A Retrospective Cohort Study of the Chinese Visceral Adiposity Index and Risk of Fatty Liver

Background Obesity is an important risk factor for the development of fatty liver. The current diagnostic indexes of obesity cannot effectively reflect the role of adipose tissue distribution in the development of fatty liver. Objective To assess the correlation of the baseline visceral adiposity index (VAI) and the Chinese visceral adiposity index (CVAI) with the risk of fatty liver. Methods This was a retrospective cohort study involving 17 086 adults receiving physical examination at the Health Management Center of Nanjing Drum Tower Hospital, the Affiliated Hospital of Nanjing University Medical School, from February 2018 to November 2021. Physical examination records were reviewed by the electronic information system of the Health Management Center, and baseline characteristics, physical examination findings, laboratory testing and equipment inspection data were collected. Fatty liver was determined by follow-up abdominal ultrasound and computed tomography (CT). Follow-up was ended in December 2022, and follow-up events were recorded as the interval from non-occurrence of fatty liver at baseline to the first onset of fatty liver or at the end of the follow-up. Subjects were divided into fatty liver group and non-fatty liver group. Baseline VAI and CVAI were compared between groups. They were further divided into four groups based on the quartiles of baseline VAI and CVAI, and the incidence of fatty liver was compared. The correlation of VAI and CVAI with fatty liver was identified by Cox regression. The potential of VAI and CVAI in predicting fatty liver was assessed by the receiver operating characteristic (ROC) curves. Differences in the area under the curve (AUC) between VAI and CVAI were compared by the Delong's test. Results The mean age, BMI and follow-up period of the subjects were (44.3±13.2) years old, (23.2±2.6) kg/m2, and (2.7±1.1) years, respectively. By the end of follow-up, fatty liver was detected in 2 523 (14.8%) subjects. The incidence of fatty liver was significantly higher in men than women [19.4% (1 837/9 461) vs. 9.0% (686/7 625), P<0.001]. The VAI and CVAI of the fatty liver group were significantly higher than those of the non-fatty liver group (P<0.05). Trend testing showed that the incidence of fatty liver significantly increased with the increasing baseline VAI (χ2trend=1 034.9) and CVAI (χ2trend=1 334.8) (both P<0.001). Multivariate Cox regression showed that the risk of fatty liver in the VAI-Q4 group and CVAI-Q4 group was 2.579 times (95%CI=2.088-3.186) and 3.375 times (95%CI=2.488-4.576) that of the VAI-Q1 group, respectively. The ROC curve showed that the AUC of CVAI in predicting fatty liver was greater than that of VAI (0.737 vs. 0.708, P<0.001). Stratified analysis showed that CVAI was significantly correlated with the occurrence of fatty liver in gender, age and BMI subgroups (P<0.001) . Conclusion Baseline CVAI is significantly correlated with fatty liver development, and superior to VAI in predicting fatty liver.