Background The prevalence of metabolic-associated fatty liver disease (MAFLD) has increased rapidly. And there is no conclusion on body fat composition, characteristics of metabolic indicators, and their predictive values for MAFLD in non-obese populations.
Objective To identify the risk factors for MAFLD by comparing body fat composition and key metabolic indicators (blood lipids, blood sugar, uric acid) between obese and non-obese MAFLD patients, and to assess their associations with MAFLD as well as their predictive values for MAFLD in non-obese patients.
Methods Physical examinees with and without liver B-ultrasound-detected fatty liver were recruited from Health Management Center, Nanjing Drum Tower Hospital from January 2018 to January 2019 after excluding those with non-MAFLD, and divided into obese group (including 129 cases with MAFLD, and 129 without fatty liver) and non-obese group (including 129 without fatty liver cases, and 129 with MAFLD) by BMI. The body fat composition and metabolic indices in non-obese MAFLD cases were compared with those of the other three subgroups. The correlation of each index with MAFLD in non-obese cases was analyzed. The independent risk factors of MAFLD in non-obese cases were identified by using Logistic regression. The predictive value of each index for MAFLD in non-obese was assessed using the receiver operating characteristic (ROC) curve.
Results (1) Comparison of body fat composition and metabolic indicators: compared with non-obese without fatty liver cases, non-obese cases with MAFLD had greater average BMI, body fat (BF), body fat ratio (BFR), visceral fat area (VFA), waist circumference (WC), waist-hip-ratio (WHR), total cholesterol (TC), triglyceride (TG), low-density lipoprotein cholesterol (LDL-C), uric acid (UA), alanine transaminase (ALT) and gamma-glutamyl transpeptidase (GGT), and lower high-density lipoprotein cholesterol (HDL-C) (P<0.05). In comparison with obese cases with MAFLD, non-obese cases with MAFLD had lower average BMI, BF, VFA, WC, WHR, fasting plasma glucose (FPG), and glycosylated hemoglobin (HbA1c) (P<0.05). Non-obese cases with MAFLD had lower average BMI, BF and WC, and higher average TG, UA, ALT and GGT than obese cases without fatty liver (P<0.05). Non-obese female cases with MAFLD had greater average age, BF, BFR and VFA and lower WC, VA, GGT than non-obese male cases with MAFLD (P<0.05). (2) Kendall's rank correlation analysis showed that the risk of MAFLD in non-obese cases increased with the growth of BFR, VFA, WHR, TC, TG, LDL-C, UA, ALT, and GGT (r=0.099, 0.092, 0.136, 0.095, 0.176, 0.092, 0.114, 0.125, 0.142, P<0.05), but decreased with the growth of HDL-C (r=-0.112, P<0.05). (3) Multivariate Logistic regression analysis showed that TG, ALT, UA, BFR and VFA were risk factors of MAFLD in non-obese cases. (4) The results of ROC analysis of the performance of five indicators predicting MAFLD in non-obese cases were as follows: BFR had an AUC of 0.853〔95%CI (0.807, 0.898) 〕, with 0.789 sensitivity, 0.770 specificity when 22.30% was chosen as the optimal cut-off value; VFA had an AUC of 0.938〔95%CI (0.906, 0.970) 〕, with 0.852 sensitivity, 0.904 specificity when 61.45 cm2 was chosen as the optimal cut-off value; TG had an AUC of 0.807〔95%CI (0.754, 0.860) 〕, with 0.822 sensitivity, 0.713 specificity when 1.02 mmol/L was chosen as the optimal cut-off value; UA had an AUC of 0.665〔95%CI (0.599, 0.731) 〕, with 0.605 sensitivity, 0.682 specificity when 356.00 μmol/L was chosen as the optimal cut-off value; ALT had an AUC of 0.752〔95%CI (0.692, 0.812) 〕, with 0.814 sensitivity, 0.770 specificity when 18.35 U/L was chosen as the optimal cut-off value.
Conclusion Compared with non-obese people without fatty liver, non-obese people with MAFLD had increased BF and visceral fat, abnormal lipid metabolism, elevated levels of UA and transaminase. The risk of MAFLD in non-obese people increased with the increase in TG, ALT, UA, BFR and VFA, but decreased with the increase in HDL-C. BFR, VFA, TG, UA and ALT could partially predict and diagnose MAFLD in non-obese people, providing evidence for the delivery of interventions as soon as possible.