SYSTEMATIC REVIEW article

Front. Endocrinol., 29 October 2025

Sec. Systems Endocrinology

Volume 16 - 2025 | https://doi.org/10.3389/fendo.2025.1660437

Association between thyroid function and metabolic associated fatty liver disease: a systemic review and meta-analysis

  • 1. The First People’s Hospital of Jiande, Hangzhou, China

  • 2. Office of Research and Innovation, The First Affiliated Hospital of Henan University of Science and Technology, Luoyang, China

  • 3. Zhejiang Key Laboratory of Medical Epigenetics, Department of Cardiology, Affiliated Hospital of Hangzhou Normal University, Hangzhou Institute of Cardiovascular Diseases, Engineering Research Center of Mobile Health Management System & Ministry of Education, Hangzhou Normal University, Hangzhou, China

  • 4. Department of Cardiology, Hangzhou Lin’an Fourth People’s Hospital, Hangzhou, Zhejiang, China

Abstract

Background:

Metabolic dysfunction–associated fatty liver disease (MAFLD) is one of the most common chronic liver diseases. The relationship between MAFLD and thyroid function parameters remains controversial.

Aim:

This study aimed to explore the influence of metabolic parameters and thyroid dysfunction on the development of MAFLD and examine the relationship between them in different age groups.

Methods:

The PubMed, Embase, Web of Science, China National Knowledge Infrastructure(CNKI), and Cochrane Library databases were searched. Standardized mean difference (SMD) and odds ratio with a 95% confidence interval (CI) were calculated.

Results:

A total of 36 studies involving 198,254 participants were eligible. Compared with controls, the patients with MAFLD had significantly higher thyroid-stimulating hormone (TSH) levels (MAFLD vs controls: SMD = 0.02, 95% CI = 0.01–0.03); significantly higher free triiodothyronine levels (MAFLD vs controls: SMD = 0.19, 95% CI = 0.18–0.20); significantly lower free thyroxine levels (MAFLD vs controls: SMD = −0.41, 95% CI = −0.42 to −0.40); significantly higher total triiodothyronine levels (MAFLD vs controls: SMD = 0.26, 95% CI = 0.23–0.29); and significantly lower total thyroxine levels (MAFLD vs controls: SMD = −0.10, 95% CI = −0.13 to −0.07).

Conclusions:

The TSH level may be an important risk factor for the occurrence and development of MAFLD. The relationship between them is influenced by age.

1 Introduction

Metabolic dysfunction–associated fatty liver disease (MAFLD) encompasses a range of liver diseases closely related to abnormal metabolic disorders. It is considered a hepatic manifestation of the metabolic syndrome (1). MAFLD is characterized by hepatic steatosis (detected by imaging techniques, blood biomarkers/scores, or liver histology) associated with type 2 diabetes mellitus (T2DM) and overweight/obesity, regardless of alcohol intake or excluding other causes of chronic liver disease (2). The introduction of the new concept of MAFLD aims to include metabolic dysfunction as a diagnostic criterion and covers a larger and more diverse population than non-alcoholic fatty liver disease (NAFLD) (3). A meta-analysis involving 3.3 million people revealed that the global prevalence of MAFLD was 38.77% (4). Compared with the general population, patients with MAFLD experience higher liver-related morbidity or mortality and are strongly associated with extrahepatic diseases such as cancer, cardiovascular events, stroke, chronic kidney disease, and obstructive sleep apnea (5–8).

The rationale for MAFLD is to include metabolic dysfunction as a diagnostic criterion, including abdominal obesity and T2DM. The thyroid gland regulates body weight, lipid-like substance metabolism, and insulin resistance. Therefore, thyroid hormone may be closely related to the occurrence and development of MAFLD (9). However, the relationship between MAFLD and thyroid function still remains controversial. Three published meta-analyses have explored the relationship between NAFLD and thyroid function, but their findings were inconsistent (10–12). The patients with NAFLD/NASH had significantly higher thyroid-stimulating hormone (TSH) levels than controls in adults. Previous studies found no significant difference in thyroid hormone levels between patients with NAFLD and non-NAFLD. This might be mainly attributed to the heterogeneity of sample sizes and patient characteristics. MAFLD covers a much broader spectrum, and no studies demonstrating an association between MAFLD and thyroid function parameters have been reported. Therefore, this study conducted a new systematic review and meta-analysis, including a lot of research and exploration of several thyroid function parameters [free triiodothyronine (FT3), free thyroxine (FT4), total triiodothyronine (TT3), total thyroxine (TT4), and TSH levels].

