnosis of overt hypothyroidism was based on use of T4 replacement therapy, TSH more than 4.0 mIU/L with a low free T4 level, or TSH more than 10.0 mIU/L.
To examine the association between T4 replacement therapy and the risk of LREs among patients with MASLD, a subgroup analysis was performed, including only those with hypothyroidism at baseline. Patients were classified as T4 users at baseline if they received T4 replacement therapy on or before the baseline date. Those who never received T4 replacement therapy or received it during follow-up were classified as nonusers at baseline.
Outcome
The primary outcome was LREs, a composite outcome consisting of HCC and cirrhotic complications. HCC was identified based on ICD-9-CM diagnosis codes and procedure codes; cirrhotic complications were determined using ICD-9-CM diagnosis and procedure codes, including the development of ascites, spontaneous bacterial peritonitis, variceal bleeding, hepatic encephalopathy, hepatorenal syndrome, and liver transplantation, as well as liver-related mortality (
Supplementary Table 1). Liver-related death was identified as International Classification of Diseases, Tenth Revision, diagnosis codes C22, and K70-K77. The definitions of cirrhosis are shown in
Supplementary Methods.
Continuous variables were expressed as mean±standard deviation (SD) or median (interquartile range [IQR]), and categorical variables in the form of n (%). Comparison of characteristics between subgroups was conducted using the Chi-squared test or Fisher’s exact test for categorical variables, and the ANOVA test or Kruskal–Wallis test for continuous variables.
The primary analysis compared hypothyroidism and hyperthyroidism with euthyroidism, with a subgroup analysis further stratifying hypothyroidism into subclinical and overt categories.
To assess the association between baseline thyroid status and the presence of baseline cirrhosis, unadjusted and adjusted logistic regression models were used to estimate odds ratios (ORs) and 95% confidence intervals (CIs).
The cumulative incidence function (CIF) of LREs was performed by using Aalen-Johansen method. Gray’s test, accounting for the competing risk of death, was applied to test for differences between the CIFs. Univariate and multivariable Fine-Gray subdistribution hazard regression models with adjustment of competing risk of death were used to estimate subdistribution hazard ratios (SHRs) and 95% CIs for the association between baseline thyroid status and the risk of LREs. The proportional subdistribution hazards assumption for the Fine-Gray subdistribution hazard models was assessed using modified weighted Schoenfeld residuals, and no significant violation was observed. Baseline TSH and free T4 levels, as continuous variables or categorised according to national guidelines and laboratory reference ranges, were also analysed using Fine-Gray models to estimate SHRs.
Time-dependent analyses were conducted to further explore the dynamic relationship between thyroid function and LRE risk during follow-up. We performed unadjusted and adjusted time-dependent cause-specific hazard models to examine the association between time-varying thyroid status and thyroid function (TSH and free T4) and the risk of LREs using cause-specific hazard ratio (CSHR) and 95% CIs. The CIF of LREs during follow-up was performed by using Simon-Makuch plot, which accounts for time-varying exposure. Mantel-Byar test was used to test for differences between the curves.
All cross-sectional and longitudinal multivariable models were adjusted for baseline age, sex, BMI, waist circumference, alanine aminotransferase (ALT), platelets, cirrhosis (only for longitudinal analysis), DM, hypertension, total cholesterol, high-density lipoprotein (HDL) cholesterol, low-density lipoprotein (LDL) cholesterol, and triglyceride levels.
Sankey diagrams were used to visualise longitudinal transitions in thyroid status over time, as determined by thyroid function tests. Patients were classified into four thyroid status categories: hyperthyroidism, euthyroidism, subclinical hypothyroidism, and overt hypothyroidism. The diagrams illustrate transitions in thyroid status from baseline through years 1, 3, and 5, up until the occurrence of LREs or death.
To assess the robustness of our findings, we performed additional sensitivity analyses. A censoring analysis was conducted to minimize the misclassification due to outdated thyroid function tests results. Specifically, follow-up was censored at 1 year after the last available thyroid function test if the interval between the last thyroid function test and the end of the follow-up date exceeded 1 year, and events occurring beyond this censoring date were treated as censored. Moreover, we conducted 1-year and 2-year landmark analysis by censoring events that occurred within the first 1 year or 2 years during the follow-up to minimize reverse causation bias. Other detailed sensitivity and subgroup analyses were presented in
Supplementary Methods.
Multiple imputation by chained equations was applied to account for missing covariates, generating five imputed datasets using the “mice” package in R. All statistical tests were two-tailed, and P-value <0.05 was considered statistically significant. Data analyses were performed using R software (version 4.4.1, R Core Team 2024).
RESULTS
Baseline characteristics
We included a total of 20,478 eligible patients (
Fig. 1). The mean age was 56.4±13.2 years, 8,990 (43.9%) patients were male, and 383 (1.9%) patients had cirrhosis (
Table 1). At baseline, 1,702 (8.3%) had hypothyroidism and 598 (2.9%) had hyperthyroidism at baseline. Among those with hypothyroidism, 1,133 (66.6%) patients were overt hypothyroid, and 493 (29.0%) patients were subclinical hypothyroid. Compared with patients with euthyroidism, patients with hypothyroidism were older, less likely to be male, more likely to have hypertension, and had higher BMI, LDL, triglycerides and aspartate aminotransferase, but had lower ALT. Over a median follow-up duration of 4.8 years (IQR, 3.9–10.4), 179 patients (0.9%) developed LREs. Among these, 36 patients (0.18%) had HCC. The most frequently observed cirrhotic complication during follow-up was ascites (124, 0.61%), followed by liver-related mortality (26, 0.13%), variceal bleeding (10, 0.05%), hepatic encephalopathy (7, 0.03%), and spontaneous bacterial peritonitis (3, 0.01%) (
Supplementary Table 3).
