Published online before print November 13, 2006, doi: 10.1161/CIRCULATIONAHA.106.615740
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
(Circulation. 2006;114:2458-2465.)
© 2006 American Heart Association, Inc.
Epidemiology |
Genetic Variation Is Associated With C-Reactive Protein Levels in the Third National Health and Nutrition Examination Survey
From the Department of Genome Sciences, University of Washington, Seattle (D.C.C., J.D.S., C.S., M.W., L.W., M.J.R., D.A.N.), and Harris Corporation/National Center for Health Statistics, CDC, Hyattsville, Md (C.L.S., X.Q.). Dr Qin is now with Business Computer Applications, Inc, Survey Operation Section/National Center for Chronic Disease Prevention and Health Promotion, Division of Adult and Community Health, Behavioral Surveillance Branch, Atlanta, Ga. Dr Crawford is now at Vanderbilt University, Center for Human Genetics Research, Nashville, Tenn.
Correspondence to Dana C. Crawford, Vanderbilt University, Center for Human Genetics Research, 515B Light Hall, 2215 Garland Ave, Nashville, TN 37232. E-mail crawford{at}chgr.mc.vanderbilt.edu
Received January 25, 2006; revision received August 30, 2006; accepted October 5, 2006.
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
Background— Increased serum C-reactive protein (CRP) is an independent risk factor for cardiovascular disease. Previous studies have suggested that genetic variation within the CRP gene is associated with serum CRP.
Methods and Results— We genotyped CRP genetic variants in 7159 individuals from the Third National Health and Nutrition Examination Survey (NHANES III). NHANES III is American population-based sample linked to hundreds of phenotypes, including CRP; however, the CRP assay used in this survey is not a high-sensitivity CRP assay, and 65% of participants (n=4679) had CRP measurements at or below the level of detection. Despite these limitations, we identified specific CRP single-nucleotide polymorphisms (SNPs) and haplotypes associated with serum CRP levels in the general population. Two variants were associated with increased levels of serum CRP: SNP rs3093058 (in linkage disequilibrium with a CRP promoter SNP rs3093062) in the non-Hispanic black sample and the triallelic promoter SNP rs3091244 in the non-Hispanic black and Mexican American samples. Two other SNPs were associated with decreased levels of serum CRP in either the non-Hispanic black (rs1205 and rs2808630) or Mexican American (rs1205) samples. Three haplotypes inferred from 7 SNPs (ATTGCGA, TTAGCGA, and AAAGAGA) were associated (P
0.01) with increased levels of serum CRP in the non-Hispanic black sample; 2 haplotypes (ATTGCGA and AAAGCGA) were associated (P<0.05) with increased levels in the Mexican American sample; and 1 haplotype (AAAGCGA) was associated (P<0.03) with increased levels in the non-Hispanic white sample. Post hoc analysis suggests that the AA genotype of the triallelic SNP rs3091244, after adjustment for covariates, was associated with prevalent coronary heart disease in the non-Hispanic white population sample.
Conclusions— Genetic variation within CRP is associated with serum CRP levels in the general population and may be associated with prevalent coronary heart disease.
Key Words: C-reactive protein • epidemiology • genes • genetics • inflammation • population
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Coronary heart disease (CHD) is the leading cause of death in the United States1 and is a significant source of morbidity and economic burden in the general population. Genes that affect lipid concentrations or are involved in inflammation have long been suspected to be involved in the development of CHD.2 Until recently,3,4 candidate gene association studies have mostly failed to identify a consistent relationship between genetic variation and the development of CHD.5,6 The lack of a consistent association could be accounted for by several factors, including genetic heterogeneity, unaccounted environmental factors, and inadequate sample size. In addition, most early studies examined a single genetic polymorphism within a candidate gene, a design that assumes that the polymorphism under study is directly involved in the expression of the phenotype or is in linkage disequilibrium (LD) with the polymorphism responsible for a proportion of the phenotype.7
Clinical Perspective p 2465
To overcome the limitations afflicting most genetic association studies, we applied candidate gene variation data to 
7000 DNAs available for study from the Third National Health and Nutrition Examination Survey (NHANES III). NHANES is a series of cross-sectional surveys designed to provide national statistics on the health and nutritional status of the US noninstitutionalized population. NHANES III, conducted between 1988 and 1994, included DNA collection in the second phase of the survey (1991–1994). This study is now one of the largest American population-based collections of DNAs linked to an extensive battery of phenotypic and demographic data available. Thus, DNA from NHANES III, when coupled with high-throughput sequencing and genotyping technologies, is an ideal and powerful resource for present-day genetic association studies.
We focus our present investigation on the factors that significantly influence serum C-reactive protein (CRP) levels, both genetic and environmental, in NHANES III. CRP is an independent risk factor for CHD,8–10 and recent evidence suggests that lowering CRP levels reduces the rate of atherosclerosis progression11 and decreases the risk of recurrent cardiovascular events.12 Given its growing importance in potentially predicting and/or preventing cardiovascular events, identifying the genetic variants that significantly affect CRP levels is an important step in understanding a possible target for CHD prevention. By combining NHANES III with strategies to represent common genetic variation within the CRP gene,13 we can report here the significant association of genetic variants within CRP and CRP levels in NHANES III. We also estimate the proportion of CRP variance explained by CRP single-nucleotide polymorphisms (SNPs) and haplotypes in a population-based setting and provide evidence that at least 1 SNP (rs3091244) may be associated with prevalent CHD.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Participants
NHANES III is a nationally representative sample of noninstitutionalized individuals surveyed between 1988 and 1994 by the National Center for Health Statistics (NCHS) at the Centers for Disease Control and Prevention (CDC). Ascertainment and survey design have been previously described.14,15 Briefly, NHANES III is a complex, multistaged survey that oversampled minorities (non-Hispanic blacks and Mexican Americans), children, and the elderly. Sample weights are provided for each individual included in the survey to correct for nonresponse and to correct for the fact that each individual was not selected for the survey with equal probability. Participants were asked to complete a household interview and a physical examination in the Mobile Examination Center. If the participant could not visit the Mobile Examination Center, a special home physical examination was arranged.
