Increased Serum Lipoprotein(a) Levels in Patients with Early Renal Failure

Methods

A total of 417 patients referred consecutively to the hypertension clinic of our department from January 1993 to October 1996 were included in a cross-sectional study. The patients seen at the clinic include persons with an established diagnosis of arterial hypertension, persons referred for evaluation of blood pressure-related problems (family history of hypertension and other cardiovascular diseases, occasional finding of high or low blood pressure, or ambulatory blood pressure monitoring), and normotensive persons who wish to have their blood pressure measured and receive counseling. Exclusion criteria were age younger than 30 years or older than 75 years, body mass index greater than 30 kg/m², diabetes mellitus or other endocrine disease, severe hypertension [15], creatinine clearance less than 30 mL/min per 1.73 m² of body surface area, urinary protein excretion greater than 1.0 g/d, and diseases or treatments that might interfere with serum lipids. Arterial hypertension was diagnosed according to World Health Organization guidelines [15]. All patients had measurements of 24-hour creatinine clearance, urinary protein excretion, and microalbuminuria [16]. Impairment of renal function was considered to be present when the 24-hour creatinine clearance was less than 90 mL/min per 1.73 m². The cause of renal failure was established by history, urinalysis, and review of medical records.

Cardiovascular status was assessed without previous knowledge of patients' creatinine clearance and levels of lipoproteins by history, physical examination, electrocardiography, or additional diagnostic procedures when appropriate [17]. The diagnoses of myocardial infarction, transient ischemic attack, prolonged reversible neurologic deficit, and atherothrombotic stroke were confirmed by retrospective review of medical records.

Therapy with antihypertensive drugs was withdrawn 1 week before measurement of blood variables and renal function. At the time of the study, patients were allowed to maintain their usual diet. Blood samples were obtained after fasting for 12 to 14 hours for analysis of serum albumin; triglycerides; total and high-density lipoprotein cholesterol; lipoprotein(a); apolipoproteins A-I, A-II, B, C-II, C-III, and E; and uric acid. Apolipoprotein(a) phenotypes were characterized in a subset of 226 patients.

Serum levels of lipids and lipoproteins were measured by using standard methods, and apolipoproteins were measured by using an immunoturbidimetric method (Eiken, Tokyo, Japan) [17]. Lipoprotein(a) was measured on samples that had been frozen at −70°C and stored for an average of 3 weeks. The lipoprotein(a) level was determined with a double-antibody enzyme-linked immunosorbent assay (Strategic Diagnostic, Newark, New Jersey), and apolipoprotein(a) phenotyping was performed by sodium dodecyl sulfate-agarose gel electrophoresis, as described elsewhere [17]. All measurements were done in duplicate.

All values are expressed as the mean ±SD unless otherwise noted. Because the distributions of triglycerides and lipoprotein(a) values were positively skewed, a log transformation was used to satisfy normality assumptions. The Student t-test for unpaired data and the Mann-Whitney test were used for comparisons between two groups. Patients with normal and decreased creatinine clearance were compared by using stepwise multiple-discriminant analysis. The chi-square goodness-of-fit test was used to compare frequency distributions. The relation between log lipoprotein(a) and other variables was examined by linear regression analysis, and the correlation was expressed by the Pearson correlation coefficient.

Because of the high number of detectable apolipoprotein(a) isoforms, many phenotypes were represented in low numbers. To obtain sufficient sample sizes for comparison, we decided a priori to combine the apolipoprotein(a) phenotypes in two subgroups according to the molecular weight of the smallest apolipoprotein(a) isoforms, as in other studies [14, 17]. The low-molecular-weight group included all patients with at least one apolipoprotein(a) isoform with 11 to 22 kringle IV repeats; the high-molecular-weight group included all patients who had only isoforms with more than 22 kringle IV repeats. A P value less than 0.05 was considered statistically significant.

Results

Creatinine clearance less than 90 mL/min per 1.73 m² was found in 160 of 417 patients. Renal failure was caused by arteriolar nephrosclerosis in 149 patients and by chronic glomerulonephritis in 11. As shown in Table 1, age, blood pressure levels, and prevalence of hypertension were significantly greater in patients with decreased creatinine clearance than in patients with normal creatinine clearance. Sex distribution, anthropometric indices, alcohol consumption, and smoking status at the time of the study were similar in the two groups. No meaningful differences in serum levels of triglycerides; total, low-density lipoprotein, and high-density lipoprotein cholesterol; and apolipoproteins were found between patients with normal and those with impaired renal function. Serum lipoprotein(a) levels were significantly greater in patients with creatinine clearance less than 90 mL/min per 1.73 m² than in those with creatinine clearance of 90 mL/min per 1.73 m² or more. The log serum lipoprotein(a) level was inversely correlated with creatinine clearance (r = 0.243; P < 0.001) (Figure 1) and was directly correlated with apolipoprotein B levels (r = 0.104; P = 0.027). Lipoprotein(a) levels were not affected by age and sex and were not correlated with blood pressure, serum albumin levels, and serum lipid levels. Discriminant analysis, which included systolic blood pressure, age, and log lipoprotein(a), indicated that these three variables were independently associated with early renal failure.

