Full text of DOI:10.1207/S15327914nc391_4

This article was downloaded by: [University of Connecticut]
On: 27 April 2013, At: 23:47
Publisher: Routledge
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,
37-41 Mortimer Street, London W1T 3JH, UK
Nutrition and Cancer
Publication details, including instructions for authors and subscription information:
http://www.tandfonline.com/loi/hnuc20
Coffee and Alcohol Intake and Risk of Ovarian Cancer:
An Italian Case-Control Study
Alessandra Tavani , Silvano Gallus , Luigino Dal Maso , Silvia Franceschi , Silvia Maurizio
Montella , Ettore Conti & Carlo La Vecchia
Version of record first published: 18 Nov 2009.
To cite this article: Alessandra Tavani , Silvano Gallus , Luigino Dal Maso , Silvia Franceschi , Silvia Maurizio Montella ,
Ettore Conti & Carlo La Vecchia (2001): Coffee and Alcohol Intake and Risk of Ovarian Cancer: An Italian Case-Control Study,
Nutrition and Cancer, 39:1, 29-34
To link to this article: http://dx.doi.org/10.1207/S15327914nc391_4
PLEASE SCROLL DOWN FOR ARTICLE
Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions
This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to
anyone is expressly forbidden.
The publisher does not give any warranty express or implied or make any representation that the contents
will be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses should
be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims,
proceedings, demand, or costs or damages whatsoever or howsoever caused arising directly or indirectly in
connection with or arising out of the use of this material.

NUTRITION AND CANCER, 39(1), 29–34
Copyright © 2001, Lawrence Erlbaum Associates, Inc.
Coffee and Alcohol Intake and Risk of Ovarian Cancer:
An Italian Case-Control Study
Alessandra Tavani, Silvano Gallus, Luigino Dal Maso, Silvia Franceschi,
Maurizio Montella, Ettore Conti, and Carlo La Vecchia
Abstract: The relation between coffee and alcohol intake
consumption was found in two case-control studies (1,2), al-
though slightly, nonsignificantly increased risks were found
in five other case-control studies (3–7) and in a cohort study
and ovarian cancer risk was analyzed in a case-control
study conducted in Italy between 1992 and 1999. Cases were
1,031 women, aged 18–79 years, with incident, histologi-
cally confirmed invasive epithelial ovarian cancer, and con-
trols were 2,411 women, aged 17–79 years, admitted to the
hospital for acute nonneoplastic non-hormone-related dis-
eases. Coffee intake (mostly espresso and mocha) was not
associated with ovarian cancer risk, with an odds ratio (OR)
of 0.93 [95% confidence interval (CI) = 0.69–1.27] in drink-
ers of ³4 cups/day compared with drinkers of <1 cup/day.
No meaningful relation was observed with cappuccino (OR
= 1.06, 95% CI = 0.85–1.32 for drinkers compared with
nondrinkers), decaffeinated coffee (OR = 0.64, 95% CI
0.42–0.96), and tea intake (OR = 0.90, 95% CI =
0.75–1.08). Total alcohol intake was not associated with
ovarian cancer risk (OR = 1.09, 95% CI = 0.76–1.57 in
drinkers of ³36 g/day compared with never drinkers). No re-
lationship was found with wine (OR = 1.03, 95% CI =
0.70–1.50 for >39 g/day compared with never drinkers),
beer, amari, grappa, and spirits. No significant heterogene-
ity was found for coffee or total alcohol intake across strata
of age, education, parity, oral contraceptive use, family his-
tory of ovarian/breast cancer, body mass index, and calorie
intake. This study, based on a large data set, provides no
support for a causal association between invasive epithelial
ovarian cancer risk and coffee and alcohol intake.
