Lipoprotein(a) and Apolipoprotein(a) Isoforms
http://www.100md.com
循环学杂志 2005年第3期
the Donald W. Reynolds Center for Clinical Cardiovascular Research (R.G., Z.Y., R.P., J.C.C., H.H.H.), Dallas, Tex
Department of Statistics (R.G., Z.Y.), Rice University, Houston, Tex
Department of Medicine (S.M.), University of Washington, Seattle, Wash
Howard Hughes Medical Institute (H.H.H.) at the University of Texas Southwestern Medical Center, Dallas, Tex.
Abstract
Background— Elevated plasma levels of lipoprotein(a) [Lp(a)] are an independent risk factor for cardiovascular disease in whites. Blacks have 2- to 3-fold higher plasma levels of Lp(a) than whites and yet do not have a correspondingly higher rate of coronary events. It remains unclear whether elevated plasma levels of Lp(a) are an independent risk factor for coronary atherosclerosis in individuals of African descent.
Methods and Results— The relationship between plasma levels of Lp(a), apolipoprotein(a) isoform sizes, and the presence of coronary calcium was examined in 761 blacks and 527 whites (men aged >40 years, women aged >45 years) from a population-based sample. No relationship was found between plasma levels of Lp(a), apolipoprotein(a) isoform size, or a combination of these 2 variables and coronary artery calcium (CAC) in whites or blacks. No correlation was observed between plasma levels of Lp(a) and coronary calcium scores in any group, although all black men with very high plasma levels of Lp(a) (>300 μmol/L; n=7) were CAC-positive. Whites with high plasma levels of Lp(a) plus elevated plasma levels of LDL cholesterol (men) or reduced levels of HDL cholesterol (men and women) or who smoked (women) had a higher prevalence of CAC. In contrast, no joint effects between plasma levels of Lp(a) and other cardiovascular risk factors on coronary calcium were found in blacks.
Conclusions— No consistent independent relationship between plasma levels of Lp(a) or apolipoprotein(a) isoform size and coronary calcium was found in whites or blacks.
Key Words: lipoproteins ; apolipoproteins ; arteriosclerosis ; coronary disease ; risk factors
Introduction
Lipoprotein(a) [Lp(a)] is a cholesteryl ester–rich lipoprotein distinguished from other lipoproteins by the presence of a unique apolipoprotein (apo), apo(a). Plasma concentrations of Lp(a) vary over a 1000-fold range, and the distribution of levels is highly skewed, with most whites and Asians having low levels.1 Until recently, controversy persisted about the status of elevated plasma levels of Lp(a) as an independent risk factor for the development of coronary atherosclerosis2,3; however, a meta-analysis of 27 prospective studies, which included 5436 coronary heart disease cases, demonstrated a consistent, positive association between high plasma levels of Lp(a) and coronary atherosclerotic events in whites.4 The distribution of plasma levels of Lp(a) in blacks is less skewed, and the median level is 2 to 4 times greater than in whites.5 Paradoxically, the higher levels of Lp(a) in blacks are not associated with a corresponding increase in the prevalence of coronary atherosclerosis.6–8
A possible explanation for the apparent dissociation between plasma Lp(a) levels and coronary atherosclerotic risk in blacks is that only a subset of individuals with high plasma levels of Lp(a), those with apo(a) isoforms of small size, are at increased risk.9–11 The apo(a) gene (APOA) is highly polymorphic in size owing to allelic differences in the number of a kringle-encoding sequence resembling kringle 4 of plasminogen.12,13 In vitro studies suggest that apo(a) interferes with plasminogen, a central zymogen in thrombosis, and that smaller apo(a) isoforms are more effective at inhibiting thrombolysis.14,15 In some studies, whites with apo(a) isoforms of small size have an increased incidence of coronary atherosclerosis.16–18 Because plasma levels of Lp(a) are inversely related to the size of the apo(a) gene and isoforms in whites,19 most whites with high plasma levels of Lp(a) have a small isoform. Therefore, it has been difficult to distinguish the relative atherogenic effects of apo(a) isoform size and plasma levels of Lp(a) in whites. In contrast, many blacks with high plasma levels of Lp(a) have apo(a) isoforms of intermediate size.20 Thus, examination of the relationship between plasma levels of Lp(a), apo(a) isoforms, and coronary atherosclerosis in blacks provides an opportunity to assess the relative contributions of isoform size and plasma levels of Lp(a) to coronary atherosclerotic risk.
In the present study, we examined the relationship between plasma levels of Lp(a), apo(a) isoform size, and coronary atherosclerotic burden in 1288 blacks and whites from the Dallas Heart Study (DHS)21 using electron-beam computed tomography (EBCT), a sensitive and noninvasive method to assess coronary atherosclerotic burden.22,23 Because Lp(a) has been demonstrated in vitro to adhere to multiple components of the extracellular matrix in the arterial wall, it has been proposed that the accumulation of Lp(a) within the lesion wall promotes accumulation of LDL.24 As the atherosclerotic lesion develops, calcium is deposited in the arterial wall.23 If the major effect of Lp(a) were to promote atherosclerosis, it would be expected that individuals with high plasma levels of Lp(a) would have a greater prevalence and/or amount of coronary calcium.
Methods
Study Population
The study subjects were obtained from the DHS, a multiethnic population-based probability sample of Dallas County, Texas,21 in which ethnicity was self-reported. The Institutional Review Board of the University of Texas Southwestern Medical Center approved the study, and all participants provided informed consent in accordance with institutional guidelines. Fasting blood samples were obtained from 3398 subjects (1760 blacks, 1068 whites, and 570 Hispanics). Blood was maintained at 4°C until the plasma was separated, placed in aliquots, and stored at –80°C. Plasma levels of Lp(a) and apo(a) isoform sizes were determined with well-validated assays.25
Of the participants who underwent phlebotomy, 2971 (50% black) completed a clinic visit at which coronary artery calcification (CAC) was evaluated by EBCT. A total of 970 blacks and 593 whites (men aged >40 years and women aged >45 years) underwent duplicate EBCT scans. A complete data set of the variables for this analysis in men aged >40 years and women aged >45 years included 1288 subjects (380 black women, 241 white women, 381 black men, and 286 white men).26
Risk Factor Profiling of Subjects
The assessment of cardiovascular risk factor status in the DHS is as described previously.21 The mean of the middle 3 (of 5) blood pressure measurements from the clinic visit was used in the present study. Subjects with blood pressure >140/90 mm Hg or taking antihypertensive medications were classified as hypertensive. The plasma lipids and lipoprotein-cholesterol (C) levels were measured by enzymatic techniques.27 Plasma Lp(a) was measured by a sandwich ELISA that is insensitive to apo(a) isoform size.25 Apo(a) isoform sizes were determined by immunoblot analysis with an apo(a)-specific antibody.28 The primary smoking variable (yes/no) was defined as current smoking (within 30 days of the interview) and a lifetime history of having smoked at least 100 cigarettes.
EBCT Scan for Coronary Calcium
Two consecutive EBCT scans were performed on each study participant as described previously.26 A focus was defined as a region of 3 or more contiguous voxels with a computed tomography number >130 Hounsfield units (HU). The voxel size was 0.586x0.586x3 mm (field of view 30 cm, matrix 512, 3-mm slice), so that 3 voxels represent a volume of 3.08 mm.29 Scans were read in a blinded fashion by one author (RP), and the results were expressed in Agatston units.29,30 The mean of the 2 scores from each participant was used as the final CAC score.
EBCT scores were highly skewed,26 and standard transformations failed to achieve approximate normality, largely because of the high frequency of subjects with scores at or near zero. For the primary analysis, EBCT scores were classified as CAC positive (CAC+) when the mean EBCT score was >10 Agatston units or CAC negative (CAC–) when the mean EBCT score was 10 Agatston units. A threshold of 10 Agatston units was selected because >95% of subjects were concordant (above or below 10 Agatston units) for repeated measurements at this threshold.26
Statistical Analysis
No mathematical transformation of the EBCT data achieved an adequate approximation of normality, and thus, standard modeling approaches could not be used to analyze quantitative EBCT measurements. Therefore, individuals were classified as either CAC+ (mean EBCT score >10 Agatston units) or CAC– (mean EBCT score 10 Agatston units), and logistic regression was used to assess the association between CAC prevalence and cardiovascular risk factors. To consider the possible loss of information associated with the use of a binary scale for EBCT scores, we also used ordinal regression analysis and classified the amount of CAC based on the atherosclerotic burden using the Rumberger scale: 0 to 10 (minimal), 10 to 100 (mild), 100 to 400 (moderate), and >400 (extensive) Agatston units.31
Continuous variables were summarized with means or medians and SDs. Two-sample t tests and ANOVA F tests were used to assess differences in means. To compare 2 skewed distributions [eg, the Lp(a) levels in black men and in white men], we assessed differences in medians. For large sample sizes, the sample median is normally distributed,32 and therefore, the difference in sample medians is normally distributed. The reported probability values for median differences are thus derived from the normal based Wald test: Z=(W–0)/SE(W), where W is a test statistic, SE(W) is the standard error for W, and 0 is the hypothesized value for W. In this case, W is the difference in sample medians and 0 is zero. The analytic expression for SE(W) is based on the asymptotic variance formula32 for a sample median. 2 tests were used for contingency table analysis of categorical variables. Lp(a) levels were summarized on their original scale and log-transformed before logistic and ordinal regression analysis because of the skewness of the distribution.
