(9) 6, 56 5 & 9 Macmillan Publishers Limited All rights reserved 95-7/9 $. www.nature.com/ejcn ORIGINAL ARTICLE Coffee consumption is not related to the metabolic syndrome at the age of 6 years: the Amsterdam Growth and Health Longitudinal Study MT Driessen, LLJ Koppes, L Veldhuis,,, D Samoocha and JWR Twisk, Department of Public and Occupational Health, EMGO Institute, VU University Medical Center, Amsterdam, The Netherlands; Faculty of Earth and Life Science, Institute of Health Sciences, Vrije Universiteit, Amsterdam, The Netherlands; Department of Public Health, Erasmus MC University Medical Center Rotterdam, Rotterdam, The Netherlands and Department of Clinical Epidemiology and Biostatistics, VU University Medical Center, Amsterdam, The Netherlands Background/Objectives: Coffee consumption has been postulated to decrease the risk of diabetes mellitus type II. The longterm effects of coffee consumption on the metabolic syndrome (MS) and its components are unknown. This study investigated the relationship of long-term coffee consumption between the age of 7 and 6 years with the prevalence of the MS at the age of 6 years. Subject/Methods: Data on coffee consumption and the MS components were derived from a healthy sample of 7 men and 9 women followed up from the age of 7 years onwards. Data analysis was performed with the use of generalized estimating equations and regression analysis. Results: At the age of 6 years, the prevalence of the MS was.%. The growth of coffee consumption did not differ significantly between subjects with or without the MS or its components. Regression analyses showed that one cup day higher coffee consumption was related to. mm Hg lower mean arterial blood pressure (P ¼.),. mg ml higher triglyceride level (P ¼ 7),. mg ml higher high-density lipoprotein cholesterol level (P ¼.5),.9% higher HbA c (P ¼.) and. cm larger waist circumference (P ¼ 7). After adjustment for physical activity, energy intake, smoking behaviour and alcohol consumption, none of the relationships between coffee consumption and the MS or its components was significant. Conclusions: Coffee consumption is not associated with the MS or its components in a healthy sample followed up for 9 years. (9) 6, 56 5; doi:.8/ejcn.8.6; published online February 8 Keywords: metabolic syndrome; coffee consumption; risk components; healthy sample Introduction The metabolic syndrome (MS) is an important risk factor of diabetes mellitus (DM) type II (Isomaa et al., ) and Correspondence: Professor Dr JWR Twisk, Department of Clinical Epidemiology and Biostatistics, VU University Medical Center, PO Box 757, Amsterdam 7 MB, The Netherlands. E-mail: jwr.twisk@vumc.nl Guarantors: JWR Twisk and MT Driessen. Contributors: MTD contributed to the manuscript and data analysis; LLJK contributed to the manuscript, study design, data collection and project planning; LV contributed to the manuscript and data analysis; DS contributed to the manuscript and data collection; JWRT contributed to the manuscript, study design, data analysis and coordinated the project. Received June 7; revised 8 December 7; accepted January 8; published online February 8 cardiovascular diseases (Wannamethee et al., 5). In the US,.7% of the adults aged X years (Ford et al., ) and 9.% of the children aged 9 years (de Ferranti et al., ) were estimated to be affected. The MS has become a major worldwide public health problem (Ford, 5). Coffee is one of the most widely consumed beverages (Schaefer, ) and various effects on human health have been postulated to coffee consumption. Recent prospective cohort studies showed that coffee consumption may lower the risk of DM (Tuomilehto et al., ; van Dam et al.,, 6; Hu et al., 6). Two reviews of cohort studies reported inverse associations between coffee consumption and the risk of DM (van Dam and Hu, 5; Greenberg et al., 6). Furthermore, it has been shown that coffee consumption was inversely associated with several glycaemic markers and
