Increased risk of incident stroke associated with the cyclooxygenase 2 (COX-2) G−765C polymorphism in African-Americans: The Atherosclerosis Risk in Communities Study

Atherosclerosis 196 (2008) 926–930

Increased risk of incident stroke associated with the cyclooxygenase 2 (COX-2) G−765C polymorphism in African-Americans: The Atherosclerosis Risk in Communities Study Shun Kohsaka a , Kelly A. Volcik b , Aaron R. Folsom c , Kenneth K. Wu d , Christie M. Ballantyne e , James T. Willerson a , Eric Boerwinkle b,∗ a Texas Heart Institute, Baylor College of Medicine, Houston, TX, United States Human Genetics Center, University of Texas Health Science Center, 1200 Herman Pressler Dr., Houston, TX 77030, United States c Division of Epidemiology and Community Health, University of Minnesota School of Public Health, Minneapolis, MN, United States d Vascular Biology Research Center and Division of Hematology, University of Texas Health Science Center, Houston, TX, United States e Department of Medicine, Baylor College of Medicine, Houston, TX, United States b

Received 13 October 2006; received in revised form 14 January 2007; accepted 8 February 2007 Available online 9 March 2007

Abstract Background: A hallmark feature of atherosclerosis is inflammation mediated by prostaglandins (PGs) catalyzed by the enzyme cyclooxygenase (COX). The present study explored whether the COX-2 G−765C polymorphism contributes to increased incidence of coronary heart disease (CHD) or stroke in the large prospective Atherosclerosis Risk in Communities (ARIC) Study. Methods: Incidences of CHD and stroke were identified through annual follow-up and hospital and death certificate surveillance. The study included 1488 incident CHD and 527 stroke events after an average of 14 years of follow-up. The frequency of the −765C variant allele was markedly different between African-Americans and whites, therefore all analyses were performed separately by race. Due to the small number of persons with the −765CC genotype, heterozygous and homozygous variant genotypes were combined for this analysis. Results: The COX-2 G−765C polymorphism was not a significant predictor of CHD in either racial group, but it was a significant predictor of incident stroke in African-Americans. After adjustment for age and gender, the hazard rate ratio for developing stroke for the CG+CC genotypes relative to the GG genotype was 1.34 (95% confidence interval [CI] 1.03–1.74, P = 0.03) in African-Americans. This result was essentially unchanged when established predictors such as smoking, diabetes and hypertension were added to the model (HRR 1.34, 95%CI 1.03–1.76, P = 0.03). Conclusion: We have found the COX-2 G−765C polymorphism to be a risk factor for incident stroke in African-Americans. This study provides additional evidence for utilizing inflammation-related genetic polymorphisms for identifying individuals at increased risk for stroke. © 2007 Elsevier Ireland Ltd. All rights reserved. Keywords: Cyclooxygenase 2; Polymorphism; Genetic; Risk factors; Cerebrovascular accident; Coronary disease

1. Background Coronary heart disease (CHD) and stroke are complex disorders characterized by variable gene expression and intermediate phenotypes, such as atherosclerosis. A hallmark feature of atherosclerosis is inflammation medi∗

Corresponding author. Tel.: +1 713 500 9800; fax: +1 713 500 0900. E-mail address: [email protected] (E. Boerwinkle).

0021-9150/$ – see front matter © 2007 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.atherosclerosis.2007.02.010

ated by prostaglandins (PGs) catalyzed by the enzyme cyclooxygenase (COX) [1]. PGs influence the development of atherosclerosis by modulating the inflammatory response, the expression of metalloproteinases, and the growth of cells implicated in the process, such as vascular smooth muscle cells [2–4]. There are at least two COX genes encoding two enzymes: COX-1 and COX-2. COX-1, expressed constitutively in most tissues, is thought to release PGs involved in the main-

