Original Article

Genes and Immunity (2012) 13, 509–513; doi:10.1038/gene.2012.26; published online 7 June 2012

Corrected online: 30 August 2012

There is a Corrigendum (1 September 2012) associated with this article.

Genome-wide association study identifies common variants at TNFRSF13B associated with IgG level in a healthy Chinese male population

M Liao1,2,7, F Ye2,7, B Zhang2,7, L Huang2,7, Q Xiao2,7, M Qin1, L Mo1, A Tan2,3, Y Gao1,2,6, Z Lu1,2, C Wu1,2, Y Zhang1,2, H Zhang2,4, X Qin2,5, Y Hu2, X Yang2,4 and Z Mo1,2

  1. 1Institute of Urology and Nephrology, First Affiliated Hospital of Guangxi Medical University, Nanning, China
  2. 2Center for Genomic and Personalized Medicine, Guangxi Medical University, Nanning, China
  3. 3Center for Metabolic Disease and Diabetes, First Affiliated Hospital of Guangxi Medical University, Nanning, China
  4. 4Department of Occupational Health and Environmental Health, School of Public Health of Guangxi Medical University, Nanning, China
  5. 5Department of Clinical Laboratory, First Affiliated Hospital of Guangxi Medical University, Nanning, China
  6. 6Medical Scientific Research Centre, Guangxi Medical University, Nanning, China

Correspondence: Professor X Yang, Department of Occupational Health and Environmental Health, School of Public Health of Guangxi Medical University, NO. 6 Shuangyong Road, Nanning, Guangxi Zhuang Autonomous Region 530021, China. E-mail: yxbo21021@163.com; Professor Z Mo, Institute of Urology and Nephrology, First Affiliated Hospital of Guangxi Medical University, NO. 22 Shuangyong Road, Nanning, Guangxi Zhuang Autonomous Region 530021, China. E-mail: zengnanmo@hotmail.com

7These authors contributed equally to this work.

Received 16 March 2012; Revised 1 May 2012; Accepted 4 May 2012
Advance online publication 7 June 2012



IgG has a crucial role in humoral immune response. Serum IgG level is mainly determined under genetic control. To explore the genetic influence on serum IgG levels, a two-stage genome-wide association study (GWAS) was performed in a healthy Chinese population of 3495 men, including 1999 unrelated subjects in the first stage and 1496 independent individuals in the second stage. Three single-nucleotide polymorphisms (SNPs) located in TNFRSF13B on 17p11.2 or nearby were significantly associated with IgG level: rs4792800 in the intron (combined P-value=1.45 × 10−12), rs12603708 in the intron (combined P-value=1.82 × 10−8) and rs3751987 at ~65kb downstream of the 5′-UTR region of TNFRSF13B (combined P-value=3.67 × 10−9). Additionally, smoking was identified to be associated with IgG level in both stages (P<0.001), but there was no significant interaction between smoking and the identified SNPs (P>0.05). The strong association between variants at TNFRSF13B and IgG level may be helpful to further explore the biological mechanism by which the serum IgG is affected by transmembrane activator, calcium modulator and cyclophilin ligand interactor encoded by TNFRSF13B.

This article has been corrected since Advance Online Publication and a corrigendum is also published in this issue.


IgG; genome-wide association studies; TNFRSF13B; TACI; smoking



IgG, constituting some 80% of total serum immunoglobulin, is the most abundant class of antibodies (Ab) in serum and extracellular fluid. It has an important role in the defense against various pathogens, toxins and even cancers.1 As the only isotype that can pass through the human placenta, IgG is able to protect the fetus in utero.2 Meanwhile, IgG gets involved in chronic inflammatory processes contributed to the destruction of healthy tissues in autoimmune diseases, such as arthritis and systemic lupus erythematosus.3

