JECH Online First, published on April 15, 2014 as 10.1136/jech-2013-203451 Research report
Genetically predicted 17β-estradiol and systemic inflammation in women: a separate-sample Mendelian randomisation analysis in the Guangzhou Biobank Cohort Study J Zhao,1 C Q Jiang,2 T H Lam,1 B Liu,2 K K Cheng,3 S Kavikondala,1 W S Zhang,2 G M Leung,1 C M Schooling1,4 ▸ Additional material is published online only. To view please visit the journal online (http://dx.doi.org/10.1136/jech2013-203451). 1
School of Public Health, Li Ka Shing Faculty of Medicine, The University of Hong Kong, Hong Kong SAR 2 Guangzhou Number 12 Hospital, Guangzhou, China 3 Department of Public Health and Epidemiology, University of Birmingham, UK 4 CUNY School of Public Health and Hunter College, New York, USA Correspondence to Dr CM Schooling, G/F Patrick Manson Building (North Wing), 7 Sassoon Road, Hong Kong; [email protected] Received 29 September 2013 Revised 11 March 2014 Accepted 26 March 2014
ABSTRACT Background Many chronic diseases are characterised by low-grade systemic inflammation. Oestrogens may promote immune response consistent with sex-specific patterns of diseases. In vitro culture and animal experiments suggest oestrogens are anti-inflammatory and might thereby protect against low-grade systemic inflammation. Evidence from epidemiological studies is limited. Using a Mendelian randomisation analysis with a separate-sample instrumental variable (SSIV) estimator, we examined the association of genetically predicted 17β-estradiol with well-established systemic inflammatory markers (total white cell count, granulocyte and lymphocyte count). Methods A genetic score predicting 17β-estradiol was developed in 237 young Chinese women (university students) from Hong Kong based on a parsimonious set of genetic polymorphisms (ESR1 (rs2175898) and CYP19A1 (rs1008805)). Multivariable linear regression was used to examine the association of genetically predicted 17β-estradiol with systemic inflammatory markers among 3096 older (50+ years) Chinese women from the Guangzhou Biobank Cohort Study. Results Predicted 17β-estradiol was negatively associated with white blood cell count (−6.3 103/mL, 95% CI −11.4 to −1.3) and granulocyte count (−4.5 103/mL, 95% CI −8.5 to −0.4) but not lymphocyte count (−1.5 103/mL, 95% CI −3.4 to 0.4) adjusted for age only. Results were similar further adjusted for education, smoking, use of alcohol, physical activity, Body Mass Index, waist-hip ratio, age of menarche, age at menopause, use of hormonal contraceptives and hormone replacement therapy. Conclusions Endogenous genetically predicted 17β-estradiol reduced low-grade systemic inflammatory markers (white blood cell count and granulocyte count), consistent with experimental and ecological evidence of 17β-estradiol promoting immune response. Replication in a larger sample is required.
INTRODUCTION To cite: Zhao J, Jiang CQ, Lam TH, et al. J Epidemiol Community Health Published Online First: [ please include Day Month Year] doi:10.1136/jech-2013203451
Low-grade systemic inflammation is increasingly recognised as an important component of noncommunicable complex, chronic diseases for which anti-inflammatories are a promising new therapy.1 Sex-specific patterns of immune function-related disease suggest a role of sex steroids in immune function. Generally women mount a more vigorous immune response than men, and are more vulnerable
Zhao J, et al. J Epidemiol Community Health(or 2014;0:1–6. doi:10.1136/jech-2013-203451 Copyright Article author their employer) 2014. Produced
to autoimmune disease,2 with the difference emerging at puberty, while men are more vulnerable to infections.3 Anti-inflammatory effects of oestrogens have been found in mechanistic studies in vitro and in animal experiments,4–6 mediated by oestrogen receptors.7 They act as transcriptional factors in the signalling of oestrogens’ anti-inflammatory responses7 in white blood cell counts,8 including granulocytes7 and lymphocytes.8 However, despite the potential etiologic relevance to the inflammatory component of non-communicable diseases, possibly due to immune dysfunction,1 and to auto-immune diseases, the effect of oestrogens on markers of lowgrade systemic inflammation in humans, such as white blood cell and/or its differentials, has rarely been examined. Hormone replacement therapy (HRT) typically combined oestrogen with progesterone,9–11 and so cannot determine the effect of oestrogens. To our knowledge, no large-scale randomised controlled trial (RCT) has reported the effect of oestrogen monotherapy on white blood cells and/or its differentials. A clinical investigation found the increase in blood monocyte count after menopause might be reversed by oestrogen replacement therapy.12 By contrast, one cross-sectional study showed no association,13 and another a positive association of 17β-estradiol with white blood cell count.14 However, observational studies are open to residual confounding and reverse causality. The known harms of oestrogen therapy for cardiovascular disease15 preclude an RCT examining the effects of oestrogen on inflammation. A Mendelian randomisation (MR) design, using sex-hormone-related genetic variants as an instrumental variable, provides a means of assessing the causal effect of oestrogens on systemic inflammatory markers. Genetic variants determined at conception, resulting in life-long differences in endogenous exposures, are unlikely to be associated with socioeconomic position or lifestyle. However, oestrogens decrease with age,16 so oestrogens at older ages may not be a good indicator of lifetime endogenous exposure. Using an MR analysis with a separate-sample instrumental variable (SSIV) estimator, we examined the association of peak oestrogen (genetically predicted 17β-estradiol) with well-established markers of systemic inflammation (total white blood cells, granulocyte, and lymphocyte counts) among women.
by BMJ Publishing Group Ltd under licence.
1
Research report MATERIALS AND METHODS Study design A separate-sample two-stage MR analysis was used. First, a genetic score predicting serum 17β-estradiol was developed in young Chinese women from Hong Kong based on a parsimonious set of genetic polymorphisms from selected single nucleotide polymorphisms (SNPs). Second, we examined the association of genetically predicted 17β-estradiol with total white blood cell count, lymphocyte, and granulocyte count among older Chinese women from the Guangzhou Biobank Cohort Study (GBCS).17
Sources of data Two groups of women of different ages from the same genetic background, that is, from Hong Kong and Guangzhou, the capital of Guangdong, in Southern China, were recruited. Most Hong Kong residents are first, second or third generation migrants from Guangdong.18 First, 237 young women students (mean age 21.0 years) were recruited from the University of Hong Kong, restricted to those with parents and at least three grandparents born in Hong Kong or Guangdong and not taking hormone-related medication. Morning blood samples were taken on the 4th day to 7th day of the menstrual cycle for 17β-estradiol assessment, by immunoassay (Ortho Clinical Diagnostics Vitros Eci), and DNA extraction. Self-administered questionnaires were used to collect information, such as socioeconomic position and health status. Second, we used a sample of older women (50+ years) from GBCS, an ongoing collaboration of Guangzhou Number 12 Hospital, the Universities of Hong Kong and Birmingham, UK.17 Recruitment of participants was in 3 phases. All participants were permanent residents of Guangzhou and members of ‘The Guangzhou Health and Happiness Association for the Respectable Elders’ (GHHARE), a community social and welfare association unofficially aligned with the municipal government. Membership is open to older people for a monthly fee of 4 Yuan (50 US cents). About 7% of permanent Guangzhou residents aged 50+ years are members of GHHARE, of whom 11% (about 10 000 participants) enrolled for each of phases one, two and three. The inclusion criteria were that they were capable of consenting, ambulatory, and not receiving treatment modalities which, if omitted, may result in immediate lifethreatening risk, such as chemotherapy or radiotherapy for cancer, or dialysis for renal failure. Fasting blood samples were collected at recruitment in phase 3 or at follow-up for participants recruited in other phases. Samples were stored, as whole blood or as buffy coat and sera, at −80°C for all, apart from a subset of phase 3 participants whose DNA was extracted from fresh blood and stored at −80°C.19 The University of Hong Kong-Hospital Authority Hong Kong West Cluster Joint Institutional Review Board approved the study. The Guangzhou Medical Ethics Committee of the Chinese Medical Association approved GBCS, including the use of genetic data. All participants gave written, informed consent prior to participation.
