Characterizing metabolomic signatures related to coffee and tea consumption and their association with incidence and dynamic progression of type 2 diabetes: A multi-state analysis

1/33 Title: Characterizing metabolomic signatures related to coffee and tea consumption and their association with incidence and dynamic progression of type 2 diabetes: A multi-state analysis Authors: Guzhengyue Zheng, Shanshan Ran, Jingyi Zhang, Zhengmin (Min) Qian, Fei Tian, Hui Shi, Michael Elliott, Maya Tabet, Yin Yang, Hualiang Lin ORCiD IDs: https://orcid.org/0000-0002-3643-9408 (Hualiang Lin) Correspondence Address: No. 74, 2nd Yat-sen Road, Yuexiu District, Guangzhou, Guangdong, China ([email protected]) Joint Authorship: N/A Affiliations: Department of Epidemiology, School of Public Health, Sun Yat-sen University, Guangzhou, China (Guzhengyue Zheng, Shanshan Ran, Jingyi Zhang, Fei Tian, Hui Shi, Yin Yang, Hualiang Lin); School of Medicine, Xizang Minzu University, Xianyang, China (Guzhengyue Zheng); Department of Epidemiology and Biostatistics, College for Public Health & Social Justice, Saint Louis University, Saint Louis, United States of America [Zhengmin (Min) Qian & Michael Elliott]; College of Global Population Health, University of Health Sciences and Pharmacy in Saint Louis, Saint Louis, United States of America (Maya Tabet). Key words: tea, coffee, metabolomic signatures, type 2 diabetes progression, multi-state regression 2/33 Acknowledgments1:This research using the UK Biobank Resource was approved under Application Number 69550. The authors appreciate all UK Biobank participants and all staff for their contribution to these studies. Funding: This work was supported, in whole or in part, by the Bill & Melinda Gates Foundation [Grant No.: INV-016826]. Under the grant conditions of the Foundation, a Creative Commons Attribution 4.0 Generic License has already been assigned to the Author Accepted Manuscript version that might arise from this submission. Conflict of Interest: The authors declare no conflicts or competing interests that could have appeared to influence the work reported in this paper. Disclaimer: N/A Data Availability Statement: The datasets generated and analyzed during the current study are available upon reasonable request to the Access Management System (AMS) through the UK Biobank website (https://www.ukbiobank.ac.uk/enable-your-research/apply-for-access). 1Study investigators, conference presentations, preprint publication information, thanks. 3/33 Abstract Our study aimed to investigate the impact of tea and coffee consumption and related metabolomic signatures on dynamic transitions from diabetes-free status to incident type 2 diabetes (T2D), and subsequently to T2D-related complications and death. We included 438,970 participants in the UK Biobank who were free of diabetes and diabetes complications at baseline. Of these, 212,146 individuals had information on all metabolic biomarkers. We identified teaand coffee-related metabolomic signatures using elastic net regression models. We examined associations of tea and coffee intake and related metabolomic signatures with the onset and progression of T2D using multi-state regression models. We observed that tea and coffee consumption and related metabolomic signatures were inversely associated with the risk of five T2D transitions. For example, HRs (95% CIs) per SD increase of the tea-related metabolomic signature were 0.87 (0.85, 0.89), 0.97 (0.95, 0.99), 0.91 (0.90, 0.92), 0.92 (0.91, 0.94), and 0.91 (0.90, 0.92) for transitions from diabetes-free state to incident T2D, from diabetes-free state to total death, from incident T2D to T2D complications, from incident T2D to death, and from T2D complications to death. These findings highlight the benefit of tea and coffee intake in reducing the risk of occurrence and progression of T2D. Keywords: tea, coffee, metabolomic signatures, type 2 diabetes progression, multi-state regression 4/33 Introduction In 2021, 537 million individuals worldwide lived with type 2 diabetes (T2D), and T2D cases are estimated to reach 783 million by 2045 (1). People with T2D face severe complications, including macrovascular disease, microvascular complications, and premature mortality (2). Thus, there is an urgent need to prevent the occurrence of T2D and