DNA from the genetic component of the NHANES survey was generated using cell lines created from blood samples of participants >12 years of age during NHANES III phase 2 (1991–1994). The total number of participants from NHANES III phase 2 was 16 530. The sample weights were recalculated using previously described methods16 for the 7159 participants for whom DNA was available to avoid nonresponse bias. The present study was approved by the CDC Ethics Review Board and the University of Washington Human Subjects Committee.
Laboratory Measures
CRP was measured in NHANES III participants using a modification of the Behring latex-enhanced CRP assay (Behring Diagnostics, Westwood, Mass) as previously described.17,18 Glucose was measured using a modified hexokinase enzymatic method. Total cholesterol and triglycerides were measured enzymatically. High-density lipoprotein was measured after other lipoproteins were precipitated using a polyanion/divalent cation. Glycosylated hemoglobin (HbA1C) was measured using ion exchange chromatography. Details of the laboratory protocols for each of these measures have been published19 and can be found on the CDC National Center for Health Statistics website at http://www.cdc.gov/nchs/data/nhanes/nhanes3/ cdrom/NCHS/MANUALS/LABMAN.PDF.
Sequencing, tagSNPs, and Genotyping
As previously reported, the CRP gene (6.8 kb), including all exons, introns, 1.7 kb upstream, and 2.8 kb downstream, was resequenced in DNA samples available through the Coriell Cell Repositories (http://locus.umdnj.edu/ccr/) by the SeattleSNPs Program for Genomic Applications (National Heart, Lung, and Blood Institute) for genetic variation discovery in 23 and 24 Americans of European and African ancestry, respectively.20,21 Overall, we detected 31 SNPs, with 29 in the African-American sample and 12 in the European-American sample.20,21 All SNPs and genotypes from this resequencing effort are available on the SeattleSNPs website (http://pga.gs.washington.edu) and dbSNP (http://www.ncbi.nlm.nih.gov/projects/SNP/).
We used the LDSelect software at default settings (r2>0.64) to select tagSNPs from all common variation within CRP (>10% minor allele frequency).13 Because LD varies across different populations,22,23 tagSNPs were determined for the European- and African-descent samples separately, and the results were combined to determine the optimum set of tagSNPs that represent both population samples. Recent evidence suggests that this optimum set of tagSNPs also represents common CRP variation in a Mexican sample.21 A total of 6 tagSNPs were selected for genotyping in CRP: rs3093058, rs3091244, rs1417938, rs3093066, rs1205, and rs2808630. In addition to the 6 tagSNPs, we also genotyped rs1800947, a synonymous SNP (Leu184Leu) previously reported to be associated with CRP levels,24,25 and 2 newly described nonsynonymous SNPs: Pro133Lys (rs34200896) and Gly166Glu (rs34340208).21
All 9 SNPs for CRP were genotyped in 7159 NHANES III samples by a single laboratory. The location of all SNPs and the allele frequencies are given in Table I of the online Data Supplement. Seven SNPs were genotyped using TaqMan Assays by Design (Applied Biosystems, Foster City, Calif) under standard conditions,26 and 1 SNP (rs1800947) was genotyped using Epoch MGB Eclipse Probe Systems (Epoch Biosciences, Bothel, Wash). The triallelic SNP rs3091244 was genotyped as previously described.20 Individuals determined to be heterozygous by TaqMan Assays by Design for the rare SNP Gly166Glu were confirmed by resequencing. The overall average genotyping success rate for the NHANES III samples was 95%. For quality control and quality assurance, two 96-well plates were provided that contained duplicate samples and no template controls. The genotyping error rate based on the blind quality control and quality assurance plates was <1%, and there was no evidence of contamination. All genotypes for NHANES III are deposited in a database maintained by the NCHS (http://www.cdc.gov/nchs/about/major/nhanes/nh3data_genetic.htm).
Statistical Analysis
All analyses were completed using SAS/Genetics 9.1 (SAS Institute, Cary, NC) and SUDAAN 9.0 (Research Triangle Institute, Research Triangle Park, NC). Haplotypes were inferred from tagSNPs using the expectation-maximization (EM) algorithm provided by SAS/Genetics 9.1, which also provides the probabilities for each individual’s inferred haplotype pair for subsequent regression analyses. Common haplotypes (>5% frequency) were determined from most likely pair of haplotypes inferred by the EM algorithm and are listed in supplementary Table II by race/ethnicity.
When appropriate, variables were transformed using the natural log to create a more normal distribution. Univariate and multiple regressions were performed in which ln(CRP) was the dependent variable. NHANES III participants with CRP levels below the limit of detection were coded as having a CRP level of 0.21 mg/dL. Most covariates included in the model were previously reported in the literature as associated with CRP levels.17,18,20,27–30 The following variables are treated as continuous variables in the present study: age, total cholesterol (mg/dL), high-density lipoprotein (mg/dL), triglycerides (mg/dL), glucose (mg/dL), glycosylated hemoglobin, systolic blood pressure (mm Hg), diastolic blood pressure (mm Hg), and body mass index (kg/m2). Total cholesterol serves as a proxy for low-density lipoprotein because it was not available for all participants. In addition, glycosylated hemoglobin is used rather than the glucose tolerance test because the latter was performed only on the morning samples of participants who fasted.31 Smoking status was coded as a categorical variable in which current smokers were individuals who answered "yes" to the question, "Do you smoke cigarettes now?" or who had cotinine levels >15 ng/mL. Alcohol consumption was coded as a categorical variable (1=<35 drinks per week, 2=
35 drinks per week. Prevalent CHD was determined by self-reported myocardial infarction or angina pectoris and treated as a categorical variable. Education was defined as having 0 to 8, 9 to 12, or >12 years of education and was treated as an ordinal variable. Self-described race/ethnicity was classified as non-Hispanic white, non-Hispanic black, Mexican American, and other.