Serum IgG level varies considerably from one person to another,4 but little when testing the same person repeatedly.5 This variation in the serum IgG concentration reflects the balance among the rates of synthesis and catabolism.6 Multitudinous environmental factors including infections, smoking and chemical compounds can contribute to this variation.7, 8, 9 Reports of pedigree studies or twin studies10, 11 have shown that genetic factors are more important than environmental exposures in determining serum immunoglobulin level in human.12, 13 The values of genetic heritability was h2=0.617+0.020 for IgG.6 Besides, several studies in a population of common variable immunodeficiency patients or IgA deficiency patients implied that there might be an association between some gene loci and serum IgG level.14, 15 However, to the best of our knowledge, no study has ever been conducted to identify the immune loci directly involved in the regulation of the IgG level.

Genome-wide association study (GWAS) is a powerful and unbiased tool for the identification of common genetic variants associated with complex traits.16 This hypothesis-free approach based on linkage disequilibrium (LD) allows the discovery of novel genetic loci, which were previously not regarded to be associated with the trait.17 Considering that the common genetic variants associated with serum IgG might be identified by the GWAS approach, we performed the present two-stage GWAS in a healthy Chinese male population. To the best of our knowledge, it is the first GWAS to search the population-specific genetic variations associated with serum IgG level.



General characteristics of the participants in the two-stage GWAS

The basic characteristics of the individuals recruited in this two-stage GWAS have been described in our prior study.18 In total, 1999 participants in stage 1 and 1496 participants in stage 2 were available for analysis. No significant difference was observed between the two stages in age distribution (37.54 versus 37.31 years, P=0.54), body mass index (23.31 versus 23.46kgm−2, P=0.18) and smoking behavior (P=0.66), excepting for alcohol consumption (P=0.02). According to the results of independent samples t-test (Table 1), we found that there were significant differences on serum IgG level between drinkers and non-drinkers in the first stage (P<0.001), whereas no significant difference was detected in the second stage (P=0.158). In addition, IgG level in smokers was significantly lower than that in non-smokers in both stages (Table 1).

Findings of the two-stage GWAS

The quantile–quantile plot of adjusted P-values indicated no systematic bias with an inflation factor of 1.047, which showed that no population substructure was observed in our GWAS population (Figure 1). When the top two Eigens were added to other covariates in the GWAS analysis, similar results were obtained. The major results of the present GWAS are shown in the Manhattan plot (Figure 2).

Figure 1.
Figure 1 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact help@nature.com or the author

The quantile–quantile plot of genome-wide QTL mapping for log-transformed IgG level.

Full figure and legend (38K)

Figure 2.
Figure 2 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact help@nature.com or the author

Manhattan plot of genome-wide association analysis for IgG level. X-axis shows chromosomal positions. Y-axis shows –log10 P-values from the linear regression. Candidate gene names are shown for the significantly associated loci. A full colour version of this figure is available at the Genes and Immunity journal online.

Full figure and legend (98K)

In the first stage, after testing the independence of each SNP through the multiple regression analysis, only the single-nucleotide polymorphisms (SNPs) that remained significant at 1.0 × 10−6 were chosen, thus one SNP per region remained. According to LD and hap-block of SNPs, five SNPs were finally selected to be further confirmed (Supplementary material and Supplementary Figures 1–3). Rs1634731 is in the intron of MUC21 on 6p21.32 (P=2.13 × 10−6) and rs2013111 is in intergenic region of gene (P=9.77 × 10−6), whrereas the other three are all in the TNFRSF13B on 17p11.2 or nearby. They are rs3751987 located at ~65kb downstream of the 5'-UTR region of TNFRSF13B (P=1.57 × 10−7), rs12603708 and rs4792800 in the intron region of TNFRSF13B (P=5.61 × 10−7 and P=3.33 × 10−11, respectively). In terms of LD, rs3751987 was in strong LD with rs4792800 located in TNFRSF13B (r2=0.54), whereas rs4792800 and rs12603708 were in moderate LD with each other (r2=0.34).

