The objective of this study was to investigate the effects of low serum bilirubin levels on the risk for future coronary artery disease (CAD) in a prospective cohort. CAD events were examined according to baseline serum bilirubin levels in a prospective large-scale, community-based Korean cohort in 2 subsequent prospective biennial surveys. A total of 8,593 subjects were included, 0.9% of whom reported newly developed CAD events during the 4 years of follow-up. Cox regression analyses showed that the lowest serum total bilirubin level category (bilirubin ≤0.32 mg/dl) was an independent risk factor for future CAD events (adjusted hazard ratio [HR] 1.890, 95% confidence interval [CI] 1.088 to 3.284; p = 0.024). Subjects with metabolic syndrome had a higher risk for future CAD events than those without metabolic syndrome (HR 3.366, 95% CI 2.079 to 5.448, p <0.001). Low bilirubin levels increased the CAD risk in subjects with metabolic syndrome further (HR 2.016, 95% CI 1.069 to 3.800; p = 0.030), with these subjects showing a >6 times higher risk for CAD than subjects with bilirubin levels >0.32 mg/dl and no metabolic syndrome (HR 6.228, 95% CI 3.118 to 12.437; p <0.001). In conclusion, the addition of low serum bilirubin levels to the traditional risk factors for CAD, such as metabolic syndrome, may yield an improvement of risk prediction.
For decades, bilirubin, an end product of heme metabolism, was considered a cytotoxic waste product. However, recent studies have demonstrated that bilirubin has anti-inflammatory and antioxidative properties, likely mediated by efficient scavenging of peroxyl radicals and inhibition of the oxidation of low-density lipoprotein (LDL), which are known factors related to atherosclerosis and coronary artery disease (CAD). Serum bilirubin concentrations have been reported to be inversely associated with CAD risk in several prospective studies, indicating that serum bilirubin may be used to predict the risk for CAD. However, it has not been elucidated whether bilirubin can provide additional information to the conventional risk assessment. Because risk prediction plays a major role in CAD prevention, it is important to identify novel risk factors for CAD and to refine and improve risk assessment methods. The present study investigated whether low bilirubin levels could serve as a novel independent biomarker when added to the traditional risk factors in a community-based prospective cohort.
Methods
The design and baseline data of the Ansung-Ansan cohort study have been published previously. Baseline examinations were performed in 2001 and 2002, and follow-up examinations have been carried out every 2 years since. The present study involved the results of 4 years of follow-up. Eligible subjects were those aged 40 to 69 years who had been residents of the survey area for ≥6 months. A total of 10,038 subjects participated in the Ansung-Ansan cohort study, comprising 5,018 subjects of the 7,192 eligible subjects from Ansung (70% response rate) and 5,020 of the 15,580 eligible subjects from Ansan (32.4% response rate). Subjects with histories of CAD, electrocardiographic abnormalities suggestive of ischemic heart disease, and highly elevated aminotransferase levels (5 times the upper normal limit; aspartate aminotransferase or alanine aminotransferase ≥200 IU/L) at baseline were excluded, resulting in a total of 8,593 subjects included in the study.
The study protocol and any amendments were approved by the ethics committee of the Korean Center for Disease Control and Prevention and the Ajou University School of Medicine Institutional Review Board, and written informed consent was obtained from all participants. The study was carried out in accordance with the Declaration of Helsinki.
Data collection and examinations were performed using the same instruments by trained researchers throughout the study. Blood pressure and anthropometric measurements were checked using the standardized method, and laboratory data were analyzed at a central laboratory. Participants completed questionnaires during face-to-face interviews regarding demographics and medical history.
The definition of a CAD event was based on the questionnaire administered at baseline and at the biennial follow-up visits. Subjects with CAD events were defined as those who reported taking medications or having undergone reperfusion therapy or bypass surgery for any CAD, including myocardial infarction. On the basis of the revised National Cholesterol Education Program Adult Treatment Panel III criteria, we defined metabolic syndrome as the presence of ≥3 of the following 5 components: abdominal obesity, waist circumference >90 cm for men and >80 cm for women, blood pressure >130/85 mm Hg, triglyceride >150 mg/dl, high-density lipoprotein (HDL) cholesterol <40 mg/dl in men and <50 mg/dl in women, and elevated blood glucose levels and fasting blood glucose >100 mg/dl. The Framingham risk score (FRS) is derived from equations of traditional risk factors, including age, gender, total and HDL cholesterol, systolic blood pressure, diabetes mellitus, and smoking, as previously described.