The joint effect of Lp(a) with other cardiovascular risk factors on the presence of CAC can be analyzed with logistic and ordinal regression models that include both main and interaction effects; however, because of the large sample size in the present study, we were able to conduct a more straightforward analysis within each race-gender subgroup without imposing model assumptions. To assess the joint effect of plasma levels of Lp(a) and a binary risk factor on CAC status, we compared the odds of being CAC+ in subjects with the risk factor and high Lp(a) to the odds of being CAC+ in subjects without the risk factor and low Lp(a). To analyze the joint effect of Lp(a) and continuous risk factors, we defined the top 30% tail of the risk factor as a positive risk factor effect and the bottom 30% as a negative risk factor effect and proceeded as above. The ORs for the risk factors were age-adjusted with logistic regression as follows. Let P denote the probability of disease. With 2 predictor variables, x1 and x2, the logistic model associates a person’s log-odds of disease with the 2 predictors: log[P/(1–P)]=;0+;1x1+;2x2. To obtain age-adjusted ORs, we defined P as the probability of being CAC+, x1 as age, and x2 as an indicator variable for risk group membership (high, low) defined by a given risk factor. The estimated value of ;1 represents the change in log-odds of CAC+ associated with a 1-year increase in age, adjusted for x2. According to the model, this change is assumed to be constant for persons of all ages. Of particular interest is the estimated value of ;2, which is interpreted as the age-adjusted log-OR for x2. Therefore, if x2 represents an indicator (yes/no) for hypertension, then ;2 is the log-OR of CAC+ for a person with hypertension relative to a person without hypertension. The age-adjusted ORs were used to analyze the joint effects of Lp(a) and other cardiovascular factors on CAC status.
The DHS is a probability-based survey with oversampling for blacks.21 Sample weights were used to extrapolate statistical inferences to the population, as described previously.21 Statistical analyses were performed with and without sample weights. Because the 2 sets of results were almost always in agreement, we reported unweighted data and results. Analyses were completed with the statistical packages S-Plus (Insightful) and Stata (Stata Corporation).
Results
Plasma levels of Lp(a) were significantly higher in blacks than in whites (Table 1). Levels of plasma Lp(a) were significantly higher (P<0.0001) in black women than in black men, but no gender difference was apparent in whites. Distributions of plasma Lp(a) by ethnicity and gender were similar to those of prior studies19 and did not differ between subjects who did and did not obtain an EBCT scan (data not shown).
To determine whether plasma levels of Lp(a) were associated with the presence of coronary calcium, the prevalence of positive CAC scores was determined for individuals within each Lp(a) quintile (Figure 1). No significant differences were found in the proportion of individuals who were CAC+ among the quintiles. Similarly, median plasma levels of Lp(a) were not significantly different in black men, black women, white men, or white women who were CAC+ and CAC– (Table 2), nor was any significant association between log-transformed plasma levels of Lp(a) and CAC status found in either logistic or ordinal regressions (data not shown). The relationship between plasma levels of Lp(a) and coronary calcium after the amount of CAC was classified for each individual with the Rumberger scale31 was also examined; no significant relationship was found between the 2 variables for any of the ethnic and gender-specific groups (data not shown). Accounting for age and smoking did not change the negative results of the logistic and ordinal regressions.
Mean plasma levels of Lp(a) were also not significantly different in the CAC+ and CAC– subjects, with 1 exception. Black men who were CAC+ had a significantly higher mean plasma level of Lp(a) (95 nmol/L) than black men without CAC (79 nmol/L; P=0.03; Table 2). To probe why the mean (but not median) plasma level of Lp(a) was significantly higher in the CAC+ black men, we examined and compared the distribution of plasma Lp(a) levels in CAC+ and CAC– black men (Figure 2). Although no overall correlation between plasma Lp(a) levels and EBCT score was found in black men (r=–0.018), inspection of the distributions revealed an excess of black men with extremely high plasma levels of Lp(a) (>300 nmol/L) in the CAC+ group (n=7) compared with the CAC– group (n=0). No significant difference in mean plasma level of Lp(a) of the 2 groups was found after elimination of these 7 individuals. Examination of the clinical characteristics of the 7 individuals with Lp(a) levels >300 nmol/L revealed no enrichment for other cardiovascular risk factors (high blood pressure, low plasma HDL-C, diabetes, smoking, or elevated levels of LDL-C or C-reactive protein) or for other factors that increase both Lp(a) and atherosclerotic risk (eg, renal insufficiency). As expected given the inverse relationship between apo(a) size and plasma levels of Lp(a), the apo(a) isoform sizes were smaller than the mean in all 7 black men (smallest isoform in each subject ranged in size from 14 to 21). Only a singe white man had a plasma level of Lp(a)>300 nmol/L, and the smallest apo(a) isoform in that individual was 17; this subject has no detectable CAC. These findings suggests that black men with extremely high levels of Lp(a) may be at increased risk of atherosclerosis, but the small number of individuals in this category precludes meaningful statistical analysis. Moreover, 9 of the 13 black women with plasma Lp(a) levels over 300 nmol/L [who all had at least 1 apo(a) isoform <19] were CAC– (data not shown).
The size of the major apo(a) isoform was inversely related to plasma Lp(a) levels (Figure 3), which were uniformly higher in blacks than in whites across the entire spectrum of isoform sizes. The slope of the regression of log-Lp(a) on apo(a) isoform size was similar in all groups. To assess the relationship between apo(a) isoform size and CAC+ status, the sizes of the major apo(a) isoforms were divided into quartiles, and the prevalence of positive CAC scores was compared in the different groups (Figure 4). No significant difference (2 P>0.05 in all subgroups) in the prevalence of CAC+ scores was apparent across the apo(a) isoform quartiles. No association was found between apo(a) isoform size and EBCT scores when analyzed as a discretized trait with either logistic or ordinal regressions (data not shown).
It has been suggested that only individuals with elevated levels of Lp(a) and small apo(a) isoforms (K4 repeats <22) are at increased risk of cardiovascular disease.9,10,18,20 To test for this possibility, we examined the relationship between the size of the major apo(a) isoform and plasma levels of Lp(a) in black and white subjects classified by their CAC status (Figure 5). A joint effect would be evident if a predominance of CAC+ individuals had smaller apo(a) isoforms and higher plasma levels of Lp(a). No such relationship was observed in any of the 4 ethnic-gender groups, which indicates the absence of a joint effect. The slopes of the regression line of log-Lp(a) on apo(a) isoform size in each CAC+ and CAC– subgroup were not significantly different (P>0.05).
Finally, we investigated the possibility that high plasma levels of Lp(a) are atherogenic in the presence of other risk factors. CAC+ scores were adjusted for age in these analyses, because age was significantly associated with CAC+ status in all groups (P<0.01; Table 2). The odds of being CAC+ in subjects with high plasma levels of Lp(a) and high risk factors were compared with the odds in subjects with low plasma levels of Lp(a) and low risk factors. The OR of high LDL to low LDL was 1.5 (P=0.24) and increased to 4.2 (P=0.03) when high-LDL/high-Lp(a) was compared with low-LDL/low-Lp(a) groups in white men (Table 3). White men also showed a strong joint effect between high plasma levels of Lp(a) and a low plasma level of HDL-C (OR=8.3, P=0.01) and high levels of plasma triglyceride (OR=5.2, P=0.03); a similar joint effect between high plasma levels of Lp(a) and low plasma levels of HDL-C was seen in white women (OR increasing from 2.5 to 4.6). No interactive effects were seen between high plasma levels of Lp(a) and high plasma levels of homocysteine in either white women or men, in contrast to what was observed previously.33 Of the discrete variables examined (diabetes, hypertension, and smoking), only smoking showed a joint effect with plasma levels of Lp(a) in white women. To assess the validity of the significant joint effects, we evaluated possible confounding factors between the high-Lp(a)/high-risk-factor subjects and the low-Lp(a)/low-risk-factor subjects; no evidence for confounding was found for diabetes, smoking, hypertension, or body mass index. No joint effects between plasma levels of Lp(a) and any of the continuous variables (plasma levels of LDL-C, HDL-C, triglycerides, body mass index, homocysteine) or the discrete variables (diabetes, hypertension and smoking) were found in black men or women (data not shown).