diabetes (Bidel et al., 6, 7). Despite the favourable effects of coffee on DM, the exact mechanism involved with the decreased risk of DM remains unclear (van Dam and Hu, 5). To date, only one cross-sectional study investigated the effect of coffee consumption on the MS (Hino et al., 7) and revealed that habitual coffee consumption was inversely associated with the MS and its components (except for high-density lipoprotein (HDL) cholesterol) in a general Japanese population. Given that the decreased risk of DM was associated with coffee consumption, it could be hypothesized that an association in the prediabetic stage such as the MS is plausible. However, no information on long-term coffee consumption and the MS and its risk components obtained from longitudinal studies is available. Therefore, the present study investigates the relationship between long-term coffee consumption and the prevalence of the MS and its components at the age of 6 years in a cohort of healthy men and women, who have been followed up from the age of 7 years onwards. Materials and methods Study design and sample This study is a part of the Amsterdam Growth and Health Longitudinal Study, an observational longitudinal study that started in 977 with a group of 6 boys and girls. Its initial goals were to describe the natural development of growth, health and lifestyle of adolescents and to investigate longitudinal relationships between biological and lifestyle variables. First- and second-year pupils from two equally large secondary schools in The Netherlands (level of education slightly above the Dutch average) were recruited and all pupils of these schools agreed to participate in the baseline measurements at the mean age of. (s.d..8) years. Pupils were not selected by the researchers and no selection criteria were used. Follow-up measurements have been performed at the age of, 5, 6,, 7, 9, and 6 years. At each follow-up, anthropometrical (body height, body weight and skin folds), biological (serum lipoprotein levels, blood pressure and physical fitness), lifestyle (nutritional habits, smoking behaviour and daily physical activity) and psychological variables were assessed. In the most recent follow up at the age of 6 years, the five components needed to identify the MS were assessed for the first time on 68 participants (9 women). An extensive assessment of coffee consumption was performed only at the ages of 7, 9, and 6 years. All participants of the Amsterdam Growth and Health Longitudinal Study with data available at the age of 6 years on the five components of the MS, and data available on coffee consumption between the age of 7 and 6 years were included in our analyses. None of the participants used lipidor blood pressure-lowering medication at the time of the study (Ferreira et al., 5a). The Amsterdam Growth and Health Longitudinal Study was approved by the medical ethics committee of the VU University Medical Center, and all subjects gave written informed consent. Identification of the MS and its components The identification of the MS and its components is based on a modified definition proposed by the National Cholesterol Education Program Adult Treatment Panel III (Grundy et al., ), namely, when three or more of the following five components are present: () a systolic blood pressure of mm Hg or higher and/or a diastolic blood pressure of 85 mm Hg or higher; () a HDL cholesterol level of less than mg ml (o. mm) in men and less than 5 mg ml (o.9 mm) in women; () a triglyceride level of 5 mg ml or more (X.69 mm); () a glycated haemoglobin (HbA c ) level higher than 6.% was used instead of fasting plasma glucose levels of mg ml or more (6. mm) (Davidson et al., 999), because glucose data were not available and (5) a waist circumference of more than 9 cm in men and more than 8 cm in women (Han et al., 995; Dobbelsteyn et al., ). Measurements of the MS and its components Blood pressure was measured on two occasions: () in a supine position with an automated device (model BP-88; Colin Press-Mate, Komaki City, Japan) blood pressure was assessed at 5-min intervals, for 5 6 min and () before a cardiopulmonary fitness test, two assessments were performed with a sphygmomanometer (SpeidlKeller; Franken & Itallie, Amsterdam, The Netherlands), after at least 5 min of rest in a sitting position and the lower value was recorded. The systolic and diastolic blood pressure values obtained from these two occasions were averaged and used in the regression analyses. HDL and triglyceride levels were determined using enzymatic techniques (Roche Diagnostics, Mannheim, Germany), and HbA c was determined using ion-exchange high-performance liquid chromatography with a Mono S column (Pharmacia, Uppsala, Sweden). Waist circumference was measured with a flexible steel tape (Martin circumeter; Franken & Itallie) at the level midway between the lowest rib margin and the iliac crest. Coffee consumption and other lifestyle covariates Data on coffee consumption were assessed on four occasions. First measurement was performed at the age of 7 years, the last (fourth) measurement was performed at the age of 6 years. Coffee consumption was not assessed before the age of 7 years. A questionnaire asking for the average number of cups of coffee consumed within eight periods of the day was used: before, with and after breakfast; with and after lunch; before and with dinner; during the evening and during the night. In the Netherlands, on average one cup is equivalent 57