All data were analyzed using SPSS version 18.0 (SPSS, Inc., Chicago, Illinois). The baseline characteristics are described as mean ± SD or number (percentage). To determine the differences in baseline clinical characteristics according to the development of CAD during the follow-up period, linear regression and logistic regression analyses adjusted for age and gender were used. Bilirubin levels were categorized into increments of 0.32 mg/dl (approximately 1 SD, a statistically relevant difference in bilirubin concentration), and the clinical characteristics according to bilirubin and gender were analyzed by analysis of covariance for continuous variables adjusted for age and by chi-square tests for categorical variables. The post hoc test for analysis of covariance was performed by the Bonferroni method. Cox proportional-hazard regression was used to assess the risk for CAD events during the follow-up period according to bilirubin levels. Statistical significance was defined as p <0.05.
Results
The baseline characteristics according to CAD events during 4 years of follow-up are listed in Table 1 . Of the 8,593 participants, 80 (35 men [0.9%] and 45 women [1.0%]) developed CAD. Among 80 subjects with CAD events during the follow-up period, 17 reported having undergone reperfusion therapy in that period. Compared with subjects who did not develop CAD, those who developed CAD at follow-up were older and had higher body mass indexes, lower bilirubin levels, and higher γ-glutamyl transferase levels at baseline. Diabetes mellitus, hypertension, and hypertriglyceridemia were also more frequently found at baseline in subjects with future CAD events. However, baseline LDL and HDL cholesterol levels did not significantly differ between subjects with and without CAD events.
| Variable | Total | CAD at Follow-up | P ∗ | ||
|---|---|---|---|---|---|
| No | Yes | ||||
| N (% Men) | 8593 (47.0%) | 8513 (47.0%) | 80 (43.8%) | 0.562 † | |
| Age (years) | 52 ± 9 | 51 ± 9 | 57 ± 8 | <0.001 | |
| Body mass index (kg/m 2 ) | 24.6 ± 3.1 | 24.6 ± 3.1 | 25.4 ± 3.2 | 0.011 | |
| Systolic blood pressure (mmHg) | 117 ± 18 | 117 ± 18 | 123 ± 16 | 0.253 | |
| Diastolic blood pressure (mmHg) | 75 ± 12 | 75 ± 12 | 76 ± 9 | 0.859 | |
| Fasting glucose (mg/dl) | 89.1 ± 22.2 | 89.0 ± 22.2 | 92.3 ± 23.9 | 0.227 | |
| Total cholesterol (mg/dl) | 193.9 ± 36.3 | 193.8 ± 36.3 | 200.6 ± 40.4 | 0.158 | |
| Low-density lipoprotein cholesterol (mg/dl) ‡ | 119.2 ± 35.4 | 119.2 ± 35.3 | 120.0 ± 36.1 | 0.906 | |
| High-density lipoprotein cholesterol (mg/dl) | Men | 44.9 ± 10.6 | 44.9 ± 10.6 | 47.8 ± 13.6 | 0.191 |
| Women | 47.7 ± 11.0 | 47.7 ± 11.1 | 46.4 ± 10.9 | 0.622 | |
| Triglyceride (mg/dl) ‡ | 157.1 ± 102.3 | 156.9 ± 102.4 | 177.5 ± 87.2 | 0.022 | |
| Creatinine (mg/dl) | 0.8 ±0.2 | 0.8 ± 0.2 | 0.8 ± 0.2 | 0.570 | |
| Bilirubin (mg/dl) ‡ | 0.60 ± 0.32 | 0.60 ± 0.32 | 0.52 ± 0.27 | 0.043 | |
| Aspartate aminotransferase (IU/L) | 27.7 ± 12.9 | 27.7 ± 12.9 | 25.8 ± 8.7 | 0.122 | |