In the second stage, the five candidate SNPs were further validated by examining 1496 healthy subjects. The three SNPs in TNFRSF13B or nearby still showed significant associations with IgG level (Table 2): rs4792800 (P=6.32 × 10–5), rs3751987 (P=6.74 × 10–4) and rs12603708 (P=6.96 × 10–4). SNP rs4792800 was the most significantly associated with IgG level (combined P-value=1.45 × 10–12), whereas the other two SNPs also reached the GWAS significance level of 5 × 10–8: rs3751987 (combined P-value=3.67 × 10–9) and rs12603708 (combined P-value=1.82 × 10–8). However, rs1634731 and rs2013111 were not confirmed with combined P-values of 3.13 × 10–5 and 7.05 × 10–4, respectively.



It has long been recognized that serum IgG level was influenced by genetic variants.12, 13 Through the present two-stage GWAS in the healthy Chinese male population, we found that serum IgG level was significantly associated with three SNPs located in the region of TNFRSF13B on 17p11.2 or nearby. They were rs4792800 (combined P-value=1.45 × 10–12) and rs12603708 (combined P-value=1.82 × 10–8) in the intron, and rs3751987 (combined P-value=3.67 × 10–9) at ~65kb downstream of the 5′-UTR region of TNFRSF13B.

TNFRSF13B encodes the tumor necrosis factor receptor (TNFR) superfamily member transmembrane activator, calcium modulator and cyclophilin ligand interactor (TACI), which was preferentially expressed on marginal zone B cells, CD27+ memory B-cell subsets and plasma cells.19, 20 TACI binds to not only the B-cell-activating factor (BAFF) encoded by TNFSF13B, but also the second ligand called a proliferative inducing ligand (APRIL) encoded by TNFSF13.21 These interactions induce class switch recombination in human and murine B cells.22, 23 Moreover, TACI is able to induce activation of the transcription factors NFAT, AP1 and NF-κ-B by interacting with the TNF ligand.24 Ambiguous biological activities of TACI in humoral immunity have been reviewed in mice and humans.21 Nevertheless, it has been suggested that TACI was associated with common variable immunodeficiency and selective IgA deficiency in human.14, 15 One study also showed that TACI might be involved in innate immune functions, rather than only in adaptive immunity.25

The controversial role of TACI in IgG production has been collectively shown in previous studies. Animal studies showed that TACI/ mice were present with defective IgG secretion.26, 27 TACI seemed to have a positive role in the regulation of IgG production, however, it might negatively regulate the IgG production. It was previously reported in vitro study that B-cell proliferation and IgG production costimulated by BAFF-R and CD40 could be suppressed by TACI.28 Moreover, in naïve TACI-deficient mice, total serum IgG subtypes produced by B cells showed a marginal increase in response to the stimulation.29 In human beings, TACI has been shown to be associated with common variable immunodeficiency dominated by the presence of markedly reduced serum level of IgG.14, 15 It probably induced disease susceptibility rather than caused disease directly.30 Collectively, these findings are indirectly consistent with our findings that TNFRSF13B mutations may be associated with serum IgG level. Nevertheless, more studies are needed to explore the detail biological mechanism howTNFRSF13B mutations influence IgG level.

Previous study has identified that smoking was associated with serum IgG level. It has been reported that the function of regulatory T lymophocyte was suppressed by cigarette smoking, thus the overall level of serum IgG was depressed.31 Besides, the interactions between smoking and genetic effects on the IgG subclasses (IgG1, IgG2 and IG3) have been demonstrated in previous study.7 Although in our two-stage GWA study, serum IgG level in smokers is significantly lower than that in non-smokers, which is consistent with previous studies;7, 9 there is no significant interaction between smoking and any of the three identified SNPs in determining the serum level of IgG. Nevertheless, the suggestion that the phenotype of TNFRSF13B mutations is affected by gene-environment factors has been proposed by several studies.32

