Reverse epidemiology of lipid-death associations in a cohort of end-stage renal disease patients.

BACKGROUND AND AIMS: Cardiovascular disease is the leading cause of death among end-stage renal disease (ESRD) patients with hypercholesterolemia as a major cause. A few studies have demonstrated counter-intuitive findings known as reverse epidemiology where normal levels of cholesterol are associated with higher levels of mortality. The purpose of this study was to determine if there are reverse epidemiological associations between lipid risk factors and mortality in ESRD patients.
METHODS: ESRD (n = 438) patients were recruited from 4 outpatient dialysis units. Patients were tracked for 36 months until study completion or death with mortality status as the outcome measure.
RESULTS: Analysis of covariance revealed significant differences at posttest and reverse epidemiological effects for total cholesterol (p = 0.0001), low-density lipoprotein cholesterol (LDL) (p = 0.023), LDL particle number (p = 0.0001), LDL size (p = 0.009), triglycerides (p = 0.0001), and very low-density lipoprotein cholesterol (p = 0.036). A step-wise linear regression revealed weak, but significant predictors of mortality with total cholesterol (β = 0.263, p = 0.017) and LDL (β = -0.177, p = 0.045). A Cox death hazard ratio revealed LDL size as a significant predictor of mortality in this study.
CONCLUSIONS: Our study discovered reverse epidemiology in a number of lipid variables. Additionally regression revealed that LDL and total cholesterol were predictors of mortality with lower levels being more predictive of death.