![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
|
|
|
|
|
||
,3
* Cardiology Clinic, School of Medicine, University of Athens, Athens, Greece and
Cardiology Division, Veterans Affairs Medical Center, George Washington University, Washington, DC
3To whom correspondence should be addressed. E-mail: d.b.panagiotakos{at}usa.net.
 |
ABSTRACT |
|---|
 TOP ABSTRACT SUBJECTS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
KEY WORDS: • coronary risk • coffee • acute coronary syndrome
Conflicting information exists regarding the effect of coffee consumption on the cardiovascular system. Some authors have reported a positive association (1–3) between coffee intake and coronary heart disease, whereas others have reported no relationship (4–6). Moreover, several investigators have suggested that the increase in coronary heart disease risk may be the result of elevated blood pressure and cholesterol levels due to unhealthy diet habits or increased smoking habits associated with coffee consumption (7–13). It seems that an independent association between coffee consumption and the risk of developing coronary heart disease has not yet been established. Furthermore, epidemiologic studies at the population level that examine the association of coffee consumption in various quantities with the risk of developing nonfatal acute coronary syndromes are rather limited in the medical literature.
To evaluate the association between various quantities of coffee consumption and the risk of developing acute coronary syndromes, we studied 
2000 coronary patients and controls from the CARDIO2000 study (14–17). The relationship was adjusted for multiple potential confounding risk factors, including the conventional cardiovascular risk factors, the presence of depression, nutritional habits and the educational level of the participants.
|
SUBJECTS AND METHODS |
|---|
|
TOPABSTRACTSUBJECTS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
The CARDIO2000 is a multicenter case-control study that explores the association between several demographic, nutritional, psychological, lifestyle and clinical factors with the risk of developing nonfatal acute coronary syndromes. According to the population distribution provided by the National Statistical Services we stratified our sampling into all regions of Greece. For each region, we randomly selected a specific number of patients from the prefectorial or the major private hospitals of each county. The patients were from approximately half of the clinics in Athens and Thessalonica (the two major cities in Greece, covering 55% of the total population), and from almost all of the clinics of the other counties. The number of participants was determined through power analysis to evaluate a minimum difference of 7% in relative risk between those who do and do not drink coffee with statistical power of 0.80 and P < 0.05. Thus, from January 2000 to August 2001, 848 of 956 patients (89% response rate) who had been randomly preselected from the hospitals with a first symptom of coronary heart disease (stable angina was excluded from the analysis) in their life agreed to participate in the study (cases). The inclusion criteria for cardiac cases were as follows: 1) diagnosis of first acute myocardial infarction that was defined by two of three features, i.e., electrocardiographic changes, or compatible clinical symptoms or/and specific diagnostic enzyme elevations (49% of the patients had myocardial infarction); or 2) diagnosis of unstable angina (i.e., one or more angina episodes at rest within the preceding 48 h with no abnormal enzyme rise) corresponding to class III of the Braunwald classification (51% of the patients had unstable angina). Patients with a previous history of cardiovascular disease, including aortic aneurysm, stroke and peripheral vascular disease, were excluded from this study.
After the selection of the cardiac patients, we randomly selected 1300 controls who were free of cardiovascular disease; 1078 of these (83% response rate) agreed to participate. The controls were frequency matched to the patients according to their age distribution (within classes of ± 3 y), their gender and the region of Greece (we matched the controls with the patients by region to reduce the potential confounding effect of culture differences among Greek citizens). Controls were mainly individuals (91% of the total number of controls) who visited the outpatient departments of the same hospital during the same time period as the coronary patients for minor surgical operations (e.g., bone fracture) or for routine examinations (e.g., medical examinations for the driving license or other official certificates such as validation of the social security handbooks). We used these types of controls to have more accurate medical information, to eliminate the potential adverse effect of several, unknown, confounders and to increase the likelihood that cases and controls shared the same study base (18). However, in a few country hospitals in which the available number of "hospitalized" controls was not sufficient for the matching procedure, we randomly selected a small proportion of people (9% of the total number of controls) from the municipality rolls. An indicator variable was used for the sensitivity analysis that was applied to explore whether the different "sources" of controls might influence our findings. All controls were people without any clinical symptoms, signs or suspicion of cardiovascular disease in their medical history, including stroke, aortic aneurysm and peripheral vascular disease, as evaluated by a cardiologist.