| Alanine aminotransferase (IU/L) | 26.0 ± 16.6 | 26.0 ± 16.6 | 25.1 ± 13.6 | 0.884 | |
| Gamma-glutamyl transferase (IU/L) ‡ | 35.1 ± 58.4 | 35.0 ± 58.5 | 41.6 ± 50.4 | 0.004 | |
| High-sensitive C-reactive protein (mg/dl) ‡ | 0.23 ± 0.53 | 0.23 ± 0.53 | 0.29 ± 0.28 | 0.008 | |
| Smoking | Non-smoker | 5045 (59.3%) | 5001 (59.4%) | 44 (55.0%) | 0.128 |
| Ex-smoker | 1313 (15.4%) | 1299 (15.4%) | 14 (17.5%) | 0.051 | |
| Smoker | 2148 (25.3%) | 2126 (25.2%) | 22 (27.5%) | 0.954 | |
| Drinker § | 4074 (47.4%) | 4040 (47.5%) | 34 (42.5%) | 0.946 | |
| Family history of CAD | 24 (0.3%) | 24 (0.3%) | 0 (0%) | 0.998 | |
| Diabetes mellitus ¶ | 1191 (13.9%) | 1167 (13.8%) | 24 (30.8%) | 0.002 | |
| Hypertension ‖ | 1813 (21.1%) | 1783 (20.9%) | 30 (37.5%) | 0.030 | |
| Current lipid-lowering agents use | 34 (0.4%) | 34 (0.4%) | 0 (0%) | 0.998 | |
| Triglyceride ≥150 mg/dl | 3397 (39.6%) | 3352 (39.4%) | 45 (56.3%) | 0.006 | |
| High-density lipoprotein cholesterol <40 mg/dl for men, <50 mg/dl for women | 4216 (49.1%) | 4177 (49.1%) | 39 (48.8%) | 0.769 | |
| Low-density lipoprotein cholesterol ≥160 mg/dl | 975 (11.4%) | 966 (11.4%) | 9 (11.3%) | 0.899 | |
| Low-density lipoprotein cholesterol ≥130 mg/dl | 2984 (34.7%) | 2957 (34.8%) | 27 (33.8%) | 0.719 | |
| Living in urban area (Ansan) | 4637 (54.0%) | 4607 (54.1%) | 30 (37.5%) | 0.186 | |
∗ Age and sex-adjusted P value between groups.
† Sex difference among groups.
‡ Logarithmically transformed before analysis.
§ Defined as who had consumed alcoholic beverage ≥1 drink per month.
¶ Defined as diagnosis of diabetes mellitus and/or as taking oral anti-diabetic agents or insulin.
‖ Defined as a systolic blood pressure ≥140 mmHg or a diastolic blood pressure ≥90 mmHg or as taking antihypertensive medications.
The mean serum total bilirubin concentration in the entire cohort was 0.60 ± 0.32 mg/dl (range 0.11 to 3.74). Bilirubin levels were categorized into 5 groups, with increments of 0.32 mg/dl: ≤0.32, 0.33 to 0.64, 0.65 to 0.96, 0.97 to 1.28 and >1.28 mg/dl. Clinical characteristics were analyzed separately by gender, because bilirubin levels in men were significantly higher than in women ( Tables 2 and 3 ). Subjects with the lowest bilirubin levels were significantly older than those in the other groups, such that all subsequent analyses were performed with adjustment for age. In the group with the lowest bilirubin levels (≤0.32 mg/dl), subjects had a tendency toward higher glycated hemoglobin, homeostasis model assessment of insulin resistance, and high-sensitivity C-reactive protein ( Tables 2 and 3 ). The proportion of smokers was higher among subjects with lower bilirubin levels. Subjects with diabetes or hypertension at baseline were also more frequently observed in the group with low bilirubin levels, although there was no statistical significance in men. Accordingly, bilirubin was found to be negatively correlated with CAD risk factors such as age, insulin resistance, high-sensitivity C-reactive protein, diabetes, and hypertension.