Discussion
The major finding of this study was that the presence of CAC is not significantly related to plasma levels of Lp(a), apo(a) isoform size, or the combination of apo(a) isoform size and plasma level of Lp(a) in whites or in black women. These data constitute the most representative sample of blacks in which this relationship has been studied. Extremely high plasma levels of Lp(a) (>300 nmol/L), but not apo(a) isoform size, were associated with a higher prevalence of coronary calcium in black men. Further analysis revealed that elevated plasma levels of Lp(a) significantly increased the odds of having detectable coronary calcium in whites with selected cardiovascular risk factors, including high plasma levels of LDL-C, elevated plasma levels of triglycerides, or low plasma levels of HDL-C in men. In white women, elevated plasma levels of Lp(a) acted synergistically with a smoking history and with low plasma levels of HDL-C to increase the odds of being CAC+. No joint effects between elevated levels of Lp(a) and other cardiovascular risk factors on coronary calcium were found in black men or black women.
Because the median level of Lp(a) in blacks is 2- to 3-fold higher than in whites, and the prevalence of established cardiovascular risk factors is equivalent or greater in blacks than in whites (Table 1), the burden of coronary atherosclerosis would be expected to be greater in the black population. Yet, most studies find the rate of preclinical atherosclerosis and ischemic events is similar in blacks and whites,8,26 although black women may have increased odds of disease compared with white women.34 A major limitation of most studies examining the relationship between plasma levels of Lp(a) and coronary atherosclerosis in blacks is that they have been small and not representative.10,35–37 The largest study assessing the effect of plasma levels of Lp(a) on cardiovascular events in blacks was the Atherosclerosis Risk in Communities Study,38 in which only a marginally significant relationship was found. The DHS is the most representative study to date that has examined the relationship between plasma levels of Lp(a) and coronary atherosclerotic burden. The present results support previous findings that plasma levels of Lp(a) are not associated with the presence of coronary atherosclerotic lesions in blacks.
An appealing possible explanation for the finding that blacks have significantly higher plasma levels of Lp(a) and yet do not have a correspondingly higher incidence of coronary atherosclerosis is that only a subset of apo(a) isoforms, those of smaller size (<22 kringle) are atherogenic.10,11,16 No evidence to support a relationship between the presence of coronary calcium and either apo(a) isoform size or plasma levels of Lp(a) in blacks or whites was found in the present study (Figures 1 and 4). Thus, if apo(a) isoforms have an independent effect on the atherogenicity of Lp(a), it does not manifest as increased calcium deposition in the coronary arteries.
The present study also represents one of the largest population-based samples in which plasma levels of Lp(a) and apo(a) isoforms have been measured in both blacks and whites with an Lp(a) assay that accurately measures the number of circulating Lp(a) particles, irrespective of the number of kringle 4 repeats.25 Black women had a significantly higher plasma level of Lp(a) than black men, as has been noted previously,20,39,40 whereas no such gender difference was present in whites. In the future, gender differences in plasma levels of Lp(a) may provide clues as to the nongenetic factors that contribute to plasma levels of Lp(a).
If the major effect of Lp(a) were to promote atherosclerosis, as has been proposed,24 individuals with high plasma levels of Lp(a) would have been expected to have a greater prevalence of coronary calcium in the present study. Yet, with the exception of black men with very high plasma levels of Lp(a), no evidence for an independent relationship between plasma levels of Lp(a) and coronary calcium, assessed either as a quantitative or qualitative trait, was seen in either blacks of whites; however, the odds of being CAC+ were greatly increased in whites who had other cardiovascular risk factors, which suggests Lp(a) may only increase the risk of coronary atherosclerosis in those already at higher risk.41 The finding that high plasma levels of Lp(a) significantly increased the odds for having significant coronary atherosclerosis in white men with high plasma levels of LDL-C has been observed in other studies.42–44 Synergistic effects of high plasma levels of Lp(a) and LDL-C on lesion development were also apparent when recombinant apo(a) was expressed in the Watanabe Heritable Hyperlipidemic rabbit.45
The present study does not conflict with the meta-analysis that found that plasma levels of Lp(a) are an independent risk factor for coronary events4 or the study that found a relationship with angina,11 because we did not measure these variables. The results of this study instead suggest that the major adverse cardiovascular effect of Lp(a) may not be due to its effects on the development of the atherosclerotic lesion but rather to its effects on lesion stability or susceptibility to thrombosis later in the transition of atherosclerosis to cardiovascular events. Additional large-scale studies need to be performed in blacks to determine whether Lp(a) levels confer risk for cardiac events.
Apo(a) isoform size cannot be used to explain the apparent paradox of blacks having higher plasma levels of Lp(a) and yet similar amounts of coronary atherosclerosis. Because coronary calcium was used as a marker for atherosclerosis in the present study, we cannot exclude the possibility that plasma levels of Lp(a) are associated with cardiovascular event rates in blacks, although there is not strong evidence to support this possibility.
The mechanisms responsible for ethnic- and gender-specific differences in the relative concentrations of Lp(a) and its association with coronary atherosclerosis remain unclear. Future studies focusing on these differences may provide much needed insights into the mysteries surrounding the physiological role and pathological effects of this enigmatic lipoprotein.19,46 It is possible that the absence of a relationship between elevated plasma levels of Lp(a) and coronary atherosclerosis in blacks (except those black men with very high levels) may reflect the action of a protective factor in this population. Identification of this factor could have therapeutic implications for other populations.
Acknowledgments
This study was supported by the Donald W. Reynolds Cardiovascular Clinical Research Center at Dallas and the National Institutes of Health (ROI-HL-47619). We thank Mujeeb Basit and Dr DuWayne Willet for their excellent assistance with use of the database for the Dallas Heart Study. We thank Drs Darren McGuire and Alice Chang for critical review of the manuscript.
References
Albers JJ, Adolphson JL, Hazzard WR. Radio-immunoassay of human plasma Lp(a) lipoprotein. J Lipid Res. 1977; 18: 331–338.
Barnathan ES. Has lipoprotein "little" (a) shrunk; JAMA. 1993; 270: 2224–2225.
Ridker PM. An epidemiologic reassessment of lipoprotein(a) and atherothrombotic risk. Trends Cardiovasc Med. 1995; 5: 225–229.
Danesh J, Collins R, Peto R. Lipoprotein(a) and coronary heart disease: meta-analysis of prospective studies. Circulation. 2000; 102: 1082–1085.
Guyton JR, Dahlèn GH, Patsch W, Kautz JA, Gotto AM Jr. Relationship of plasma lipoprotein Lp(a) levels to race and to apolipoprotein B. Arteriosclerosis. 1985; 5: 265–272.
Heiss G, Schonfeld G, Johnson JL, Heyden S, Hames CG, Tyroler HA. Black-white differences in plasma levels of apolipoproteins: the Evans County Heart Study. Am Heart J. 1984; 108: 807–814.
Keil JE, Sutherland SE, Knapp RG, Lackland DT, Gazes PC, Tyroler HA. Mortality rates and risk factors for coronary disease in Black as compared with White men and women. N Engl J Med. 1993; 329: 73–78.
Otten MW Jr, Teutsch SM, Williamson DF, Marks JS. The effect of known risk factors on the excess mortality of Black adults in the United States. JAMA. 1990; 263: 845–850.
Marcovina SM, Koschinsky ML. Lipoprotein(a) concentration and apolipoprotein(a) size: a synergistic role in advanced atherosclerosis; Circulation. 1999; 100: 1151–1153.
Paultre F, Pearson TA, Weil HF, Tuck CH, Myerson M, Rubin J, Francis CK, Marx HF, Philbin EF, Reed RG, Berglund L. High levels of Lp(a) with a small apo(a) isoform are associated with coronary artery disease in African American and white men. Arterioscler Thromb Vasc Biol. 2000; 20: 2619–2624.
Rifai N, Ma J, Sacks FM, Ridker PM, Hernandez WJ, Stampfer MJ, Marcovina SM. Apolipoprotein(a) size and lipoprotein(a) concentration and future risk of angina pectoris with evidence of severe coronary atherosclerosis in men: the Physicians’ Health Study. Clin Chem. 2004; 50: 1364–1371.
McLean JW, Tomlinson JE, Kuang W-J, Eaton DL, Chen EY, Fless GM, Scanu AM, Lawn RM. cDNA sequence of human apolipoprotein(a) is homologous to plasminogen. Nature. 1987; 330: 132–137.
Lackner C, Boerwinkle E, Leffert CC, Rahmig T, Hobbs HH. Molecular basis of apolipoprotein(a) isoform size heterogeneity as revealed by pulsed-field gel electrophoresis. J Clin Invest. 1991; 87: 2153–2161.