delay the development of complications. A healthy diet has played a fundamental role in T2D prevention and complication control (3). Specific dietary components, such as tea and coffee, have been linked to T2D (4, 5), with mixed findings (6, 7). Some studies have demonstrated that tea intake was negatively linked to the risk of T2D (8, 9), while others have not (10). Several epidemiological studies have reported the preventive effects of coffee consumption on the onset and progression of T2D (11-13). Tea is rich in tea polyphenols, and coffee is high in bioactive compounds such as caffeine, phenolics, diterpenes, and chlorogenic acids, which may prevent the onset and progression of T2D through antioxidative, anti-inflammatory, antibacterial, antiviral, antimutagenic, and thermogenic properties and by promoting metabolism(5, 14). However, previous studies analysing the associations of tea and coffee intake with the onset and progression of T2D have typically only examined disease status (15, 16), and have not examined the progression trajectories of T2D. 5/33 In recent years, studies have reported that high-throughput metabolomic techniques can objectively measure the host’s metabolic response to dietary intake (17-19). Several epidemiological studies and randomized controlled trials have identified metabolomic signatures related to tea and coffee (20, 21). However, few studies have investigated the associations of metabolomic signatures reflecting tea and coffee intake with T2D-related outcomes (21). In addition, no studies have investigated the associations of metabolomic signatures reflecting tea and coffee intake with the progression trajectories of T2D. Therefore, we conducted multi-state regression models to examine the effects of tea and coffee consumption, as well as their corresponding metabolomic signatures, on the transitions from a diabetes-free state to incident T2D, subsequently to T2D complications, and finally to death among adults from the UK Biobank. Methods Study population We used data from the UK Biobank, an ongoing prospective cohort study in the United Kingdom (22). The UK Biobank was established during 2006-2010 with over 0.5 million participants aged 37-73 years. Participants in the UK Biobank cohort provided information on socioeconomic characteristics, lifestyle exposures, family history of diseases, dietary intake, plasma sample, and history of medication and disease via physical and biological measurements, verbal interviews, and 6/33 questionnaires at baseline (2006-2010) (23). Among 502,461 participants, we excluded participants with no follow-up (1298 participants), prevalent cases of any types of diabetes or diabetes complications at baseline (59,025 participants), and those with incomplete data on tea and coffee consumption (3168 participants) (Figure S1). Notably, at baseline, prevalent cases of diabetes and diabetes complications were identified in three ways: hospital inpatient and primary care records through the International Classification of Disease, Tenth version (ICD-10) codes, self-reported data, and diabetes-related medication status (24). The final analytic cohort of 438,970 individuals (the primary sample) had comparable baseline characteristics to the 502,461 participants initially recruited to the cohort (Table S1). To identify metabolomic signatures reflecting tea and coffee intake and investigate the associations of metabolomic signatures with the dynamic progression of T2D, we used a subsample of 212,146 participants who had data for all metabolic biomarkers (Figure S1). Assessment of tea and coffee consumption Dietary data, including tea and coffee intake, were collected using a food frequency questionnaire (FFQ). Four rounds of FFQ dietary surveys on coffee and tea intake were conducted from 2006 to 2022 (Appendix S1and Figure S2). Most participants only completed the FFQ survey at baseline. Nevertheless, comparing tea and coffee intake at the second, third, and fourth dietary assessments versus at baseline revealed that most participants who developed T2D during follow-up did not alter their tea and 7/33 coffee consumption (Table S2). Therefore, we only used tea and coffee consumption at baseline as dietary exposure. Tea and coffee consumption was considered a continuous variable (cups/day) when