2 Methods

2.1 Protocol

This systematic review and meta-analysis was conducted based on the meta-analysis of observational studies following epidemiology guidelines.

2.2 Eligibility criteria

The inclusion criteria were as follows: (1) original studies that explored the association between MAFLD, NAFLD, and NASH with thyroid function; (2) studies independent of sex and age; and (3)studies in which MAFLD was diagnosed by histology (liver biopsy), imaging modalities (such as ultrasonography, CT, or MRI), or validated biochemical indices. The exclusion criteria were as follows: (1) reviews and meta-analyses, (2) irrelevant pieces of literature, (3) duplicate studies, (4) experimental studies, (5) editorials/comments and case reports, and (6) original studies including patients with MAFLD/NAFLD/NASH but with competing etiologies for steatosis and coexisting causes for other chronic liver diseases, such as significant alcohol consumption, hepatitis C, and medications.

2.3 Search strategy

We searched the PubMed, Embase, Web of Science, CNKI, and Cochrane Library databases from the inception dates to March 1, 2025, with no restrictions on language. Conference abstracts were hand searched to identify potentially eligible studies. The search strategy was as follows: (“MAFLD” [All Fields] OR “Metabolic associated fatty liver disease” [All Fields]) OR (“NASH” [All Fields] OR “Nonalcoholic Steatohepatitis” [All Fields]) OR (“NAFLD” [All Fields] OR “Non-alcoholic Fatty Liver Disease” [All Fields]) AND (“thyroid” [All Fields] OR “thyroid-stimulating hormone” [All Fields] OR “free thyroxine” [All Fields] OR “free triiodothyronine” [All Fields] OR “total triiodothyronine” [All Fields] OR “total thyroxine” [All Fields]).

2.4 Study selection

Two reviewers (J.T. and C.C.) independently screened the studies based on the titles and abstracts. If the studies were potentially fit, full texts were further retrieved and screened. Disagreements were resolved by consensus.

2.5 Data extraction

Two reviewers (J.T. and C.C.) extracted all the data independently. We extracted the following data from each study: study characteristics, first author, year of publication, country, study design, study population, number of participants, age, MAFLD diagnosis, and thyroid function parameters (TSH, FT3, FT4, TT3, and TT4). If necessary, the reviewers contacted for relevant data.

If some data were expressed as mean with a 95% confidence interval (CI) alone, the methods provided by the Cochrane Handbook were used to calculate the standard deviation (SD). The formula used is as follows:

If some data were expressed as median with range or interquartile range alone, the methods provided by Wan et al. (13) and Luo et al. (14) were used to estimate the mean and SD. All formulas are available on the website http://www.comp.hkbu.edu.hk/∼xwan/MedianRange.html.

2.6 Statistical analysis

The association of MAFLD/NASH/NAFLD with thyroid function parameters (including FT3, FT4, TT3, TT4, and TSH) was assessed. For continuous data, standardized mean difference (SMD) with 95% CI was calculated using the inverse variance statistical method. For dichotomous variable data, the odds ratio (OR) with 95% CI was calculated using the Mantel–Haenszel statistical method. The data were pooled using a random-effects model to obtain a more conservative effect estimate. The I2 statistic and Q test were used to measure the heterogeneity across the included studies, where I2 >50% and P value <0.1 were considered to be of significant heterogeneity. If more than five eligible studies were included in a meta-analysis, the publication bias was assessed using funnel plots. All statistical analyses were performed using Review Manager version 5.2 software.

3 Results

3.1 Study selection

A total of 1246 studies were retrieved, 36 studies were eligible (Figure 1). MAFLD was diagnosed by hepatic steatosis (including imaging techniques, blood biomarkers/scores, or liver histological tests). Patients without MAFLD rather than healthy individuals were randomly selected as controls. Ultimately, 198,254 participants were included in the quantitative analysis.