Hancock MA, Boffa MB, Marcovina SM, Nesheim ME, Koschinsky ML. Inhibition of plasminogen activation by lipoprotein(a): critical domains in apolipoprotein(a) and mechanism of inhibition on fibrin and degraded fibrin surfaces. J Biol Chem. 2003; 278: 23260–23269.
Kang C, Dominguez M, Loyau S, Miyata T, Durlach V, Angles-Cano E. Lp(a) particles mold fibrin-binding properties of apo(a) in size-dependent manner: a study with different-length recombinant apo(a), native Lp(a), and monoclonal antibody. Arterioscler Thromb Vasc Biol. 2002; 22: 1232–1238.
Sandholzer C, Saha N, Kark JD, Rees A, Jaross W, Dieplinger H, Hoppichler F, Boerwinkle E, Utermann G. Apo(a) isoforms predict risk for coronary heart disease: a study in six populations. Arterioscler Thromb. 1992; 12: 1214–1226.
Kraft H, Lingenhel A, Kochl S, Hoppichler F, Kronenberg F, Abe A, Muhlberger V, Schonitzer D, Utermann G. Apolipoprotein(a) kringle IV repeat number predicts risk for coronary heart disease. Arterioscler Thromb Vasc Biol. 1996; 16: 713–719.
Kronenberg F, Kronenberg MF, Kiechl S, Trenkwalder E, Santer P, Oberhollenzer F, Egger G, Utermann G, Willeit J. Role of lipoprotein(a) and apolipoprotein(a) phenotype in atherogenesis: prospective results from the Bruneck study. Circulation. 1999; 100: 1154–1160.
Utermann G. The mysteries of lipoprotein(a). Science. 1989; 246: 904–910.
Marcovina SM, Albers JJ, Wijsman E, Zhang Z, Chapman NH, Kennedy H. Differences in Lp[a] concentrations and apo[a] polymorphs between black and white Americans. J Lipid Res. 1996; 37: 2569–2585.
Victor RG, Haley RW, Willett D, Peshock RM, Vaeth PC, Leonard D, Basit M, Cooper R, Hobbs HH, for the Dallas Heart Study Investigators. A population-based probability sample for the multidisciplinary study of ethnic disparities in cardiovascular disease: recruitment and validation in the Dallas Heart Study. Am J Cardiol. 2004; 93: 1473–1480.
Sangiorgi G, Rumberger JA, Severson A, Edwards WD, Gregoire J, Fitzpatrick LA, Schwartz RS. Arterial calcification and not lumen stenosis is highly correlated with atherosclerotic plaque burden in humans: a histologic study of 723 coronary artery segments using nondecalcifying methodology. J Am Coll Cardiol. 1998; 31: 126–133.
Doherty TM, Asotra K, Fitzpatrick LA, Qiao JH, Wilkin DJ, Detrano RC, Dunstan CR, Shah PK, Rajavashisth TB. Calcification in atherosclerosis: bone biology and chronic inflammation at the arterial crossroads. Proc Natl Acad Sci U S A. 2003; 100: 11201–11206.
Bihari-Varga M, Gruber E, Rotheneder M, Zechner R, Kostner GM. Interaction of lipoprotein Lp(a) and low density lipoprotein with glycosaminoglycans from human aorta. Arteriosclerosis. 1988; 8: 851–857.
Marcovina SM, Albers JJ, Gabel B, Koschinsky ML, Gaur VP. Effect of the number of apolipoprotein(a) kringle 4 domains on immunochemical measurements of lipoprotein(a). Clin Chem. 1995; 41: 246–255.
Jain T, Yu Z, McGuire D, Peshock R, Willett D, Vega GL, Guerra R, Hobbs HH, Grundy SM, the DHS collaborators. African Americans and Caucasians have a similar prevalence of coronary calcium in the Dallas Heart Study. J Am Coll Cardiol. 2004; 44: 1011–1017.
Mostaza JM, Schulz I, Vega GL, Grundy SM. Comparison of pravastatin with crystalline nicotinic acid monotherapy in treatment of combined hyperlipidemia. Am J Cardiol. 1997; 79: 1298–1301.
Marcovina SM, Hobbs HH, Albers JJ. Relation between number of apolipoprotein(a) kringle 4 repeats and mobility of isoforms in agarose gel: a standardized isoform nomenclature. Clin Chem. 1996; 42: 436–439.
Agatston AS, Janowitz WR, Hildner FJ, Zusmer NR, Viamonte MJ, Detrano RC. Quantification of coronary artery calcium using ultrafast computed tomography. J Am Coll Cardiol. 1990; 15: 827–832.
Janowitz WR, Agatston AS, Kaplan G, Viamonte MJ. Differences in prevalence and extent of coronary artery calcium detected by ultrafast computed tomography in asymptomatic men and women. Am J Cardiol. 1993; 72: 247–254.
Rumberger JA, Brundage BH, Rader DJ, Kondos G. Electron beam computed tomographic coronary calcium scanning: a review and guidelines for use in asymptomatic persons. Mayo Clin Proc. 1999; 74: 243–252.
Casella G, Berger RL. Statistical Inference. 2nd ed. Pacific Grove, Calif: Duxbury; 2002.
Foody JM, Milberg JA, Robinson K, Pearce GL, Jacobsen DW, Sprecher DL. Homocysteine and lipoprotein(a) interact to increase CAD risk in young men and women. Arterioscler Thromb Vasc Biol. 2000; 20: 493–499.
Jha AK, Varosy PD, Kanaya AM, Hunninghake DB, Hlatky MA, Waters DD, Furberg CD, Shlipak MG. Differences in medical care and disease outcomes among black and white women with heart disease. Circulation. 2003; 108: 1089–1094.
Moliterno DJ, Jokinen EV, Miserez AR, Lange RA, Willard JE, Boerwinkle E, Hillis D, Hobbs HH. No association between plasma lipoprotein(a) concentrations and the presence of absence of coronary atherosclerosis in African-Americans. Arterioscler Thromb Vasc Biol. 1995; 15: 850–855.
Sorrentino MJ, Vielhauer C, Eisenbart JD, Fless GM, Scanu AM, Feldman T. Plasma lipoprotein(a) protein concentration and coronary artery disease in black patients compared with white patients. Am J Med. 1992; 93: 658–662.
Knapp RG, Schreiner PJ, Sutherland SE, Keil JE, Gilbert GE, Klein RL, Hames C, Tyroler HA. Serum lipoprotein(a) levels in elderly black and white men in the Charleston heart study. Clin Genet. 1993; 44: 225–231.
Sharrett AR, Ballantyne CM, Coady SA, Heiss G, Sorlie PD, Catellier D, Patsch W. Coronary heart disease prediction from lipoprotein cholesterol levels, triglycerides, lipoprotein(a), apolipoproteins A-I and B, and HDL density subfractions: the Atherosclerosis Risk in Communities (ARIC) Study. Circulation. 2001; 104: 1108–1113.
Rotimi CN, Cooper RS, Marcovina SM, McGee D, Owoaje E, Ladipo M. Serum distribution of lipoprotein(a) in African Americans and Nigerians: potential evidence for a genotype-environmental effect. Genet Epidemiol. 1997; 14: 157–168.
Ali S, Bunker CH, Aston CE, Ukoli FA, Kamboh MI. Apolipoprotein A kringle 4 polymorphism and serum lipoprotein (a) concentrations in African blacks. Hum Biol. 1998; 70: 477–490.
von Eckardstein A, Schulte H, Cullen P, Assmann G. Lipoprotein(a) further increases the risk of coronary events in men with high global cardiovascular risk. J Am Coll Cardiol. 2001; 37: 434–439.
Armstrong V, Cremer P, Eberle E, Manke A, Schulze F, Wieland H, Kreuzer H, Seidel D. The association between serum Lp(a) concentrations and angiographically assessed coronary atherosclerosis: dependence on serum LDL levels. Atherosclerosis. 1986; 62: 249–257.
Maher VM, Brown BG. Lipoprotein (a) and coronary heart disease. Curr Opin Lipidol. 1995; 6: 229–235.
Luc G, Bard JM, Arveiler D, Ferrieres J, Evans A, Amouyel P, Fruchart JC, Ducimetiere P. Lipoprotein (a) as a predictor of coronary heart disease: the PRIME Study. Atherosclerosis. 2002; 163: 377–384.
Sun H, Unoki H, Wang X, Liang J, Ichikawa T, Arai Y, Shiomi M, Marcovina SM, Watanabe T, Fan J. Lipoprotein(a) enhances advanced atherosclerosis and vascular calcification in WHHL transgenic rabbits expressing human apolipoprotein(a). J Biol Chem. 2002; 277: 47486–47492.
