Follicle-stimulating hormone is negatively associated with nonalcoholic fatty liver disease in a Chinese elderly population: a retrospective observational study

Abstract

Background: Several studies have explored the connection between follicle-stimulating hormone (FSH) and nonalcoholic fatty liver disease (NAFLD). However, the impact of FSH elevation on NAFLD remains a topic of debate. Hence, this investigation aimed to evaluate the potential correlation between FSH levels and NAFLD in the aging population. Methods: This was a retrospective observational cross-sectional study between July 2017 and August 2018 in our hospital. We used data obtained from 455 patients over 60 years old. Anthropometrics and laboratory tests were performed for each patient. NAFLD was diagnosed by sonographic features and the fatty liver index (LFI). Results: Of the 455 patients, 200 (43.96%) had NAFLD on their ultrasound and 169 (37.14%) had NAFLD according to the LFI. An intraclass correlation coefficient of the two methods was 80.4% (P < 0.001). People with NAFLD on their ultrasound showed lower FSH levels (52.68 vs. 61.39 IU/L) and more unfavorable metabolic profiles. FSH was negatively correlated with age, alanine aminotransferase, estradiol, testosterone, systolic blood pressure, waist, body mass index, fasting blood glucose, postload plasma glucose and positive associated with total cholesterol, high-density lipoprotein-cholesterol and low-density lipoprotein-cholesterol by Spearman correlation analysis (all P < 0.05). By controlling for all confounding factors, the odds ratios (OR) of FSH for NAFLD were determined in elderly individuals, both men and women, aged 60–70 years and over 70 years. These ORs were found to be 0.937, 0.982, 0.983, and 0.973, respectively, with corresponding 95% confidence intervals (CI) of 0.892–0.984 (P = 0.009), 0.971–0.993 (P = 0.002), 0.967–0.999 (P = 0.033), and 0.958–0.989 (P = 0.001). In addition, our findings demonstrated no significant correlation between FSH and advanced fibrosis when adjusting for potential covariates. The OR for advanced fibrosis was 0.979 (95% CI, 0.938–1.022, P = 0.339). Additionally, ROC curve analysis showed an optimal cut-off value of 66.91 for women and 15.25 for men for NAFLD diagnosis. Conclusions: There was an inverse relationship observed between levels of FSH in the blood serum and NAFLD in the elderly population. These findings suggest that reduced FSH levels might serve as a potential risk factor or biomarker for NAFLD in the elderly.