Figure 1

3.2 Study characteristics

The characteristics of the 36 included studies are depicted(Table 1). Among these, 7 were cohort studies (15–21), 25 were cross-sectional studies (22–46), and 4 were case–control studies (47–50). These clinical trials were conducted globally: 6 in Europe [Germany (23, 26), Italy (18, 47), Netherlands (30), and Spain (39)], 28 in Asia (5 in Turkey (15–17, 27, 29), 2 in South Korea (24, 36), 1 in India (20), 1 in Iran (22), 19 in China [19, 21, 25, 28, 31–35, 37, 38, 40–46)], 2 in Africa [Egypt (48, 50)], and 1 in the United States (49). MAFLD was diagnosed by different methods: through liver biopsies in 5 studies (18, 37, 47, 49, 50), through computed tomography (CT) in 1 study (41), and through ultrasound technology in 31 studies. All the 36 included studies reported detailed data regarding thyroid function parameters (Table 2). All studies reported detailed data on the TSH parameter, 22 studies reported data on the FT3 parameter, 28 on the FT4 parameter, and 6 on the TT3 and TT4 parameters.

Table 1

First author(Year)CountryStudy designNo. participantsEnrollment periodDiagnosis of MALFDThyroid function parameters
Ambrosio 2021ItalyCross-sectional study532014−2016liver biopsyTSH, FT3, FT4
Basarir 2021TurkeyCross-sectional study2082015.02−2017.02ultrasonographyTSH, FT4
Berg 2017NetherlandsCross-sectional study20289NAultrasonographyTSH, FT3, FT4
Bilgin 2014TurkeyCross-sectional study1202010.04−2011.06ultrasonographyTSH, FT3, FT4
Carulli 2013ItalyCase-control study692001−2008liver biopsyTSH, FT3, FT4
Chao 2021ChinaCross-sectional survey554482017.01−2019.02ultrasonographyTSH, FT3, FT4
Chen 2019ChinaCross-sectional study18382013.04−2016.09ultrasonographyTSH, FT3, FT4
Du 2021ChinaCross-sectional study14222014.12−2019.10ultrasonographyTSH, FT3, FT4
Escudé 2020SpainCross-sectional study101162009−2013ultrasonographyTSH
Eshraghian 2013IranCross-sectional study8322011.09−2012.09ultrasonographyTSH, FT3, FT4
Fan 2022ChinaCross-sectional study184272015.01−2021.12ultrasonographyTSH, FT3, FT4, TT3, TT4
Fatma 2016TurkeyCross-sectional study612014.7−2015.01ultrasonographyTSH, FT3, FT4
Feng 2019ChinaCross-sectional study5892015.01−2017.12ultrasonographyTSH, FT3, FT4
Gu 2021ChinaCross-sectional study64622013−2019.ultrasonographyTSH, FT3, FT4
Guo 2021ChinaCross-sectional study34972017.11−2019.12ultrasonographyTSH, FT3, FT4
Gupta 2020IndiaCross-sectional study1002018.09−2019.10ultrasonographyTSH, TT3, TT4
Kaltenbach 2017GermanyCross-sectional study3322000.02−2001.05ultrasonographyTSH, TT3, TT4
Kim 2022KoreaCross-sectional study15892013.01−2015.12ultrasonographyTSH, FT4
Lee 2015KoreaCohort study185442008.01−2012.12ultrasonographyTSH, FT4
Li 2022ChinaCross-sectional study1292017.01−2021.01liver biopsyTSH, FT3, FT4
Liang 2018ChinaCross-sectional study325732016.01−2016.12ultrasonographyTSH, FT4
Liu 2014ChinaCross-sectional study25762013.01−2013.11ultrasonographyTSH, FT3, FT4
Liu 2018ChinaCross-sectional study17732015.01−2015.12ultrasonographyTSH, FT3, FT4
LIU 2019ChinaCross-sectional study156602013.01−2018.10ultrasonographyTSH, FT3, FT4
Ludwig 2015GermanyCross-sectional study1276NAultrasonographyTSH, TT3, TT4
Moustafa 2009EgyptCase-control study33NAultrasonographyTSH, TT3, TT4
Naguib 2021EgyptCase-control study100NAliver biopsyTSH, FT4
Pagadala 2012USACase-control study6632006.10−2009.06Liver biopsyTSH
Pan 2019ChinaCross-sectional study1292013−2014ultrasonographyTSH
Sert 2013TurkeyCross-sectional study1002011.04−2012.04ultrasonographyTSH, FT4
Shao 2022Chinacross-sectional study1152019.01−2021.10computed tomography (CT)TSH, FT3, FT4, TT3, TT4
Tao 2014Chinacross-sectional study7392012.09−2013.02ultrasonographyTSH, FT3, FT4
Torun 2014TurkeyCross-sectional study722011.01−2013.02ultrasonographyTSH, FT3, FT4
Xu 2011ChinaCross-sectional study8782008.01−2008.12ultrasonographyTSH, FT3, FT4
ZHANG 2012ChinaCross-sectional study(female)7772010.01 to 2010.04ultrasonographyTSH
Cross-sectional study
(male)
5452010.01 to 2010.04ultrasonographyTSH
Zhang 2022ChinaCross-sectional study1202016.01 to 2022.09ultrasonographyTSH, FT3, FT4

Characteristics of included studies.