Hobbs HH, White AL. Lipoprotein(a): intrigues and insights. Curr Opin Lipidol. 1999; 10: 225–236.(Rudy Guerra, PhD; Zhaoxia)
Department of Statistics (R.G., Z.Y.), Rice University, Houston, Tex
Department of Medicine (S.M.), University of Washington, Seattle, Wash
Howard Hughes Medical Institute (H.H.H.) at the University of Texas Southwestern Medical Center, Dallas, Tex.
Abstract
Background— Elevated plasma levels of lipoprotein(a) [Lp(a)] are an independent risk factor for cardiovascular disease in whites. Blacks have 2- to 3-fold higher plasma levels of Lp(a) than whites and yet do not have a correspondingly higher rate of coronary events. It remains unclear whether elevated plasma levels of Lp(a) are an independent risk factor for coronary atherosclerosis in individuals of African descent.
Methods and Results— The relationship between plasma levels of Lp(a), apolipoprotein(a) isoform sizes, and the presence of coronary calcium was examined in 761 blacks and 527 whites (men aged >40 years, women aged >45 years) from a population-based sample. No relationship was found between plasma levels of Lp(a), apolipoprotein(a) isoform size, or a combination of these 2 variables and coronary artery calcium (CAC) in whites or blacks. No correlation was observed between plasma levels of Lp(a) and coronary calcium scores in any group, although all black men with very high plasma levels of Lp(a) (>300 μmol/L; n=7) were CAC-positive. Whites with high plasma levels of Lp(a) plus elevated plasma levels of LDL cholesterol (men) or reduced levels of HDL cholesterol (men and women) or who smoked (women) had a higher prevalence of CAC. In contrast, no joint effects between plasma levels of Lp(a) and other cardiovascular risk factors on coronary calcium were found in blacks.
Conclusions— No consistent independent relationship between plasma levels of Lp(a) or apolipoprotein(a) isoform size and coronary calcium was found in whites or blacks.
Key Words: lipoproteins ; apolipoproteins ; arteriosclerosis ; coronary disease ; risk factors
Introduction
Lipoprotein(a) [Lp(a)] is a cholesteryl ester–rich lipoprotein distinguished from other lipoproteins by the presence of a unique apolipoprotein (apo), apo(a). Plasma concentrations of Lp(a) vary over a 1000-fold range, and the distribution of levels is highly skewed, with most whites and Asians having low levels.1 Until recently, controversy persisted about the status of elevated plasma levels of Lp(a) as an independent risk factor for the development of coronary atherosclerosis2,3; however, a meta-analysis of 27 prospective studies, which included 5436 coronary heart disease cases, demonstrated a consistent, positive association between high plasma levels of Lp(a) and coronary atherosclerotic events in whites.4 The distribution of plasma levels of Lp(a) in blacks is less skewed, and the median level is 2 to 4 times greater than in whites.5 Paradoxically, the higher levels of Lp(a) in blacks are not associated with a corresponding increase in the prevalence of coronary atherosclerosis.6–8
A possible explanation for the apparent dissociation between plasma Lp(a) levels and coronary atherosclerotic risk in blacks is that only a subset of individuals with high plasma levels of Lp(a), those with apo(a) isoforms of small size, are at increased risk.9–11 The apo(a) gene (APOA) is highly polymorphic in size owing to allelic differences in the number of a kringle-encoding sequence resembling kringle 4 of plasminogen.12,13 In vitro studies suggest that apo(a) interferes with plasminogen, a central zymogen in thrombosis, and that smaller apo(a) isoforms are more effective at inhibiting thrombolysis.14,15 In some studies, whites with apo(a) isoforms of small size have an increased incidence of coronary atherosclerosis.16–18 Because plasma levels of Lp(a) are inversely related to the size of the apo(a) gene and isoforms in whites,19 most whites with high plasma levels of Lp(a) have a small isoform. Therefore, it has been difficult to distinguish the relative atherogenic effects of apo(a) isoform size and plasma levels of Lp(a) in whites. In contrast, many blacks with high plasma levels of Lp(a) have apo(a) isoforms of intermediate size.20 Thus, examination of the relationship between plasma levels of Lp(a), apo(a) isoforms, and coronary atherosclerosis in blacks provides an opportunity to assess the relative contributions of isoform size and plasma levels of Lp(a) to coronary atherosclerotic risk.
In the present study, we examined the relationship between plasma levels of Lp(a), apo(a) isoform size, and coronary atherosclerotic burden in 1288 blacks and whites from the Dallas Heart Study (DHS)21 using electron-beam computed tomography (EBCT), a sensitive and noninvasive method to assess coronary atherosclerotic burden.22,23 Because Lp(a) has been demonstrated in vitro to adhere to multiple components of the extracellular matrix in the arterial wall, it has been proposed that the accumulation of Lp(a) within the lesion wall promotes accumulation of LDL.24 As the atherosclerotic lesion develops, calcium is deposited in the arterial wall.23 If the major effect of Lp(a) were to promote atherosclerosis, it would be expected that individuals with high plasma levels of Lp(a) would have a greater prevalence and/or amount of coronary calcium.
Methods
Study Population
The study subjects were obtained from the DHS, a multiethnic population-based probability sample of Dallas County, Texas,21 in which ethnicity was self-reported. The Institutional Review Board of the University of Texas Southwestern Medical Center approved the study, and all participants provided informed consent in accordance with institutional guidelines. Fasting blood samples were obtained from 3398 subjects (1760 blacks, 1068 whites, and 570 Hispanics). Blood was maintained at 4°C until the plasma was separated, placed in aliquots, and stored at –80°C. Plasma levels of Lp(a) and apo(a) isoform sizes were determined with well-validated assays.25
Of the participants who underwent phlebotomy, 2971 (50% black) completed a clinic visit at which coronary artery calcification (CAC) was evaluated by EBCT. A total of 970 blacks and 593 whites (men aged >40 years and women aged >45 years) underwent duplicate EBCT scans. A complete data set of the variables for this analysis in men aged >40 years and women aged >45 years included 1288 subjects (380 black women, 241 white women, 381 black men, and 286 white men).26
Risk Factor Profiling of Subjects
The assessment of cardiovascular risk factor status in the DHS is as described previously.21 The mean of the middle 3 (of 5) blood pressure measurements from the clinic visit was used in the present study. Subjects with blood pressure >140/90 mm Hg or taking antihypertensive medications were classified as hypertensive. The plasma lipids and lipoprotein-cholesterol (C) levels were measured by enzymatic techniques.27 Plasma Lp(a) was measured by a sandwich ELISA that is insensitive to apo(a) isoform size.25 Apo(a) isoform sizes were determined by immunoblot analysis with an apo(a)-specific antibody.28 The primary smoking variable (yes/no) was defined as current smoking (within 30 days of the interview) and a lifetime history of having smoked at least 100 cigarettes.
EBCT Scan for Coronary Calcium
Two consecutive EBCT scans were performed on each study participant as described previously.26 A focus was defined as a region of 3 or more contiguous voxels with a computed tomography number >130 Hounsfield units (HU). The voxel size was 0.586x0.586x3 mm (field of view 30 cm, matrix 512, 3-mm slice), so that 3 voxels represent a volume of 3.08 mm.29 Scans were read in a blinded fashion by one author (RP), and the results were expressed in Agatston units.29,30 The mean of the 2 scores from each participant was used as the final CAC score.
EBCT scores were highly skewed,26 and standard transformations failed to achieve approximate normality, largely because of the high frequency of subjects with scores at or near zero. For the primary analysis, EBCT scores were classified as CAC positive (CAC+) when the mean EBCT score was >10 Agatston units or CAC negative (CAC–) when the mean EBCT score was 10 Agatston units. A threshold of 10 Agatston units was selected because >95% of subjects were concordant (above or below 10 Agatston units) for repeated measurements at this threshold.26
Statistical Analysis
No mathematical transformation of the EBCT data achieved an adequate approximation of normality, and thus, standard modeling approaches could not be used to analyze quantitative EBCT measurements. Therefore, individuals were classified as either CAC+ (mean EBCT score >10 Agatston units) or CAC– (mean EBCT score 10 Agatston units), and logistic regression was used to assess the association between CAC prevalence and cardiovascular risk factors. To consider the possible loss of information associated with the use of a binary scale for EBCT scores, we also used ordinal regression analysis and classified the amount of CAC based on the atherosclerotic burden using the Rumberger scale: 0 to 10 (minimal), 10 to 100 (mild), 100 to 400 (moderate), and >400 (extensive) Agatston units.31
Continuous variables were summarized with means or medians and SDs. Two-sample t tests and ANOVA F tests were used to assess differences in means. To compare 2 skewed distributions [eg, the Lp(a) levels in black men and in white men], we assessed differences in medians. For large sample sizes, the sample median is normally distributed,32 and therefore, the difference in sample medians is normally distributed. The reported probability values for median differences are thus derived from the normal based Wald test: Z=(W–0)/SE(W), where W is a test statistic, SE(W) is the standard error for W, and 0 is the hypothesized value for W. In this case, W is the difference in sample medians and 0 is zero. The analytic expression for SE(W) is based on the asymptotic variance formula32 for a sample median. 2 tests were used for contingency table analysis of categorical variables. Lp(a) levels were summarized on their original scale and log-transformed before logistic and ordinal regression analysis because of the skewness of the distribution.
