Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Clin Chem. Author manuscript; available in PMC 2010 April 19.
Published in final edited form as:
PMCID: PMC2856608

Relation of Genetic Variation in the Gene Coding for C-Reactive Protein with its Plasma Protein Concentrations: Findings from the Women’s Health Initiative Observational Cohort



Although common genetic variants of the CRP gene have been associated with plasma CRP concentrations (hsCRP) in several cohorts of European Americans, relatively few studies have comprehensively assessed this association in well-characterized multiethnic populations.


In a case-control study of diabetes nested in the Women’s Health Initiative Observational Cohort, we comprehensively evaluated the association of genetic variation in the CRP gene with plasma hsCRP concentrations. Thirteen tagging single nucleotide polymorphisms (tSNPs) were identified and subsequently genotyped in 3,782 postmenopausal women.


The allele frequencies for these tSNPs and haplotype blocks defined by these tSNPs varied significantly by ethnic group. Consistent with prior studies among Whites, rs3093068, rs1130864 and rs1417938 were significantly associated with higher hsCRP concentrations (geometric mean per minor allele change: 1.20–1.25 mg/L), while rs1205 and rs1800947 were significantly associated with lower hsCRP values (1.28–1.48 mg/L). The associations with rs3093068 and rs1205 appeared to be stronger in Asian/Pacific Islanders than in Whites with a geometric mean increase of 1.65 mg/L compared to 1.25 mg/L, respectively. The minor alleles, rs3093075 and rs3093059, were associated with substantially increased hsCRP concentrations whereas rs1800947 was associated with lower hsCRP values. All haplotype-based association results tended to be consistent with the associations seen with single CRP SNPs.


Our large multiethnic case-control study of postmenopausal women provides evidence that common genetic variants in CRP gene are substantially associated with plasma hsCRP concentrations in this case-control subcohort. There are also suggestions of ethnic variations in these associations.

Keywords: C-Reactive Protein, CRP, Women’s Health Initiative, WHI


C-Reactive Protein (CRP) is an acute phase reactant secreted by the liver that has long been used to characterize systemic inflammation. Increased circulating concentrations of high-sensitivity CRP (hsCRP) have been associated with obesity, type 2 diabetes (T2DM)(14), and atherosclerotic cardiovascular disease(511). In a prospective case-control study of 3,782 postmenopausal women nested in the Women’s Health Initiative observational study (WHI-OS), increased circulating concentrations of hsCRP were significantly associated with an increased risk of clinical diabetes(3). Further, racial and ethnic differences in plasma hsCRP values have also been noted in these populations(1214). More recently, common genetic variants in the CRP gene were found to be associated with plasma hsCRP concentrations in several cohorts of European Americans(1520). To date, however, relatively few studies have examined the associations of common variants in the gene coding for CRP with its plasma hsCRP concentrations in a multiethnic-population. Therefore, we conducted a comprehensive assessment of the association of genetic variation in the CRP gene with plasma hsCRP values in a large case-control study nested in the WHI-OS, an ethnically diverse cohort of postmenopausal women aged over 50 years including White/Caucasian, Black/African, Hispanic and Asian/Pacific Islander participants. In particular, we examined whether any genetic effect on plasma hsCRP concentrations differ by ethnicity.


Study Participants

Details regarding our case-control study design have been described elsewhere(3;21). Briefly, of the 93,676 postmenopausal women enrolled in the WHI-OS(21), approximately 82,069 had no prior history of cardiovascular disease and/or diabetes at baseline. Following the principle of risk-set sampling, 1,584 diabetes cases were individually matched with 2,198 controls on age (±2.5 years), ethnicity, clinical center, time of blood draw (±0.10 hours), and length of follow-up. The ethnic groups represented in this study include Whites (n=1,936), Blacks (n=1,098), Hispanics (n=455), and Asian/Pacific Islanders (n=293). Further details about the ethnic distribution and demographics have already been described previously(3). The Institutional Review Board (IRB) at UCLA has previously approved the original study, and all study subjects have signed informed consent forms prior to study enrollment in the WHI.