Both hypothyroidism (adjusted OR, 1.44; 95% CI, 1.05–1.99;
P=0.025) and hyperthyroidism (adjusted OR, 1.67; 95% CI, 1.04–2.69;
P=0.034) were associated with a higher risk of cirrhosis, compared with euthyroidism (
Table 2). When hypothyroidism was stratified by severity, overt hypothyroidism remained significantly associated with cirrhosis (adjusted OR, 1.52; 95% CI, 1.04–2.21;
P=0.029), whereas subclinical hypothyroidism was not (adjusted OR, 1.29; 95% CI, 0.71–2.35;
P=0.401).
The incidence of LREs did not differ significantly across baseline thyroid status (Gray’s test,
P=0.107) (
Supplementary Fig. 1A), nor among the hypothyroidism subgroups (Gray’s test,
P=0.098) (
Supplementary Fig. 1B). Patients with baseline hypothyroidism had a significantly higher risk of developing LREs compared to those with euthyroidism (SHR, 1.59; 95% CI, 1.04–2.43;
P=0.032) in the univariate model (
Supplementary Table 4). After classification into subtypes, overt hypothyroidism was associated with a 1.67-fold increased risk of incident LREs (SHR, 1.67; 95% CI, 1.02–2.71,
P=0.040). However, these associations were attenuated after adjustment (hypothyroidism: adjusted SHR, 1.43; 95% CI, 0.93–2.22;
P=0.100, overt hypothyroidism: adjusted SHR, 1.54; 95% CI, 0.93–2.55;
P=0.091).
The cumulative incidence of LREs showed no statistically significant difference across the three baseline TSH categories (Gray’s test,
P=0.991) (
Supplementary Fig. 2).
Supplementary Table 5 shows that lower TSH levels and higher TSH levels were not risk factors for developing LREs (<0.4 mIU/L: adjusted SHR, 0.86; 95% CI, 0.38–1.94;
P=0.710; >4.0 mIU/L: adjusted SHR, 1.02; 95% CI, 0.50–2.10;
P=0.950). Similarly, baseline free T4 categories (categorised as low, normal, or high based on laboratory assays) showed no significant differences in LRE incidence (Gray’s test,
P=0.841;
Supplementary Fig. 3). Compared with normal free T4, neither low nor high free T4 level groups were significantly associated with incident LREs (low: adjusted SHR, 1.53; 95% CI, 0.45–5.15;
P=0.490; high: adjusted SHR, 1.03; 95% CI, 0.32–3.25;
P=0.960) (
Supplementary Table 6). When analysed as continuous variables, baseline TSH and free T4 levels were also not associated with risk of LRE (baseline TSH: adjusted SHR, 1.00; 95% CI, 0.97–1.02;
P=0.880; baseline free T4: adjusted SHR, 0.96; 95% CI, 0.89–1.03;
P=0.250) (
Supplementary Table 7).
However, when TSH was treated as a time-varying covariate, higher TSH levels were significantly associated with an increased risk of LREs (adjusted CSHR, 1.02; 95% CI, 1.00–1.04;
P=0.014). In contrast, free T4 levels were inversely associated with the risk of LREs, although the association did not reach statistical significance (adjusted CSHR, 0.97; 95% CI, 0.91–1.04;
P=0.438) (
Supplementary Table 7). Given the observed association between higher time-varying TSH levels and increased LRE risk, we subsequently categorised TSH during follow-up into normal (0.4–4.0 mIU/L, reference), high (4–10 mIU/L), and very high (>10 mIU/L) time-varying groups to further assess risk stratification. Compared with normal TSH levels, patients with high TSH level had a 2.49-fold increased risk of developing LREs (adjusted CSHR, 2.49; 95% CI, 1.51–4.13;
P<0.001), and those with very high TSH level had an even greater risk (adjusted CSHR, 4.91; 95% CI, 1.56–15.47;
P=0.006) (
Table 3).
Next, to further explore the impact of dynamic changes in thyroid status, we also created multiple Sankey plots and conducted a time-dependent analysis. Over the first 5 years of follow-up, transitions in thyroid status occurred in 265 of 450 patients (58.9%) with subclinical hypothyroidism, of whom 247 became euthyroid. Among 157 patients with overt hypothyroidism, 118 (75.2%) transitioned, with 93 becoming euthyroid. Among 515 patients with hyperthyroidism, 296 (57.5%) transitioned, and 270 became euthyroid. Most patients (88.1%) who were euthyroid at baseline remained euthyroid. Overall, 409 patients (4.3%) died, and 36 patients (0.4%) developed LREs during the follow-up (
Fig. 2). Similar transition patterns were observed at 1-year and 3-year follow-up, as shown in
Supplementary Figure 4A and
4B, respectively. Compared with euthyroidism, time-varying hypothyroidism was significantly associated with an increased risk of LREs (adjusted CSHR, 3.19; 95% CI, 2.06–4.94;
P<0.001), with both subclinical (adjusted CSHR, 2.58; 95% CI, 1.56–4.29;
P<0.001) and overt hypothyroidism (adjusted CSHR, 6.41; 95% CI, 2.96–13.85;
P<0.001) independently conferring elevated risks. However, time-varying hyperthyroidism showed no association with LREs (adjusted CSHR, 1.21; 95% CI, 0.59–2.48;
P=0.599) (
Fig. 3 and
Table 3).
All sensitivity analyses yielded consistent results (
Supplementary Results,
Supplementary Tables 9–
22). In particular, the association between time-varying hypothyroidism and risk of LRE remained significant when applying the censoring approach for outdated thyroid function test results (
Table 4). Similarly, the 1-year and 2-year landmark analyses also showed that the results remained unchanged after censoring LREs occurring within the first 1 or 2 years of follow-up (
Table 5 and
Supplementary Tables 16,
17).
DISCUSSION
In this territory-wide retrospective cohort study, baseline overt hypothyroidism was significantly associated with an increased risk of cirrhosis in patients with MASLD. In contrast, baseline thyroid function was not associated with LRE risk. When thyroid status was assessed as a time-varying exposure, both subclinical and overt hypothyroidism were significantly associated with increased LRE risk, and higher time-varying TSH levels independently predicted incident LREs. T4 therapy was associated with a numerically lower risk of LREs.