Limitations in our study should be evaluated objectively. First, this study was restricted to male participants, because it was derived from the male cohort in Fangchenggang Area Male Health and Examination Survey (FAMHES)18 We just explored the genetic influence on serum IgG level with a healthy male population but did not replicate in females or patients, which might lead to a relative selection bias.7 Moreover, we did not measure the IgG subclasses, IgA and IgM in the present study. Considering that TNFRSF13B has been linked with specific IgG subclasses in recent animal33 and human studies,34 and its mutations have been described in a higher frequency in common variable immunodeficiency patients who display low levels of IgG, IgA and IgM, following GWAS in terms of IgG subclass, IgA or IgM is expected. In addition, the total number of population was still limited, which might make some SNPs not to reach the significance level of GWAS as expected. For instance, rs1634731 in MUC21 showing a potential trend to be associated with IgG level just reached the combined P-value of 3.13 × 10–5.

In summary, to our knowledge, we are the first to perform the two-stage GWAS in Chinese male population to explore the genetic influence on serum IgG level. The three SNPs with the most significant association with IgG level are located in TNFRSF13B on 17p11.2 or nearby. The result indicates a significant association between common variants at TNFRSF13B and serum IgG level, and may provide insights into the biological mechanism how the serum IgG level is affected by TACI encoded by TNFRSF13B.


Materials and methods

Study population

A two-stage GWAS was performed to identify the genes/loci that influence serum IgG level. The participants in stage 1 (n=2012) were randomly selected from the men (n=4303) who participated in physical examinations in the Medical Centre of Fangchenggang First People’s Hospital from September 2009 to December 2009. They were unrelated healthy Chinese males aged 20–69 years from the FAMHES, which has been described elsewhere.35

The stage 2 was conducted with 1496 healthy Chinese men aged 20–69 years. They were randomly chosen from the men who participated in physical examinations from September 2009 to September 2010 in the Medical Centre of Fangchenggang First People’s Hospital, Guigang People’s Hospital and Yulin First People’s Hospital. The design of this two-stage GWAS had been described in detail elsewhere.18 Both smoking and drinking in two stages were assessed on the basis of a self-administered life-style questionnaire according to the same protocol. Respondents that reported smoking currently (daily smoking >6 moths) were coded as smokers, and those reported drinking any beverage ‘more than once a year’ were coded as drinkers, whereas others were non-drinkers.18, 35 Our study research protocol was approved by the local Ethics Committee and Informed consent of each participant was obtained with written documentation.

Measurement of IgG

The description of the laboratory test has been previously reported in detail.18 Briefly, about 10ml overnight fasting venous blood specimens were collected between 0800 and 1100h in the morning and were transported frozen to the testing center of Department of Clinical Laboratory at the First Affiliated Hospital of Guangxi Medical University in Nanning in two hours, which were centrifuged within 15 to 25 min and stored at −80 until analysis. IgG was measured with electrochemiluminescence immunoassay on COBAS 6000 system E601 (Elecsys module) immunoassay analyzer (Roche Diagnostics, GmbH, Mannheim, Germany) with the same batch of reagents, and the inter-assay coefficient of variation was 4.97%.

SNP genotyping and quality control

In our study, two different platforms were used for SNP genotyping. Genome-wide SNP genotyping was performed using the Illumina Omni 1 platform in stage 1, and the SequenomiPLEX system (Sequenom, Inc., San Diego, CA, USA) was used in stage 2. Polymerase chain reaction and extension primers were designed using Mass ARRAY Assay Design 3.1 software (Sequenom, Inc.). Genotyping procedures were performed following the manufacturer’s iPLEX Application Guide (Sequenom Inc.).

In stage 1, quality control (QC) procedures were first applied to 2012 individuals. Total 1999 samples passed the call rate of 95% and were included in the final GWAS analysis. Those SNPs were filtered if they showed violation of the Hardy–Weinberg equilibrium (P<0.001), genotype call rates <95%, or minor allele frequency <1%. Following these criteria, 709211 SNPs were retained. We used the IMPUTE computer program36 to infer the genotypes of SNPs that were not directly genotyped with a posterior probability of >0.90 applying to call genotypes. After applying the same QC criteria, as used above, a total of 1940243 SNPs remained in the final analysis.