Tagging SNP Selection

As described previously(22), we implemented a two-stage approach in choosing haplotype tagging Single Nucleotide Polymorphisms (tSNPs). First, we surveyed common genetic variation using the National Center for Biotechnology Information database SNP (NCBI dbSNP) supplemented by HapMap database(23). Our goal was to capture the common haplotype patterns in the genetic region covering the 30 kb 5′ upstream and 30 kb 3′ downstream of the CRP gene (3.1kb). The initial set of SNPs were selected based on the following criteria: 1) Functionality priority: nonsynonymous coding SNPs (cSNPs) and splicing-site SNPs (ssSNPs) must be kept; the priority order of picking SNPs based on their potential functions is: nonsynonymous SNPs > ssSNPs > synonymous SNPs > 5′ UTR SNPs > 3′ UTR SNPs > intronic SNPs; 2) minor allele frequency (MAF) ≥ 5% in at least one ethnic group; and 3) SNP density: SNPs should be relatively evenly spaced across the gene region(24).

In the second stage, we identified tSNPs based on the Linkage Disequilibrium (LD) patterns of those selected SNPs among 61 individuals from each ethnic population. Pair-wise LD between SNPs was assessed using the squared correlation statistic r2. The Haploview program was used to calculate the LD coefficient and define haplotype blocks(25;26). From the initial set of SNPs, a total of 13 tSNPs that account for most of the genetic variation within the CRP locus across all four ethnic groups were selected and genotyped in all the case-control samples.

SNP Genotyping Method

For these 13 tSNPs, large-scale genotyping was performed using the TaqMan allelic discrimination method. Specific primers and probes were custom-designed by Applied Biosystems (Applied Biosystems, ABI). Following polymerase chain reaction (PCR) amplification, end-point fluorescence was read using the ABI Primer 7900HT instrument and genotypes were scored using SDS2.2.2 Allelic Discrimination Software (Applied Biosystems). To evaluate reproducibility we performed blind duplicate genotyping in a randomly-selected subset of 5% of samples. Concordance rate was >99%.

Measurement of hsCRP concentrations

Since all incident cases and controls were free of clinical diabetes at baseline, baseline hsCRP concentrations were not influenced by clinical diabetes status. According to a standardized protocol, fasting blood specimens were collected from each participant at baseline and processed locally into separate aliquots containing serum, plasma, and buffy coat. Detailed sample preparation and handling procedures are described elsewhere(3). hsCRP was measured on the Hitachi 911 analyzer (Roche Diagnostics) using an immunoturbidimetric assay with reagents and calibrators from Denka Seiken Co Ltd. The coefficient of variation was 1.61% for hsCRP(3).

Statistical Analysis

We assessed each SNP for allele frequency and Hardy-Weinberg Equilibrium (HWE) test using the Proc Allele procedure in SAS (v9.1.3, SAS Institute, Cary, NC). We also tested for heterogeneity of genotype distributions across ethnic groups using the Chi-square test. For hsCRP concentrations with markedly skewed distributions, we made logarithmic transformations to enhance compliance with normality assumption then calculated geometric mean differences. To determine the effect of genetic variants on plasma hsCRP concentrations, we calculated the geometric mean differences in plasma hsCRP according to each tSNP genotyped by fitting general linear models (GLM) treating plasma hsCRP concentration as a dependent variable and tSNPs as independent variables. An additive inheritance model was used in single gene analysis, and the results of this analysis were expressed as the geometric mean increase in hsCRP per each additional copy of the minor allele. All the linear models included matching factors and potential confounders (case/control status, alcohol consumption, cigarette smoking, body mass index (BMI), hormone replacement therapy (HRT) used, and physical activity) as covariates in each of the four ethnic groups. Likelihood ratio tests were used to test the interaction effects between the genotypes and ethnicity on diabetes risk.

The Haploview program was used to characterize the LD patterns between tSNPs(25;26). In haplotype-based analysis, we estimated haplotype frequency using an “expectation-substitution” algorithm(27), which treated expected haplotype scores (calculated under a user-specified inheritance model) as observed covariates in analysis, and stored these expected scores in an auxiliary data set for customized analyses. Only haplotypes with estimated frequencies ≥ 1% in the combined controls were included for further analyses. To examine the association between the resulting haplotype/haplotype combinations and hsCRP concentrations, the estimate of haplotype dosage was treated as a surrogate variable for the true haplotype.

