Back

Subclinical Atherosclerosis Links Premature and Late-Onset Coronary Artery Disease

Guo, M.; Zhao, M.; Zhao, Y.; Wang, A.; Guo, X.; Tao, L.; Liu, J.

2025-11-09 epidemiology
10.1101/2025.11.06.25339722 medRxiv
Show abstract

BackgroundDespite established relationship between atherosclerosis and coronary artery disease (CAD), evidence on subclinical atherosclerosis and its differential associations with premature coronary artery disease (PCAD) versus late-onset coronary artery disease (LCAD) remains limited. AimsThis study aims to delve deeper into the associations between subclinical atherosclerosis and the incidence risk of both PCAD and LCAD. MethodsUsing UK Biobank data, we identified PCAD (male <55/female <65 years; n=7,398) and LCAD (male [&ge;]55/female [&ge;]65 years; n=39,085) cohorts. Conditional inference tree classification optimized carotid intima-media thickness (cIMT) stratification in both cohorts. ResultsConditional inference tree categorized the PCAD cohort into two subgroups: cIMT [&le;]700m and cIMT >700m, with the latter demonstrating a HR of 2.079 for cardiovascular risk. In the LCAD cohort, four cIMT strata were identified: [&le;] 620m, 620-763m (HR=1.401), 763-1054m (HR=1.810), and >1054m (HR=2.850). Multivariable-adjusted Cox models demonstrated significant associations between subclinical atherosclerosis and PCAD (HR=2.079, 95%CI:1.477-2.925) and LCAD (HR=1.776, 95%CI:1.455-2.169), highlighting the prognostic value of cIMT stratification in coronary artery disease risk assessment. ConclusionsA UK Biobank prospective cohort study revealed subclinical atherosclerosis significantly associated with PCAD and LCAD risks. Even within conventional cIMT "safe thresholds", incremental increases predicted elevated risks, underscoring limitations of current thresholds. Future research need to develop multimodal frameworks integrating dynamic cIMT trajectories to refine risk stratification and early interventions.

Matching journals

The top 5 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.