Longitudinal coronary CT angiography-derived body composition changes and subsequent cardiovascular events: a landmark cohort study

Clinically indicated coronary computed tomography angiography (CCTA) examinations contain extracoronary body-composition information, but the prognostic relevance of longitudinal within-person changes remains uncertain. We evaluated whether longitudinal changes in CCTA-derived body-composition measures were associated with subsequent major adverse cardiovascular events (MACE). This retrospective landmark cohort included 352 participants with at least two clinically indicated, quality-eligible CCTA examinations more than 1 year apart. In the primary analysis, the first quality-eligible examination was designated as baseline and the latest quality-eligible examination as the index date and time zero for outcome follow-up. Epicardial, visceral, intermuscular, subcutaneous, and skeletal muscle measures were quantified at both selected examinations. Annualized changes were standardized and evaluated in separate Cox models adjusted for clinical risk factors, baseline body mass index, and the corresponding baseline imaging measure. Additional sensitivity analyses evaluated robustness to interscan duration. During follow-up, 74 participants experienced MACE. After multivariable adjustment and false-discovery-rate correction, annualized increases in epicardial adipose tissue volume index, epicardial adipose tissue density, intermuscular adipose tissue index, visceral adipose tissue index, and visceral adipose tissue density were associated with higher MACE risk, whereas higher skeletal muscle density was associated with lower risk. Epicardial adipose tissue density had the largest point estimate (hazard ratio per 1-standard-deviation increase, 1.85; 95% confidence interval, 1.49–2.29). The corresponding hazard ratio for skeletal muscle density was 0.65 (95% confidence interval, 0.54–0.78). Longitudinal CCTA-derived changes in adipose tissue and skeletal muscle were associated with subsequent composite MACE in this selected single-center repeat-CCTA cohort. These findings do not establish prognostic value independent of coronary atherosclerotic disease burden. Because the measures were extracted from clinically indicated examinations, the findings support further evaluation of opportunistic cardiometabolic phenotyping rather than additional CCTA acquisition. Larger multicenter cohorts incorporating detailed coronary disease burden and external validation are required.