Evaluation of various biomarkers as potential mediators of the association between coffee consumption and incident type 2 diabetes in the EPIC-Potsdam Study

Evaluation of various biomarkers as potential mediators of the association between coffee consumption and incident type 2 diabetes in the EPIC-Potsdam Study1–3 Simone Jacobs, Janine Kr€ oger, Anna Floegel, Heiner Boeing, Dagmar Drogan, Tobias Pischon, Andreas Fritsche, Cornelia Prehn, Jerzy Adamski, Berend Isermann, Cornelia Weikert, and Matthias B Schulze ABSTRACT Background: The inverse association between coffee consumption and the risk of type 2 diabetes (T2D) is well established; however, little is known about potential mediators of this association. Objective: We aimed to investigate the association between coffee consumption and diabetes-related biomarkers and their potential role as mediators of the association between coffee consumption and T2D. Design: We analyzed a case-cohort study (subcohort: n = 1610; verified incident T2D cases: n = 417) nested within the European Prospective Investigation into Cancer and Nutrition–Potsdam study involving 27,548 middle-aged participants. Habitual coffee consumption was assessed with a validated, semiquantitative food-frequency questionnaire. We evaluated the association between coffee consumption and several T2D-related biomarkers, such as liver markers (reflected by g-glutamyltransferase, fetuin-A, and sex hormone–binding globulin), markers of dyslipidemia (high-density lipoprotein cholesterol and triglycerides), inflammation [C-reactive protein (CRP)], an adipokine (adiponectin), and metabolites, stratified by sex. Results: Coffee consumption was inversely associated with diacylphosphatidylcholine C32:1 in both sexes and with phenylalanine in men, as well as positively associated with acyl-alkyl-phosphatidylcholines C34:3, C40:6, and C42:5 in women. Furthermore, coffee consumption was inversely associated with fetuin-A (P-trend = 0.06) and CRP in women and g-glutamyltransferase and triglycerides in men. Coffee consumption tended to be inversely associated with T2D risk in both sexes, reaching significance only in men [HR (95% CI): women: $4 compared with .0 to ,2 cups coffee/d: 0.78 (0.46, 1.33); men: $5 compared with .0 to ,2 cups coffee/d: 0.40 (0.19, 0.81)]. The association between coffee consumption and T2D risk in men was slightly reduced after adjustment for phenylalanine or lipid markers. Conclusions: Coffee consumption was inversely associated with a diacyl-phosphatidylcholine and liver markers in both sexes and positively associated with certain acyl-alkyl-phosphatidylcholines in women. Furthermore, coffee consumption showed an inverse trend with CRP in women and with triglycerides and phenylalanine in men. However, these markers explained only to a small extent the inverse association between long-term coffee consumption and T2D risk. Am J Clin Nutr 2014;100:891–900. INTRODUCTION Brewed coffee is among the most widely consumed beverages in the world (1). During recent years, a number of epidemiologic studies have found an inverse relationship between consumption of coffee and the risk of developing type 2 diabetes (T2D)4 (2). The association does not depend on race, sex, obesity, or region of the study population (2). Explanations for the protective effect of coffee consumption on the risk of T2D have been related to varying coffee components such as magnesium, lignans, and chlorogenic acid (2). However, little is known about the underlying biologic mechanisms of this association. A meta-analysis of intervention studies suggests that consumption of unfiltered, but not filtered, coffee increases serum concentrations of triglycerides (3). Evidence from clinical trials suggests furthermore that coffee intake increases adiponectin (4–6), decreases fetuin-A concentrations (5), and has no effect on C-reactive protein (CRP) (5–7), the sex hormone–binding globulin (SHBG) 1 From the Departments of Molecular Epidemiology (SJ, JK, and MBS) and Epidemiology (A Floegel, HB, CW, and DD), German Institute of Human Nutrition Potsdam-Rehbruecke, Nuthetal, Germany; Institute of Experimental Genetics, Genome Analysis Center, Helmholtz Zentrum München, German Research Center for Environmental Health, Neuherberg, Germany (CP and JA); Molecular Epidemiology Group, Max Delbrueck Center for Molecular Medicine Berlin-Buch, Berlin, Germany (TP); the Department of Internal Medicine, Division of Endocrinology, Diabetology, Nephrology, Vascular Disease, and Clinical Chemistry, University Hospital of the Eberhard Karls University, Tübingen, Germany (A Fritsche); Institute for Diabetes Research and Metabolic Diseases of the Helmholz Centre Munich at the University of Tübingen (IDM), Tübingen, Germany (A Fritsche); the Department for Clinical Chemistry and Pathobiochemistry, Otto-von-Guericke-University Magdeburg, Magdeburg, Germany (BI); and the German Center for Diabetes Research (DZD), Neuherberg, Germany (SJ, A Fritsche, JK, and MBS). 2 Supported by the Federal Ministry of Science, Germany (01 EA 9401) and the European Union (SOC 95201408 05F02), the German Cancer Aid (grant 70-2488-Ha I) and the European Community (grant SOC 98200769 05 F02), and in part by a grant from the German Federal Ministry of Education and Research (BMBF) to the German Centre for Diabetes Research (DZD). 3 Address correspondence to MB Schulze, German Institute of Human Nutrition Potsdam-Rehbruecke, Department of Molecular Epidemiology, ArthurScheunert-Allee 114-116, 14558 Nuthetal, Germany. E-mail: . 4 Abbreviations used: CRP, C-reactive protein; EPIC, European Prospective Investigation into Cancer and Nutrition; FFQ, food-frequency questionnaire; GGT, g-glutamyltransferase; SHBG, sex hormone–binding globulin; T2D, type 2 diabetes. Received November 20, 2013. Accepted for publication June 9, 2014. First published online July 23, 2014; doi: 10.3945/ajcn.113.080317. Am J Clin Nutr 2014;100:891–900. Printed in USA. Ó 2014 American Society for Nutrition 891 892 JACOBS ET AL (8), or HDL cholesterol (3). However, these intervention studies are limited by their short durations and high coffee dosages. In prospective studies, adjustment for total cholesterol and triglycerides (9) or SHBG (10) led to an attenuation of the association between coffee intake and T2D risk, suggesting that this association might be partly mediated by dyslipidemia or SHBG. To our knowledge, a comprehensive investigation of the association between coffee consumption and diabetes-related biomarkers reflecting different biologic pathways under real-life conditions is lacking. Moreover, a potential mediating role of various biomarkers on the inverse association between coffee consumption and T2D risk has not been investigated. Identification of potential mediators can help to develop hypotheses regarding which metabolic pathways may play a role in the association between coffee consumption and T2D risk. Therefore, the aim of this study was to evaluate the association between coffee consumption and various diabetes-related biomarkers, reflecting different metabolic pathways, includ (...truncated)