Common carotid artery inima-media thickness trajectories over time are more strongly associated with incident cardiovascular disease than a one off measurement

Atherosclerotic cardiovascular disease (CVD) remains the leading cause of mortality and morbidity globally. Several studies have previously looked into whether carotid artery intima-media thickness (CCA-IMT) measured at a single point in time is associated with CVD risk. More recent analyses have considered progression of CCA-IMT over time as a risk predictor, but measurement of progression is particularly prone to measurement error. Modelling the trajectories allows estimating the true strength of the aetiological association with CVD risk more accurately.

The aim of the present study is (i) to model within-person trajectories of CCA-IMT incorporating repeat measurements taken over an extended period of time, thereby reducing measurement error, and (ii) to quantify the association of within-person CCA-IMT trajectories over time with the risk of a CVD event.

Individual participant data from studies within the Proof-ATHERO consortium were used for this analysis. The Proof-ATHERO consortium collates data from studies with information on atherosclerosis measures, CV risk factors and CV outcomes. Study-specific joint longitudinal and survival models were fitted to model the association between CCA-IMT trajectory and CVD risk, accounting for conventional CV risk factors including: age, sex, systolic blood pressure (SBP), high-density lipoprotein cholesterol (HDL-C), total cholesterol, smoking status, history of diabetes and history of CVD. Study-specific hazard ratios were pooled in a random effects meta-analysis.

Data from 38 studies were analysed, involving 68,077 individuals and 153,643 CCA-IMT measurements. 51% of the individuals were female, the weighted mean (SD) age of participants at baseline was 61 (9) years, 13% of individuals had a history of CVD. Over a median (5th percentile, 95th percentile) follow-up of 7.3 (0, 28.6) years, 13,116 (19.3%) individuals suffered an incident CVD event. At baseline, the weighted mean (SD) CCA-IMT was 0.83 (0.19) mm, SBP was 135 (19) mmHg, HDL-C was 1.29 (0.40) mmol/L, total cholesterol was 5.49 (1.06) mmol/L; 26% were current smokers and 16% had a history of diabetes. Trajectories of CCA-IMT were independently associated with CVD risk after accounting for conventional CVD risk factors. The pooled adjusted hazard ratio for CVD (95% CI) per study-specific SD higher CCA-IMT value was 1.20 (1.14-1.27) (p<0.001). In comparison, a one off measurement of CCA-IMT at baseline was associated with a pooled hazard ratio of 1.11 (1.06-1.16) per study-specific SD higher value (p<0.001).

The association of CCA-IMT trajectories based on repeated measurements with incident CVD risk is stronger than the association of a one off measurement.