This is not the document you are looking for? Use the search form below to find more!

Report home > Science

Association of Coffee Drinking with Total and Cause-Specific Mortality

0.00 (0 votes)
Document Description
Association of Coffee Drinking with Total and Cause-Specific Mortality
File Details
  • Added: May, 17th 2012
  • Reads: 538
  • Downloads: 1
  • File size: 647.48kb
  • Pages: 14
  • Tags: cofee, mortality, lower
  • content preview
Submitter
  • Name: neurociencianews
Embed Code:

Add New Comment




Related Documents

Association of Coffee Drinking with Total and Cause-Specific Mortality

by: neurocienciasnews, 14 pages

Association of Coffee Drinking with Total and Cause-Specific Mortality

Kenneth Smith The Association Of Latino Professionals In Finance And Accounting

by: tomaj, 3 pages

KENNETH SMITH 1130 North Franklin Street Pittsburgh, PA 15233 412-321-6163 Home ...

The association of placenta previa with history of cesarean delivery and abortion: A metaanalysis

by: shinta, 13 pages

Objective: Our purpose was to determine the incidence of placenta previa based on the available epidemiologic evidence and to quantify the risk of placenta previa based on the presence ...

Grape (Vitis vinifera L.) BAC Library Construction, Preliminary STC Analysis, and Identification of Clones Associated with Flavonoid and Stilbene Biosynthesis

by: shinta, 5 pages

We have constructed a grape BAC library using the cultivar Syrah. The library contains 55,296 clones stored in 144 384-well microtiter plates. A random sampling of 381 BACs indicated an average ...

[Kamal Koirala] Glossary; Compilation of Terminologies related with Conflict and Peace Studies

by: Kamal Koirala, 269 pages

What I found……………….. As Conflict, Peace & Development Studies is a new multidisciplinary subject in the academic field of South Asia (especially in the post ...

[Kamal Koirala] Glossary; Compilation of Terminologies related with Conflict and Peace Studies

by: Kamal Koirala, 269 pages

What I found……………….. As Conflict, Peace & Development Studies is a new multidisciplinary subject in the academic field of South Asia (especially in the post ...

Effects of Coffee Consumption against the Development of Liver Dysfunction: A 4-Year Follow-Up Study of Middle-Aged Japanese Male Office Workers

by: shinta, 4 pages

The association of coffee consumption with the development of increased serum aspartate aminotransferase (AST) and/or alanine aminotransferase (ALT) activities over 4 years was studied in ...

Glossary, Compilation of terminology related with conflict and peace studies

by: kamal, 269 pages

What I found……………….. As Conflict, Peace & Development Studies is a new multidisciplinary subject in the academic field of South Asia (especially in the post ...

Coffee Mug With Photo

by: printlandteam, 9 pages

Printed coffee mugs are not only used for personal purposes, but some companies also use them for their promotional activities. Today, the corporate firms like to advertise their name with logo ...

Operations with Integers and Rational Numbers

by: nishagoyal, 3 pages

If we look at any integer say -7, we find that -7 can be written as -7 /1, where -7 and 1 both are integers and the denominator 1 < > 0. So -7 is absolutely expressed in form of a rational number. ...

Content Preview
T h e n e w e ng l a n d j o u r na l o f m e dic i n e
original article
Association of Coffee Drinking with Total
and Cause-Specific Mortality
Neal D. Freedman, Ph.D., Yikyung Park, Sc.D., Christian C. Abnet, Ph.D.,
Albert R. Hollenbeck, Ph.D., and Rashmi Sinha, Ph.D.
Abs tr act
Background
Coffee is one of the most widely consumed beverages, but the association between From the Division of Cancer Epidemiology
and Genetics, National Cancer Institute,
coffee consumption and the risk of death remains unclear.
National Institutes of Health, Department
of Health and Human Services, Rockville,
Methods
MD (N.D.F., Y.P., C.C.A., R.S.); and AARP,
Washington, DC (A.R.H.). Address reprint
We examined the association of coffee drinking with subsequent total and cause- requests to Dr. Freedman at the Nutri-
specific mortality among 229,119 men and 173,141 women in the National Institutes tional Epidemiology Branch, Division of
of Health-AARP Diet and Health Study who were 50 to 71 years of age at baseline. Cancer Epidemiology and Genetics, 6120
Executive Blvd., EPS/320, MSC 7232,
Participants with cancer, heart disease, and stroke were excluded. Coffee consumption Rockville, MD 20852, or at freedmanne@
was assessed once at baseline.
mail.nih.gov.
N Engl J Med 2012;366:1891-904.
Results
Copyright (c) 2012 Massachusetts Medical Society.
During 5,148,760 person-years of follow-up between 1995 and 2008, a total of
33,731 men and 18,784 women died. In age-adjusted models, the risk of death was
increased among coffee drinkers. However, coffee drinkers were also more likely to
smoke, and, after adjustment for tobacco-smoking status and other potential con-
founders, there was a significant inverse association between coffee consumption
and mortality. Adjusted hazard ratios for death among men who drank coffee as
compared with those who did not were as follows: 0.99 (95% confidence interval
[CI], 0.95 to 1.04) for drinking less than 1 cup per day, 0.94 (95% CI, 0.90 to 0.99) for
1 cup, 0.90 (95% CI, 0.86 to 0.93) for 2 or 3 cups, 0.88 (95% CI, 0.84 to 0.93) for 4 or
5 cups, and 0.90 (95% CI, 0.85 to 0.96) for 6 or more cups of coffee per day (P<0.001
for trend); the respective hazard ratios among women were 1.01 (95% CI, 0.96 to
1.07), 0.95 (95% CI, 0.90 to 1.01), 0.87 (95% CI, 0.83 to 0.92), 0.84 (95% CI, 0.79 to
0.90), and 0.85 (95% CI, 0.78 to 0.93) (P<0.001 for trend). Inverse associations were
observed for deaths due to heart disease, respiratory disease, stroke, injuries and ac-
cidents, diabetes, and infections, but not for deaths due to cancer. Results were
similar in subgroups, including persons who had never smoked and persons who re-
ported very good to excellent health at baseline.
Conclusions
In this large prospective study, coffee consumption was inversely associated with
total and cause-specific mortality. Whether this was a causal or associational finding
cannot be determined from our data. (Funded by the Intramural Research Program
of the National Institutes of Health, National Cancer Institute, Division of Cancer
Epidemiology and Genetics.)
n engl j med 366;20 nejm.org may 17, 2012
1891
The New England Journal of Medicine
Downloaded from nejm.org at Hinari Phase 2 sites on May 17, 2012. For personal use only. No other uses without permission.
Copyright (c) 2012 Massachusetts Medical Society. All rights reserved.

