Study On The Association Of Thyroid Hormones And Thyroid Stimulating Hormone With Nonalcoholic Fatty Liver

BackgroundThyroid hormone plays an important role on regulation of energy expenditure and tissue metabolism, including maintaining lipid and glucose homeostasis. It affects expression of gene in liver, skeletic muscle and adipose tissues. Thyroid dysfunction often leads to various metabolic disorders, for example, hypothyroidism often results in elevation of cholesterol and triglycerides; low thyroid function had been shown to be associated with metabolic syndrome and its components, including obesity, insulin resistance,dyslipidemia and hypertension.Nonalcoholic fatty liver disease (NAFLD) is characterized by increased accumulation of lipid in liver. With the development of economy, the prevalence of NAFLD grows rapidly and becomes the most common liver disease in the world. NAFLD is closely related to metabolic diseases such as obesity, hyperlipidemia and diabetes.Therefore, NAFLD is considered as the manifestation of metabolic syndrome on the liver. Given higher prevalence of NAFLD in patients with metabolic syndrome, the relationship between hypothyroidism and NAFLD has gained attention. The higher prevalence of hypothyroidism among patients with NAFLD has been reported by several studies. A cross-sectional study on the relationship between the broad spectrum of hypothyroidism and NAFLD showed that both subclinical hypothyroidism and overt hypothyroidism were statistically associated NAFLD independent of known metabolic risk factors.Recently, these associations between thyroid hormone and hepatic steatosis have been extended to euthyroid individuals by three studies, but with somewhat inconsistent results. Several studies showed that low free thyroxine (FT4) level or high thyroid-stimulating hormone (TSH) level was significantly associated with hepatic steatosis. For free triiodothyronine (FT3), only one study showed that FT3 concentrations were positively associated with hyperechogenicliver pattern in the people with normal ALT. The inconsistencies may partly be attributable to clinical heterogeneity and small sample. A large-sample study was warranted to further address this issue.Objective1. To assess the relationships between the thyroid parameters and NAFLD and metabolic parameters in a large-sample middle-aged euthyroid population.2. To assess whether gender, menopause and diabetes affect the interactions between thyroid function and NAFLD.MethodsSubjects who visited Provincial Hospital, affiliated to Shandong University (Jinan, China) for a routine health check-up were reviewed retrospectively. All participants were asked to complete a self-reported questionnaire and provided an overnight fasting blood sample between 0900 and 1000 h.The following subjects were excluded: subjects with abnormal serum levels of TSH, FT4 or FT3; those with excessive alcohol intake (>20 g/day); those with hepatitis B virus antigen or hepatitis C virus antibody or known liver diseases; those with chronic renal diseases; pregnant women and subjects with any diseases or taking any medicine that might affect their thyroid status and hepatic steatosis (thyroid hormone, antithyroid drugs, iodine, amiodarone, estrogens, androgens, steroid hormones, β-adrenoceptor blockers, antiepileptic drugs, tamoxifen, methotrexate, statins, and fibrates). Besides, because diabetes mellitus was reported to be accompanied by changes in thyroid hormone dynamic(14), subjects with diabetes were also excluded.Anthropometric and laboratory measurementsEssential informations including gender, age, drinking, smoker, hypertension, diabetes and other past medical history were obtained from the self-reported questionnaire.Body weight and height were measured with an autoanthropometer. Body mass index (BMI) was calculated by dividing the weight (Kg) with the square of height (m2). All patients were asked to rest for at least 30 minutes and then in a sitting position, the blood pressures of their right arms were measured twice with a desk-model sphygmomanometer. There was a 3-minutes interval between the two measurements for each participant, and the mean value of the two measurements was used as systolic blood pressure (SBP) and diastolic blood pressure (DBP).Fasting venous blood specimens were collected to measure the parameters of liver function, lipid profile, blood glucose, uric acid, urea nitrogen (BUN), creatinine and thyroid hormone.Hepatic