Metabolic (dysfunction-) associated fatty liver disease (MAFLD), formerly known as non-alcoholic fatty liver disease (NAFLD), is characterized by metabolic dysfunction and excessive liver triglyceride accumulation. MAFLD significantly increases mortality risk. Currently, there are no approved pharmacological treatments, with lifestyle modifications—dietary control and increased physical activity (PA)—being the primary intervention. While combined approaches are most effective, many struggle to maintain both a healthy diet and regular exercise due to various barriers, including physiological, psychological, and socio-environmental factors. Low dietary quality is an independent risk factor for NAFLD-related mortality, often assessed using the Healthy Eating Index (HEI). This study aimed to investigate the interaction between diet quality (HEI score), PA levels, and mortality in MAFLD patients, focusing on physically inactive individuals using publicly accessible longitudinal data from NHANES III.

Existing research highlights the increased mortality risk associated with MAFLD (Kim et al., 2021; Nguyen et al., 2021; Lin et al., 2021). Lifestyle modification, encompassing dietary control and increased physical activity, is the recommended first-line treatment (Eslam et al., 2020; Younossi et al., 2021). However, adherence to these lifestyle changes is challenging (Glass et al., 2022; Burgess et al., 2017; Kim et al., 2020), with studies showing lower compliance with physical activity guidelines in individuals with MAFLD (Kim et al., 2020). Poor dietary quality is independently linked to increased mortality in NAFLD (Paik et al., 2022), highlighting the need to understand the combined impact of diet and PA on MAFLD outcomes. The HEI, a quantitative measure of diet quality, has been used to evaluate dietary patterns and their association with health outcomes (Kennedy et al., 1995; Guenther et al., 2008). However, research on the interaction between diet quality and PA levels in influencing MAFLD outcomes remains limited, making this study crucial.

This study utilized data from the Third National Health and Nutrition Examination Survey (NHANES III, 1988–1994) and linked mortality data from 2019. The NHANES III is a nationally representative cross-sectional survey assessing the health and nutritional status of non-institutionalized US individuals. Data were publicly accessible. Participants were followed until December 31, 2015. Hepatic steatosis was assessed using ultrasonography data from NHANES III. Physical activity was calculated by multiplying self-reported exercise frequency by its metabolic equivalent (MET). The population was divided into active and inactive groups based on the median total activity score. Dietary quality was assessed using the HEI score from 24-hour dietary recall data within the NHANES database. Income was based on the poverty income ratio; low income was defined as a ratio below 1. Education was categorized as low (<12 years). MAFLD diagnosis followed the international expert consensus statement, including ultrasonography-confirmed hepatic steatosis with either overweight/obesity, type 2 diabetes, or at least two metabolic risk factors in non-obese individuals. Liver fibrosis was estimated using FIB-4 and NFS scores. Other indicators included waist-to-hip ratio (WHR), BMI, HbA1c, ALT, AST, serum cholesterol, triglycerides, and eGFR. Quantitative variables were compared using t-tests or Mann-Whitney U-tests; qualitative variables were compared using χ² tests. Cox proportional hazard models were used to assess associations between HEI, PA, and mortality, adjusting for potential confounders. Restricted cubic splines (RCS) were used to visualize the relationship between HEI scores and mortality risk. Analyses were conducted using R software.

The final analysis included 3709 participants with MAFLD, with a median follow-up of 26.2 years. 1549 (41.8%) deaths were recorded. Multivariate Cox regression, adjusted for confounders, showed inverse correlations between both HEI score and PA level with all-cause mortality (P<0.05). Subgroup analysis stratified by PA level revealed that a higher HEI score was associated with lower all-cause, cardiovascular, and cancer-related mortality in the physically inactive MAFLD group (P<0.05), but these correlations were not observed in the physically active group. Kaplan-Meier curves showed better prognosis for those with active PA levels (log-rank P=0.024). In the physically inactive subgroup, the inverse association between HEI score and all-cause mortality remained significant in multivariate Cox regression analysis (HR=0.992, 95% CI: 0.986-0.997, P=0.002), whereas no significant association was found in the physically active subgroup (HR=0.999, 95% CI: 0.994-1.004, P=0.672). RCS and contour plots further illustrated these findings. While the overall analysis showed no correlation between HEI/PA levels and cause-specific (cardiovascular or cancer-related) mortality, subgroup analysis revealed an inverse relationship between HEI score and both cardiovascular and cancer-related mortality in the inactive group but not in the active group. Baseline characteristics differed between the physically active and inactive groups, with the active group having more males, lower educational levels, lower family income, lower BMI, and less severe metabolic derangements. The active group also tended to have higher HEI scores.

This study demonstrates the differential impact of dietary quality and physical activity as lifestyle interventions for MAFLD. In individuals with low PA levels, a healthy diet significantly improves long-term survival. However, dietary quality doesn't provide additional survival benefits for those already physically active. This suggests that the benefits of exercise might offset the negative effects of a poor diet. Previous research supports this, indicating that exercise can mitigate the harmful effects of poor diet (Bock et al., 2020; Duval et al., 2017; Krogh-Madsen et al., 2014; Bidwell et al., 2014). Conversely, for inactive individuals, diet becomes a crucial factor influencing long-term outcomes. Poor dietary quality is linked to oxidative stress and inflammation, impacting survival (Aleksandrova et al., 2021; Jayanama et al., 2021; Onvani et al., 2017). The study's results align with meta-analyses showing improved liver enzyme levels, liver fat, and histology with exercise, either alone or combined with dietary interventions (Katsagoni et al., 2017). Discrepancies with previous studies might stem from differing definitions of fatty liver disease (MAFLD vs. NAFLD) and the adjustment for PA levels. The study highlights the importance of individualized lifestyle recommendations, prioritizing either physical activity or dietary improvements based on an individual's lifestyle and capabilities.

This study highlights the importance of high-quality diet for improving survival outcomes in MAFLD patients, especially those with low physical activity levels. While physical activity is beneficial, a healthy diet is crucial for sedentary individuals. Future research should employ more objective measures of diet and physical activity to confirm these findings and investigate the underlying mechanisms. Longitudinal studies tracking lifestyle changes are needed to better understand the dynamic relationship between diet, physical activity, and long-term outcomes in MAFLD.

The study's limitations include reliance on self-reported data for diet and physical activity, which may introduce recall and reporting biases. The HEI, calculated from a single 24-hour dietary recall, provides a snapshot rather than a comprehensive assessment of long-term dietary habits. The method for categorizing physical activity, using the median calculated PA score, could lead to misclassification. Additionally, the study only assessed baseline exercise and diet; changes over time were not considered. Future studies should utilize more objective measures, such as wearable activity trackers and detailed food diaries, and incorporate repeated assessments to capture longitudinal changes in lifestyle factors and their impact on MAFLD outcomes.