Previous studies have reported that hypothyroidism is associated with MASLD and its severity and progression. Numerous cross-sectional and longitudinal studies have shown that thyroid dysfunction, including both overt and subclinical hypothyroidism, is associated with an increased risk of MASLD [
9,
19]. Additionally, liver biopsy data suggest that hypothyroidism is linked to more severe histological forms of MASLD, including steatohepatitis and advanced fibrosis [
9,
19-
21]. Some studies have also reported the presence of hypothyroidism in cirrhosis [
13,
22] and an increased risk of HCC among patients with hypothyroidism, potentially mediated through MASLD-related pathways [
11,
12]. Meanwhile, MASLD is increasingly recognised as a progressive liver disease that can lead to cirrhosis, HCC, and other liver-related complications, including hepatic decompensation and liver-related mortality. In this context, approximately 20% of MASLD patients progress to MASH, of whom another 20% may progress to cirrhosis [
23]. The rate of progression averages 14 years per fibrosis stage in MASLD [
23], and fibrosis stage and steatohepatitis are the strongest predictors of LREs [
24,
25]. Our study builds upon these findings by demonstrating a longitudinal association between hypothyroidism and risk of LREs, which represent clinical consequences of MASLD progression.
The interplay between hypothyroidism and MASLD outcomes is multifaceted, reflecting overlapping and sustained pathophysiological mechanisms. Hypothyroidism contributes to hepatic steatosis by impaired lipid metabolism and worsened insulin resistance. Specifically, increased TSH levels upregulate sterol regulatory element binding protein-1c activity in hepatocyte, decrease thyroid hormones and increase de novo lipogenesis [
4]. The accumulation of free fatty acids (FFAs) overloads mitochondria and peroxisomes (in endoplasmic reticulum [ER]) to cause oxidative stress and generate reactive oxygen species (ROS) and toxic intermediates [
26]. To mitigate lipotoxicity, hepatic FFAs are re-esterified into triglycerides, assembled into very low-density lipoproteins or stored in lipid droplets within hepatocytes [
27]. While acute FFA overload is reduced, sustained accumulation still leads to steatosis. Meanwhile, in response to liver injury, hepatocytes and stromal cells alter deiodinase expression by increasing DIO3 and decreasing DIO1 activity, leading to enhanced inactivation of thyroid hormones and reduced intrahepatic thyroid hormone signalling [
28,
29]. Furthermore, thyroid hormone is essential for fatty acid catabolism through hepatic lipophagy, which traffics lipid to lysosomes for degradation. Impaired autophagy of lipids in hypothyroidism disrupts this process, activates mammalian target of rapamycin pathway, and decreases the overall quality of autophagy of damaged mitochondria, thereby exacerbating oxidative and ER stress [
29-
31]. In the setting of prolonged hypothyroidism, these compensatory mechanisms become overwhelmed, contributing to persistent hepatocellular injury, systemic inflammation, and oxidative DNA damage. This pathological environment further promotes fibrogenesis via activation of transforming growth factor β and hepatic stellate cells [
26,
31,
32]. Overall, these mechanisms may drive the simple steatosis progression to end-stage liver disease. The intricate interaction between thyroid dysfunction and liver disease progression suggests that hypothyroidism may act be a potential accelerator of LREs.
In this study, we observed a significant association between time-varying hypothyroidism and the risk of LREs, whereas baseline thyroid function was not as predictive. This highlights the importance of capturing dynamic changes in thyroid status when evaluating its impact on MASLD progression, as a single baseline measurement is static and may not reflect long-term dysfunction. For example, a patient who is euthyroid at baseline may subsequently develop subclinical or overt hypothyroidism during follow-up, given the natural fluctuation of TSH and free T4 levels over time [
17]. Even for clinical diagnosis, repeated measurements over 2–3 months are typically required to confirm hypothyroidism [
33]. Moreover, MASLD is a chronic, progressive disease in which thyroid status may change over time due to disease progression or treatment. Without accounting for longitudinal changes, the risk of liver-related outcomes in patients who develop hypothyroidism during follow-up may be underestimated. Time-varying models address this by capturing whether thyroid dysfunction persists, worsens, or resolves over time. This distinction is important because LREs are more likely driven by the cumulative burden of prolonged hypothyroidism rather than transient alteration. Our Sankey plot shows fluctuations in thyroid status, with many patients transitioning between euthyroid and hypothyroid states, which further highlights the limitations of a one-off baseline assessment used in previous studies [
8,
9]. Furthermore, we also observed a non-significant trend toward reduced LRE risk with T4 therapy, suggesting a potential protective effect. The lack of statistical significance is likely due to the complex interplay between thyroid function and treatment, with T4 therapy often adjusted according to thyroid function tests. This aligns with previous studies showing that thyroid hormone replacement could reduce hepatic fat content and may influence outcomes by restoring euthyroidism, thereby modifying the MASLD progression [
6,
7]. These findings prompted further exploration of whether thyroid function normalization through T4 replacement treatment reduces LRE risk. No significant difference was observed, suggesting that the increased risk associated with hypothyroidism may be reversible with endocrine restoration. However, despite most patients ultimately achieving normalization during follow-up, treatment initiation time varied widely, resulting in some patients in a prolonged untreated state. This variation in untreated duration may influence outcomes, as the cumulative burden of hypothyroidism may differ across individuals influenced by their hepatic reserve, metabolic condition and treatment response. Given the limitations of previous studies and the statistically non-significant findings in our study, further prospective studies are warranted to investigate the therapeutic role of T4 replacement treatment in preventing LRE. Taken together, a one-off assessment may not reflect the cumulative exposure, therapeutic modification, disease progression, or prognostic impact of thyroid dysfunction. Dynamic monitoring enables more accurate risk stratification, early detection of LRE risk, and better assessment of treatment effects and thyroid-liver interplay.
Our study has several strengths, which includes a large sample size, use of territory-wide data representative of the local population and dynamic thyroid function assessments. This study also has limitations. First, MASLD was identified using ICD-9-CM codes, which are under-coded due to under-recognition or lack of formal diagnosis. Second, serum T3, thyroglobulin, anti-TPO, and anti-TG antibody data were not analysed due to high missingness. Third, although the sample size was adequate for assessing clinical outcomes, the median follow-up of 4.8 years may be relatively short given the slow progression of chronic liver disease toward cirrhosis and complications. Fourth, residual confounding cannot be fully excluded given the observational design. However, the large sample size partially offset this by allowing multivariable adjustment. Finally, the generalizability is limited by the homogeneous Hong Kong population and requires validation in other ethnic groups.
