Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Testing relationship between tea intake and the risk of rheumatoid arthritis and systemic lupus erythematosus: a Mendelian randomization study

Testing relationship between tea intake and the risk of rheumatoid arthritis and systemic lupus... Background We performed Mendelian randomization (MR) to assess the causal effect of tea intake on rheumatoid arthritis (RA) and systemic lupus erythematosus (SLE). Methods Genetic instruments for tea intake were obtained from a large genome‑ wide association study (GWAS) dataset of the UK Biobank. Genetic association estimates for RA (6236 cases and 147,221 controls) and SLE (538 cases and 213,145 controls) were obtained from the FinnGen study through the IEU GWAS database. Results MR analyses using the inverse‑ variance weighted method showed that tea intake was not associated with risk of RA [odds ratio (OR) per standard deviation increment in genetically predicted tea intake = 0.997, 95% confidence interval (CI) 0.658–1.511] and SLE (OR per standard deviation increment in genetically predicted tea intake = 0.961, 95% CI 0.299–3.092). Weighted median, weighted mode, MR‑Egger, leave ‑ one‑ out and multivariable MR controlling for several confounding factors including current tobacco smoking, coffee intake, and alcoholic drinks per week yielded completely consistent results. No evidence of heterogeneity and pleiotropy was found. Conclusion Our MR study did not suggest a causal effect of genetically predicted tea intake on RA and SLE. Keywords Mendelian randomization, Rheumatoid arthritis, Systemic lupus erythematosus, Tea intake sexes and all ethnicities [2]. The hallmark of SLE includes Introduction the production of autoantibodies, aberrant formation of Rheumatoid arthritis (RA) and systemic lupus erythema- immune complexes, increased levels of pro-inflammatory tosus (SLE) are two common immune mediated inflam - cytokines, and chronic inflammation of multiple organ matory disorders. RA affects approximately 0.5–1% of the systems [3]. The pathogenesis of RA and SLE is complex. global population, characterized by symmetrical polyar- Genetic, environmental, and lifestyle factors can affect thritis, chronic synovial inflammation, gradual joint dete - the risk of RA and SLE. Although lifestyle factors such as rioration, and chronic disability [1]. SLE can affect both alcohol drinking and cigarette smoking have been exten- sively studied, the effect of tea intake on the risk of RA *Correspondence: and SLE remains largely unknown. Jian Huang huangjian_chn@163.com Tea is one of the most commonly consumed beverages Department of Orthopedic Trauma and Hand Surgery, The First Affiliated all over the world. It consists of a number of bioactive Hospital of Guangxi Medical University, Nanning 530021, China compounds including polyphenols and caffeine. These Clinical Laboratory Center, The First Affiliated Hospital of Guangxi Medical University, Nanning 530021, China compounds possess antioxidative, anti-inflammatory, © The Author(s) 2023. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/. Lu and Huang Advances in Rheumatology (2023) 63:10 Page 2 of 8 immunosuppressive, and cancer-preventative proper- Biobank Axiom array and imputed against the UK10K, ties [4–7]. Some experimental studies showed that tea 1000 Genomes Phase 3 and Haplotype Reference Con- and its active ingredients such as epigallocatechin-3-gal- sortium panels [26]. The GWAS was adjusted for age, sex, late (EGCG) could ameliorate animal models of RA by genotyping array, and the principal components by the regulating Th17/Treg balance, inducing B cell apopto - original GWAS researchers. sis, decreasing pro-inflammatory cytokine production, Single nucleotide polymorphisms (SNPs) that strongly and inhibiting proliferation of activated synovial fibro - associated with tea intake at genome-wide significance −8 blasts [8–11]. Evidence from experimental studies also (P < 5 × 10 ) were selected as instrumental variables. demonstrated that tea extract alleviated experimental Before performing MR analyses, SNP instruments were lupus nephritis through promotion of the nuclear fac- clumped at r < 0.001 using the R package “TwoSam- tor E2-related factor 2 signaling pathway, inhibition of pleMR” version 0.5.6 (https:// github. com/ MRCIEU/ renal nucleotide-binding domain-like receptor protein 3 TwoSa mpleMR). In addition, palindromic SNPs with activation, and regulation of systemic regulatory T-cell intermediate allele frequencies were removed. Additional activity [12]. However, observational clinical studies eval- file  1: Table S1 shows the information on genetic variants uating the association of tea intake with RA and SLE have included as instruments for tea intake. yielded inconsistent results [13–19]. Current tobacco smoking summary-level data were Mendelian randomization (MR) is an analytical derived from the Medical Research Council-Integrative method applying genetic variants as instrumental vari- Epidemiology Unit (MRC-IEU) consortium (462,434 ables to examine causal associations from observational participants), which is available at https:// gwas. mrcieu. data. Genetic variants are randomly distributed at con-ac. uk/ datas ets/ ukb-b- 223/. Summary statistics for alco- ception based on Mendel’s law. MR analyses are thus less holic drinks per week were obtained from the GWAS susceptible to possible confounding factors and reverse and Sequencing Consortium of Alcohol and Nicotine causation. In the epidemiology field, MR is increasingly use (335,394 participants) at https:// gwas. mrcieu. ac. uk/ used for assessing the causal effect of a risk or protective datas ets/ ieu-b- 73/ [27]. Genetic instruments for cof- factor on an outcome. Prior MR studies have evaluated fee intake were derived from the MRC-IEU consortium the causal relationships of several lifestyle factors such (428,860 participants) at https:// gwas. mrcieu. ac. uk/ datas as coffee intake, alcohol drinking, and smoking with the ets/ ukb-b- 5237/. All data used were from individuals of risk of RA and SLE [20–24], whereas the role of tea intake European ancestry. Additional file  2: Table S2, Additional remains unclear. In the present study, we applied a two- file  3: Table  S3 and Additional file  4: Table  S4 show the sample MR approach to assess the causal effect of geneti - information on genetic variants included as instruments cally predicted tea intake on the risk of RA and SLE. for current tobacco smoking, alcoholic drinks per week, and coffee intake. Methods Two-sample MR was performed using summary-level Genetic data for outcomes data from published genome-wide association studies The association of genetic instruments with RA was (GWASs). Ethical approval was not necessary for this MR retrieved from the FinnGen study (6236 RA cases and study; we did not use individual-level data. 147,221 controls) at https:// gwas. mrcieu. ac. uk/ datas ets/ finn-b- M13_ RHEUMA/. The association of genetic Genetic data for exposures instruments with SLE was also obtained from the The GWAS summary statistics for tea intake in individu - FinnGen study (538 SLE cases and 213,145 controls), als of European ancestry (447,485 participants) based which is available at https:// gwas. mrcieu. ac. uk/ datas on the UK Biobank were obtained from the IEU GWAS ets/ finn-b- M13_ SLE/. Launched in 2017, the FinnGen database at https:// gwas. mrcieu. ac. uk/ datas ets/ ukb- study is a nationwide cohort study which aims to collect b- 6066/. Briefly, the UK Biobank is a large prospective and analyze genome and health data from 500,000 Finn- cohort of more than 500,000 participants (aged 40 to ish biobank participants. RA and SLE were identified 69 years); they were recruited across England, Wales, and according to International classification of diseases (ICD) Scotland between 2006 and 2010 [25]. As part of the UK codes retrieved from nationwide registries in Finland. Biobank diet Survey, self-reported tea intake was ascer- Genetic associations in FinnGen were evaluated with tained using a touchscreen question: “How many cups of adjustment for age, sex, genotype batch, and 10 princi- tea do you drink each day? (include black and