Prevalence of Subclinical Hypothyroidism and Longitudinal Thyroid Stimulating Hormone Changes in Youth with Metabolic Dysfunction-Associated Steatotic Liver Disease: an Observational Study

Abstract

Background:

Studies on adults have shown an association between overt or subclinical hypothyroidism and metabolic dysfunction-associated steatotic liver disease (MASLD). The goal of this study was to assess the relationship between thyroid-stimulating hormone (TSH) levels and the histological characteristics of MASLD in youth.

Methods:

This observational study used prospectively collected liver biopsy and clinical data from youth enrolled in two pediatric clinical trials in the Nonalcoholic Steatohepatitis Clinical Research Network (NASH CRN). Thyroid assays were compared between youth with MASLD and population-based controls aged ≤ 18 years from the National Health and Nutrition Examination Survey (NHANES). Individuals with overt hypothyroidism, abnormal anti-thyroid antibody, or thyroid-related medications were excluded. Subclinical hypothyroidism was defined as TSH between 4.5 and 10.0 uIU/L. Multinomial logistic regression was used to test the association between TSH and MASLD histological changes at baseline, adjusting for age, sex, race/ethnicity, and body mass index. Mixed-effect models, including treatment and time, were used for the longitudinal analysis.

Results:

Mean TSH, total thyroxine (T4), total triiodothyronine (T3), and free T4 levels were higher (p < 0.001) in the NASH CRN cohort (n= 218; 421 observations) than in the NHANES cohort (n=2,198). TSH levels were positively associated with increased steatosis over time (p = 0.03). Subclinical hypothyroidism was associated with borderline or definite metabolic-associated steatohepatitis on histology at baseline (p=0.03) and with changes in fibrosis over time (p = 0.01).

Conclusion:

The association between TSH and steatosis severity in individuals with normal thyroid hormone concentrations suggests an independent role of TSH in MASLD.

Introduction

Metabolic dysfunction-associated steatotic liver disease (MASLD) is the most common chronic liver disease in children and parallels the prevalence of obesity^1,2. Subclinical hypothyroidism (SH) is defined as mildly elevated thyroid stimulating hormone (TSH) with normal circulating thyroid hormone levels. SH is often observed in individuals with obesity, although the underlying mechanism or cause-effect relationship is not known³.

Thyroid hormones are important regulators of hepatic lipid metabolism through the induction of lipogenesis, coupling of autophagy to mitochondrial fat oxidation leading to ketogenesis, and reversal of cholesterol transport^4,5. Adult studies have suggested a close relationship between SH and cardiometabolic risk factors, including MASLD^6-8. European studies have shown an association between SH and MASLD identified by ultrasound in children and/or adolescents with obesity^9-12. We confirmed a positive association between higher TSH levels and the presence of biopsy-confirmed MASLD, independent of body mass index (BMI) z-score, in a single-center study of Hispanic/ Latino children compared to age-matched controls from primary care clinics¹³. Whether this association applies widely beyond a single center to a large and ethnically diverse cohort of children with MASLD, and how it relates to the histological characteristics and severity of liver disease, is unknown.

We hypothesized that elevated TSH levels play an important role in the development and severity of histological changes of MASLD and that understanding this relationship will contribute to better defining the complex role of thyroid function in steatotic liver disease. To test this hypothesis, we used data from two large cohorts: the Nonalcoholic Steatohepatitis Clinical Research Network (NASH CRN) and the population-based National Health and Nutrition Examination Survey (NHANES).

Methods

This study was considered exempt for review by the Columbia University Irving Medical Center (CUIMC) Institutional Review Board (IRB) because of the use of de-identified data and samples (Protocol IRB-AAAS7067).

Study populations

Data and samples from two population groups were utilized for this study.

NHANES cohort

The NHANES is a cross-sectional, multistage complex nationally representative survey that uses interviews, physical examinations, and laboratory data to assess the health and nutritional status of the non-institutionalized civilian US population¹⁴. This study used data collected in five 2-year NHANES cycles where thyroid function assays were measured (1999-2000, 2001-2002 and 2007-2008 through 2011-2012) in youth aged ≤ 18 years. Demographic information included age, sex, self-reported race/ethnicity, and classification as non-Hispanic white (NHW), non-Hispanic black (NHB), Hispanic, or other. Data pertaining to thyroid function, lipid levels, liver enzymes, and anthropometric measurements were obtained. Participants with alanine ALT levels > 22 U/L (females) or > 26 U/L (males), according to the biologic upper normal limit, were excluded¹⁵.