(8). This led to the conclusion by a working group of the In
-
ternational Agency for Research on Cancer that caffeine was
not a carcinogen for the ovary but that a marginally in-
creased risk could not be excluded (9). Since then, a Norwe-
gian cohort study found a nonstatistically significant
increase in risk (relative risk = 2.0 for ³7 cups/day) (10), and
a hospital-based case-control study from Greater Athens
(11), based on 189 women, found no consistent association
of coffee intake with ovarian cancer risk. As for tea con-
sumption, there was no association with ovarian cancer risk
in case-control studies (3,6,7,12).
Most studies, including the Third National Cancer Survey
(13), investigating the role of alcohol consumption on the eti-
ology of ovarian cancer, found no association (2,3,11,14–17),
leading to the conclusion that alcohol intake was not associ-
ated with ovarian cancer risk (18). However, two case-control
studies found a slightly increased risk of ovarian cancer with
the amount of alcohol drunk (5,19), and another case-control
study (20) and the Iowa Women’s Health Study (21) found a
somewhat decreased risk.
We analyzed data from a large case-control study con-
ducted between 1992 and 1999 in Italy. These data are of
specific interest, given the specific pattern, quantity and type
of coffee consumption (largely espresso and mocha, i.e.,
nonfiltered coffee with a limited caffeine content) (22), and
the relatively high alcohol intake (mainly wine) in this popu-
lation. Moreover, the present large sample data set allowed
adequate adjustment for known and likely risk factors and
meaningful analyses in subgroups of patients.
Introduction
Coffee consumption and alcohol intake are among the
most common exposures worldwide, and in view of their
high prevalence of use, even modest associations with ovar-
ian cancer would have substantial public health relevance.
The relationship between coffee intake and ovarian can-
cer risk has been extensively studied. In the 1980s, a signifi-
cant increased risk of ovarian cancer with elevated coffee
Subjects and Methods
The data were derived from a case-control study of ovar-
ian cancer, conducted between January 1992 and September
A. Tavani, S. Gallus, and C. La Vecchia are affiliated with the Istituto di Ricerche Farmacologiche “Mario Negri,” 20157 Milan, Italy. L. Dal Maso is
affiliated with the Centro di Riferimento Oncologico, Aviano (Pordenone), Italy. S. Franceschi is affiliated with the Field and Intervention Studies Unit,
International Agency for Research on Cancer, Lyon Cedex, France. M. Montella is affiliated with the Servzio di Epidemiologia, Istituto Tumori “Fondazione
Pascale,” Naples, Italy. E. Conti is affiliated with the Servizio di Epidemiologia e Oncogenesi, Istituto Regina Elena per lo Studio e la Cura dei Tumori, Rome,
Italy. C. La Vecchia is also affiliated with the Istituto di Statistica Medica e Biometria, Università degli Studi di Milano, Milan, Italy.

take. The reproducibility of wine and total alcohol intake
showed correlation coefficients (r) >0.75, whereas the valid-
ity was somewhat higher for wine (r ~ 0.70) than for other
alcoholic beverages and total alcohol intake (26).
1999 in four Italian areas: Greater Milan, the provinces of
Pordenone, Padua, and Gorizia (Northeastern Italy), the
province of Latina (Central Italy), and the urban area of Na-
ples (Southern Italy). The interviewers were centrally
trained, and the structured questionnaire was tested for reli-
ability and reproducibility (23,24). Less than 4% of cases
and controls approached refused interview, and the response
rates did not vary across hospitals and geographic areas.
Cases were 1,031 women (median age 56, range 18–79 yr)
with incident, histologically confirmed invasive epithelial
ovarian cancer, admitted to the major teaching and general
hospitals in the areas under surveillance.
Controls were 2,411 women (median age 57, range
17–79 years) residing in the same geographical areas and ad-
mitted to the same network of hospitals as cases for a wide
spectrum of acute conditions unrelated to known or potential
risk factors for ovarian cancer. Women were specifically ex-
cluded if they were admitted to the hospital for hormonal
and gynecological diseases and if they had overgone bilat-
eral ovariectomy. Among controls, 26% had traumatic con-
ditions (mostly fractures and sprains), 28% nontraumatic
orthopedic disorders (mostly low back pain and disk disor-
ders), 15% acute surgical conditions (mostly abdominal,
such as acute appendicitis or strangulated hernia), and 31%
miscellaneous other illnesses (such as eye, ear, nose, throat,
and dental disorders).