green tea)” pal components by the FinnGen researchers. Linkage Participants who answered > 99 or < 0 were excluded, and disequilibrium proxies (r > 0.8) were applied if the SNP those answering > 20 were asked to confirm. Genotyping instruments were missing from the outcome dataset. We of the participants was undertaken using Affymetrix UK L u and Huang Advances in Rheumatology (2023) 63:10 Page 3 of 8 harmonised the exposure and outcome datasets to ensure To further control for potential pleiotropic effects, a that the genetic associations reflect the same effect allele. multivariable MR was carried out [35]. We took into account several covariates including current tobacco smoking, alcoholic drinks per week, and coffee intake. Statistical analysis The inverse-variance weighted method was used for Our MR study is based on three key assumptions (Addi- multivariable MR. All statistical tests were two sided. All tional file  5: Figure S1). To assess if the SNP instruments MR analyses were undertaken using R software version obtained were robustly associated with tea intake, we cal- 4.0.4 (R Foundation for Statistical Computing, Vienna, culated the F statistic as previously described by Noyce Austria). and colleagues [28]. We considered an F-statistic of more than 10 as sufficient to conduct MR analyses, which is Results conventionally accepted in the field [29]. In the main MR As shown in Additional file  1: Table  S1, among genetic analyses, we combined the causal estimates for all instru- variants included as instruments for tea intake, mental variables with the use of the inverse-variance rs2472297 has the strongest association with tea intake –109 weighted method. It provides reliable causal estimates (P = 2.30 × 10 ). The 40 SNPs for tea intake corre - when directional pleiotropy is absent [30]. To assess sponded to a F-statistic of 62.99, explaining 0.56% of the robustness of our main MR results to pleiotropy, the variance in tea intake. In the primary MR analyses, the weighted median, weighted mode, MR-Egger, and tea intake was not associated with risk of RA (OR per MR pleiotropy residual sum and outlier (MR-PRESSO) standard deviation increment in genetically predicted tea methods were used in sensitivity analyses. The weighted intake = 0.997, 95% CI 0.658–1.511, P = 0.988) and SLE median method can provide a robust evaluation as long (OR per standard deviation increment in genetically pre- as at least 50% of the weight stems from valid instru- dicted tea intake = 0.961, 95% CI 0.299–3.092, P = 0.947) ments satisfying the MR assumptions [31]. The weighted using the inverse-variance weighted method (Table 1 and mode method can estimate the causal relationship allow- Fig.  1). The MR-Egger, weighted median and weighted ing for even the majority of instrumental variables to be mode MR estimates yielded similar results to the inverse- pleiotropic [32]. The MR-Egger approach may be used variance weighted analyses (Table  1 and Fig.  1), suggest- even though all genetic variants are invalid instrumental ing that genetically predicted tea intake was not causally variables [30]. We estimated the average directional plei- associated with RA and SLE. In the main MR analyses, otropy across instrumental variables with the intercept we found absence of heterogeneity across the individual P value of the MR-Egger regression. The MR-PRESSO MR estimates derived from the 40 SNPs (tea intake and method (https:// github. com/ rondo lab/ MR- PRESSO) can RA: Q-statistic = 49.267, P = 0.126; tea intake and SLE: detect horizontal pleiotropic outliers [33]. Besides these Q-statistic = 39.853, P = 0.432). The MR-Egger intercept robust methods, to minimize any possible confounding test indicated no evidence of directional pleiotropy (tea introduced by individual SNPs, a leave-one-out analysis intake and RA: intercept of 0.003, P = 0.724; tea intake was conducted in our sensitivity checks. To look for signs and SLE: intercept of 0.007, P = 0.775) (Table 1). In addi- of heterogeneity, Cochran’s Q statistic and I were calcu- tion, using the MR-PRESSO method, we did not find the lated [34]. presence of pleiotropy (tea intake and RA: P-value global Table 1 Univariable MR estimates of genetically predicted tea intake on the risk of rheumatoid arthritis and systemic lupus erythematosus Exposure Outcome Number of MR method OR Association P value Pleiotropy P value Heterogeneity P value instruments 95% CI Egger intercept Q−statistic Tea intake Rheumatoid arthritis 40 Inverse‑ variance weighted 0.997 0.658–1.511 0.988 49.267 0.126 MR−Egger 0.859 0.341–2.162 0.748 0.003 0.724 49.103 0.107 Weighted median 0.770 0.422–1.403 0.393 Weighted mode 0.880 0.429–1.805 0.728 Tea intake Systemic lupus erythematosus 40 Inverse‑ variance weighted 0.961 0.299–3.092 0.947 39.853 0.432 MR−Egger 0.685 0.051–9.154 0.776 0.007 0.775 39.766 0.391 Weighted median 0.407 0.069–2.396 0.320 Weighted mode 0.475 0.072–3.145 0.445 CI confidence interval, OR odds ratio, MR Mendelian randomization Lu and Huang Advances in Rheumatology (2023) 63:10 Page 4 of 8 Fig. 1 Univariable Mendelian randomization analyses for the association of genetically instrumented tea intake with RA and SLE. CI, confidence interval; IVW, inverse‑ variance weighted; OR, odds ratio; RA, rheumatoid arthritis; SLE, systemic lupus erythematosus test 0.122; tea intake and SLE: P-value global test 0.44). In summary, the main MR analyses using the inverse- The leave-one-out analyses did not identify any outliers variance weighted method and sensitivity analyses using for all estimates (Additional file  6: Table  S5), indicating several other reliable MR methods, leave-one-out analy- that the MR results were not driven by any single SNP. ses, and multivariable MR provided consistent results, Additional file  7: Table  S6 shows the inverse-variance suggesting that  genetically predicted  tea intake was not weighted MR results of the causal effect of genetically causally associated with RA and SLE. predicted current tobacco smoking, alcoholic drinks per week, and coffee intake on RA and SLE. Discussion We further carried out a multivariable MR to evaluate Tea intake is among the most widely consumed beverages the causal effect of tea intake on RA and SLE by adjusting in the world. Our study estimated the causal association for current tobacco smoking, alcoholic drinks per week, of tea intake with RA and SLE in individuals of European and coffee intake. The results did not suggest any causal ancestry. Across the primary and all sensitivity MR esti- association of tea intake with RA after adjusting for cur- mates, our study provided no support for a causal effect rent tobacco smoking (OR = 0.992, 95% CI 0.676–1.459, of tea intake on the risk of RA and SLE. P = 0.968), alcoholic drinks per week (OR = 1.046, 95% CI Tea and its extract such as caffeine and polyphenols 0.680–1.613, P = 0.837), and coffee intake (OR = 0.980, have been showed to display antiinflammatory, immuno - 95% CI 0.660–1.191, P = 0.902) (Fig.  2). Similarly, there suppressive, and antioxidative properties. Although tea was no causal association between tea intake and SLE or its extract was reported to ameliorate animal models after adjusting for current tobacco smoking (OR = 1.097, of RA and SLE in some experimental studies, observa- 95% CI 0.343–3.521, P = 0.876), alcoholic drinks per tional clinical studies failed to replicate animal findings in week (OR = 1.173, 95% CI 0.391–3.525, P = 0.774), and humans [14, 16, 36]. Several studies even observed that coffee intake (OR = 1.246, 95% CI 0.347–4.436, P = 0.738) tea intake was associated with increased risk of RA and (Fig. 2). SLE [15, 18, 37]. Observational studies are susceptible to L u and Huang Advances in Rheumatology (2023) 63:10 Page 5 of 8 Fig. 2 Multivariable Mendelian randomization analyses for the association of genetically instrumented tea intake with RA and SLE. CI, confidence interval; OR, odds ratio; RA, rheumatoid arthritis; SLE, systemic lupus erythematosus biases such as residual confounding and selection bias observational studies are not able to determine the causal [38]. It is possible that the prior observational studies association of tea intake with the risk of RA and