ultrasonography was performed by experienced radiologists who were blinded to the clinical presentations and laboratory findings of the subjects. The diagnosis of NAFLD was based on the presence of fatty liver disease determined by ultrasonography in the absence of the following: (1) seropositivity for hepatitis B surface antigen or antibody to hepatitis C virus, (2) excessive alcohol intake (>20 g/day), (3) other causes of liver disease, and (4) medications known to produce fatty liver disease..Statistical analysisContinuous data were represented as mean±tandard deviation variables while categorical data were expressed as numbers or percentages. Inter-group analysis was done by using independent t test, Mann-Whitney U test and chi-square test. Analysis of covariance was performed to adjust for age and sex. Partial correlation analysis was used to control age and sex while evaluating the relationship between levels of thyroid hormone and metabolic parameters or laboratory measurements. Multivariate logistic regression analysis was done to determine risk factors relating to NAFLD. Statistical significance was set at p<0.05. All the above analyses were conducted with SPSS software in version 17.0.ResultsResult IClinical characteristics of total subjects and comparisonsA total of 3151subjects meeting the inclusion criteria were finally included. The mean age was 46.42±10.60 of year (age range: 36-60 years). A total of 38.4% of the patients presented NAFLD. Old age and percentage of male, smoker and hypertension were much higher in NAFLD group compared with non-NAFLD group. Compared with participants in non-NAFLD group, those in NAFLD group had higher levels of BMI and blood pressure, and higher serum levels of glutamic-pyruvic transaminase (ALT), glutamic-oxalacetic transaminase (ALT), glutamyltranspetidase (GGT), glucose, total cholesterol (TC), triglyceride (TG), low density lipoprotein-cholesterol (LDL-c) and uric acid. Levels of TSH and FT4 were similar between NAFLD and non-NAFLD group (TSH: 2.20±0.93 vs 2.12±0.90 mIU/L; FT4: 16.23±2.07 vs 16.47±2.08 pmol/L, respectively), while level of FT3 was much higher in NAFLD group (5.11±0.67 vs 4.83±0.64 pmol/L, adjusted p=0.000).Partial correlation analysisThe correlations between thyroid hormone and metabolic measurements were calculated by using partial correlation analysis adjusting for age and sex. FT3 showed positive correlation with BMI, systolic blood pressure (SBP), diastolic blood pressure (DBP), glucose, TG, BUN, LDL-c, creatinine, uric acid, ALT, AST and Sqrt GGT, and negative correlation with HDL-c. Levels of FT4 were positively correlated with glucose, BUN and uric acid, and negatively correlated with ALT and AST. Variables which were positively associated with TSH included creatinine, uric acid, ALT and AST. TSH was negatively correlated with HDL-c.Multiple logistic regression analysisThe multiple logistic regression analysis showed FT3 was an independent risk factor for NAFLD, with an adjusted odds ratio (OR) of 1.324 (95% CI:1.079-1.624, p=0.007) in the presence of relatively high serum FT3 level. Elevation of serum FT4 level was significantly associated with the reduction of risk for NAFLD (OR: 0.949, 95% CI:0.900-1.000,p=0.049). Other independent factors for predicting NAFLD included BMI, glucose, TG, creatinine and uric acid. TSH level was not significantly associated with NAFLD (1.003,95% CI:0.896-1.121, p=0.964).We performed an additional analysis of FT3, FT4 and TSH in relation to NAFLD separately in men and women, and also in pre- and post-menopausal women. The results showed that FT3 levels were not significantly associated with NAFLD in men or women after adjusting for age, BMI, lipids, creatinine, BUN and uric acid et al. Similarly, no significant associations between FT3 and NAFLD were observed in pre- and post-menopausal women. With respect to the correlation of FT4 and TSH with NAFLD, multiple logistic regression analyses performed in subgroups according to sex also generated no statistically significant results. However, menopausal status affected the interactions between FT4 and NAFLD in women, with negative association in pre-menopausal women (OR:0.787,95% CI: 0.628-0.988,p=0.0039) and null association in post-menopausal women OR:1.006, 95% CI:0.829-1.220,p=0.951). Menopausal status did not change the interactions between TSH or FT3 and NAFLD.Result ⅡIt was reported that diabetes mellitus was accompanied by changes in thyroid