58 to 5 ml (van Dam et al., ). We used all the available data on coffee consumption. Physical activity was assessed with a validated interview and expressed in metabolic equivalents min week, expressing the intensity of activity in comparison to resting energy expenditure, with one metabolic equivalent equal to the standard for resting expenditure (Ainsworth et al., 99). Energy intake was assessed with a cross-check dietary history interview (except at the age of 9 years). Subjects were asked to recall their usual dietary intake during the previous month, by reporting the frequency, amounts and methods of preparation of foods and drinks consumed, and whether the subjects were alcohol drinkers or abstainers (Koppes et al., ). Smoking behaviour (yes/no) was measured with a questionnaire (Bernaards et al., ). More detailed information of the methods used in the Amsterdam Growth and Health Longitudinal Study can be found elsewhere (Ferreira et al., 5b). Statistical analysis Generalized estimation equations (GEEs) were used to investigate the association between the 9-year growth of coffee consumption and the presence of the MS at the age of 6 years. GEE adjusts for the correlation between repeated observations taken in the same subjects and has the advantage of handling longitudinal data on subjects with varying numbers of unequally spaced observations (Twisk, ). An exchangeable correlation structure was assumed in all GEE analyses and time was treated as a categorical variable and represented by dummies. The presence or absence of the MS was added as an independent variable and the interaction between this variable and the time dummies was added to investigate the differences in the growth of coffee consumption between subjects with and without the MS. Due to the Poisson distribution of the number of cups of coffee day, Poisson GEE analyses were used. Because no significant interactions with gender were present, the analyses were performed for men and women together. Besides crude analyses, all models were additionally adjusted for gender, physical activity, energy intake, smoking behaviour and alcohol consumption. To assess the associations between long-term coffee consumption and the MS and its components, the average coffee consumption over the period from the age of 7 to 6 years was calculated. Associations were analysed by logistic regression and linear regression analyses were used for the MS components. Since all MS components were not normally distributed, log transformations were used. Two analyses were performed: () adjusting for gender and () adjusting for gender and lifestyle covariates. Physical activity and energy intake were also averaged over the period from the age of 7 to 6 years. For smoking behaviour and alcohol consumption, dichotomous information at the age of 6 years was used. The GEE analyses were performed with STATA version 7, and the linear and logistic regression analyses were performed with SPSS version.. Results Table presents descriptive information about the participants regarding gender distribution, coffee consumption, the prevalence of the MS and its components and the lifestyle covariates. A total of 9 women and 7 men were measured, having an average coffee consumption of. cups day. No excessive values on lifestyle covariates were found. The prevalence of the MS at the age of 6 years was.%. Figure a presents the growth of coffee consumption for subjects with or without the MS at the age of 6 years. Subjects with the MS had a higher coffee consumption at the age of 7, 9 and 6 years as compared with subjects without Table Characteristics of the participants (n ¼ 68) at the age of 6 years Characteristics Values MS and its components Prevalence (%) Arterial blood pressure, mean (s.d.), mm Hg a 9. (8.8) X/85 mm Hg a. Triglyceride, mean (s.d.), mm. (.9) X5 mg ml 7.7 HDL cholesterol, mean (s.d.), mm. (.) o mg ml in men.8 o5 mg ml in women HbA c, mean (s.d.), % 5. (.6) 6.%.9 Waist circumference, mean (s.d.), cm 78.6 (.) 9 cm in men 5. 8 cm in women MS. Covariates Women, no. (%) 9 (5.7) Coffee consumption, mean (s.d.), cups day. (.) Physical activity, mean (s.d.), MET min week (8) Energy intake, mean (s.d.), MJ day. (.7) Current smoking, no. (%) 86 (.) Current drinking, no. (%) 7 (88.9) Abbreviations: HbA c, glycated haemoglobin; HDL, high-density lipoprotein; MET, metabolic equivalent; MS, metabolic syndrome. a Mean arterial blood pressure was calculated as ( diastolic þ systolic blood pressure)/.