SLE. may not reliably control for these biases. In some case– Therefore, reliable conclusions may not be drew from control studies, control subjects were not selected for conventional observational studies. matching RA or SLE patients on confounding factors. MR is a promising approach in genetic epidemiology. For example, the study by Kiyohara et  al. [15] recruited It can minimize potential biases in conventional obser- university students and staffs from nursing homes and vational studies and perform causal inference. To clarify clinics as control subjects; they had different character - the causal association, we performed a two-sample MR istics compared with SLE patients. In addition, some using summary-level statistics from large GWAS data- observational studies collected information on tea intake sets. We selected SNPs that were strongly associated with and confounding factors only at baseline, which may lead tea intake as instruments; the potential for weak instru- to misclassification. Furthermore, tea is frequently con - ment bias was reduced since all instruments had suf- sumed in combination with cigarettes and alcohol. It is ficiently large F-statistics (> 10). The main MR analyses thought that conventional observational studies have using the inverse-variance weighted method showed null difficulty in separating the effect of tea from cigarettes causal associations of tea intake with RA and SLE and no and alcohol when evaluating the effect of tea intake evidence of heterogeneity. The sensitivity analyses using on RA and SLE. Besides these shortness, conventional the MR-Egger, weighted median, weighted mode, and Lu and Huang Advances in Rheumatology (2023) 63:10 Page 6 of 8 multivariable MR  analyses yielded completely consist- out completely the possibility that the null association ent results with those of the main MR analyses. The MR- of tea intake with RA and SLE was owing to a small Egger intercept test, the MR-PRESSO global test, and the causal role which was undetectable in this MR analysis. leave-one-out analyses indicated low likelihood of plei- However, it is certain that our MR results did not suffer otropy. Thus, our MR findings were robust against viola - from weak instrument bias since the instruments used tions of the MR assumptions. were strong (F-statistic > 10). Fourthly, since different These findings are in agreement with two prior meta- ethnic populations such as Europeans and Asians have analyses evaluating the association between tea and risk different tea intake habits, more research is required to of RA [39, 40]. Lee et  al. [39] summarized available evi- investigate the generalizability of our results to races dence from three studies (cohort or case–control design) other than Europeans. including 638 patients with RA and 113,998 controls. They found no association between tea intake and the incidence of RA in the overall meta-analysis (relative risk Conclusion (RR) = 0.88, 95% CI = 0.62–1.24, P = 0.463) and subgroup In sum, we did not find any effect of genetically pre - analysis according to study design (case–control design: dicted tea intake on the risk of RA and SLE based on RR = 1.00, 95% CI = 0.58–1.72, P = 1.000; cohort design: summary-level data from large GWAS in individuals of RR = 0.68, 95% CI = 0.27–1.71, P = 0.410). Similarly, the European ancestry. meta-analysis by Asoudeh et  al. [40] using prospective cohort studies (823 RA cases and 191,313 controls) did Supplementary Information not observe significant relation between tea consump - The online version contains supplementary material available at https:// doi. org/ 10. 1186/ s42358‑ 023‑ 00290‑7. tion and the risk of RA (RR = 1.05, 95% CI 0.73–1.53). There were no meta-analyses estimating the relationship Additional file 1: Table S1. Summary results of the SNPs associated with between tea and SLE risk. Our findings were also in line tea intake. with a recently published MR study by Chen and col- Additional file 2: Table S2. Summary results of the SNPs associated with leagues who showed no effect of tea consumption on RA current tobacco smoking. (sample size: 58,284, OR = 1.24, 95% CI 0.81–1.91) [41]. Additional file 3: Table S3. Summary results of the SNPs associated with alcoholic drinks per week. Compared with that MR study, we increased the statisti- cal power by using a much larger study sample (447,485 Additional file 4: Table S4. Summary results of the SNPs associated with coffee intake. participants). Additionally, we controlled for several Additional file 5: Figure S1. Schematic representation of the Mendelian confounding lifestyle factors including smoking, alcohol randomisation design. Our Mendelian randomization study is based on drinking, and coffee consumption using multivariable three key assumptions: (1) instruments are associated with tea intake; (2) MR, which was important but neglected by Chen and instruments have no associations with confounding factors; and (3) instru‑ ments directly affect rheumatoid arthritis and systemic lupus erythemato ‑ coworkers [41]. sus through tea intake. This MR study is subject to some limitations. Firstly, Additional file 6: Table S5. Leave ‑ one‑ out analyses using the IVW we assessed the effects of black tea and green tea method. together but did not separate their effects, due to the Additional file 7: Table S6. The IVW MR estimates on the causal effect questionnaire design in the original GWAS study. Black of genetically predicted current tobacco smoking, alcoholic drinks per tea and green tea differ in terms of chemical composi - week, and coffee intake on rheumatoid arthritis and systemic lupus erythematosus. tion [41]. The major polyphenolic compounds of black tea include thearubigins and theaflavins, whereas green Acknowledgements tea is rich in catechins such as epicatechin-3-gallate Not applicable. and epicatechin and EGCG [42]. Limited evidence showed that consumption of black tea and green tea Author contributions JH, conceptualization; JH, methodology; RBL and JH, resources; RBL and JH, exhibited a different association with SLE risk [15]. We validation and formal analysis; RBL and JH, writing‑ original draft prepara‑ hope that future MR studies can evaluate the causal tion; JH, supervision; JH, writing‑review and editing. Both authors read and effect of black tea and green tea intake on RA and approved the final manuscript. SLE separately when more data becomes available. Funding Secondly, we were not able to assess non-linear asso- There is no funding for this study. ciations since two-sample MR can only assess linear Availability of data and materials associations. Future MR analyses are needed to evalu- This Mendelian randomization study is based on summary‑level data that ate the non-linear relationships of tea intake with RA have been made publicly available in the IEU GWAS database (https:// gwas. and SLE. Thirdly, it is known that one of the shortcom - mrcieu. ac. uk/). ings of MR is limited statistical power. We cannot rule L u and Huang Advances in Rheumatology (2023) 63:10 Page 7 of 8 18. Lamichhane D, Collins C, Constantinescu F, Walitt B, Pettinger M, Parks Declarations C, et al. Coffee and tea consumption in relation to risk of rheumatoid arthritis in the women’s health initiative observational cohort. J Clin Ethics approval and consent to participate Rheumatol. 2019;25(3):127–32. Not applicable. 19. Westerlind H, Palmqvist I, Saevarsdottir S, Alfredsson L, Klareskog L, Di Giuseppe D. Is tea consumption associated with reduction of risk of Consent for publication rheumatoid arthritis? A Swedish case‑ control study. Arthritis Res Ther. Not applicable. 2021;23(1):209. 20. Bae SC, Lee YH. Coffee consumption and the risk of rheumatoid arthritis Competing interests and systemic lupus erythematosus: a Mendelian randomization study. The authors have no competing interests to declare. Clin Rheumatol. 2018;37(10):2875–9. 21. Bae SC, Lee YH. Alcohol intake and risk of rheumatoid arthritis: a Mende‑ lian randomization study. Z Rheumatol. 2019;78(8):791–6. Received: 5 July 2022 Accepted: 24 February 2023 22. Jiang X, Alfredsson L. Modifiable environmental exposure and risk of rheumatoid arthritis‑ current evidence from genetic studies. Arthritis Res Ther. 2020;22(1):154. 