The joint effect of Lp(a) with other cardiovascular risk factors on the presence of CAC can be analyzed with logistic and ordinal regression models that include both main and interaction effects; however, because of the large sample size in the present study, we were able to conduct a more straightforward analysis within each race-gender subgroup without imposing model assumptions. To assess the joint effect of plasma levels of Lp(a) and a binary risk factor on CAC status, we compared the odds of being CAC+ in subjects with the risk factor and high Lp(a) to the odds of being CAC+ in subjects without the risk factor and low Lp(a). To analyze the joint effect of Lp(a) and continuous risk factors, we defined the top 30% tail of the risk factor as a positive risk factor effect and the bottom 30% as a negative risk factor effect and proceeded as above. The ORs for the risk factors were age-adjusted with logistic regression as follows. Let P denote the probability of disease. With 2 predictor variables, x1 and x2, the logistic model associates a person’s log-odds of disease with the 2 predictors: log[P/(1–P)]=;0+;1x1+;2x2. To obtain age-adjusted ORs, we defined P as the probability of being CAC+, x1 as age, and x2 as an indicator variable for risk group membership (high, low) defined by a given risk factor. The estimated value of ;1 represents the change in log-odds of CAC+ associated with a 1-year increase in age, adjusted for x2. According to the model, this change is assumed to be constant for persons of all ages. Of particular interest is the estimated value of ;2, which is interpreted as the age-adjusted log-OR for x2. Therefore, if x2 represents an indicator (yes/no) for hypertension, then ;2 is the log-OR of CAC+ for a person with hypertension relative to a person without hypertension. The age-adjusted ORs were used to analyze the joint effects of Lp(a) and other cardiovascular factors on CAC status.
The DHS is a probability-based survey with oversampling for blacks.21 Sample weights were used to extrapolate statistical inferences to the population, as described previously.21 Statistical analyses were performed with and without sample weights. Because the 2 sets of results were almost always in agreement, we reported unweighted data and results. Analyses were completed with the statistical packages S-Plus (Insightful) and Stata (Stata Corporation).
Results
Plasma levels of Lp(a) were significantly higher in blacks than in whites (Table 1). Levels of plasma Lp(a) were significantly higher (P<0.0001) in black women than in black men, but no gender difference was apparent in whites. Distributions of plasma Lp(a) by ethnicity and gender were similar to those of prior studies19 and did not differ between subjects who did and did not obtain an EBCT scan (data not shown).
To determine whether plasma levels of Lp(a) were associated with the presence of coronary calcium, the prevalence of positive CAC scores was determined for individuals within each Lp(a) quintile (Figure 1). No significant differences were found in the proportion of individuals who were CAC+ among the quintiles. Similarly, median plasma levels of Lp(a) were not significantly different in black men, black women, white men, or white women who were CAC+ and CAC– (Table 2), nor was any significant association between log-transformed plasma levels of Lp(a) and CAC status found in either logistic or ordinal regressions (data not shown). The relationship between plasma levels of Lp(a) and coronary calcium after the amount of CAC was classified for each individual with the Rumberger scale31 was also examined; no significant relationship was found between the 2 variables for any of the ethnic and gender-specific groups (data not shown). Accounting for age and smoking did not change the negative results of the logistic and ordinal regressions.
Mean plasma levels of Lp(a) were also not significantly different in the CAC+ and CAC– subjects, with 1 exception. Black men who were CAC+ had a significantly higher mean plasma level of Lp(a) (95 nmol/L) than black men without CAC (79 nmol/L; P=0.03; Table 2). To probe why the mean (but not median) plasma level of Lp(a) was significantly higher in the CAC+ black men, we examined and compared the distribution of plasma Lp(a) levels in CAC+ and CAC– black men (Figure 2). Although no overall correlation between plasma Lp(a) levels and EBCT score was found in black men (r=–0.018), inspection of the distributions revealed an excess of black men with extremely high plasma levels of Lp(a) (>300 nmol/L) in the CAC+ group (n=7) compared with the CAC– group (n=0). No significant difference in mean plasma level of Lp(a) of the 2 groups was found after elimination of these 7 individuals. Examination of the clinical characteristics of the 7 individuals with Lp(a) levels >300 nmol/L revealed no enrichment for other cardiovascular risk factors (high blood pressure, low plasma HDL-C, diabetes, smoking, or elevated levels of LDL-C or C-reactive protein) or for other factors that increase both Lp(a) and atherosclerotic risk (eg, renal insufficiency). As expected given the inverse relationship between apo(a) size and plasma levels of Lp(a), the apo(a) isoform sizes were smaller than the mean in all 7 black men (smallest isoform in each subject ranged in size from 14 to 21). Only a singe white man had a plasma level of Lp(a)>300 nmol/L, and the smallest apo(a) isoform in that individual was 17; this subject has no detectable CAC. These findings suggests that black men with extremely high levels of Lp(a) may be at increased risk of atherosclerosis, but the small number of individuals in this category precludes meaningful statistical analysis. Moreover, 9 of the 13 black women with plasma Lp(a) levels over 300 nmol/L [who all had at least 1 apo(a) isoform <19] were CAC– (data not shown).
The size of the major apo(a) isoform was inversely related to plasma Lp(a) levels (Figure 3), which were uniformly higher in blacks than in whites across the entire spectrum of isoform sizes. The slope of the regression of log-Lp(a) on apo(a) isoform size was similar in all groups. To assess the relationship between apo(a) isoform size and CAC+ status, the sizes of the major apo(a) isoforms were divided into quartiles, and the prevalence of positive CAC scores was compared in the different groups (Figure 4). No significant difference (2 P>0.05 in all subgroups) in the prevalence of CAC+ scores was apparent across the apo(a) isoform quartiles. No association was found between apo(a) isoform size and EBCT scores when analyzed as a discretized trait with either logistic or ordinal regressions (data not shown).
It has been suggested that only individuals with elevated levels of Lp(a) and small apo(a) isoforms (K4 repeats <22) are at increased risk of cardiovascular disease.9,10,18,20 To test for this possibility, we examined the relationship between the size of the major apo(a) isoform and plasma levels of Lp(a) in black and white subjects classified by their CAC status (Figure 5). A joint effect would be evident if a predominance of CAC+ individuals had smaller apo(a) isoforms and higher plasma levels of Lp(a). No such relationship was observed in any of the 4 ethnic-gender groups, which indicates the absence of a joint effect. The slopes of the regression line of log-Lp(a) on apo(a) isoform size in each CAC+ and CAC– subgroup were not significantly different (P>0.05).
Finally, we investigated the possibility that high plasma levels of Lp(a) are atherogenic in the presence of other risk factors. CAC+ scores were adjusted for age in these analyses, because age was significantly associated with CAC+ status in all groups (P<0.01; Table 2). The odds of being CAC+ in subjects with high plasma levels of Lp(a) and high risk factors were compared with the odds in subjects with low plasma levels of Lp(a) and low risk factors. The OR of high LDL to low LDL was 1.5 (P=0.24) and increased to 4.2 (P=0.03) when high-LDL/high-Lp(a) was compared with low-LDL/low-Lp(a) groups in white men (Table 3). White men also showed a strong joint effect between high plasma levels of Lp(a) and a low plasma level of HDL-C (OR=8.3, P=0.01) and high levels of plasma triglyceride (OR=5.2, P=0.03); a similar joint effect between high plasma levels of Lp(a) and low plasma levels of HDL-C was seen in white women (OR increasing from 2.5 to 4.6). No interactive effects were seen between high plasma levels of Lp(a) and high plasma levels of homocysteine in either white women or men, in contrast to what was observed previously.33 Of the discrete variables examined (diabetes, hypertension, and smoking), only smoking showed a joint effect with plasma levels of Lp(a) in white women. To assess the validity of the significant joint effects, we evaluated possible confounding factors between the high-Lp(a)/high-risk-factor subjects and the low-Lp(a)/low-risk-factor subjects; no evidence for confounding was found for diabetes, smoking, hypertension, or body mass index. No joint effects between plasma levels of Lp(a) and any of the continuous variables (plasma levels of LDL-C, HDL-C, triglycerides, body mass index, homocysteine) or the discrete variables (diabetes, hypertension and smoking) were found in black men or women (data not shown).