NASH CRN cohort

The data for this project were collected by the NASH CRN prior to the 2023 nomenclature revision of nonalcoholic fatty liver disease (NAFLD) to MASLD and nonalcoholic steatohepatitis (NASH) to metabolic dysfunction-associated steatohepatitis (MASH)¹⁶. NASH CRN used the validated NAFLD activity score (NAS) to measure the severity of histological disease. The terms used did not change at the time of writing. This study used liver biopsy and associated clinical and demographic data collected at baseline and end-of-treatment from participants of two multicenter pediatric clinical trials conducted by the NASH CRN: Treatment of NAFLD in Children (TONIC, NCT00063635) and Cysteamine Bitartrate Delayed-Release for the Treatment of Nonalcoholic Fatty Liver Disease in Children (CyNCh, NCT01529268). The TONIC was a randomized, double-masked, placebo-controlled trial that enrolled 8- to 17-year-old children with histologically confirmed MASLD and persistently elevated ALT levels. Participants were randomized to metformin (500 mg twice daily), vitamin E (400 IU twice daily), or placebo for 96 weeks between September 2005 and March 2010. The secondary outcomes included changes in liver histology. The primary outcome was a sustained reduction in the ALT level ≤50% of the baseline level or ≤40 U/L at each visit from 48 to 96 weeks^17,18. CyNCh was a placebo-controlled, randomized phase II b clinical trial that enrolled 8–17-year-old children with histologically confirmed moderate-to-severe MASLD conducted between June 2012 and January 2014¹⁹. The primary outcome was the proportion of children with histological improvement, defined as a ≥ 2-point reduction in NAS and no worsening of fibrosis, after 52 weeks of treatment. In both studies, the initial liver biopsy was performed for clinical diagnosis and MASLD staging, whereas the end-of-treatment liver biopsy was obtained under an IRB-approved research protocol at each site. The treatments used in these studies had no known effect on thyroid function; therefore, baseline and follow-up data from participants in both the placebo and treatment arms were used. All the data were provided by the NASH CRN Data Coordinating Center.

Thyroid assays

Baseline and end-of-treatment (96 wk for the TONIC study and 52 wK for the CyNCh study) serum samples from the enrolled participants were used to measure TSH, free T4, total T4, total T3, and thyroid peroxidase (TPO) antibodies using an Immulite solid-phase enzyme-labeled chemiluminescent competitive immunoassay (Siemens Healthcare Diagnostics) at the Irving Institute Biomarker core at CUIMC. The reference ranges, coefficients of variation, and international system (SI) units for each test are shown in Supplemental Table 1. Thyroid function assays for NHANES samples were performed at a designated central laboratory^3,20. The percentile curves for TSH levels in NHANES youth³ were used to define subclinical hypothyroidism as TSH between 4.5 – 10 uIU/L.

MASLD histology

The formulation and validation of histological criteria for NAS used by NASH CRN have been previously described²¹. Briefly, NAS was defined as the unweighted sum of the histological scores for steatosis (grade 0-3), lobular inflammation (0-3), and ballooning (0-2), resulting in a total score between 0 and 8. The inclusion criterion for CyNCh was NAS ≥ 4, whereas the average NAS for TONIC participants was 4.6. Fibrosis severity, a result of chronic disease activity, is not included as a component of the NAS, but is staged separately into four stages: stage 0 (no fibrosis); stage 1 (zone 3 perisinusoidal fibrosis or portal/periportal fibrosis), stage 2 (zone 3 perisinusoidal fibrosis and portal/periportal fibrosis), stage 3 (bridging fibrosis), and stage 4 (cirrhosis). Stage 1 was further subdivided into 1a (mild/delicate zone 3 perisinusoidal fibrosis), 1b (moderate/dense zone 3 perisinusoidal fibrosis), and 1c (portal/periportal fibrosis only).

The NASH CRN histology committee developed consensus criteria for definite NASH (now termed MASH), borderline NASH (now MASH), and NAFLD without NASH (now MASLD without MASH) based on the aggregate presence and degree of individual MASLD histological features²¹. A typical set of minimum criteria for the diagnosis of definite MASH includes steatosis, lobular inflammation, and hepatocellular ballooning.

BMI z-score calculation

The BMI z-score was calculated using the lambda mu sigma method by the Centers for Disease Control 2000 growth charts, which include extended BMI calculations²².