23. Jiang X, Zhu Z, Manouchehrinia A, Olsson T, Alfredsson L, Kockum I. Alcohol consumption and risk of common autoimmune inflammatory References diseases‑ evidence from a large‑scale genetic analysis totaling 1 million 1. Cush JJ. Rheumatoid arthritis: early diagnosis and treatment. Rheum Dis individuals. Front Genet. 2021;12:687745. Clin North Am. 2022;48(2):537–47. 24. Wang P, Dan YL, Wu Q, Tao SS, Yang XK, Wang DG, et al. Non‑ causal effects 2. Barber MRW, Drenkard C, Falasinnu T, Hoi A, Mak A, Kow NY, et al. Global of smoking and alcohol use on the risk of systemic lupus erythematosus. epidemiology of systemic lupus erythematosus. Nat Rev Rheumatol. Autoimmun Rev. 2021;20(9):102890. 2021;17(9):515–32. 25. Bycroft C, Freeman C, Petkova D, Band G, Elliott LT, Sharp K, et al. The UK 3. Mathias LM, Stohl W. Systemic lupus erythematosus (SLE): emerging Biobank resource with deep phenotyping and genomic data. Nature. therapeutic targets. Expert Opin Ther Targets. 2020;24(12):1283–302. 2018;562(7726):203–9. 4. Neyestani TR, Gharavi A, Kalayi A. Selective effects of tea extract and its 26. Cole JB, Florez JC, Hirschhorn JN. Comprehensive genomic analysis of phenolic compounds on human peripheral blood mononuclear cell dietary habits in UK Biobank identifies hundreds of genetic associations. cytokine secretions. Int J Food Sci Nutr. 2009;60(Suppl 1):79–88. Nat Commun. 2020;11(1):1467. 5. Peairs A, Dai R, Gan L, Shimp S, Rylander MN, Li L, et al. Epigallocatechin‑ 27. Liu M, Jiang Y, Wedow R, Li Y, Brazel DM, Chen F, et al. Association studies 3‑ gallate (EGCG) attenuates inflammation in MRL/lpr mouse mesangial of up to 1.2 million individuals yield new insights into the genetic etiol‑ cells. Cell Mol Immunol. 2010;7(2):123–32. ogy of tobacco and alcohol use. Nat Genet. 2019;51(2):237–44. 6. Bañuls‑Mirete M, Ogdie A, Guma M. Micronutrients: essential treatment 28. Noyce AJ, Kia DA, Hemani G, Nicolas A, Price TR, De Pablo‑Fernandez for inflammatory arthritis? Curr Rheumatol Rep. 2020;22(12):87. E, et al. Estimating the causal influence of body mass index on risk 7. Song C, Xu S, Chang L, Zhao X, Mei X, Ren X, et al. Preparation of EGCG of Parkinson disease: a Mendelian randomisation study. PLoS Med. decorated, injectable extracellular vesicles for cartilage repair in rat arthri‑ 2017;14(6):e1002314. tis. Regen Biomater. 2021;8(6):67. 29. Burgess S, Thompson SG. CRP CHD genetics collaboration avoiding bias 8. Marotte H, Ruth JH, Campbell PL, Koch AE, Ahmed S. Green tea extract from weak instruments in Mendelian randomization studies. Int J Epide‑ inhibits chemokine production, but up‑regulates chemokine receptor miol. 2011;40(3):755–64. expression, in rheumatoid arthritis synovial fibroblasts and rat adjuvant ‑ 30. Bowden J, Davey Smith G, Haycock PC, Burgess S. Consistent estima‑ induced arthritis. Rheumatology0. 2010;49(3):467–79. tion in mendelian randomization with some invalid instruments using a 9. Liu D, Li P, Song S, Liu Y, Wang Q, Chang Y, et al. Pro‑apoptotic effect of weighted median estimator. Genet Epidemiol. 2016;40(4):304–14. epigallo‑ catechin‑3‑ gallate on B lymphocytes through regulating BAFF/ 31. Burgess S, Bowden J, Fall T, Ingelsson E, Thompson SG. Sensitivity analyses PI3K/Akt/mTOR signaling in rats with collagen‑induced arthritis. Eur J for robust causal inference from mendelian randomization analyses with Pharmacol. 2012;690(1–3):214–25. multiple genetic variants. Epidemiology. 2017;28(1):30–42. 10. Wu D, Wang J, Pae M, Meydani SN. Green tea EGCG, T cells, and T cell‑ 32. Hartwig FP, Davey Smith G, Bowden J. Robust inference in summary data mediated autoimmune diseases. Mol Aspects Med. 2012;33(1):107–18. Mendelian randomization via the zero modal pleiotropy assumption. Int J 11. Riegsecker S, Wiczynski D, Kaplan MJ, Ahmed S. Potential benefits of Epidemiol. 2017;46(6):1985–98. green tea polyphenol EGCG in the prevention and treatment of vascular 33. Verbanck M, Chen CY, Neale B, Do R. Detection of widespread horizontal inflammation in rheumatoid arthritis. Life Sci. 2013;93(8):307–12. pleiotropy in causal relationships inferred from Mendelian randomization 12. Tsai PY, Ka SM, Chang JM, Chen HC, Shui HA, Li CY, et al. Epigallocatechin‑ between complex traits and diseases. Nat Genet. 2018;50(5):693–8. 3‑ gallate prevents lupus nephritis development in mice via enhancing 34. Bowden J, Del Greco MF, Minelli C, Davey Smith G, Sheehan N, Thompson the Nrf2 antioxidant pathway and inhibiting NLRP3 inflammasome J. A framework for the investigation of pleiotropy in two‑sample sum‑ activation. Free Radic Biol Med. 2011;51(3):744–54. mary data Mendelian randomization. Stat Med. 2017;36(11):1783–802. 13. Mikuls TR, Cerhan JR, Criswell LA, Merlino L, Mudano AS, Burma M, 35. Burgess S, Thompson SG. Multivariable Mendelian randomization: the use et al. Coffee, tea, and caffeine consumption and risk of rheumatoid of pleiotropic genetic variants to estimate causal effects. Am J Epidemiol. arthritis: results from the Iowa Women’s Health Study. Arthritis Rheum. 2015;181(4):251–60. 2002;46(1):83–91. 36. Pattison DJ, Symmons DP, Lunt M, Welch A, Luben R, Bingham SA, et al. 14. Karlson EW, Mandl LA, Aweh GN, Grodstein F. Coffee consumption and Dietary risk factors for the development of inflammatory polyarthritis: evi‑ risk of rheumatoid arthritis. Arthritis Rheum. 2003;48(11):3055–60. dence for a role of high level of red meat consumption. Arthritis Rheum. 15. Kiyohara C, Washio M, Horiuchi T, Asami T, Ide S, Atsumi T, et al. Modifying 2004;50(12):3804–12. effect of N‑acetyltransferase 2 genotype on the association between 37. Mazzucca CB, Scotti L, Cappellano G, Barone‑Adesi F, Chiocchetti A. Nutri‑ systemic lupus erythematosus and consumption of alcohol and caffeine ‑ tion and rheumatoid arthritis onset: a prospective analysis using the UK rich beverages. Arthritis Care Res. 2014;66(7):1048–56. biobank. Nutrients. 2022;14(8):1554. 16. Washio M, Fujii T, Kuwana M, Kawaguchi Y, Mimori A, Horiuchi T, et al. 38. Khasawneh LQ, Al‑Mahayri ZN, Ali BR. Mendelian randomization in Lifestyle and other related factors for the development of mixed connec‑ pharmacogenomics: the unforeseen potentials. Biomed Pharmacother. tive tissue disease among Japanese females in comparison with systemic 2022;150:112952. lupus erythematosus. Mod Rheumatol. 2014;24(5):788–92. 39. Lee YH, Bae SC, Song GG. Coffee or tea consumption and the risk of rheu‑ 17. Rambod M, Nazarinia M, Raieskarimian F. The impact of dietary habits matoid arthritis: a meta‑analysis. Clin Rheumatol. 2014;33(11):1575–83. on the pathogenesis of rheumatoid arthritis: a case‑ control study. Clin Rheumatol. 2018;37(10):2643–8. Lu and Huang Advances in Rheumatology (2023) 63:10 Page 8 of 8 40. Asoudeh F, Dashti F, Jayedi A, Hemmati A, Fadel A, Mohammadi H. Caf‑ feine, coffee, tea and risk of rheumatoid arthritis: systematic review and dose‑response meta‑analysis of prospective cohort studies. Front Nutr. 2022;9:822557. 41. Chen S, Chen T, Chen Y, Huang D, Pan Y, Chen S. Causal association between tea consumption and bone health: a mendelian randomization study. Front Nutr. 2022;9:872451. 42. Khan N, Mukhtar H. Tea polyphenols in promotion of human health. Nutrients. 2018;11(1):39. Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in pub‑ lished maps and institutional affiliations. Re Read ady y to to submit y submit your our re researc search h ? Choose BMC and benefit fr ? Choose BMC and benefit from om: : fast, convenient online submission thorough peer review by experienced researchers in your field rapid publication on acceptance support for research data, including large and complex data types • gold Open Access which fosters wider collaboration and increased citations maximum visibility for your research: over 100M website views per year At BMC, research is always in progress. Learn more biomedcentral.com/submissions http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Advances in Rheumatology Springer Journals

Testing relationship between tea intake and the risk of rheumatoid arthritis and systemic lupus erythematosus: a Mendelian randomization study

Advances in Rheumatology , Volume 63 (1) – Mar 10, 2023

Loading next page...