S. Kohsaka et al. / Atherosclerosis 196 (2008) 926–930

tenance of gastrointestinal tract and vascular homeostasis. COX-2, undetectable in most tissues, can be upregulated by bacterial lipopolysaccharides, cytokines, growth factors, and tumor promoters, suggesting its relevance to inflammation and cancer [5]. Although COX-2 expression is enhanced in endothelial cells of patients with CHD leading to enhanced production of matrix metalloproteinases in atherosclerotic plaques, the inhibition of COX-2 function to regulate atherogenesis has been controversial [6,7]. A recent placebo-controlled trial of rofecoxib, a COX inhibitor, revealed a two-fold increase of myocardial infarction and stroke, which led to the withdrawal of the drug from the market [8]. Recently, a functional COX-2 promoter polymorphism of a guanine to cytosine substitution at position −765 (G−765C) has been recognized and proposed to alter Sp1 binding to this region [9–11]. A single nested case–control study suggested the G−765C polymorphism to be protective for myocardial infarction and stroke, but this finding has not been replicated [12]. The present study explored whether the COX-2 G−765C polymorphism was associated with incident CHD or stroke in the Atherosclerosis Risk in Communities (ARIC) study.

927

revascularization. Incident stroke cases were defined as validated definite or probable hospitalized hemorrhagic, embolic, or thrombotic strokes. For the current study, a total of 1488 CHD cases and 527 stroke cases were identified after an average of 14 years of follow-up. 2.2. Baseline examination and laboratory measures Seated blood pressure was measured three times with a random-zero sphygmomanometer and the last two measurements were averaged. Hypertension was defined as systolic blood pressure ≥140 mm Hg or diastolic blood pressure ≥90 mm Hg or current use of antihypertensive medications. Questionnaires and in-person interviews were used to assess use of antihypertensive medications. Diabetes was defined by a fasting glucose level ≥126 mg/dL, a non-fasting glucose level ≥200 mg/dL, and/or history of or treatment for diabetes. Cigarette-smoking status was analyzed by comparing current smokers to individuals who had formerly or never smoked. Body mass index (BMI, kg/m2 ) was calculated from height and weight measurements. Plasma total cholesterol was measured by an enzymatic method [14]. High-density lipoprotein (HDL) cholesterol was measured after dextran–magnesium precipitation of non-HDL lipoproteins.

2. Methods

2.3. Genotype determination

2.1. The ARIC Study

Genotyping of the COX-2 G−765C polymorphism was performed in a multiplex reaction utilizing the MassARRAY system (Sequenom, San Diego, CA). A 117 bp product was amplified using 0.05 mM each of the forward primer 5 -ACGTTGGATGACAGGGTAACTGCTTAGGAC-3 and the reverse primer 5 -ACGTTGGATGACTGTTCTCCGTACCTTCAC-3 , 7.5 ng DNA, 2.5 mM MgCl2 , and 0.1 U of AmpliTaq Gold DNA Polymerase (Applied Biosystems, Foster City, CA) in a 5.0 ␮L reaction volume. After an initial step of 15 min at 94 ◦ C to activate the AmpliTaq Gold, products were amplified using 45 cycles of 94 ◦ C for 2 s, 56 ◦ C for 3 s and 72 ◦ C for 1 min. PCR was followed by the addition of a 2.0 ␮L shrimp alkaline phosphatase (SAP) enzyme reaction mix, SAP inactivation, and an extension reaction using 1.0 ␮M of the extension primer 5 -GGAGAATTTACCTTTCCC-3 and 0.5 U of Thermo Sequenase (GE Healthcare, Fairfield, CT). Extension products were desalted with SpectroCLEAN resin and 12 nL of the analyte was transferred onto a 384-well SpectroCHIP (Sequenom, San Diego, CA). Mass spectrometry analysis was performed using the SpectroREADER (Bruker Biflex III, MALDI-TOF instrument), the mass spectra were collected, and SpectroCALLER (Sequenom, San Diego, CA) automatically assigned the genotype calls.

Study participants were selected from the ARIC Study, a prospective investigation of atherosclerosis and its clinical sequelae involving 15,792 individuals aged 45–64 years at recruitment (1987–1989). Institutional Review Boards approved the ARIC Study, and all participants provided their written informed consent. A detailed description of the ARIC Study design and methods, as well as details on quality assurance for ascertainment and classification of CHD and stroke events, have been published elsewhere [13]. Briefly, subjects were selected by probability sampling from four communities: Forsyth County, North Carolina; Jackson, Mississippi; northwestern suburbs of Minneapolis, Minnesota; and Washington County, Maryland. Incidence of CHD and stroke were determined by contacting participants annually to identify hospitalizations during the previous year and by surveying discharge lists from local hospitals and death certificates from state vital statistics offices for potential cardiovascular and cerebrovascular events [13]. Participants were excluded from analyses (n = 2362) if they had (1) a positive or unknown history of prevalent CHD or stroke or a history of transient ischemic attack (TIA)/stroke symptoms at the initial clinic visit, (2) prohibited use of their DNA for research purposes, (3) an ethnic background other than white or African-American, or (4) insufficient genotype information. Incident CHD cases were defined as a definite or probable myocardial infarction (MI), a silent MI between examinations by ECG, a definite CHD death, or a coronary