identifying their corresponding metabolomic signatures. To determine whether tea and coffee consumption was associated with T2D progression, we classified participants into the following three categories, based on the distribution of tea and coffee intake: tea (<1 cup/day, 1-2 cups/day, and ≥3 cups/day) and coffee (<1 cup/day, 1-3 cups/day, and ≥4 cups/day). Metabolomics profiling Details regarding metabolic biomarker profiling, including plasma sample collection, metabolomics analysis, the multistep processing procedure, and quality control, can be obtained from UK Biobank online resources (25) and have been described previously (26, 27). In short, biomarker data from approximately 275,000 participants at baseline recruitment (2006-2010) and 17,000 participants at repeated assessment (2012-2013) were collected using high-throughput nucleic magnetic resonance (NMR) spectroscopy on Nightingale Health’s metabolic biomarker platform in Finland. Of note, approximately 15,500 individuals participated in both a baseline and repeated assessment. A total of 251 metabolic measures, including 168 absolute levels (mmol/L), 81 ratio measures, and two newly derived variables, were quantified per ethylenediaminetetraacetic acid (EDTA) plasma sample (Table S3). These metabolic measures spanned glycolysis related metabolites, amino acids, lipids, lipoproteins (14 subclasses), ketone bodies, fatty acids, fluid balance, and inflammation. Based on 8/33 previous studies (21, 28, 29), we included all 251 metabolites in the subsequent elastic net regularized logistic regression models. Follow-up for outcomes Follow-up time was counted from the date when participants consented to join the UK Biobank until the date of loss to follow-up, death, or the last follow-up (31 December 2022), whichever came first. In the current study, the outcomes of interest were incident T2D, T2D complications, T2D-specific death, and total death. Outcomes that occurred during follow-up were identified through primary care, hospital admissions, and death registries linked to the UK Biobank (22). We defined the outcomes using the ICD-10 codes (Appendix S2). Covariates Data on potential confounders were collected from baseline questionnaires, including age (years, continuous), sex (male/female), ethnicity (White/non-White), residence (urban/rural), income-to-poverty ratio (<1.0/1.0-2.5/2.5-3.9/≥4.0), education attainment (college degree or higher/any school degree/vocational qualifications and other), family history of diabetes (yes/no), alcohol intake (g/week, continuous), smoking status (current/previous/never), physical activity [metabolic equivalent of task (MET) in weekly hours, continuous], dietary supplement (yes/no), body mass index (BMI) (≥30 kg/m2/25-29.9 kg/m2/18.5-24.9 kg/m2/<18.5 kg/m2), hyperlipidemia (yes/no), and hypertension (yes/no) (Appendix S3). 9/33 Statistical analysis Continuous variables are described as the mean (standard deviation, SD), and categorical variables are presented as the frequency (proportion). Multiple imputation techniques were applied to fill in the missing values for all covariates (30, 31) (Appendix S4). Baseline characteristics are shown by five disease transitions. Spearman’s correlation coefficients were used to assess the correlation between tea and coffee consumption and related metabolomic signatures. The present study included a two-phase analysis: we assessed the associations of tea and coffee consumption with T2D progression in phase Ⅰ analysis, and we identified metabolomic signatures related to tea and coffee intake and explored their associations with T2D progression in phase Ⅱ analysis. The identification of metabolomic signatures reflecting tea and coffee intake was conducted through several steps: (1) the biomarker values were Zscore standardized; (2) metabolites reflecting tea and coffee intake were selected from 251 metabolites, and their corresponding regression coefficients (weights) were obtained using elastic net regularized logistic regression models (Tables S3-S5); and (3) teaand coffee-related metabolomic signatures were obtained by multiplying the selected metabolite standardized values with the corresponding weights and then summing them together. We applied multi-state regression models to estimate hazard ratios (HRs) and 95% 