 
/lp/springer-journals/testing-relationship-between-tea-intake-and-the-risk-of-rheumatoid-icMJYhRB4m
Publisher
Springer Journals
Copyright
Copyright © The Author(s) 2023
eISSN
2523-3106
DOI
10.1186/s42358-023-00290-7
Publisher site
See Article on Publisher Site

Abstract

Background We performed Mendelian randomization (MR) to assess the causal effect of tea intake on rheumatoid arthritis (RA) and systemic lupus erythematosus (SLE). Methods Genetic instruments for tea intake were obtained from a large genome‑ wide association study (GWAS) dataset of the UK Biobank. Genetic association estimates for RA (6236 cases and 147,221 controls) and SLE (538 cases and 213,145 controls) were obtained from the FinnGen study through the IEU GWAS database. Results MR analyses using the inverse‑ variance weighted method showed that tea intake was not associated with risk of RA [odds ratio (OR) per standard deviation increment in genetically predicted tea intake = 0.997, 95% confidence interval (CI) 0.658–1.511] and SLE (OR per standard deviation increment in genetically predicted tea intake = 0.961, 95% CI 0.299–3.092). Weighted median, weighted mode, MR‑Egger, leave ‑ one‑ out and multivariable MR controlling for several confounding factors including current tobacco smoking, coffee intake, and alcoholic drinks per week yielded completely consistent results. No evidence of heterogeneity and pleiotropy was found. Conclusion Our MR study did not suggest a causal effect of genetically predicted tea intake on RA and SLE. Keywords Mendelian randomization, Rheumatoid arthritis, Systemic lupus erythematosus, Tea intake sexes and all ethnicities [2]. The hallmark of SLE includes Introduction the production of autoantibodies, aberrant formation of Rheumatoid arthritis (RA) and systemic lupus erythema- immune complexes, increased levels of pro-inflammatory tosus (SLE) are two common immune mediated inflam - cytokines, and chronic inflammation of multiple organ matory disorders. RA affects approximately 0.5–1% of the systems [3]. The pathogenesis of RA and SLE is complex. global population, characterized by symmetrical polyar- Genetic, environmental, and lifestyle factors can affect thritis, chronic synovial inflammation, gradual joint dete - the risk of RA and SLE. Although lifestyle factors such as rioration, and chronic disability [1]. SLE can affect both alcohol drinking and cigarette smoking have been exten- sively studied, the effect of tea intake on the risk of RA *Correspondence: and SLE remains largely unknown. Jian Huang huangjian_chn@163.com Tea is one of the most commonly consumed beverages Department of Orthopedic Trauma and Hand Surgery, The First Affiliated all over the world. It consists of a number of bioactive Hospital of Guangxi Medical University, Nanning 530021, China compounds including polyphenols and caffeine. These Clinical Laboratory Center, The First Affiliated Hospital of Guangxi Medical University, Nanning 530021, China compounds possess antioxidative, anti-inflammatory, © The Author(s) 2023. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/. Lu and Huang Advances in Rheumatology (2023) 63:10 Page 2 of 8 immunosuppressive, and cancer-preventative proper- Biobank Axiom array and imputed against the UK10K, ties [4–7]. Some experimental studies showed that tea 1000 Genomes Phase 3 and Haplotype Reference Con- and its active ingredients such as epigallocatechin-3-gal- sortium panels [26]. The GWAS was adjusted for age, sex, late (EGCG) could ameliorate animal models of RA by genotyping array, and the principal components by the regulating Th17/Treg balance, inducing B cell apopto - original GWAS researchers. sis, decreasing pro-inflammatory cytokine production, Single nucleotide polymorphisms (SNPs) that strongly and inhibiting proliferation of activated synovial fibro - associated with tea intake at genome-wide significance −8 blasts [8–11]. Evidence from experimental studies also (P < 5 × 10 ) were selected as instrumental variables. demonstrated that tea extract alleviated experimental Before performing MR analyses, SNP instruments were lupus nephritis through promotion of the nuclear fac- clumped at r < 0.001 using the R package “TwoSam- tor E2-related factor 2 signaling pathway, inhibition of pleMR” version 0.5.6 (https:// github. com/ MRCIEU/ renal nucleotide-binding domain-like receptor protein 3 TwoSa mpleMR). In addition, palindromic SNPs with activation, and regulation of systemic regulatory T-cell intermediate allele frequencies were removed. Additional activity [12]. However, observational clinical studies eval- file  1: Table S1 shows the information on genetic variants uating the association of tea intake with RA and SLE have included as instruments for tea intake. yielded inconsistent results [13–19]. Current tobacco smoking summary-level data were Mendelian randomization (MR) is an analytical derived from the Medical Research Council-Integrative method applying genetic variants as instrumental vari- Epidemiology Unit (MRC-IEU) consortium (462,434 ables to examine causal associations from observational participants), which is available at https:// gwas. mrcieu. data. Genetic variants are randomly distributed at con-ac. uk/ datas ets/ ukb-b- 223/. Summary statistics for alco- ception based on Mendel’s law. MR analyses are thus less holic drinks per week were obtained from the GWAS susceptible to possible confounding factors and reverse and Sequencing Consortium of Alcohol and Nicotine causation. In the epidemiology field, MR is increasingly use (335,394 participants) at https:// gwas. mrcieu. ac. uk/ used for assessing the causal effect of a risk or protective datas ets/ ieu-b- 73/ [27]. Genetic instruments for cof- factor on an outcome. Prior MR studies have evaluated fee intake were derived from the MRC-IEU consortium the causal relationships of several lifestyle factors such (428,860 participants) at https:// gwas. mrcieu. ac. uk/ datas as coffee intake, alcohol drinking, and smoking with the ets/ ukb-b- 5237/. All data used were from individuals of risk of RA and SLE [20–24], whereas the role of tea intake European ancestry. Additional file  2: Table S2, Additional remains unclear. In the present study, we applied a two- file  3: Table  S3 and Additional file  4: Table  S4 show the sample MR approach to assess the causal effect of geneti - information on genetic variants included as instruments cally predicted tea intake on the risk of RA and SLE. for current tobacco smoking, alcoholic drinks per week, and coffee intake. Methods Two-sample MR was performed using summary-level Genetic data for outcomes data from published genome-wide association studies The association of genetic instruments with RA was (GWASs). Ethical approval was not necessary for this MR retrieved from the FinnGen study (6236 RA cases and study; we did not use individual-level data. 147,221 controls) at https:// gwas. mrcieu. ac. uk/ datas ets/ finn-b- M13_ RHEUMA/. The association of genetic Genetic data for exposures instruments with SLE was also obtained from the The GWAS summary statistics for tea intake in individu - FinnGen study (538 SLE cases and 213,145 controls), als of European ancestry (447,485 participants) based which is available at https:// gwas. mrcieu. ac. uk/ datas on the UK Biobank were obtained from the IEU GWAS ets/ finn-b- M13_ SLE/. Launched in 2017, the FinnGen database at https:// gwas. mrcieu. ac. uk/ datas ets/ ukb- study is a nationwide cohort study which aims to collect b- 6066/. Briefly, the UK Biobank is a large prospective and analyze genome and health data from 500,000 Finn- cohort of more than 500,000 participants (aged 40 to ish biobank participants. RA and SLE were identified 69 years); they were recruited across England, Wales, and according to International classification of diseases (ICD) Scotland between 2006 and 2010 [25]. As part of the UK codes retrieved from nationwide registries in Finland. Biobank diet Survey, self-reported tea intake was ascer- Genetic associations in FinnGen were evaluated with tained using a touchscreen question: “How many cups of adjustment for age, sex, genotype batch, and 10 princi- tea do you drink each day? (include