2.4. Statistical analyses All statistical analyses were conducted utilizing STATA Version 8.0 (College Station, TX). Allele frequencies were

928

S. Kohsaka et al. / Atherosclerosis 196 (2008) 926–930

estimated by gene counting. Hardy–Weinberg equilibrium expectations were tested using a χ2 goodness-of-fit test. Cox proportional hazards (PH) models were used to estimate the hazard rate ratios (HRRs) of incident CHD and stroke. Due to the low frequency of individuals having the homozygous −765CC genotype, initial analyses were performed by comparing combined variant heterozygote and homozygote individuals (−765GC + −765CC) to individuals homozygous for the wild-type allele (−765GG) (dominant model). A recessive model comparing CC homozygous individuals to combined heterozygote and GG homozygous individuals was also performed. All Cox PH models were analyzed separately for African-Americans and whites. For analyses of incident CHD/stroke cases, follow-up time intervals were defined as the time between the initial clinical visit and the date of the first CHD/stroke event. For non-cases, follow-up continued until 31 December 2002, the date of death, or the date of last contact if lost to follow-up, whichever came first. For incident CHD analysis models, covariates included age, gender, field center, BMI, HDL and total cholesterol, smoking, diabetes and hypertension status. For incident stroke analysis models, covariates (identified by the National Institute of Neurological Disorders and Stroke, www.ninds.nih.gov) included age, gender, field center, smoking, diabetes and hypertension status. Covariates were assessed for statistical significance in the models by the Wald χ2 statistic.

3. Results Genotype frequencies of the COX-2 G−765C polymorphism for incident CHD cases and non-cases, in addition to incident stroke cases and non-cases, are presented in Table 1 by race. Due to the marked variant allele frequency difference between whites (0.15) and African-Americans (0.32), subsequent CHD and stroke analyses were performed separately by racial group. Genotype frequencies for the COX-2 G−765C polymorphism did not differ between CHD cases and non-cases in whites or African-Americans. Genotype frequencies between stroke cases and non-cases were significantly different in African-Americans, with the combined variant genotypes more frequent in the stroke cases (61%)

compared to non-cases (54%). No genotype frequency differences were observed between white stroke cases and non-cases. Results from Cox proportional hazards (PH) models used to estimate the hazard rate ratios (HRRs) of incident CHD for individuals with or without genotypes carrying the COX2 variant allele are presented in Table 2 by racial group. After adjustment for age and gender (Model 1), genotypes containing the COX-2 −765C allele were not a significant predictor of CHD in African-Americans or whites (P = 0.4 and 0.5, respectively). Similarly, no association was observed in African-Americans or whites after further adjustment for multiple CHD risk factors (Model 2, P = 0.8 and 0.6, respectively). Results from Cox PH models used to estimate the HRRs of incident stroke for individuals with versus without genotypes carrying the COX-2 variant allele are presented in Table 2 by racial group. After adjustment for age and gender (Model 1), genotypes containing the COX-2 −765C allele were a significant predictor of stroke in African-Americans (P = 0.03), but not in whites (P = 0.3). After further adjustment for multiple stroke risk factors, the COX-2 −765C variant allele remained a significant predictor of incident stroke in AfricanAmericans (Model 2, P = 0.03). A recessive model comparing CC homozygous individuals to combined heterozygote and GG homozygous individuals was also performed, with no association with incident stroke observed. Additional analyses were performed for ischemic and hemorrhagic stroke separately, but the results were not different: both subgroups showed increased risk, with a slightly higher risk observed for hemorrhagic strokes.

