Atherosclerosis (AS) is a vascular disorder in which the arterial walls thicken as a result of the accumulation of fatty materials. AS is also a chronic inflammatory disease caused by the deposition of activated lymphocytes and macrophages and is promoted by low-density lipoproteins. When AS progresses, cytokines such as interleukin-6 (IL-6), tumor necrosis factor-α (TNF-α), and C-reactive protein (CRP) increase.^1,2) High sensitivity CRP (hsCRP) is associated with instability of atheromatous plaque and is a predictive indicator for cardiovascular morbidity.³⁾ Thus, an elevated level of hsCRP is associated with a risk of myocardial infarction in older people, patients with angina who have a high risk of cardiovascular disease (CVD), and also in healthy adults.⁴⁾ Additionally, similar to serum total cholesterol and the ratio of serum total cholesterol to high density lipoprotein (TC/HDL-C), an elevated level of hsCRP is an independent predictive indicator of cardiovascular morbidity⁵⁾ and is a stronger predictive risk factor for CVD than classical risk factors.⁶⁾ Schwertner et al.⁷⁾ reported that serum total bilirubin (sTB) was inversely correlated with risk of CVD, and Hunt et al.⁸⁾ found that the patients who had lower levels of sTB had up to an 80% increase in the risk of CVD compared to patients who had higher levels of sTB. However, there are few hypotheses about the relationship between sTB and CVD. According to a previous Korean study, sTB is a valuable indicator for AS of coronary arteries in patients with asymptomatic type II diabetes mellitus (DM). But there is little research regarding the relationship between sTB and CVD in Koreans. Therefore, we conducted the present study to analyze the relationship between sTB and hsCRP in middle-aged Korean men who visited a hospital health promotion center.

Table 1 shows the baseline characteristics of the 6,800 study subjects. Subjects were classified into two groups: the low-risk group (n = 5,100; hsCRP < 1.0 mg/L) and the average to high-risk group (n = 1,789; hsCRP > 1.0 mg/L) for CVD. Compared to the average to high-risk group, the low-risk group had lower body mass index (BMI); lower prevalence of current smokers and patients with DM and HT; and lower levels of FPG, systolic BP, ALT, and TC/HDL-C ratio (all P < 0.001). Additionally, the low-risk group had a lower level of sTB than the average to high-risk group (P = 0.03).

Table 2 shows the baseline characteristics of the four sTB groups (quartiles), which were analyzed using ANOVA and chi-square tests. The mean age of the group with the lowest sTB (sTB < 13.58 µmol/L) was significantly lower than the highest sTB group (sTB > 20.70 µmol/L) (44.31 ± 3.95 years vs. 44.09 ± 3.74 years, P < 0.001). BMI, prevalence of current smokers, and the TC/HDL-C ratio significantly decreased with increasing levels of sTB (P < 0.012 and P = 0.01, respectively). Levels of hsCRP also significantly decreased with increasing levels of sTB (P < 0.001). However, the frequency of alcohol consumption, percentage of patients who had DM or HT, levels of FPG, systolic BP, and ALT were not significantly different among the four groups.

We performed a multivariate logistic regression analysis to estimate ORs and 95% CIs of increased hsCRP according to quartiles of sTB levels and used three logistic regression models to adjust for various factors (Table 3). In model I, which adjusted only for age, when compared to the lowest quartile of sTB, the ORs (95% CI) for hsCRP in the 2nd, 3rd, and 4th sTB quartiles were 0.82 (0.70 to 0.95), 0.73 (0.63 to 0.85), and 0.61 (0.52 to 0.71), respectively, and these were statistically significant (P for trend < 0.001). In both model II (adjusted for age, smoking status, and alcohol consumption) and model III (adjusted for BMI, ALT, FPG, systolic BP, and TC/HDL-C ratio as well as the factors in model II), when compared to the lowest quartile of serum total bilirubin, the ORs (95% CI) for hsCRP in the 2nd, 3rd, and 4th sTB quartiles significantly decreased (all P for trend < 0.001).

In this study, hsCRP was inversely associated with sTB. In addition, hsCRP had significant positive associations with BMI; smoking status; medical history of DM and HT; and levels of systolic BP, FPG, ALT, and the TC/HDL-C ratio, but was not associated with age or alcohol consumption.