T h e n e w e ng l a n d j o u r na l o f m e dic i n e
Coffee is one of the most widely con- ample power to detect even modest associations
sumed beverages, both in the United States and allowed for subgroup analyses according to
and worldwide. Since coffee contains caf- important baseline factors, including the pres-
feine, a stimulant, coffee drinking is not generally ence or absence of adiposity and diabetes, as well
considered to be part of a healthy lifestyle. How- as cigarette-smoking status.
ever, coffee is a rich source of antioxidants1 and
other bioactive compounds, and studies have
Methods
shown inverse associations between coffee con-
sumption and serum biomarkers of inflamma- Study Population
tion2 and insulin resistance.3,4
The NIH-AARP Diet and Health Study has been
Considerable attention has been focused on described previously.32 Between 1995 and 1996, a
the possibility that coffee may increase the risk total of 617,119 AARP members, 50 to 71 years of
of heart disease,5,6 particularly since drinking age, returned a comprehensive questionnaire as-
coffee has been associated with increased low- sessing diet and lifestyle. Participants resided in
density lipoprotein cholesterol levels7 and short- six states (California, Florida, Louisiana, New Jer-
term increases in blood pressure.8 Results from sey, North Carolina, and Pennsylvania) and two
a number of studies have been inconsistent.9,10 metropolitan areas (Atlanta and Detroit). Of the
The heterogeneous findings may be due to dif- respondents, 566,401 completed the question-
ferences between case-control and prospective naire satisfactorily. Completion of the self-admin-
study designs and possibly also to inconsistent istered questionnaire was considered to imply in-
control for important confounders such as to- formed consent to participate in the study.
bacco smoking. In addition, the numbers of
We excluded from these analyses 15,760 per-
deaths have been small in most studies. Cohort sons whose questionnaires were completed by a
studies do not support a positive association spouse or other surrogate correspondent, as well
between coffee drinking and mortality, however, as 51,234 persons with cancer, 65,044 with heart
and some even suggest a modest inverse asso- disease, 10,459 who had had a previous stroke,
ciation.11-15
2082 who did not provide information on coffee
Previous studies have also investigated the as- use, 15,820 who did not provide information on
sociation between coffee consumption and other cigarette smoking, 3731 with an extremely low or
major causes of death, and they have shown in- high caloric consumption (two times as high as
verse associations with diabetes,4 inflammatory the 75th percentile of caloric intake or two times
diseases,11 stroke,16-18 and injuries and acci- as low as the 25th percentile of caloric intake), and
dents,19,20 although associations with cancer have 11 who died before their completed question-
generally been null.15,20-22 The results of studies naire was received. The resulting analytic cohort
of coffee consumption and total mortality have included 229,119 men and 173,141 women. The
been mixed,15,20-31 with associations that have NIH-AARP Diet and Health Study was approved
been consistent with either the null hypothesis or by the Special Studies Institutional Review Board
a modest inverse effect. Data are lacking to of the National Cancer Institute.
clarify the association between coffee drinking
and mortality, to determine whether there is a Assessment of Exposure
dose-response relationship, and to assess wheth- Participants completed a baseline questionnaire
er associations are consistent across various sub- that assessed demographic and lifestyle charac-
groups.
teristics and 124 dietary items, as previously de-
We used data from a very large study, the scribed.32 Consumption of fruits, vegetables, red
National Institutes of Health (NIH)-AARP Diet meat, white meat, and saturated fat were adjusted
and Health Study (ClinicalTrials.gov number, for total energy intake with the use of the nutrient-
NCT00340015), to determine whether coffee con- density approach (i.e., measured per 1000 kcal per
sumption is associated with total or cause-specific day for food groups and as a percentage of total
mortality.32 The current analysis, involving more energy for saturated fat).
than 400,000 participants and 52,000 deaths, had
Coffee consumption was assessed according to
1892
n engl j med 366;20 nejm.org may 17, 2012
The New England Journal of Medicine
Downloaded from nejm.org at Hinari Phase 2 sites on May 17, 2012. For personal use only. No other uses without permission.
Copyright (c) 2012 Massachusetts Medical Society. All rights reserved.

