High-density lipoprotein cholesterol and causes of death in chronic kidney disease

Abstract

Recent data suggest a U-shaped association between high-density lipoprotein cholesterol (HDL-c) and death in chronic kidney disease (CKD). However, whether the increased mortality in patients with extreme levels is explained by specific causes of death remains unclear. We studied the associations between HDL-c and cause-specific deaths in CKD. We included 38,377 patients with estimated glomerular filtration rate 15-59mL/min/1.73m2. We classified deaths into 3 major categories: (1) cardiovascular; (2) malignant; and (3) noncardiovascular/nonmalignant causes. We fitted Cox regression models for overall mortality and separate competing risk models for each major cause of death category to evaluate their respective associations with categories of HDL-c (≤30, 31-40, 41-50 [referent], 51-60, >60mg/dL). Separate analyses were conducted for men and women. During a median follow-up of 4.5years, 9665 patients died. After adjusting for relevant covariates, in both sexes, HDL-c 31 to 40mg/dL and ≤30mg/dL were associated with higher risk of all-cause mortality, cardiovascular mortality, malignancy-related deaths, and noncardiovascular/nonmalignancy-related deaths. HDL-c >60mg/dL was associated with lower all-cause (hazard ratio: 0.75, 95% confidence interval: 0.69, 0.81), cardiovascular, malignancy-related, and noncardiovascular/nonmalignancy-related deaths among women but not in men. Similar results were noted when HDL-c was examined as a continuous measure. In a non-dialysis-dependent CKD population, HDL-c ≤40mg/dL was associated with risk of higher all-cause, cardiovascular, malignant, and noncardiovascular/nonmalignant mortality in men and women. HDL >60mg/dL was associated with lower risk of all-cause, cardiovascular, malignant, and noncardiovascular/nonmalignant mortality in women but not in men.