Defining pediatric MASLD

In addition to the biopsy confirmation obtained from the NASH CRN histological criteria mentioned above, the pediatric cardiometabolic criteria from the NAFLD Nomenclature Consensus Group were used to define MASLD¹⁶.

Exclusions

Participants with a history of thyroid disease, thyroid supplements, or drugs that influence thyroid function, such as steroids, lithium, amiodarone, or beta-blockers, were excluded. Additionally, individuals who had clinical hypothyroidism at the time of the assay (TSH > 10 IU/L), hyperthyroidism (total T4 > 13.2 ng/mL or TSH < 0.4 IU/L) or had TPO antibodies above the assay reference range were excluded.

Statistical analysis

The data from the two NASH CRN trial cohorts were grouped together for analysis, as were the data from the five NHANES cycles for population-based controls. Summary statistics were derived based on the distribution of variables.

Thyroid assays were compared between the NHANES and baseline NASH CRN cohorts using the Wilcoxon rank-sum test or Pearson chi-square test. TSH levels were compared across steatosis grades and fibrosis stages in the baseline NASH CRN samples using the Kruskal‒Wallis test. The chi-square test of independence was used to assess the relationship between SH and the baseline MASLD histological measures. Multinomial regression was used to assess the relationship between the baseline and categorical MASLD histological measures. In the base model, age, race, sex, and BMI z-scores were used as covariates. In two alternate models, the homeostasis model of assessment-insulin resistance (HOMA-IR) and non-high-density lipoprotein cholesterol (non-HDL-C) levels were added to the baseline model. The model fit was tested using the log-likelihood metric, likelihood ratio chi-square test, and pseudo-R-square.

Linear mixed-effect models (LMMs) were used to assess the relationship between changes in log TSH and histological MASLD scores over time. The fixed effects included age, sex, race/ethnicity, treatment, and follow-up time with the participant as a random effect. The BMI z-score, HOMA-IR, and non-HDL-C levels were additional fixed effects in the alternate models. Akaike Information Criterion, Bayesian Information Criteria, and likelihood ratio tests were used to assess the model fit. Statistical significance was set at p value < 0.05. Statistical analyses were performed using SAS version 9.4, M8, and R version 4.3.1.

Results

Study populations

NASH CRN cohort

Baseline observations and serum samples were obtained from 281 youth. Individuals with TSH > 10 mIU/L (n=3), TPO antibodies ≥ 35 IU/mL (n=47), and medications related to the thyroid (n = 13) were removed, resulting in a total of 218 unique participants with a total of 421 observations (Supplementary Figure 1a). In addition to biopsy-proven steatohepatitis, all 218 participants had a BMI ≥ 85^th percentile, meeting the pediatric consensus criteria for MASLD. At baseline, 67 (30.7%) participants had definite MASH, 101 (46.3%) had borderline MASH, and 50 (22.9%) had MASLD, but not MASH. The median NAS in this group was 5 (interquartile range 4-6). There were 42 (19%) participants with NAS < 4 and 22 (10%) with NAS > 6.

NHANES cohort

Among the 3,115 children enrolled in the five NHANES cycles, TSH or ALT data were missing for 321. Individuals with TSH > 10 mIU/mL (n=5), total T4 > 13.2 ng/mL (n=10), positive TPO antibodies (n = 156), thyroid-related medications (n=17), and high ALT levels (n = 407) were excluded, resulting in a total of 2,198 population-based controls (Supplementary Figure 1b).

Demographic and biochemical comparison of the NHANES and NASH CRN cohorts

Table 1 presents the demographic and biochemical characteristics of the participants in each cohort. The NASH CRN participants were younger, had a greater percentage of males and Hispanic individuals, and had a higher BMI z-score than the NHANES cohort. TSH levels and the prevalence of SH were higher in the NASH CRN cohort as were the biochemical cardiometabolic parameters associated with obesity. Supplemental Figure 2 compares the distribution of the thyroid assay results in the two cohorts. The differences in TSH levels remained significant (p < 0.001) by analysis of covariance (ANCOVA) after adjusting for age, sex, race/ethnicity, and BMI z-score. The odds of SH in the NASH CRN cohort were 2.53 (95% CI 1.10-5.81, p value = 0.03) compared to those in the NHANES after adjusting for the same covariates.

Table 1.