With the MS or the risk component at the age of 6 years Without the MS or the risk component at the age of 6 years 59 Coffee consumption, cups day -.5.5.5 Metabolic Syndrome 7 9 6 Coffee consumption, cups day -.5.5.5 Blood pressure /85 mmhg 7 9 6 Coffee consumption, cups day -.5.5.5 Triglyceride level 5mg ml - 7 9 6 Coffee consumption, cups day -.5.5 HDL cholesterol level < mg ml - in men or <5 mg ml - in women 7 9 6 Coffee consumption, cups day -.5.5 HbAc level >6.% 7 9 6 Age, years Coffee consumption, cups day -.5.5.5 Waist circumference >9 cm in men or >8 cm in women 7 9 6 Age, years Figure Growth of long-term coffee consumption (cups day ) of subjects with and without the MS (a) and with and without the MS components (b f) at the age of 6 years. the MS. At the age of years, however, the opposite was found. All differences were not significant and adjustment for gender and lifestyle covariates did not change these results. No significant differences in the growth of coffee consumption between subjects with and without the MS components were found (Figures b f). The large difference in the growth of coffee consumption between subjects with and without high HbA c was not significant due to the low prevalence of subjects with HbA c levels 6.% (that is,.9%). As for the MS, adjustment for gender and lifestyle covariates did not influence the findings on the MS components. Table shows the results of the regression analyses regarding the relationship between the 9-year average
5 Table Associations between average coffee consumption between 7 and 6 years of age and the metabolic syndrome and its components Variables Gender adjusted Gender and lifestyle adjusted Effect estimate P-value Effect estimate P-value Metabolic syndrome.9 (OR).6.96 (OR).6 Components Blood pressure, mm Hg a...9.8 Triglyceride level, mg ml. 7..9 HDL cholesterol level, mg ml..5.5.5 HbA c, %.9..8.7 Waist circumference, cm. 7.5. Abbreviations: HbA c, glycated haemoglobin; HDL, high-density lipoprotein; OR, odds ratio. a Mean arterial blood pressure was calculated as ( diastolic þ systolic blood pressure)/. amount of coffee consumption and the MS and its components at the age of 6 years. No significant association with the MS was found. Long-term coffee consumption was inversely associated with blood pressure (P ¼.) in the gender-adjusted analysis. This association, however, decreased and was no longer significant after additional adjustment for lifestyle covariates. For all the other components, no significant relationships with long-term coffee consumption were found. Discussion In the present study, no relationships were found between the 9-year growth of coffee consumption and the MS as well as the risk components of the MS. Only, coffee consumption appeared to be somewhat significantly inversely associated with blood pressure, but this relationship disappeared after adjustment for lifestyle covariates. Probably the best evidence for a positive health effect of coffee consumption has been found for DM. A meta-analyses reported that consuming more than 6 7 cups of coffee day, compared with a reference group ( cups of coffee day ), was associated with a 5% lower risk of DM (van Dam and Hu, 5). Inverse associations of coffee consumption on DM were found for caffeinated as well as for decaffeinated coffee (Salazar-Martinez et al., ; Greenberg et al., 5; van Dam et al., 6). These findings have been confirmed by a review (Greenberg et al., 6). Recently, a large cohort study revealed even stronger inverse associations for decaffeinated coffee than for regular coffee on the risk of DM in older women (Pereira et al., 6), indicating that coffee components other than caffeine are responsible for the inverse association with DM risk. A cross-sectional study (n ¼ 9) investigating the effects of coffee consumption on the MS in a general Japanese population showed that coffee consumption was inversely associated with the MS and all (except for HDL) its components (Hino et al., 7). According to the authors, caffeine was responsible for the associations. Caffeine is an adenosine receptor antagonist, and all tissues with adenosine receptors can be affected by caffeine exposure (van Soeren and Graham, 998). Furthermore, findings show that caffeine increases fat oxidation and mobilization of glycogen in muscle (Spriet et al., 99), increases lipolysis (Astrup et al., 99) and decreases body fat storage (Astrup et al., 99). In the present study, a weak inverse relationship was found between long-term coffee consumption and mean arterial blood pressure, which