The progression and regression of metabolic dysfunction-associated fatty liver disease are associated with the development of subclinical atherosclerosis: A prospective analysis

Background Metabolic dysfunction-associated fatty liver disease (MAFLD) has been proposed and diagnosed based on modified criteria. However, evidence for the risks of developing subclinical atherosclerosis with MAFLD transitions according to its new definition has never been reported. Methods Using data from a community-based cohort, 6232 participants aged 40 years or older were included and were followed up for a median of 4.3 years during 2010–2015. Participants were categorized into four groups (stable non-MAFLD, MAFLD regressed to non-MAFLD, non-MAFLD progressed to MAFLD, and stable MAFLD). Subclinical atherosclerosis was defined as elevated carotid intima-media thickness (CIMT), elevated brachial-ankle pulse wave velocity (ba-PWV), or microalbuminuria. Results Compared with the stable non-MAFLD category, participants who progressed to MAFLD at follow-up visit had a 1.356-fold increased risk of developing elevated CIMT [odds ratio (OR) = 1.356; 95% confidence interval (CI) = 1.134–1.620], and a 1.458-fold increased risk of incident microalbuminuria (OR = 1.458; 95% CI = 1.034–2.056) after adjustment for confounders, respectively. In addition, participants with stable MAFLD showed 17.6%, 32.4%, and 35.4% increased risks of developing elevated CIMT, elevated ba-PWV and microalbuminuria, respectively. Compared with the stable MAFLD category, participants with MAFLD and low probability of fibrosis at baseline who regressed to non-MAFLD at follow-up visit had a 29.4% decreased risk of developing elevated CIMT (OR = 0.706; 95% CI = 0.507–0.984), a 43.1% decreased risk of developing elevated ba-PWV (OR = 0.569; 95% CI = 0.340–0.950), but was not significantly associated with incident microalbuminuria (OR = 0.709; 95% CI = 0.386–1.301). The decreased risks attributed to MAFLD regression were more evident in participants without diabetes or dyslipidemia, as well as in those with 0–1 metabolic risk abnormalities, respectively. Conclusions MAFLD was significantly associated with higher risks of developing subclinical atherosclerosis. Moreover, the regression of MAFLD might modify the risks of developing subclinical atherosclerosis, especially among those with low probability of fibrosis or less metabolic risk abnormalities. Since 40% of baseline participants with missing data on MAFLD measurement at follow-up were excluded, the conclusions should be speculated with caution.