black and green tea)” pal components by the FinnGen researchers. Linkage Participants who answered > 99 or < 0 were excluded, and disequilibrium proxies (r > 0.8) were applied if the SNP those answering > 20 were asked to confirm. Genotyping instruments were missing from the outcome dataset. We of the participants was undertaken using Affymetrix UK L u and Huang Advances in Rheumatology (2023) 63:10 Page 3 of 8 harmonised the exposure and outcome datasets to ensure To further control for potential pleiotropic effects, a that the genetic associations reflect the same effect allele. multivariable MR was carried out [35]. We took into account several covariates including current tobacco smoking, alcoholic drinks per week, and coffee intake. Statistical analysis The inverse-variance weighted method was used for Our MR study is based on three key assumptions (Addi- multivariable MR. All statistical tests were two sided. All tional file  5: Figure S1). To assess if the SNP instruments MR analyses were undertaken using R software version obtained were robustly associated with tea intake, we cal- 4.0.4 (R Foundation for Statistical Computing, Vienna, culated the F statistic as previously described by Noyce Austria). and colleagues [28]. We considered an F-statistic of more than 10 as sufficient to conduct MR analyses, which is Results conventionally accepted in the field [29]. In the main MR As shown in Additional file  1: Table  S1, among genetic analyses, we combined the causal estimates for all instru- variants included as instruments for tea intake, mental variables with the use of the inverse-variance rs2472297 has the strongest association with tea intake –109 weighted method. It provides reliable causal estimates (P = 2.30 × 10 ). The 40 SNPs for tea intake corre - when directional pleiotropy is absent [30]. To assess sponded to a F-statistic of 62.99, explaining 0.56% of the robustness of our main MR results to pleiotropy, the variance in tea intake. In the primary MR analyses, the weighted median, weighted mode, MR-Egger, and tea intake was not associated with risk of RA (OR per MR pleiotropy residual sum and outlier (MR-PRESSO) standard deviation increment in genetically predicted tea methods were used in sensitivity analyses. The weighted intake = 0.997, 95% CI 0.658–1.511, P = 0.988) and SLE median method can provide a robust evaluation as long (OR per standard deviation increment in genetically pre- as at least 50% of the weight stems from valid instru- dicted tea intake = 0.961, 95% CI 0.299–3.092, P = 0.947) ments satisfying the MR assumptions [31]. The weighted using the inverse-variance weighted method (Table 1 and mode method can estimate the causal relationship allow- Fig.  1). The MR-Egger, weighted median and weighted ing for even the majority of instrumental variables to be mode MR estimates yielded similar results to the inverse- pleiotropic [32]. The MR-Egger approach may be used variance weighted analyses (Table  1 and Fig.  1), suggest- even though all genetic variants are invalid instrumental ing that genetically predicted tea intake was not causally variables [30]. We estimated the average directional plei- associated with RA and SLE. In the main MR analyses, otropy across instrumental variables with the intercept we found absence of heterogeneity across the individual P value of the MR-Egger regression. The MR-PRESSO MR estimates derived from the 40 SNPs (tea intake and method (https:// github. com/ rondo lab/ MR- PRESSO) can RA: Q-statistic = 49.267, P = 0.126; tea intake and SLE: detect horizontal pleiotropic outliers [33]. Besides these Q-statistic = 39.853, P = 0.432). The MR-Egger intercept robust methods, to minimize any possible confounding test indicated no evidence of directional pleiotropy (tea introduced by individual SNPs, a leave-one-out analysis intake and RA: intercept of 0.003, P = 0.724; tea intake was conducted in our sensitivity checks. To look for signs and SLE: intercept of 0.007, P = 0.775) (Table 1). In addi- of heterogeneity, Cochran’s Q statistic and I were calcu- tion, using the MR-PRESSO method, we did not find the lated [34]. presence of pleiotropy (tea intake and RA: P-value global Table 1 Univariable MR estimates of genetically predicted tea intake on the risk of rheumatoid arthritis and systemic lupus erythematosus Exposure Outcome Number of MR method OR Association P value Pleiotropy P value Heterogeneity P value instruments 95% CI Egger intercept Q−statistic Tea intake Rheumatoid arthritis 40 Inverse‑ variance weighted 0.997 0.658–1.511 0.988 49.267 0.126 MR−Egger 0.859 0.341–2.162 0.748 0.003 0.724 49.103 0.107 Weighted median 0.770 0.422–1.403 0.393 Weighted mode 0.880 0.429–1.805 0.728 Tea intake Systemic lupus erythematosus 40 Inverse‑ variance weighted 0.961 0.299–3.092 0.947 39.853 0.432 MR−Egger 0.685 0.051–9.154 0.776 0.007 0.775 39.766 0.391 Weighted median 0.407 0.069–2.396 0.320 Weighted mode 0.475 0.072–3.145 0.445 CI confidence interval, OR odds ratio, MR Mendelian randomization Lu and Huang Advances in Rheumatology (2023) 63:10 Page 4 of 8 Fig. 1 Univariable Mendelian randomization analyses for the association of genetically instrumented tea intake with RA and SLE. CI, confidence interval; IVW, inverse‑ variance weighted; OR, odds ratio; RA, rheumatoid arthritis; SLE, systemic lupus erythematosus test 0.122; tea intake and SLE: P-value global test 0.44). In summary, the main MR analyses using the inverse- The leave-one-out analyses did not identify any outliers variance weighted method and sensitivity analyses using for all estimates (Additional file  6: Table  S5), indicating several other reliable MR methods, leave-one-out analy- that the MR results were not driven by any single SNP. ses, and multivariable MR provided consistent results, Additional file  7: Table  S6 shows the inverse-variance suggesting that  genetically predicted  tea intake was not weighted MR results of the causal effect of genetically causally associated with RA and SLE. predicted current tobacco smoking, alcoholic drinks per week, and coffee intake on RA and SLE. Discussion We further carried out a multivariable MR to evaluate Tea intake is among the most widely consumed beverages the causal effect of tea intake on RA and SLE by adjusting in the world. Our study estimated the causal association for current tobacco smoking, alcoholic drinks per week, of tea intake with RA and SLE in individuals of European and coffee intake. The results did not suggest any causal ancestry. Across the primary and all sensitivity MR esti- association of tea intake with RA after adjusting for cur- mates, our study provided no support for a causal effect rent tobacco smoking (OR = 0.992, 95% CI 0.676–1.459, of tea intake on the risk of RA and SLE. P = 0.968), alcoholic drinks per week (OR = 1.046, 95% CI Tea and its extract such as caffeine and polyphenols 0.680–1.613, P = 0.837), and coffee intake (OR = 0.980, have been showed to display antiinflammatory, immuno - 95% CI 0.660–1.191, P = 0.902) (Fig.  2). Similarly, there suppressive, and antioxidative properties. Although tea was no causal association between tea intake and SLE or its extract was reported to ameliorate animal models after adjusting for current tobacco smoking (OR = 1.097, of RA and SLE in some experimental studies, observa- 95% CI 0.343–3.521, P = 0.876), alcoholic drinks per tional clinical studies failed to replicate animal findings in week (OR = 1.173, 95% CI 0.391–3.525, P = 0.774), and humans [14, 16, 36]. Several studies even observed that coffee intake (OR = 1.246, 95% CI 0.347–4.436, P = 0.738) tea intake was associated with increased risk of RA and (Fig. 2). SLE [15, 18, 37]. Observational studies are susceptible to L u and Huang Advances in Rheumatology (2023) 63:10 Page 5 of 8 Fig. 2 Multivariable Mendelian randomization analyses for the association of genetically instrumented tea intake with RA and SLE. CI, confidence interval; OR, odds ratio; RA, rheumatoid arthritis; SLE, systemic lupus erythematosus biases such as residual confounding and selection bias observational studies are not able to determine the causal [38]. It is possible that the prior observational