Coffee and Mortality
10 frequency categories, ranging from 0 to 6 or (ICD-9, 480-487 and 490-496; ICD-10, J10-J18
more cups per day. In addition, 96.5% of coffee and J40-J47), stroke (ICD-9, 430-438; ICD-10,
drinkers provided information on whether they I60-I69), injuries and accidents (e.g., accident,
drank caffeinated or decaffeinated coffee more suicide, and homicide) (ICD-9, 800-978; ICD-10,
than half the time, and we used this information V01-X59, Y85-Y86, U03, X60-X84, Y87.0, U01-U02,
to categorize coffee drinkers.
X85-Y09, Y35, Y87.1, and Y89.0), diabetes (ICD-9,
In a subgroup of 1953 study participants who 250; ICD-10, E10-E14), infections (e.g., tuberculo-
also completed a 24-hour dietary-recall question- sis, septicemia, and other infectious and parasitic
naire on 2 nonconsecutive days,33 the Spearman diseases) (ICD-9, 001-139; ICD-10, A00-B99), and
coefficient for the correlation between coffee con- all other causes.
sumption as assessed with this questionnaire and
State data on the incidence of cancer were ob-
coffee consumption as assessed with the baseline tained from the Arizona Cancer Registry, the
food-frequency questionnaire was 0.80. The re- Georgia Center for Cancer Statistics, the Califor-
spective Spearman correlation coefficients for caf- nia Cancer Registry, the Michigan Cancer Surveil-
feinated and decaffeinated coffee were 0.64 and lance Program, the Florida Cancer Data System,
0.48, respectively. Among participants who com- the Louisiana Tumor Registry, the New Jersey
pleted the 24-hour dietary-recall questionnaire, State Cancer Registry, the North Carolina Central
79% drank ground coffee, 19% drank instant cof- Cancer Registry, the Pennsylvania Cancer Regis-
fee, 1% drank espresso coffee, and 1% did not try, and the Texas Cancer Registry.
specify the type of coffee they consumed.
Statistical Analysis
Cohort Follow-up
Coffee consumption was tabulated according to a
Participants were followed from baseline (1995- number of dietary and lifestyle factors. Hazard ra-
1996) until the date of death or December 31, tios and 95% confidence intervals for mortality
2008, whichever came first, by means of linkage to associated with coffee consumption were estimat-
the National Change of Address database main- ed with the use of Cox proportional-hazards re-
tained by the U.S. Postal Service, specific change- gression models, with person-years as the under-
of-address requests from participants, and up- lying time metric; results calculated with age as
dated addresses returned during other mailings. the underlying time metric were similar. We tested
Vital status was assessed by periodic linkage of the proportional-hazards assumption by model-
the cohort to the Social Security Administration ing the interaction of follow-up time with coffee
Death Master File, linkage with cancer registries, consumption and observed no significant devia-
questionnaire responses, and responses to other tions. Analyses were conducted with the use of
mailings.
SAS software, version 9.1. Statistical tests were
two-sided, and P values of less than 0.05 were con-
Causes of Death
sidered to indicate statistical significance.
Specific causes of death were obtained through
We present risk estimates separately for men
follow-up linkage to the National Death Index Plus, and women. Multivariate models were adjusted for
maintained by the National Center for Health the following baseline factors: age; body-mass
Statistics. We used the International Classification of index (BMI); race or ethnic group; level of educa-
Diseases, Ninth Revision (ICD-9), and International tion; alcohol consumption; the number of ciga-
Classification of Diseases, 10th Revision (ICD-10) rettes smoked per day, use or nonuse of pipes or
codes to classify the underlying cause of death cigars, and time of smoking cessation (<1 year,
(obtained from death certificates) as follows: 1 to <5 years, 5 to <10 years, or 10 years before
cancer (ICD-9, 140-239; ICD-10, C00-C97 and baseline); health status; presence or absence of
D00-D48), heart disease (ICD-9, 390-398, 401- diabetes; marital status; level of physical activity;
404, 410-429, and 440-448; ICD-10, I00-I13, total energy intake; consumption of fruits, vegeta-
I20-I51, and I70-I78), respiratory disease (e.g., bles, red meat, white meat, and saturated fat; and
pneumonia, influenza, chronic obstructive pul- use of any vitamin supplement (yes vs. no). In ad-
monary disease, and associated conditions) dition, risk estimates for death from cancer were
n engl j med 366;20 nejm.org may 17, 2012
1893
The New England Journal of Medicine
Downloaded from nejm.org at Hinari Phase 2 sites on May 17, 2012. For personal use only. No other uses without permission.
Copyright (c) 2012 Massachusetts Medical Society. All rights reserved.

T h e n e w e ng l a n d j o u r na l o f m e dic i n e
adjusted for history of cancer (other than non- women. Several differences (P<0.05) would be
melanoma skin cancer) in a first-degree relative expected by chance alone.
(yes vs. no). For women, status with respect to
postmenopausal hormone therapy was also in-
R esults
cluded in multivariate models. Less than 5% of
the cohort lacked any single covariate; for each Association of Coffee Consumption
covariate, we included an indicator for missing with Dietary and Lifestyle Factors
data in the regression models, if necessary. In Coffee consumption at baseline was associated
a sensitivity analysis, we adjusted for propen- with several other dietary and lifestyle factors
sity scores34 that reflected associations of cof- (Table 1). As compared with persons who did not
fee consumption with the other variables in the drink coffee, coffee drinkers were more likely to
multivariate-adjusted models. Results obtained smoke cigarettes and consume more than three
with the use of propensity-score adjustment were alcoholic drinks per day, and they consumed more
very similar to those from multivariate-adjusted red meat. Coffee drinkers also tended to have a
models (Table 1 in the Supplementary Appen- lower level of education; were less likely to engage
dix, available with the full text of this article at in vigorous physical activity; and reported lower
NEJM.org).
levels of consumption of fruits, vegetables, and
Hazard ratios for death associated with cate- white meat. However, coffee drinkers, especially
gories of coffee consumption (<1, 1, 2 or 3, 4 or women who drank coffee, were less likely to re-
5, and 6 cups per day), as compared with no cof- port having diabetes. About two thirds of coffee
fee consumption, were estimated from a single drinkers reported drinking predominantly caf-
model. Tests of linear trend across categories of feinated coffee.
coffee consumption were performed by assigning
participants the midpoint of their coffee-consump- Coffee Consumption and Total Mortality
tion category and entering this new variable into During 14 years of follow-up (median, 13.6 years;
a separate Cox proportional-hazards regression total person-years, 5,148,760), 33,731 men and
model.
18,784 women died. In age-adjusted analyses, cof-
In secondary analyses, we determined risk es- fee consumption was associated with increased
timates for categories of consumption of caffein- mortality among both men (Table 2) and women
ated and decaffeinated coffee and examined asso- (Table 3). However, after multivariate adjustment
ciations among prespecified baseline subgroups for potential confounders, particularly smoking
based on the following: follow-up time; age; (Table 1 in the Supplementary Appendix), a mod-
cigarette-smoking status; presence or absence of est inverse association between coffee drinking
diabetes; BMI; alcohol consumption; self-reported and total mortality was observed for both sexes.
health; high or low consumption of red meat, Hazard ratios for death among men who drank
white meat, fruits, and vegetables; use or nonuse coffee, as compared with men who did not drink
of any vitamin supplement; and, in women, use or coffee, were as follows: 0.99 (95% confidence inter-
nonuse of postmenopausal hormone therapy. For val [CI], 0.95 to 1.04) for less than 1 cup of coffee
these analyses, we combined the categories of per day, 0.94 (95% CI, 0.90 to 0.99) for 1 cup, 0.90
4 or 5 cups of coffee per day and 6 or more cups (95% CI, 0.86 to 0.93) for 2 or 3 cups, 0.88 (95% CI,
per day to preserve numbers in the top category 0.84 to 0.93) for 4 or 5 cups, and 0.90 (95% CI,
of consumption. P values for interactions were 0.85 to 0.96) for 6 or more cups (P<0.001 for trend
computed by means of likelihood-ratio tests com- across categories) (Table 2). Hazard ratios among
paring Cox proportional-hazards models with and women who drank coffee, as compared with those
those without cross-product terms for each level who did not, were as follows: 1.01 (95% CI, 0.96
of the baseline stratifying variable, with coffee to 1.07) for less than 1 cup of coffee per day, 0.95
consumption as an ordinal variable. For total (95% CI, 0.90 to 1.01) for 1 cup, 0.87 (95% CI,
mortality, we performed 12 interaction tests for 0.83 to 0.92) for 2 or 3 cups, 0.84 (95% CI, 0.79
men and 13 interaction tests for women. We also to 0.90) for 4 or 5 cups, and 0.85 (95% CI, 0.78 to
performed interaction tests for smoking status 0.93) for 6 or more cups (P<0.001 for trend across
with eight different outcomes for both men and categories) (Table 3).
1894
n engl j med 366;20 nejm.org may 17, 2012
The New England Journal of Medicine
Downloaded from nejm.org at Hinari Phase 2 sites on May 17, 2012. For personal use only. No other uses without permission.
Copyright (c) 2012 Massachusetts Medical Society. All rights reserved.