For all regression analyses, pregnant women were excluded (n=130) because pregnancy elevates CRP levels.32,33 For the non-CRP model, all variables listed above were allowed to enter the model for the multivariate regression. We also included interaction terms (gender-by-triglycerides and gender-by–diastolic blood pressure). We excluded glucose because this variable had a substantial amount of missing data in NHANES III because of fasting requirements. Systolic blood pressure was excluded because this measure, although significant, did not contribute substantially to the model. Probability values also were calculated using the Satterthwaite method. The Taylor series was implemented for the variance estimation method, and standard errors were calculated using a robust method.34
Genetic covariates were added to the model in 3 different ways: dominant (genotypes coded as 0 or 1, where 0 is homozygous for the major allele and 1 is heterozygous plus homozygous for the minor allele), general genotype (all genotypes), or haplotypes (7 tagSNPs phased). Models including SNPs as covariates are reported for each individual tagSNP (except for rs34200896 and rs34340208, the latter of which had <5 heterozygous individuals identified) adjusted for non-CRP covariates. For the model including haplotypes, we used the probabilities of correct haplotype inference provided by the EM algorithm to adjust for haplotype uncertainty in the model. Therefore, each haplotype was considered a continuous variable to account for the uncertainty in haplotype inference by the EM algorithm. Because haplotypes inferred from the same tagSNPs were found to be associated with CRP levels in a previous report,20 we chose to include the same major 8 haplotypes (H1 through H8) in the models in an attempt to replicate and further define these associations using our data set. To test for associations between prevalent CHD and CRP tagSNPs and haplotypes, logistic regressions were performed after adjustment for covariates, including CRP levels.
All regression models, including genetic covariates, were stratified by race/ethnicity for 2 reasons. First, tests for effect modification between the race/ethnicity and CRP SNP covariates were significant for all tagSNPs except rs1417938 (P=0.055). Furthermore, race/ethnicity is significant in all models when included as a covariate in the multivariate regressions (data not shown). Second, stratification by race/ethnicity takes into account differing population demography and history that ultimately affect the structure of LD across the human genome.22,23 Because SNPs were chosen for genotyping on the basis of r2, any resulting association must be interpreted within the context of a population with a population history or demography similar to that from which the original tagSNP collection was derived. When population structure is ignored, the designation "tagSNP" for the genotyped SNP is no longer valid, and the ability to trace the association to the putative causal SNP is lost.
C.L. Sanders has had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. The authors had full access to the output data and take full responsibility for their integrity. All authors have read and agree to the manuscript as written.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
The NHANES III phase 2 study participants included in the genetic portion of the study are described in Table 1. Approximately 50% of the study participants were female, and the overall mean age was 40.8 years. The mean CRP level for NHANES III was 0.40 mg/dL (SE, 0.01 mg/dL). The majority of CRP levels measured (4679) were below or at the level of detection. We found that 3.38% of the participants had self-reported myocardial infarction or angina pectoris.
|
Univariate regressions were performed for each variable listed in Table 1. All covariates except alcohol consumption were significantly associated with serum CRP levels (P<0.01). Multiple regressions were then performed to determine the final model for non-CRP factors that affect CRP levels in NHANES III. All variables listed in Table 1 except for glucose, alcohol consumption, and systolic blood pressure remained in the final non-CRP model (supplementary Table III). On the basis of previous findings,20 first-order interactions (gender-by-triglycerides and gender-by–diastolic blood pressure) also were included in the final model. Collectively, non-CRP factors account for less than a quarter of the total variance in CRP levels (supplementary Table III; R2=0.1915).
Regressions were then performed for each SNP defined as an independent variable and CRP levels as a dependent variable with adjustment for non-CRP variables. All models were stratified by race/ethnicity (see Methods). Results for the dominant model are given in Supplementary Tables IV and V. For the general genotype model, all SNPs except rs1417938 were significantly associated with serum CRP levels in at least 1 of the 3 population samples after adjustment for non-CRP covariates (Table 2 and supplementary Table V).
|
Two notable associations were observed for SNP rs3093058 and rs3091244, both of which were associated with increased levels of CRP. Specifically, for rs3093058, the AT genotype was significantly associated (P<0.0001) with increased levels of CRP compared with the AA referent genotype in the non-Hispanic black sample. For the triallelic rs3091244, several genotypes were significantly associated with increased levels of serum CRP levels in the non-Hispanic black (AA, TT, AT, CT; P<0.01) and Mexican American (TT, CT; P<0.05) samples compared with the referent CC genotype (Table 2).
We also identified 2 tagSNPs, rs1205 and rs2808630, that were significantly associated with decreased CRP levels in the non-Hispanic black and Mexican American samples (P<0.05; Table 2). The significant associations for tagSNPs rs1800947 and rs3093066 are based on very small sample sizes (n<5) and therefore may be reliable. Mean CRP levels by genotype for all 7 tagSNPs are given in supplementary Table VI.
Because the single SNP regressions demonstrated that multiple sites within the CRP gene significantly affect CRP levels, haplotypes were inferred to capture possible allelic associations. A total of 8 haplotypes (H1 through H8) were included in the model, with the most common haplotype across the 3 populations (ACAGCGA; H2) as the referent. Rare haplotypes were coded in aggregate as H9. The resulting models suggest that 4 haplotypes were significantly associated with increased serum CRP levels after adjustment for non-CRP variables: haplotypes ATTGCGA (H5), TTAGCGA (H6), and AAAGAGA (H8) in the non-Hispanic black sample; haplotypes ATTGCGA (H5) and AAAGCGA (H7) in the Mexican American sample; and haplotype AAAGCGA (H7) in the non-Hispanic white sample (Table 3 and supplementary Table VII). Compared with the non-CRP models for the non-Hispanic white, non-Hispanic black, and Mexican American samples (R2 = 0.1858, 0.2039, and 0.1923, respectively), models with the addition of haplotypes experienced only a small increase in the proportion of variation of CRP levels explained (R2=0.1889, 0.2324, and 0.2039, respectively). For comparison, models including haplotypes but excluding body mass index as a covariate (R2=0.1519, 0.1752, and 0.1568, respectively) suggest that CRP genetic variants contribute less to the variance of observed CRP levels compared with non-CRP variables in all 3 population samples studied here.
|
Given that increased CRP levels are associated with an increased risk for CHD8–10 and given that we have identified CRP SNPs and haplotypes associated with CRP levels, we performed logistic regressions in a post hoc analysis to determine whether CRP SNPs or haplotypes are associated with prevalent CHD in NHANES III. In multivariate analyses, the triallelic SNP rs3091244 had the most striking association with prevalent CHD. After adjustment for non-CRP covariates, the AA genotype was associated with prevalent CHD (odds ratio [OR], 30.11; 95% confidence interval [CI], 3.26 to 278.08) compared with the CC referent genotype in the non-Hispanic white population sample (P=0.0043; supplementary Table VIII). Surprisingly, the association of the AT genotype with prevalent CHD was in the opposite direction (OR, 0.10; 95% CI, 0.02 to 0.57) in the same population sample (P=0.0115; supplementary Table VIII). These associations, however, should be interpreted with caution because the number of cases with the AA or AT genotype (each n<5) compared with the referent CC genotype (n=52) for rs3091244 in the non-Hispanic white population sample was very small. Finally, we identified only 1 other association among CRP tagSNP with prevalent CHD: rs1800947 in the Mexican American population sample (P=0.0323; supplementary Table VIII). Compared with the referent GG genotype, the CG genotype at rs1800947 was associated with decreased prevalent CHD (OR, 0.11; 95% CI, 0.01 to 0.81). No haplotypes were significantly associated with prevalent CHD (data not shown).
