Dynamic versus Static Metabolic Models for Predicting ECG-Defined Cardiovascular Risk in Elderly MAFLD: A Three-Year Cohort Study

Purpose: Given the high and rising prevalence of metabolic dysfunction-associated fatty liver disease (MAFLD) among the aging population and its established link to cardiovascular risk, this study aimed to evaluate the predictive value of dynamic metabolic trajectories for ECG-defined cardiovascular risk in elderly patients with MAFLD. Patients and Methods: This three-year longitudinal study enrolled 1086 elderly patients with MAFLD from Weifang City, Shandong Province, China. Group-based trajectory modeling (GBTM) was applied to identify dynamic changes in 12 metabolic indicators. The predictive performance of the metabolic trajectory model was compared with that of the cross-sectional model using 5-fold cross-validation. Multivariable logistic regression was employed to evaluate the independent associations between specific metabolic trajectories and ECG-defined cardiovascular risk. Results: During follow-up, 877 participants (80.76%) developed new ECG abnormalities. The trajectory model demonstrated a modest but statistically significant improvement in discrimination over the cross-sectional model (ΔAUC = 0.054). Specific progressively worsening metabolic trajectories were strongly associated with increased risk: the “Obesity-Increasing” BMI, “Moderate Hypertension-Increasing” SBP, and “High Level-Increasing” TC trajectories. Notably, a “legacy effect” of liver injury was evident: patients whose elevated AST later declined (“High-Decreasing” trajectory) still faced substantially elevated cardiovascular risk (aOR = 4.15; aRR = 1.20). Conversely, the “Moderate Diabetes-Decreasing” FPG trajectory (aOR = 0.49; aRR = 0.82) and adherence to a predominantly vegetarian diet (aOR = 0.22; aRR = 0.61) were associated with significantly lower risk. Advanced age remained a strong independent risk factor. Conclusion: Dynamic metabolic trajectories offer incremental predictive value over static measures in predicting ECG-defined cardiovascular risk in elderly MAFLD patients. Clinical management should shift from state-based to trend-based intervention, focusing on early control of adverse trends, long-term vigilance for patients with a history of liver injury, and active improvement of reversible risk factors. Tailored dietary interventions are also recommended. These findings provide an evidence-based foundation for developing precise and proactive risk prevention strategies in this high-risk population. Plain Language Summary: Why was the study done? Heart disease is common in older adults with MAFLD. Current risk assessments often use single measurements, which may miss important health trends over time. We conducted this study to see whether tracking metabolic changes over several years could better predict heart risk. What did the researchers do and find? We followed 1086 older adults with MAFLD for three years. We identified patterns in how 12 metabolic indicators changed over time, grouping people based on improving, worsening, or stable trends. We then tested whether these trend-based groups predicted future heart risk (measured by ECG changes) better than a single measurement. We found that the trajectory-based model predicted CVD risk much better than one-time measurements. Patients with worsening groups—especially increasing obesity, rising blood pressure, or climbing cholesterol—faced the highest risk. Even after liver enzyme levels normalized, those with past liver injury were linked to a lasting higher risk. Conversely, people who improved their blood sugar or mostly ate vegetarian foods had much lower risk. What do these results mean? For elderly patients with MAFLD, tracking health trends over time provides a clearer picture of future heart risk than a single check. This approach enables earlier and more personalized care—like helping patients avoid worsening trends, offering long-term monitoring for those with past liver issues, and promoting beneficial diets. These steps can prevent disease, improve well-being, and use healthcare resources more effectively. Keywords: group-based trajectory modeling, electrocardiography, risk prediction, precision prevention, geriatrics