Table 2

First author(Year)GroupsAgeNo. PtsTSH (mIU/l)FT3 (pmol/l)FT4 (pmol/l)TT3 (nmol/l)TT4 (nmol/l)
Ambrosio 2021NAFLD58±28.9282.38±26.895.2±2.0716.3±11.93NANA
NO-NAFLD52±32.59252.17±7.825.6±2.2916.2±4.59NANA
Basarir 2021NAFLD12.6±2.6942.5±1.3NA13±3.9NANA
NO-NAFLD10.8±2.71142.3±1.2NA13±3.9NANA
Berg 2017NAFLD46±8.8942742.04±0.845.3±0.5215.4±1.85NANA
NO-NAFLD42±11.85160152.01±0.865.2±0.5215.7±1.78NANA
Bilgin 2014NAFLD12.68±2.13803.04±1.436.67±1.0516.34±2.45NANA
NO-NAFLD11.90±2.52402.75±1.255.71±0.7917.63±2.06NANA
Carulli 2013NAFLD44.41 ±11.43442.21±0.955.21±0.6913.44±2.95NANA
NO-NAFLD44.40 ±11.38251.67±0.745.1±0.5112.83±2.42NANA
Chao 2021NAFLD48.8 ±9.79190031.85±1.554.13±1.2926.13±1.92NANA
NO-NAFLD46.0 ±10.7364451.88±1.673.99±1.3928.34±2NANA
Chen 2019NAFLD57.3±6.49142.64±0.834.96±0.5115.54±1.72NANA
NO-NAFLD56.8±6.69242.61±0.814.85±0.4715.56±1.68NANA
Du 2021MAFLD53.07±13.328362.05±1.174.21±1.516.42±7.46NANA
NO-MAFLD57.28±12.135862.01±1.214.5±1.0116.69±10.95NANA
Escudé 2020NAFLD60±1067902.3±1.9NANANANA
NO-NAFLD61±1033262.2±1.9NANANANA
Eshraghian 2013NAFLD48.20±12.821272.02±1.353.8±0.7315.84±2.76NANA
NO-NAFLD36.97 ±18.767052.29±1.474.09±1.6416.51±5.63NANA
Fan 2022MAFLD51.09±15.7849521.98±0.94.66±0.513.98±1.871.63±0.2494.39±15.91
NO-MAFLD45.59±16.13134751.95±0.94.5±0.5114.11±1.871.56±0.2495.53±15.84
Fatma 2016NAFLD49.78 ±12.77361.71±0.964.39±0.89.27±2.57NANA
NO-NAFLD46.52 ±13.80251.5±0.834.48±0.5210.3±2.06NANA
Feng 2019NAFLD64.42±13.682172.46±0.994.76±0.5616.64±1.94NANA
NO-NAFLD72.60±14.563721.76±0.754.25±0.6416.86±1.99NANA
Gu 2021MAFLD51.50±7.9016751.93±0.875.05±0.6215.73±2.21NANA
NO-MAFLD51.0±8.4847872.05±0.94.9±0.6215.47±2.21NANA
Guo 2021NAFLD49.49 ± 9.8621732.2±0.845.08±0.5917.06±1.95NANA
NO-NAFLD49.28 ±10.4613242.25±0.834.77±0.4816.96±1.98NANA
Gupta 2020NAFLD10.52±2.2623.93±1.4NANA2.22±0.3882.27±31.18
NO-NAFLD9.32±2.4384.47±1.8NANA8.49±2.682.5±25.74
Kaltenbach2017NAFLD14.1±1.9992.8±1.1NANA1.7±0.4103±18.0
NO-NAFLD13.9±1.82332.5±1.4NANA1.6±0.3103±18