Statistical analysis

Log-transformed values of IgG were used to ensure a normal distribution and to make the results more robust. In order to identify the effect of smoking or drinking on IgG level, univariate t-test was used to compare the difference of IgG level in subgroups divided by smoking or drinking. The linear regression model implemented in PLINK 37 was used to analyze the association between quantitative traits of IgG and SNP genotypes. Population stratification was estimated by a principal component approach, as implemented in EIGENSTRAT software.38 The top two Eigens were adjusted as covariates in the linear regression model.

Multivariate linear regression analysis was applied to test the independence of the respective SNPs. Only the SNPs that remained significant at 10–6 in the multivariate analysis were selected. The combined analysis of two-stage data was performed using a linear regression, adjusting for the covariates (population stratification, age and smoking) and stage information.

The effect of genotypes and environmental factor (smoking) on serum IgG level was investigated through a linear regression model, in which the dependent variable was log-transformed IgG level and the independent variables were the genotypes and the environmental factor (smoking). Covariates that were adjusted in the individual SNP analysis were also included in this model. Smoking was tested as binary variables (yes/no).


Conflict of interest

The authors declare no conflict of interest.



  1. Gornik O, Pavic T, Lauc G. Alternative glycosylation modulates function of IgG and other proteins - Implications on evolution and disease. Biochim Biophys Acta (e-pub ahead of print 13 December 2011; doi:10.1016/j.bbagen.2011.12.004). | Article |
  2. Palmeira P, Quinello C, Silveira-Lessa AL, Zago CA, Carneiro-Sampaio M. IgG placental transfer in healthy and pathological pregnancies. Clin Dev Immunol 2012; 2012: 985646. | Article | PubMed |
  3. Albert H, Collin M, Dudziak D, Ravetch JV, Nimmerjahn F. In vivo enzymatic modulation of IgG glycosylation inhibits autoimmune disease in an IgG subclass-dependent manner. Proc Natl Acad Sci USA 2008; 105: 15005–15009. | Article | PubMed |
  4. Allansmith M, McClellan BH, Butterworth M, Maloney JR. The development of immunoglobulin levels in man. J Pediatr 1968; 72: 276–290. | Article | PubMed |
  5. Allansmith M, McClellan B, Butterworth M. Stability of human immunoglobulin levels. Proc Soc Exp Biol Med 1967; 125: 404–407. | PubMed |
  6. Hatagima A, Cabello PH, Krieger H. Causal analysis of the variability of IgA, IgG, and IgM immunoglobulin levels. Hum Biol 1999; 71: 219–229. | PubMed |
  7. Gunsolley JC, Pandey JP, Quinn SM, Tew J, Schenkein HA. The effect of race, smoking and immunoglobulin allotypes on IgG subclass concentrations. J Periodontal Res 1997; 32: 381–387. | Article | PubMed |
  8. Lange A, Smolik R, Zatonski W, Szymanska J. Serum immunoglobulin levels in workers exposed to benzene, toluene and xylene. Int Arch Arbeitsmed 1973; 31: 37–44. | Article | PubMed |
  9. Gonzalez-Quintela A, Alende R, Gude F, Campos J, Rey J, Meijide LM et al. Serum levels of immunoglobulins (IgG, IgA, IgM) in a general adult population and their relationship with alcohol consumption, smoking and common metabolic abnormalities. Clin Exp Immunol 2008; 151: 42–50. | Article | PubMed |
  10. Kalff MW, Hijmans W. Serum immunoglobulin levels in twins. Clin Exp Immunol 1969; 5: 469–477. | PubMed |
