epicardial adipose tissue (eAt) is associated with cardiovascular risk. the longitudinal change in eAt volume (eAtv) and density (eAtd), and potential modulators of these parameters, has not been described. We prospectively recruited 90 patients with non-obstructive coronary atherosclerosis on baseline computed tomography coronary angiography (ctcA) performed for suspected coronary artery disease to undergo a repeat research CTCA. EATv in millilitres (mL) and EATd in Hounsfield units (HU) were analysed and multivariable regression analysis controlling for traditional cardiovascular risk factors (cVRf) performed to assess for any predictors of change. Secondary analysis was performed based on statin therapy. The median duration between CTCA was 4.3years. Mean EATv increased at follow-up (72 ± 33 mL to 89 ± 43 mL, p < 0.001) and mean EATd decreased (baseline −76 ± 6 HU vs. −86 ± 5 HU, p < 0.001). There were no associations between baseline variables of body mass index, age, sex, hypertension, hyperlipidaemia, diabetes or smoking on change in EATv or EATd. No difference in baseline, follow-up or delta EATv or EATd was seen in patients with (60%) or without baseline statin therapy. in this select group of patients, eAtv consistently increased and eAtd consistently decreased at long-term follow-up and these changes were independent of cVRf, age and statin use. together with the knowledge of strong associations between EAT and cardiac disease, these findings may suggest that eAt is an independent parameter rather than a surrogate for cardiovascular risk. open Scientific RepoRtS | (2020) 10:7109 | https://doi.org/10.1038/s41598-020-63135-z www.nature.com/scientificreports www.nature.com/scientificreports/ Scientific RepoRtS | (2020) 10:7109 | https://doi.org/10.1038/s41598-020-63135-zwww.nature.com/scientificreports www.nature.com/scientificreports/ however, this is reflective of the current literature in examining relevant associations of EAT. Finally, there is potential for error in using delta EAT values with potential overlap from test-retest variability. Our previous work has demonstrated limits of agreement up to 10 mL higher or lower between observers with a mean bias however of only 1 mL, however our inter-observer correlation was excellent at 0.98 with assessors blinded to scan timing and patient details. conclusion Epicardial adipose tissue volume and density demonstrate significant longitudinal changes in patients with non-obstructive coronary artery disease with a consistent increase in EAT volume and consistent decrease in EAT density. There are no clinical risk factors that appear to associate with the change in EAT parameters and this effect is also independent of statin therapy. This finding may suggest that EAT is an independent marker, rather than surrogate of cardiovascular risk.