View larger version:

• In this window
• In a new window

Figure 1. A significant inverse correlation ( = 0.243; < 0.001) was seen. Relation between creatinine clearance and log lipoprotein(a).rP

We found clinical and laboratory evidence of one or more events attributed to atherosclerosis in 8.6% of patients with normal renal function and 24.4% of patients with early renal failure (P < 0.001). The prevalence of coronary artery disease (18.8% and 6.6%; P < 0.001), cerebrovascular disease (10.0% and 3.9%; P = 0.033), and peripheral vascular disease (7.5% and 2.3%; P = 0.031) was significantly higher in patients with decreased creatinine clearance than in patients with normal creatinine clearance. Serum lipoprotein(a) levels were significantly greater in patients with evidence of atherosclerotic disease (29.3 ± 18.2 mg/dL; n = 61) than in patients with no evidence of atherosclerotic disease (15.7 ± 21.1 mg/dL; n = 356) (P < 0.001).

Apolipoprotein(a) phenotyping demonstrated that the frequency of low-molecular-weight apolipoprotein(a) isoforms was similar in patients with normal (29.0%) and impaired (29.4%) renal function. This suggests that elevated lipoprotein(a) levels in patients with early renal failure have no genetic basis.

Discussion

Our results show that serum lipoprotein(a) levels are increased in patients with early renal failure and that this increase is directly related to the severity of renal functional impairment. Elevated lipoprotein(a) levels and early impairment of renal function were associated with an increased prevalence of atherosclerotic disease. That the frequency distribution of apolipoprotein(a) isoforms is similar to that of normal patients suggests that elevated lipoprotein(a) levels in early renal failure have no genetic basis.

In recent years, many studies have investigated the relation between lipoprotein(a) and advanced renal disease [12] and have yielded conflicting results. Inconsistent findings were probably related to lack of statistical power due to small numbers of patients. The wide range of serum lipoprotein(a) levels and the skewed distribution of these levels require cautious statistical evaluation, and a minimum of 100 patients per group is required for the detection of a 30% difference at a probability of less than 5% [12]. All studies including more than 100 patients and controls have consistently reported an increase in serum lipoprotein(a) levels in patients with end-stage renal disease [13, 14, 18]. Furthermore, the few studies that have had an adequate number of patients with end-stage renal disease and have evaluated apolipoprotein(a) phenotypes have shown that increased serum lipoprotein(a) levels are not the result of a shift in the spectrum of size polymorphism of apolipoprotein(a) [13, 14]. This suggests that elevation lipoprotein(a) levels are secondary to renal failure rather than genetic causes.

Although many studies have examined lipoprotein(a) levels in patients with end-stage renal disease, limited data are available on lipoprotein(a) and apolipoprotein(a) isoforms in patients with early renal failure. Haffner and colleagues [19] studied 66 patients with renal failure not requiring dialysis and creatinine clearance values from 20 to 52 mL/min and found that serum lipoprotein(a) levels were greater in these patients than in controls. In this study, no correlation was seen between lipoprotein(a) levels and creatinine clearance. In addition to the limitation imposed by the small sample size, the patients included in this study were highly unselected; patients of different ethnicities and those with diabetes and severe obesity were included. Moreover, the selection criterion used to define the presence of renal failure (plasma creatinine concentration > 1.2 mg/dL) cannot exclude the possibility that a small number of patients with renal disease was included in the control group. Arnadottir and coworkers [20] found an inverse correlation between serum lipoprotein(a) levels and glomerular filtration rate in 72 patients with renal failure not requiring dialysis. In this study, as in the study by Haffner and colleagues, apolipoprotein(a) isoforms were not evaluated. Our study provides clear evidence that serum lipoprotein(a) levels are increased in early stages of renal failure. Furthermore, the inverse correlation between log lipoprotein(a) and creatinine clearance suggests that the kidney may be involved in the metabolism of lipoprotein(a), although it is also possible that the biochemical abnormalities associated with renal failure inhibit lipoprotein(a) catabolism at extrarenal sites.

Some limitations of our study should be stressed. First, the use of a clinic sample may limit the generalizability of the conclusions to the general population because of bias in the referral of patients to the source of care. Second, because the underlying renal disease in most patients with decreased creatinine clearance was arteriolar nephrosclerosis, results should be extrapolated to other types of renal disease with caution. Third, cardiovascular end points were determined retrospectively and in an ad hoc manner; this may have introduced ascertainment bias. Finally, the apparent increase in cardiovascular risk in patients with early renal failure could be ascribed to the greater prevalence of arterial hypertension in these patients.

In conclusion, serum lipoprotein(a) levels are increased in patients with early renal failure. This increase does not seem to have a genetic basis; rather, it is secondary to renal function impairment, suggesting that the kidney has a role in lipoprotein(a) catabolism. Lipoprotein(a) may represent an important determinant of cardiovascular damage at all levels of renal function, but this idea needs further evaluation in longitudinal studies.