Further, to account for potential false positives due to multiple comparisons, we calculated the false discovery rate (FDR) by incorporating all p-values from multiple tests performed for SNPs and haplotypes in the GLM analysis(28). Proc Multtest procedure in SAS 9.1.3 was used to obtain the adjusted p-values.


LD Patterns in the CRP gene defined by tagging SNPs

Table 1 summarizes the genetic variation within the CRP locus defined by the 13 tSNPs in each of the four ethnic groups. The allele frequencies for all tSNPs varied significantly by ethnicity (all P < 0.0001). The specific locations of these thirteen tSNPs are schematically presented in Figure 1, based on NCBI Entrez gene for the CRP gene structure (NT_004487.18), and contig position for each SNP based on NCBI build 36 and dbSNP b129. None of the SNPs, except for rs3093068 in Whites, showed statistically significant deviation from HWE among controls in each ethnic group (data not shown).

Figure 1
Human C-Reactive Protein (CRP) gene (Chromosome 1q21-q23)
Table 1
The location, relative distances, minor allele frequencies (MAF) of genotyped tagging SNPs in CRP Gene in Controls

Single-marker association analysis

Of the 13 tSNPs genotyped, eleven tSNPs were significantly associated with plasma hsCRP concentrations after adjustment for matching factors and other potential confounders (see Supplemental Data Table 1 available at with the online version of this paper). Supplemental Data Figure 1 shows the different pattern in the mean differences (without logarithmic transformations) and 95% confidence intervals (CIs) of hsCRP concentrations for each SNP, as estimated by an additive genetic model. In an additive model, geometric mean differences in plasma hsCRP concentrations were associated with an increasing number of minor alleles as compared to those homozygous major alleles (i.e. the genotypes were coded as 0=homozygous major alleles, 1=heterozygotes, and 2=homozygous minor alleles). For six tSNPs, carriers of each additional copy of the minor allele were associated with decreased hsCRP concentrations: the geometric mean per-allele decrease ranging from 1.07 to 1.46 mg/L(all P < 0.05). In contrast, the minor alleles of five tSNPs (rs3093059, rs1417938, rs1130864, rs3093075, and rs3093068) were associated with increased hsCRP concentrations with the geometric mean per-allele increase ranging from 1.17 to 1.27mg/L (all P < 0.05). The tSNPs that were associated with plasma hsCRP concentrations also varied by ethnicity (Supplemental Data Table 2). Figure 2 shows the different pattern in the mean differences (without logarithmic transformations) and 95% CIs of hsCRP values for each SNP by ethnicity, as estimated by an additive genetic model. Nine out of 13 tSNPs were significantly associated with plasma hsCRP concentrations in Whites and Asian/Pacific Islanders, and eight of these nine tSNPs were common to both ethnic groups. The one significantly different tSNP, rs1417938 in Whites, rs4275453 in Asian/Pacific Islanders, was associated with increased hsCRP concentrations. The association of these tSNPs with higher hsCRP concentrations in Whites and Asian/Pacific Islanders was significantly different from that in Blacks and Hispanics where no association with hsCRP concentrations was observed. In contrast, only four of the 13 tSNPs among Blacks and Hispanics were significantly associated with plasma hsCRP concentrations. There was no overlap in the four significant tSNPs between Blacks and Hispanics.

Figure 2
The adjusted mean difference of plasma hsCRP concentrations (95% confidence intervals) according to CRP gene variation based on an additive genetic model, by ethnicity

Haplotype association analysis

The lengths and locations of specific haplotype blocks by ethnicity are presented in Figure 3. Haplotype frequencies and the specific haplotype blocks varied with ethnicity (Figure 3 and Table 2). In Whites, there were three haplotypes in block 1 and five haplotypes in block 2. The 0-0 carriers in block 1 and carriers of 0-1-1-0-0-0-0-0-1-0 in block 2 (where 0 coded for major allele and 1 coded for minor allele) exhibited the highest hsCRP values with geometric mean difference in hsCRP concentrations of 1.27 mg/L and 1.21 mg/L compared with non-carriers, respectively. We only identified 3 haplotypic blocks of this CRP gene in other ethnicities, each containing two to five haplotypes within each block. Blacks and Hispanics had haplotype blocks with similar boundaries whereas block 2 in Asian/Pacific Islanders was significantly different from the other ethnic groups. In block 2 for Blacks and Hispanics, haplotype 0-0-0-0-0-0-0 carriers had increased hsCRP concentrations (1.23 and 1.58 mg/L, respectively) compared with non-carriers.