DNA extraction and analysis of SNPs DNA was extracted using QIAamp DNA Blood Midi Kit (Catalog No. 51185) for fresh blood in Hong Kong, phenolchloroform extraction for fresh blood in GBCS and magnetic bead extraction for previously stored specimens in GBCS.19 SNPs were selected from genes (ESR1, ESR2 and CYP19A1) functionally relevant to estradiol or breast cancer,20–23 with minor allele frequency >5% in Chinese24 (see online appendix 2
figure 1). Five SNPs (rs2175898 from ESR1, rs1256030 and rs1256031 from ESR2, and rs10046 and rs1008805 from CYP19A1) were analysed at the Centre for Genomic Sciences of the University of Hong Kong, for the Hong Kong sample, and a commercial company (Beijing Capital Bio Corporation) in Beijing, for the GBCS sample, using a Mass ARRAY system (Sequenom, San Diego, California). For DNA quality analysis we used spectrophotometry for most of the samples and gel electrophoresis for four duplicate check controls and six randomly selected samples in each DNA sample plate. The determined sample concentration and A260/280 ratios were 10– 20 ng/μL and 1.7–2.0, respectively. A call rate 95%.
Statistical analysis Hardy–Weinberg equilibrium was tested with an exact test on a contingency table of observed-versus-expected frequencies. Linkage disequilibrium among SNPs in each gene was tested with Haploview 4.2. SNPs with D’ close to 1.0 and r2 greater than 0.7 were considered to be strongly linked, suggesting that one SNP can represent another. Stepwise linear regression with all candidate SNPs (excluding those deviating from Hardy– Weinburg equilibrium, or in linkage disequilibrium) was used to find a parsimonious set of SNPs which best predicted log 17β-estradiol in the young women, because the distribution of 17β-estradiol was skewed. Replication in 1000 bootstrapping samples was used for internal validation. The significance level was set at 0.20 to ensure the initial inclusion criterion was not too restrictive. In the older women, estimated log 17β-estradiol was calculated according to the genetic prediction equation established in the young women. Each value of genetically predicted log 17β-estradiol was antilogged to give genetically predicted 17β-estradiol, for ease of interpretation. Multivariable linear regression was used to assess the adjusted association of genetically predicted 17β-estradiol with markers of systemic inflammation. Although genetically predicted 17β-estradiol should not be confounded, potential confounders were included to achieve more precise estimates.25 Model 1 adjusted for age. Model 2 additionally adjusted for education, smoking, use of alcohol and physical activity. Model 3 additionally adjusted for Body Mass Index and waist-hip ratio. Model 4 additionally adjusted for age of menarche, age at menopause, use of hormonal contraceptives and HRT. Model 5 used bootstrapping with 1000 replications for internal validation. We assessed whether the association of genetically predicted 17β-estradiol with inflammation varied with age from the relevant interaction term. All statistical analyses were conducted using Stata 10.1 (StataCorp LP, College Station, Texas, USA).
RESULTS Genetic prediction of 17β-estradiol in young women Among the 237 young women, none of the SNPs (rs2175898, rs1256031, rs1256030, rs10046 and rs1008805) deviated from Hardy–Weinberg equilibrium (see online appendix table 1); rs1256031 and rs1256030 were in linkage disequilibrium; only rs1256031 was kept. The four remaining SNPs were included in a stepwise linear regression of log 17β-estradiol; rs1008805 in CYP19A1 and rs2175898 in ESR1 were selected for genetic prediction of log 17β-estradiol as 0.1×rs1008805-0.1×rs2175898 +4.7 (see online appendix tables 2 and 3). Genotypes were coded 0, 1 and 2, and the number of major alleles was used as a continuous variable because the R2 was larger than using the genotypes as categorical. The proportion of variance in log Zhao J, et al. J Epidemiol Community Health 2014;0:1–6. doi:10.1136/jech-2013-203451
Research report Table 1 Mean and SD of genetically predicted 17β-estradiol by sociodemographic characteristics in 3096 Women (50+), Guangzhou Biobank Cohort Study, 2003–2008 Genetically predicted 17β-estradiol (pmol/L) Characteristics
n
Age group, years 50–54 55–59 60–64 65–69 70–74 75–79 ≥80 Education Less than primary school Primary school Junior middle school Senior middle school Junior college College Smoking status Never smoker Ex-smoker Current smoker Use of alcohol Never 10 indicates adequate strength of the genetic instrument. Replication using an unweighted genetic score and bootstrapping also provides some validation. The selected SNPs would not be expected to affect inflammation directly (pleiotropy) other than via oestrogen, because the SNPs are from genes functionally relevant to estradiol, and no genome-wide association study (GWAS) indicates an association with inflammatory markers. The genetic score was not associated with testosterone (data not shown), or confounders (table 1). Eighth, MR assumes linearity.46 However, there was no evidence of a nonlinear association, and results were similar assuming a log-linear association. Ninth, genetically predicted 17β-estradiol is different from serum 17β-estradiol measurement and only explained a small proportion of the variance in the exposure, so MR estimates are less precise (although less biased) than those from traditional regression.30 Although adjustment for potential confounders may improve precision,25 replication in a larger sample is required. Tenth, we cannot determine the clinical significance of the small effect size, because MR estimates should be interpreted as hypothesis testing, rather than indicating the size of causal effects.47 However, that does not preclude relevance from other perspectives. Zhao J, et al. J Epidemiol Community Health 2014;0:1–6. doi:10.1136/jech-2013-203451
Research report From a public health perspective, our findings of genetically predicted 17β-estradiol promoting immune response provide a novel theoretically based explanation for the difference in autoimmune disease susceptibility between China and the West,48 with corresponding etiologic implications for prevention and treatment in both settings. Chinese women have lower 17β-estradiol levels than women in the West,49 which may contribute to reduced immune response. This difference may partially explain why Chinese women are less vulnerable to autoimmune diseases, such as rheumatoid arthritis,48 than Western women. Moreover, very preliminary evidence suggests these differences may partially arise in childhood,50 with corresponding implications for intervention. Given that inflammation is an important component of non-communicable chronic diseases,1 experimental studies and our findings might indicate a potential beneficial effect of oestrogens on these diseases. However, such an effect has not been substantiated in RCTs,16 17 most likely because any beneficial effects of oestrogen on inflammation are outweighed by adverse cardiovascular effects via other mechanisms, rather than casting doubt on the causal role of low-grade systemic inflammation in complex, chronic non-communicable diseases. Ongoing RCTs of antiinflammatory therapies operating by other mechanisms51 52 will provide more definitive evidence concerning the causal role of immune function and systemic inflammation in cardiovascular diseases. Our findings suggest endogenous genetically predicted 17β-estradiol may reduce markers of low-grade systemic inflammation, such as white blood cell and granulocyte count, consistent with experimental and ecological evidence indicating 17β-estradiol promotes immune response. Replication in a larger sample is required.
