Comparison of several blood lipid-related indexes in the screening of non-alcoholic fatty liver disease in women: a cross-sectional study in the Pearl River Delta region of southern China

Abstract

Background: Lipid metabolism disorders play a critical role in the progression of non-alcoholic fatty liver disease (NAFLD). However, the number of studies on the relationships among blood lipid-related indexes and NAFLD is limited, and few studies have emphasized the comparison of blood lipid-related indexes in the same population to identify the optimal index for NAFLD screening. This study aimed to investigate the relationships among several blood lipid-related indexes and NAFLD, and to find the index with the best screening value for NAFLD.

Methods: Based on a general health examination at community health service agencies in the Pearl River Delta region of China in 2015, 3239 women were recruited in this cross-sectional study. The relationships among blood lipid-related indexes and NAFLD were assessed separately by constructing multivariate logistic regression models. Receiver operating characteristic analysis was used to evaluate and compare the screening abilities of the indexes for NAFLD. All data analyses were conducted in SPSS and MedCalc software.

Results: Whether in the crude model or each model adjusted for possible confounding factors, the risk of NAFLD significantly rose with increasing cardiometabolic index (CMI), triglyceride glucose index (TyG), triglycerides (TG) to high-density lipoprotein cholesterol (HDL-C) ratio (TG/HDL-C), total cholesterol (TC) to HDL-C ratio (TC/HDL-C) and low-density lipoprotein (LDL-C) to HDL-C ratio (LDL-C/HDL-C). Moreover, the area under the curve (AUC) of CMI was 0.744, which was better than that of TyG (0.725), TG/HDL-C (0.715), TC/HDL-C (0.650), and LDL-C/HDL-C (0.644) (P < 0.001). In addition, the optimal cut-off points were 0.62 for CMI, 8.55 for TyG, 1.15 for TG/HDL-C, 4.17 for TC/HDL-C, and 2.22 for LDL-C/HDL-C.

Conclusions: CMI is easy to obtain, is a recommended index in the screening of NAFLD in women and may be useful for detecting populations that are at high risk of NAFLD.

Keywords: Blood lipid-related indexes; Cross-sectional study; Non-alcoholic fatty liver disease; Screening ability; Southern China; Women.

Fig. 1

The relationships between blood lipid-related indexes and the risk of NAFLD. These indexes were regarded as continuous variables in the multivariate logistic regression models in this figure. Because the indexes were converted to z scores in the multivariate logistic regression model, OR increase with each 1-unit increase in SD for every index. Crude model: unadjusted; Model I: adjusted for age, hypertension, diabetes; Model II: adjusted for Model I plus SBP, DBP, and BMI; Model III: adjusted for model II plus RBC, WBC, Hb, PLT, FPG, HbA1c, ALT, AST, ALP, GGT, TBIL, Scr, BUN and UA

Fig. 2

ROC curves of different blood lipid-related indexes in the screening of NAFLD in women