As mentioned, to reduce the unbalanced distribution of several measured or unmeasured confounders, both patients and controls were randomly selected. The randomization procedure was based on a sequence of binary numbers (i.e., 1, 0, 0, 0, 1, 1 ...) that was first applied in the cardiology clinics’ admission lists. Thus, the coronary patients assigned the number 1 were selected and asked to participate (this procedure ensured the selection of approximately the half of the cardiac patients that visited each cardiology clinic). The same procedure (i.e., admission listings) was applied for the controls, after taking into account the matching and exclusion criteria. In the case of population controls, the random selection was obtained through the municipal rolls.
Investigated measurements.
All participants were asked to describe their usual frequency of consumption of coffee over the last year. Based on the distribution of coffee consumption, we categorized usual coffee consumption as none, up to 300 mL/d (moderate use), or up to 600 mL/d (heavy use) and >600 mL/d (very heavy use). All reported types of coffee (instant coffee, "Greek" type, filtered or "cappuccino") were adjusted for one cup (150 mL) of coffee and a caffeine concentration of 27.5% (19). We did not include the consumption of decaffeinated coffee, tea and caffeine-containing drinks (Coca-Cola) or chocolate. Ceasing to drink coffee during the last year (in months of abstinence) was recorded and taken into account as a covariate for the analysis.
We defined current smokers as those who smoked at least 1 cigarette/d. Former smokers were defined as subjects who had stopped smoking for >1 y. Individuals who reported that they had never smoked a cigarette were considered to be never smokers. Occasional smokers were excluded from the analysis due the small number in cases (n = 12) and controls (n = 21). Educational level was measured by years of schooling. Clinical symptoms of depression during the previous month were evaluated using the Center of Epidemiological Studies-Depression scale (range 0–60). Values > 15 on this scale indicate significant depressive episodes (20). Physical activity was defined as any type of nonoccupational physical exercise, at least once per week during the past year, and was gradated in qualitative terms such as light (expended energy <16 kJ/min, e.g., walking slowly, cycling stationary, light stretching), moderate (expended energy 16–28 kJ/min, e.g., walking briskly, cycling outdoors, swimming with moderate effort) and vigorous (expended energy >28 kJ/min, e.g., walking briskly uphill, long distance running, cycling fast or racing, swimming fast crawl). The rest of the subjects were defined as physically inactive. Also, the duration (in minutes) per session and the years of leisure time physical activity were taken into account as covariates in the multivariate analysis. The evaluation of usual nutritional habits and the classification as those who did or did not follow the Mediterranean diet (characterized by a pattern that is high in fruits, vegetables, bread, cereals, potatoes, poultry, beans, nuts and fish, little red meat and dairy products, and moderate alcohol consumption, and with olive oil as an important source of fat) was based on a questionnaire from the Department of Nutrition of the National School of Public Health (19). Finally, alcohol consumption was measured by daily ethanol intake, in wine glasses of 100 mL, and adjusted for 12% ethanol concentration.
In keeping with the long-standing classification criteria used in several epidemiologic studies, subjects who reported that they have been told they had hypertension, were currently taking any antihypertensive medication and whose blood pressure was >140/90 mm Hg were classified as hypertensive (14). Moreover, during the first 12 h of hospitalization for the patients and at the end of the interview for the controls, we measured serum fasting total cholesterol and blood glucose levels. Also, the majority of the controls (86%) and patients (83%) had at least one laboratory measurement in their medical records during the past year. The average values of total cholesterol and blood glucose levels, as well as the patients’ medical records and reports assisted us in classifying the participants as hypercholesterolemics and diabetics. In particular, according to National Cholesterol Education Program Adult Treatment Panel II criteria, hypercholesterolaemics were those who received special hypolipidemic treatment or had fasting blood measurement of serum total cholesterol > 5.68 mmol/L or > 5.16 mmol/L in the cases of subjects having > 2 other risk factors for coronary heart disease (hypertension, cigarette smoking, diabetes and positive family history for primary coronary heart disease). Diabetics were those with fasting blood glucose > 6.94 mmol/L or who were receiving a special diet or treatment. BMI was calculated in all participants as weight (kg)/height2 (m2). Finally, premature history of myocardial infarction or sudden death in first- or second-degree relatives (<55 y old for men and <65 y old for women) led to a classification in the positive family history group.