Acknowledgements The Guangzhou Biobank Cohort Study investigators include: Guangzhou No 12 Hospital, Dr Zhang WS, Dr Zhu T, Dr Liu B, Prof. Jiang CQ (Co-Principal Investigator (PI)); University of Hong Kong-Dr C M Schooling, Professor SM McGhee SM, Prof R Fielding, Prof GM Leung, Prof TH Lam (Co-PI); University of Birmingham, Dr G N Thomas, Dr P Adab, Professor KK Cheng (Co-PI). The authors thank Drs Au Yeung SL (PhD, the University of Hong Kong) for convoking the student sample, Drs Garcia-Barcelo, So Man, Xu L (PhD, the University of Hong Kong) and Dr Zhu T (Master of Immunology) for facilitating the DNA extraction. Contributors CMS designed this specific study and acquired some data with the help of SK. THL, KKC and CQJ designed the original study and acquired the GBCS data with the help of GML, WSZ and BL. JZ implemented the analysis, interpreted the data and drafted the manuscript with assistance from CMS. All authors revised the manuscript critically for important intellectual content; and gave final approval of the version to be submitted. Funding This work was supported by the University of Hong Kong Foundation for Development and Research (Hong Kong, China); the University of Hong Kong University Research Committee Strategic Research Theme of Public Health (Hong Kong, China); Guangzhou Public Health Bureau (Guangzhou, China), Guangzhou Science and Technology Bureau (Guangzhou, China) and the University of Birmingham (Birmingham, UK). This substudy was funded by the Research Grant Council General Research Fund (grant number 769710), Research Grant Council of Hong Kong, Hong Kong SAR, People’s Republic of China. The funders had no role in data collection and analysis, the decision to publish, or preparation of the manuscript. Competing interests None. Patient consent Obtained. Ethics approval The University of Hong Kong-Hospital Authority Hong Kong West Cluster Joint Institutional Review Board; The Guangzhou Medical Ethics Committee of the Chinese Medical Association. Provenance and peer review Not commissioned; externally peer reviewed.
REFERENCES 1 2
3
What is already known on this subject? ▸ Many complex, chronic diseases are characterised by low-grade systemic inflammation. ▸ Some experimental studies suggest oestrogens are anti-inflammatory and might thereby protect against low-grade systemic inflammation. ▸ Evidence from trials and epidemiological studies is limited.
4
5
6
7
8
What this study adds? ▸ Using a separate sample Mendelian randomisation analysis in Chinese women, we found a negative association of endogenous genetically predicted 17β-estradiol with markers of low-grade systemic inflammation. ▸ The negative association of endogenous genetically predicted 17β-estradiol with white blood cell count and granulocyte count is consistent with experimental and ecological evidence of 17β-estradiol promoting immune cell function. ▸ From a public health perspective, our findings provide a novel theoretically based explanation for the difference in autoimmune disease susceptibility between China and the West, with corresponding etiologic implications for prevention and treatment in both settings. Zhao J, et al. J Epidemiol Community Health 2014;0:1–6. doi:10.1136/jech-2013-203451
9
10 11 12
13 14
15
16
Tabas I, Glass CK. Anti-inflammatory therapy in chronic disease: challenges and opportunities. Science 2013;339:166–72. Ansar Ahmed S, Penhale WJ, Talal N. Sex hormones, immune responses, and autoimmune diseases. Mechanisms of sex hormone action. Am J Pathol 1985;121:531–51. Marriott I, Huet-Hudson YM. Sexual dimorphism in innate immune responses to infectious organisms. Immunol Res 2006;34:177–92. Nadkarni S, Cooper D, Brancaleone V, et al. Activation of the annexin A1 pathway underlies the protective effects exerted by estrogen in polymorphonuclear leukocytes. Arterioscler Thromb Vasc Biol 2011;31:2749–59. Abu-Taha M, Rius C, Hermenegildo C, et al. Menopause and ovariectomy cause a low grade of systemic inflammation that may be prevented by chronic treatment with low doses of estrogen or losartan. J Immunol 2009;183:1393–402. Hughes EL, Cover PO, Buckingham JC, et al. Role and interactions of annexin A1 and oestrogens in the manifestation of sexual dimorphisms in cerebral and systemic inflammation. Br J Pharmacol 2013;169:539–53. Molero L, Garcia-Duran M, Diaz-Recasens J, et al. Expression of estrogen receptor subtypes and neuronal nitric oxide synthase in neutrophils from women and men: regulation by estrogen. Cardiovasc Res 2002;56:43–51. Cunningham M, Gilkeson G. Estrogen receptors in immunity and autoimmunity. Clin Rev Allergy Immunol 2011;40:66–73. Fahlman MM, Boardley D, Flynn MG, et al. Effects of hormone replacement therapy on selected indices of immune function in postmenopausal women. Gynecol Obstet Invest 2000;50:189–93. Bergstrom H, Tonstad S. Combined hormone replacement therapy improves lipids but may increase leucocytes. J Intern Med 2006;259:537–8. Meng YX, Jiang HY, Chen AJ, et al. Hemostatic changes in women using a monthly injectable contraceptive for one year. Contraception 1990;42:455–66. Ben-Hur H, Mor G, Insler V, et al. Menopause is associated with a significant increase in blood monocyte number and a relative decrease in the expression of estrogen receptors in human peripheral monocytes. Am J Reprod Immunol 1995;34:363–9. Frohlich M, Muhlberger N, Hanke H, et al. Markers of inflammation in women on different hormone replacement therapies. Ann Med 2003;35:353–61. Folsom AR, Golden SH, Boland LL, et al. Association of endogenous hormones with C-reactive protein, fibrinogen, and white blood count in post-menopausal women. Eur J Epidemiol 2005;20:1015–22. Rossouw JE, Anderson GL, Prentice RL, et al. Risks and benefits of estrogen plus progestin in healthy postmenopausal women: principal results From the Women’s Health Initiative randomized controlled trial. JAMA 2002;288:321–33. Straub RH. The complex role of estrogens in inflammation. Endocr Rev 2007;28:521–74.