In conclusion, time-varying hypothyroidism is significantly associated with increased incidence of LRE in patients with MASLD. The association remains robust after stratified into subclinical and overt hypothyroidism. These findings suggest the prognostic value of dynamic thyroid function assessment over single baseline measurements and support dynamic monitoring and early intervention as potential strategies to reduce the risk of liver-related complications.
FOOTNOTES
-
Authors’ contributions
All authors were responsible for the study concept and design. Xinrui Jin, Terry Cheuk-Fung Yip, Vincent Wai-Sun Wong, Grace Lai-Hung Wong, Sherlot Juan Song and Nana Peng responsible for the data collection. Xinrui Jin, Terry Cheuk-Fung Yip, Vincent Wai-Sun Wong were responsible for data analysis, had full access to all of the data in the study, and took responsibility for the integrity of the data and the accuracy of the data analysis. All authors were responsible for the interpretation of data. Xinrui Jin, Terry Cheuk-Fung Yip, Vincent Wai-Sun Wong were responsible for the manuscript drafting, and the critical revision of the manuscript for important intellectual content. All authors reviewed and approved the final manuscript.
-
Acknowledgements
The study was supported in part by the General Research Fund from HKSAR Government (project reference 14106923).
-
Conflicts of Interest
Terry Cheuk-Fung Yip has served as an advisory committee member and a speaker for Gilead Sciences. He has received a research grant from Gilead Sciences. Grace Lai-Hung Wong has served as an advisory committee member for AstraZeneca, Barinthus Biotherapeutics, Gilead Sciences, GlaxoSmithKline, Janssen and Virion Therapeutics, and as a speaker for Abbott, AbbVie, Ascletis, Bristol-Myers Squibb, Echosens, Ferring, Gilead Sciences, GlaxoSmithKline, Janssen, and Roche. She has also received research grants from Gilead Sciences. Vincent WaiSun Wong has served as an advisory board member for AbbVie, AstraZeneca, Boehringer Ingelheim, Echosens, Gilead Sciences, Intercept, Inventiva, Merck, Novo Nordisk, Pfizer, Sagimet Biosciences, TARGET PharmaSolutions, and Visirna; and a speaker for Abbott, AbbVie, Echosens, Gilead Sciences, Novo Nordisk, and Unilab. He has received a research grant from Gilead Sciences, and is a cofounder of Illuminatio Medical Technology. Jimmy CheTo Lai has served as a speaker for Gilead Sciences and an advisory board member for Gilead Sciences and Boehringer Ingelheim. Alice Pik-Shan Kong has received research grants and/or speaker honoraria from Abbott, Astra Zeneca, Bayer, Boehringer Ingelheim, Dexcom, Eli-Lilly, Kyowa Kirin, Merck Serono, Merck Sharp & Dohme, Nestle, NovoNordisk, Pfizer, Sanofi and Zuellig Pharma.
SUPPLEMENTARY MATERIAL
Supplementary material is available at Clinical and Molecular Hepatology website (
http://www.e-cmh.org).
Supplementary Figure 1.
Cumulative risks of incident LRE in accordance with baseline thyroid status (A) in participants with hypothy-roidism, euthyroidism, and hyperthyroidism. (B) in patients with overt hypothyroidism, subclinical hypothyroidism, and euthyroidism. LRE, liver-related event.
Supplementary Figure 2.
Cumulative risks of incident LRE in accordance with baseline TSH status. When TSH is <0.4, which is flagged as “Low”, the normal range of TSH is 0.4–4 mIU/L; when TSH is >4 mIU/L, which is flagged as “High”. LRE, liver-related event; TSH, thyroid-stimulating hormone.
Supplementary Figure 3.
Cumulative risks of incident LRE in accordance with baseline free T4 status. The normal range of free T4 is based on laboratory assay. Free T4, free thyroxine; LRE, liver-related event.
Supplementary Figure 4.
Transitions between thyroid states during the first year and third year. (A) First year: total number 12,631; baseline, hyperthyroidism (n=647), euthyroidism (n=11,182), subclinical hypothyroidism (n=600), overt hypothyroidism (n=202); Year 1, hyperthyroidism (n=525), euthyroidism (n=11,332), subclinical hypothyroidism (n=530), overt hypothyroidism (n=133), LRE (n=8), and died (n=103). (B) Third year: total number =11,986, baseline, hyperthyroidism (n=617), euthyroidism (n=10,600), subclinical hypothyroidism (n=576), overt hypothyroidism (n=193); Year 3, hyperthyroidism (n=469), euthyroidism (n=10,682), subclinical hypothyroidism (n=465), overt hypothyroidism (n=92), LRE (n=28), and died (n=250). Thyroid status was defined by thyroid function test. LRE, liver-related event.
Supplementary Figure 5.
Cumulative risks of incident LRE in accordance with or without thyroxine replacement. Included patients with hypothyroidism. LRE, liver-related event.
Supplementary Figure 6.
Cumulative incidence of liver-related events among patients with euthyroidism vs. treated hypothyroidism with normalization, (A) before propensity score matching and (B) after propensity score matching.
Supplementary Table 12.
Association of thyroid status during follow-up and risk of LRE excluding patients without thyroid status assessment within 1 year before the diagnosis of MASLD
Supplementary Table 18.
Association between TSH levels and status at baseline and during follow-up and risk of LRE stratified by thy-
roxine replacement treatment status
Figure 1.Patient flowchart for selection of study subjects. Free T4, free thyroxine; HBV, hepatitis B virus; HCC, hepatocellular carcinoma; HCV, hepatitis C virus; HIV, human immunodeficiency virus; MASLD, metabolic dysfunction-associated steatotic liver disease; TSH, thyroid-stimulating hormone.