All participants were informed about the purposes of the study and agreed to participate. The CARDIO2000 study was approved by the Institutional Ethics Committee Further details regarding the design and methodology of the study were published previously (14–17).
Statistical analysis.
Continuous variables are presented as means ± 1 SD. Qualitative variables are presented as absolute and relative frequencies. Contingency tables with calculation of the 
2 test, as well as application of Student’s t test, evaluated associations between the categorical and normally distributed continuous variables. The estimates of the relative risks of developing acute coronary syndromes were made by calculating odds ratios (OR) and their corresponding 95% CI through conditional logistic regression analysis. First we evaluated all of the interactions between the exposure variables and the main factor of interest (coffee consumption). Then we explored the potential confounding effect of the other exposure variables (by adding and removing each one from the model). Both elimination procedures were based on the Wald statistic. Significant confounders as well as interactions were retained in the model. Deviance residuals were calculated to evaluate the model’s goodness-of-fit. A sensitivity analysis was also applied to evaluate the hypothesis tested in various subgroups of participants.
The "link function" (association) between the OR and coffee consumption (none, moderate, and heavy) was evaluated through second-order polynomial interpolation using Lagrange multipliers (21). All reported probability values were based on two-sided tests and compared with a significance level of P = 0.05. STATA 6 software was used for the all the calculations (Stata, College Station, TX).
|
RESULTS |
|---|
|
TOPABSTRACTSUBJECTS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTSUBJECTS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Although several studies in the past have attempted to assess the role of coffee consumption on cardiovascular disease incidence, the information remains inconsistent. Based on a large sample of coronary patients and cardiovascular disease–free controls, this study is one of the few in the literature that suggests a J-shaped association between coffee consumption and the risk of developing acute coronary syndromes. This finding may explain in part the contradictory results reported by earlier studies (1–7). It is important to mention that different brewing methods alter the concentrations of kahweol and cafestol, the most likely cholesterol-raising factors in unfiltered coffee; this may explain the discrepant results in studies from different parts of the world (22,23). In addition, Marchioli et al. (24) suggested that cohort studies that evaluate the effect of nutritional habits on human health cannot readily account for changes in lifestyle habits that occur during the observation period. Thus, they suggest that negative results of cohort studies could be associated with a modification of lifestyle habits (including stopping drinking coffee and lowering intake in addition to changing the type of coffee consumed) that is due either to aging of the cohort or to a worsening health status during follow-up. Furthermore, Lang et al. (7) investigated the relationships and possible interactions between coffee consumption and several potential confounders such as tobacco consumption, blood pressure levels, age and BMI, in a cross-sectional epidemiologic study. They suggested that the contradictory interactions observed among blood pressure levels, cigarette smoking and coffee drinking might help to explain the conflicting reports concerning the association between coffee consumption and ischemic heart disease. However, the present study revealed the J-shaped association between coffee consumption and coronary risk, independently of the aforementioned confounding factors.
A discussion of the mechanisms responsible for the effect of coffee consumption on the coronary risk is beyond the scope of this study. However, some investigators have reported that an increase in homocysteine levels as a result of coffee consumption could at least partially explain its positive effect on coronary risk (25) and thromboembolic stroke risk in hypertensives (26). Thus, it remains to be investigated whether the observed J-shaped association is explained in part by the association between coffee consumption and homocysteine levels. It has also been reported that a lipid fraction of boiled coffee has been shown to significantly raise cholesterol and LDL levels in a dose-dependent manner, whereas filtered coffee does not contain this lipid-elevating fraction (27). We speculate that the observed J-shaped association between boiled coffee consumption and the risk of developing acute coronary syndromes could be due to this dose-response association between lipid levels and boiled coffee intake. However, the present analysis revealed that the variant effect of coffee consumption on coronary risk was independent of serum total cholesterol and blood glucose levels of the participants; thus the aforementioned potential explanation may not hold. On the other hand, our inability to take into account the method of coffee preparation may limit the value of the previous hypothesis. Moreover, unfiltered coffee may have other, nonlipid-related effects on coronary risk that were not investigated in the present study. Finally, Roginsky and Barsukova (28) reported that the protective effects of coffee on atherosclerosis and other diseases can be attributed to the antioxidant activity of polyphenols. In particular, they observed that the inclusion of soluble coffee and other beverages in the testing system resulted in a pronounced retardation of methyl linoleate oxidation. Unfortunately, the present study could not test the antioxidant activity of polyphenols in the link between coffee and coronary risk, thereby explaining the observed association by this mechanism.