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Consistent with previous reports,20,35–39 we demonstrate that both specific CRP SNPs and haplotypes are associated with serum CRP levels in the general population. Specifically, tagSNP rs3093058 was associated with increased levels of CRP in the non-Hispanic black population, whereas 2 SNPs were associated with decreased levels of CRP in the non-Hispanic black (rs1205 and rs2808630) and Mexican American populations (rs1205). Genotypes for the triallelic SNP rs3091244 were associated with increased levels of serum CRP in the non-Hispanic black and Mexican American samples and were associated with prevalent CHD in the non-Hispanic white population sample. SNP rs1800947 was not associated with serum CRP but was associated with decreased prevalent CHD in the Mexican American population sample.
The association between tagSNPs and CRP levels could represent a direct causal relationship or an indirect relationship that serves as a proxy for the true causal SNP. The latter situation is more likely for tagSNP rs3093058; the former situation is more likely for the triallelic rs3091244. SNP rs3093058 is in LD with other CRP SNPs (rs3093061, rs3093062, rs3093068, and rs3093076). SNP rs3093062 and rs3091244 lie within an evolutionarily conserved region of the CRP promoter and are predicted to alter a transcription factor E box binding element.20,40 Furthermore, in vitro assays have demonstrated the functional significance of rs3093062 and rs3091244 in the promoter region of CRP.20,40
The functional significance of the association with the other SNPs in CRP (rs1800947, rs1205, and rs2808630) is more difficult to understand. SNP rs1205 tags only itself and is located distal to the 3' untranslated region of CRP and in the MLT1K repeat. SNP rs2808630 also is located distal to the 3' untranslated region and is in LD with SNP rs2794521 in the non-Hispanic white and Mexican American samples. None of these tagSNPs (or their associated SNPs) occur in an evolutionarily conserved region (http://ecrbrowswer.dcode.org); thus, their functional significance remains obscure.
In contrast to the above SNPs, the synonymous SNP rs1800947 lies in an exon (Leu184Leu in exon 2) and, as expected, in an evolutionarily conserved region of CRP. Unlike the rare nonsynonymous CRP SNPs predicted to alter protein function,21,35 SNP rs1800947 is not predicted to alter protein function. Many reports in the literature,20,24,25,35,39,41 but not all,37,42–44 suggest that the C allele of SNP rs1800947 is associated with decreased CRP levels in European-descent populations. We did not identify an association with the C allele at rs1800947 and decreased serum CRP; however, we did identify an association with decreased prevalent CHD and the CG genotype in the Mexican American population. SNP rs1800947 is not in LD with any other CRP SNP. Furthermore, examination of extended LD using the Perlegen data set45 for European Americans identified only 1 intergenic SNP (rs12049404) in LD with SNP rs1800947 (r2>0.40) within a 100-kb region proximal and distal to the CRP locus (data not shown). Very few examples of functional synonymous SNPs exist in the literature46,47; however, given that no other SNP is in strong LD to rs1800947 in at least the European-descent population, it may be that this presumably silent SNP is functional.
Four haplotypes also were significantly associated with CRP levels: ATTGCGA, TTAGCGA, AAAGCGA, and AAAGAGA. These haplotypes are identical to haplotypes H5, H6, H7, and H8 described by Carlson et al,20 who reported that all were associated with increased CRP levels in the Coronary Artery Risk Development In young Adults (CARDIA) study. Here, all 4 haplotypes also are associated with increased levels of serum CRP (Table 3). Compared with a common haplotype observed in all population samples (the referent ACAGCGA), all 4 haplotypes differ by 2 SNPs: rs3091244 (A and T alleles) and rs1205 (G allele). Three of these haplotypes also differed from the referent haplotype by 3 other SNPs: rs1417938 (H5), rs3093058 (H6), and rs3093066 (H8). It is interesting to note that the AA genotype for rs3091244 is not associated with increased levels of CRP in the non-Hispanic white sample (P=0.0770; Table 2), whereas H7 containing the A allele at rs3091244 is associated with increased CRP levels in the same population (Table 3). Although associations between CRP variants and CRP levels could be identified by genotyping rs3091244 alone, these results underscore the importance of modeling both SNP and haplotype associations because neither approach alone is sufficient to identify as many of the genetic variants as possible that contribute significantly to the variance of the phenotype for different populations.
Although the present study has been successful in identifying genetic variants that contribute to CRP levels in the general population, the overall contribution of these variants is small compared with already known contributors such as body mass index. The variance for CRP levels explained by CRP haplotypes is <5% for all population studies here. These estimates are similar to estimates from other cohorts for CRP haplotypes.20,35,37,38 In comparison, the estimates for the contribution of CRP haplotypes are far lower than the recent estimates for VKORC1 haplotypes, which explain 25% of the variance in warfarin dosing.48
Despite the fact that the genetic contributions measured here are very small, the possibility of larger genetic contributions yet to be discovered is still real because the full CRP genetic model leaves most of the variance unexplained (80%). In addition, heritability estimates suggest that 20% to 60% of the variance in basal CRP levels can be attributable to genetics.43,49–51 CRP is 1 component of a complex acute-phase response that involves several genes, including its transcriptional regulators IL-6 and IL1B and the protein kinase C pathway.52,53 Preliminary studies suggest that polymorphisms in IL650,54 and IL1B55–57 are associated with CRP levels, but the associations are somewhat inconsistent,43,56,58 and the size of the contribution is unclear. Nevertheless, it may be that genes in these CRP regulatory pathways have a greater impact on CRP levels in the general population compared with genetic variants within CRP, warranting further study of CRP-related pathways in large population-based surveys such as NHANES III.