Figure 3
Haploview plot defining ethnic-specific LD structures between the 13 tSNPs within or near CRP gene among control participants. Each diamond for each SNP combination indicates the pairwise LD between all tSNPs. LD strength between the chosen SNPs is determined ...
Table 2
Associations between ethnic-specific haplotypes of CRP gene and plasma hsCRP concentrations

In addition to examining the haplotype block identified in our study population, we also replicated the haplotypes found in other studies (i.e., rs1130864-rs1205-rs3093068 by Dehghan et al(16), and rs1180947-rs1130864-rs1205 by Timpson et al(17) (Table 3 and Supplemental Data Table 3). Among the rs1130864-rs1205-rs3093068 haplotypes, 0-0-1 carriers appeared to have increased plasma hsCRP compared with non-carriers, and this relationship was consistently observed in all four ethnicities (geometric mean difference = 1.24, 1.36, 1.31, and 1.60 mg/L, respectively).

Table 3
Geometric mean differences in plasma hsCRP concentrations according to specific CRP haplotype and ethnicity estimated by an additive genetic effect model


In this large multiethnic cohort of postmenopausal women, common genetic variants in the CRP gene were highly associated with plasma hsCRP concentrations. In particular, these genotype-phenotype relationships appeared to vary significantly across ethnic groups.

Several tSNPs (rs1130864, rs1205, rs3093068, rs1800947, rs2794521) identified in our study have been reported previously in both White and African American men and women (1520). Although the specific effects of these variants on plasma hsCRP concentrations remain to be determined, it has been estimated in a previous family study that 30–40% of the variation in plasma hsCRP concentrations may be due to genes(29). In our multi-ethnic cohort, common genetic variants in the CRP gene accounted for 4% of the total variation of plasma hsCRP in Whites, 5% of the variation in Blacks, and 10% of the variation in Hispanics and Asian/Pacific Islanders.

Consistent with prior studies among Whites (19), carriers of minor alleles of rs3093068, rs1130864 and rs1417938 were found to have significantly higher hsCRP concentrations, whereas rs1205 and rs1800947 were associated with lower plasma hsCRP concentrations compared to non-carriers. The tSNPs rs1130864 and rs1205 appeared to have even greater effects on plasma hsCRP concentrations in Asian/Pacific Islanders (hsCRP concentrations increase by 1.62 mg/L with rs1130864 and decrease by 1.57 mg/L with rs1205). Similar ethnic differences were also found for rs3093075 and rs3093059 where greater magnitude of associations appeared more evident in Asians and Hispanics than that in Whites (i.e. geometric mean hsCRP values increase by 1.5 to 1.7 mg/L vs. 1.2 mg/L). In contrast, rs1800947 was significantly associated with lower hsCRP concentrations in Whites but not in Hispanics, a finding that has been reported previously(20). The tSNP rs2794521 has not been found to be associated with plasma hsCRP concentrations in other studies but has been reported to be associated with T2DM in Pima Indians(30) and coronary disease in ethnic Han Chinese(31). In the current study, both rs2794521 and rs2808634 were significantly associated with lower hsCRP concentrations in Hispanics but not in Whites. The direction and magnitude of associations for rs1205, rs3093075, rs2808629 and rs1470515 were similar in Blacks compared to Whites.

Several reasons for the ethnic differences of the gene-phenotype relationships have been noted previously. The apparently distinct LD patterns of this CRP gene by ethnicity were a novel observation and a unique strength of our multiethnic study, as previous studies have mainly examined tSNPs or haplotypes in European Americans. In prior reports(15;32), rs3091244 was found to be a functional triallelic SNP in the CRP promoter region with high LD to other SNPs (i.e., rs113084, rs1417938 and rs2795421), possibly explaining the significant associations found in those studies. There may be further population stratification (3335) within each ethnic subgroup.I If such stratification is present, the significant ethnic differences we found in our study would only be biased toward the null.

