The nonlinear relationship between thyroid function parameters and metabolic dysfunction-associated fatty liver disease

Yingying Hu; Fan Zhou; Fang Lei; Lijin Lin; Xuewei Huang; Tao Sun; Weifang Liu; Xingyuan Zhang; Jingjing Cai; Zhi-Gang She; Hongliang Li

doi:10.3389/fendo.2023.1115354

The nonlinear relationship between thyroid function parameters and metabolic dysfunction-associated fatty liver disease

Front Endocrinol (Lausanne). 2023 Feb 22:14:1115354. doi: 10.3389/fendo.2023.1115354. eCollection 2023.

Authors

Yingying Hu^{1

2}, Fan Zhou^{3

4}, Fang Lei^{2

5}, Lijin Lin^{1

2}, Xuewei Huang⁶, Tao Sun^{1

2}, Weifang Liu^{1

2}, Xingyuan Zhang^{2

5}, Jingjing Cai^{2

6}, Zhi-Gang She^{1

2}, Hongliang Li^{1

2}

Affiliations

¹ Department of Cardiology, Renmin Hospital of Wuhan University, Wuhan, China.
² Institute of Model Animal, Wuhan University, Wuhan, China.
³ Department of Gastroenterology, Huanggang Central Hospital of Yangtze University, Huanggang, China.
⁴ Huanggang Institute of Translational Medicine, Huanggang, China.
⁵ School of Basic Medical Science, Wuhan University, Wuhan, China.
⁶ Department of Cardiology, The Third Xiangya Hospital, Central South University, Changsha, China.

Abstract

Background: The relationship between thyroid function parameters and metabolic dysfunction-associated fatty liver disease (MAFLD) remains controversial. Additionally, little is known about the relationship between thyroid function parameters and MAFLD in the Chinese population.

Methods: We conducted a retrospective cross-sectional study involving 177,540 individuals with thyroid function tests and MAFLD diagnosis from 2010-2018. The association between thyroid function parameters and MAFLD was evaluated on a continuous scale with restricted cubic spline (RCS) models and by the prior-defined centile categories with multivariable-adjusted logistic regression models. Thyroid function parameters included free triiodothyronine (FT3), free tetra-iodothyronine (FT4), and thyroid stimulating hormone (TSH). Additionally, fully adjusted RCS models stratified by sex, age, and location were studied.

Results: In the RCS models, the risk of MAFLD increased with higher levels of FT3 when FT3 <5.58pmol/L, while the risk of MAFLD decreased with higher levels of FT3 when FT3 ≥5.58pmol/L (P nonlinearity <0.05). While RCS analysis suggested that the FT4 levels had a negative association with MAFLD (P nonlinearity <0.05), indicating an increase in FT4 levels was associated with a decreased risk of MAFLD. RCS analysis suggested an overall positive association between the concentration of TSH and MAFLD risk (P nonlinearity <0.05). The rising slope was sharper when the TSH concentration was less than 1.79uIU/mL, which indicated the association between TSH and MAFLD risk was tightly interrelated within this range. The multivariable logistic regression showed that populations in the 81st-95th centile had the highest risk of MAFLD among all centiles of FT3/TSH, with the 1st-5th centile as the reference category.

Conclusions: Our study suggested nonlinear relationships between thyroid function parameters and MAFLD. Thyroid function parameters could be additional modifiable risk factors apart from the proven risk factors to steer new avenues regarding MAFLD prevention and treatment.

Keywords: metabolic dysfunction-associated fatty liver disease; multivariable-adjusted logistic regression models; nonlinear relationship; restricted cubic spline analysis; thyroid function parameters.

Publication types

Research Support, Non-U.S. Gov't

MeSH terms

Cross-Sectional Studies
Humans
Non-alcoholic Fatty Liver Disease* / metabolism
Retrospective Studies
Thyroid Gland* / metabolism
Thyrotropin / metabolism
Thyroxine / metabolism

Substances

Thyroxine
Thyrotropin

Grants and funding

This work was supported by grants from the Hubei Province Innovation Platform Construction Project (20204201117303072238), and the National Science Foundation of China (81770053, 81970364, 81870171, 82170436).