The study presented here has several strengths and limitations. The major strength of the present study is the use of a large, population-based survey and its uniformly collected laboratory and examination variables. This comprehensive data set allows examination of both genetic and environmental variables in several population samples that contribute to the phenotype of interest.
A limitation is that the measure of CRP is not the high-sensitivity test now available for investigation. The test used for NHANES III has a lower limit of 0.21 mg/dL, which makes the distribution for NHANES III truncated at the lower limit of the measure. Despite this limitation in phenotype resolution, we were able to confirm that CRP genetic variants are associated with serum CRP levels. Another limitation is that CRP levels were measured only once in this cross-sectional study and could possibly have changed if a second measurement were taken. However, on the basis of recent results from CARDIA that demonstrated that CRP levels are stable when compared at different time points several years apart,20 we do not anticipate that a second measurement of CRP levels would be significantly different than the original measurement in NHANES III.
Finally, only prevalent CHD could be measured in this cross-sectional study, and the total number of reported events among NHANES III participants was low (n=250), decreasing our power to identify meaningful associations. Surprisingly, we were able to identify a significant association between the triallelic SNP rs3091244 and prevalent CHD in the non-Hispanic white population sample after adjusting for covariates in a post hoc analysis. To the best of our knowledge, this is the first report suggesting an association between prevalent CHD and rs3091244. Other reports examining the relationship between CRP SNPs and prevalent CHD or related phenotypes either did not genotype rs3091244 directly39,59,60 or did genotype rs3091244 but did not identify an association.37,38 Replication of this finding, particularly in cohorts in whom incident CHD can be measured, is necessary to confirm the rs3091244 association and its potential usefulness in predicting new CHD events.
|
Acknowledgments |
|---|
We thank Alex Reiner, MD, MS (Cardiovascular Health Research Unit, University of Washington); Tushar Bhangale, MBBS, MTech (Departments of Bioengineering and Genome Sciences, University of Washington); and Lisa Mirel, MS (National Center for Health Statistics, CDC), for critical reading of this manuscript and statistical advice.
Sources of Funding
This work was funded by grants from the National Heart, Lung, and Blood Institute’s SeattleSNPs Program for Genomic Applications (U01 HL66682) and the National Institute of Environmental Health Science’s Environmental Genome Project (N01 ES15478).
Disclosures
None.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
1. Hoyert DL, Arias E, Smith BL, Murphy SL, Kochanek KD. Deaths: final data for 1999. Natl Vital Stat Rep. 2001; 49: 1–113.[Medline] [Order article via Infotrieve]
2. Ellsworth DL, Sholinsky P, Jaquish C, Fabsitz RR, Manolio TA. Coronary heart disease: at the interface of molecular genetics and preventive medicine. Am J Prev Med. 1999; 16: 122–133.[CrossRef][Medline] [Order article via Infotrieve]
3. Ozaki K, Ohnishi Y, Iida A, Sekine A, Yamada R, Tsunoda T, Sato H, Sato H, Hori M, Nakamura Y, Tanaka T. Functional SNPs in the lymphotoxin-alpha gene that are associated with susceptibility to myocardial infarction. Nat Genet. 2002; 32: 650–654.[CrossRef][Medline] [Order article via Infotrieve]
4. Helgadottir A, Manolescu A, Thorleifsson G, Gretarsdottir S, Jonsdottir H, Thorsteinsdottir U, Samani NJ, Gudmundsson G, Grant SF, Thorgeirsson G, Sveinbjornsdottir S, Valdimarsson EM, Matthiasson SE, Johannsson H, Gudmundsdottir O, Gurney ME, Sainz J, Thorhallsdottir M, Andresdottir M, Frigge ML, Topol EJ, Kong A, Gudnason V, Hakonarson H, Gulcher JR, Stefansson K. The gene encoding 5-lipozygenase activating protein confers risk of myocardial infarction and stroke. Nat Genet. 2004; 36: 233–239.[CrossRef][Medline] [Order article via Infotrieve]
5. Ioannidis JP, Ntzani EE, Trikalinos TA, Contopoulos-Ioannidis DG. Replication validity of genetic associations studies. Nat Genet. 2001; 29: 306–309.[CrossRef][Medline] [Order article via Infotrieve]
6. Lusis AJ, Mar R, Pajukanta P. Genetics of atherosclerosis. Annu Rev Genomics Hum Genet. 2004; 5: 189–218.[CrossRef][Medline] [Order article via Infotrieve]
7. Cardon LR, Bell JI. Association study designs for complex diseases. Nat Rev Genet. 2001; 2: 91–99.[CrossRef][Medline] [Order article via Infotrieve]
8. Ridker PM, Stampfer MJ, Rifai N. Novel risk factors for systemic atherosclerosis: a comparison of C-reactive protein, fibrinogen, homocysteine, lipoprotein(a), and standard cholesterol screening as predictors of peripheral arterial disease. JAMA. 2001; 285: 2481–2485.
9. Danesh J, Whincup P, Walker M, Lennon L, Thomson A, Appleby P, Gallimore JR, Pepys MB. Low grade inflammation and coronary heart disease: prospective study and updated meta-analysis. BMJ. 2000; 321: 204–207.
10. Danesh J, Wheeler JG, Hirschfield GM, Eda S, Eiriksdottir G, Rumley A, Lowe GDO, Pepys MB, Gudnason V. C-reactive protein and other circulating markers of inflammation in the prediction of coronary heart disease. N Engl J Med. 2004; 350: 1387–1397.
11. Nissen SE, Tuzcu EM, Schoenhagen P, Crowe T, Sasiela WJ, Tsai J, Orazem J, Magorien RD, O’Shaughnessy C, Ganz P, for the Reversal of Atherosclerosis With Aggressive Lipid Lowering (REVERSAL) Investigators. Statin therapy, LDL cholesterol, C-reactive protein, and coronary artery disease. N Engl J Med. 2005; 352: 29–38.