5
Research report 17
18 19
20
21
22
23
24
25 26
27 28
29 30 31 32 33
34 35
6
Jiang C, Thomas GN, Lam TH, et al. Cohort profile: the Guangzhou Biobank Cohort Study, a Guangzhou-Hong Kong-Birmingham collaboration. Int J Epidemiol 2006;35:844–52. Tsang S. A modern history of Hong Kong. Hong Kong: Hong Kong University Press, 2004, xi, 340. Au Yeung SL, Jiang CQ, Cheng KK, et al. Evaluation of moderate alcohol use and cognitive function among men using a Mendelian randomization design in the Guangzhou biobank cohort study. Am J Epidemiol 2012;175:1021–8. Breast and Prostate Cancer Cohort Consortium, Cox DG, Bretsky P, et al. Haplotypes of the estrogen receptor beta gene and breast cancer risk. Int J Cancer 2008;122:387–92. Lundin E, Wirgin I, Lukanova A, et al. Selected polymorphisms in sex hormone-related genes, circulating sex hormones and risk of endometrial cancer. Cancer Epidemiol 2012;36:445–52. Udler MS, Azzato EM, Healey CS, et al. Common germline polymorphisms in COMT, CYP19A1, ESR1, PGR, SULT1E1 and STS and survival after a diagnosis of breast cancer. Int J Cancer 2009;125:2687–96. Talbott KE, Gammon MD, Kibriya MG, et al. A CYP19 (aromatase) polymorphism is associated with increased premenopausal breast cancer risk. Breast Cancer Res Treat 2008;111:481–7. National Center for Biotechnology Information, US National Library of Medicine. NCBI Single Nucleotide Polymorphism database. http://www.ncbi.nlm.nih.gov/snp (accessed 10 Dec 2013). Burgess S, Thompson SG, Collaboration CCG. Avoiding bias from weak instruments in Mendelian randomization studies. Int J Epidemiol 2011;40:755–64. Naessen T, Bergquist J, Lind L, et al. Higher endogenous estrogen levels in 70-year-old women and men: an endogenous response to counteract developing atherosclerosis? Menopause 2012;19:1322–8. Ovesen L. Adolescence: a critical period for long-term tracking of risk for coronary heart disease? Ann Nutr Metab 2006;50:317–24. Kadlubar FF, Berkowitz GS, Delongchamp RR, et al. The CYP3A4*1B variant is related to the onset of puberty, a known risk factor for the development of breast cancer. Cancer Epidemiol Biomarkers Prev 2003;12:327–31. MacMahon B, Cole P, Brown JB, et al. Urine oestrogen profiles of Asian and North American women. Int J Cancer 1974;14:161–7. Sheehan NA, Didelez V, Burton PR, et al. Mendelian randomisation and causal inference in observational epidemiology. PLoS Med 2008;5:e177. Cutolo M, Sulli A, Capellino S, et al. Sex hormones influence on the immune system: basic and clinical aspects in autoimmunity. Lupus 2004;13:635–8. Mundy GR. Osteoporosis and inflammation. Nutr Rev 2007;65(12 Pt 2):S147–51. Xia X, Zhang S, Yu Y, et al. Effects of estrogen replacement therapy on estrogen receptor expression and immunoregulatory cytokine secretion in surgically induced menopausal women. J Reprod Immunol 2009;81:89–96. Jain SK, Kannan K, Prouty L, et al. Progesterone, but not 17beta-estradiol, increases TNF-alpha secretion in U937 monocytes. Cytokine 2004;26:102–5. Hou L, Lloyd-Jones DM, Ning H, et al. White blood cell count in young adulthood and coronary artery calcification in early middle age: coronary artery risk development in young adults (CARDIA) study. Eur J Epidemiol 2013;28:735–42.
36
37 38
39
40
41
42 43
44
45
46
47 48
49 50
51
52
Danesh J, Collins R, Appleby P, et al. Association of fibrinogen, C-reactive protein, albumin, or leukocyte count with coronary heart disease: meta-analyses of prospective studies. JAMA 1998;279:1477–82. Kolaczkowska E, Kubes P. Neutrophil recruitment and function in health and inflammation. Nat Rev Immunol 2013;13:159–75. Li H, Beeghly-Fadiel A, Wen W, et al. Gene-environment interactions for breast cancer risk among Chinese women: a report from the Shanghai Breast Cancer Genetics Study. Am J Epidemiol 2013;177:161–70. Tchetgen Tchetgen EJ, Walter S, Glymour MM. Commentary: building an evidence base for mendelian randomization studies: assessing the validity and strength of proposed genetic instrumental variables. Int J Epidemiol 2013;42:328–31. Dunning AM, Dowsett M, Healey CS, et al. Polymorphisms associated with circulating sex hormone levels in postmenopausal women. J Natl Cancer Inst 2004;96:936–45. Haiman CA, Dossus L, Setiawan VW, et al. Genetic variation at the CYP19A1 locus predicts circulating estrogen levels but not breast cancer risk in postmenopausal women. Cancer Res 2007;67:1893–7. Bennett DA. An introduction to instrumental variables--part 2: mendelian randomisation. Neuroepidemiology 2010;35:307–10. Burgess S, Butterworth A, Thompson SG. Mendelian randomization analysis with multiple genetic variants using summarized data. Genet Epidemiol 2013;37:658–65. Schuit SC, de Jong FH, Stolk L, et al. Estrogen receptor alpha gene polymorphisms are associated with estradiol levels in postmenopausal women. Eur J Endocrinol 2005;153:327–34. Olson SH, Bandera EV, Orlow I. Variants in estrogen biosynthesis genes, sex steroid hormone levels, and endometrial cancer: a HuGE review. Am J Epidemiol 2007;165:235–45. Freeman G, Cowling BJ, Schooling CM. Power and sample size calculations for Mendelian randomization studies using one genetic instrument. Int J Epidemiol 2013;42:1157–63. Schooling CM, Au Yeung SL, Freeman G. Mendelian randomization estimates may be inflated. J Am Coll Cardiol 2013;61:1931. Dai SM, Han XH, Zhao DB, et al. Prevalence of rheumatic symptoms, rheumatoid arthritis, ankylosing spondylitis, and gout in Shanghai, China: a COPCORD study. J Rheumatol 2003;30:2245–51. Key TJ, Chen J, Wang DY, et al. Sex hormones in women in rural China and in Britain. Br J Cancer 1990;62:631–6. Wong IO, Cowling BJ, Leung GM, et al. Trends in mortality from septicaemia and pneumonia with economic development: an age-period-cohort analysis. PLoS ONE 2012;7:e38988. Ridker PM. Closing the loop on inflammation and atherothrombosis: why perform the cirt and cantos trials? Trans Am Clin Climatol Assoc 2013;124:174–90. Goldfine AB, Conlin PR, Halperin F, et al. A randomised trial of salsalate for insulin resistance and cardiovascular risk factors in persons with abnormal glucose tolerance. Diabetologia 2013;56:714–23.