16 /33 (Tables S7 –S20). Stratified analyses by sex and age showed most observed interactions lacked major implications (Tables S21–S24). Discussion We identified 32 and 38 metabolites related to the consumption of tea and coffee, respectively. Both teaand coffee-related metabolites were primarily lipids and lipoproteins, amino acids, and glycolysis related metabolites. We observed that higher consumption of tea and coffee and the related metabolomic signatures were associated with a lower risk of all five disease transitions, namely from a diabetes-free state to incident T2D, from a diabetes-free state to T2D complications, from incident T2D to T2D complications, from incident T2D to death, and from T2D complications to death. Comparison with other studies and explanations We observed that tea consumption was associated with a reduced risk of five stages of T2D progression. Previous studies examining the association of tea intake with the onset and progression of T2D, have typically only examined disease status and have yielded inconsistent results. The EPIC-InterAct case-cohort study, the Health Professionals Follow-Up Study, and the Nurses' Health Study found that regular tea intake has a protective effect on the onset of T2D (36, 37), which was consistent with our results. Ma et al. reported that the consumption of tea had an inverse association with cardiovascular diseases (CVD) outcomes and all-cause mortality among adults 17 /33 with T2D (38), which supports our findings. However, a double-blind, randomized study among 49 patients with T2D found no significant association between the extract of black and green tea and changes in glycosylated hemoglobin (10). Three possible reasons for the nonsignificant association include the fact that the intervention factor studied was the extract of tea, the limited study time frame of 3 months, and the blinded, placebo-controlled, and randomized study design (10).Nevertheless, despite mixed findings, most studies have shown that tea intake could reduce the risk of the onset and progression of T2D. We also found that the consumption of coffee was associated with a reduced risk of the five stages of T2D progression. Prior research in this area is in line with our findings. The Nurses’ Health Study and the Health Professionals Follow-Up study found that consumption of coffee is inversely associated with incident CVD [HR for the highest versus the lowest intake = 0.82 (95% CI: 0.69, 0.98)] and total death [HR for the highest versus the lowest intake = 0.74 (95% CI: 0.63, 0.86)] among patients with T2D (38). The Korean National Health and Nutrition study showed that coffee intake is inversely associated with the risk of diabetic retinopathy in adults with diabetes (13). Hossein et al. reported that drinking coffee is associated with a reduced risk of mortality in patients with T2D (39). We found that most of the teaand coffee-related metabolites were lipoproteins and lipids, amino acids, and glycolysis related metabolites, which was also supported by 18 /33 previous evidence that tea and coffee can have an effect on lipids and lipoproteins, amino acids, and glycolysis related metabolites (40-42). In addition, we observed that teaand coffee-related metabolomic signatures were associated with the incidence and progression of T2D. Consistent with our results, a nested case‒control study documented that 34 coffee-related plasma metabolites were predominantly lipids and were associated with the risk of T2D (21); another nested case‒control study reported an inverse association between the metabolite panel of filtered coffee and the risk of T2D (43). While these studies were limited to examining T2D status, our study focused on five dynamic stages of T2D progression. Further, no study has investigated the associations of tea-related metabolomic signatures with T2D outcomes. Given that 24 of the metabolites associated with coffee intake were also associated with tea intake in the present study, it is not surprising that the corresponding metabolomic signature of tea was associated with T2D outcomes. Future studies are needed to further assess the associations of tea-related metabolomic signatures and T2D risk. Although associations between tea and coffee intake and T2D-related outcomes have been assessed