Figure 2.Transitions between thyroid states during the fifth year. Total number 9,411; baseline, hyperthyroidism (n=515), euthyroidism (n=8,289), subclinical hypothyroidism (n=450), overt hypothyroidism (n=157); Year 5, hyperthyroidism (n=362), euthyroidism (n=8,178), subclinical hypothyroidism (n=346), overt hypothyroidism (n=80), LRE (n=36), and died (n=409). Thyroid status was defined by thyroid function test. LRE, liver-related events.
Figure 3.Cumulative risk of incident LRE in accordance with thyroid status during follow-up. Thyroid status was defined by thyroid function test. LRE, liver-related events.
Table 1.
Table 1.
|
Characteristics |
N |
Overall |
Hypothyroidism |
Euthyroidism |
Hyperthyroidism |
P-value |
|
Number |
20,478 |
|
1,702 (8.3) |
18,178 (88.8) |
598 (2.9) |
|
|
Age (yr) |
20,478 |
56.4±13.2 |
57.3±13.5 |
56.2±13.1 |
56.4±13.2 |
<0.001 |
|
Male sex |
|
8,990 (43.9) |
494 (29.0) |
8,260 (45.4) |
236 (39.5) |
<0.001 |
|
FIB-4*
|
6,964 |
1.1 (0.7-1.6) |
1.2 (0.8-1.9) |
1.0 (0.7-1.6) |
1.1 (0.8-1.8) |
<0.001 |
|
Missing |
|
13,514 (66.0) |
1,094 (64.3) |
12,049 (66.3) |
371 (62.0) |
|
|
Cirrhosis†
|
|
383 (1.9) |
49 (2.9) |
313(1.7) |
21 (3.5) |
<0.001 |
|
TSH (mlU/L) |
15,359 |
1.5 (1.0-2.2) |
3.7 (1.3-5.0) |
1.5 (1.0-2.1) |
0.3 (0.1-0.4) |
<0.001 |
|
<0.4 |
|
612 (4.0) |
142 (8.6) |
NA |
470 (79.3) |
|
|
0.4-4 |
|
13,953 (90.8) |
719 (43.8) |
13,123 (100) |
111 (18.7) |
|
|
4-10 |
|
719 (4.7) |
709 (43.2) |
NA |
12 (2) |
|
|
>10 |
|
74 (0.5) |
73 (4.4) |
NA |
NA |
|
|
Missing |
|
5,119 (25.0) |
59 (3.5) |
5,055 (27.8) |
5 (0.8) |
|
|
Free T4 (pmol/L) |
5,367 |
13.9 (12.2-16.2) |
13.9 (11.8-16.5) |
13.8 (12.2-1 5.7) |
16.2 (13.0-20.5) |
<0.001 |
|
Missing |
|
15,111 (73.8) |
395 (23.2) |
14,594 (80.3) |
122 (20.4) |
|
|
Free T3 (pmol/L) |
341 |
4.4 (3.7-5.0) |
4.4 (3.7-5.2) |
4.2 (3.6-4.8) |
4.8 (4.1-5.8) |
<0.001 |
|
Missing |
|
20,137 (98.3) |
1,622 (95.3) |
18,004 (99.0) |
511 (85.5) |
|
|
Thyroglobulin (ug/L) |
26 |
0.5 (0.2-20.3) |
0.2 (0.2-6.0) |
62.5 (4.2-794.0) |
14.7 (14.7-14.7) |
0.029 |
|
Missing |
|
20,452 (99.9) |
1,683 (98.9) |
18,172 (99.9) |
597 (99.8) |
|
|
Anti-TPO positive |
340 |
117 (34.4) |
67 (3.9) |
16 (0.1) |
34 (5.7) |
<0.001 |
|
Missing |
|
20,138 (98.3) |
1,560 (91.7) |
18,074 (99.4) |
597 (84.3) |
|
|
Anti-TG positive |
350 |
93 (26.6) |
51 (3.0) |
20 (0.1) |
22 (3.7) |
0.006 |
|
Missing |
|
20,128 (98.3) |
1,548 (91.0) |
18,076 (99.4) |
504 (84.3) |
|
|
Use of medication for hypothyroidism |
20,478 |
746 (3.6) |
746 (43.8) |
NA |
NA |
<0.001 |
|
Type 2 diabetes mellitus |
20,478 |
11,579 (56.5) |
956 (56.2) |
10,273 (56.5) |
350 (58.5) |
0.600 |
|
Hypertension |
20,478 |
15,568 (76.0) |
1,332 (78.3) |
13,743 (75.6) |
487 (81.4) |
<0.001 |
|
BMI (kg/m2) |
7,902 |
28.3±4.8 |
28.5±5.3 |
28.3±4.7 |
27.5±5.0 |
0.024 |
|
Male |
|
28.7±4.7 |
28.4±5 |
28.7±4.6 |
28.5±5.8 |
|
|
Female |
|
28.0±4.8 |
28.5±5.5 |
28.0±4.8 |
26.9±4.4 |
|
|
Missing |
|
12,576 (61.4) |
1,064 (62.5) |
11,146 (61.3) |
366 (61.2) |
|
|
Waist circumference (cm) |
6,557 |
95.9±11.2 |
95.9±11.6 |
96.0±11.1 |
93.8±11.5 |
0.017 |
|
Male |
|
99.3±11.1 |
99.1±10.4 |
99.4±11 |
99.3±12.4 |
|
|
Female |
|
93.6±10.6 |
94.8±11.8 |
93.5±10.5 |
90.5±9.4 |
|
|
Missing |
|
13,901 (67.9) |
1,203 (70.7) |
12,274 (67.5) |
424 (70.9) |
|
|
SBP (mmHg) |
6,492 |
133.0±15.5 |
134.0±16.5 |
132.9±15.4 |
133.9±15.6 |
0.130 |
|
Missing |
|
13,986 (68.3) |
1,208 (71.0) |
12,354 (68.0) |
424 (70.9) |
|
|
DBP (mmHg) |