Zhao J, et al. J Epidemiol Community Health 2014;0:1–6. doi:10.1136/jech-2013-203451
Genetically predicted 17β-estradiol and systemic inflammation in women: a separate-sample Mendelian randomisation analysis in the Guangzhou Biobank Cohort Study J Zhao,1 C Q Jiang,2 T H Lam,1 B Liu,2 K K Cheng,3 S Kavikondala,1 W S Zhang,2 G M Leung,1 C M Schooling1,4 ▸ Additional material is published online only. To view please visit the journal online (http://dx.doi.org/10.1136/jech2013-203451). 1
School of Public Health, Li Ka Shing Faculty of Medicine, The University of Hong Kong, Hong Kong SAR 2 Guangzhou Number 12 Hospital, Guangzhou, China 3 Department of Public Health and Epidemiology, University of Birmingham, UK 4 CUNY School of Public Health and Hunter College, New York, USA Correspondence to Dr CM Schooling, G/F Patrick Manson Building (North Wing), 7 Sassoon Road, Hong Kong; [email protected] Received 29 September 2013 Revised 11 March 2014 Accepted 26 March 2014
ABSTRACT Background Many chronic diseases are characterised by low-grade systemic inflammation. Oestrogens may promote immune response consistent with sex-specific patterns of diseases. In vitro culture and animal experiments suggest oestrogens are anti-inflammatory and might thereby protect against low-grade systemic inflammation. Evidence from epidemiological studies is limited. Using a Mendelian randomisation analysis with a separate-sample instrumental variable (SSIV) estimator, we examined the association of genetically predicted 17β-estradiol with well-established systemic inflammatory markers (total white cell count, granulocyte and lymphocyte count). Methods A genetic score predicting 17β-estradiol was developed in 237 young Chinese women (university students) from Hong Kong based on a parsimonious set of genetic polymorphisms (ESR1 (rs2175898) and CYP19A1 (rs1008805)). Multivariable linear regression was used to examine the association of genetically predicted 17β-estradiol with systemic inflammatory markers among 3096 older (50+ years) Chinese women from the Guangzhou Biobank Cohort Study. Results Predicted 17β-estradiol was negatively associated with white blood cell count (−6.3 103/mL, 95% CI −11.4 to −1.3) and granulocyte count (−4.5 103/mL, 95% CI −8.5 to −0.4) but not lymphocyte count (−1.5 103/mL, 95% CI −3.4 to 0.4) adjusted for age only. Results were similar further adjusted for education, smoking, use of alcohol, physical activity, Body Mass Index, waist-hip ratio, age of menarche, age at menopause, use of hormonal contraceptives and hormone replacement therapy. Conclusions Endogenous genetically predicted 17β-estradiol reduced low-grade systemic inflammatory markers (white blood cell count and granulocyte count), consistent with experimental and ecological evidence of 17β-estradiol promoting immune response. Replication in a larger sample is required.
INTRODUCTION To cite: Zhao J, Jiang CQ, Lam TH, et al. J Epidemiol Community Health Published Online First: [ please include Day Month Year] doi:10.1136/jech-2013203451
Low-grade systemic inflammation is increasingly recognised as an important component of noncommunicable complex, chronic diseases for which anti-inflammatories are a promising new therapy.1 Sex-specific patterns of immune function-related disease suggest a role of sex steroids in immune function. Generally women mount a more vigorous immune response than men, and are more vulnerable
Zhao J, et al. J Epidemiol Community Health(or 2014;0:1–6. doi:10.1136/jech-2013-203451 Copyright Article author their employer) 2014. Produced
to autoimmune disease,2 with the difference emerging at puberty, while men are more vulnerable to infections.3 Anti-inflammatory effects of oestrogens have been found in mechanistic studies in vitro and in animal experiments,4–6 mediated by oestrogen receptors.7 They act as transcriptional factors in the signalling of oestrogens’ anti-inflammatory responses7 in white blood cell counts,8 including granulocytes7 and lymphocytes.8 However, despite the potential etiologic relevance to the inflammatory component of non-communicable diseases, possibly due to immune dysfunction,1 and to auto-immune diseases, the effect of oestrogens on markers of lowgrade systemic inflammation in humans, such as white blood cell and/or its differentials, has rarely been examined. Hormone replacement therapy (HRT) typically combined oestrogen with progesterone,9–11 and so cannot determine the effect of oestrogens. To our knowledge, no large-scale randomised controlled trial (RCT) has reported the effect of oestrogen monotherapy on white blood cells and/or its differentials. A clinical investigation found the increase in blood monocyte count after menopause might be reversed by oestrogen replacement therapy.12 By contrast, one cross-sectional study showed no association,13 and another a positive association of 17β-estradiol with white blood cell count.14 However, observational studies are open to residual confounding and reverse causality. The known harms of oestrogen therapy for cardiovascular disease15 preclude an RCT examining the effects of oestrogen on inflammation. A Mendelian randomisation (MR) design, using sex-hormone-related genetic variants as an instrumental variable, provides a means of assessing the causal effect of oestrogens on systemic inflammatory markers. Genetic variants determined at conception, resulting in life-long differences in endogenous exposures, are unlikely to be associated with socioeconomic position or lifestyle. However, oestrogens decrease with age,16 so oestrogens at older ages may not be a good indicator of lifetime endogenous exposure. Using an MR analysis with a separate-sample instrumental variable (SSIV) estimator, we examined the association of peak oestrogen (genetically predicted 17β-estradiol) with well-established markers of systemic inflammation (total white blood cells, granulocyte, and lymphocyte counts) among women.
by BMJ Publishing Group Ltd under licence.
1
Research report MATERIALS AND METHODS Study design A separate-sample two-stage MR analysis was used. First, a genetic score predicting serum 17β-estradiol was developed in young Chinese women from Hong Kong based on a parsimonious set of genetic polymorphisms from selected single nucleotide polymorphisms (SNPs). Second, we examined the association of genetically predicted 17β-estradiol with total white blood cell count, lymphocyte, and granulocyte count among older Chinese women from the Guangzhou Biobank Cohort Study (GBCS).17
Sources of data Two groups of women of different ages from the same genetic background, that is, from Hong Kong and Guangzhou, the capital of Guangdong, in Southern China, were recruited. Most Hong Kong residents are first, second or third generation migrants from Guangdong.18 First, 237 young women students (mean age 21.0 years) were recruited from the University of Hong Kong, restricted to those with parents and at least three grandparents born in Hong Kong or Guangdong and not taking hormone-related medication. Morning blood samples were taken on the 4th day to 7th day of the menstrual cycle for 17β-estradiol assessment, by immunoassay (Ortho Clinical Diagnostics Vitros Eci), and DNA extraction. Self-administered questionnaires were used to collect information, such as socioeconomic position and health status. Second, we used a sample of older women (50+ years) from GBCS, an ongoing collaboration of Guangzhou Number 12 Hospital, the Universities of Hong Kong and Birmingham, UK.17 Recruitment of participants was in 3 phases. All participants were permanent residents of Guangzhou and members of ‘The Guangzhou Health and Happiness Association for the Respectable Elders’ (GHHARE), a community social and welfare association unofficially aligned with the municipal government. Membership is open to older people for a monthly fee of 4 Yuan (50 US cents). About 7% of permanent Guangzhou residents aged 50+ years are members of GHHARE, of whom 11% (about 10 000 participants) enrolled for each of phases one, two and three. The inclusion criteria were that they were capable of consenting, ambulatory, and not receiving treatment modalities which, if omitted, may result in immediate lifethreatening risk, such as chemotherapy or radiotherapy for cancer, or dialysis for renal failure. Fasting blood samples were collected at recruitment in phase 3 or at follow-up for participants recruited in other phases. Samples were stored, as whole blood or as buffy coat and sera, at −80°C for all, apart from a subset of phase 3 participants whose DNA was extracted from fresh blood and stored at −80°C.19 The University of Hong Kong-Hospital Authority Hong Kong West Cluster Joint Institutional Review Board approved the study. The Guangzhou Medical Ethics Committee of the Chinese Medical Association approved GBCS, including the use of genetic data. All participants gave written, informed consent prior to participation.