Since comprehensive screening of SNPs was conducted in the current study, we were able to define haplotype structures for each ethnic group beyond those reported in previous studies(16;17). The haplotype frequencies for Whites were similar to those reported for European Americans(16;17). For example, Timpson et al(17) showed that carriers of C-A-C in rs1800947, rs1130864, and rs1205 (i.e. 0-1-0 in Table 3 and Supplemental Data Table 3) have the highest plasma hsCRP values whereas GGT carriers have the lowest hsCRP concentrations. In our study, plasma hsCRP concentrations were 1.21 mg/L (Standard error (SE)=1.04) higher in haplotype 0-1-0 carriers, and 1.49 mg/L (SE=1.07) lower in 1-0-1 carriers compared with that of non-carriers. Similar trends were also found in Black and Asian carriers (i.e., carriers of haplotype 0-1-0 carriers had higher hsCRP values compared to non-carriers). In contrast, carriers of the same haplotypes had lower hsCRP concentrations than did non-carriers among Hispanics (p>0.05). In the report by Dehghan et al(16), the highest contrast of plasma hsCRP values was found in comparing carriers of the CCG haplotype (defined by three SNPs, rs1130754, rs1205, and rs3093068, same as 0-0-1 in Table 3 and Supplemental Data Table 3) with carriers of the most common haplotype CTC (i.e., 0-1-0 in Supplemental Data Table 3).

More importantly, minor allele frequencies, haplotype structure, and LD patterns varied by ethnic group. For example, we observed similar haplotype frequencies in Whites as well as similar magnitudes of effects on plasma hsCRP concentrations associated with the haplotypes to those seen in the Dehghan et al study(16) (i.e., increased hsCRP concentrations by 1.24 mg/L in carriers of 0-0-1 and decreased hsCRP concentrations by 1.29 mg/L in carriers of 0-1-0 haplotype). However, no significant associations were found between hsCRP and these haplotypes in the other ethnic groups. In general, our haplotype-based analysis was consistent with the single SNP- plasma hsCRP associations in terms of direction and magnitude. We also observed that the direction of the effect remained the same but the magnitude of the effect varied with ethnicity.

In conclusion, we found that common genetic variants in the CRP gene were substantially and independently associated with plasma hsCRP concentrations in this large multiethnic case–control study of postmenopausal women. These associations may differ by ethnicity. Further studies will be warranted to confirm the ethnic differences of CRP gene variance and the potential direct role of these genetic variants on T2DM risk.

Supplementary Material

Supplementary Data


We would like to acknowledge all WHI centers, principal investigators and committed participants in this research. The WHI is funded by the National Heart, Lung, and Blood Institute, and U.S. Department of Health and Human Services. This study was supported by the National Institute of Diabetes and Digestive and Kidney Diseases R01 DK062290 from the National Institutes of Health.


C-Reactive Protein
haplotype tagging SNPs
high-sensitivity C-Reactive Protein
type 2 diabetes mellitus
Institutional Review Board
University of California, Los Angeles,
Women’s Health Initiative Observational Study
single nucleotide polymorphism
National Center for Biotechnology Information database SNP
splicing-site SNPs
nonsynonymous coding SNPs
Untranslated Region
polymerase chain reaction
Applied Biosystems
minor allele frequency
linkage disequilibrium
Hardy-Weinberg Equilibrium
general linear models
body mass index
hormone replacement therapy
false discovery rate
standard error


HUMAN GENES DISCUSSED IN THIS MANUSCRIPT: CRP with HUGO approved names “C-reactive protein” and “pentraxin-related” (HGNC: 2367). Alias: PTX1.

CONFLICT OF INTEREST: None of the authorshad any conflicts of interest.

AUTHOR CONTRIBUTIONS: Drs. You and Liu had full access to the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Dr. Lee and Dr. You contributed equally to the manuscript as joint first authors.

