Thyroid Function Could Help Predict NAFLD Risk in Diabetes


A study reveals a link between low free thyroxine and fatty liver risk in type 2 diabetes patients, suggesting a gender-specific thyroid-liver interaction.

Michael Roden, MD | Credit: PanNASH Initiative

Michael Roden, MD
Credit: PanNASH Initiative

Results from a recent study are highlighting the association between low free thyroxine and steatosis risk in patients with type 2 diabetes, further suggesting a sex-specific interaction between thyroid function and hepatic lipid metabolism.

“As subclinical hypothyroidism and even low-normal thyroid function have been associated with increased mortality in individuals with NAFLD, early detection of abnormal thyroid function is of high clinical relevance,” wrote investigators.1

Nonalcoholic fatty liver disease (NAFLD) is the most common chronic liver condition in the US, estimated to affect about 25% of adults.2 Risk factors include high cholesterol, high levels of triglycerides in the blood, metabolic syndrome, obesity, type 2 diabetes, and hypothyroidism.3

To investigate associations between free thyroxine or thyroid-stimulating hormones and NAFLD in recent onset diabetes, Michael Roden, MD, director of the Deutsches Diabetes Zentrum and the Division of Endocrinology and Diabetology at Düsseldorf University Hospital, and a team of investigators launched the current study to estimate incidence for patients with recent onset type 1 diabetes, type 2 diabetes, or without diabetes. Investigators identified a study cohort using participants of the German Diabetes Study who had complete sets of free thyroxine and thyroid-stimulating hormone data were selected for analysis.1

To be included in the study, participants were required to have a known diabetes duration of <12 months or without diabetes and aged 18 to 69 years. Patients were excluded if they had diabetes of other causes, pregnancy, acute or severe chronic heart, hepatic, renal or psychiatric diseases or immunosuppressive treatment and alcohol abuse. In total, investigators identified 1129 participants, 358 with type 1 diabetes, 596 with type 2 diabetes, and 175 without diabetes.1

Patients with type 2 diabetes had a mean age of 53.6 years (46.3 to 59.8) compared to 34.6 years (26.6 to 45.4) in the type 1 diabetes group and 45.7 years (29.5 to 54.2) in the group without diabetes. BMI was also greater among the type 2 diabetes group (30.7; 27.0 to 35.1) compared to patients with type 1 diabetes (24; 22.0 to 26.6) and patients without diabetes (26.2; 23.1 to 29.5).1

Investigators performed ANCOVA-like regression models, allowing for different intercepts, slopes, and residual variances in Fatty Liver Index, free thyroxine and thyroid-stimulating hormones levels between groups. ANCOVA-like regression analyses were also performed allowing for different intercepts, slopes, and residual variances in HbA1c, blood glucose, C-peptide secretion, M-value and free thyroxine and thyroid-stimulating hormones. Serum HDL-cholesterol and triglycerides were also included in regression models. Investigators performed tests for sex interactions and skewed data were log-transformed before analysis.1

Results showed a positive correlation between Fatty Liver Index and hepatocellular lipid content by 1H-MRS (ß = .715, P < .001). In contrast, Fatty Liver Index was inversely associated with free thyroxine only among those with type 2 diabetes (ß = −.110, P < .01) independent of age, sex and BMI, but the correlation was not significant after adjustment for insulin sensitivity (ß = −.021, P = .67). Thyroid-stimulating hormones did not correlate with Fatty Liver Index in any group. Further analyses showed a negative relationship between hepatocellular lipid content and free thyroxine (ß = −.092; P ≤ .05), but not with thyroid-stimulating hormones (ß = .01, P = .83) in all 3 groups.1

Investigators divided the entire cohort into steatosis categories according to the Fatty Liver Index. In the low steatosis risk category, patients without diabetes had the lowest free thyroxine (without diabetes vs type 1 diabetes: P < .01; without diabetes vs type 2 diabetes: P < .001). Patients without diabetes also had the lowest free thyroxine in the high steatosis risk category (without diabetes vs type 1 diabetes: P < .05; without diabetes vs type 2 diabetes: P < .001), while free thyroxine was similar across all groups of the intermediate steatosis risk category.1

Type 2 diabetes within the high steatosis risk category had lower free thyroxine than those of the low steatosis risk category (P < .01), who had higher free thyroxine than those within the intermediate steatosis risk category (P < .05). Of note, no differences were found for thyroid-stimulating hormones between the Fatty Liver Index categories within any group.1

Overall, 9.4% of males and 20.6% of females with type 2 diabetes had low steatosis risk according to cut-offs of the Fatty Liver Index, while prevalence of NAFLD was 49.6% for males and 63.6% for females with type 2 diabetes. Males with type 2 diabetes showed similar free thyroxine (P = .06) and thyroid-stimulating hormone levels (P = .24), but higher Fatty Liver Index values (P < .001) relative to females with type 2 diabetes.1

Fatty Liver Index further associated with free thyroxine in males (ß = −.139, p < .01), but not in females with type 2 diabetes (ß = −.086, P = .26). Investigators pointed out the association in males remained after adjustments for age and BMI (ß = −.117, P < .05), but disappeared upon adjusting for insulin sensitivity (ß = −.033, P = .56).1

Additionally, Fatty Liver Index was associated with thyroid-stimulating hormones in males (ß = .116, P < .05), but not in females with type 2 diabetes (ß = −.057, P = .45) The correlation between Fatty Liver Index and thyroid-stimulating hormones in males with type 2 diabetes was insignificant after adjustment for age and BMI (ß = .052, P = .30).1

“The findings underline the need of thyroid function screening as early as possible in type 2 diabetes, but calls for further validation in other cohorts and mechanistic clinical studies,” concluded investigators.1


  1. Saatmann N, Schön M, Zaharia O-P, et al. Association of thyroid function with non-alcoholic fatty liver disease in recent-onset diabetes. Liver Int. 2023; 00: 1-12. doi:10.1111/liv.15723
  2. American Liver Foudnation. NASH Definition & Prevalence. What is fatty liver disease? Accessed September 15, 2023.
  3. Mayo Clinic. Nonalcoholic fatty liver disease. Diseases and Conditions. Accessed September 15, 2023.
Related Videos
Video 10 - "Future Treatment Landscape for COPD"
Video 9 - "Emerging Treatment Approaches in COPD"
Ghada Bourjeily, MD: Research Gaps on Sleep Issues During Pregnancy
John Winkelman, MD, PhD: When to Use Low-Dose Opioids for Restless Legs Syndrome
Bhanu Prakash Kolla, MBBS, MD: Treating Sleep with Psychiatric Illness
Jennifer Martin, PhD: Boosting CPAP Adherence in Women with Sleep Apnea
Video 2 -  4 KOLs are featured in, "Educating Primary Care Clinicians on Outpatient HE Management and Ammonia Testing"
Video 1 - 4 KOLs are featured in, "Exploring the Impact of Hepatic Encephalopathy on Patients and Their Families "
© 2024 MJH Life Sciences

All rights reserved.