studies association of tea intake with the risk of RA and SLE. may not reliably control for these biases. In some case– Therefore, reliable conclusions may not be drew from control studies, control subjects were not selected for conventional observational studies. matching RA or SLE patients on confounding factors. MR is a promising approach in genetic epidemiology. For example, the study by Kiyohara et  al. [15] recruited It can minimize potential biases in conventional obser- university students and staffs from nursing homes and vational studies and perform causal inference. To clarify clinics as control subjects; they had different character - the causal association, we performed a two-sample MR istics compared with SLE patients. In addition, some using summary-level statistics from large GWAS data- observational studies collected information on tea intake sets. We selected SNPs that were strongly associated with and confounding factors only at baseline, which may lead tea intake as instruments; the potential for weak instru- to misclassification. Furthermore, tea is frequently con - ment bias was reduced since all instruments had suf- sumed in combination with cigarettes and alcohol. It is ficiently large F-statistics (> 10). The main MR analyses thought that conventional observational studies have using the inverse-variance weighted method showed null difficulty in separating the effect of tea from cigarettes causal associations of tea intake with RA and SLE and no and alcohol when evaluating the effect of tea intake evidence of heterogeneity. The sensitivity analyses using on RA and SLE. Besides these shortness, conventional the MR-Egger, weighted median, weighted mode, and Lu and Huang Advances in Rheumatology (2023) 63:10 Page 6 of 8 multivariable MR  analyses yielded completely consist- out completely the possibility that the null association ent results with those of the main MR analyses. The MR- of tea intake with RA and SLE was owing to a small Egger intercept test, the MR-PRESSO global test, and the causal role which was undetectable in this MR analysis. leave-one-out analyses indicated low likelihood of plei- However, it is certain that our MR results did not suffer otropy. Thus, our MR findings were robust against viola - from weak instrument bias since the instruments used tions of the MR assumptions. were strong (F-statistic > 10). Fourthly, since different These findings are in agreement with two prior meta- ethnic populations such as Europeans and Asians have analyses evaluating the association between tea and risk different tea intake habits, more research is required to of RA [39, 40]. Lee et  al. [39] summarized available evi- investigate the generalizability of our results to races dence from three studies (cohort or case–control design) other than Europeans. including 638 patients with RA and 113,998 controls. They found no association between tea intake and the incidence of RA in the overall meta-analysis (relative risk Conclusion (RR) = 0.88, 95% CI = 0.62–1.24, P = 0.463) and subgroup In sum, we did not find any effect of genetically pre - analysis according to study design (case–control design: dicted tea intake on the risk of RA and SLE based on RR = 1.00, 95% CI = 0.58–1.72, P = 1.000; cohort design: summary-level data from large GWAS in individuals of RR = 0.68, 95% CI = 0.27–1.71, P = 0.410). Similarly, the European ancestry. meta-analysis by Asoudeh et  al. [40] using prospective cohort studies (823 RA cases and 191,313 controls) did Supplementary Information not observe significant relation between tea consump - The online version contains supplementary material available at https:// doi. org/ 10. 1186/ s42358‑ 023‑ 00290‑7. tion and the risk of RA (RR = 1.05, 95% CI 0.73–1.53). There were no meta-analyses estimating the relationship Additional file 1: Table S1. Summary results of the SNPs associated with between tea and SLE risk. Our findings were also in line tea intake. with a recently published MR study by Chen and col- Additional file 2: Table S2. Summary results of the SNPs associated with leagues who showed no effect of tea consumption on RA current tobacco smoking. (sample size: 58,284, OR = 1.24, 95% CI 0.81–1.91) [41]. Additional file 3: Table S3. Summary results of the SNPs associated with alcoholic drinks per week. Compared with that MR study, we increased the statisti- cal power by using a much larger study sample (447,485 Additional file 4: Table S4. Summary results of the SNPs associated with coffee intake. participants). Additionally, we controlled for several Additional file 5: Figure S1. Schematic representation of the Mendelian confounding lifestyle factors including smoking, alcohol randomisation design. Our Mendelian randomization study is based on drinking, and coffee consumption using multivariable three key assumptions: (1) instruments are associated with tea intake; (2) MR, which was important but neglected by Chen and instruments have no associations with confounding factors; and (3) instru‑ ments directly affect rheumatoid arthritis and systemic lupus erythemato ‑ coworkers [41]. sus through tea intake. This MR study is subject to some limitations. Firstly, Additional file 6: Table S5. Leave ‑ one‑ out analyses using the IVW we assessed the effects of black tea and green tea method. together but did not separate their effects, due to the Additional file 7: Table S6. The IVW MR estimates on the causal effect questionnaire design in the original GWAS study. Black of genetically predicted current tobacco smoking, alcoholic drinks per tea and green tea differ in terms of chemical composi - week, and coffee intake on rheumatoid arthritis and systemic lupus erythematosus. tion [41]. The major polyphenolic compounds of black tea include thearubigins and theaflavins, whereas green Acknowledgements tea is rich in catechins such as epicatechin-3-gallate Not applicable. and epicatechin and EGCG [42]. Limited evidence showed that consumption of black tea and green tea Author contributions JH, conceptualization; JH, methodology; RBL and JH, resources; RBL and JH, exhibited a different association with SLE risk [15]. We validation and formal analysis; RBL and JH, writing‑ original draft prepara‑ hope that future MR studies can evaluate the causal tion; JH, supervision; JH, writing‑review and editing. Both authors read and effect of black tea and green tea intake on RA and approved the final manuscript. SLE separately when more data becomes available. Funding Secondly, we were not able to assess non-linear asso- There is no funding for this study. ciations since two-sample MR can only assess linear Availability of data and materials associations. Future MR analyses are needed to evalu- This Mendelian randomization study is based on summary‑level data that ate the non-linear relationships of tea intake with RA have been made publicly available in the IEU GWAS database (https:// gwas. and SLE. Thirdly, it is known that one of the shortcom - mrcieu. ac. uk/). ings of MR is limited statistical power. We cannot rule L u and Huang Advances in Rheumatology (2023) 63:10 Page 7 of 8 18. Lamichhane D, Collins C, Constantinescu F, Walitt B, Pettinger M, Parks Declarations C, et al. Coffee and tea consumption in relation to risk of rheumatoid arthritis in the women’s health initiative observational cohort. J Clin Ethics approval and consent to participate Rheumatol. 2019;25(3):127–32. Not applicable. 19. Westerlind H, Palmqvist I, Saevarsdottir S, Alfredsson L, Klareskog L, Di Giuseppe D. Is tea consumption associated with reduction of risk of Consent for publication rheumatoid arthritis? A Swedish case‑ control study. Arthritis Res Ther. Not applicable. 2021;23(1):209. 20. Bae SC, Lee YH. Coffee consumption and the risk of rheumatoid arthritis Competing interests and systemic lupus erythematosus: a Mendelian randomization study. The authors have no competing interests to declare. Clin Rheumatol. 2018;37(10):2875–9. 21. Bae SC, Lee YH. Alcohol intake and risk of rheumatoid arthritis: a Mende‑ lian randomization study. Z Rheumatol. 