extensively, associations with T2D dynamic transitions remain unknown. Previous studies analysing the associations of tea and coffee intake with the onset and progression of T2D have typically only examined a single disease outcome. Our study adds to the available evidence by evaluating the effects of tea and coffee intake on the dynamic progression of T2D. Notably, we found that a significantly reduced risk of developing complications from incident T2D and mortality from T2D 19 /33 or T2D complications among regular tea drinkers of ≥4 cups per day and coffee consumers of ≥3 cups per day. When limiting the sample to individuals with T2D at baseline, we observed significant inverse associations between the consumption of tea (or coffee) and the risk of disease progression from T2D to its complications, from T2D to mortality, and from T2D complications to mortality. These findings provide further supporting evidence that drinking tea or coffee might decrease the risk of complications occurring among individuals with T2D and fatal events among patients with T2D or T2D complications. The potential mechanism for the preventive effect of tea consumption on T2D risk is mainly attributed to tea polyphenols, such as catechins, anthocyanins, flavones, and phenolic acids (5). Tea polyphenols may prevent the onset and progression of T2D in three ways: first, tea polyphenol intake can inhibit gluconeogenesis and prevent glucose absorption (44); second, tea polyphenol intake may directly enhance glucose metabolism by improving insulin sensitivity and increasing insulin secretion (45); third, tea polyphenols have lipolytic, antioxidative, anti-inflammatory, and anti-obesity effects, which are important in the aetiology of T2D (14). The beneficial effects of coffee on T2D risk are possibly due to the presence of caffeine, phenolics, chlorogenic acids, trigonelline, diterpenes, melanoidins, and polyphenols. These beneficial bioactive compounds have been suggested to improve lipolysis (46), inhibit gluconeogenesis, delay glucose uptake (47), and reduce systematic inflammation (12) and oxidative stress (48). The abovementioned mechanisms relating to glucose, lipid, 20 /33 and amino acid metabolism are consistent with our findings that the metabolites associated with tea and coffee are mainly glycolysis-related metabolites, lipoproteins and lipids, and amino acids. Strengths and limitations Strengths of this study include the use of the multi-state model, the large population-based cohort with long-term follow-up, detailed information on diet and confounding factors, and on the use of a series of sensitivity analyses to confirm the robustness of our findings. Nevertheless, our study has several limitations. First, we only used consumption of tea and coffee at baseline as exposure because most participants only completed the FFQ survey at baseline. However, we compared tea and coffee intake at the second, third, and fourth dietary assessments versus at baseline; most participants who developed T2D during follow-up did not alter their tea and coffee consumption. Second, we cannot exclude the possibility of residual confounding given the observational nature of this study. Third, the UK Biobank study participants were mainly White and healthy volunteers. This may limit the generalizability of our results. Fourth, we could not adjust for the total energy consumption in the primary analyses since this information was unavailable from the UK Biobank’s FFQ surveys. However, we conducted a sensitivity analysis where we further controlled for total energy intake and consumption of sugary drinks in a sample with 24-hour dietary recall surveys; the 21 /33 results were consistent with those from the primary analysis. Finally, only approximately 45% UK Biobank participants had general practitioner data, which may lead to a less precise estimate of the risk factor-outcome associations for incident T2D and T2D complications. Conclusions In conclusion, our study showed that regular consumption of tea and coffee was associated with a reduced risk of T2D incidence and progression. Metabolomic signatures reflecting tea and coffee intake were associated with T2D progression. These results not only suggest that increasing tea and coffee intake could be beneficial for primary, secondary, and tertiary prevention of T2D but also suggest that measuring metabolites representative of tea and coffee intake might enhance precision in predicting T2D progression. Autor contributions GZ was responsible for conceptualization, methodology, formal analysis, software, and writing the original draft; ZQ, ME, and MT were responsible for language modification and polishing; FT and HS contributed to methodology and software; SR, JZ, and YY were responsible for data curation; HL worked on conceptualization, data curation, project administration, resources, supervision, and re-drafting. All authors have read and approved the final manuscript. 