Kim 2022NAFLD55±7.413782.3±1.78NA15.47±1.95NANA
NO-NAFLD56 ±7.4112112.21±1.13NA15.6±1.95NA
Lee 2015NAFLD39.2±5.923482.3±4.3NA16.47±2.06NANA
NO-NAFLD37.8±5.7161962.3±2.6NA16.21±1.93NANA
Li 2022NAFLD51.5±14.07402.87±1.275.11±0.4915.67±2.18NANA
NO-NAFLD43.0 ±20.74892.11±1.164.74±0.7815.75±2.23NANA
Liang 2018NAFLD48.8±13.451832.95±1.49NA16.32±2.01NANA
NO-NAFLD43.23±13.7273902.86±1.48NA16.47±1.93NANA
Liu 2014NAFLD46.90± 9.419882.13±0.95.12±0.5816.41±2.04NANA
NO-NAFLD44.42±10.7415882.2±0.934.84±0.5816.18±2.06NANA
Liu 2018NAFLD50.42±8.906381.67±0.884.63±0.5512.79±1.33NANA
NO-NAFLD49.30±11.3611351.58±0.934.43±0.5312.71±1.48NANA
LIU 2019NAFLD49.3±10.270951.65±1.214.51±0.713.07±1.98NANA
NO-NAFLD47.3±11.985651.6±1.14.35±0.6813.27±2NANA
Ludwig 2015NAFLD47.7 ± 11.53491.8±1.4NANA1.6±0.383.2±15.6
NO-NAFLD38.0 ± 12.19271.8±3.5NANA1.6±0.392±17.4
Moustafa 2009NAFLD55±4.6132.1±0.75NANA1.84±0.27100.39±14.8
NO-NAFLD51±9.3201.75±0.9NANA1.89±0.3894.72±15.44
Naguib 2021NAFLD47.9 ±3503.3±1.2NA13±3.9NANA
NO-NAFLD45.7 ±4502.2±0.8NA15.6±3.9NANA
Pagadala 2012NAFLD50.4±11.12332.27±1.33NANANANA
NO-NAFLD51.0 ±14.14301.7±0.67NANANANA
Pan 2019NAFLD12.47 ± 3.24176.17±9.29NANANANA
NO-NAFLD11.97 ± 3.391124.01±3.4NANANANA
Sert 2013NAFLD13.13 ± 1.26583.56±1.34NA15.44±1.93NANA
NO-NAFLD13.65 ± 1.4421.77±0.63NA16.09±1.93NANA
Shao 2022NAFLD44±15.56811.77±0.875.13±0.6711.38±1.971.52±0.27105.3±16.12
NO-NAFLD34±12.59341.82±1.134.93±0.4413.01±1.761.54±0.23106.62±19.75
Tao 2014NAFLD48.9± 8.41961.96±1.074.86±0.7316.87±1.77NANA
NO-NAFLD47.5± 8.15431.81±0.924.88±0.5717.5±1.87NANA
Torun 2014NAFLD12.2 ±3.6284.2±3.95.7±1.112.9±4.4NANA
NO-NAFLD12.8 ±1.7442.2±1.435.4±0.913.2±2.3NANA
Xu 2011NAFLD71.2±3.82272±0.934.6±0.4411.12±1.43NANA
NO-NAFLD71.9±4.26511.82±0.884.58±0.4911.58±1.47NANA
ZHANG 2012(1)NAFLDNA1293.26±1.33NANANANA
(female)NO-NAFLDNA6482.98±1.25NANANANA
ZHANG 2012(2)NAFLDNA1372.56±1.16NANANANA
(male)NO-NAFLDNA4082.34±1.08NANANANA
Zhang 2022MAFLD57±8.89471.78±1.075.04±0.5817±2.59NANA
NO-MAFLD55±12.59731.96±1.334.68±0.5916.9±2.89NANA