| Variable | Serum Total Bilirubin (mg/dl) Levels | P | |||||
|---|---|---|---|---|---|---|---|
| ≤0.32 | 0.33–0.64 | 0.65–0.96 | 0.97–1.28 | >1.28 | |||
| N | 287 | 1869 | 1236 | 417 | 227 | ||
| Age (years) | 55.3 ± 9.0 ∗ , † , ‡ , § | 52.3 ± 8.7 † , ‡ , § | 50.7 ± 8.4 | 49.4 ± 8.3 | 49.0 ± 8.0 | <0.001 | |
| Body mass index (kg/m 2 ) | 23.2 ± 3.1 ∗ , † , ‡ , § | 24.2 ± 3.0 | 24.5 ± 2.8 | 24.3 ± 2.8 | 24.3 ± 2.7 | <0.001 | |
| Systolic blood pressure (mmHg) | 121.4 ± 18.1 | 117.8 ± 16.5 | 116.3 ± 16.2 | 116.3 ± 16.9 | 114.5 ± 15.5 | 0.125 | |
| Diastolic blood pressure (mmHg) | 76.7 ± 10.8 | 76.0 ± 11.0 | 76.2 ± 11.3 | 76.7 ± 11.6 | 75.4 ± 11.0 | 0.383 | |
| HbA1c (%) | 5.9 ± 0.9 § | 5.9 ± 0.9 § | 5.8 ± 0.9 § | 5.7 ± 1.0 § | 5.5 ± 0.7 | <0.001 | |
| Triglyceride (mg/dl) ¶ | 166.2 ± 98.4 | 177.6 ± 114.2 ‡ , § | 175.2 ± 126.2 | 161.0 ± 93.8 | 159.8 ± 118.7 | 0.001 | |
| Fasting insulin (μIU/ml) | 7.9 ± 6.8 † , ‡ , § | 7.4 ± 4.8 † , ‡ , § | 6.9 ± 4.1 | 6.2 ± 3.1 | 6.2 ± 3.3 | <0.001 | |
| Homeostasis model assessment of insulin resistance ‖ | 1.7 ± 1.8 ‡ , § | 1.7 ± 1.2 ‡ , § | 1.6 ± 1.0 | 1.5 ± 0.9 | 1.4 ± 0.9 | <0.001 | |
| Aspartate aminotransferase (IU/L) | 29.5 ± 13.7 § | 29.2 ± 12.8 † , ‡ , § | 31.0 ± 14.9 § | 31.4 ± 16.1 § | 38.2 ± 25.1 | <0.001 | |
| Alanine aminotransferase (IU/L) | 28.8 ± 17.2 | 30.5 ± 18.0 | 31.9 ± 19.9 | 30.8 ± 18.2 | 34.0 ± 22.1 | 0.129 | |
| Gamma-glutamyl transferase (IU/L) ¶ | 48.4 ± 66.1 † | 50.9 ± 80.0 | 54.0 ± 75.0 | 52.6 ± 67.4 | 60.2 ± 109.5 | 0.013 | |
| High-sensitive C-reactive protein (mg/dl) ¶ | 0.35 ± 0.64 ‡ , § | 0.26 ± 0.51 ‡ , § | 0.23 ± 0.35 ‡ | 0.20 ± 0.64 | 0.25 ± 0.73 | <0.001 | |
| Smoking | Non-smoker | 42 (14.8%) | 309 (16.6%) | 275 (22.3%) | 102 (24.6%) | 54 (23.8%) | <0.001 |
| Ex-smoker | 63 (22.2%) | 506 (27.2%) | 433 (35.1%) | 159 (38.3%) | 94 (41.4%) | ||
| Smoker | 179 (63.0%) | 1044 (56.2%) | 526 (42.6%) | 154 (37.1%) | 79 (34.8%) | ||
| Diabetes mellitus | 44 (15.6%) | 284 (15.3%) | 191 (15.5%) | 62 (14.9%) | 23 (10.3%) | 0.352 | |
| Hypertension | 60 (20.9%) | 383 (20.5%) | 251 (20.3%) | 84 (20.1%) | 37 (16.3%) | 0.673 | |
∗ Significantly different from subjects with serum total bilirubin level 0.33–0.64 mg/dl.
† Significantly different from subjects with serum total bilirubin level 0.65–0.96 mg/dl.
‡ Significantly different from subjects with serum total bilirubin level 0.97–1.28 mg/dl.
§ Significantly different from subjects with serum total bilirubin level >1.28 mg/dl.
¶ Logarithmically transformed before analysis.