4. Discussion In this study, we show that the COX-2 G−765C polymorphism is associated with risk of stroke in African-Americans belonging to a population-based cohort. Individuals carrying the −765C allele had a 1.34-fold increased risk of stroke compared to those homozygous for the −765G allele. In both whites and African-Americans from this study, the COX2 G−765C polymorphism was not associated with incident

Table 1 COX-2 genotype frequencies for incident CHD cases, incident stroke cases and non-cases COX-2 genotype

Incident CHD cases, n (%)

Non-cases, n (%)

Pa

Incident stroke cases, n (%)

Non-cases, n (%)

Pa

Whites GG CG CC

825 (71) 296 (26) 33 (3)

6391 (73) 2202 (25) 221 (2)

Ref 0.5

220 (74) 64 (22) 11 (4)

6996 (72) 2434 (25) 243 (3)

Ref 0.4

African-Americans GG CG CC

160 (48) 146 (44) 28 (8)

1411 (45) 1391 (45) 326 (10)

Ref 0.3

90 (39) 114 (49) 28 (12)

1481 (46) 1423 (44) 326 (10)

Ref 0.03

a P-value comparing genotype frequencies between cases and non-cases, combining variant homozygous and heterozygous genotypes (CG + CC), with GG genotype serving as referent group.

S. Kohsaka et al. / Atherosclerosis 196 (2008) 926–930

929

Table 2 Hazard rate ratios (HRRs) relating COX-2 variant genotypes to incident CHD and stroke Incident CHD genotype status††

Model 1a , HRR (95%CI), P

Model 2b , HRR (95%CI), P

Whites GG vs. CG GG vs. CC GG vs. CG + CC GG + CG vs. CC

1.03 (0.90–1.17), 0.7 1.18 (0.84–1.68), 0.3 1.04 (0.92–1.18), 0.5 1.17 (0.83–1.66), 0.4

1.03 (0.90–1.18), 0.7 1.07 (0.75–1.53), 0.7 1.03 (0.91–1.18), 0.6 1.06 (0.75–1.51), 0.7

African-Americans CC GG vs. CG GG vs. CC GG vs. CG + CC GG + CG vs. CC

0.94 (0.75–1.18), 0.6 0.81 (0.54–1.21), 0.3 0.92 (0.74–1.14), 0.4 0.83 (0.57–1.23), 0.4

1.00 (0.79–1.26), 1.0 0.84 (0.55–1.26), 0.4 0.97 (0.78–1.21), 0.8 0.84 (0.56–1.24), 0.4

Incident stroke genotype status††

Model 1a , HRR (95%CI), P

Model 2† , HRR (95%CI), P

Whites GG vs. CG GG vs. CC GG vs. CG + CC GG + CG vs. CC

0.82 (0.62–1.09), 0.2 1.49 (0.81–2.73), 0.2 0.88 (0.68–1.15), 0.3 1.56 (0.85–2.85), 0.1

0.80 (0.60–1.05), 0.1 1.49 (0.81–2.74), 0.2 0.85 (0.66–1.11), 0.2 1.58 (0.86–2.89), 0.1

African-Americans CC GG vs. CG GG vs. CC GG vs. CG + CC GG + CG vs. CC

1.32 (1.00–1.73), 0.05 1.44 (0.94–2.20), 0.09 1.34 (1.03–1.74), 0.03 1.25 (0.84–1.85), 0.3

1.34 (1.02–1.78), 0.04 1.34 (0.86–2.10), 0.2 1.34 (1.03–1.76), 0.03 1.15 (0.76–1.74), 0.5

a b † ††

Adjusted for age and gender. Adjusted for age, gender, center, BMI, smoking, diabetes and hypertension status, HDL and total cholesterol. Adjusted for age, gender, center, smoking, diabetes and hypertension status. GG genotype serves as the wild-type genotype.

CHD. To our knowledge, there are no previous association studies evaluating the COX-2 G−765C polymorphism in African-Americans, and there are no published longitudinal cohort studies examining this polymorphism and incident disease. Contrary to the findings presented here, Cipollone and colleagues reported that the COX-2 −765C allele was associated with a protective effect when a combined endpoint of myocardial infarction and cerebrovascular ischemia was used in their case–control study [12]. The reason for this discrepancy is unknown. Undetected population substructure is a potential problem for every association study and differences in genetic ancestry can result in false-positive genetic associations, especially in case–control studies. Furthermore, the pathogenesis of atherosclerotic disease is extremely complex and the clinical role of COX-2 may not be unidirectional. The COX-2 G−765C polymorphism resides in a putative Sp1 binding site upstream of the COX gene and influences COX2 gene expression [11,12] and levels of several inflammatory markers [15]. Therefore, COX-2 is widely accepted to be a proinflammatory agonist but, increasing evidence supports that COX-2 also has other roles, including anti-inflammatory and anti-atherogenic properties [16]. COX-2 seems to have a dual role in inflammation by initiating the process of inflammation and then later aiding in its resolution. There has been increased interest in examining the relationship between sequence variations in candidate genes and phenotypes related to atherosclerosis. PGE2 produc-