hormone dynamic, so subjects with diabetes were further excluded and the statistical analyses were performed again to obtain the Result Ⅱ.Clinical characteristics of total subjects and comparisonsA total of 2576 subjects meeting the inclusion criteria were finally included. The basic characteristics of all subjects and comparisons between NAFLD non-NAFLD group were displayed in Table 4. The mean age was 45.37±10.32 of year (age range: 36-60 years). A total of 38.4% of the patients presented NAFLD. Old age and percentage of male, smoker and hypertension were much higher in NAFLD group compared with non-NAFLD group. Compared with participants in non-NAFLD group, those in NAFLD group had higher levels of BMI and blood pressure, and higher serum levels of glutamic-pyruvic transaminase (ALT), glutamic-oxalacetic transaminase (ALT), glutamyltranspetidase (GGT), glucose, total cholesterol (TC), triglyceride (TG), low density lipoprotein-cholesterol (LDL-c) and uric acid. Levels of TSH and FT4 were similar between NAFLD and non-NAFLD group (TSH: 2.13±0.90 vs 2.20±0.93 mIU/L; FT4: 16.41±2.04 vs 16.18±2.06 pmol/L, respectively), while level of FT3 was much higher in NAFLD group (5.12±0.58 vs 4.84±0.58 pmol/L, adjusted p=0.000). No significant differences on creatinine between the two groups after adjustment for age and sex were observed.Partial correlation analysisThe correlations between thyroid hormone and metabolic measurements were calculated by using partial correlation analysis adjusting for age and sex. FT3 showed positive correlation with BMI, glucose, TG, LDL-c, uric acid, ALT, AST and GGT, and negative correlation with TC and HDL-c. Levels of FT4 were positively correlated with glucose, BUN and uric acid, and negatively correlated with ALT and AST. Besides uric acid, TSH did not significantly correlate with the variables.Multiple logistic regression analysisThe multiple logistic regression analysis showed FT3 was an independent risk factor for NAFLD, with an adjusted odds ratio (OR) of 1.253 (95% CI:1.010-1.556, p=0.040) in the presence of relatively high serum FT3 level. Other independent factors for predicting NAFLD included BMI, glucose, TG, creatinine and uric acid. Neither FT4 nor TSH level was significantly associated with NAFLD.We performed an additional analysis of FT3, FT4 and TSH in relation to NAFLD separately in men and women, and also in pre- and post-menopausal women. The results showed that FT3 levels were not significantly associated with NAFLD in men or women after adjusting for age, BMI, lipids, creatinine, BUN and uric acid et al. Similarly, no significant associations between FT3 and NAFLD were observed in pre- and post-menopausal women. With respect to the correlation of FT4 and TSH with NAFLD, multiple logistic regression analyses performed in subgroups according to sex also generated no statistically significant results. However, menopausal status affected the interactions between FT4 and NAFLD in women, with negative association in pre-menopausal women (OR: 0.777,95% CI: 0.617-0.979,p=0.0032) and null association in post-menopausal women (OR: 1.037,95% CI: 0.841-1.277, p=0.736). Menopausal status did not change the interactions between TSH or FT3 and NAFLD.Conclusions1. High levels of FT3 within the reference range were positively associated with NAFLD in middle-aged euthyroid subjects, independent of gender, age, BMI, smoking, glucose, lipids, uric acid, BUN and creatinine. Exclusion of diabetes did not change the results significantly.2. Levels of FT4 within the reference range were negatively associated with NAFLD in middle-aged euthyroid subjects, independent of gender, age, BMI, smoking, glucose, lipids, uric acid, BUN and creatinine. This significant association disappeared after excluding subjects with diabetes.3. A negative association between FT4 and NAFLD in pre-menopausal women and a null association in post-menopausal women were observed. Our results showed that the interactions between serum FT4 level and NAFLD were affected by menopausal status.4. No significant associations between TSH and NAFLD were observed.5. High levels of uric acid were positively associated with NAFLD in middle-aged euthyroid subjects, independent of gender, age, BMI, smoking, glucose, lipids, BUN,creatinine and thyroid function.6. High level of serum creatinine seemed to be a protective factor for NAFLD independent of gender, age, BMI, smoking, glucose, lipids, uric acid, BUN and thyroid function.