DNA extraction and analysis of SNPs DNA was extracted using QIAamp DNA Blood Midi Kit (Catalog No. 51185) for fresh blood in Hong Kong, phenolchloroform extraction for fresh blood in GBCS and magnetic bead extraction for previously stored specimens in GBCS.19 SNPs were selected from genes (ESR1, ESR2 and CYP19A1) functionally relevant to estradiol or breast cancer,20–23 with minor allele frequency >5% in Chinese24 (see online appendix 2
figure 1). Five SNPs (rs2175898 from ESR1, rs1256030 and rs1256031 from ESR2, and rs10046 and rs1008805 from CYP19A1) were analysed at the Centre for Genomic Sciences of the University of Hong Kong, for the Hong Kong sample, and a commercial company (Beijing Capital Bio Corporation) in Beijing, for the GBCS sample, using a Mass ARRAY system (Sequenom, San Diego, California). For DNA quality analysis we used spectrophotometry for most of the samples and gel electrophoresis for four duplicate check controls and six randomly selected samples in each DNA sample plate. The determined sample concentration and A260/280 ratios were 10– 20 ng/μL and 1.7–2.0, respectively. A call rate 95%.
Statistical analysis Hardy–Weinberg equilibrium was tested with an exact test on a contingency table of observed-versus-expected frequencies. Linkage disequilibrium among SNPs in each gene was tested with Haploview 4.2. SNPs with D’ close to 1.0 and r2 greater than 0.7 were considered to be strongly linked, suggesting that one SNP can represent another. Stepwise linear regression with all candidate SNPs (excluding those deviating from Hardy– Weinburg equilibrium, or in linkage disequilibrium) was used to find a parsimonious set of SNPs which best predicted log 17β-estradiol in the young women, because the distribution of 17β-estradiol was skewed. Replication in 1000 bootstrapping samples was used for internal validation. The significance level was set at 0.20 to ensure the initial inclusion criterion was not too restrictive. In the older women, estimated log 17β-estradiol was calculated according to the genetic prediction equation established in the young women. Each value of genetically predicted log 17β-estradiol was antilogged to give genetically predicted 17β-estradiol, for ease of interpretation. Multivariable linear regression was used to assess the adjusted association of genetically predicted 17β-estradiol with markers of systemic inflammation. Although genetically predicted 17β-estradiol should not be confounded, potential confounders were included to achieve more precise estimates.25 Model 1 adjusted for age. Model 2 additionally adjusted for education, smoking, use of alcohol and physical activity. Model 3 additionally adjusted for Body Mass Index and waist-hip ratio. Model 4 additionally adjusted for age of menarche, age at menopause, use of hormonal contraceptives and HRT. Model 5 used bootstrapping with 1000 replications for internal validation. We assessed whether the association of genetically predicted 17β-estradiol with inflammation varied with age from the relevant interaction term. All statistical analyses were conducted using Stata 10.1 (StataCorp LP, College Station, Texas, USA).
RESULTS Genetic prediction of 17β-estradiol in young women Among the 237 young women, none of the SNPs (rs2175898, rs1256031, rs1256030, rs10046 and rs1008805) deviated from Hardy–Weinberg equilibrium (see online appendix table 1); rs1256031 and rs1256030 were in linkage disequilibrium; only rs1256031 was kept. The four remaining SNPs were included in a stepwise linear regression of log 17β-estradiol; rs1008805 in CYP19A1 and rs2175898 in ESR1 were selected for genetic prediction of log 17β-estradiol as 0.1×rs1008805-0.1×rs2175898 +4.7 (see online appendix tables 2 and 3). Genotypes were coded 0, 1 and 2, and the number of major alleles was used as a continuous variable because the R2 was larger than using the genotypes as categorical. The proportion of variance in log Zhao J, et al. J Epidemiol Community Health 2014;0:1–6. doi:10.1136/jech-2013-203451
Research report Table 1 Mean and SD of genetically predicted 17β-estradiol by sociodemographic characteristics in 3096 Women (50+), Guangzhou Biobank Cohort Study, 2003–2008 Genetically predicted 17β-estradiol (pmol/L) Characteristics
n
Age group, years 50–54 55–59 60–64 65–69 70–74 75–79 ≥80 Education Less than primary school Primary school Junior middle school Senior middle school Junior college College Smoking status Never smoker Ex-smoker Current smoker Use of alcohol Never 10 indicates adequate strength of the genetic instrument. Replication using an unweighted genetic score and bootstrapping also provides some validation. The selected SNPs would not be expected to affect inflammation directly (pleiotropy) other than via oestrogen, because the SNPs are from genes functionally relevant to estradiol, and no genome-wide association study (GWAS) indicates an association with inflammatory markers. The genetic score was not associated with testosterone (data not shown), or confounders (table 1). Eighth, MR assumes linearity.46 However, there was no evidence of a nonlinear association, and results were similar assuming a log-linear association. Ninth, genetically predicted 17β-estradiol is different from serum 17β-estradiol measurement and only explained a small proportion of the variance in the exposure, so MR estimates are less precise (although less biased) than those from traditional regression.30 Although adjustment for potential confounders may improve precision,25 replication in a larger sample is required. Tenth, we cannot determine the clinical significance of the small effect size, because MR estimates should be interpreted as hypothesis testing, rather than indicating the size of causal effects.47 However, that does not preclude relevance from other perspectives. Zhao J, et al. J Epidemiol Community Health 2014;0:1–6. doi:10.1136/jech-2013-203451
Research report From a public health perspective, our findings of genetically predicted 17β-estradiol promoting immune response provide a novel theoretically based explanation for the difference in autoimmune disease susceptibility between China and the West,48 with corresponding etiologic implications for prevention and treatment in both settings. Chinese women have lower 17β-estradiol levels than women in the West,49 which may contribute to reduced immune response. This difference may partially explain why Chinese women are less vulnerable to autoimmune diseases, such as rheumatoid arthritis,48 than Western women. Moreover, very preliminary evidence suggests these differences may partially arise in childhood,50 with corresponding implications for intervention. Given that inflammation is an important component of non-communicable chronic diseases,1 experimental studies and our findings might indicate a potential beneficial effect of oestrogens on these diseases. However, such an effect has not been substantiated in RCTs,16 17 most likely because any beneficial effects of oestrogen on inflammation are outweighed by adverse cardiovascular effects via other mechanisms, rather than casting doubt on the causal role of low-grade systemic inflammation in complex, chronic non-communicable diseases. Ongoing RCTs of antiinflammatory therapies operating by other mechanisms51 52 will provide more definitive evidence concerning the causal role of immune function and systemic inflammation in cardiovascular diseases. Our findings suggest endogenous genetically predicted 17β-estradiol may reduce markers of low-grade systemic inflammation, such as white blood cell and granulocyte count, consistent with experimental and ecological evidence indicating 17β-estradiol promotes immune response. Replication in a larger sample is required.