weakened after adjustment for lifestyle covariates. Regarding blood pressure, a meta-analysis showed that chronic coffee consumption was associated with an average increase of. mm Hg systolic and. mm Hg diastolic blood pressures (Jee et al., 999). More recently, it was suggested that caffeine ingested by coffee consumption only led to a small effect on blood pressure (. mm Hg systolic and.7 mm Hg diastolic blood pressures) (Noordzij et al., 5). For all other risk components in the present study, no associations with coffee consumption were found. In the literature, however, beneficial effects of coffee on glycaemic markers (Bidel et al., 6) have been found and positive associations between coffee consumption and insulin sensitivity have been reported (Agardh et al., ; Arnlov et al., ; van Dam et al., ). Furthermore, coffee consumption of both caffeinated and decaffeinated coffee was associated with a decrease in C-peptide levels in women, indicating that chronic coffee consumption reduces insulin secretion (Wu et al., 5). In our study, unfortunately, no information on insulin sensitivity or fasting glucose levels was available. However, there was information present on HbA c. Excessive glycosylation is important in the genesis and development of diabetes microvascular complications (Bucala and Vlassara, 995; Wolffenbuttel et al., 996). Therefore, we expected the growth of coffee consumption to differ between subjects with high risk on HbA c level and subjects with normal HbA c levels. However, the results concerning the growth of coffee consumption and the longterm exposure revealed no significant associations with HbA c. Regarding waist circumference, in a long-term study (over a period of years) it was pointed out that an increase in
caffeine intake mainly due to coffee consumption was associated with a smaller weight gain (reported as body mass index) compared with subjects who decreased their caffeine intake in the past years (Lopez-Garcia et al., 6a). Controversially, other studies found a relationship between coffee consumption and body mass index in an opposite direction (van Dam and Feskens, ; Rosengren et al., ; Tuomilehto et al., ). With respect to cholesterol, consuming six cups of coffee day was associated with increased low-density lipoprotein (LDL) cholesterol levels and not with HDL cholesterol levels (Jee et al., ). Raised LDL cholesterol levels were the result of the pot boiled method of coffee preparation, which may increase the risk of coronary heart disease (Zock et al., 99; Jee et al., ). Our study sample drank, however, most of the time drip-filtered caffeinated coffee, which was not found to be associated (whether caffeinated or decaffeinated) with plasma levels of total, LDL and HDL cholesterol levels (Lopez-Garcia et al., 6b). There are several reasons for not finding strong associations between coffee consumption and the MS and its risk components. First of all, the prevalence of the MS in our study sample was.%, while in the US these rates are.7% (Ford et al., ), with increasing prevalence rates of the MS when age increases. The low MS prevalence rate of % found by Hino et al. (7) was in agreement with the prevalence rates found in our study; however, they investigated an older cohort (mean 6.7 years). Our findings indicate that the age of 6 years might be too young to find a relation between the MS and coffee consumption. Second, the meta-analyses of van Dam and Hu (5) suggested a stronger inverse association between high amount of coffee consumption (6 or 7) and DM incidence. The study of Hino et al. (7), on the other hand, reported effects on the MS when on average one cup of coffee day was consumed. Our study sample consumed per person on average. cups of coffee day and only a few subjects drank seven or more cups of coffee day. Additional analysis of our data showed that the few subjects with (extreme) high coffee intake were not likely to have less risk components of the MS. Third, most of the evidence of the inverse association between coffee consumption and DM were obtained from large cross-sectional studies. We, however, studied a smaller study sample (n ¼ 68), therefore it was plausible to find fewer participants with the MS or its components. Finally, we did not obtain information about coffee additives that may have played a role in the observed relationships. Participants might differ in coffee-drinking habits (for example, adding milk, full-fat cream or sugar in their coffee), which could have led to variability in glucose, free fatty acids and daily caloric