Lipoprotein(a) and prevalent cardiovascular disease in a dialysis population: The Choices for Healthy Outcomes in Caring for ESRD (CHOICE) study

Abstract

Background: Levels of lipoprotein(a) [Lp(a)], an atherogenic lipoprotein, are elevated in patients with end-stage renal disease and inversely related to the size of apolipoprotein(a) [apo(a)], a glycoprotein bound to Lp(a). We studied the association of Lp(a) level and apo(a) size with prevalent atherosclerotic cardiovascular disease (ASCVD) in 871 incident dialysis patients (261 blacks, 565 whites, 45 other).

Methods: Lp(a) was measured by an apo(a) size-independent enzyme-linked immunoassay; and apo(a) size was measured by sodium dodecyl sulfate-agarose gel electrophoresis. Prevalent ASCVD, derived from medical records, was defined as coronary heart disease or cerebral or peripheral vascular disease. Adjustment variables included age, sex, race, dialysis modality, diabetes, serum creatinine level, albumin level, and low-density lipoprotein cholesterol level.

Results: ASCVD prevalence was 58%. Median Lp(a) levels for those with compared with those without ASCVD were 38 versus 35 nmol/L for whites (P = 0.49) and 100 versus 74 nmol/L for blacks, respectively (P = 0.35). Lp(a) level was associated with ASCVD among those younger than 60 years (odds ratio [OR] for +1 interquartile range of Lp(a), 1.5; P = 0.02), but not among those 60 years and older (OR, 1.0; P = 0.82; P(interaction) by age, 0.08). ORs were 1.3 for all whites (P = 0.03) and 1.1 for all blacks (P = 0.87; P(interaction)by race = 0.53). ORs of ASCVD were 1.7 for whites younger than 60 years (P = 0.01) and 1.2 for blacks younger than 60 years (P = 0.77; P(interaction) by race = 0.42). No association between apo(a) isoform size and ASCVD was present.

Conclusion: In an incident dialysis cohort, Lp(a) level was associated with prevalent ASCVD among whites younger than 60 years, but not among blacks or those older than 60 years. Apo(a) isoform size was not associated with prevalent ASCVD. These data suggest that baseline ASCVD is unlikely to strongly confound the potential associations of Lp(a) level and prospectively ascertained ASCVD among incident dialysis patients.