Baseline Characteristics of Study Population

Characteristic	NASH CRN, N = 2181	NHANES, N = 2,1981	p-value2
Age(years)	12.87 (11.40, 15.16)	15.33 (13.67, 17.00)	<0.001
Gender			<0.001
Male	167 (77%)	1,083 (49%)
Female	51 (23%)	1,115 (51%)
Race/Ethnicity			<0.001
NHW	50 (23%)	627 (29%)
NHB	4 (2%)	641 (29%)
Other	5 (2%)	141 (6%)
Hispanic	159 (73%)	789 (36%)
BMI z-score	2.33 (2.04, 2.51)	0.53 (−0.22, 1.31)	<0.001
% of BMI 95pct	127 (113, 140)	79 (70, 92)	<0.001
Free T4 (ng/dL)	1.11 (1.01, 1.22)	0.80 (0.70, 0.90)	<0.001
Total T4 (ug/dL)	8.30 (7.31, 9.28)	7.60 (6.70, 8.50)	<0.001
Total T3 (ng/dL)	160 (136, 177)	129 (115, 145)	<0.001
TSH (uIU/mL)	2.41 (1.68, 3.19)	1.39 (0.94, 1.98)	<0.001
Subclinical Hypo	18 (8.3%)	33 (1.5%)	<0.001
TBG (ug/mL)	28 (24, 36)	9 (6, 14)	<0.001
ALT (IU/L)	87 (67, 143)	15 (13, 18)	<0.001
AST (IU/L)	53 (40, 80)	22 (19, 26)	<0.001
Cholesterol (mg/dL)	168 (145, 188)	154 (138, 173)	<0.001
LDL-C (mg/dL)	102 (82, 121)	84 (71, 101)	<0.001
Triglycerides (mg/dL)	133 (96, 176)	72 (53, 102)	<0.001
HDL-C (mg/dL)	38 (32, 44)	49 (43, 58)	<0.001
HOMA-IR	6.39 (4.27, 9.56)	2.65 (1.88, 3.91)	<0.001

¹Median (IQR); n (%)

²Wilcoxon rank sum test; Pearson’s Chi-squared test; Fisher’s exact test

Abbreviations: NHW, Non-hispanic white; NHB, Non-hispanic black; % of BMI 95pct, percent of BMi 95^th percentile for age and sex; TBG, thyroid binding globulin; Subclinical Hypo, subclinical hypothyroidism; HOMA-IR, homeostatic index of insulin resistance.

Associations between baseline TSH levels and MASLD histology

Figure 1 shows the trend in TSH levels across the steatosis grades and stages of fibrosis. There was a statistically significant difference between the lowest and highest steatosis grade as well as fibrosis score, suggesting a gradient in TSH levels across severity, albeit with a threshold effect such that there is no statistical difference between steatosis grade 2 and 3 and fibrosis scores above 1a. There was no difference in the demographic or biochemical features according to the steatosis grade, as shown in Supplementary Table 2. At baseline, 18 individuals had SH. There was a significant relationship between SH and histologically borderline or definite MASH χ² (1, N = 218) = 4.5, p = 0.03, and marginally with fibrosis χ² (5, N = 218) = 4.5, p = 0.07, but not with other histological metrics. In multinomial regression models, log TSH was associated with grade 2 (OR = 2.60 [95% CI 1.09, 6.22], p value = 0.03) and grade 3 steatosis (OR = 2.39 [95% CI 1.08, 5.30], p value = 0.03) compared to grade 1 after adjusting for covariates. This relationship remained significant after adjusting for age, race, sex, BMI z-score, HOMA-IR and non-HDL levels. No association was detected between log TSH and the other histological measures.

Figure 1.

TSH levels across groups by MASLD histological characteristics: a. by steatosis grades, and b. by fibrosis levels

Associations between TSH levels and longitudinal changes in MASLD histology

In longitudinal analyses, TSH levels increased with increasing steatosis grade over time (Supplementary Figure 3). Linear mixed-effects models demonstrated an association between log TSH and steatosis grade and NAS (Table 2). The treatments used in the clinical trials (Cysteamine Bitartrate, Metformin and Vitamin E) were not statistically significant. While the BMI z-score was not significant, the interaction between time and BMI z-score was positively associated with steatosis grade (β = 0.01, 95% CI 0.01-0.02, p value < 0.001). The addition of HOMA-IR or non- HDL-C to the multivariable models did not change the steatosis grade results. However, the association with NAS was attenuated after the addition of the BMI z-score to the model.

Table 2.

Association of TSH levels with steatosis and NAS, and subclinical hypothyroidism with fibrosis.

