Appendix 1.
Cohort studies addressing the association between coffee consumption and total and cause-specific mortality.*
| Author | Sex, age (years) | N | Exposure | Outcome | No. of cases | Mean follow-up (years) | Adjustment for confounding | Multivariate-adjusted result | Comments |
|---|---|---|---|---|---|---|---|---|---|
| Iwai et al. (1) | M/F, 40-79 | 2855 | Coffee at baseline | Death from all causes | 361 | 9.9 | Age, physical activity, education, history of chronic diseases.SMK, alcohol (only in M). | For consumption of ≥2 cups/d, the RR was 0.43 (95% CI 0.30-0.63) for M and 0.76 (0.45-1.27) for F. | Inverse association between coffee and total mortality in M. |
|
| |||||||||
| Kleemola et al. (2) | M/F, 30-59 | 20179 | Coffee at baseline | Death from all causes Death from CHD |
1645 975 |
10 | Age, SMK, cholesterol, blood pressure, history of MI | M: for consumption of ≥7 cups/d, the RR of total death was 1.01 (0.84-1.22) and the RR of CHD death was 1.22 (0.90-1.65). F: for consumption of ≥7 cups/d, the RR of total death was 0.62 (0.44-0.87) and the RR of CHD death was 0.57 (0.28-1.16). |
Inverse association between coffee and total mortality in F.No association between coffee and CHD mortality in M/F. |
|
| |||||||||
| Woodward et al. (3) | M/F, 40-59 | 11000 | Coffee at baseline | Death from all causes Death from CHD |
573 357 |
7.7 | Age, housing tenure, activity at work, activity in leisure, SMK, BMI, Bortner score, cotinine, SBP, fibrinogen, total cholesterol, HDL, TG, alcohol, vit.C, tea. | M: for consumption of ≥5 cups/d, the RR of total death was 0.77 and the RR of CHD death was 0.58. F: for consumption of ≥5 cups/d, the RR of total death was 0.79 and the RR of CHD death was 1.18. |
Inverse association between coffee and total mortality in M/F. Inverse association between coffee and CHD mortality in M. |
|
| |||||||||
| Rosengren et al. (4) | M, 51-59 | 6765 | Coffee at baseline | Death from all causes Death from CHD Death from cancer |
478 169 Not available |
7.1 | Age, SBP, BMI, diabetes, alcohol abuse, family history of MI, mental stress, physical activity, occupational class. | For consumption of ≥9 cups/d, the RR of total death was 0.6 (0.3-0.9), the RR of CHD death was 1.4 (0.8-2.4), and the RR of cancer death was 0.9 (0.4-2.1). | Inverse association between coffee and total mortality. No association between coffee and CHD and cancer mortality. |
|
| |||||||||
| Murray et al. (5) | M, ≥35 | 16911 | Coffee at baseline | Death from all causes other than CHD Death from CHD |
985 721 |
11.5 | Age, SMK, urban/rural residence | For consumption of ≥7 cups/d, the RR of total death was 0.86, and the RR of CHD death was 0.91. | Inverse association between coffee and total and CHD mortality. |
|
| |||||||||
| Andersen et a. (6) | F, 55-69 | 41836 | Coffee at baseline | Death from all causes Death from CVD Death from cancer Death from inflammatory diseases |
4265 1411 1733 713 |
15 | Age, SMK, alcohol, BMI, waist-hip ratio, education, physical activity, use of estrogens, MV, energy intake, whole and refined grain, red meat, fish and seafood, total fruit and vegetables. | For consumption of 4-5 cups/d, the RR of total death was 0.81 (0.72-0.91), the RR of CVD death was 0.81 (0.66-0.99), the RR of cancer death was 0.84 (0.70-1.02), and the RR of inflammatory death was 0.67 (0.50-0.90). | U-shape association between coffee and total and CVD mortality; no association with cancer mortality; inverse association with inflammatory mortality. |
|
| |||||||||
| Lopez-Garcia et al. (8) | M/F, 30-55 | 128493 | Cumulative exposure | Death from CHD | 1417 | 14 (M), 20 (F) | Age, SMK, BMI, physical activity, alcohol, parental history of MI, menopausal status, use of hormone therapy, aspirin, MV, vit. E, hypertension, hypercholesterolemia, diabetes mellitus. | M: For consumption of ≥6 cups/d, the RR of CHD death was 0.60 (0.26-1.36). F: For consumption of ≥6 cups/d, the RR of CHD death was 0.61 (0.37-1.02). |
Weak inverse association between coffee and CHD mortality. |
|
| |||||||||
| Legrady et al. (25) | M, 40-56 | 1910 | Coffee at 1 year baseline | Death from all causes Death from CHD Death from all causes other than CHD |
452 220 232 |
19 | Age, DBP, cholesterol, SMK. | For consumption of ≥6 cups/d, the RR of total death was 1.33 (1.07-1.65), the RR of CHD death was 1.71 (1.27-2.30), and the RR of non-CHD death was 1.02 (0.73-1.41). | Direct association between coffee and total and CHD mortality. |
|
| |||||||||
| Tverdal et al. (26) | M/F, 35-54 | 38564 | Coffee at baseline | Death from CHD | 174 | 6.4 | Age, cholesterol, HDL, SBP, SMK | M: For consumption of ≥9 cups/d, the RR of CHD death was 2.2 (1.1-4.5). F: For consumption of ≥9 cups/d, the RR of CHD death was 5.1 (0.4-60.3). |
Direct association between coffee and CHD mortality. |
|
| |||||||||
