Abstract
Background The age-related trends in the predictive ability of carotid intima-media thickness (CIMT) for cardiovascular risk remain unclear. We aimed to identify the age-related trends in the predictive value of CIMT for cardiovascular death. Methods and Results In a prospective cohort of adults aged 35 to 75 years without history of cardiovascular disease who were enrolled between 2014 and 2020, we measured CIMT at baseline and collected the vital status and cause of death. We divided the study population into 4 age groups (35-44, 45-54, 55-64, and 65-75 years). Competing risk models were fitted to estimate the associations between CIMT and cardiovascular death. The added values of CIMT in prediction were assessed by the differences of the Harrell's concordance index and the net reclassification improvement index. We included 369 478 adults and followed them for a median of 4.7 years. A total of 4723 (1.28%) cardiovascular deaths occurred. After adjusting for the traditional risk factors, the hazard ratios for CIMT(mean) per SD decreased with age, from 1.27 (95% CI, 1.17-1.37) in the 35 to 44 years age group to 1.14 (95% CI, 1.10-1.19) in the 65 to 75 years age group (P for interaction <0.01). Meanwhile, the net reclassification improvement indexes for CIMT(mean) were attenuated with age, from 22.60% (95% CI, 15.56%-29.64%) in the 35 to 44 years age group to 7.00% (95% CI, -6.82% to 20.83%) in the 65 to 75 years age group. Similar results were found for maximum CIMT in all age groups. Conclusions CIMT may improve cardiovascular risk prediction in the young and middle-aged populations, rather than those aged ≥55 years.