Discussion
The major finding of this study was that the presence of CAC is not significantly related to plasma levels of Lp(a), apo(a) isoform size, or the combination of apo(a) isoform size and plasma level of Lp(a) in whites or in black women. These data constitute the most representative sample of blacks in which this relationship has been studied. Extremely high plasma levels of Lp(a) (>300 nmol/L), but not apo(a) isoform size, were associated with a higher prevalence of coronary calcium in black men. Further analysis revealed that elevated plasma levels of Lp(a) significantly increased the odds of having detectable coronary calcium in whites with selected cardiovascular risk factors, including high plasma levels of LDL-C, elevated plasma levels of triglycerides, or low plasma levels of HDL-C in men. In white women, elevated plasma levels of Lp(a) acted synergistically with a smoking history and with low plasma levels of HDL-C to increase the odds of being CAC+. No joint effects between elevated levels of Lp(a) and other cardiovascular risk factors on coronary calcium were found in black men or black women.
Because the median level of Lp(a) in blacks is 2- to 3-fold higher than in whites, and the prevalence of established cardiovascular risk factors is equivalent or greater in blacks than in whites (Table 1), the burden of coronary atherosclerosis would be expected to be greater in the black population. Yet, most studies find the rate of preclinical atherosclerosis and ischemic events is similar in blacks and whites,8,26 although black women may have increased odds of disease compared with white women.34 A major limitation of most studies examining the relationship between plasma levels of Lp(a) and coronary atherosclerosis in blacks is that they have been small and not representative.10,35–37 The largest study assessing the effect of plasma levels of Lp(a) on cardiovascular events in blacks was the Atherosclerosis Risk in Communities Study,38 in which only a marginally significant relationship was found. The DHS is the most representative study to date that has examined the relationship between plasma levels of Lp(a) and coronary atherosclerotic burden. The present results support previous findings that plasma levels of Lp(a) are not associated with the presence of coronary atherosclerotic lesions in blacks.
An appealing possible explanation for the finding that blacks have significantly higher plasma levels of Lp(a) and yet do not have a correspondingly higher incidence of coronary atherosclerosis is that only a subset of apo(a) isoforms, those of smaller size (<22 kringle) are atherogenic.10,11,16 No evidence to support a relationship between the presence of coronary calcium and either apo(a) isoform size or plasma levels of Lp(a) in blacks or whites was found in the present study (Figures 1 and 4). Thus, if apo(a) isoforms have an independent effect on the atherogenicity of Lp(a), it does not manifest as increased calcium deposition in the coronary arteries.
The present study also represents one of the largest population-based samples in which plasma levels of Lp(a) and apo(a) isoforms have been measured in both blacks and whites with an Lp(a) assay that accurately measures the number of circulating Lp(a) particles, irrespective of the number of kringle 4 repeats.25 Black women had a significantly higher plasma level of Lp(a) than black men, as has been noted previously,20,39,40 whereas no such gender difference was present in whites. In the future, gender differences in plasma levels of Lp(a) may provide clues as to the nongenetic factors that contribute to plasma levels of Lp(a).
If the major effect of Lp(a) were to promote atherosclerosis, as has been proposed,24 individuals with high plasma levels of Lp(a) would have been expected to have a greater prevalence of coronary calcium in the present study. Yet, with the exception of black men with very high plasma levels of Lp(a), no evidence for an independent relationship between plasma levels of Lp(a) and coronary calcium, assessed either as a quantitative or qualitative trait, was seen in either blacks of whites; however, the odds of being CAC+ were greatly increased in whites who had other cardiovascular risk factors, which suggests Lp(a) may only increase the risk of coronary atherosclerosis in those already at higher risk.41 The finding that high plasma levels of Lp(a) significantly increased the odds for having significant coronary atherosclerosis in white men with high plasma levels of LDL-C has been observed in other studies.42–44 Synergistic effects of high plasma levels of Lp(a) and LDL-C on lesion development were also apparent when recombinant apo(a) was expressed in the Watanabe Heritable Hyperlipidemic rabbit.45
The present study does not conflict with the meta-analysis that found that plasma levels of Lp(a) are an independent risk factor for coronary events4 or the study that found a relationship with angina,11 because we did not measure these variables. The results of this study instead suggest that the major adverse cardiovascular effect of Lp(a) may not be due to its effects on the development of the atherosclerotic lesion but rather to its effects on lesion stability or susceptibility to thrombosis later in the transition of atherosclerosis to cardiovascular events. Additional large-scale studies need to be performed in blacks to determine whether Lp(a) levels confer risk for cardiac events.
Apo(a) isoform size cannot be used to explain the apparent paradox of blacks having higher plasma levels of Lp(a) and yet similar amounts of coronary atherosclerosis. Because coronary calcium was used as a marker for atherosclerosis in the present study, we cannot exclude the possibility that plasma levels of Lp(a) are associated with cardiovascular event rates in blacks, although there is not strong evidence to support this possibility.
The mechanisms responsible for ethnic- and gender-specific differences in the relative concentrations of Lp(a) and its association with coronary atherosclerosis remain unclear. Future studies focusing on these differences may provide much needed insights into the mysteries surrounding the physiological role and pathological effects of this enigmatic lipoprotein.19,46 It is possible that the absence of a relationship between elevated plasma levels of Lp(a) and coronary atherosclerosis in blacks (except those black men with very high levels) may reflect the action of a protective factor in this population. Identification of this factor could have therapeutic implications for other populations.
Acknowledgments
This study was supported by the Donald W. Reynolds Cardiovascular Clinical Research Center at Dallas and the National Institutes of Health (ROI-HL-47619). We thank Mujeeb Basit and Dr DuWayne Willet for their excellent assistance with use of the database for the Dallas Heart Study. We thank Drs Darren McGuire and Alice Chang for critical review of the manuscript.
References
Albers JJ, Adolphson JL, Hazzard WR. Radio-immunoassay of human plasma Lp(a) lipoprotein. J Lipid Res. 1977; 18: 331–338.
Barnathan ES. Has lipoprotein "little" (a) shrunk; JAMA. 1993; 270: 2224–2225.
Ridker PM. An epidemiologic reassessment of lipoprotein(a) and atherothrombotic risk. Trends Cardiovasc Med. 1995; 5: 225–229.
Danesh J, Collins R, Peto R. Lipoprotein(a) and coronary heart disease: meta-analysis of prospective studies. Circulation. 2000; 102: 1082–1085.
Guyton JR, Dahlèn GH, Patsch W, Kautz JA, Gotto AM Jr. Relationship of plasma lipoprotein Lp(a) levels to race and to apolipoprotein B. Arteriosclerosis. 1985; 5: 265–272.
Heiss G, Schonfeld G, Johnson JL, Heyden S, Hames CG, Tyroler HA. Black-white differences in plasma levels of apolipoproteins: the Evans County Heart Study. Am Heart J. 1984; 108: 807–814.
Keil JE, Sutherland SE, Knapp RG, Lackland DT, Gazes PC, Tyroler HA. Mortality rates and risk factors for coronary disease in Black as compared with White men and women. N Engl J Med. 1993; 329: 73–78.
Otten MW Jr, Teutsch SM, Williamson DF, Marks JS. The effect of known risk factors on the excess mortality of Black adults in the United States. JAMA. 1990; 263: 845–850.
Marcovina SM, Koschinsky ML. Lipoprotein(a) concentration and apolipoprotein(a) size: a synergistic role in advanced atherosclerosis; Circulation. 1999; 100: 1151–1153.
Paultre F, Pearson TA, Weil HF, Tuck CH, Myerson M, Rubin J, Francis CK, Marx HF, Philbin EF, Reed RG, Berglund L. High levels of Lp(a) with a small apo(a) isoform are associated with coronary artery disease in African American and white men. Arterioscler Thromb Vasc Biol. 2000; 20: 2619–2624.
Rifai N, Ma J, Sacks FM, Ridker PM, Hernandez WJ, Stampfer MJ, Marcovina SM. Apolipoprotein(a) size and lipoprotein(a) concentration and future risk of angina pectoris with evidence of severe coronary atherosclerosis in men: the Physicians’ Health Study. Clin Chem. 2004; 50: 1364–1371.
McLean JW, Tomlinson JE, Kuang W-J, Eaton DL, Chen EY, Fless GM, Scanu AM, Lawn RM. cDNA sequence of human apolipoprotein(a) is homologous to plasminogen. Nature. 1987; 330: 132–137.
Lackner C, Boerwinkle E, Leffert CC, Rahmig T, Hobbs HH. Molecular basis of apolipoprotein(a) isoform size heterogeneity as revealed by pulsed-field gel electrophoresis. J Clin Invest. 1991; 87: 2153–2161.