Table of thyroid function indicators.

NA, not applicable.

3.2.1 Thyroid-stimulating hormone

A total of 36 studies with 198,254 patients were included in the meta-analysis. The patients with MAFLD had significantly higher levels of TSH than controls (SMD = 0.02; 95% CI = 0.01, 0.03; P < 0.001) (Figure 2). Heterogeneity was statistically significant (I2 = 89%, P < 0.001). The publication bias was not statistically significant (Figure 3). Meanwhile, it was found that compared with the controls, the patients with MAFLD had significantly higher levels of TSH in the children/youth group according to the age subgroup analysis (SMD = 0.32, 95% CI = 0.19, 0.045, P < 0.001); no significant difference was observed in the level of TSH in the middle-aged group (SMD = 0.01, 95% CI = 0.00, 0.02, P = 0.02); and significantly higher levels of TSH was observed in the elderly group (SMD = 0.10, 95% CI = 0.06, 0.14, P < 0.001).

Figure 2

Figure 3

3.2.2 Free triiodothyronine

A total of 22 studies with 121,169 patients were included in the meta-analysis. The patients with MAFLD had significantly higher levels of FT3 than controls (SMD = 0.19, 95% CI = 0.18, 0.20, P < 0.001) (Figure 4). Heterogeneity was statistically significant (I2 = 89%, P < 0.001). Compared with the controls, the patients with MAFLD had significantly higher levels of FT3 in the children/youth group (SMD = 0.70, 95% CI = 0.39, 1.01, P < 0.001); the middle-aged group (SMD = 0.19, 95% CI = 0.18, 0.2, P < 0.001); and the elderly group (SMD = 0.38, 95% CI = 0.27, 0.50, P < 0.001).

Figure 4

3.2.3 Free thyroxine

A total of 28 studies with 184,283 patients were included in the meta-analysis. The patients with MAFLD had significantly lower levels of FT4 than controls (SMD = −0.41, 95% CI = −0.42, −0.40, P < 0.001) (Figure 5). Heterogeneity was statistically significant (I2 = 100%, P < 0.001). Compared with controls, the patients with MAFLD had significantly lower levels of FT4 in the children/youth group (SMD = −0.20, 95% CI = −0.38, −0.02, P < 0.03); the middle-aged group (SMD = −0.41, 95% CI = −0.42, −0.40, P < 0.001); and the elderly group (SMD = −0.22, 95% CI = −0.34, −0.11, P < 0.001).

Figure 5

3.2.4 Total triiodothyronine

A total of 6 studies with 20,283 patients were included in the meta-analysis. The patients with MAFLD had significantly higher levels of TT3 than controls (SMD = 0.26, 95% CI = 0.23, 0.29, P < 0.001) (Figure 6). Heterogeneity was statistically significant (I2 = 97%, P < 0.001).

Figure 6

3.2.5 Total thyroxine

A total of 6 studies involving 20,283 patients were included in the meta-analysis. The patients with MAFLD had significantly lower levels of TT3 than controls (SMD = −0.10; 95% CI = −0.13, −0.07; P < 0.001) (Figure 7). Heterogeneity was statistically significant (I2 = 90%, P < 0.001).

Figure 7

4 Discussion

A total of 36 studies with 198,254 participants were included in the systematic review and meta-analysis. The findings revealed the following. (1) Increased TSH levels were significantly associated with a higher risk of MAFLD. However, no significant difference in the TSH level was found between the patients with MAFLD and the controls in the middle-aged population. In the adolescent and elderly population, the TSH level of patients with MAFLD significantly differed from that of the controls. (2) The increased levels of FT3 and TT3 were significantly associated with the risk of MAFLD in all populations. (3) The decreased levels of TT4 and FT4 were significantly associated with the risk of MAFLD in all populations.

The “second strike” hypothesis has been widely considered to be the classical pathogenesis of MAFLD. The progressive accumulation of lipids in the cytoplasm of hepatocytes triggers a cytotoxic event that triggers an inflammatory response. The liver, being the central organ regulating endocrine function and maintaining triglyceride metabolism, plays a crucial role. Abnormal triglyceride accumulation in hepatocytes forms fatty liver. Some studies have shown a positive correlation between TSH and triglyceride concentrations (51, 52), even within the normal reference range (53). TSH might play a key role in regulating triglycerides and cholesterol, thus promoting the development of MAFLD (54).