Conception and design: Liu, You

Acquisition of data: Liu, Tinker, Nathan, Manson

Analysis and interpretation of data: You, Liu, Lee, Song, Hsu

Drafting of manuscript: Lee, You, Liu

Critical revision for intellectual content: Lee, You, Song, Hsu, Tinker, Manson, Nathan, Liu

Statistical expertise: You, Song, Hsu, Liu

Obtained funding: Liu

Administrative, technical, or material support: Liu, Manson, Tinker, Nathan

Study supervision: Liu

Reference List

1. Hu FB, Meigs JB, Li TY, Rifai N, Manson JE. Inflammatory markers and risk of developing type 2 diabetes in women. Diabetes. 2004;53:693–700. [PubMed]
2. Pradhan AD, Manson JE, Rifai N, Buring JE, Ridker PM. C-reactive protein, interleukin 6, and risk of developing type 2 diabetes mellitus. JAMA. 2001;286:327–34. [PubMed]
3. Liu S, Tinker L, Song Y, Rifai N, Bonds DE, Cook NR, et al. A prospective study of inflammatory cytokines and diabetes mellitus in a multiethnic cohort of postmenopausal women. Arch Intern Med. 2007;167:1676–85. [PubMed]
4. Zee RY, Germer S, Thomas A, Raji A, Rhees B, Ridker PM, et al. C-reactive protein gene variation and type 2 diabetes mellitus: a case-control study. Atherosclerosis. 2008;197:931–6. [PubMed]
5. Pai JK, Pischon T, Ma J, Manson JE, Hankinson SE, Joshipura K, et al. Inflammatory markers and the risk of coronary heart disease in men and women. N Engl J Med. 2004;351:2599–610. [PubMed]
6. Ridker PM, Hennekens CH, Buring JE, Rifai N. C-reactive protein and other markers of inflammation in the prediction of cardiovascular disease in women. N Engl J Med. 2000;342:836–43. [PubMed]
7. Ridker PM, Cushman M, Stampfer MJ, Tracy RP, Hennekens CH. Plasma concentration of C-reactive protein and risk of developing peripheral vascular disease. Circulation. 1998;97:425–8. [PubMed]
8. Ridker PM, Buring JE, Shih J, Matias M, Hennekens CH. Prospective study of C-reactive protein and the risk of future cardiovascular events among apparently healthy women. Circulation. 1998;98:731–3. [PubMed]
9. Ridker PM, Haughie P. Prospective studies of C-reactive protein as a risk factor for cardiovascular disease. J Investig Med. 1998;46:391–5. [PubMed]
10. Ridker P, Rifai N, Koenig W, Blumenthal RS. C-reactive protein and cardiovascular risk in the Framingham Study. Arch Intern Med. 2006;166:1327–8. [PubMed]
11. Pai JK, Mukamal KJ, Rexrode KM, Rimm EB. C-reactive protein (CRP) gene polymorphisms, CRP levels, and risk of incident coronary heart disease in two nested case-control studies. PLoS ONE. 2008;3:e1395. [PMC free article] [PubMed]
12. Ranjit N, Diez-Roux AV, Shea S, Cushman M, Ni H, Seeman T. Socioeconomic position, race/ethnicity, and inflammation in the multi-ethnic study of atherosclerosis. Circulation. 2007;116:2383–90. [PubMed]
13. Albert MA, Glynn RJ, Buring J, Ridker PM. C-reactive protein levels among women of various ethnic groups living in the United States (from the Women’s Health Study) Am J Cardiol. 2004;93:1238–42. [PubMed]
14. Albert MA, Ridker PM. C-reactive protein as a risk predictor: do race/ethnicity and gender make a difference? Circulation. 2006;114:e67–e74. [PubMed]
15. Miller DT, Zee RY, Suk DJ, Kozlowski P, Chasman DI, Lazarus R, et al. Association of common CRP gene variants with CRP levels and cardiovascular events. Ann Hum Genet. 2005;69:623–38. [PubMed]
16. Dehghan A, Kardys I, de Maat MP, Uitterlinden AG, Sijbrands EJ, Bootsma AH, et al. Genetic variation, C-reactive protein levels, and incidence of diabetes. Diabetes. 2007;56:872–8. [PubMed]