In retrospective case-control studies, two main sources of systematic errors may exist, i.e., the selection and the recall bias. To eliminate selection bias, we tried to set objective criteria for both patients and controls. However, insignificant misclassification may exist because a small percentage of asymptomatic coronary patients may be wrongly assigned as controls, even if a cardiologist had evaluated them. The use of "hospitalized" controls instead of a population-based group may increase the accuracy of the biomedical evaluation of the participants, but could overestimate the findings because they may differ from the general population by being more health conscious. Also, one of main limitations of case-control studies is their failure to recognize that in the case of myocardial infarction (likely less for unstable angina), a substantial proportion of cases never reach the hospital because they die outside of the hospital setting. Furthermore, in case-control studies, it is usually observed that patients who had a recent adverse event are more likely to place greater emphasis (overestimate) on several factors related to the disease than the control group (recall bias). To reduce this type of bias and analyze accurate information, we made an effort to obtain precise information from the patients as well as from their relatives or accompanying persons and compare their responses with Kendall’s ô-criterion. However, overestimation of the effect of the investigated combination may still exist. For the medical information, we tried to avoid recall bias by obtaining accurate and detailed data from the subjects’ medical records. However, over-/underestimation may exist, especially in the measurement of nutritional and smoking habits, and the onset of the investigated cardiovascular risk factors. Moreover, the coronary patients who died at entry or during the first 12 h were not included in the study. This bias could influence our results, but because the proportion of deaths during d 1 was estimated to be between 2 and 4% by the physicians of the study, we believe that the inability to include the fatal events did not significantly alter our findings. Furthermore, the potential effect of uncontrolled/unknown confounders was reduced by using the same study base for both patients and controls.
In conclusion, our findings offer a message for public health, i.e., moderate coffee consumption may be of benefit because it seems to be associated with lower coronary heart disease risk, whereas heavy consumption is associated with a significant increase of the previous risk. The J-shaped association noted may be a partial explanation for the contradictory findings reported by other studies. However, it is hard to claim that our findings suggest evidence of causality and explain the pathophysiologic mechanisms that link coffee consumption and coronary risk. Thus, prospective studies in the field of clinical epidemiology are required to refute or confirm our findings.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
2 Supported by research grants from the Hellenic Heart Foundation (ELIKAR-11/1999–2002).
Manuscript received 8 June 2003. Initial review completed 27 June 2003. Revision accepted 7 July 2003.
|
LITERATURE CITED |
|---|
|
TOPABSTRACTSUBJECTS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
1. Klag, M. J., Mead, L. A., La Croix, A. Z., Wang, N. Y., Coresh, J., Liang, K. Y., Pearson, T. A. & Levine, D. M. (1994) Coffee intake and coronary heart disease. Ann. Epidemiol. 4:425-433.[Medline]
2. Battig, K. (1992) Cardiovascular effects of everyday coffee consumption. Schweiz. Med. Wochenschr. 122:1536-1543.[Medline]
3. Schwarz, B., Bischof, H. P. & Kunze, M. (1990) Coffee and cardiovascular risk: epidemiological findings in Austria. Int. J. Epidemiol. 19:894-898.
4. Palmer, J. R., Rosenberg, L., Rao, R. S. & Shapiro, S. (1995) Coffee consumption and myocardial infarction in women. Am. J. Epidemiol. 141:724-731.
5. Grobbee, D. E., Rimm, E. B., Giovannucci, E., Colditz, G., Stampfer, M. & Willett, W. (1990) Coffee, caffeine, and cardiovascular disease in men. N. Engl. J. Med. 323:1026-1032.[Abstract]
6. Gyntelberg, F., Hein, H. O., Suadicani, P. & Sorensen, H. (1995) Coffee consumption and risk of ischemic heart disease—a settled issue?. J. Intern. Med. 237:55-61.[Medline]
7. Lang, T., Bureau, J. F., Degoulet, P., Salah, H. & Benattar, C. (1983) Blood pressure, coffee, tea and tobacco consumption: an epidemiological study in Algiers. Eur. Heart J. 4:602-607.