Serum bilirubin is known to have an antioxidant effect and an inverse relationship with the risk for CVD.^11-14) Bilirubin is a final metabolite of heme, but the function of bilirubin is not clear. Vitek et al.¹⁵⁾ reported that patients with Gilbert syndrome, who have a slightly elevated level of serum direct bilirubin, had a lower risk of CVD compared to healthy subjects. According to the NHANES study (1999-2004),¹³⁾ a 1.71 micromol/L (0.1 mg/dL) increment in bilirubin level was associated with a 9% reduction in the odds of stroke (OR, 0.91; 95% CI, 0.86 to 0.96) among the civilian population and with a 10% reduction in the odds of stroke (OR, 0.90; 95% CI, 0.80 to 1.00) among patients with a history of stroke. Nobotny and Vitek¹⁴⁾ reported that serum bilirubin in men had an inverse relationship with the severity of AS in a meta-analysis of 11 studies. We found that after adjusting for other risk factors of CVD, sTB was inversely associated with hsCRP, which is a predictive indicator for CVD, and this finding is consistent with the results of previous studies. Although this was a cross-sectional study and our participants were middle-aged Korean men who visited a health promotion center, we believe the number of study participants was sufficient to evaluate the relationship between sTB and hsCRP.

There have been several studies about the relationship between hsCRP and risk factors for CVD. Nakanishi et al.¹⁶⁾ reported that hsCRP had a significant relationship with age, smoking status, serum fasting glucose, BMI, BP, HDL-C, and triglycerides. In a previous Korean study,¹⁷⁾ hsCRP had a positive relationship with serum fasting glucose, age, and BP. However, hsCRP was not significantly associated with age in our study. This result is possibly due to the narrow age distribution of our study sample (40 to 59 years old).

Mild to moderate alcohol consumption reduces hsCRP. Several studies^18-20) reported that men who drank one to four drinks of alcoholic beverages per week had the lowest levels of hsCRP, and the relationship between alcohol consumption and hsCRP was U shaped. However, this study did not detect a significant association between alcohol consumption and hsCRP. We did not collect specific information about the amount of alcohol consumed by study subjects because of our inability to validate the information. Serum bilirubin is affected by aging, smoking status, and fasting state, among other factors, and has an inverse relationship with risk factors for CVD, such as aging, obesity, smoking, HT, and dyslipidemia.^7,21,22) We found that sTB had an inverse association with age, BMI, smoking status, and the TC/HCL-C ratio. However, in this study, sTB did not demonstrate a significant relationship with HT or systolic BP. A previous Korean study²³⁾ reported that serum bilirubin had a significant relationship with HT in women but not in men. Therefore, we suggest that future studies about the association between sTB and BP or HT in the general population should include women and older people.

We transformed the levels of sTB measured in this study to SI units (µmol/L). The SI unit for sTB levels contains more information and is better at detecting changes of serum bilirubin within physiologic levels.¹⁴⁾

However, further studies about the relationships between other forms of serum bilirubin and risk factors for CVD are necessary. It is not clear how serum bilirubin lowers the risk of CVD. One potential mechanism is the antioxidant and anti-inflammatory effects of bilirubin through inhibition of oxidation of low-density lipoprotein cholesterol (LDL-C) and oxygen radicals, which may consequently prevent AS. Another possible mechanism is the association of high-density serum bilirubin with lower levels of inflammation.^24-26)

One of the limitations of this study was that we included only the total level of serum bilirubin in the analysis to investigate the association with hsCRP and risk factors for CVD; serum direct and indirect bilirubin were not included in the analysis. Because all forms of bilirubin (free or albumin-bound) inhibit hyperoxidation of LDL-C,²⁷⁾ this limitation might not have affected our results. The other limitation was the age distribution of the study sample. The study subjects were middle-aged Korean men (40-59 years old); therefore, the results may not be necessarily generalized to other populations in or outside of Korea. The third limitation was that this study was a cross-sectional study, which makes it difficult to detect a causal relationship between sTB and hsCRP and risk factors for CVD.

In conclusion, this study showed that sTB was inversely associated with hsCRP and risk factors for CVD. Despite some limitations, this study is meaningful because it is a large study of subjects who visited a health promotion center. We anticipate that the results of this study will be used in future studies.