| Stensvold et al. (27) | M/F, 35-54 | 38564 | Coffee at baseline | Death from CHD | 476 | 12 | Age, cholesterol, HDL, SBP, SMK | For consumption of ≥9 cups/d, the RR of CHD death was 1.3. | No association between coffee and CHD mortality when Tverdal et al. study (25) was continued for 6 more years. |
|
| |||||||||
| Greenberg et al. (40) | M/F, 32-86 | 6594 | Caffeinated beverages at baseline | Death from CVD Death from CHD Death from cerebrovascular disease |
426 347 79 |
8.8 | Age, SMK, BMI, sex, race, physical activity, alcohol, income, education, American-style diet. | In ≥65 y old, for consumption of ≥4 cups/d the RR of CVD death was 0.53 (0.38-0.75), the RR of CHD death was 0.47 (0.32-0.69), and the RR of cerebrovascular death was 0.88 (0.42-1.83). | Inverse association between caffeinated beverages and CVD and CHD mortality among elderly participants but not among younger participants. |
|
| |||||||||
| Hart et al. (41) | M, 35-64 | 5766 | Coffee at baseline | Death from CHD | 625 | 17 | Age, DBP, cholesterol, SMK, social class, education, BMI, angina. | For consumption of ≥4.5 cups/d, the RR of CHD death was 1.49 (0.89-2.47). | No association between coffee and CHD mortality. |
|
| |||||||||
| Jazbec et al. (42) | M/F, 35-59 | 3364 | Coffee at baseline | Death from all causes Death from CVD |
950 435 |
27 | Age, region, SMK, DBP, feeling of well-being, history of stomach ulcer | M: for consumption of 1-2 cups/d, the RR of total death was 0.78 (0.61-0.98), and the RR of CVD was 0.82 (0.58-1.16). F: for consumption of 1-2 cups/d, the RR of total death was 0.63 (0.46-0.86), and the RR of CVD was 0.67 (0.43-1.05). |
Inverse association between coffee and total mortality. |
|
| |||||||||
| Lindsted et al. (43) | M, ≥30 | 9484 | Coffee at baseline | Death from all causes Death from CVD |
Not available | 25 | Age, BMI, heart disease, hypertension, race, exercise, sleep, marital status, education, SMK, diet history. | For consumption of ≥1-2 cups/d, the RR of total death was 1.15 (1.05-1.26), and the RR of CVD was 1.09 (0.82-1.46). | Weak direct association between coffee and total mortality and CVD mortality. |
|
| |||||||||
| Jacobsen et al. (44) | M/F, not available | 16555 | Coffee at baseline | Death from all causes Death from cancer |
4032 886 |
11.5 | Age, residence, SMK | For consumption of ≥7 cups/d, the RR of total death was 0.95 (p=0.15), and the RR of cancer was 1.14 (p=0.49). | No association between coffee and total and cancer death. |
|
| |||||||||
| Happonen et al. (45) | M, 42-60 | 1971 | Coffee at baseline | Death from CHD | 269 | 14 | Age. SMK, ischemia, diabetes, income, serum insulin, DBP, HDL, LDL, maximal oxygen uptake, WHR. | Heavy drinkers (≥814 mL/d) had a RR of CHD death of 1.43 (1.06-1.94). | Direct association between coffee and CHD mortality. |
|
| |||||||||
| Paganini-Hill et al. (46) | M/F, 44-101 | 13624 | Coffee at baseline | Death from all causes | 11386 | 23 | Age, sex, SMK, physical activity BMI, alcohol, hypertension, angina, heart attack, stroke, diabetes, rheumatoid arthritis, cancer. | For consumption of 2-3 cups/d, the RR of total death was 0.89 (0.85-0.94). | U-shape association between coffee and total and mortality. |
|
| |||||||||
| Dawber et al. (47) | M/F, 30-62 | 5209 | Coffee at baseline | Death from all causes Death from CHD |
321 Not available |
12 | Age, SMK | M: for consumption of ≥6 cups/d, the RR of total death was 1.01 and the RR of CHD death was 0.92. | Direct association between coffee and total mortality in M. No association between coffee and CHD mortality in M/F. |
|
| |||||||||
| Vandenbroucke et al. (48) | M/F, 40-65 | 3091 | Coffee at baseline | Death from all causes | Not available | 25 | Age, SMK, alcohol, BMI, living parents, cholesterol SBP. | M: for consumption of ≥5 cups/d, the RR of total death was 1.42 (0.94-2.15). F: for consumption of ≥5 cups/d, the RR of total death was 0.83 (0.52-1.30). |
Direct association between coffee and total mortality in M. |
|
| |||||||||
| Kurozawa et al. (49) | M/F, 40-79 | 110688 | Coffee at baseline | Death from hepatocellular carcinoma | 258 | 11 | Age, gender, education, history of diabetes and liver diseases, SMK, alcohol. | For consumption of ≥1 cups/d, the RR of cancer death was 0.50 (0.31-0.79). | Inverse association between coffee and risk of death from hepatocellular carcinoma. |
*
Obtained from a MEDLINE search through February 2008 using the key words coffee and caffeine in combination with mortality and death.
M: men; F: female; SMK: smoking; CHD: coronary heart disease; MI: myocardial infarction; BMI: body mass index; SBP: systolic blood pressure; HDL: high density lipoprotein; TG: tryglicerides; vit.C: vitamin C; MV: multivitamins; vit.E: vitamin E; DBP: diastolic blood pressure; LDL: low density lipoprotein; WHR: waist-hip ratio.