8. Burke, V., Beilin, L. J., German, R., Grosskopf, G. S., Ritchie, J., Puddey, I. B. & Rogers, P. (1992) Association of lifestyle and personality characteristics with blood pressure and hypertension: a cross-sectional study in the elderly. J. Clin. Epidemiol. 45:1061-1070.[Medline]
9. Lowik, M.R.H., Hofman, Z., Kok, F. J., Wedel, M., Hulshof, K.F.A., Odink, J. & Schaafsma, G. (1991) Nutrition and blood pressure among elderly men and women (Dutch Nutrition Surveillance System). J. Am. Coll. Nutr. 10:149-155.[Abstract]
10. Salvaggio, A., Periti, M., Miano, L. & Zambelli, C. (1990) Association between habitual coffee consumption and blood pressure levels. J. Hypertens. 8:585-590.[Medline]
11. Salvaggio, A., Periti, M., Quaglia, G., Marzorati, D. & Tavanelli, M. (1992) The independent effect of habitual cigarette and coffee consumption on blood pressure. Eur. J. Epidemiol. 8:776-782.[Medline]
12. Kirchhoff, M., Torp-Pedersen, C., Hougaard, K., Jacobsen, T. J., Sjol, A., Munch, M., Tingleff, J., Jorgensen, T., Schroll, M. & Olsen, M. E. (1994) Casual blood pressure in a general Danish population: relation to age, sex, weight, height, diabetes, serum lipids and consumption of coffee, tobacco and alcohol. J. Clin. Epidemiol. 47:469-474.[Medline]
13. Klag, M. J., Wang, N.-Y., Meoni, L., Brancati, F., Cooper, L., Liang, K.-Y., Young, H. & Ford, D. (2002) Coffee intake and risk of hypertension: The Johns Hopkins Precursors Study. Arch. Intern. Med. 162:657-662.
14. Panagiotakos, D. B., Pitsavos, C., Chrysohoou, C., Stefanadis, C. & Toutouzas, P. K. (2001) Risk stratification of coronary heart disease through established and emerging lifestyle factors, in a Mediterranean population: CARDIO2000 Epidemiological Study. J. Cardiovasc. Risk 8:329-335.[Medline]
15. Panagiotakos, D. B., Pitsavos, C., Chrysohoou, C., Moraiti, A., Stefanadis, C. & Toutouzas, P.K. (2001) The effect of short-term depressive episodes in the risk stratification of acute coronary syndromes: a case-control study in Greece (CARDIO2000). Acta Cardiol. 56:357-365.[Medline]
16. Panagiotakos, D. B., Chrysohoou, C., Pitsavos, C., Tzioumis, K., Papaioannou, I., Stefanadis, C. & Toutouzas, P. K. (2002) The association of Mediterranean diet with lower risk of acute coronary syndromes, in hypertensive subjects. Int. J. Cardiol. 82:141-147.[Medline]
17. Pitsavos, C., Panagiotakos, D. B., Chrysohoou, C., Skoumas, J., Stefanadis, C. & Toutouzas, P. K. (2002) Education and acute coronary syndromes: results from the CARDIO2000 epidemiological study. Bull. WHO 80:371-377.[Medline]
18. MacMahon, B. & Trichopoulos, D. (1996) Epidemiology. Principles and Methods 2nd ed. 1996 Little, Brown and Company New York, NY.
19. Trichopoulou, A. (1992) Composition of Greek Foods and Dishes (in Greek and English) 1992 Athens School of Public Health Athens, Greece.
20. Radloff, L. S. (1977) The CES-D scale: a self-report depression scale for research in the general population. Appl. Psychol. Meas. 1:385-401.