intake. Subsequently, free fatty acids are a major contributor to the development of insulin resistance and have influence on cholesterol levels. In conclusion in this study sample, the 9-year growth of coffee consumption, measured from the age of 7 to 6 years, is not different for subjects with or without the MS, or with or without its components. The 9-year average amount of coffee consumption and the MS and its components at the age of 6 years was not only inversely associated with mean arterial blood pressure and not with the MS or the other components. However, the relationship with mean arterial blood pressure was weak and disappeared after correction for lifestyle covariates. As a result of the small and relatively young study sample, it is premature to conclude that coffee has no inverse association with the MS. More information on this topic obtained from large (longitudinal) cohorts including various age groups is needed. References Agardh EE, Carlsson S, Ahlbom A, Efendic S, Grill V, Hammar N et al. (). Coffee consumption, type diabetes and impaired glucose tolerance in Swedish men and women. J Intern Med 55, 65 65. Ainsworth BE, Haskell WL, Leon AS, Jacobs Jr DR, Montoye HJ, Sallis JF et al. (99). Compendium of physical activities: classification of energy costs of human physical activities. Med Sci Sport Exerc 5, 7 8. Arnlov J, Vessby B, Riserus U (). Coffee consumption and insulin sensitivity. JAMA 9, 99. Astrup A, Andersen T, Christensen NJ, Bülow J, Madsen J, Breum L et al. (99). Impaired glucose-induced thermogenesis and arterial norepinephrine response persist after weight reduction in obese humans. Am J Clin Nutr 5, 7. Astrup A, Breum L, Toubro S, Hein P, Quaade F (99). The effect and safety of an ephedrine/caffeine compound compared to ephedrine, caffeine and placebo in obese subjects on a energy restricted diet. A double blind trial. Int J Obes Relat Metab Disord 6, 69 77. Bernaards CM, Twisk JW, van Mechelen W, Snel J, Kemper HC (). Comparison between self-report and a dipstick method (NicCheck) to assess nicotine intake. Eur Addict Res, 6 67. Bidel S, Hu G, Sundvall J, Karpio J, Tuemilehto J (6). Effects of coffee consumption on glucose tolerance, serum glucose and insulin levels a cross-sectional analysis. Horm Metab Res 8, 8. Bidel S, Silventoinen K, Hu G, Lee DH, Kaprio J, Tuomilehto J (7). Coffee consumption, serum gamma-glutamyltransferase and risk of type II diabetes. Eur J Clin Nutr, 8 [Epub ahead of print]. Bucala R, Vlassara H (995). Advanced glycosylation end products in diabetic renal and vascular disease. Am J Kidney Dis 6, 875 888. Davidson MB, Schriger DL, Peters AL, Lorber B (999). Relationship between fasting plasma glucose and glycosylated hemoglobin: potential for false-positive diagnoses of type diabetes using new diagnostic criteria. JAMA 8,. de Ferranti SD, Gauvreau K, Ludwig DS, Neufeld EJ, Newburger JW, Rifai N (). Prevalence of the metabolic syndrome in American adolescents: findings from the Third National Health and Nutrition Examination Survey. Circulation, 9 97. Dobbelsteyn CJ, Joffres MR, MacLean DR, Flowerdew G (). A comparative evaluation of waist circumference, waist-to-hip ratio and body mass index as indicators of cardiovascular risk factors. The Canadian Heart Health Surveys. Int J Obes Relat Metab Disord 5, 65 66. Ferreira I, Henry RM, Twisk JW, van Mechelen W, Kemper HC, Stehouwer CD (5a). The metabolic syndrome, cardiopulmonary fitness, and subcutaneous trunk fat as independent determinants of arterial stiffness: the Amsterdam Growth and Health Longitudinal Study. Arch Intern Med 65, 875 88. Ferreira I, Twisk JW, van MW, Kemper HC, Stehouwer CD (5b). Development of fatness, fitness, and lifestyle from adolescence to the age of 6 years: determinants of the metabolic syndrome in 5
5 young adults: the Amsterdam Growth and Health Longitudinal Study. Arch Intern Med 65, 8. Ford ES (5). Prevalence of the metabolic syndrome defined by the International Diabetes Federation among adults in the US. Diabetes Care 8, 75 79. Ford ES, Giles WH, Dietz WH (). Prevalence of the metabolic syndrome among US adults: findings from the third National Health and Nutrition Examination Survey. JAMA 87, 56 59. Greenberg JA, Axen KV, Schnoll R, Boozer CN (5). Coffee, tea and diabetes: the role of weight loss and caffeine. Int J Obes (Lond) 9, 9. Greenberg JA, Boozer CN, Geliebter A (6). Coffee, diabetes, and weight control. Am J Clin Nutr 8, 68 69. Grundy SM, Brewer Jr HB., Cleeman JI, Smith Jr SC., Lenfant C (). Definition of metabolic syndrome: report of the National Heart, Lung, and