Low-normal serum unconjugated bilirubin levels are associated with late but not early carotid atherosclerotic lesions in T2DM subjects.

Aims: We aimed to examine the association of serum unconjugated bilirubin (UCB) within normal limits with carotid atherosclerosis in Chinese patients with type 2 diabetes mellitus (T2DM).

Methods: This cross-sectional, real-world study was performed in 8,006 hospitalized T2DM patients including 4,153 men and 3,853 women with normal UCB. The subjects were stratified into quintiles based on serum UCB levels (<6.2, 6.2-7.9, 8.0-8.9, 9.0-10.9, and >10.9 μmol/l, respectively). Carotid atherosclerotic lesions detected by ultrasonography, including carotid intima-media thickness (CIMT), carotid plaque, and stenosis, were compared among the five groups. The associations of serum UCB levels and quintiles with carotid atherosclerotic lesions were also determined by multiple logistic regression.

Results: The prevalence of carotid plaque (55.3%, 49.5%, 47.4%, 43.8%, and 37.5%, respectively; < 0.001 for trend) and stenosis (15.2%, 12.2%, 9.1%, 7.7%, and 5.4%, respectively; < 0.001 for trend) was progressively lower across the UCB quintiles even after adjusting for age, sex, and duration of diabetes. Results of a fully adjusted multiple logistic regression analysis revealed that serum UCB levels and quintiles were significantly associated with carotid plaque and stenosis. Compared with the subjects in the lowest UCB quintile, the risk of carotid plaque decreased by 25.5%, 28.7%, 33.5%, and 42.8%, and that of carotid stenosis by 24.6%, 37.4%, 44.9%, and 47.3%, respectively, in those from the second to highest UCB quintiles. High serum UCB within the normal range was a protective factor against carotid plaque [odds ratio (OR) 0.810, 95% confidence interval (CI) 0.747-0.878; < 0.001] and stenosis [OR 0.722, 95% CI 0.647-0.805; < 0.001]. However, no significant association was observed between serum UCB and CIMT in T2DM patients. Furthermore, C-reactive protein (CRP) levels were significantly higher in the subjects with carotid atherosclerosis than in those without carotid atherosclerosis and clearly decreased across the UCB quintiles.

Conclusions: Serum UCB within normal limits is inversely associated with late carotid atherosclerotic lesions including carotid plaque and stenosis but not CIMT, an early carotid atherosclerotic lesion in T2DM patients. High-normal UCB may be protective against carotid atherosclerosis by its anti-inflammation effect, which was indicated by significantly decreased CRP levels from the lowest to highest UCB quintiles.