Ceramides improve accuracy in prediction of preclinical atherosclerosis

Background: Ceramides are known to predict clinical cardiovascular diseases, but their role in prediction of subclinical atherosclerosis is unknown.

Subjects and methods: We investigated significance of plasma ceramides using LC-MS/MS technique in prediction of subclinical atherosclerosis assessed by carotid intima-media thickness (cIMT) in Young Finns Study cohort in 2007 (number:2197, age:30-45 years, women:45%). Statistical analysis was done with dichotomized cIMT data (cases: value ³ 90^th percentile vs. controls: value <90^th percentile). We focused on: i) risk ceramides- Cer(d18:1/16:0), Cer(d18:1/18:0), Cer(d18:1/24:1) and their ratios with protective ceramide - Cer(d18:1/24:0), and ii) Coronary Event Risk Test (CERT) score derived from these individual ceramides and their ratios. Reference predictive model with major traditional risk factors and two test predictive models with the ceramides or CERT scores added to the reference model were built. Risk factors were selected with bootstrap backward-stepwise algorithm. Model fitting and validation was done for 1000 bootstraps of the original data by: i) fitting models to training data (70% data), ii) testing the models on test data (30% data), and iii) calculate accuracy measure (ROC AUC).

Results: Our results suggest moderate (reference AUC 0.762, 95% CI [0.699, 0.822], ceramides test AUC 0.77, 95% CI [0.711, 0.826], CERT test AUC 0.769, 95%CI [0.707, 0.824]) but significant (reference vs ceramides t-test P-value: 7.91e-10, reference vs. CERT t-test P-value: 3.49e-07) added value of the ceramides in predictive model for subclinical atherosclerosis.

Conclusions: CERT and ceramide testing may help in predicting subclinical atherosclerosis for primary prevention more effectively than traditional risk factors alone.