22 /33 Acknowledgments: This research using the UK Biobank Resource was approved under Application Number 69550. The authors appreciate all UK Biobank participants and all staff for their contribution to these studies. Funding: This work was supported, in whole or in part, by the Bill & Melinda Gates Foundation [Grant No.: INV-016826]. Under the grant conditions of the Foundation, a Creative Commons Attribution 4.0 Generic License has already been assigned to the Author Accepted Manuscript version that might arise from this submission. Conflicts of Interest: The authors declare no conflicts or competing interests that could have appeared to influence the work reported in this paper. Ethics approval and consent to participate: The project has approval from the North West Multi-Centre Research Ethics Committee (MREC) (REC reference: 21/NW/0157). Each participant in the UK Biobank provided written informed consent. Data Availability Statement: The datasets generated and analyzed during the current study are available upon reasonable request to the Access Management System (AMS) through the UK Biobank website (https://www.ukbiobank.ac.uk/enable-your-research/apply-for-access). References 1. Ogurtsova K, Guariguata L, Barengo NC, et al. IDF diabetes Atlas: Global estimates of undiagnosed diabetes in adults for 2021. Diabetes Res Clin Pract 2022;183:109118. 2. Addendum. 11. Microvascular Complications and Foot Care: Standards of Medical Care in Diabetes-2021: Diabetes Care 2021;44(Suppl. 1):S151-S167. Diabetes Care 2021;44(9):2186-7. 3. Merino J, Guasch-Ferre M, Li J, et al. Polygenic scores, diet quality, and type 2 diabetes risk: 23 /33 An observational study among 35,759 adults from 3 US cohorts. PLoS Med 2022;19(4):e1003972. 4. Alperet DJ, Rebello SA, Khoo EY, et al. The effect of coffee consumption on insulin sensitivity and other biological risk factors for type 2 diabetes: a randomized placebo-controlled trial. Am J Clin Nutr 2020;111(2):448-58. 5. Wang P, Ma XM, Geng K, et al. Effects of Camellia tea and herbal tea on cardiometabolic risk in patients with type 2 diabetes mellitus: A systematic review and meta-analysis of randomized controlled trials. Phytother Res 2022;36(11):4051-62. 6. Mahmoud F, Haines D, Al-Ozairi E, et al. Effect of Black Tea Consumption on Intracellular Cytokines, Regulatory T Cells and Metabolic Biomarkers in Type 2 Diabetes Patients. Phytother Res 2016;30(3):454-62. 7. Mousavi A, Vafa M, Neyestani T, et al. The effects of green tea consumption on metabolic and anthropometric indices in patients with Type 2 diabetes. J Res Med Sci 2013;18(12):1080-6. 8. Dinh TC, Thi Phuong TN, Minh LB, et al. The effects of green tea on lipid metabolism and its potential applications for obesity and related metabolic disorders - An existing update. Diabetes Metab Syndr 2019;13(2):1667-73. 9. Kondo Y, Goto A, Noma H, et al. Effects of Coffee and Tea Consumption on Glucose Metabolism: A Systematic Review and Network Meta-Analysis. Nutrients 2018;11(1). 10. Mackenzie T, Leary L, Brooks WB. The effect of an extract of green and black tea on glucose control in adults with type 2 diabetes mellitus: double-blind randomized study. Metabolism 2007;56(10):1340-4. 11. Yang J, Tobias DK, Li S, et al. Habitual coffee consumption and subsequent risk of type 2 diabetes in individuals with a history of gestational diabetes - a prospective study. Am J Clin Nutr 2022;116(6):1693-703. 12. Ochoa-Rosales C, van der Schaft N, Braun KVE, et al. C-reactive protein partially mediates the inverse association between coffee consumption and risk of type 2 diabetes: The UK Biobank and the Rotterdam study cohorts. Clin Nutr 2023;42(5):661-9. 