tion, which is closely associated with monocyte COX-2 mRNA expression, was noted to be increased with increasing number of risk factors, mainly smoking and diabetes [17]. Another study demonstrated lower COX-2 expression among hypercholesterolemic subjects [15]. However, none of these clinical risk factors had significant correlation with the COX2 G−765C polymorphism in our study cohort (results not shown) and the interaction with the clinical risk factor alone could not explain the discrepancies between the findings in the present study and the other reports. At least two SNPs in the promoter of COX-2 impact on the transcriptional activation associated with the COX-2 promoter in vitro. Of these, G−1195A had a greater impact than G−765C mediating susceptibility to esophageal cancer [11]. However, G−1195A is in strong linkage disequilibrium with the G−765C SNP analyzed in the current study (D coefficient = 0.95) [18]. Because of this almost complete genotypic concordance, these two SNPs are expected to yield similar results in the association analyses. Accumulating evidence suggests that the COX-2 G−765C polymorphism has a larger effect on stroke and other cerebrovascular diseases than on CHD. Interestingly, the G−765C polymorphism was found to be closely related to COX-2 gene expression and enzyme activity in carotid plaque macrophages [12]. In addition to hypertension, there is a growing tide supporting a dominant role for inflammation in stroke and cerebrovascular disease [19]. Future studies will be examining the role of novel anti-inflammatory

930

S. Kohsaka et al. / Atherosclerosis 196 (2008) 926–930

agents in the prevention and treatment of stroke. However, the results from clinical trials of COX-2 inhibitors to prevent coronary artery disease and stroke have been discouraging [5,7]. This may be related to variable individual pharmacodynamic response to COX-2 inhibition. Although the precise mechanism(s) by which COX-2 polymorphisms affect the risk of stroke remains unclear, persons with the G−765C polymorphism may constitute a subgroup with different drug responses. It is interesting that we detected a significant association between the COX-2 G−765C polymorphism and stroke in African-Americans, but not in whites in the ARIC study. Stroke is the third leading cause of death and the leading cause of neurologic disability in the United States, with African-Americans having a higher risk compared to whites. In addition, almost a third of strokes occur in younger individuals of the African-American population [9]. Multiple studies have demonstrated that the risk factors for stroke are different between African-Americans and whites [20,21]. In particular, plasma cholesterol levels are less influential and hypertension is more influential in African-Americans compared to whites [22]. Quantitative estimates of the fraction of vascular events arising from risk factor combinations, including genetic polymorphisms, are valuable data for the potential benefits of health interventions and prevention in African-Americans. Taken together, the COX-2 G−765C polymorphism may aid in risk assessment amongst African-Americans when combined with other clinical and genetic risk factors.

Acknowledgements The Atherosclerosis Risk in Communities Study is carried out as a collaborative study supported by National Heart, Lung, and Blood Institute contracts N01-HC-55015, N01-HC-55016, N01-HC-55018, N01-HC-55019, N01-HC55020, N01-HC-55021, and N01-HC-55022. The authors thank the staff and participants of the ARIC study for their important contributions.

References [1] Belton O, Byrne D, Kearney D, Leahy A, Fitzgerald DJ. Cyclooxygenase-1 and -2-dependent prostacyclin formation in patients with atherosclerosis. Circulation 2000;102(8):840–5. [2] Cyrus T, Sung S, Zhao L, Funk CD, Tang S, Pratico D. Effect of low-dose aspirin on vascular inflammation, plaque stability, and atherogenesis in low-density lipoprotein receptor-deficient mice. Circulation 2002;106(10):1282–7. [3] Belton OA, Duffy A, Toomey S, Fitzgerald DJ. Cyclooxygenase isoforms and platelet vessel wall interactions in the apolipoprotein E knockout mouse model of atherosclerosis. Circulation 2003;108(24):3017–23.