  11. Kohler PF, Rivera VJ, Eckert ED, Bouchard TJ, Heston LL. Genetic regulation of immunoglobulin and specific antibody levels in twins reared apart. J Clin Invest 1985; 75: 883–888. | Article | PubMed |
  12. Allansmith M, McClellan B, Butterworth M. The influence of heredity and environment on human immunoglobulin levels. J Immunol 1969; 102: 1504–1510. | PubMed |
  13. Grundbacher FJ. Heritability estimates and genetic and environmental correlations for the human immunoglobulins G, M, and A. Am J Hum Genet 1974; 26: 1–12. | PubMed | CAS |
  14. Salzer U, Chapel HM, Webster AD, Pan-Hammarstrom Q, Schmitt-Graeff A, Schlesier M et al. Mutations in TNFRSF13B encoding TACI are associated with common variable immunodeficiency in humans. Nat Genet 2005; 37: 820–828. | Article | PubMed | ISI | CAS |
  15. Castigli E, Wilson SA, Garibyan L, Rachid R, Bonilla F, Schneider L et al. TACI is mutant in common variable immunodeficiency and IgA deficiency. Nat Genet 2005; 37: 829–834. | Article | PubMed | ISI | CAS |
  16. Visscher PM, Montgomery GW. Genome-wide association studies and human disease: from trickle to flood. JAMA 2009; 302: 2028–2029. | Article | PubMed | ISI |
  17. Chan KY, Wong CM, Kwan JS, Lee JM, Cheung KW, Yuen MF et al. Genome-wide association study of hepatocellular carcinoma in Southern Chinese patients with chronic hepatitis B virus infection. PLoS One 2011; 6: e28798. | Article | PubMed |
  18. Tan A, Sun J, Xia N, Qin X, Hu Y, Zhang S et al. A genome-wide association and gene-environment interaction study for serum triglycerides levels in a healthy Chinese male population. Hum Mol Genet 2012; 21: 1658–1664. | Article | PubMed |
  19. Avery DT, Kalled SL, Ellyard JI, Ambrose C, Bixler SA, Thien M et al. BAFF selectively enhances the survival of plasmablasts generated from human memory B cells. J Clin Invest 2003; 112: 286–297. | Article | PubMed | ISI | CAS |
  20. Darce JR, Arendt BK, Wu X, Jelinek DF. Regulated expression of BAFF-binding receptors during human B cell differentiation. J Immunol 2007; 179: 7276–7286. | PubMed | CAS |
  21. Salzer U, Jennings S, Grimbacher B. To switch or not to switch--the opposing roles of TACI in terminal B cell differentiation. Eur J Immunol 2007; 37: 17–20. | Article | PubMed |
  22. Castigli E, Wilson SA, Scott S, Dedeoglu F, Xu S, Lam KP et al. TACI and BAFF-R mediate isotype switching in B cells. J Exp Med 2005; 201: 35–39. | Article | PubMed | ISI | CAS |
  23. He B, Xu W, Santini PA, Polydorides AD, Chiu A, Estrella J et al. Intestinal Bacteria Trigger T Cell-Independent Immunoglobulin A2 Class Switching by Inducing Epithelial-Cell Secretion of the Cytokine APRIL. Immunity 2007; 26: 812–826. | Article | PubMed | ISI | CAS |
  24. von Bulow GU, Russell H, Copeland NG, Gilbert DJ, Jenkins NA, Bram RJ. Molecular cloning and functional characterization of murine transmembrane activator and CAML interactor (TACI) with chromosomal localization in human and mouse. Mamm Genome 2000; 11: 628–632. | Article | PubMed | ISI | CAS |
  25. Sazzini M, Zuntini R, Farjadian S, Quinti I, Ricci G, Romeo G et al. An evolutionary approach to the medical implications of the tumor necrosis factor receptor superfamily member 13B (TNFRSF13B) gene. Genes Immun 2009; 10: 566–578. | Article | PubMed |
  26. von Bülow G-U, van Deursen JM, Bram RJ. Regulation of the T-Independent Humoral Response by TACI. Immunity 2001; 14: 573–582. | Article | PubMed | ISI | CAS |