17. Timpson NJ, Lawlor DA, Harbord RM, Gaunt TR, Day IN, Palmer LJ, et al. C-reactive protein and its role in metabolic syndrome: mendelian randomisation study. Lancet. 2005;366:1954–9. [PubMed]
18. Carlson CS, Aldred SF, Lee PK, Tracy RP, Schwartz SM, Rieder M, et al. Polymorphisms within the C-reactive protein (CRP) promoter region are associated with plasma CRP levels. Am J Hum Genet. 2005;77:64–77. [PubMed]
19. Lange LA, Burdon K, Langefeld CD, Liu Y, Beck SR, Rich SS, et al. Heritability and expression of C-reactive protein in type 2 diabetes in the Diabetes Heart Study. Ann Hum Genet. 2006;70:717–25. [PubMed]
20. Crawford DC, Yi Q, Smith JD, Shephard C, Wong M, Witrak L, et al. Allelic spectrum of the natural variation in CRP. Hum Genet. 2006;119:496–504. [PMC free article] [PubMed]
21. Design of the Women’s Health Initiative clinical trial and observational study. The Women’s Health Initiative Study Group. Control Clin Trials. 1998;19:61–109. [PubMed]
22. Hao K, Liu S, Niu T. A sparse marker extension tree algorithm for selecting the best set of haplotype tagging single nucleotide polymorphisms. Genet Epidemiol. 2005;29:336–52. [PMC free article] [PubMed]
23. A haplotype map of the human genome. Nature. 2005;437:1299–320. [PMC free article] [PubMed]
24. Wang L, Liu S, Niu T, Xu X. SNPHunter: a bioinformatic software for single nucleotide polymorphism data acquisition and management. BMC Bioinformatics. 2005;6:60. [PMC free article] [PubMed]
25. Gabriel SB, Schaffner SF, Nguyen H, Moore JM, Roy J, Blumenstiel B, et al. The structure of haplotype blocks in the human genome. Science. 2002;296:2225–9. [PubMed]
26. Barrett JC, Fry B, Maller J, Daly MJ. Haploview: analysis and visualization of LD and haplotype maps. Bioinformatics. 2005;21:263–5. [PubMed]
27. Kraft P, Cox DG, Paynter RA, Hunter D, De VI. Accounting for haplotype uncertainty in matched association studies: a comparison of simple and flexible techniques. Genet Epidemiol. 2005;28:261–72. [PubMed]
28. Benjamini Y, Drai D, Elmer G, Kafkafi N, Golani I. Controlling the false discovery rate in behavior genetics research. Behav Brain Res. 2001;125:279–84. [PubMed]
29. 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;154:681–9. [PubMed]
30. Wolford JK, Gruber JD, Ossowski VM, Vozarova B, Antonio TP, Bogardus C, Hanson RL. A C-reactive protein promoter polymorphism is associated with type 2 diabetes mellitus in Pima Indians. Mol Genet Metab. 2003;78:136–44. [PubMed]
31. Chen J, Zhao J, Huang J, Su S, Qiang B, Gu D. -717A>G polymorphism of human C-reactive protein gene associated with coronary heart disease in ethnic Han Chinese: the Beijing atherosclerosis study. J Mol Med. 2005;83:72–8. [PubMed]
32. Szalai AJ, Wu J, Lange EM, McCrory MA, Langefeld CD, Williams A, et al. 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;83:440–7. [PubMed]
33. Serre D, Montpetit A, Pare G, Engert JC, Yusuf S, Keavney B, et al. Correction of population stratification in large multi-ethnic association studies. PLoS ONE. 2008;3:e1382. [PMC free article] [PubMed]
34. Pritchard JK, Stephens M, Donnelly P. Inference of population structure using multilocus genotype data. Genetics. 2000;155:945–59. [PubMed]
35. Hao K, Li C, Rosenow C, Wong WH. Detect and adjust for population stratification in population-based association study using genomic control markers: an application of Affymetrix Genechip Human Mapping 10K array. Eur J Hum Genet. 2004;12:1001–6. [PubMed]