6,492 |
76.4±10.3 |
75.8±10.9 |
76.4±10.2 |
75.9±11.3 |
0.400 |
|
Missing |
|
13,986 (68.3) |
1,208 (71.0) |
12,354 (68.0) |
424 (70.9) |
|
|
Total cholesterol |
19,968 |
4.6 (3.9-5.5) |
4.7 (3.9-5.5) |
4.6 (3.9-5.4) |
4.6 (3.8-5.4) |
0.002 |
|
Missing |
|
510 (2.5) |
44 (2.6) |
424 (2.3) |
42 (7.0) |
|
|
HDL cholesterol (mmol/L) |
19,750 |
1.2±0.3 |
1.2±0.3 |
1.2±0.3 |
1.2±0.3 |
0.060 |
|
Missing |
|
728 (3.6) |
64 (3.8) |
620 (3.4) |
44 (7.4) |
|
|
LDL cholesterol (mmol/L) |
19,846 |
2.7±0.9 |
2.8±1.0 |
2.7±0.9 |
2.7±0.9 |
0.027 |
|
Missing |
|
632 (3.1) |
62 (3.6) |
524 (2.9) |
46 (7.7) |
|
|
Triglyceride (mmol/L) |
19,911 |
1.6 (1.2–2.3) |
1.7 (1.2–2.4) |
1.6 (1.2–2.3) |
1.6 (1.1–2.3) |
0.016 |
|
Missing |
|
567 (2.8) |
55 (3.2) |
477 (2.6) |
35 (5.9) |
|
|
Fasting glucose (mmol/L) |
19,664 |
5.9 (5.2–7.0) |
5.8 (5.2–7.0) |
5.9 (5.2–7.0) |
5.9 (5.2–7.1) |
0.200 |
|
Missing |
|
816 (4.0) |
64 (3.8) |
711 (3.9) |
39 (6.5) |
|
|
ALT (U/L) |
20,347 |
41 (26–64) |
40 (24–65) |
41 (26–64) |
37 (23–60) |
0.001 |
|
Missing |
|
131 (0.6) |
7 (0.4) |
122 (0.7) |
2 (0.3) |
|
|
AST (U/L) |
7,580 |
32 (24–47) |
35 (25–53) |
32 (24–46) |
30 (22–47) |
0.001 |
|
Missing |
|
12,898 (63.0) |
1,054 (61.9) |
11,486 (63.2) |
358 (59.9) |
|
|
Platelets (×109/L) |
17,368 |
257±73 |
256±75 |
257±73 |
255±81 |
0.200 |
|
Missing |
|
3,110 (15.2) |
197 (11.6) |
2,869 (15.8) |
44 (7.4) |
|
|
Follow-up duration (yr) |
20,478 |
4.8 (3.9–10.4) |
4.9 (4.0–12.0) |
4.8 (3.9–10.2) |
4.9 (4.0–10.8) |
<0.001 |
Table 2.Association between baseline thyroid status and risk of cirrhosis at baseline
Table 2.
|
Variables |
Prevalence of cirrhosis (%) |
Univariate model |
Multivariable model 1 |
Multivariable model 2 |
|
OR (95% CI) |
P-value |
Adjusted OR (95% CI) |
P-value |
Adjusted OR (95% CI) |
P-value |
|
Hypothyroidism |
2.88 |
1.69 (1.25–2.30) |
<0.001 |
1.44 (1.05–1.99) |
0.025 |
1.44 (1.05–1.99) |
0.025 |
|
Subclinical |
2.64 |
1.55 (0.88–2.71) |
0.129 |
1.34 (0.73–2.44) |
0.344 |
1.29 (0.71–2.35) |
0.401 |
|
Overt |
3.00 |
1.77 (1.23–2.53) |
0.002 |
1.51 (1.04–2.21) |
0.031 |
1.52 (1.04–2.21) |
0.029 |
|
Euthyroidism |
1.72 |
1.00 (ref) |
|
1.00 (ref) |
|
1.00 (ref) |
|
|
Hyperthyroidism |
3.51 |
2.08 (1.33–3.26) |
0.001 |
1.63 (1.01–2.63) |
0.044 |
1.67 (1.04–2.69) |
0.034 |
Table 3.Association between TSH status and thyroid status during follow-up and risk of LRE
Table 3.
|
Variables |
Univariate model |
Multivariable model 1 |
Multivariable model 2 |
|
CSHR (95% CI) |
P-value |
CSHR (95% CI) |
P-value |
CSHR (95% CI) |
P-value |
|
TSH status |
|
|
|
|
|
|
|
0.4–4 mIU/L |
1.00 (ref) |
|
1.00 (ref) |
|
1.00 (ref) |
|
|
4–10 mIU/L |
2.67 (1.61–4.42) |
<0.001 |
2.44 (1.47–4.04) |
<0.001 |
2.49 (1.51–4.13) |
<0.001 |
|
>10 mIU/L |
5.14 (1.64–16.13) |
0.005 |
4.77 (1.52–15.03) |
0.008 |
4.91 (1.56–15.47) |
0.006 |
|
Thyroid status |
|
|
|
|
|
|
|
Hypothyroidism |
3.51 (2.27–5.43) |
<0.001 |
3.14 (2.02–4.85) |
<0.001 |
3.19 (2.06–4.94) |
<0.001 |
|
Subclinical |
2.88 (1.74–4.78) |
<0.001 |
2.58 (1.56–4.29) |
<0.001 |
2.58 (1.56–4.29) |
<0.001 |
|
Overt |
7.49 (3.50–16.03) |
<0.001 |
6.52 (3.02–14.10) |
<0.001 |
6.41 (2.96–13.85) |
<0.001 |
|
Euthyroidism |
1.00 (ref) |
|
1.00 (ref) |
|
1.00 (ref) |
|
|
Hyperthyroidism |
1.41 (0.69–2.89) |
0.342 |
1.19 (0.58–2.43) |
0.636 |
1.21 (0.59–2.48) |
0.599 |
Table 4.Association of thyroid status during follow-up and risk of LRE in the sensitivity analysis based on the method of censoring
Table 4.
