Recent animal studies support close associations of Periostin with hepatosteatosis and steatohepatitis. This study is to evaluate the role of serum periostin in non-alcoholic fatty liver disease (NAFLD). A hospital-based age-/sex-matched case-control study was conducted. Binary logistic regression and receiver operating characteristic (ROC) curve were performed. Serum adipokines were measured by Adipokine Magnetic Bead Panel kits. The serum concentration of Periostin in NAFLD (1914.16 [1323.59-2654.88] ng/ml, P < 0.001) was higher than it in control (1244.94 [837.87-2028.55] ng/ml). The frequency of NAFLD grew (29.8, 52.6, and 67.2%, P < 0.001), as Periostin concentration increased among its tertiles. Compared with the 1st tertile, the 2nd and the 3rd tertiles of Periostin indicated significant associations with higher odds of NAFLD [adjusted odds ratio = 2.602 (95% confidence interval (CI) 1.030-6.575), P = 0.043 and 2.819 (95% CI 1.629-4.878), P < 0.001]. ROC curve of Periostin was developed to predict the presence of NAFLD (area under ROC = 0.693 [95% CI 0.614-0.771], P < 0.001). Lastly, Periostin correlated with several adipokines, including Resistin (r = 0.269, P = 0.018), Adiponectin (r = -0.352, P = 0.002), Interleukin (IL)-6 (r = 0.359, P = 0.001), IL-8 (r = 0.364, P = 0.001), Lipocalin-2 (r = 0.623, P < 0.001), Hepatocyte growth factor (r = 0.522, P < 0.001), and Nerve growth factor (r = 0.239, P = 0.036). It suggests Periostin as a potential biomarker in the management of NAFLD.
Keywords: Adipokine; Biomarker; Case–control; Non-alcoholic fatty liver disease (NAFLD); Periostin.