‖ Calculated using the following formula: (fasting plasma glucose [mmol/L] × fasting serum insulin [mU/L])/22.5.
| Variable | Serum Total Bilirubin (mg/dl) Levels | P | |||||
|---|---|---|---|---|---|---|---|
| ≤0.32 | 0.33–0.64 | 0.65–0.96 | 0.97–1.28 | >1.28 | |||
| N | 839 | 2727 | 740 | 190 | 61 | ||
| Age (years) | 54.0 ± 9.2 ∗ , † , ‡ , § | 52.1 ± 8.9 ‡ , § | 51.2 ± 8.7 | 49.9 ± 8.8 | 48.5 ± 8.2 | <0.001 | |
| Body mass index (kg/m 2 ) | 24.8 ± 3.4 | 24.9 ± 3.3 ‡ | 24.9 ± 3.1 ‡ | 24.0 ± 3.1 | 24.9 ± 3.4 | 0.009 | |
| Systolic blood pressure (mmHg) | 120.0 ± 18.6 † , ‡ , § | 116.8 ± 19.3 | 114.1 ± 19.7 | 112.1 ± 18.1 | 107.4 ± 16.0 | <0.001 | |
| Diastolic blood pressure (mmHg) | 75.1 ± 11.1 † , ‡ | 73.6 ± 12.0 | 72.0 ± 12.3 | 70.7 ± 12.6 | 68.7 ± 12.6 | <0.001 | |
| HbA1c (%) | 5.9 ± 1.0 ∗ , † , § | 5.8 ± 0.9 | 5.7 ± 0.9 | 5.7 ± 1.0 § | 5.5 ± 0.6 | <0.001 | |
| Triglyceride (mg/dl) ¶ | 150.9 ± 80.1 † , § | 144.0 ± 91.9 | 132.5 ± 69.1 | 135.4 ± 93.1 | 120.1 ± 76.5 | 0.001 | |
| Fasting insulin (μIU/ml) | 9.2 ± 6.3 ∗ , † , ‡ , § | 8.0 ± 4.5 ‡ | 7.5 ± 5.6 ‡ | 6.3 ± 3.4 | 6.3 ± 3.8 | <0.001 | |
| Homeostasis model assessment of insulin resistance ‖ | 2.0 ± 1.5 ∗ , † , ‡ , § | 1.7 ± 1.1 ‡ | 1.7 ± 2.3 | 1.4 ± 0.9 | 1.3 ± 0.8 | <0.001 | |
| Aspartate aminotransferase (IU/L) | 24.2 ± 7.8 † , ‡ , § | 24.7 ± 9.0 † , ‡ , § | 26.3 ± 11.2 ‡ , § | 29.2 ± 18.0 § | 34.2 ± 19.0 | <0.001 | |
| Alanine aminotransferase (IU/L) | 20.8 ± 9.8 | 21.6 ± 13.1 | 22.1 ± 14.0 | 22.7 ± 14.5 | 22.4 ± 13.0 | 0.038 | |
| Gamma-glutamyl transferase (IU/L) ¶ | 18.8 ± 16.8 | 19.8 ± 22.8 | 20.4 ± 21.7 | 21.7 ± 37.4 | 17.8 ± 10.1 | 0.091 | |
| High-sensitive C-reactive protein (mg/dl) ¶ | 0.32 ± 1.04 ∗ | 0.20 ± 0.35 | 0.19 ± 0.25 | 0.21 ± 0.44 | 0.19 ± 0.32 | 0.009 | |
| Smoking | Non-smoker | 763 (93.4%) | 2543 (94.7%) | 714 (97.1%) | 184 (97.4%) | 59 (96.7%) | 0.034 |
| Ex-smoker | 11 (1.3%) | 38 (1.4%) | 7 (1.0%) | 1 (0.5%) | 1 (1.6%) | ||
| Smoker | 43 (5.3%) | 104 (3.9%) | 14 (1.9%) | 4 (2.1%) | 1 (1.6%) | ||
| Diabetes mellitus | 132 (15.9%) | 337 (12.4%) | 92 (12.5%) | 23 (12.1%) | 3 (4.9%) | 0.029 | |
| Hypertension | 201 (24.0%) | 618 (22.7%) | 139 (18.8%) | 28 (14.7%) | 12 (19.7%) | 0.011 | |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
