MAFLD better predicts the progression of atherosclerotic cardiovascular risk than NAFLD: Generalized estimating equation approach

• Published in: Hepatology research Vol. 51; no. 11; pp. 1115 - 1128
• Main Authors: Tsutsumi, Tsubasa, Eslam, Mohammed, Kawaguchi, Takumi, Yamamura, Sakura, Kawaguchi, Atsushi, Nakano, Dan, Koseki, Masahiro, Yoshinaga, Shinobu, Takahashi, Hirokazu, Anzai, Keizo, George, Jacob, Torimura, Takuji
• Format: Journal Article
• Language: English
• Published: Hoboken Wiley Subscription Services, Inc 01.11.2021
• Subjects: Alcohol; alcohol consumption; Arteriosclerosis; Atherosclerosis; cardiovascular disease risk; Cardiovascular diseases; Fatty liver; generalized estimating equation approach; Liver diseases; metabolic associated fatty liver disease; metabolic dysfunction; Metabolism; non‐alcoholic fatty liver disease; Regression analysis
• ISSN: 1386-6346; 1872-034X
• DOI: 10.1111/hepr.13685

Aim Metabolic associated fatty liver disease (MAFLD) partly overlaps with non‐alcoholic fatty liver disease (NAFLD). Thus, using a generalized estimating equation (GEE) approach, we aimed to investigate the difference in worsening of atherosclerotic cardiovascular disease (ASCVD) risk between patients with MAFLD and NAFLD. We also investigated factors related to the difference between the two groups. Methods We enrolled 2306 subjects with fatty liver (MAFLD 80.7%, NAFLD 63.4%). Subjects with MAFLD/NAFLD were sub‐classified into three groups: NAFLD with no metabolic dysfunction (non‐Met NAFLD), overlapping, and MAFLD with moderate alcohol consumption (mod‐Alc MAFLD). ASCVD risk was estimated by non‐invasive tests, including the Suita score. An event was defined as worsening of these scores from the low‐risk to the high‐risk group. Independent factors for the event were analyzed by Cox regression analysis with the GEE. Results In Cox regression analysis, MAFLD (HR 1.08, 95% CI 1.02–1.15, p = 0.014) and alcohol consumption (20–39 g/day; HR 1.73, 95% CI 1.26–2.36, p = 0.001) were independently associated with worsening of the Suita score. In a subanalysis, the incidence of the event was significantly lower in non‐Met NAFLD than in the overlapping group (HR 0.70, 95% CI 0.50–0.98, p = 0.042). However, no significant difference was observed in the incidence between the overlapping and mod‐Alc MAFLD group (HR 1.19, 95% CI 0.89–1.58, p = 0.235). Conclusions The GEE approach demonstrates that MAFLD better identifies patients with worsening of ASCVD risk than NAFLD. Moreover, the superiority of MAFLD over NAFLD was due to the presence of metabolic dysfunction rather than moderate alcohol consumption.