Coffee and Mortality

Cups
4.4
3.0
1.2
2.0

=

5,152)
61.0
97.2
52.6
37.8
24.3
25.0
48.1
15.4
10.0
1642
29.6
23.9
62.4
45.1
6 (N
56.4-65.3

=

0.33).

(P
men
Cups
5
in
61.6
96.5
52.7
43.5
28.7
25.1
4.5
31.1
3.3
16.0
8.2
1.4
2.2
1540
28.4
27.3
66.1
51.2
or
4 (N=17,434)
57.0-65.8

status

=

173,141)
Cups
5.2
3.1
8.2
1.6
2.2
health
(N
3
62.2
93.9
52.2
46.1
31.0
25.5
15.9
16.1
1463
27.1
29.5
67.6
55.3
or = 68,250)
2 (N
57.5-66.3
Women

self-reported
Cup
7.0
8.8
2.3
1.8
2.3
1 = 31,355)
62.9
87.2
51.1
46.6
31.3
25.8
16.6
10.2
1432
25.6
30.1
67.0
54.8
and
(N
58.1-66.7


=

0.002)
Coffee
7.4
8.1
1.8
1.8
2.3
(P

=

20,865)
60.9
89.8
51.3
45.4
33.9
25.9
18.2
12.2
1444
25.6
28.3
68.4
52.4
No (N
56.1-65.6
men
in

Cups
60.3
97.3
48.5
83.4
37.2
26.6
8.0
34.7
11.6
20.6
9.3
0.8
1.7
diabetes
2175
40.5
21.5
55.5
6 = 10,139)
for
(N
55.9-64.8

except
Cups
5
61.0
96.9
48.6
85.2
44.9
26.7
8.1
20.3
13.1
19.7
7.0
1.1
1.8
2007
38.3
24.4
57.6
or = 32,084)
4 (N
56.6-65.5
meters.
Consumption.*

categories,
in
Cups
Coffee

=

229,119)
3
62.5
95.1
48.2
86.4
46.6
26.6
8.0
10.9
12.7
20.5
6.6
1.2
1.9
1882
36.5
26.0
58.1
across
category. height
(N
or = 97,144)
57.7-66.5
the
Daily
2 (N
this of
to
Men

trends
for
from
Cup
8.6
6.5
7.5
1.4
1.9
square
1 = 33,961)
63.5
91.1
46.3
85.5
46.7
26.5
10.4
21.3
1830
34.6
26.1
58.2
According
(N
58.6-67.2
the

P<0.001
excluded by
with
was
Coffee
8.4
4.8
6.3
7.5
1.5
1.9

=

21,080)
61.1
91.1
46.3
83.0
53.0
26.4
24.8
1869
33.1
25.8
59.6
divided
Participants,
No (N
56.4-65.7
relative
drinking,
Study
(%)
kilograms
the
coffee
in
of
(%)
(%)
first-degree
with
a
times/wk
in weight
5
(kcal/day)
postmenopausal
self-reported.
(%)
(%)
health
of (%)
the
(%)
was
is
index
intake
food
supplement
use
associated
cancer
Characteristics
(%)
activity
range
cancer
(%)
(%)
of
therapy
white
of
(servings/day)
(g/day)
group skin index
drinks/day
energy
(g/day)
were
self-reported
current
married
vitamin
(%)
physical
(servings/day)
meat
or
history
graduate
body-mass
smoker
fair
total
servings
meat
any
hormone
ethnic
(yr)
or
or
Median
Interquartile
Alcoholic
Fruits
Vegetables
Red
White
of
exposures
body-mass
T
a
b
l
e

1
.

Baseline

Characteristic
Age
Non-Hispanic
Family
Currently
College
Median
Diabetes
Current
>3
Vigorous
Poor
Median
Median
Use
Previous

Race Nonmelanoma
*

All The
n engl j med 366;20 nejm.org may 17, 2012
1895
The New England Journal of Medicine
Downloaded from nejm.org at Hinari Phase 2 sites on May 17, 2012. For personal use only. No other uses without permission.
Copyright (c) 2012 Massachusetts Medical Society. All rights reserved.

