BACKGROUND:

There is a very high mortality especially from cardiovascular disease among patients who are on dialysis. Atherosclerotic cardiovascular disease (ASCVD) is a leading contributor to the high morbidity and mortality among end-stage renal disease (ESRD) patients, accounting for 36 percent of ESRD deaths (total annual mortality of 23 percent). The high mortality among dialysis patients is a major public health problem and the ability to identify and treat risk factors that may reduce morbidity and mortality would be of substantial benefit. There are well-known risk factors for cardiovascular disease that certainly are related to morbidity and mortality among dialysis patients, i.e., relationship to hypertension, diabetes, smoking, etc. However, some of the new risk factors could also contribute to the excess mortality.

DESIGN NARRATIVE:

The prospective study tested the hypothesis that higher levels of several novel risk factors (Lp(a) levels and apo(a) isoforms; homocysteine and related vitamins; Chlamydia pneumoniae and cytomegalovirus; and C-reactive protein and fibrinogen) and traditional risk factors predicted higher risk of ASCVD in 925 incident dialysis patients recruited within three months of starting dialysis. Although these factors had been implicated in the etiology of ASCVD in ESRD patients, little prospective data existed. The cohort had already been recruited through a collaboration between Johns Hopkins and 80 Dialysis Clinics Incorporated (DCI) clinics; many of the important predictors and possible confounders had been measured. Long-term follow-up was obtained by extending mean followup of 2.4 years by four more years.

The investigators 1) extended specimen collection, and follow-up, and instituted standardized review of ASCVD events; 2) characterized baseline associations of novel and traditional factors with each other, dialysis modality and dose, nutritional status, and ASCVD prevalence in the full cohort using a cross-sectional design; 3) determined whether baseline levels of risk factors predicted subsequent incidence of ASCVD events, and total mortality using a prospective cohort study design and tested a priori hypothesized interactions between risk factors and the risk of ASCVD; 4) studied the variability of risk factors over time using annual measurements in a random subset of 180 patients (subcohort) using a longitudinal design; and lastly, 5) used a case-cohort design, utilizing the subcohort, to test whether the most recent level before an ASCVD event, the baseline level, or the mean level of each risk factor was most predictive of ASCVD risk. Baseline data collection included a patient health questionnaire and a standardized review of comorbidity using dialysis chart records. Serum, plasma and DNA were stored at -80 degrees C. from patient visits at recruitment (month 0), and followup (months 1,2,3,6,12,8,24, etc.). ASCVD was assessed by review of hospital charts, patients and care providers questionnaires, and HCFA death forms.