Acknowledgements The Guangzhou Biobank Cohort Study investigators include: Guangzhou No 12 Hospital, Dr Zhang WS, Dr Zhu T, Dr Liu B, Prof. Jiang CQ (Co-Principal Investigator (PI)); University of Hong Kong-Dr C M Schooling, Professor SM McGhee SM, Prof R Fielding, Prof GM Leung, Prof TH Lam (Co-PI); University of Birmingham, Dr G N Thomas, Dr P Adab, Professor KK Cheng (Co-PI). The authors thank Drs Au Yeung SL (PhD, the University of Hong Kong) for convoking the student sample, Drs Garcia-Barcelo, So Man, Xu L (PhD, the University of Hong Kong) and Dr Zhu T (Master of Immunology) for facilitating the DNA extraction. Contributors CMS designed this specific study and acquired some data with the help of SK. THL, KKC and CQJ designed the original study and acquired the GBCS data with the help of GML, WSZ and BL. JZ implemented the analysis, interpreted the data and drafted the manuscript with assistance from CMS. All authors revised the manuscript critically for important intellectual content; and gave final approval of the version to be submitted. Funding This work was supported by the University of Hong Kong Foundation for Development and Research (Hong Kong, China); the University of Hong Kong University Research Committee Strategic Research Theme of Public Health (Hong Kong, China); Guangzhou Public Health Bureau (Guangzhou, China), Guangzhou Science and Technology Bureau (Guangzhou, China) and the University of Birmingham (Birmingham, UK). This substudy was funded by the Research Grant Council General Research Fund (grant number 769710), Research Grant Council of Hong Kong, Hong Kong SAR, People’s Republic of China. The funders had no role in data collection and analysis, the decision to publish, or preparation of the manuscript. Competing interests None. Patient consent Obtained. Ethics approval The University of Hong Kong-Hospital Authority Hong Kong West Cluster Joint Institutional Review Board; The Guangzhou Medical Ethics Committee of the Chinese Medical Association. Provenance and peer review Not commissioned; externally peer reviewed.
REFERENCES 1 2
3
What is already known on this subject? ▸ Many complex, chronic diseases are characterised by low-grade systemic inflammation. ▸ Some experimental studies suggest oestrogens are anti-inflammatory and might thereby protect against low-grade systemic inflammation. ▸ Evidence from trials and epidemiological studies is limited.
4
5
6
7
8
What this study adds? ▸ Using a separate sample Mendelian randomisation analysis in Chinese women, we found a negative association of endogenous genetically predicted 17β-estradiol with markers of low-grade systemic inflammation. ▸ The negative association of endogenous genetically predicted 17β-estradiol with white blood cell count and granulocyte count is consistent with experimental and ecological evidence of 17β-estradiol promoting immune cell function. ▸ From a public health perspective, our findings provide a novel theoretically based explanation for the difference in autoimmune disease susceptibility between China and the West, with corresponding etiologic implications for prevention and treatment in both settings. Zhao J, et al. J Epidemiol Community Health 2014;0:1–6. doi:10.1136/jech-2013-203451
9
10 11 12
13 14
15
16
Tabas I, Glass CK. Anti-inflammatory therapy in chronic disease: challenges and opportunities. Science 2013;339:166–72. Ansar Ahmed S, Penhale WJ, Talal N. Sex hormones, immune responses, and autoimmune diseases. Mechanisms of sex hormone action. Am J Pathol 1985;121:531–51. Marriott I, Huet-Hudson YM. Sexual dimorphism in innate immune responses to infectious organisms. Immunol Res 2006;34:177–92. Nadkarni S, Cooper D, Brancaleone V, et al. Activation of the annexin A1 pathway underlies the protective effects exerted by estrogen in polymorphonuclear leukocytes. Arterioscler Thromb Vasc Biol 2011;31:2749–59. Abu-Taha M, Rius C, Hermenegildo C, et al. Menopause and ovariectomy cause a low grade of systemic inflammation that may be prevented by chronic treatment with low doses of estrogen or losartan. J Immunol 2009;183:1393–402. Hughes EL, Cover PO, Buckingham JC, et al. Role and interactions of annexin A1 and oestrogens in the manifestation of sexual dimorphisms in cerebral and systemic inflammation. Br J Pharmacol 2013;169:539–53. Molero L, Garcia-Duran M, Diaz-Recasens J, et al. Expression of estrogen receptor subtypes and neuronal nitric oxide synthase in neutrophils from women and men: regulation by estrogen. Cardiovasc Res 2002;56:43–51. Cunningham M, Gilkeson G. Estrogen receptors in immunity and autoimmunity. Clin Rev Allergy Immunol 2011;40:66–73. Fahlman MM, Boardley D, Flynn MG, et al. Effects of hormone replacement therapy on selected indices of immune function in postmenopausal women. Gynecol Obstet Invest 2000;50:189–93. Bergstrom H, Tonstad S. Combined hormone replacement therapy improves lipids but may increase leucocytes. J Intern Med 2006;259:537–8. Meng YX, Jiang HY, Chen AJ, et al. Hemostatic changes in women using a monthly injectable contraceptive for one year. Contraception 1990;42:455–66. Ben-Hur H, Mor G, Insler V, et al. Menopause is associated with a significant increase in blood monocyte number and a relative decrease in the expression of estrogen receptors in human peripheral monocytes. Am J Reprod Immunol 1995;34:363–9. Frohlich M, Muhlberger N, Hanke H, et al. Markers of inflammation in women on different hormone replacement therapies. Ann Med 2003;35:353–61. Folsom AR, Golden SH, Boland LL, et al. Association of endogenous hormones with C-reactive protein, fibrinogen, and white blood count in post-menopausal women. Eur J Epidemiol 2005;20:1015–22. Rossouw JE, Anderson GL, Prentice RL, et al. Risks and benefits of estrogen plus progestin in healthy postmenopausal women: principal results From the Women’s Health Initiative randomized controlled trial. JAMA 2002;288:321–33. Straub RH. The complex role of estrogens in inflammation. Endocr Rev 2007;28:521–74.