T h e n e w e ng l a n d j o u r na l o f m e dic i n e

Value Trend
P
0.02
0.03
0.004
0.57
0.003
0.93
0.02
0.001
for
<0.001
<0.001
<0.001
<0.001
<0.001
<0.001
Cups
(18.8)
(8.7)
(4.4)
(1.8)
(0.6)
(0.5)
(0.3)

6 = 10,139)
(1.51-1.69)
(0.85-0.96)
(1.95-2.34)
(0.98-1.19)
(1.28-1.62)
(0.78-1.00)
(2.13-3.24)
(0.65-1.00)
(0.89-1.64)
(0.61-1.14)
(0.72-1.40)
(0.51-1.02)
(0.46-1.10)
(0.39-0.94)
880
442
185
63
50
27
(N
1904
1.60
0.90
2.13
1.08
1.44
0.88
2.63
0.81
1.21
0.83
1.00
0.72
0.71
0.60

Cups
(6.9)
(3.7)
5
(15.5)
(1.3)
(0.4)
(0.5)
(0.3)
or = 32,084)
(1.15-1.27)
(0.84-0.93)
(1.46-1.70)
(0.96-1.12)
(1.02-1.23)
(0.79-0.96)
(1.37-1.96)
(0.69-1.00)
(0.60-0.97)
(0.51-0.84)
(0.78-1.26)
(0.68-1.12)
(0.62-1.10)
(0.60-1.08)
4
409
141
168
107
(N
4966
2219
1184
1.21
0.88
1.57
1.04
1.12
0.87
1.64
0.83
0.76
0.65
1.00
0.87
0.83
0.80

Cups
(14.5)
(6.0)
(3.5)
(1.1)
3
(0.6)
(0.5)
(0.3)
or = 97,144)
(0.99-1.07)
(0.86-0.93)
(1.17-1.34)
(0.93-1.07)
(0.87-1.03)
(0.79-0.94)
(1.03-1.43)
(0.70-0.98)
(0.72-1.06)
(0.68-1.02)
(0.73-1.10)
(0.71-1.09)
(0.57-0.92)
(0.59-0.96)
2
555
492
310
(N
5804
3353
1046
14,115
1.03
0.90
1.25
1.00
0.95
0.86
1.21
0.83
0.87
0.84
0.90
0.88
0.73
0.75

(5.4)
(3.7)
Men.*
Cup
(14.9)
(1.0)
(0.7)
(0.6)
(0.5)
1 = 33,961)
(0.94-1.03)
(0.90-0.99)
(0.98-1.15)
(0.89-1.04)
(0.86-1.03)
(0.84-1.01)
(0.89-1.28)
(0.77-1.11)
(0.73-1.14)
(0.73-1.15)
(0.80-1.27)
(0.80-1.29)
(0.75-1.27)
(0.76-1.31)
352
222
202
154
(N
5049
1824
1243
0.99
0.94
1.06
0.96
0.94
0.92
1.07
0.93
0.91
0.92
1.01
1.02
0.98
1.00
229,119

among
Cup
(14.2)
(5.0)
(3.4)
(1.0)
(0.6)
(0.5)
(0.5)
<1 = 34,710)
(0.98-1.07)
(0.95-1.04)
(0.98-1.14)
(0.93-1.09)
(0.87-1.05)
(0.85-1.02)
(0.97-1.40)
(0.87-1.27)
(0.80-1.24)
(0.79-1.24)
(0.76-1.21)
(0.77-1.24)
(0.85-1.42)
(0.82-1.39)
351
221
186
165
(N
4931
1729
1193
Mortality
1.02
0.99
1.06
1.01
0.96
0.93
1.17
1.05
0.99
0.99
0.96
0.98
1.10
1.07

Coffee
(13.1)
(4.5)
(3.4)
(0.8)
(0.6)
(0.5)
(0.4)

=

21,080)

1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
No
946
712
169
125
113
87
Cause-Specific
(N
2766
and

Total
All
850
33,731
13,402
8,127
2,512
1,327
1,211
with
Participants
CI)
CI)
CI)
CI)
CI)
CI)
CI)
(95%
(95%
(95%
(95%
(95%
(95%
(95%
Consumption
CI)
CI)
CI)
CI)
CI)
CI)
CI)
ratio
ratio
ratio
ratio
ratio
ratio
ratio
Coffee
(95%
(95%
(95%
(95%
(95%
(95%
(95%
ratio
hazard
ratio
hazard
ratio
hazard
ratio
hazard
ratio
hazard
ratio
hazard
ratio
hazard
Daily
of

(%)
hazard
(%)
hazard
(%)
hazard
(%)
hazard
(%)
hazard
(%)
hazard
(%)
hazard
disease
accidents
Death
deaths
deaths
deaths
deaths
deaths
deaths
deaths
of
of
of
of
of
of
and
of
of
disease
causes
No.
Age-adjusted
Multivariate-adjusted
No.
Age-adjusted
Multivariate-adjusted
No.
Age-adjusted
Multivariate-adjusted
No.
Age-adjusted
Multivariate-adjusted
No.
Age-adjusted
Multivariate-adjusted
No.
Age-adjusted
Multivariate-adjusted
No.
Age-adjusted
Multivariate-adjusted
T
a
b
l
e

2
.

Association

Cause
All
Cancer
Heart
Respiratory
Stroke
Injuries
Diabetes
1896
n engl j med 366;20 nejm.org may 17, 2012
The New England Journal of Medicine
Downloaded from nejm.org at Hinari Phase 2 sites on May 17, 2012. For personal use only. No other uses without permission.
Copyright (c) 2012 Massachusetts Medical Society. All rights reserved.