12. Ridker PM, Cannon CP, Morrow D, Rifai N, Rose LM, McCabe CH, Pfeffer MA, Braunwald E, for the Pravastatin or Atorvastatin Evaluation and Infection Therapy-Thrombolysis in Myocardial Infarction 22 (PROVE IT-TIMI 22) Investigators. C-reactive protein levels and outcomes after statin therapy. N Engl J Med. 2005; 352: 20–28.
13. Carlson CS, Eberle MA, Rieder MJ, Yi Q, Kruglyak L, Nickerson DA. Selecting a maximally informative set of single-nucleotide polymorphisms for association analyses using linkage disequilibrium. Am J Hum Genet. 2004; 74: 106–120.[CrossRef][Medline] [Order article via Infotrieve]
14. Centers for Disease Control and Prevention. Plan and Operation of the Third National Health and Nutrition Examination Survey, 1988–94. Bethesda, Md: Centers for Disease Control and Prevention; 1994.
15. Third National Health and Nutrition Examination Survey, 1988–94, Plan and Operations Procedures Manuals [CD-ROM]. Hyattsville, Md.: US Department of Health and Human Services (DHHS), National Center for Health Statistics, Centers for Disease Control and Prevention; 1996.
16. Lohr SL. Sampling: Design and Analysis. 1st ed. Pacific Grove, Calif: Duxbury Press; 1999.
17. Visser M, Bouter LM, McQuillan GM, Wener MH, Harris TB. Elevated C-reactive protein levels in overweight and obese adults. JAMA. 1999; 282: 2131–2135.
18. Wong ND, Pio J, Valencia R, Thakal G. Distribution of C-reactive protein and its relation to risk factors and coronary heart disease risk estimation in the National Health and Nutrition Examination Survey (NHANES) III. Prev Cardiol. 2001; 4: 109–114.[Medline] [Order article via Infotrieve]
19. Gunter EW, Lewis BG, Koncikowski SM. Laboratory Procedures Used for the Third National Health and Nutrition Examination Survey (NHANES III), 1988–1994. Hyattsville, Md: US Department of Health and Human Services, Centers for Disease Control and Prevention; 1996.
20. Carlson CS, Aldred SF, Lee PK, Tracy RP, Schwartz SM, Rieder M, Liu K, Williams OD, Iribarren C, Lewis EC, Fornage M, Boerwinkle E, Gross M, Jaquish C, Nickerson DA, Myers RM, Siscovick DS, Reiner AP. Polymorphisms within the C-reactive protein (CRP) promoter region are associated with plasma CRP levels. Am J Hum Genet. 2005; 77: 64–77.[CrossRef][Medline] [Order article via Infotrieve]
21. Crawford DC, Yi Q, Smith JD, Shephard C, Wong M, Witrak L, Livingston RJ, Rieder MJ, Nickerson DA. Allelic spectrum of natural variation in CRP. Hum Genet. 2006; 119: 496–504.[CrossRef][Medline] [Order article via Infotrieve]
22. Wall JD, Pritchard JK. Haplotype blocks and linkage disequilibrium in the human genome. Nat Rev Genet. 2003; 4: 587–597.[CrossRef][Medline] [Order article via Infotrieve]
23. Evans DM, Cardon LR. A comparison of linkage disequilibrium patterns and estimated population recombination rates across multiple populations. Am J Hum Genet. 2005; 76: 681–687.[CrossRef][Medline] [Order article via Infotrieve]
24. Zee RY, Ridker PM. Polymorphism in the human C-reactive protein (CRP) gene, plasma concentrations of CRP, and the risk of future arterial thrombosis. Atherosclerosis. 2002; 162: 217–219.[CrossRef][Medline] [Order article via Infotrieve]
25. Russell AI, Cunninghame Graham DS, Shepherd C, Roberton CA, Whittaker J, Meeks J, Powell RJ, Isenberg DA, Walport MJ, Vyse TJ. Polymorphism at the C-reactive protein locus influences gene expression and predisposes to systemic lupus erythematosus. Hum Mol Genet. 2004; 13: 134–147.
26. De La Vega FM, Lazaruk KD, Rhodes MD, Wenz MH. Assessment of two flexible and compatible SNP genotyping platforms: TaqMan SNP genotyping assays and the SNPlex genotyping system. Mutat Res. 2005; 573: 111–135.[Medline] [Order article via Infotrieve]
27. Wener MH, Daum PR, McQuillan GM. The influence of age, sex, and race on the upper reference limit of serum C-reactive protein concentration. J Rheumatol. 2000; 27: 2351–2359.[Medline] [Order article via Infotrieve]
28. Ford ES, Giles WH, Myers GL, Mannino DM. Population distribution of high-sensitivity C-reactive protein among US men: findings from National Health and Nutrition Examination Survey 1999–2000. Clin Chem. 2003; 49: 686–690.
29. Sesso HD, Buring JE, Rifai N, Blake GJ, Gaziano JM, Ridker PM. C-reactive protein and the risk of developing hypertension. JAMA. 2003; 290: 2945–2951.
30. Miller M, Zhan M, Havas S. High attributable risk of elevated C-reactive protein level to conventional coronary heart disease risk factors. Arch Intern Med. 2005; 165: 2063–2068.
31. Davidson MB, Schriger DL, Peters AL, Lorber B. Relationship between fasting plasma glucose and glycosylated hemoglobin: potential for false-positive diagnosis of type 2 diabetes using new diagnostic criteria. JAMA. 1999; 281: 1203–1210.
32. Watts DH, Krohn MA, Wener MH, Eschenbach DA. C-reactive protein in normal pregnancy. Obstet Gynecol. 1991; 77: 176–180.[Medline] [Order article via Infotrieve]
33. Wood WG, Ludemann J, Mitusch R, Heinrich J, Maass R, Frick U. Evaluation of a sensitive immunoluminometric assay for the determination of C-reactive protein (CRP) in serum and plasma and the establishment of reference ranges for different groups of subjects. Clin Lab. 2000; 46: 131–140.[Medline] [Order article via Infotrieve]
34. Binder DA. On the variances of asymptotically normal estimators from complex surveys. Int Stat Rev. 1983; 51: 279–292.