[4] Connolly E, Bouchier-Hayes DJ, Kaye E, Leahy A, Fitzgerald D, Belton O. Cyclooxygenase isozyme expression and intimal hyperplasia in a rat model of balloon angioplasty. J Pharmacol Exp Ther 2002;300(2):393–8. [5] Fitzgerald GA, Patrono C. The coxibs, selective inhibitors of cyclooxygenase-2. N Engl J Med 2001;345(6):433–42. [6] Cipollone F, Prontera C, Pini B, et al. Overexpression of functionally coupled cyclooxygenase-2 and prostaglandin E synthase in symptomatic atherosclerotic plaques as a basis of prostaglandin E(2)dependent plaque instability. Circulation 2001;104(8):921–7. [7] Egan KM, Wang M, Fries S, et al. Cyclooxygenases, thromboxane, and atherosclerosis: plaque destabilization by cyclooxygenase-2 inhibition combined with thromboxane receptor antagonism. Circulation 2005;111(3):334–42. [8] Fitzgerald GA. Coxibs and cardiovascular disease. N Engl J Med 2004;351(17):1709–11. [9] Lucas JW, Schiller JS, Benson V. Summary health statistics for U.S. adults: National Health Interview Survey 2001. Vital Health Stat 2004;10(218):1–134. [10] Colaizzo D, Fofi L, Tiscia G, et al. The Cox-2G/C−765 polymorphism may modulate the occurrence of cerebrovascular ischemia. Blood Coagul Fibrinolysis 2006;17(2):93–6. [11] Zhang X, Miao X, Tan W, et al. Identification of functional genetic variants in cyclooxygenase-2 and their association with risk of esophageal cancer. Gastroenterology 2005;129(2):565–76. [12] Cipollone F, Toniato E, Martinotti S, et al. Identification of new elements of plaque stability (INES) study group. A polymorphism in the cyclooxygenase 2 gene as an inherited protective factor against myocardial infarction and stroke. J Am Med Assoc 2004;291(18):2221–8. [13] White A, Folsom A, Chambless L, et al. Community surveillance of coronary heart disease in the atherosclerosis risk in communities (ARIC) study: methods and initial 2 years experience. J Clin Epidemiol 1996;49:223–33. [14] Siedel J, Hagele EO, Ziegenhorn J, Wahlefeld AW. Reagent for the enzymatic determination of serum total cholesterol with improved lipolytic efficiency. Clin Chem 1983;29:1075–85. [15] Orbe J, Beloqui O, Rodriguez JA, Belzunce MS, Roncal C, Paramo JA. Protective effect of the G−765C COX-2 polymorphism on subclinical atherosclerosis and inflammatory markers in asymptomatic subjects with cardiovascular risk factors. Clin Chim Acta 2006;368(1/2):138–43. [16] Antman EM, DeMets D, Loscalzo J. Cyclooxygenase inhibition and cardiovascular risk. Circulation 2005;112:759–70. [17] Beloqui O, P´aramo JA, Orbe J, et al. Monocyte cyclooxygenase-2 overactivity: a new marker of subclinical atherosclerosis in asymptomatic subjects with cardiovascular risk factors? Eur Heart J 2005;26:153–8. [18] Konheim YL, Wolford JK. Association of a promoter variant in the inducible cyclooxygenase-2 gene (PTGS2) with type 2 diabetes mellitus in Pima Indians. Hum Genet 2003;113(5):377–81. [19] Lucas SM, Rothwell NJ, Gibson RM. The role of inflammation in CNS injury and disease. Br J Pharmacol 2006;147(S1):S232–40. [20] Gaines K, Burke G, for the SECORDS Investigators. Ethnic differences in stroke: black-white differences in the United States population. Neuroepidemiology 1995;14:209–39. [21] Abel GA, Sacco RL, Lin I-F, Boden-Albala B, Kargman DE, Paik MC. Race-ethnic variability in etiologic fraction for stroke risk factors: the Northern Manhattan Stroke study. In: Proceedings of the program and abstracts of the 23rd international joint conference on stroke and cerebral circulation. 1998. [22] Moliterno DJ, Jokinen EV, Miserez AR, et al. No association between plasma lipoprotein(a) concentrations and the presence or absence of coronary atherosclerosis in African-Americans. Arterioscler Thromb Vasc Biol 1995;15(7):850–5.