Remnant Cholesterol is Prospectively Associated With Cardiovascular Disease Events and All-Cause Mortality in Kidney Transplant Recipients


Reuben William Horace; Mary Roberts; Theresa I. Shireman; Basma Merhi; Paul Jacques; Andrew G. Bostom; Simin Liu; Charles B. Eaton


Nephrol Dial Transplant. 2022;37(2):382-389. 

In This Article

Materials and Methods

The completed FAVORIT study was a unique multicenter, multiethnic controlled clinical trial[5,6] that provided definitive evidence that high-dose, folic acid–based lowering of plasma homocysteine—a putatively thromboatherogenic, amino acid by-product of methionine metabolism[5,6] did not reduce hard, centrally adjudicated CVD events (both fatal and nonfatal), graft failure or total mortality in chronic stable KTRs.[5] FAVORIT randomized 4110 stable KTRs from August 2002 through January 2007 to either a standard multivitamin with high doses of folic acid (5 mg), vitamin B6 (pyridoxine; 50 mg) and vitamin B12 (cyanocobalamin; 1 mg) versus a multivitamin containing low doses of vitamin B6 (1.4 mg) and vitamin B12 (2 μg) with no folic acid. Follow-up contacts occurred every 6 months through 31 January 2010 to obtain study-related outcomes through 24 June 2009. Both men and women ages 35–75 years who had a kidney transplant for at least 6 months were screened for eligibility at 30 transplantation centers located in the USA, Canada and Brazil. Entry criteria included stable kidney function and CVD event.

Analytic Cohort

Among 4110 enrolled participants, 200 were excluded for missing baseline remnant cholesterol levels and 98 with other missing covariates or outcomes (CVD), leaving 3812 participants described in the analysis. KTRs were included in the study population if they had clinically stable kidney function and valid remnant cholesterol data (Figure 1). Demographics data (age, sex and race), smoking status (yes or no), past medical history (diabetes mellitus), physical examination findings [body mass index (BMI), systolic and diastolic blood pressure (SBP and DBP)] and laboratory variables (creatinine and lipoproteins) were collected at study enrollment. Baseline hypertension was defined by SBP ≥140 mmHg, DBP ≥90 mmHg or antihypertensive medication use. Diabetes was defined by the use of insulin or oral hypoglycemic medications or patient history. Smoking status was defined as a current smoker. Race was defined as non-Hispanic white, non-Hispanic black or other. eGFR was estimated using the Chronic Kidney Disease Epidemiology Collaboration equation, a newer estimating equation with better performance at higher GFR levels.[8,10] BMI was calculated using the formula weight (kg)/height (m).[2] Serum remnant cholesterol was defined as total cholesterol (TC = HDL + LDL). Nonfasting plasma lipids were measured by standard enzymatic methods with the following intra- and interassay coefficients of variation: TC, intra-assay 1.6% and interassay 2.8%; HDL cholesterol, intra-assay 3.0% and interassay 4.0%; total triglycerides, intra-assay 2.0% and inter-assay 3.4%. Nonfasting plasma LDL cholesterol was estimated using the Friedewald equation at triglyceride levels <400 mg/dL. Consistent with the recent findings of Jorgensen et al.,[9] who evaluated nonfasting direct versus estimated LDL among individuals with triglyceride of 88–885 mg/dL (R 2 = 0.85, P < 0.001), comparison of LDL values within a mildly to moderately hypertriglyceridemic subsample of our population (i.e. n = 227 subjects with triglyceride levels of 400–<1000 mg/dL), direct versus Friedewald equation-estimated, yielded similar results (R 2 = 0.74, P < 0.001), notwithstanding the accepted performance limitations of these estimates in this range of triglycerides. All values of remnant cholesterol were log transformed and categorized into quartiles. Quartiles were based on log transformation prior to exclusion, therefore quartiles have different numbers of participants due to differential exclusions.

Figure 1.

Complete remnant cholesterol cohort. Eligible FAVORIT participants were the 3812 who were not missing baseline remnant cholesterol, baseline CVD, race and BMI.

The primary FAVORIT outcome was pooled incident or recurrent CVD that comprised CVD death, myocardial infarction, resuscitated sudden death, stroke, coronary artery revascularization, lower extremity revascularization or amputation above the ankle for severe arterial disease, carotid endarterectomy or angioplasty, abdominal aortic aneurysm repair or renal artery revascularization.[5–7] The first four components of this primary outcome were centrally reviewed and adjudicated by the FAVORIT Clinical Endpoints Committee; the remaining outcomes were identified through medical record abstraction. The Clinical Endpoints Committee also reviewed records for unstable angina cases and urgent coronary revascularization procedures in search of myocardial infarctions that were not identified by the clinical site staff. All-cause mortality and dialysis-dependent kidney graft failure were secondary outcomes.[5]

Jonckheere–Terpstra tests for ordered alternatives (for continuous variables) and a Wilcoxon-type test for trend (for categorical variables) were used to compare baseline characteristics across remnant cholesterol quartiles. Kaplan–Meier survival analysis was used to estimate the survival function among study participants by remnant cholesterol quartiles. Remnant cholesterol was log transformed and categorized into four groups using the respective median value for the analysis. Cox proportional hazards regression was used to evaluate the association of baseline remnant cholesterol levels with time to primary and secondary study outcomes. These models were adjusted for HMG CoA reductase inhibitor (statin) or other lipid-lowering drug use, body mass index, SBP, DBP, albuminuria, age, race, gender, diabetes, graft vintage, smoking status, mechanistic target of rapamycin (mTOR) inhibitors, donor type and treatment arm. Maintenance of proportional hazards assumptions was assessed for the primary and secondary outcome analyses by examination of log–log plotting of survival probability. Additionally, two sensitivity analyses were performed. The first examined the relationship of remnant cholesterol and CVD outcomes with HDL, LDL and triglycerides in the adjusted model. The second sensitivity analysis examined remnant cholesterol and all-cause mortality outcome with HDL, LDL and triglycerides in the extended models. Additionally, we performed a mediation analysis on graft loss against transplant failure outcomes. Analyses were performed using Stata version 15 software (StataCorp, College Station, TX, USA). Two-sided P-values <0.05 were considered statistically significant for all analyses.