35. Miller DT, Zee RYL, Danik JS, Kozlowski P, Chasman DI, Lazarus R, Cook NR, Ridker PM, Kwiatkowski DJ. Association of common CRP gene variants with CRP levels and cardiovascular events. Ann Hum Genet. 2005; 69: 623–638.[CrossRef][Medline] [Order article via Infotrieve]
36. Timpson NJ, Lawlor DA, Harbord RM, Gaunt TR, Day IN, Palmer LJ, Hattersley AT, Ebrahim S, Lowe GD, Rumley A, Smith GD. C-reactive protein and its role in metabolic syndrome: mendelian randomisation study. Lancet. 2005; 366: 1954–1959.[CrossRef][Medline] [Order article via Infotrieve]
37. Kathiresan S, Larson MG, Vasan RS, Guo C-Y, Gona P, Keaney JF, Wilson PWF, Newton-Cheh C, Musone SL, Camargo AL, Drake JA, Levy D, O’Donnell CJ, Hirschhorn JN, Benjamin EJ. Contribution of clinical correlates in 13 C-reactive protein gene polymorphisms to interindividual variability in serum C-reactive protein level. Circulation. 2006; 113: 1415–1423.
38. Wang Q, Hunt SC, Xu Q, Province MA, Eckfeldt JH, Pankow JS, Song Q. Association study of CRP gene polymorphisms with serum CRP level and cardiovascular risk in the NHLBI Family Heart Study. Am J Physiol Heart Circ Physiol. May 26, 2006. DOI: 10.1152/ajpheart.01164.2005. Available at: http://ajpheart.physiology.org.
39. Lange LA, Carlson C, Lange EM, Zeng D, Lin D, Hindorff L, Cushman M, Walston J, Bis J, Kuller L, Psaty B, Tracy R, Nickerson D, Durda P, Reiner A. Polymorphisms in the CRP gene are associated with plasma C-reactive protein levels and cardiovascular events: the Cardiovascular Health Study. JAMA. In press.
40. Szalai AJ, Wu J, Lange EM, McCrory MA, Langefeld CD, Williams A, Zakharkin SO, George V, Allison DB, Cooper GS, Xie F, Fan Z, Edberg JC, Kimberly RP. Single-nucleotide polymorphisms in the C-reactive protein (CRP) gene promoter that affect transcription factor binding, alter transcriptional activity, and associate with differences in baseline serum CRP level. J Mol Med. 2005; 83: 440–447.[CrossRef][Medline] [Order article via Infotrieve]
41. Suk HJ, Ridker PM, Cook NR, Zee RYL. Relation of polymorphism within the C-reactive protein gene and plasma CRP levels. Atherosclerosis. 2005; 178: 139–145.[CrossRef][Medline] [Order article via Infotrieve]
42. Brull DJ, Serrano N, Zito F, Jones L, Montgomery HE, Rumley A, Sharma P, Lowe GDO, World MJ, Humphries SE, Hingorani AD. Human CRP gene polymorphism influences CRP levels: implications for the prediction and pathogenesis of coronary heart disease. Arterioscler Thromb Vasc Biol. 2003; 23: 2063–2069.
43. de Maat MPM, Bladbjerg EM, von Bornemann Hjelmborg J, Bathum L, Jespersen J, Christensen K. Genetic influence on inflammation variables in the elderly. Arterioscler Thromb Vasc Biol. 2004; 24: 2168–2173.
44. Kovacs A, Green F, Hannson L-O, Lundman P, Samnegard A, Boquist S, Ericsson C-G, Watkins H, Hamsten A, Tornvall P. A novel common single nucleotide polymorphism in the promoter region of the C-reactive protein gene associated with the plasma concentration of C-reactive protein. Atherosclerosis. 2005; 178: 193–198.[CrossRef][Medline] [Order article via Infotrieve]
45. Hinds DA, Stuve LL, Nilsen GB, Halperin E, Eskin E, Ballinger DG, Frazer KA, Cox DR. Whole-genome patterns of common DNA variation in three human populations. Science. 2005; 307: 1072–1079.
46. Shen LX, Basilion JP, Stanton VP Jr. Single-nucleotide polymorphisms can cause different structural folds of mRNA. Proc Natl Acad Sci U S A. 1999; 96: 7871–7876.
47. Cartegni L, Chew SL, Krainer AR. Listening to silence and understanding nonsense: exonic mutations that affect splicing. Nat Rev Genet. 2002; 3: 285–298.[CrossRef][Medline] [Order article via Infotrieve]
48. Rieder MJ, Reiner AP, Gage BF, Nickerson DA, Eby CS, McLeod HL, Blough DK, Thummel KE, Veenstra DL, Rettie AE. Effect of VKORC1 haplotypes on transcriptional regulation and warfarin dose. N Engl J Med. 2005; 352: 2285–2293.
49. Pankow JS, Folsom AR, Cushman M, Borecki IB, Hopkins PN, Eckfeldt JH, Tracy RP. Familial and genetic determinants of systemic markers of inflammation: the NHLBI Family Heart Study. Atherosclerosis. 2001; 154: 681–689.[CrossRef][Medline] [Order article via Infotrieve]
50. Vickers MA, Green FR, Terry C, Mayosi BM, Julier C, Lathrop M, Ratcliffe PJ, Watkins HC, Keavney B. Genotype at a promoter polymorphism of the interleukin-6 gene is associated with baseline levels of plasma C-reactive protein. Cardiovasc Res. 2002; 53: 1029–1034.
51. MacGregor AJ, Gallimore JR, Spector TD, Pepys MB. Genetic effects on baseline values of C-reactive protein and serum amyloid a protein: a comparison of monozygotic and dizygotic twins. Clin Chem. 2004; 50: 130–134.
52. Black S, Kushner I, Samols D. C-reactive protein. J Biol Chem. 2004; 279: 48487–48490.
53. Ivashchenko Y, Kramer F, Schafer S, Bucher A, Veit K, Hombach V, Busch A, Ritzeler O, Dedio J, Torzewski J. Protein kinase C pathway is involved in transcriptional regulation of C-reactive protein synthesis in human hepatocytes. Arterioscler Thromb Vasc Biol. 2005; 25: 186–192.
54. Ferrari SL, Ahn-Luong L, Garnero P, Humphries SE, Greenspan SL. Two promoter polymorphisms regulating interleukin-6 gene expression are associated with circulating levels of C-reactive protein and markers of bone resorption in postmenopausal women. J Clin Endocrinol Metab. 2003; 88: 255–259.