Hancock MA, Boffa MB, Marcovina SM, Nesheim ME, Koschinsky ML. Inhibition of plasminogen activation by lipoprotein(a): critical domains in apolipoprotein(a) and mechanism of inhibition on fibrin and degraded fibrin surfaces. J Biol Chem. 2003; 278: 23260–23269.
Kang C, Dominguez M, Loyau S, Miyata T, Durlach V, Angles-Cano E. Lp(a) particles mold fibrin-binding properties of apo(a) in size-dependent manner: a study with different-length recombinant apo(a), native Lp(a), and monoclonal antibody. Arterioscler Thromb Vasc Biol. 2002; 22: 1232–1238.
Sandholzer C, Saha N, Kark JD, Rees A, Jaross W, Dieplinger H, Hoppichler F, Boerwinkle E, Utermann G. Apo(a) isoforms predict risk for coronary heart disease: a study in six populations. Arterioscler Thromb. 1992; 12: 1214–1226.
Kraft H, Lingenhel A, Kochl S, Hoppichler F, Kronenberg F, Abe A, Muhlberger V, Schonitzer D, Utermann G. Apolipoprotein(a) kringle IV repeat number predicts risk for coronary heart disease. Arterioscler Thromb Vasc Biol. 1996; 16: 713–719.
Kronenberg F, Kronenberg MF, Kiechl S, Trenkwalder E, Santer P, Oberhollenzer F, Egger G, Utermann G, Willeit J. Role of lipoprotein(a) and apolipoprotein(a) phenotype in atherogenesis: prospective results from the Bruneck study. Circulation. 1999; 100: 1154–1160.
Utermann G. The mysteries of lipoprotein(a). Science. 1989; 246: 904–910.
Marcovina SM, Albers JJ, Wijsman E, Zhang Z, Chapman NH, Kennedy H. Differences in Lp[a] concentrations and apo[a] polymorphs between black and white Americans. J Lipid Res. 1996; 37: 2569–2585.
Victor RG, Haley RW, Willett D, Peshock RM, Vaeth PC, Leonard D, Basit M, Cooper R, Hobbs HH, for the Dallas Heart Study Investigators. A population-based probability sample for the multidisciplinary study of ethnic disparities in cardiovascular disease: recruitment and validation in the Dallas Heart Study. Am J Cardiol. 2004; 93: 1473–1480.
Sangiorgi G, Rumberger JA, Severson A, Edwards WD, Gregoire J, Fitzpatrick LA, Schwartz RS. Arterial calcification and not lumen stenosis is highly correlated with atherosclerotic plaque burden in humans: a histologic study of 723 coronary artery segments using nondecalcifying methodology. J Am Coll Cardiol. 1998; 31: 126–133.
Doherty TM, Asotra K, Fitzpatrick LA, Qiao JH, Wilkin DJ, Detrano RC, Dunstan CR, Shah PK, Rajavashisth TB. Calcification in atherosclerosis: bone biology and chronic inflammation at the arterial crossroads. Proc Natl Acad Sci U S A. 2003; 100: 11201–11206.
Bihari-Varga M, Gruber E, Rotheneder M, Zechner R, Kostner GM. Interaction of lipoprotein Lp(a) and low density lipoprotein with glycosaminoglycans from human aorta. Arteriosclerosis. 1988; 8: 851–857.
Marcovina SM, Albers JJ, Gabel B, Koschinsky ML, Gaur VP. Effect of the number of apolipoprotein(a) kringle 4 domains on immunochemical measurements of lipoprotein(a). Clin Chem. 1995; 41: 246–255.
Jain T, Yu Z, McGuire D, Peshock R, Willett D, Vega GL, Guerra R, Hobbs HH, Grundy SM, the DHS collaborators. African Americans and Caucasians have a similar prevalence of coronary calcium in the Dallas Heart Study. J Am Coll Cardiol. 2004; 44: 1011–1017.
Mostaza JM, Schulz I, Vega GL, Grundy SM. Comparison of pravastatin with crystalline nicotinic acid monotherapy in treatment of combined hyperlipidemia. Am J Cardiol. 1997; 79: 1298–1301.
Marcovina SM, Hobbs HH, Albers JJ. Relation between number of apolipoprotein(a) kringle 4 repeats and mobility of isoforms in agarose gel: a standardized isoform nomenclature. Clin Chem. 1996; 42: 436–439.
Agatston AS, Janowitz WR, Hildner FJ, Zusmer NR, Viamonte MJ, Detrano RC. Quantification of coronary artery calcium using ultrafast computed tomography. J Am Coll Cardiol. 1990; 15: 827–832.
Janowitz WR, Agatston AS, Kaplan G, Viamonte MJ. Differences in prevalence and extent of coronary artery calcium detected by ultrafast computed tomography in asymptomatic men and women. Am J Cardiol. 1993; 72: 247–254.
Rumberger JA, Brundage BH, Rader DJ, Kondos G. Electron beam computed tomographic coronary calcium scanning: a review and guidelines for use in asymptomatic persons. Mayo Clin Proc. 1999; 74: 243–252.
Casella G, Berger RL. Statistical Inference. 2nd ed. Pacific Grove, Calif: Duxbury; 2002.
Foody JM, Milberg JA, Robinson K, Pearce GL, Jacobsen DW, Sprecher DL. Homocysteine and lipoprotein(a) interact to increase CAD risk in young men and women. Arterioscler Thromb Vasc Biol. 2000; 20: 493–499.
Jha AK, Varosy PD, Kanaya AM, Hunninghake DB, Hlatky MA, Waters DD, Furberg CD, Shlipak MG. Differences in medical care and disease outcomes among black and white women with heart disease. Circulation. 2003; 108: 1089–1094.
Moliterno DJ, Jokinen EV, Miserez AR, Lange RA, Willard JE, Boerwinkle E, Hillis D, Hobbs HH. No association between plasma lipoprotein(a) concentrations and the presence of absence of coronary atherosclerosis in African-Americans. Arterioscler Thromb Vasc Biol. 1995; 15: 850–855.
Sorrentino MJ, Vielhauer C, Eisenbart JD, Fless GM, Scanu AM, Feldman T. Plasma lipoprotein(a) protein concentration and coronary artery disease in black patients compared with white patients. Am J Med. 1992; 93: 658–662.
Knapp RG, Schreiner PJ, Sutherland SE, Keil JE, Gilbert GE, Klein RL, Hames C, Tyroler HA. Serum lipoprotein(a) levels in elderly black and white men in the Charleston heart study. Clin Genet. 1993; 44: 225–231.
Sharrett AR, Ballantyne CM, Coady SA, Heiss G, Sorlie PD, Catellier D, Patsch W. Coronary heart disease prediction from lipoprotein cholesterol levels, triglycerides, lipoprotein(a), apolipoproteins A-I and B, and HDL density subfractions: the Atherosclerosis Risk in Communities (ARIC) Study. Circulation. 2001; 104: 1108–1113.
Rotimi CN, Cooper RS, Marcovina SM, McGee D, Owoaje E, Ladipo M. Serum distribution of lipoprotein(a) in African Americans and Nigerians: potential evidence for a genotype-environmental effect. Genet Epidemiol. 1997; 14: 157–168.
Ali S, Bunker CH, Aston CE, Ukoli FA, Kamboh MI. Apolipoprotein A kringle 4 polymorphism and serum lipoprotein (a) concentrations in African blacks. Hum Biol. 1998; 70: 477–490.
von Eckardstein A, Schulte H, Cullen P, Assmann G. Lipoprotein(a) further increases the risk of coronary events in men with high global cardiovascular risk. J Am Coll Cardiol. 2001; 37: 434–439.
Armstrong V, Cremer P, Eberle E, Manke A, Schulze F, Wieland H, Kreuzer H, Seidel D. The association between serum Lp(a) concentrations and angiographically assessed coronary atherosclerosis: dependence on serum LDL levels. Atherosclerosis. 1986; 62: 249–257.
Maher VM, Brown BG. Lipoprotein (a) and coronary heart disease. Curr Opin Lipidol. 1995; 6: 229–235.
Luc G, Bard JM, Arveiler D, Ferrieres J, Evans A, Amouyel P, Fruchart JC, Ducimetiere P. Lipoprotein (a) as a predictor of coronary heart disease: the PRIME Study. Atherosclerosis. 2002; 163: 377–384.
Sun H, Unoki H, Wang X, Liang J, Ichikawa T, Arai Y, Shiomi M, Marcovina SM, Watanabe T, Fan J. Lipoprotein(a) enhances advanced atherosclerosis and vascular calcification in WHHL transgenic rabbits expressing human apolipoprotein(a). J Biol Chem. 2002; 277: 47486–47492.
Hobbs HH, White AL. Lipoprotein(a): intrigues and insights. Curr Opin Lipidol. 1999; 10: 225–236.(Rudy Guerra, PhD; Zhaoxia)