In terms of pathophysiology, the presence of thyroid hormones can contribute to the development of MAFLD through multiple pathways. Thyrotropin hormone regulates the thyrotropin receptor (TSHR) through highly specific interactions with thyroxine. TSHR is expressed primarily on the cell membrane of thyroid follicular cells and is also expressed to varying degrees in numerous extrathyroidal tissues and cells, such as liver, kidney, bone, adipose tissue, and fibroblasts (55, 56). The binding of TSH to TSHR on hepatocytes can increase triglyceride levels, leading to hepatic steatosis. Thus, the potential mechanism of TSH-induced hepatic triglyceride accumulation involves the binding of TSH to its receptor TSHR, which then triggers hepatic Sterol Regulatory Element Binding Protein-1c(SREBP-1c)activity via the cAMP/PKA/PPARa pathway, resulting in reduced AMPK activity and increased expression of genes associated with insulin resistance and adipogenesis (57). Meanwhile, another study showed that TSH could be involved in intrahepatic lipolysis by the activation of autophagy and β-oxidation of fatty acids (58). Thus, elevated TSH levels indicated hypothyroidism and reduced activity of hepatic lipase, leading to fat cell accumulation in hepatocytes (59). A study demonstrated that foreign bodies from TSH-stimulated hepatocytes increased and showed specific protein profile changes, many of which were involved in metabolism, signal transduction, apoptosis, and inflammation (60). TSH is also associated with the regulation of microrne (61). Adipokine levels were altered in patients with hypothyroidism (62). Recently, a high TSH level was significantly correlated with NASH in individuals with normal thyroid function and NAFLD confirmed by biopsy (63). In this study, TSH increased the risk of MAFLD significantly. However, the results of this study were controversial. The subgroup analysis in this study showed no significant association between TSH and MAFLD in the middle-aged population, but found an association between TSH and MAFLD in adolescents and older adults.

A feedback loop exists between TSH and thyroid gland in the body. An increase in the level of TSH usually suggests a low thyroid function (i.e., hypothyroidism) (12). Ferrandino et al. demonstrated the development of NAFLD in mice with mild hypothyroidism without the downregulation of hepatic TH signaling or reduced hepatic lipid utilization. This condition led to an increased shuttling of fatty acids (FAs) to the liver, where they were esterified and accumulated as triglycerides. On the contrary, mice with severe hypothyroid exhibited a downregulation of hepatic TH signaling and a severe inhibition of adipose tissue lipolysis, which reduced the delivery of FAs to the liver. The resulting lack of triglyceride esterification substrates protected mice with severely hypothyroid against NAFLD (64). We speculated that MAFLD was induced by insufficient release of TSH in adolescents and the elderly, when the thyroid was immature or in the process of degeneration, respectively. In adults, the pituitary-thyroid axis may be mature enough to maintain adequate TSH secretion, which helps regulate lipid metabolism in the liver.The association between TSH and MAFLD might be influenced by age. Selin et al. demonstrated that children with metabolic disease had lower levels of FT4 and higher levels of TSH than normal children (65). Gu et al. found that FT3 was a higher risk factor for NAFLD than those without NAFLD, while FT4 and TSH were not significantly correlated with NAFLD in middle-aged and elderly people (age >40) (46). This also explained why such a wide divergence was observed in the existing studies, with some suggesting no association and others suggesting an association. This might depend on the age distribution of the population included in previous studies and the differences in TSH concentrations.

The thyroid hormone receptor (TR) is of two types, TRα and TRβ, and it is the main receptor in the liver (66). Studies have shown that M1 macrophages polarizably induce secreted phosphoprotein 1 (SPP1) secretion and downregulate hepatocyte TRβ in a paracrine manner, exacerbating lipid deposition in the liver and compensating for increased serum TSH. Increased levels of TSH can lead to SPP1 secretion by M1 macrophages. The positive feedback interaction between the thyroid and liver may play a significant role in maintaining and amplifying the pathological process of MAFLD (67). Therefore, TRβ agonists have been investigated as potential therapies for serum dyslipidemia and MAFLD (68). Chaves et al. demonstrated that mutations in the THR-beta gene of TR can induce that signal transduction damage in the liver, leading to hepatic steatosis. This indicates the influence of thyroid hormones on lipid metabolism in the liver (69). Our findings supported the effect of thyroxine on MAFLD. This study found that elevated levels of FT3 and TT3 were significantly associated with the risk of MAFLD, and decreased levels of TT4 and FT4 were significantly associated with the risk of MAFLD. Further clinical studies are needed to investigate whether TSH affects FT3, FT4, TT3, and TT4 and thus MAFLD.

In addition, recent evidence suggests that thyroid dysfunction may not only be associated with hepatic steatosis but also with ectopic fat deposition outside the liver. For example, Bayyigit et al. (Diabetes Metab Res Rev, 2024) reported that both hypothyroidism and subclinical hypothyroidism were significantly associated with an increased prevalence of non-alcoholic fatty pancreas disease (NAFPD) as well as MAFLD. Their findings indicate a potential bidirectional relationship between thyroid dysfunction and ectopic fat accumulation, further underscoring the need for future studies to comprehensively investigate the interplay among hypothyroidism, MAFLD, and NAFPD.