2019;78(8):791–6. Received: 5 July 2022 Accepted: 24 February 2023 22. Jiang X, Alfredsson L. Modifiable environmental exposure and risk of rheumatoid arthritis‑ current evidence from genetic studies. Arthritis Res Ther. 2020;22(1):154. 23. Jiang X, Zhu Z, Manouchehrinia A, Olsson T, Alfredsson L, Kockum I. Alcohol consumption and risk of common autoimmune inflammatory References diseases‑ evidence from a large‑scale genetic analysis totaling 1 million 1. Cush JJ. Rheumatoid arthritis: early diagnosis and treatment. Rheum Dis individuals. Front Genet. 2021;12:687745. Clin North Am. 2022;48(2):537–47. 24. Wang P, Dan YL, Wu Q, Tao SS, Yang XK, Wang DG, et al. Non‑ causal effects 2. Barber MRW, Drenkard C, Falasinnu T, Hoi A, Mak A, Kow NY, et al. Global of smoking and alcohol use on the risk of systemic lupus erythematosus. epidemiology of systemic lupus erythematosus. Nat Rev Rheumatol. Autoimmun Rev. 2021;20(9):102890. 2021;17(9):515–32. 25. Bycroft C, Freeman C, Petkova D, Band G, Elliott LT, Sharp K, et al. The UK 3. Mathias LM, Stohl W. Systemic lupus erythematosus (SLE): emerging Biobank resource with deep phenotyping and genomic data. Nature. therapeutic targets. Expert Opin Ther Targets. 2020;24(12):1283–302. 2018;562(7726):203–9. 4. Neyestani TR, Gharavi A, Kalayi A. Selective effects of tea extract and its 26. Cole JB, Florez JC, Hirschhorn JN. Comprehensive genomic analysis of phenolic compounds on human peripheral blood mononuclear cell dietary habits in UK Biobank identifies hundreds of genetic associations. cytokine secretions. Int J Food Sci Nutr. 2009;60(Suppl 1):79–88. Nat Commun. 2020;11(1):1467. 5. Peairs A, Dai R, Gan L, Shimp S, Rylander MN, Li L, et al. Epigallocatechin‑ 27. Liu M, Jiang Y, Wedow R, Li Y, Brazel DM, Chen F, et al. Association studies 3‑ gallate (EGCG) attenuates inflammation in MRL/lpr mouse mesangial of up to 1.2 million individuals yield new insights into the genetic etiol‑ cells. Cell Mol Immunol. 2010;7(2):123–32. ogy of tobacco and alcohol use. Nat Genet. 2019;51(2):237–44. 6. Bañuls‑Mirete M, Ogdie A, Guma M. Micronutrients: essential treatment 28. Noyce AJ, Kia DA, Hemani G, Nicolas A, Price TR, De Pablo‑Fernandez for inflammatory arthritis? Curr Rheumatol Rep. 2020;22(12):87. E, et al. Estimating the causal influence of body mass index on risk 7. Song C, Xu S, Chang L, Zhao X, Mei X, Ren X, et al. Preparation of EGCG of Parkinson disease: a Mendelian randomisation study. PLoS Med. decorated, injectable extracellular vesicles for cartilage repair in rat arthri‑ 2017;14(6):e1002314. tis. Regen Biomater. 2021;8(6):67. 29. Burgess S, Thompson SG. CRP CHD genetics collaboration avoiding bias 8. Marotte H, Ruth JH, Campbell PL, Koch AE, Ahmed S. Green tea extract from weak instruments in Mendelian randomization studies. Int J Epide‑ inhibits chemokine production, but up‑regulates chemokine receptor miol. 2011;40(3):755–64. expression, in rheumatoid arthritis synovial fibroblasts and rat adjuvant ‑ 30. Bowden J, Davey Smith G, Haycock PC, Burgess S. Consistent estima‑ induced arthritis. Rheumatology0. 2010;49(3):467–79. tion in mendelian randomization with some invalid instruments using a 9. Liu D, Li P, Song S, Liu Y, Wang Q, Chang Y, et al. Pro‑apoptotic effect of weighted median estimator. Genet Epidemiol. 2016;40(4):304–14. epigallo‑ catechin‑3‑ gallate on B lymphocytes through regulating BAFF/ 31. Burgess S, Bowden J, Fall T, Ingelsson E, Thompson SG. Sensitivity analyses PI3K/Akt/mTOR signaling in rats with collagen‑induced arthritis. Eur J for robust causal inference from mendelian randomization analyses with Pharmacol. 2012;690(1–3):214–25. multiple genetic variants. Epidemiology. 2017;28(1):30–42. 10. Wu D, Wang J, Pae M, Meydani SN. Green tea EGCG, T cells, and T cell‑ 32. Hartwig FP, Davey Smith G, Bowden J. Robust inference in summary data mediated autoimmune diseases. Mol Aspects Med. 2012;33(1):107–18. Mendelian randomization via the zero modal pleiotropy assumption. Int J 11. Riegsecker S, Wiczynski D, Kaplan MJ, Ahmed S. Potential benefits of Epidemiol. 2017;46(6):1985–98. green tea polyphenol EGCG in the prevention and treatment of vascular 33. Verbanck M, Chen CY, Neale B, Do R. Detection of widespread horizontal inflammation in rheumatoid arthritis. Life Sci. 2013;93(8):307–12. pleiotropy in causal relationships inferred from Mendelian randomization 12. Tsai PY, Ka SM, Chang JM, Chen HC, Shui HA, Li CY, et al. Epigallocatechin‑ between complex traits and diseases. Nat Genet. 2018;50(5):693–8. 3‑ gallate prevents lupus nephritis development in mice via enhancing 34. Bowden J, Del Greco MF, Minelli C, Davey Smith G, Sheehan N, Thompson the Nrf2 antioxidant pathway and inhibiting NLRP3 inflammasome J. A framework for the investigation of pleiotropy in two‑sample sum‑ activation. Free Radic Biol Med. 2011;51(3):744–54. mary data Mendelian randomization. Stat Med. 2017;36(11):1783–802. 13. Mikuls TR, Cerhan JR, Criswell LA, Merlino L, Mudano AS, Burma M, 35. Burgess S, Thompson SG. Multivariable Mendelian randomization: the use et al. Coffee, tea, and caffeine consumption and risk of rheumatoid of pleiotropic genetic variants to estimate causal effects. Am J Epidemiol. arthritis: results from the Iowa Women’s Health Study. Arthritis Rheum. 2015;181(4):251–60. 2002;46(1):83–91. 36. Pattison DJ, Symmons DP, Lunt M, Welch A, Luben R, Bingham SA, et al. 14. Karlson EW, Mandl LA, Aweh GN, Grodstein F. Coffee consumption and Dietary risk factors for the development of inflammatory polyarthritis: evi‑ risk of rheumatoid arthritis. Arthritis Rheum. 2003;48(11):3055–60. dence for a role of high level of red meat consumption. Arthritis Rheum. 15. Kiyohara C, Washio M, Horiuchi T, Asami T, Ide S, Atsumi T, et al. Modifying 2004;50(12):3804–12. effect of N‑acetyltransferase 2 genotype on the association between 37. Mazzucca CB, Scotti L, Cappellano G, Barone‑Adesi F, Chiocchetti A. Nutri‑ systemic lupus erythematosus and consumption of alcohol and caffeine ‑ tion and rheumatoid arthritis onset: a prospective analysis using the UK rich beverages. Arthritis Care Res. 2014;66(7):1048–56. biobank. Nutrients. 2022;14(8):1554. 16. Washio M, Fujii T, Kuwana M, Kawaguchi Y, Mimori A, Horiuchi T, et al. 38. Khasawneh LQ, Al‑Mahayri ZN, Ali BR. Mendelian randomization in Lifestyle and other related factors for the development of mixed connec‑ pharmacogenomics: the unforeseen potentials. Biomed Pharmacother. tive tissue disease among Japanese females in comparison with systemic 2022;150:112952. lupus erythematosus. Mod Rheumatol. 2014;24(5):788–92. 39. Lee YH, Bae SC, Song GG. Coffee or tea consumption and the risk of rheu‑ 17. Rambod M, Nazarinia M, Raieskarimian F. The impact of dietary habits matoid arthritis: a meta‑analysis. Clin Rheumatol. 2014;33(11):1575–83. on the pathogenesis of rheumatoid arthritis: a case‑ control study. Clin Rheumatol. 2018;37(10):2643–8. Lu and Huang Advances in Rheumatology (2023) 63:10 Page 8 of 8 40. Asoudeh F, Dashti F, Jayedi A, Hemmati A, Fadel A, Mohammadi H. Caf‑ feine, coffee, tea and risk of rheumatoid arthritis: systematic review and dose‑response meta‑analysis of prospective cohort studies. Front Nutr. 2022;9:822557. 41. Chen S, Chen T, Chen Y, Huang D, Pan Y, Chen S. Causal association between tea consumption and bone health: a mendelian randomization study. Front Nutr. 2022;9:872451. 42. Khan N, Mukhtar H. Tea polyphenols in promotion of human health. Nutrients. 2018;11(1):39. Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in pub‑ lished maps and institutional affiliations. Re Read ady y to to submit y submit your our re researc search h ? Choose BMC and benefit fr ? Choose BMC and benefit from om: : fast, convenient online submission thorough peer review by experienced researchers in your field rapid publication on acceptance support for research data, including large and complex data types • gold Open Access which fosters wider collaboration and increased citations maximum visibility for your research: over 100M website views per year At BMC, research is always in progress. Learn more biomedcentral.com/submissions

Journal

Advances in RheumatologySpringer Journals

Published: Mar 10, 2023

Keywords: Mendelian randomization; Rheumatoid arthritis; Systemic lupus erythematosus; Tea intake

References