In a retrospective multicenter analysis of 42,842 patients undergoing myocardial perfusion imaging, age- and sex-specific percentile curves for epicardial adipose tissue (EAT) revealed striking cardiovascular risk gradients. Patients above the 95th percentile for EAT density faced a 70% higher risk of nonfatal myocardial infarction or death (aHR 1.70, 95% CI: 1.51–1.92) versus those below the 50th percentile over 3.6 years median follow-up. High BSA-indexed EAT volume carried a more modest but significant 30% elevated hazard. Crucially, EAT density identified elevated risk even in normal-BMI individuals.
Epicardial fat sits directly adjacent to coronary arteries, secreting pro-inflammatory adipokines and contributing to plaque vulnerability — mechanisms that BMI and waist circumference cannot capture. The finding that density outperforms volume as a prognostic marker aligns with emerging evidence that metabolically active, low-attenuation pericardial fat drives inflammation more potently than sheer tissue bulk. The cohort size here is among the largest ever assembled for EAT research, lending statistical credibility that smaller single-center studies lacked. However, critical limitations apply: this is observational and retrospective, causality cannot be established, and the population is enriched for those already referred for nuclear imaging — likely sicker than the general public. The online calculator is a practical translational step, though clinical uptake requires prospective validation. As a preprint not yet peer-reviewed, these findings should be interpreted cautiously. If confirmed, population-specific EAT percentiles could meaningfully upgrade routine cardiac CT reads.
