Association between LDL-C and risk of myocardial infarction in CKD

Abstract

LDL cholesterol (LDL-C) is an important marker of coronary risk in the general population, but its utility in people with CKD is unclear. We studied 836,060 adults from the Alberta Kidney Disease Network with at least one measurement of fasting LDL-C, estimated GFR (eGFR), and proteinuria between 2002 and 2009. All participants were free of stage 5 CKD at cohort entry. We followed participants from first eGFR measurement to March 31, 2009; we used validated algorithms applied to administrative data to ascertain primary outcome (hospitalization for myocardial infarction) and Cox regression to calculate adjusted hazard ratios (HRs) for myocardial infarction by LDL-C categories within eGFR strata. During median follow-up of 48 months, 7762 patients were hospitalized for myocardial infarction, with incidence highest among participants with the lowest eGFR. Compared with 2.6-3.39 mmol/L (referent), the risk associated with having LDL-C above 4.9 mmol/L seemed greatest for GFR≥90 ml/min per 1.73 m(2) and least for eGFR=15-59.9 ml/min per 1.73 m(2). Specifically, the adjusted HRs (95% confidence intervals) of myocardial infarction associated with LDL-C of ≥4.9 compared with 2.6-3.39 mmol/L in participants with eGFR=15-59.9, 60-89.9, and ≥90 ml/min per 1.73 m(2) were 2.06 (1.59, 2.67), 2.30 (2.00, 2.65), and 3.01 (2.46, 3.69). In conclusion, the association between higher LDL-C and risk of myocardial infarction is weaker for people with lower baseline eGFR, despite higher absolute risk of myocardial infarction. Increased LDL-C may be less useful as a marker of coronary risk among people with CKD than the general population.

Figure 1.

Prevalence of LDL-C categories by eGFR level. The figure shows that a higher proportion of participants with eGFR≥90 ml/min per 1.73 m² had LDL-C<2.6 mmol/L, as compared to those with lower levels of eGFR.

Figure 2.

Adjusted relation between LDL-C and HR of myocardial infarction by eGFR as a continuous variable. LDL-C was modeled by a restricted cubic spline with four knots at the 5th, 35th, 65th, and 95th percentiles, corresponding to LDL-C values of 1.8, 2.8, 3.4, and 4.6 using a Cox regression model. The LDL value of 2.6 mmol/L was used as the referent. Because eGFR was treated as a continuous variable, we selected eGFR values of 15, 30, 45, 60, and 90 to show the splines. Fully adjusted for age, sex, diabetes, hypertension, Aboriginal status, socioeconomic status, proteinuria categories, statin use, and Charlson comorbidities (cancer, cerebrovascular disease, congestive heart failure, chronic pulmonary disease, dementia, metastatic solid tumor, myocardial infarction, liver disease, hemiplegia/paraplegia, peptic ulcer disease, peripheral vascular disease, and rheumatic disease).