Blood Institute/American Heart Association conference on scientific issues related to definition. Arterioscler Thromb Vasc Biol, e e8. Han TS, van Leer EM, Seidell JC, Lean ME (995). Waist circumference action levels in the identification of cardiovascular risk factors: prevalence study in a random sample. BMJ, 5. Hino A, Adachi H, Enomoto M, Furuki K, Shigetoh Y, Ohtsuka M et al. (7). Habitual coffee but not green tea consumption is inversely associated with metabolic syndrome: an epidemiological study in a general Japanese population. Diabetes Res Clin Pract 76, 8 89. Hu G, Jousilahti P, Peltonen M, Bidel S, Tuomilehto J (6). Joint association of coffee consumption and other factors to the risk of type diabetes: a prospective study in Finland. Int J Obes (Lond), 7 79. Isomaa B, Almgren P, Tuomi T, Forsen B, Lahti K, Nissen M et al. (). Cardiovascular morbidity and mortality associated with the metabolic syndrome. Diabetes Care, 68 689. Jee SH, He J, Appel LJ, Whelton PK, Suh I, Klag MJ (). Coffee consumption and serum lipids: a meta-analysis of randomized controlled clinical trials. Am J Epidemiol 5, 5 6. Jee SH, He J, Whelton PK, Suh I, Klag MJ (999). The effect of chronic coffee drinking on blood pressure: a meta-analysis of controlled clinical trials. Hypertension, 67 65. Koppes LL, Twisk JW, Snel J, van Mechelen W, Kemper HC (). Blood cholesterol levels of -year-old alcohol consumers are better than of nonconsumers. Pharmacol Biochem Behav 66, 6 67. Lopez-Garcia E, van Dam RM, Rajpathak S, Willett WC, Manson JE, Hu FB (6a). Changes in caffeine intake and long-term weight change in men and women. Am J Clin Nutr 8, 67 68. Lopez-Garcia E, van Dam RM, Willet WC, Rimm EB, Manson JE, Stampfer MJ et al. (6b). Coffee consumption and coronary heart disease in men and women. A prospective cohort study. Circulation, 5 5. Noordzij M, Uiterwaal CS, Arends LR, Kok FJ, Grobbee DE, Geleijnse JM (5). Blood pressure response to chronic intake of coffee and caffeine: a meta-analysis of randomized controlled trials. J Hypertens, 9 98. Pereira MA, Parker ED, Folsom AR (6). Coffee consumption and risk of type diabetes mellitus: an -year prospective study of 8 8 postmenopausal women. Arch Intern Med 66, 6. Rosengren A, Dotevall A, Wilhelmsen L, Thelle D, Johansson S (). Coffee and incidence of diabetes in Swedish women: a prospective 8-year follow-up study. J Intern Med 55, 89 95. Salazar-Martinez E, Willett WC, Ascherio A, Manson JE, Leitzmann MF, Stampfer MJ et al. (). Coffee consumption and risk for type diabetes mellitus. Ann.Intern Med, 8. Schaefer B (). Coffee consumption and type diabetes mellitus. Ann Intern Med,. Spriet LL, MacLean DA, Dyck DJ, Hultman E, Cederblad G, Graham TE (99). Caffeine ingestion and muscle metabolism during prolonged exercise in humans. Am J Physiol 6, E89 E898. Tuomilehto J, Hu G, Bidel S, Lindstrom J, Jousilahti P (). Coffee consumption and risk of type diabetes mellitus among middleaged Finnish men and women. JAMA 9, 9. Twisk JWR (). Applied Longitudinal Data Analysis for Epidemiology: a Practical Guide. Cambridge University Press: Cambridge, England. van Dam RM, Dekker JM, Nijpels G, Stehouwer CD, Bouter LM, Heine RJ (). Coffee consumption and incidence of impaired fasting glucose, impaired glucose tolerance, and type diabetes: the Hoorn Study. Diabetologia 7, 5 59. van Dam RM, Feskens EJ (). Coffee consumption and risk of type diabetes mellitus. Lancet 6, 77 78. van Dam RM, Hu FB (5). Coffee consumption and risk of type diabetes: a systematic review. JAMA 9, 97. van Dam RM, Willett WC, Manson JE, Hu FB (6). Coffee, caffeine, and risk of type diabetes: a prospective cohort study in younger and middle-aged U.S. women. Diabetes Care 9, 98. van Soeren MH, Graham TE (998). Effect of caffeine on metabolism, exercise endurance, and catecholamine responses after withdrawal. J Appl Physiol 85, 9 5. Wannamethee SG, Shaper AG, Lennon L, Morris RW (5). Metabolic syndrome vs Framingham Risk Score for prediction of coronary heart disease, stroke, and type diabetes mellitus. Arch Intern Med 65, 6 65. Wolffenbuttel BH, Girdano D, Founds HW, Bucala R (996). Longterm assessment of glucose control by haemoglobin-age measurement. Lancet 7, 5 55. Wu T, Willett WC, Hankinson SE, Giovannucci E (5). Caffeinated coffee, decaffeinated coffee, and caffeine in relation to plasma C-peptide levels, a marker of insulin secretion, in U.S. women. Diabetes Care 8, 9 96. Zock PL, Katan MB, Merkus MP, van DM, Harryvan JL (99). Effect of a lipid-rich fraction from boiled coffee on serum cholesterol. Lancet 5, 5 7.