  27. Kanswal S, Katsenelson N, Selvapandiyan A, Bram RJ, Akkoyunlu M. Deficient TACI expression on B lymphocytes of newborn mice leads to defective Ig secretion in response to BAFF or APRIL. J Immunol 2008; 181: 976–990. | PubMed | CAS |
  28. Sakurai D, Kanno Y, Hase H, Kojima H, Okumura K, Kobata T. TACI attenuates antibody production costimulated by BAFF-R and CD40. Eur J Immunol 2007; 37: 110–118. | Article | PubMed | CAS |
  29. Yan M, Wang H, Chan B, Roose-Girma M, Erickson S, Baker T et al. Activation and accumulation of B cells in TACI-deficient mice. Nat Immunol 2001; 2: 638–643. | Article | PubMed | ISI | CAS |
  30. Poodt AE, Driessen GJ, de Klein A, van Dongen JJ, van der Burg M, de Vries E. TACI mutations and disease susceptibility in patients with common variable immunodeficiency. Clin Exp Immunol 2009; 156: 35–39. | Article | PubMed |
  31. Holt PG. Immune and inflammatory function in cigarette smokers. Thorax 1987; 42: 241–249. | Article | PubMed | ISI | CAS |
  32. Conley ME. Genetics of hypogammaglobulinemia: what do we really know? Curr Opin Immunol 2009; 21: 466–471. | Article | PubMed |
  33. Bacchelli C, Buckland KF, Buckridge S, Salzer U, Schneider P, Thrasher AJ et al. The C76R transmembrane activator and calcium modulator cyclophilin ligand interactor mutation disrupts antibody production and B-cell homeostasis in heterozygous and homozygous mice. J Allergy Clin Immunol 2011; 127: 1253–1259 e13. | Article | PubMed |
  34. Speletas M, Mamara A, Papadopoulou-Alataki E, Iordanakis G, Liadaki K, Bardaka F et al. TNFRSF13B/TACI alterations in Greek patients with antibody deficiencies. J Clin Immunol 2011; 31: 550–559. | Article | PubMed |
  35. Tan A, Gao Y, Yang X, Zhang H, Qin X, Mo L et al. Low serum osteocalcin level is a potential marker for metabolic syndrome: results from a Chinese male population survey. Metabolism 2011; 60: 1186–1192. | Article | PubMed |
  36. Marchini J, Howie B, Myers S, McVean G, Donnelly P. A new multipoint method for genome-wide association studies by imputation of genotypes. Nat Genet 2007; 39: 906–913. | Article | PubMed | ISI | CAS |
  37. Purcell S, Neale B, Todd-Brown K, Thomas L, Ferreira MA, Bender D et al. PLINK: a tool set for whole-genome association and population-based linkage analyses. Am J Hum Genet 2007; 81: 559–575. | Article | PubMed | ISI | CAS |
  38. Price AL, Patterson NJ, Plenge RM, Weinblatt ME, Shadick NA, Reich D. Principal components analysis corrects for stratification in genome-wide association studies. Nat Genet 2006; 38: 904–909. | Article | PubMed | ISI | CAS |


We express our sincere thanks to the local research teams from Fangchenggang First People’s Hospital, Fangchenggang, China, for the recruitment of study subjects. We thank Y-M Wu, Y Sun and Z-X Li, for their assistance in the study recruitment. We thank X-W Zou, H-C Zheng and O Li at the Genergy Biotechnology (Shanghai) Co., Ltd, for their assistance in the genotyping. Finally, we thank all study subjects for participating in the study. The work described in this article is supported by grants from the National Natural Science Foundation of China (Grant no. 30945204) and the Provincial departments of Finance and Education, Guangxi Zhuang Autonomous Region, China (Grant no. 2009GJCJ150).

Supplementary Information accompanies the paper on Genes and Immunity website