Background. A prior cross sectional study reported that higher soluble thrombomodulin (sTM) was associated with subclinical atherosclerosis only in the presence of higher soluble intercellular adhesion molecule-1, sICAM-1. We evaluated this interaction longitudinally with regard to coronary artery calcium (CAC) in the MESA cohort.
Methods: MESA is a multi-center longitudinal study with baseline measurements of endothelial biomarkers in a random sample (n=1000) of the baseline cohort. In the random sample, 374 men and 496 women had CT measurement for CAC (Agatston score) at baseline and a median follow-up of 2.9 years. Among those with no baseline CAC (n=490), we estimated the relative risk of detectable CAC at follow-up using general linear models with Gaussian error and robust standard errors. Among those with detectable baseline CAC (n=380), change in CAC on follow-up was modeled using robust regression that down-weights outliers. Models with log(sTM) as predictor were adjusted for follow-up time, sex, ethnicity and baseline age, BMI, smoking, diabetes, systolic blood pressure, total cholesterol, use of BP or lipid-lowering medications and for hormone replacement (HRT) in women in separate models. Interactions between sTM and sICAM-1, sex and ethnicity was assessed.
Results: The median [interquartile range] of sTM was 38 [28 to 47] ng/mL for persons with incident detectable CAC, and 30 [23 to 41] ng/mL in those without (rank sum p <0.001). On adjustment, a 2-fold higher sTM at baseline (e.g., from 23 to 46 ng/mL) was associated with a 1.36-fold increased risk of detectable CAC on follow-up (95% CI: 1.03, 1.79, p=0.031). Among those with baseline detectable CAC, the rank correlation between sTM level and change in CAC score was 0.10 (p = 0.042), however, on adjustment, a 2-fold higher sTM was associated with a non-significant 0.86 Agatston units greater CAC change (p = 0.84). Higher sICAM-1 did not modify the association of sTM with CAC incidence (interaction p = 0.15) or change (interaction p = 0.26). There was no significant heterogeneity by sex, or ethnicity, or confounding by HRT in women.
Conclusion: High circulating levels of sTM were independently associated with incident calcification of the coronary arteries.
PERBANDINGAN REGRESI ROBUST PENDUGA MM DENGAN METODE RANDOM SAMPLE CONSENSUS DALAM MENANGANI PENCILAN