Association of Thyroid Hormone Levels with NAFLD in Recent-Onset Diabetes

Nonalcoholic fatty liver disease (NAFLD) is a growing health concern associated with obesity and type 2 diabetes (T2D). Thyroid hormones are important regulators of lipid homeostasis. Indeed, people with hypothyroidism are at greater risk for NAFLD and it has been suggested that humans with T2D are more likely to develop hypothyroidism. However, little is known about the relationship of thyroid function with NAFLD in type 1 diabetes (T1D) and T2D. We hypothesized that (i) low free thyroxin (fT4) and high thyroid-stimulating hormone (TSH) levels are associated with increased hepatocellular lipid content (HCL) and (ii) that fT4 and TSH can identify persons at high-risk for hepatic steatosis in people with recent-onset diabetes. We examined people with T1D (n=358) or T2D (n=596) and 175 healthy participants of the German Diabetes Study (GDS). First, validation of the fatty liver index (FLI) against quantification of HCL by 1H-magnetic resonance spectroscopy revealed a close correlation (ß=0.715, p<0.001; n=446). Second, FLI negatively correlated with fT4 in males (ß=−0.139, p<0.01), but not in females with T2D (ß=−0.086, p=0.26). Likewise, TSH associated positively with FLI in males (ß=0.116, p<0.05), but not in females with T2D (ß=−0.057, p=0.45). TSH and FLI were differently associated between males and females before (p<0.05), but not after (p=0.82) adjustment for BMI. Finally, fT4 had a low diagnostic precision for steatosis (by FLI) as derived from the respective area under the receiver operating curve (0.57, 95% CI [0.53;0.62]; p<0.001) in T2D and was not significant in other groups. TSH had no diagnostic precision for steatosis in any groups. In conclusion, FLI can be used as a surrogate measure of HCL in recent-onset diabetes. The correlation of FLI with lower thyroid function in male T2D suggests a sex-specific interaction between thyroid and liver lipid metabolism, which is mainly driven by body mass. Both, fT4 and TSH offer no relevant efficacy to detect steatosis at least in recent-onset diabetes. Disclosure N. Saatmann: None. M. Schön: None. O.P. Zaharia: None. M. Huttasch: None. K. Strassburger: None. S. Trenkamp: None. Y. Kupriyanova: None. V. Schrauwen-Hinderling: None. S. Kahl: None. V. Burkart: None. R. Wagner: Speaker's Bureau; Novo Nordisk, Sanofi. Advisory Panel; Daiichi Sankyo. M. Roden: Advisory Panel; Eli Lilly and Company. Research Support; Boehringer-Ingelheim, Novo Nordisk, Novartis. Consultant; TARGET PharmaSolutions, Inc. Research Support; Sanofi.