In conclusion, TSH plays a key role in the development and progression of MAFLD and may be influenced by age, which needs to be further elucidated by more clinical trials in different age groups. As thyroid replacement therapy improves elevated TSH levels, whether treating hypothyroidism in patients with MAFLD improves liver function or disease progression outcomes should be further elucidated by placebo-controlled clinical trials.

This study had some limitations. First, heterogeneity was significant in some meta-analyses. We tried to explore heterogeneity between studies through sensitivity and meta-regression analyses and pooled data through random-effects models to obtain more conservative effect estimates. Second, several noninvasive methods [e.g., elevated liver enzymes, fatty liver index, ultrasonography, CT magnetic resonance imaging, and spectroscopy] were commonly used to diagnose MAFLD because liver biopsy was not feasible in the general population (70). While liver biopsy remains the diagnostic gold standard, non-invasive approaches such as the Hepatic Steatosis Index or FibroScan, although practical, may introduce a degree of misclassification bias compared with histology, and this potential bias should be considered when interpreting our results.Third, we did not perform a subgroup analysis of TSH concentration because the upper normal limit of TSH concentration had long been controversial, and the diagnosis of subclinical hypothyroidism depended crucially on the definition of the upper normal limit of TSH concentration (71). We only compared TSH values between the MAFLD and the controls, irrespective of whether the population had hypothyroidism.

Statements

Data availability statement

The original contributions presented in the study are included in the article/supplementary material. Further inquiries can be directed to the corresponding authors.

Author contributions

WX: Writing – original draft, Investigation, Methodology. HF: Writing – review & editing, Resources, Software. JT: Formal Analysis, Methodology, Investigation, Writing – original draft. QY: Visualization, Resources, Validation, Software, Writing – review & editing. ZX: Writing – review & editing, Software, Resources. TT: Software, Writing – review & editing. XR: Methodology, Writing – review & editing, Investigation. LF: Methodology, Investigation, Writing – review & editing. YS: Validation, Writing – review & editing, Software. XT: Resources, Writing – review & editing. XF: Writing – review & editing, Visualization, Validation. XZ: Project administration, Supervision, Software, Writing – review & editing. MW: Project administration, Writing – review & editing, Methodology, Formal Analysis, Funding acquisition, Writing – original draft, Conceptualization. WG: Supervision, Conceptualization, Writing – review & editing, Funding acquisition, Formal Analysis, Project administration, Writing – original draft.

Funding

The author(s) declare financial support was received for the research and/or publication of this article. This study was supported by Hangzhou biomedicine and health industry development support science and technology project (No. 2022WJCY024; No.2021WJCY001; No. 2021WJCY238; No. 2021WJCY047;No. 2021WJCY115); Hangzhou Normal University Dengfeng Project “Clinical Medicine Revitalization Plan” Jiande Hospital Special Project (No. LCYXZXJH001); Medical and Technology Project of Zhejiang Province (No. 2024KY1348; No. 2024KY200); Zhejiang Traditional Chinese Medicine Scientific Research Fund Project (No. 2024ZL723); Hangzhou Natural Science Foundation of China under Grant (No.2024SZRZDH250001).

Acknowledgments

We thank EditorBar (https://www.editorbar.com/) for editing this manuscript.

Conflict of interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Generative AI statement

The author(s) declare that no Generative AI was used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

References

Summary

Keywords

metabolic associated fatty liver disease, thyroid function, meta-analysis, systemic review, MAFLD

Citation

Xiong W, Fan H, Tang J, Yu Q, Xin Z, Tang T, Rao X, Feng L, Shi Y, Tong X, Fu X, Zhang X, Wang M and Gan W (2025) Association between thyroid function and metabolic associated fatty liver disease: a systemic review and meta-analysis. Front. Endocrinol. 16:1660437. doi: 10.3389/fendo.2025.1660437

Received

06 July 2025

Accepted

13 October 2025

Published

29 October 2025

Volume

16 - 2025

Edited by

Birsen Yilmaz, Çukurova University, Türkiye

Reviewed by

Mohamed Adel Elbasiony, Mansoura University, Egypt

Akif Bayyigit, Prof. Dr. Cemil Tascioglu City Hospital, Türkiye

Updates

Copyright

*Correspondence: Mingwei Wang, ; Wentao Gan,

†These authors have contributed equally to this work

Disclaimer

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.

Outline

Figures

Cite article

Copy to clipboard


Export citation file


Share article

Article metrics