rss
J Epidemiol Community Health 66:A49 doi:10.1136/jech-2012-201753.126
  • Friday 14 September 2012, Parallel Session D
  • Poster Programme

PS27 Serum Bilirubin and Risk of Cardiovascular Events and Death in a Statin-Treated Population: A Cohort Study

  1. I Nazareth
  1. Primary Care and Population Health, UCL, London, UK

Abstract

Background Serum bilirubin is an endogenous antioxidant with a strong genetic component that may be a marker for future cardiovascular disease (CVD) risk. We examined the relationship between serum bilirubin levels recorded prior to statin prescription, primarily as test for liver function, and the diagnosis or death from CVD. We also examined whether bilirubin levels increased following statin prescription independently of liver enzymes, which has been shown in animal models and could contribute to the purported health benefits of statins in addition to cholesterol reduction.

Methods All patients with liver function tests three months prior to first statin treatment between January 1st 2000 and December 31st 2010 and no history of liver disease or CVD were extracted from The Health Improvement Network (THIN) primary care database. Restricted cubic spline Poisson regressions were fitted on bilirubin levels and adjusted for traditional cardiovascular risk factors to estimate incidence rate ratios.

Results In total 130,052 patients met the inclusion criteria and after a median follow-up of 43 months, there were 5,938 coronary heart disease (CHD) events, 2,438 stroke events, and 5,185 deaths from any cause. In men, the incidence of CHD in the lowest decile category of bilirubin (1–6 µmol/L) was 176 per 10,000 person years (PYs) compared with 139 per 10,000 PYs in the highest decile (19–40 µmol/L). The result for stroke was 72 versus 52 per 10,000 PYs and for death 139 versus 92 per 10,000 PYs. Similar differences were seen for women. The adjusted associations with bilirubin were L-shaped with a negative relationship up to around 10–15 µmol/L. The models predicted that, compared to patients with the median bilirubin level (10 µmol/L), those with a similar CVD risk profile but a bilirubin level of 5 µmol/L had a 19% (95% CI: 11–27%) higher rates of CHD, a 22% (95% CI: 10–36%) higher rates of stroke, and a 26% (95% CI: 17–35%) higher rate of death. A dose-dependent increase in mean bilirubin level was seen following atorvastatin prescription but not for simvastatin.

Conclusion Low bilirubin prior to statin prescription is an independent risk factor for CVD and death. Further work is needed to examine whether pleotrpic effects of statins can be explained by alterations in bilirubin production and/or elimination.