	Steatosis			NAS			Fibrosis*
	n	Estimate (95% CI)	p-value	n	Estimate (95% CI)	p-value	n	Estimate (95% CI)	p-value
Model 1	420	0.25 (0.06, 0.43)	0.008	420	0.32 (0.02, 0.63)	0.04	421	0.44 (0.11, 0.77)	0.01
Model 1 + BMI z-score	418	0.24 (0.06, 0.42)	0.009	418	0.23 (−0.07, 0.54)	0.14	419	0.46 (0.13, 0.79)	0.006
Model 1 + BMI z-score + non-HDL-C	417	0.24 (0.06, 0.42)	0.008	417	0.25 (−0.06, 0.55)	0.12	418	0.46 (0.13, 0.79)	0.006
Model 1 + BMI z-score + HOMA-IR	413	0.23 (0.05, 0.41)	0.01	413	0.19 (−0.11, 0.50)	0.22	414	0.48 (0.15, 0.81)	0.004

Model 1: Predictors: log(TSH), gender, race, age, treatment and follow-up time with subject as the random variable.

^*In fibrosis as the outcome, model 1 predictors were subclinical hypothyroidism (binary) with the covariates as above.

Abbreviations: NAS, Non-alcoholic fatty liver disease Activity Score

SH was associated with a change in fibrosis compared with that in the euthyroid group (Table 2). SH or log TSH were not associated with changes in lobular inflammation, chronic portal inflammation, or ballooning score on histological examination. No association was observed between the measured total or free T4 levels and the histological outcomes, and the addition of these parameters to the TSH models did not change the results (data not shown).

Discussion

We found that TSH levels were higher in youth with biopsy-proven MASLD than in population-based controls after adjusting for age, sex, race/ethnicity, BMI z-score, HOMA-IR, and non HDL-C levels. To our knowledge, this is the first pediatric study to evaluate the relationship between TSH levels and histological features of MASLD in a large cohort of children. At baseline, we observed an association between TSH level and steatosis, as well as SH and histologically borderline or definite MASH. We demonstrated a relationship between increasing TSH level and changes in steatosis and between SH and changes in fibrosis over time. As expected, there was an association between increased steatosis and BMI z-score over time.

Thyroxine regulates metabolism and hypothyroidism has been identified as a risk factor for MASLD^23,24. The active thyroid hormone triiodothyronine (T3) is involved in hepatic gluconeogenesis and hepatic insulin resistance, and likely plays a role in the pathogenesis of MASLD²⁵. T3 supplementation and thyroid hormone receptor-β (THRβ) agonists reverse or prevent MASLD in murine models of experimentally induced MASLD^26-28, and treatment with THRβ agonists is a promising therapy for MASLD²⁹.

Elevated TSH level is the most sensitive indicator of hypothyroidism. Epidemiological studies have shown an association between high TSH levels and MASLD in both adults and children. In a population-based prospective Rotterdam study in adults, Bano et al reported 49% greater odds (OR 1.49, 95% CI 1.04-2.15) of clinically relevant MASLD or fibrosis (measured by transient elastography) at follow-up of median 10 years in those with higher TSH levels, while adjusting for the relevant confounders³⁰. More recently, Kouvari et al. reported association of SH with significant fibrosis (OR 3.47, 95% CI 1.50-8.05) and MASLD (OR 3.44, 95% CI 1.48 – 7.98) in 677 adults from three centers with biopsy proven steatohepatitis, albeit without unified histological criteria across the included centers³¹. In cross-sectional studies of children with and without obesity, we and others have shown an association between higher TSH levels and SH with MASLD diagnosed by ALT levels, ultrasound, elastography or histology^9-11,13,32. Due to the lack of liver biopsy data, these studies were unable to establish a relationship with the severity of the histological changes. In the current study, we also identified higher odds of SH in children with MASH. Additionally, we demonstrate an association between TSH levels and the histological severity of steatosis and fibrosis. Unlike the Rotterdam study, we found no association between free T4 levels, either at baseline or in longitudinal analysis, and the severity of histological changes. This could be due to the younger age of our cohort with varied ethnicities and a pre-existing diagnosis of MASLD. Furthermore, individuals with TSH levels > 10 mIU/mL were excluded from our study and the follow-up period was shorter. It is also possible that over time, persistently higher TSH levels in the SH range in individuals with MASLD will be accompanied by the development of lower circulating free T4 levels as noted in Rotterdam study. These hypotheses need validation in future prospective longitudinal studies in children with MASLD.