13. Lee HJ, Park JI, Kwon SO, et al. Coffee consumption and diabetic retinopathy in adults with diabetes mellitus. Sci Rep 2022;12(1):3547. 14. Suzuki T, Pervin M, Goto S, et al. Beneficial Effects of Tea and the Green Tea Catechin Epigallocatechin-3-gallate on Obesity. Molecules 2016;21(10). 15. Hemmingsen B, Gimenez-Perez G, Mauricio D, et al. Diet, physical activity or both for prevention or delay of type 2 diabetes mellitus and its associated complications in people at increased risk of developing type 2 diabetes mellitus. Cochrane Database Syst Rev 2017;12:CD003054. 16. Satija A, Bhupathiraju SN, Rimm EB, et al. Plant-Based Dietary Patterns and Incidence of Type 2 Diabetes in US Men and Women: Results from Three Prospective Cohort Studies. PLoS Med 2016;13(6):e1002039. 17. Guasch-Ferre M, Bhupathiraju SN, Hu FB. Use of Metabolomics in Improving Assessment of Dietary Intake. Clin Chem 2018;64(1):82-98. 18. Li J, Guasch-Ferre M, Chung W, et al. The Mediterranean diet, plasma metabolome, and cardiovascular disease risk. Eur Heart J 2020;41(28):2645-56. 19. Shah RV, Steffen LM, Nayor M, et al. Dietary metabolic signatures and cardiometabolic risk. Eur Heart J 2023;44(7):557-69. 24 /33 20. Madrid-Gambin F, Garcia-Aloy M, Vazquez-Fresno R, et al. Metabolic Signature of a Functional High-Catechin Tea after Acute and Sustained Consumption in Healthy Volunteers through (1)H NMR Based Metabolomics Analysis of Urine. J Agric Food Chem 2019;67(11):3118-24. 21. Hang D, Zeleznik OA, He X, et al. Metabolomic Signatures of Long-term Coffee Consumption and Risk of Type 2 Diabetes in Women. Diabetes Care 2020;43(10):2588-96. 22. Sudlow C, Gallacher J, Allen N, et al. UK biobank: an open access resource for identifying the causes of a wide range of complex diseases of middle and old age. PLoS Med 2015;12(3):e1001779. 23. Palmer LJ. UK Biobank: bank on it. Lancet 2007;369(9578):1980-2. 24. Han H, Cao Y, Feng C, et al. Association of a Healthy Lifestyle With All-Cause and Cause-Specific Mortality Among Individuals With Type 2 Diabetes: A Prospective Study in UK Biobank. Diabetes Care 2022;45(2):319-29. 25. Nightingale Health Metabolic Biomarkers (https://biobank.ndph.ox.ac.uk/showcase/ukb/docs/NMR_companion_phase2.pdf). (Accessed 11 November 2023). 26. Bell JA, Richardson TG, Wang Q, et al. Effects of general and central adiposity on circulating lipoprotein, lipid, and metabolite levels in UK Biobank: A multivariable Mendelian randomization study. Lancet Reg Health Eur 2022;21:100457. 27. Joshi AD, McCormick N, Yokose C, et al. Prediagnostic Glycoprotein Acetyl Levels and Incident and Recurrent Flare Risk Accounting for Serum Urate Levels: A Population-Based, Prospective Study and Mendelian Randomization Analysis. Arthritis Rheumatol 2023;75(9):1648-57. 28. Williamson G. Protection against developing type 2 diabetes by coffee consumption: assessment of the role of chlorogenic acid and metabolites on glycaemic responses. Food Funct 2020;11(6):4826-33. 29. Seow WJ, Low DY, Pan WC, et al. Coffee, Black Tea, and Green Tea Consumption in Relation to Plasma Metabolites in an Asian Population. Mol Nutr Food Res 2020:e2000527. 30. Blazek K, van Zwieten A, Saglimbene V, et al. A practical guide to multiple imputation of missing data in nephrology. Kidney Int 2021;99(1):68-74. 31. Gray L, McCartney G, White IR, et al. Use of record-linkage to handle non-response and improve alcohol consumption estimates in health survey data: a study protocol. BMJ Open 2013;3(3). 32. Putter H. Tutorial in biostatistics: Competing risks and multi-state models Analyses using the mstate package. 2020. 33. Putter H, Fiocco M, Geskus RB. Tutorial in biostatistics: competing risks and multi-state models. Stat Med 2007;26(11):2389-430. 34. Cai M, Zhang S, Lin X, et al. Association of Ambient Particulate Matter Pollution of Different Sizes With In-Hospital Case Fatality Among Stroke Patients in China. Neurology 2022. 35. Wang X, Guo Y, Cai M, et al. Constituents of fine particulate matter and asthma in 6 lowand middle-income countries. J Allergy Clin Immunol 2022;150(1):214-22 e5. 36. InterAct C, van Woudenbergh GJ, Kuijsten A, et al. Tea consumption and incidence of type 2 diabetes in Europe: the EPIC-InterAct case-cohort study. PLoS One 2012;7(5):e36910. 