Coffee and Mortality
-
Coffee Consumption and Cause-Specific
Mortality

cessa fruits,
0.08
0.001
0.002
<0.001
age;
of
Specific causes of death were also examined. Af-
cancer
of
ter multivariate adjustment, coffee appeared to
smoking
be inversely associated with most major causes of
baseline: of
history
death in both men and women, including heart
at
(0.2)
(2.3)
time consumption for
disease, respiratory disease, stroke, injuries and
25
(0.54-1.35)
(0.37-0.95)
(0.86-1.16)
(0.61-0.84)
232
and
accidents, diabetes, and infections. In contrast,
factors
0.85
0.59
1.00
0.72
intake;
there was no significant association between
adjusted
cigars,
coffee consumption and deaths from cancer in
following or energy were
women. There was a borderline positive associa-
(0.2)
(2.1)
the
tion in men: among 13,402 deaths from cancer,
pipes total
80
(0.57-1.09)
(0.50-0.98)
(0.73-0.91)
(0.63-0.80)
658
for
cancer
880 deaths were reported among men who drank
of
0.79
0.70
0.81
0.71
6 or more cups of coffee per day (hazard ratio for
activity; from
the comparison with men who did not drink cof-
adjusted nonuse
fee, 1.08; 95% CI, 0.98 to 1.19; P = 0.02 for trend).
or
death
(0.3)
(2.3)
were
physical
use
for
(0.63-1.07)
(0.63-1.10)
(0.76-0.92)
(0.74-0.89)
Subgroup Analyses
276
2279
day,
0.82
0.83
0.84
0.81
status;
In analyses stratified according to the predomi-
analyses per
estimates
nant type of coffee consumed (caffeinated or de-
marital
caffeinated), the association of coffee drinking
risk
smoked
with total mortality and individual causes of death
(0.4)
(2.7)
no);
(0.74-1.34)
(0.76-1.40)
(0.81-1.01)
(0.82-1.02)
Multivariate
appeared to be similar for the two types of coffee
124
928
vs. addition,
(Fig. 1, and Tables 2 and 3 in the Supplementary
0.99
1.03
0.91
0.92
cigarettes (yes In
Appendix).
follow-up. of
Coffee consumption was associated with a
number of risk factors for death. Results of sub-
(0.3)
(2.8)
diabetes
(0.70-1.28)
(0.70-1.29)
(0.92-1.13)
(0.92-1.13)
during number
group analyses are shown in Figure 2, with more
112
974
supplements.
the
detailed results shown in Tables 4 and 5 in the
0.95
0.95
1.02
1.02
died
status;
Supplementary Appendix. Associations between
who
health vitamin
coffee consumption and mortality were generally
of no).
(0.3)
(2.6)
similar across subgroups stratified according to
1.00
1.00
1.00
1.00
vs.
68
546
consumption;
duration of follow-up and the following baseline
(yes
participants
baseline); nonuse
factors: age; diabetes (yes vs. no); BMI; alcohol
or
for
consumption; high or low consumption of red
alcohol
use
are
before
relative
meat, white meat, fruits, and vegetables; use or
685
5,617
and
nonuse of any vitamin supplement; and, in women,
years fat;
use or nonuse of postmenopausal hormone ther-
deaths education;
of

10
of
apy. The largest differences across strata were
CI)
CI)
or
first-degree
a
observed for cigarette-smoking status, with stron-
level
in
(95%
(95%
saturated
ger inverse associations between coffee drinking
years,
CI)
CI)
numbers
and
and mortality among men and women who never
ratio
ratio
The group; <10
cancer)
smoked or were former smokers than among those
(95%
(95%
to
5 meat, skin
who were current smokers (P<0.001 for interaction
ratio
hazard
ratio
hazard
ethnic
interval.
in men and P = 0.002 for interaction in women),
or years, white
and for self-reported health at baseline, with
(%)
hazard
(%)
hazard
race <5 to
stronger associations among participants report-
meat,
1
ing good or very good to excellent health than
nonmelanoma
deaths
deaths
confidence index;
red
year,
among those reporting poor to fair health (P<0.001
of
of
causes
than
for interaction in both men and women).
No.
Age-adjusted
Multivariate-adjusted
No.
Age-adjusted
Multivariate-adjusted
(<1
denotes
We further examined associations between cof-
Infections
Other
body-mass tion vegetables, (other
*

CI
fee consumption and deaths from cancer and
n engl j med 366;20 nejm.org may 17, 2012
1897
The New England Journal of Medicine
Downloaded from nejm.org at Hinari Phase 2 sites on May 17, 2012. For personal use only. No other uses without permission.
Copyright (c) 2012 Massachusetts Medical Society. All rights reserved.

T h e n e w e ng l a n d j o u r na l o f m e dic i n e

Value Trend
0.66
0.67
0.82
0.05
0.004
0.03
P for
<0.001
<0.001
<0.001
<0.001
<0.001
<0.001
<0.001
<0.001

Cups
(15.0)
(7.3)
(2.5)
(2.0)
(0.6)
(0.2)

=

5152)

(0.2)
6
(1.39-1.64)
(0.78-0.93)
(1.77-2.27)
(0.90-1.16)
(0.95-1.41)
(0.59-0.88)
(1.85-2.99)
(0.61-0.99)
(0.79-1.74)
(0.56-1.25)
(0.42-1.54)
(0.29-1.10)
9
(0.27-1.07)
(0.28-1.16)
(N
775
375
127
104
32
11
1.51
0.85
2.00
1.03
1.16
0.72
2.35
0.77
1.18
0.84
0.81
0.57
0.53
0.57

Cups
(5.8)
5
(12.0)
(2.2)
(1.2)
(0.5)
(0.2)
(0.2)
or = 17,434)
(1.07-1.20)
(0.79-0.90)
(1.38-1.67)
(0.92-1.12)
(0.83-1.09)
(0.68-0.90)
(1.06-1.57)
(0.53-0.79)
91
(0.70-1.21)
(0.62-1.09)
36
(0.49-1.12)
(0.42-0.98)
38
(0.42-0.93)
(0.54-1.23)
4
378
208
(N
2099
1016
1.13
0.84
1.52
1.01
0.95
0.78
1.29
0.65
0.92
0.82
0.74
0.64
0.63
0.82