|
Variables |
Univariate model |
Multivariable model 1 |
Multivariable model 2 |
|
CSHR (95% CI) |
P-value |
Adjusted CSHR (95% CI) |
P-value |
Adjusted CSHR (95% CI) |
P-value |
|
Hypothyroidism |
4.56 (2.81–7.39) |
<0.001 |
4.06 (2.50–6.60) |
<0.001 |
4.10 (2.52–6.67) |
<0.001 |
|
Subclinical |
3.54 (2.00–6.27) |
<0.001 |
3.17 (1.79–5.62) |
<0.001 |
3.18 (1.80–5.64) |
<0.001 |
|
Overt |
9.57 (4.41–20.76) |
<0.001 |
7.63 (3.43–16.95) |
<0.001 |
7.67 (3.46–17.00) |
<0.001 |
|
Euthyroidism |
1.00 (ref) |
|
1.00 (ref) |
|
1.00 (ref) |
|
|
Hyperthyroidism |
1.28 (0.52–3.17) |
0.589 |
1.14 (0.46–2.81) |
0.782 |
1.15 (0.46–2.85) |
0.760 |
Table 5.Association of thyroid status during follow-up and risk of LRE using 1-year and 2-year landmark analyses
Table 5.
|
Variables |
Univariate model |
Multivariable model 1 |
Multivariable model 2 |
|
CSHR (95% CI) |
P-value |
Adjusted CSHR (95% CI) |
P-value |
Adjusted CSHR (95% CI) |
P-value |
|
1-year landmark analysis |
|
|
|
|
|
|
|
Hypothyroidism |
3.37 (2.14–5.30) |
<0.001 |
3.04 (1.93–4.78) |
<0.001 |
3.07 (1.95–4.84) |
<0.001 |
|
Subclinical |
2.82 (1.68–4.75) |
<0.001 |
2.54 (1.51–4.28) |
<0.001 |
2.54 (1.51–4.29) |
<0.001 |
|
Overt |
6.75 (2.97–15.33) |
<0.001 |
5.92 (2.58–13.58) |
<0.001 |
5.83 (2.54–13.38) |
<0.001 |
|
Euthyroidism |
1.00 (ref) |
|
1.00 (ref) |
|
1.00 (ref) |
|
|
Hyperthyroidism |
1.30 (0.61–2.79) |
0.496 |
1.09 (0.51–2.34) |
0.820 |
1.10 (0.51–2.37) |
0.799 |
|
2-year landmark analysis |
|
|
|
|
|
|
|
Hypothyroidism |
3.06 (1.86–5.03) |
<0.001 |
2.79 (1.69–4.59) |
<0.001 |
2.81 (1.71–4.63) |
<0.001 |
|
Subclinical |
2.35 (1.30–4.27) |
0.005 |
2.12 (1.17–3.85) |
0.014 |
2.12 (1.17–3.86) |
0.013 |
|
Overt |
7.45 (3.27–16.93) |
<0.001 |
7.33 (3.20–16.80) |
<0.001 |
7.23 (3.16–16.61) |
<0.001 |
|
Euthyroidism |
1.00 (ref) |
|
1.00 (ref) |
|
1.00 (ref) |
|
|
Hyperthyroidism |
1.23 (0.54–2.79) |
0.624 |
1.03 (0.45–2.35) |
0.943 |
1.03 (0.45–2.36) |
0.935 |
Abbreviations
antithyroglobulin antibody
anti-thyroid peroxidase antibody
aspartate aminotransferase
Clinical Data Analysis and Reporting System
cause-specific hazard ratio
human immunodeficiency virus
International Classification of Diseases
metabolic dysfunction-associated steatotic liver disease
metabolic dysfunction-associated steatohepatitis
subdistribution hazard ratio
thyroid-stimulating hormone
REFERENCES
- 1. Younossi ZM, Golabi P, Paik JM, Henry A, Van Dongen C, Henry L. The global epidemiology of nonalcoholic fatty liver disease (NAFLD) and nonalcoholic steatohepatitis (NASH): a systematic review. Hepatology 2023;77:1335-1347.
- 2. Powell EE, Wong VW, Rinella M. Non-alcoholic fatty liver disease. Lancet 2021;397:2212-2224.
- 3. Wong RJ, Aguilar M, Cheung R, Perumpail RB, Harrison SA, Younossi ZM, et al. Nonalcoholic steatohepatitis is the second leading etiology of liver disease among adults awaiting liver transplantation in the United States. Gastroenterology 2015;148:547-555.
- 4. Lonardo A, Ballestri S, Mantovani A, Nascimbeni F, Lugari S, Targher G. Pathogenesis of hypothyroidism-induced NAFLD: Evidence for a distinct disease entity? Dig Liver Dis 2019;51:462-470.
- 5. Byrne CD, Targher G, Tilg H. Thyroid hormone receptor-beta agonists: new MASLD therapies on the horizon. Gut 2024;73:573-581.
- 6. Liu L, Yu Y, Zhao M, Zheng D, Zhang X, Guan Q, et al. Benefits of levothyroxine replacement therapy on nonalcoholic fatty liver disease in subclinical hypothyroidism patients. Int J Endocrinol 2017;2017:5753039.
- 7. Bruinstroop E, Dalan R, Cao Y, Bee YM, Chandran K, Cho LW, et al. Low-dose levothyroxine reduces intrahepatic lipid content in patients with type 2 diabetes mellitus and NAFLD. J Clin Endocrinol Metab 2018;103:2698-2706.
- 8. Bano A, Chaker L, Plompen EP, Hofman A, Dehghan A, Franco OH, et al. Thyroid function and the risk of nonalcoholic fatty liver disease: the rotterdam study. J Clin Endocrinol Metab 2016;101:3204-3211.
- 9. Mantovani A, Csermely A, Bilson J, Borella N, Enrico S, Pecoraro B, et al. Association between primary hypothyroidism and metabolic dysfunction-associated steatotic liver disease: an updated meta-analysis. Gut 2024;73:1554-1561.
- 10. Untalan M, A Crimmins N, Yates KP, A Mencin A, A Xanthakos S, Thaker VV. Prevalence of subclinical hypothyroidism and longitudinal thyroid-stimulating hormone changes in youth with metabolic dysfunction-associated steatotic liver disease: an observational study. Hepatology 2025;82:155-164.