37. Bhupathiraju SN, Pan A, Malik VS, et al. Caffeinated and caffeine-free beverages and risk of type 2 diabetes. Am J Clin Nutr 2013;97(1):155-66. 25 /33 38. Ma L, Hu Y, Alperet DJ, et al. Beverage consumption and mortality among adults with type 2 diabetes: prospective cohort study. BMJ 2023;381:e073406. 39. Shahinfar H, Jayedi A, Khan TA, et al. Coffee consumption and cardiovascular diseases and mortality in patients with type 2 diabetes: A systematic review and dose-response meta-analysis of cohort studies. Nutr Metab Cardiovasc Dis 2021;31(9):2526-38. 40. Law WS, Huang PY, Ong ES, et al. Metabonomics investigation of human urine after ingestion of green tea with gas chromatography/mass spectrometry, liquid chromatography/mass spectrometry and (1)H NMR spectroscopy. Rapid Commun Mass Spectrom 2008;22(16):2436-46. 41. Scalbert A, Brennan L, Manach C, et al. The food metabolome: a window over dietary exposure. Am J Clin Nutr 2014;99(6):1286-308. 42. Jacobs S, Kroger J, Floegel A, et al. Evaluation of various biomarkers as potential mediators of the association between coffee consumption and incident type 2 diabetes in the EPIC-Potsdam Study. Am J Clin Nutr 2014;100(3):891-900. 43. Shi L, Brunius C, Johansson I, et al. Plasma metabolite biomarkers of boiled and filtered coffee intake and their association with type 2 diabetes risk. J Intern Med 2019;287(4):405-21. 44. Yasui K, Tanabe H, Miyoshi N ST, et al. Effects of (-)-epigallocatechin-3-O-gallate on expression of gluconeogenesisrelated genes in the mouse duodenum. Biomed Res 2011;32: 313-20. 45. Chen YK, Cheung C, Reuhl KR, et al. Effects of green tea polyphenol (-)-epigallocatechin-3-gallate on newly developed high-fat/Western-style diet-induced obesity and metabolic syndrome in mice. J Agric Food Chem 2011;59(21):11862-71. 46. Blaak EE. Fatty acid metabolism in obesity and type 2 diabetes mellitus. P Nutr Soc 2007;62(3):753-60. 47. Kajikawa M, Maruhashi T, Hidaka T, et al. Coffee with a high content of chlorogenic acids and low content of hydroxyhydroquinone improves postprandial endothelial dysfunction in patients with borderline and stage 1 hypertension. Eur J Nutr 2019;58(3):989-96. 48. Hang D, Kvaerner AS, Ma W, et al. Coffee consumption and plasma biomarkers of metabolic and inflammatory pathways in US health professionals. Am J Clin Nutr 2019;109(3):635-47. 32 /33 Table 3. HRs (95% CIs) of tea-related and coffee-related metabolic signature with risk of five progressions of T2D using the multi-state model (n= 212,146) a. Tea-related metabolic signature (per SD increment) Coffee-related metabolic signature (per SD increment) Transitions Model 1 b Model 2 c Model 1 b Model 2 c Baseline →Incident T2D 0.888 (0.872, 0.904) 0.872 (0.851, 0.893) 0.875 (0.860, 0.891) 0.886 (0.865, 0.906) Baseline → Death 0.991 (0.978, 1.003) 0.969 (0.954, 0.985) 0.972 (0.962, 0.982) 0.940 (0.926, 0.955) Incident T2D → T2DCs 0.913 (0.908, 0.917) 0.910 (0.904, 0.916) 0.910 (0.905, 0.914) 0.908 (0.902, 0.914) Incident T2D → Death 0.923 (0.910, 0.936) 0.921 (0.905, 0.938) 0.919 (0.908, 0.929) 0.923 (0.907, 0.940) T2DCs → Death 0.910 (0.903, 0.918) 0.906 (0.896, 0.915) 0.905 (0.898, 0.912) 0.902 (0.893, 0.911) aHRs and 95% CIs of T2D progression risk per standard deviation increment in the metabolic signature. bModel 1 was adjusted for age, sex, ethnicity, income-to-poverty ratio, education attainment, and residence. cModel 2 was based on model 1 and additionally adjusted for smoking status, physical activity, alcohol consumption, family history of diabetes, and dietary supplement, hyperlipidemia, hypertension, BMI, and mutually adjusted for tea-related metabolic signature and coffee-related metabolic signature. Abbreviations: CIs, confidence intervals; HRs, hazard ratios; SD, standard deviation; T2D, type 2 diabetes; T2DCs, type 2 diabetes complications. 33 /33 Figure titles Figure 1 Transitions from diabetes-free status to incident T2D, T2D complications, and all-cause death. Diabetes complications included diabetic eye diseases, diabetic kidney diseases, diabetic neuropathy diseases, cardiovascular diseases, peripheral vascular diseases, and metabolic events. State-specific number of events was reported in boxes, and the transition-specific number of events and percentages (within brackets) were reported on arrows. Abbreviations: T2D, type 2 diabetes.