Cups
(4.6)
(2.0)
3
(10.5)
(1.0)
(0.5)
(0.2)
(0.2)
or = 68,250)
(0.89-0.98)
(0.83-0.92)
(1.05-1.23)
(0.90-1.06)
(0.75-0.93)
(0.76-0.95)
(0.88-1.22)
(0.67-0.93)
(0.72-1.10)
(0.75-1.15)
(0.57-1.05)
(0.56-1.06)
(0.42-0.75)
(0.57-1.03)
2
698
369
153
140
(N
7140
3110
1379
0.93
0.87
1.14
0.98
0.83
0.85
1.03
0.79
0.89
0.93
0.77
0.77
0.56
0.77

Women.*
(4.2)
Cup
(10.8)
(2.2)
(0.9)
(0.5)
(0.4)
(0.3)
1 = 31,355)
(0.88-0.98)
(0.90-1.01)
(0.93-1.11)
(0.92-1.10)
683
(0.76-0.96)
(0.81-1.03)
279
(0.71-1.03)
(0.69-1.01)
168
(0.66-1.06)
(0.70-1.13)
114
(0.88-1.67)
(0.92-1.76)
93
(0.58-1.08)
(0.67-1.25)
(N
3388
1313
173,141
0.93
0.95
1.02
1.01
0.86
0.91
0.85
0.84
0.84
0.89
1.22
1.27
0.79
0.91

among
Cup
(10.7)
(3.8)
(2.2)
(1.0)
(0.6)
(0.3)
(0.3)
<1 = 30,085)
(0.94-1.05)
(0.96-1.07)
(0.90-1.08)
(0.90-1.08)
673
(0.86-1.08)
(0.89-1.13)
315
(0.92-1.33)
(0.91-1.31)
191
(0.87-1.38)
(0.91-1.45)
91
(0.77-1.49)
(0.80-1.55)
95
(0.66-1.22)
(0.73-1.36)
(N
3221
1153
Mortality
0.99
1.01
0.99
0.99
0.96
1.00
1.11
1.09
1.09
1.15
1.07
1.11
0.90
1.00
Coffee
(10.4)
(3.8)
(2.2)
(0.9)
(0.6)
(0.3)
(0.3)

=20,865)

1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
No
783
461
187
115
57
71
Cause-Specific
(N
2161
and

All

966
462
446
Total
18,784
7,750
3,701
1,791
Participants
with
CI)
CI)
CI)
CI)
CI)
CI)
CI)
(95%
(95%
(95%
(95%
(95%
(95%
(95%
Consumption
CI)
CI)
CI)
CI)
CI)
CI)
CI)
ratio
ratio
ratio
ratio
ratio
ratio
ratio
Coffee
(95%
(95%
(95%
(95%
(95%
(95%
(95%
ratio
hazard
ratio
hazard
ratio
hazard
ratio
hazard
ratio
hazard
ratio
hazard
ratio
hazard
Daily
of

(%)
hazard
(%)
hazard
(%)
hazard
(%)
hazard
(%)
hazard
(%)
hazard
(%)
hazard
disease
accidents
Death
deaths
deaths
deaths
deaths
deaths
deaths
deaths
of
of
of
of
of
of
and
of
of
disease
causes
No.
Age-adjusted
Multivariate-adjusted
No.
Age-adjusted
Multivariate-adjusted
No.
Age-adjusted
Multivariate-adjusted
No.
Age-adjusted
Multivariate-adjusted
No.
Age-adjusted
Multivariate-adjusted
No.
Age-adjusted
Multivariate-adjusted
No.
Age-adjusted
Multivariate-adjusted
T
a
b
l
e

3
.

Association

Cause
All
Cancer
Heart
Respiratory
Stroke
Injuries
Diabetes
1898
n engl j med 366;20 nejm.org may 17, 2012
The New England Journal of Medicine
Downloaded from nejm.org at Hinari Phase 2 sites on May 17, 2012. For personal use only. No other uses without permission.
Copyright (c) 2012 Massachusetts Medical Society. All rights reserved.

Coffee and Mortality
-
other causes according to smoking status (Tables
cessa fruits,
6 and 7 in the Supplementary Appendix). The re-
for
0.63
0.13
age;
of
sults were similar for most outcomes across cat-
<0.001
<0.001
egories of smoking status, with the exception of
smoking
estimates
death from heart disease, with associations that
baseline: of
at
appeared to be null in current smokers (P = 0.002
risk
(0.4)
(1.9)
time consumption
for interaction in men and P = 0.05 for interaction
21
(0.83-2.22)
(0.58-1.61)
96
(0.77-1.20)
(0.57-0.90)
factors and
in women). We also noted significant interactions
1.36
0.97
0.96
0.72
intake; addition,
between smoking and coffee consumption with
cigars,
In
respect to the overall risk of death from cancer;
following or energy
associations appeared to be modestly inverse for
(0.2)
(1.7)
the
total therapy.
men and women who had never smoked, but not
pipes
37
(0.44-0.99)
(0.40-0.91)
295
(0.70-0.94)
(0.64-0.87)
for of
for those who were former or current smokers.
0.66
0.60
0.81
0.74
However, associations between coffee consump-
activity; hormone
tion and death from cancer were not significant
adjusted nonuse or
for any single category of coffee consumption.
no).
(0.2)
(1.7)
were
physical
use
vs.
152
(0.50-0.89)
(0.51-0.94)
(0.68-0.85)
(0.71-0.89)
Discussion
1139
day, status;
(yes
0.66
0.69
0.76
0.79
analyses per
postmenopausal
In this large, prospective U.S. cohort study, we ob-
of
marital
relative
served a dose-dependent inverse association be-
smoked
tween coffee drinking and total mortality, after
(0.3)
(2.0)
no); nonuse
(0.64-1.20)
(0.68-1.29)
(0.79-1.01)
(0.83-1.07)
Multivariate
adjusting for potential confounders (smoking sta-
96
642
vs. or first-degree
tus in particular). As compared with men who did
0.88
0.93
0.89
0.94
cigarettes (yes use a
not drink coffee, men who drank 6 or more cups
in
follow-up. of
and
of coffee per day had a 10% lower risk of death,
diabetes
whereas women in this category of consumption
(0.3)
(2.1)
during number
cancer)
had a 15% lower risk. Similar associations were ob-
77
(0.57-1.10)
(0.59-1.15)
626
(0.87-1.11)
(0.90-1.16)
the status;
skin
served whether participants drank predominantly
0.79
0.82
0.98
1.02
died
supplements;
caffeinated or decaffeinated coffee. Inverse associa-
who
health
tions persisted among many subgroups, including
(0.3)
(2.0)
vitamin
participants who had never smoked and those
1.00
1.00
1.00
1.00
65
422
consumption;
of nonmelanoma
who were former smokers and participants with
participants
baseline);
a normal BMI and those with a high BMI. Asso-
than
for
nonuse
ciations were also similar for deaths that occurred
alcohol
448
are
before or
in the categories of follow-up time examined (0 to
3,220
(other
use
<4 years, 4 to <9 years, and 9 to 14 years).
years
deaths
fat;
Our study was larger than prior studies, and the
education;
cancer
of