5
Research report 17
18 19
20
21
22
23
24
25 26
27 28
29 30 31 32 33
34 35
6
Jiang C, Thomas GN, Lam TH, et al. Cohort profile: the Guangzhou Biobank Cohort Study, a Guangzhou-Hong Kong-Birmingham collaboration. Int J Epidemiol 2006;35:844–52. Tsang S. A modern history of Hong Kong. Hong Kong: Hong Kong University Press, 2004, xi, 340. Au Yeung SL, Jiang CQ, Cheng KK, et al. Evaluation of moderate alcohol use and cognitive function among men using a Mendelian randomization design in the Guangzhou biobank cohort study. Am J Epidemiol 2012;175:1021–8. Breast and Prostate Cancer Cohort Consortium, Cox DG, Bretsky P, et al. Haplotypes of the estrogen receptor beta gene and breast cancer risk. Int J Cancer 2008;122:387–92. Lundin E, Wirgin I, Lukanova A, et al. Selected polymorphisms in sex hormone-related genes, circulating sex hormones and risk of endometrial cancer. Cancer Epidemiol 2012;36:445–52. Udler MS, Azzato EM, Healey CS, et al. Common germline polymorphisms in COMT, CYP19A1, ESR1, PGR, SULT1E1 and STS and survival after a diagnosis of breast cancer. Int J Cancer 2009;125:2687–96. Talbott KE, Gammon MD, Kibriya MG, et al. A CYP19 (aromatase) polymorphism is associated with increased premenopausal breast cancer risk. Breast Cancer Res Treat 2008;111:481–7. National Center for Biotechnology Information, US National Library of Medicine. NCBI Single Nucleotide Polymorphism database. http://www.ncbi.nlm.nih.gov/snp (accessed 10 Dec 2013). Burgess S, Thompson SG, Collaboration CCG. Avoiding bias from weak instruments in Mendelian randomization studies. Int J Epidemiol 2011;40:755–64. Naessen T, Bergquist J, Lind L, et al. Higher endogenous estrogen levels in 70-year-old women and men: an endogenous response to counteract developing atherosclerosis? Menopause 2012;19:1322–8. Ovesen L. Adolescence: a critical period for long-term tracking of risk for coronary heart disease? Ann Nutr Metab 2006;50:317–24. Kadlubar FF, Berkowitz GS, Delongchamp RR, et al. The CYP3A4*1B variant is related to the onset of puberty, a known risk factor for the development of breast cancer. Cancer Epidemiol Biomarkers Prev 2003;12:327–31. MacMahon B, Cole P, Brown JB, et al. Urine oestrogen profiles of Asian and North American women. Int J Cancer 1974;14:161–7. Sheehan NA, Didelez V, Burton PR, et al. Mendelian randomisation and causal inference in observational epidemiology. PLoS Med 2008;5:e177. Cutolo M, Sulli A, Capellino S, et al. Sex hormones influence on the immune system: basic and clinical aspects in autoimmunity. Lupus 2004;13:635–8. Mundy GR. Osteoporosis and inflammation. Nutr Rev 2007;65(12 Pt 2):S147–51. Xia X, Zhang S, Yu Y, et al. Effects of estrogen replacement therapy on estrogen receptor expression and immunoregulatory cytokine secretion in surgically induced menopausal women. J Reprod Immunol 2009;81:89–96. Jain SK, Kannan K, Prouty L, et al. Progesterone, but not 17beta-estradiol, increases TNF-alpha secretion in U937 monocytes. Cytokine 2004;26:102–5. Hou L, Lloyd-Jones DM, Ning H, et al. White blood cell count in young adulthood and coronary artery calcification in early middle age: coronary artery risk development in young adults (CARDIA) study. Eur J Epidemiol 2013;28:735–42.
36
37 38
39
40
41
42 43
44
45
46
47 48
49 50
51
52
Danesh J, Collins R, Appleby P, et al. Association of fibrinogen, C-reactive protein, albumin, or leukocyte count with coronary heart disease: meta-analyses of prospective studies. JAMA 1998;279:1477–82. Kolaczkowska E, Kubes P. Neutrophil recruitment and function in health and inflammation. Nat Rev Immunol 2013;13:159–75. Li H, Beeghly-Fadiel A, Wen W, et al. Gene-environment interactions for breast cancer risk among Chinese women: a report from the Shanghai Breast Cancer Genetics Study. Am J Epidemiol 2013;177:161–70. Tchetgen Tchetgen EJ, Walter S, Glymour MM. Commentary: building an evidence base for mendelian randomization studies: assessing the validity and strength of proposed genetic instrumental variables. Int J Epidemiol 2013;42:328–31. Dunning AM, Dowsett M, Healey CS, et al. Polymorphisms associated with circulating sex hormone levels in postmenopausal women. J Natl Cancer Inst 2004;96:936–45. Haiman CA, Dossus L, Setiawan VW, et al. Genetic variation at the CYP19A1 locus predicts circulating estrogen levels but not breast cancer risk in postmenopausal women. Cancer Res 2007;67:1893–7. Bennett DA. An introduction to instrumental variables--part 2: mendelian randomisation. Neuroepidemiology 2010;35:307–10. Burgess S, Butterworth A, Thompson SG. Mendelian randomization analysis with multiple genetic variants using summarized data. Genet Epidemiol 2013;37:658–65. Schuit SC, de Jong FH, Stolk L, et al. Estrogen receptor alpha gene polymorphisms are associated with estradiol levels in postmenopausal women. Eur J Endocrinol 2005;153:327–34. Olson SH, Bandera EV, Orlow I. Variants in estrogen biosynthesis genes, sex steroid hormone levels, and endometrial cancer: a HuGE review. Am J Epidemiol 2007;165:235–45. Freeman G, Cowling BJ, Schooling CM. Power and sample size calculations for Mendelian randomization studies using one genetic instrument. Int J Epidemiol 2013;42:1157–63. Schooling CM, Au Yeung SL, Freeman G. Mendelian randomization estimates may be inflated. J Am Coll Cardiol 2013;61:1931. Dai SM, Han XH, Zhao DB, et al. Prevalence of rheumatic symptoms, rheumatoid arthritis, ankylosing spondylitis, and gout in Shanghai, China: a COPCORD study. J Rheumatol 2003;30:2245–51. Key TJ, Chen J, Wang DY, et al. Sex hormones in women in rural China and in Britain. Br J Cancer 1990;62:631–6. Wong IO, Cowling BJ, Leung GM, et al. Trends in mortality from septicaemia and pneumonia with economic development: an age-period-cohort analysis. PLoS ONE 2012;7:e38988. Ridker PM. Closing the loop on inflammation and atherothrombosis: why perform the cirt and cantos trials? Trans Am Clin Climatol Assoc 2013;124:174–90. Goldfine AB, Conlin PR, Halperin F, et al. A randomised trial of salsalate for insulin resistance and cardiovascular risk factors in persons with abnormal glucose tolerance. Diabetologia 2013;56:714–23.
Zhao J, et al. J Epidemiol Community Health 2014;0:1–6. doi:10.1136/jech-2013-203451
