People with denser chest and back muscles are significantly less likely to have a heart attack, not because the muscles themselves protect the heart, but because their density signals an overall active lifestyle. That is the conclusion of a new analysis of the SCOT-HEART trial, published June 30 in Radiology by a team led by the University of Edinburgh.
The study used a deep learning tool called TotalSegmentator to automatically segment 104 anatomical structures from coronary CT angiography scans of 1,722 patients, none of whom had known heart disease at enrollment. The tool measured the volume and radiodensity, measured in Hounsfield units, of every organ and tissue in the scan field, enabling a comprehensive body composition analysis that would have been impractical by manual reading.
Over a 10-year follow-up period, 133 patients (7.7%) died and 106 (6.2%) had a heart attack. The strongest body composition predictor of both outcomes was skeletal muscle attenuation, the radiodensity of torso muscles, reflecting their composition rather than their size.
Density, not mass
For every 10-unit increase in skeletal muscle attenuation, meaning denser muscle with less intramuscular fat, the risk of heart attack fell by 31% (hazard ratio 0.69, 95% CI 0.54 to 0.87) and the risk of death from any cause fell by 39% (HR 0.61, 95% CI 0.47 to 0.79). Patients with muscle attenuation below the median had 1.85 times the mortality risk and 1.58 times the heart attack risk of those above it.
Muscle size was not protective. Larger muscles, greater cross-sectional area, showed no significant association with either outcome. This distinction is important: bulky torso muscles from targeted weight training appear to confer none of the cardiovascular benefit associated with lean, low-fat muscle tissue, which reflects overall metabolic health.
The association remained significant even after adjusting for coronary artery calcium score, one of the strongest established predictors of cardiac risk. Skeletal muscle density provided prognostic information beyond what standard plaque burden measures could capture.
Torso fat and liver fat showed no significant link to heart attack risk.
What it likely means
The researchers interpret muscle density primarily as a marker of overall physical activity. People with dense torso muscles are likely to be more active generally, and that broader activity, not the chest and back strength itself, is what protects the heart.
“It is likely that people who exercise enough to have strong muscles in their upper body have a healthy lifestyle which protects their heart in other ways,” said senior author Michelle C. Williams of the University of Edinburgh.
Bryan Williams, chief scientific officer at the British Heart Foundation, which partly funded the study, noted that “regular exercise can reduce the risk of cardiovascular disease by up to a third” and emphasized that all forms of exercise, not just strength training, can improve muscle density.
Caveats
The study is observational and retrospective. It cannot establish causation, whether improving muscle density through exercise directly reduces heart attack risk, or whether muscle density is simply a biomarker for people who are already healthier. The cohort was limited to patients who had been referred for CT angiography for chest pain, and the results may not generalize to asymptomatic people, younger populations, or non-white groups. Clinical thresholds for muscle density have not been established, and the authors call for validation in diverse cohorts before considering the metric for clinical decision-making.
The SCOT-HEART trial was originally designed to evaluate the role of CT angiography in chest pain assessment, not body composition. The machine learning analysis of existing scans revealed a robust association that the trial was never designed to test, statistically stable, mechanistically suggestive, but not yet ready for clinical deployment.
Source: Guimaraes AR, Williams SE, Macmillan MT, et al. Machine learning multiorgan analysis of coronary CT angiography body composition, myocardial infarction, and mortality in the SCOT-HEART trial. Radiology. 2026;319(3):e251821. doi:10.1148/radiol.251821