Associations of traditional healthy lifestyle and sleep quality with metabolic dysfunction-associated fatty liver disease: two population-based studies

Abstract

Aims: To examine whether an extended lifestyle metrics incorporating sleep quality improves risk stratification for metabolic dysfunction-associated fatty liver disease (MAFLD), at-risk metabolic dysfunction-associated steatohepatitis (MASH) and significant fibrosis.

Methods: A total of 5011 participants with abdominal ultrasound from Imaging sub-cohort of South China Cohort (ISSCC) and 3672 participants underwent vibration controlled transient elastography from US National Health and Nutrition Examination Survey (US NHANES) were included. Liver Essential 5 was constructed by incorporating sleep quality into traditional healthy lifestyles (HLS).

Results: A total of 4.66-17.72% of the association between traditional HLS and MAFLD was mediated by sleep quality regardless of the detection techniques, and their joint associations on MAFLD were significant in both cohorts. ORs for individuals with poor sleep and unfavorable HLS were 1.72 (1.29-2.30) in ISSCC and 2.25 (1.55-3.26) in US NHANES, respectively. Around half of the participants previously considered as following a favorable HLS were re-classified by Liver Essential 5 with significantly higher prevalences of MAFLD in both cohorts (P < 0.001). Similar results were also found on at-risk MASH and significant fibrosis in US NHANES. ORs of participants with per one increment increase in Liver Essential 5 were 0.82 (0.77-0.89) and 0.79 (0.70-0.88) for MAFLD in ISSCC and US NHANES, 0.62 (0.48-0.78) for at-risk MASH and 0.78 (0.65-0.93) for significant fibrosis.

Conclusions: Liver Essential 5, which incorporates sleep quality and traditional lifestyle factors, provides additional risk stratification for MAFLD-related outcomes.

Fig. 1. Joint associations of sleep quality and traditional healthy lifestyles with the prevalence of MAFLD.

A, C The Imaging sub-cohort of South China Cohort (ISSCC). B, D The US National Health and Nutrition Examination Survey (NHANES). Models included US population and study design weights to account for the complex survey design. Odds ratio (95%CI) were adjusted for age, gender, race (only US NHANES), education, marriage and income level. Multiplicative interaction was evaluated using odds ratios for the product term between sleep quality and traditional healthy lifestyles, and additive interaction was evaluated using relative excess risk due to interaction (RERI), the attributable proportion due to interaction (AP) and the synergy index (S) between sleep quality (poor versus good) and traditional healthy lifestyle (0 to 1 behavior versus 3 to 4 behaviors). AP attributable proportion due to interaction, HLS healthy lifestyle, MAFLD metabolic dysfunction-associated fatty liver disease, OR odds ratio, RERI relative excess risk due to interaction, S synergy index.

Fig. 2. Risk classification of Liver Essential 5 and its association with MAFLD.

Liver Essential 5 showed risk re-classification potential on MAFLD on top of traditional healthy lifestyles in (A) the Imaging subcohort of South China Cohort (ISSCC) and (B) the US National Health and Nutrition Examination Survey (NHANES). Association of Liver Essential 5 and MAFLD in (C) the Imaging subcohort of South China Cohort (ISSCC) and (D) the US National Health and Nutrition Examination Survey (NHANES). All models were adjusted for age, gender, race (US NHANES), education, marriage and income level. The odds ratio (95%CI) and P values in US NHANES included the US population and study design weights to account for the complex survey design. MAFLD metabolic dysfunction-associated fatty liver disease, OR odds ratio.

Fig. 3. Subgroup analysis of the association between Liver Essential 5 and MAFLD.

A The Imaging sub-cohort of South China Cohort (ISSCC). B The US National Health and Nutrition Examination Survey (NHANES). All models were adjusted for age, gender, race (US NHANES), education, marriage and income level, except for itself. The odds ratio (95%CI) and P values in US NHANES included the US population and study design weights to account for the complex survey design. OR odds ratio.