- 11. Hassan MM, Kaseb A, Li D, Patt YZ, Vauthey JN, Thomas MB, et al. Association between hypothyroidism and hepatocellular carcinoma: a case-control study in the United States. Hepatology 2009;49:1563-1570.
- 12. Reddy A, Dash C, Leerapun A, Mettler TA, Stadheim LM, Lazaridis KN, et al. Hypothyroidism: a possible risk factor for liver cancer in patients with no known underlying cause of liver disease. Clin Gastroenterol Hepatol 2007;5:118-123.
- 13. Sahin T, Oral A, Turker F, Kocak E. Can hypothyroidism be a protective factor for hepatocellular carcinoma in cirrhosis? Medicine (Baltimore) 2020;99:e19492.
- 14. Nemeth K, Gautam A, Rosen H. S4262 Progression of MASLD to cirrhosis secondary to prolonged severe hypothyroidism. Am J Gastroenterol 2024;119:S2743-S2744.
- 15. Zhang X, Wong GL, Yip TC, Tse YK, Liang LY, Hui VW, et al. Angiotensin-converting enzyme inhibitors prevent liver-related events in nonalcoholic fatty liver disease. Hepatology 2022;76:469-482.
- 16. Yip TCF, Li G, Wong GLH, Tse YK, Wong V. Accuracy of administrative coding for nonalcoholic fatty liver disease and related hepatic complications: a single-center validation study. Hepatology 2021;74:928A-929A.
- 17. Jonklaas J, Bianco AC, Bauer AJ, Burman KD, Cappola AR, Celi FS, et al.; American Thyroid Association Task Force on Thyroid Hormone Replacement. Guidelines for the treatment of hypothyroidism: prepared by the american thyroid association task force on thyroid hormone replacement. Thyroid 2014;24:1670-1751.
- 18. Pearce SH, Brabant G, Duntas LH, Monzani F, Peeters RP, Razvi S, et al. 2013 ETA guideline: management of subclinical hypothyroidism. Eur Thyroid J 2013;2:215-228.
- 19. Mantovani A, Nascimbeni F, Lonardo A, Zoppini G, Bonora E, Mantzoros CS, et al. Association between primary hypothyroidism and nonalcoholic fatty liver disease: a systematic review and meta-analysis. Thyroid 2018;28:1270-1284.
- 20. Kim D, Kim W, Joo SK, Bae JM, Kim JH, Ahmed A. Subclinical hypothyroidism and low-normal thyroid function are associated with nonalcoholic steatohepatitis and fibrosis. Clin Gastroenterol Hepatol 2018;16:123-131.e1.
- 21. Kouvari M, Valenzuela-Vallejo L, Axarloglou E, Verrastro O, Papatheodoridis G, Mingrone G, et al. Thyroid function, adipokines and mitokines in metabolic dysfunction-associated steatohepatitis: a multi-centre biopsy-based observational study. Liver Int 2024;44:848-864.
- 22. Punekar P, Sharma AK, Jain A. A study of thyroid dysfunction in cirrhosis of liver and correlation with severity of liver disease. Indian J Endocrinol Metab 2018;22:645-650.
- 23. Singh S, Allen AM, Wang Z, Prokop LJ, Murad MH, Loomba R. Fibrosis progression in nonalcoholic fatty liver vs nonalcoholic steatohepatitis: a systematic review and meta-analysis of paired-biopsy studies. Clin Gastroenterol Hepatol 2015;13:643-654.e1-9 quiz e39-40.
- 24. Ekstedt M, Hagström H, Nasr P, Fredrikson M, Stål P, Kechagias S, et al. Fibrosis stage is the strongest predictor for disease-specific mortality in NAFLD after up to 33 years of follow-up. Hepatology 2015;61:1547-1554.
- 25. Huang DQ, Singal AG, Kono Y, Tan DJH, El-Serag HB, Loomba R. Changing global epidemiology of liver cancer from 2010 to 2019: NASH is the fastest growing cause of liver cancer. Cell Metab 2022;34:969-977.e2.
- 26. Loomba R, Friedman SL, Shulman GI. Mechanisms and disease consequences of nonalcoholic fatty liver disease. Cell 2021;184:2537-2564.
- 27. Friedman SL, Neuschwander-Tetri BA, Rinella M, Sanyal AJ. Mechanisms of NAFLD development and therapeutic strategies. Nat Med 2018;24:908-922.
- 28. Bohinc BN, Michelotti G, Xie G, Pang H, Suzuki A, Guy CD, et al. Repair-related activation of hedgehog signaling in stromal cells promotes intrahepatic hypothyroidism. Endocrinology 2014;155:4591-4601.
- 29. Ratziu V, Scanlan TS, Bruinstroop E. Thyroid hormone receptor-β analogues for the treatment of metabolic dysfunction-associated steatohepatitis (MASH). J Hepatol 2025;82:375-387.
- 30. Chi HC, Tsai CY, Tsai MM, Yeh CT, Lin KH. Molecular functions and clinical impact of thyroid hormone-triggered autophagy in liver-related diseases. J Biomed Sci 2019;26:24.
- 31. Sharma N, Singh L, Sharma A, Kumar A, Mahajan D. NAFLD-associated hepatocellular carcinoma (HCC) - a compelling case for repositioning of existing mTORc1 inhibitors. Pharmacol Res 2024;208:107375.
- 32. Parola M, Pinzani M. Liver fibrosis in NAFLD/NASH: from pathophysiology towards diagnostic and therapeutic strategies. Mol Aspects Med 2024;95:101231.
- 33. Liu Y, Shan Z; Endocrine Metabolic Diseases Group of the Chinese Geriatrics Society; Thyroid Group of the Chinese Society of Endocrinology, Chinese Medical Association. Expert consensus on diagnosis and treatment for elderly with thyroid diseases in China (2021). Aging Med (Milton) 2021;4:70-92.
, Terry Cheuk-Fung Yip1,2,3,4