55. Berger P, McConnell JP, Nunn M, Kornman KS, Sorrell J, Stephenson K, Duff GW. C-reactive protein levels are influenced by common IL-1 gene variations. Cytokine. 2002; 17: 171–174.[CrossRef][Medline] [Order article via Infotrieve]
56. Eklund C, Jahan F, Pessi T, Lehtimaki T, Hurme M. IL1B gene polymorphism is associated with baseline C-reactive protein levels in healthy individuals. Eur Cytokine Netw. 2003; 14: 168–171.[Medline] [Order article via Infotrieve]
57. Eklund C, Lehtimaki T, Hurme M. Epistatic effect of C-reactive protein (CRP) single nucleotide polymorphism (SNP) +1059 and interleukin-1B SNP +3954 on CRP concentration in healthy male blood donors. Int J Immunogenet. 2005; 32: 229–232.[Medline] [Order article via Infotrieve]
58. Margaglione M, Bossone A, Cappucci C, Colaizzo D, Grandone E, Di Minno G. The effect of the interleukin-6 C/G-174 polymorphism and circulating interleukin-6 on fibrinogen plasma levels. Haematologica. 2001; 86: 199–204.
59. Casas JP, Shah T, Cooper J, Hawe E, McMahon AD, Gaffney D, Packard CJ, O’Reilly DS, Juhan-Vague I, Yudkin JS, Tremoli E, Margaglione M, Di Minno G, Hamsten A, Kooistra T, Stephens JW, Hurel SJ, Livingstone S, Colhoun HM, Miller GJ, Bautista LE, Meade T, Sattar N, Humphries SE, Hingorani AD. Insight into the nature of the CRP-coronary event association using mendelian randomization. Int J Epidemiol. 2006; 35: 922–931.
60. Kardys I, de Maat MPM, Uitterlinden AG, Hofman A, Witteman JCM. C-reactive protein gene haplotypes and risk of coronary heart disease: the Rotterdam Study. Eur Heart J. 2006; 27: 1331–1337.
|
Footnotes |
|---|
The findings and conclusions in this report are those of the authors and do not necessarily represent the views of the Centers for Disease Control and Prevention.
The online-only Data Supplement, consisting of tables, is available with this article at http://circ.ahajournals.org/cgi/content/full/CIRCULATIONAHA.106.615740/DC1.
This article has been cited by other articles:
 |
|
 |
|
  A. Bertoli, L. Vila, G. Alarcon, G McGwin, J. Edberg, M Petri, R Ramsey-Goldman, J. Reveille, R. Kimberly, and for the PROFILE Study Group Factors Associated with Arterial Vascular Events in PROFILE: A Multiethnic Lupus Cohort Lupus, October 1, 2009; 18(11): 958 - 965. [Abstract] [PDF]
|
|
|
|
 |
|
 M. Bochud, F. Marquant, P.-M. Marques-Vidal, P. Vollenweider, J. S. Beckmann, V. Mooser, F. Paccaud, and V. Rousson Association between C-Reactive Protein and Adiposity in Women J. Clin. Endocrinol. Metab., October 1, 2009; 94(10): 3969 - 3977. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 O. P Almeida, P. E Norman, R. Allcock, F. v. Bockxmeer, G. J Hankey, K. Jamrozik, and L. Flicker Polymorphisms of the CRP gene inhibit inflammatory response and increase susceptibility to depression: The Health in Men Study Int. J. Epidemiol., August 1, 2009; 38(4): 1049 - 1059. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. E. Keebler, C. L. Sanders, A. Surti, C. Guiducci, N. P. Burtt, and S. Kathiresan Association of Blood Lipids With Common DNA Sequence Variants at 19 Genetic Loci in the Multiethnic United States National Health and Nutrition Examination Survey III Circ Cardiovasc Genet, June 1, 2009; 2(3): 238 - 243. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Shen and J. M. Ordovas Impact of Genetic and Environmental Factors on hsCRP Concentrations and Response to Therapeutic Agents Clin. Chem., February 1, 2009; 55(2): 256 - 264. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. B. Grammer, W. Marz, W. Renner, B. O. Bohm, and M. M. Hoffmann C-reactive protein genotypes associated with circulating C-reactive protein but not with angiographic coronary artery disease: the LURIC study Eur. Heart J., January 2, 2009; 30(2): 170 - 182. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Kelley-Hedgepeth, D. M. Lloyd-Jones, A. Colvin, K. A. Matthews, J. Johnston, M. R. Sowers, B. Sternfeld, R. C. Pasternak, C. U. Chae, and for the SWAN Investigators Ethnic Differences in C-Reactive Protein Concentrations Clin. Chem., June 1, 2008; 54(6): 1027 - 1037. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Kolz, W. Koenig, M. Muller, M. Andreani, S. Greven, T. Illig, N. Khuseyinova, D. Panagiotakos, G. Pershagen, V. Salomaa, et al. DNA variants, plasma levels and variability of C-reactive protein in myocardial infarction survivors: results from the AIRGENE study Eur. Heart J., May 2, 2008; 29(10): 1250 - 1258. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. C. Edberg, J. Wu, C. D. Langefeld, E. E. Brown, M. C. Marion, G. McGwin Jr, M. Petri, R. Ramsey-Goldman, J. D. Reveille, S. G. Frank, et al. Genetic variation in the CRP promoter: association with systemic lupus erythematosus Hum. Mol. Genet., April 15, 2008; 17(8): 1147 - 1155. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Gallicchio, H. Chang, D. K. Christo, L. Thuita, H.-Y. Huang, P. Strickland, I. Ruczinski, S. C. Hoffman, and K. J. Helzlsouer Single Nucleotide Polymorphisms in Inflammation-related Genes and Mortality in a Community-based Cohort in Washington County, Maryland Am. J. Epidemiol., April 1, 2008; 167(7): 807 - 813. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. G. Hage and A. J. Szalai C-Reactive Protein Gene Polymorphisms, C-Reactive Protein Blood Levels, and Cardiovascular Disease Risk J. Am. Coll. Cardiol., September 18, 2007; 50(12): 1115 - 1122. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||