10
of
of
number of deaths (>52,000) was more than twice
CI)
CI)
or
that in the largest previous study.22 Whereas the
level
(95%
(95%
saturated
results of previous small studies have been incon-
years,
CI)
CI)
numbers
history
and for
sistent, our results are similar to those of several
ratio
ratio
The group; <10 to
larger, more recent studies, including the Health
(95%
(95%
5 meat,
Professionals Follow-up Study and the Nurses'
ratio
hazard
ratio
hazard
ethnic
interval.
adjusted
or
Health Study.21 In the Health Professionals Fol-
years, white
low-up Study, the hazard ratio for death among
(%)
were
hazard
(%)
hazard
race <5 to meat,
men who drank 6 or more cups of coffee per
1
day, as compared with men who drank less
deaths
deaths
confidence index;
red cancer
year,
than 1 cup per month, was 0.80 (95% CI, 0.62 to
of
of
causes
from
(<1
1.04). In the Nurses' Health Study, the corre-
No.
Age-adjusted
Multivariate-adjusted
No.
Age-adjusted
Multivariate-adjusted
denotes
sponding hazard ratio for women was 0.83 (95%
Infections
Other
body-mass tion vegetables, death
*

CI
CI, 0.73 to 0.95). In the Japan Collaborative Cohort
n engl j med 366;20 nejm.org may 17, 2012
1899
The New England Journal of Medicine
Downloaded from nejm.org at Hinari Phase 2 sites on May 17, 2012. For personal use only. No other uses without permission.
Copyright (c) 2012 Massachusetts Medical Society. All rights reserved.

T h e n e w e ng l a n d j o u r na l o f m e dic i n e
Subgroup
Men
Women
All causes of death
Any coffee
Caffeinated
Decaffeinated
Cancer
Any coffee
Caffeinated
Decaffeinated
Heart disease
Any coffee
Caffeinated
Decaffeinated
Respiratory disease
Any coffee
Caffeinated
Decaffeinated
Stroke
Any coffee
Caffeinated
Decaffeinated
Injuries and accidents
Any coffee
Caffeinated
Decaffeinated
Diabetes
Any coffee
Caffeinated
Decaffeinated
Infections
Any coffee
Caffeinated
Decaffeinated
Other causes of death
Any coffee
Caffeinated
Decaffeinated
0.25
0.50
1.00
1.50
0.25
0.50
1.00
1.50
Hazard Ratio
Hazard Ratio
Coffee Protective
Coffee a Risk
Coffee Protective
Coffee a Risk
Figure 1. Subgroup Analysis of Associations between the Consumption of 4 or More Cups of Coffee per Day and Total
and Cause-Specific Mortality.
Hazard ratios for death from all causes and from specific causes are for the comparison of men and women who
drank 4 or more cups of coffee per day with those who did not drink coffee. Participants were classified as drinking
caffeinated or decaffeinated coffee according to whether they reported drinking caffeinated or decaffeinated coffee
more than half the time. Risk estimates for other categories of coffee consumption are shown in Tables 2 and 3 in the
Supplementary Appendix. Risk estimates were adjusted for the following factors at baseline: age; body-mass index;
race or ethnic group; level of education; alcohol consumption; the number of cigarettes smoked per day, use or non-
use of pipes or cigars, and time of smoking cessation (<1 year, 1 to <5 years, 5 to <10 years, or 10 years before
baseline); health status; diabetes (yes vs. no); marital status; physical activity; total energy intake; consumption of
fruits, vegetables, red meat, white meat, and saturated fat; and use or nonuse of vitamin supplements. In addition,
risk estimates for death from cancer were adjusted for history of cancer (other than nonmelanoma skin cancer) in a
first-degree relative (yes vs. no). In women, risk estimates were also adjusted for use or nonuse of postmenopausal
hormone therapy. Horizontal lines represent 95% confidence intervals.
1900
n engl j med 366;20 nejm.org may 17, 2012
The New England Journal of Medicine
Downloaded from nejm.org at Hinari Phase 2 sites on May 17, 2012. For personal use only. No other uses without permission.
Copyright (c) 2012 Massachusetts Medical Society. All rights reserved.

Download
Association of Coffee Drinking with Total and Cause-Specific Mortality

 

 

Your download will begin in a moment.
If it doesn't, click here to try again.

Share Association of Coffee Drinking with Total and Cause-Specific Mortality to:

Insert your wordpress URL:

example:

http://myblog.wordpress.com/
or
http://myblog.com/

Share Association of Coffee Drinking with Total and Cause-Specific Mortality as:

From:

To:

Share Association of Coffee Drinking with Total and Cause-Specific Mortality.

Enter two words as shown below. If you cannot read the words, click the refresh icon.

loading

Share Association of Coffee Drinking with Total and Cause-Specific Mortality as:

Copy html code above and paste to your web page.

loading