This study showed an association between TSH levels and MASLD severity. It is difficult to separate the role of TSH from that of circulating thyroid hormones in humans because of the intricate feedback loop. Rodent studies have identified TSH receptors on hepatocytes that induce hepatic steatosis via sterol regulatory element binding protein (SREBP1C) with specific TSH-induced stimulation³³. TSH upregulates genes involved in hepatic gluconeogenesis (phosphoenolpyruvate carboxykinase and glucose 6 phosphatase) via c-AMP-regulated transcriptional coactivators³⁴. TSH also suppresses the synthesis of hepatic bile acid via the SREBP2-hepatocyte nuclear factor 4α–CYP7A1 signaling pathway³⁵. TSH inhibits cholesterol synthesis by increasing AMP kinase-mediated phosphorylation of hydroxy methyl glutaryl Co-A reductase³⁶. Collectively, these findings support the notion that TSH can regulate both hepatic lipid and cholesterol homeostasis. Future studies on tissue-specific activation of TSH receptors limited to the liver, without changing circulating thyroid hormone levels, may clarify the independent role of TSH. Since adequate circulating thyroid hormones suppress TSH production via a negative feedback loop, our results and those of rodent studies suggest that part of the therapeutic benefit of THRβ agonists may be derived from lowering TSH levels. Collectively, the results of our study and existing literature on the relationship of thyroid function and MASLD emphasize the need for assessment of TSH in adults and children with MASLD, and appropriate treatment of hypothyroidism. Further study is needed to determine the optimal frequency of monitoring thyroid function and whether subclinical hypothyroidism should be treated.

The strength of this study is the use of a large cohort of children with a well-characterized histological spectrum of MASLD, paired thyroid assays, and standardized liver biopsy readings at two time points. We included a large population-based pediatric control group from the NHANES. While the NHANES data were collected in a similar time frame representative of the US population and included oversampling of the Latino/Hispanic population, the NASH CRN cohort had a preponderance of Hispanic males with a higher BMI z-score, as is commonly noted in epidemiological studies of MASLD. The statistical models were adjusted for age, race/ethnicity, and BMI z-score to account for the variance from these differences. The NASH CRN cohort was derived from children seen in Hepatology and Gastroenterology clinics with a histologically confirmed diagnosis of MASLD. The entry criteria for the clinical trials required either an elevated ALT level (TONIC) or a NAS > 4 (CyNCh) introducing potential ascertainment bias. However, in total, nearly 20% of the children in the NASH CRN cohort had NAS < 4, ensuring that milder histological cases of MASLD were included. The thyroid assay platforms for the two cohorts were different, and future studies should replicate these findings using a uniform platform. Although we excluded children with elevated ALT levels from the control group, we did not have liver biopsy data to definitively exclude MASLD. In a subset of youth in the 2017–2018 NHANES data, a small proportion (4%) had increased liver stiffness measured by transient elastography, which could be indicative of significant fibrosis. Of these, 21% had elevated ALT, suggesting the presence of a small proportion of youth with clinically silent liver disease in our control group³⁷. Thus, our population-based control group may have some individuals with silent MASLD and potentially higher TSH levels biasing the results towards null. Nonetheless, we identified a strong positive relationship between TSH, SH, and MASLD, emphasizing the strength of the association. Despite these limitations, this study adds to the growing body of literature on the relevance of thyroid function in pediatric MASLD and provides evidence to support the use of THRβ agonists in youth.

Conclusions

This study confirmed that TSH levels were higher in children with MASLD than in those with normal ALT levels and were associated with steatosis severity. We found an association between changes in TSH levels over time and increased steatosis, while the association with NAS was attenuated after adjusting for BMI z-score. We also identified an association between subclinical hypothyroidism and MASH score at baseline and with changes in the fibrosis stage over time. The results of this study support the association of TSH with MASLD severity, that may partially account for the success of THRβ agonist treatment. The authors advocate for regular thyroid function monitoring in youth with MASLD and thyroid supplementation for elevated TSH levels.

Abbreviations:

ALT

alanine aminotransferase

BMI

body mass index

Chol

cholesterol

HbA1c

hemoglobin A1c

HDL-C

high-density lipoprotein cholesterol

LDL-C

low-density lipoprotein cholesterol

HTN

hypertension

MASLD

metabolic dysfunction-associated steatotic liver disease

NAFLD

non-alcoholic fatty liver disease

NAS

NAFLD activity score

pct95

95^th percentile

SH

subclinical hypothyroidism

TG

triglyceride

VLDL-C

very low-density lipoprotein cholesterol