21. Curtis, G. (1978) Applied Numerical Analysis 1978 Addison-Wesley Reading, MA.
22. Greenland, S. (1993) A meta-analysis of coffee, myocardial infarction and coronary death. Epidemiology 4:366-374.[Medline]
23. Thelle, D. S. (1995) Coffee, tea, and coronary heart disease. Curr. Opin. Lipidol. 6:25-27.[Medline]
24. Marchioli, R., Di Mascio, R., Marfisi, R. M., Vitullo, F. & Tognoni, G. (1996) Coffee intake and death from coronary heart disease [Letter]. Br. Med. J. 312:1539.
25. Nieto, F. J., Comstock, G. W., Chamblers, L. E. & Malinow, R. M. (1997) Coffee consumption and plasma homocysteine results from the Atherosclerosis Risk in Communities Study. Am. J. Clin. Nutr. 66:1475-1477.
26. Hakim, A. A., Ross, G. W., Curb, J. D., Rodriquez, B. L., Burchfiel, C. M., Sharp, D. S., Yano, K. & Abbott, R. D. (1998) Coffee consumption in hypertensive men in older middle-age and the risk of stroke: the Honolulu Heart Program. J. Clin. Epidemiol. 51:487-494.[Medline]
27. Pirich, C., O’Grady, J. & Sinzinger, H. (1993) Coffee, lipoproteins and cardiovascular disease. Wien. Klin. Wochenschr 105:3-6.[Medline]
28. Roginsky, V. & Barsukova, T. (2001) Chain-breaking antioxidant capability of some beverages as determined by the Clark electrode technique. J. Med. Food 4:219-229.[Medline]
This article has been cited by other articles:
|
|
 |
|
  M. G. Silletta, R. Marfisi, G. Levantesi, A. Boccanelli, C. Chieffo, M. Franzosi, E. Geraci, A. P. Maggioni, G. Nicolosi, C. Schweiger, et al. Coffee Consumption and Risk of Cardiovascular Events After Acute Myocardial Infarction: Results From the GISSI (Gruppo Italiano per lo Studio della Sopravvivenza nell'Infarto miocardico)-Prevenzione Trial Circulation, December 18, 2007; 116(25): 2944 - 2951. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. Natella, M. Nardini, F. Belelli, and C. Scaccini Coffee drinking induces incorporation of phenolic acids into LDL and increases the resistance of LDL to ex vivo oxidation in humans Am. J. Clinical Nutrition, September 1, 2007; 86(3): 604 - 609. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. A. Rosner, A. Akesson, M. J. Stampfer, and A. Wolk Coffee Consumption and Risk of Myocardial Infarction among Older Swedish Women Am. J. Epidemiol., February 1, 2007; 165(3): 288 - 293. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Vlachopoulos, N. Alexopoulos, I. Dima, K. Aznaouridis, I. Andreadou, and C. Stefanadis Acute Effect of Black and Green Tea on Aortic Stiffness and Wave Reflections J. Am. Coll. Nutr., June 1, 2006; 25(3): 216 - 223. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. Lopez-Garcia, R. M. van Dam, W. C. Willett, E. B. Rimm, J. E. Manson, M. J. Stampfer, K. M. Rexrode, and F. B. Hu Coffee Consumption and Coronary Heart Disease in Men and Women: A Prospective Cohort Study Circulation, May 2, 2006; 113(17): 2045 - 2053. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. C. Cornelis, A. El-Sohemy, E. K. Kabagambe, and H. Campos Coffee, CYP1A2 Genotype, and Risk of Myocardial Infarction JAMA, March 8, 2006; 295(10): 1135 - 1141. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Vlachopoulos, D. Panagiotakos, N. Ioakeimidis, I. Dima, and C. Stefanadis Chronic coffee consumption has a detrimental effect on aortic stiffness and wave reflections Am. J. Clinical Nutrition, June 1, 2005; 81(6): 1307 - 1312. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. E. Robinson, S. Savani, D. S. Battram, D. H. McLaren, P. Sathasivam, and T. E. Graham Caffeine Ingestion Before an Oral Glucose Tolerance Test Impairs Blood Glucose Management in Men with Type 2 Diabetes J. Nutr., October 1, 2004; 134(10): 2528 - 2533. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. Happonen, S. Voutilainen, and J. T. Salonen Coffee Drinking Is Dose-Dependently Related to the Risk of Acute Coronary Events in Middle-Aged Men J. Nutr., September 1, 2004; 134(9): 2381 - 2386. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