Data Analysis
Odds ratios (ORs) of ovarian cancer and the correspond-
ing 95% confidence intervals (CIs) for subsequent levels of
coffee and alcohol drinking were derived using uncondi-
tional multiple logistic regression, fitted by the method of
maximum likelihood (27), after allowance for study center,
year of interview, quinquennia of age, education, parity, age
at menopause, oral contraceptive use, family history of ovar-
ian/breast cancer in first-degree relatives, body mass index,
and energy intake.
Results
The distribution of ovarian cancer cases and the compari-
son group according to selected variables is given in Table 1.
Table 1. Distribution of 1,031 Cases of Ovarian Cancer
and 2,411 Controls According to Selected Variables: Italy
(1992–99)
All interviews were conducted in the hospital using a
structured questionnaire that included information on age,
education and other socioeconomic factors, anthropometric
variables, physical activity, smoking and several other life-
style habits, dietary habits, a problem-oriented medical
history, history of cancer in first-degree relatives, gyne-
cological and obstetric data, and history of use of oral con-
traceptives and hormone replacement therapy. Information
on intake of methylxanthine-containing beverages included
frequency of consumption of regular coffee (in our popula-
tion, almost totally espresso and mocha), cappuccino (coffee
and milk), decaffeinated coffee, and tea.
The section on alcohol consumption included questions
on the daily number of drinks of wine (125 ml), beer (330 ml),
grappa (a typical Italian spirit, 30 ml), amari and digestives
(other types of Italian liquors drunk after meals, 30 ml), and
spirits (whisky, cognac, brandy, 30 ml). The ethanol content
per 100 ml of beverage was estimated as 10.5 g for wine, 2.8 g
for beer, 33.6 g for grappa, 26.9 g for amari and digestives,
and 32.0 g for spirits, on the basis of a specifically designed
food and beverages composition database (25). Daily intakes
of alcohol from different beverages were computed using the
number of days per week on which each type of beverage was
consumed and the average number of drinks per day. Infor-
mation about age at first and last use and stopping drinking
was also collected. Never drinkers were defined as life-long
alcohol nondrinkers, and occasional drinkers were defined as
those drinking 10–40 g/mo of ethanol (corresponding to 1–3
drinks/mo). A satisfactory level of reproducibility and valid-
ity of the pattern of alcohol consumption was observed
across different levels and types of alcoholic beverage in-
Cases
Controls
n
%
n
%
Age
<45 yr
183
287
325
236
17.8
27.8
31.5
22.9
443
615
724
629
18.4
25.5
30.0
26.1
45–54 yr
55–64 yr
³65 yr
Educationa
<7 yr
570
227
227
55.6
22.2
22.2
1,417
620
349
59.4
26.0
14.6
7–11 yr
>11 yr
Body mass indexb
<22.31 kg/m2
22.31–24.77 kg/m2
24.78–27.68 kg/m2
>27.68 kg/m2
Energy intakeb
<1,685 kcal/day
1,685–2,079 kcal/day
2,080–2,526 kcal/day
>2,526 kcal/day
Family historyc
No
256
264
239
280
24.6
25.4
23.0
27.0
608
601
614
588
25.2
24.9
25.5
24.4
173
265
303
298
16.7
25.5
29.2
28.7
687
599
561
564
28.5
24.8
23.3
23.4
902
129
87.5
12.5
2,291
120
95.0
5.0
Yes
Oral contraceptive use
Never
Ever
Number of children
0
921
110
89.3
10.7
2,142
269
88.8
11.2
184
572
275
17.9
55.5
26.7
381
1,268
762
15.8
52.6
31.6
1–2
³3
a: Sum does not add up to the total, because some values are missing.
b: Approximate quartiles.
c: Family history of ovarian/breast cancer in 1st-degree relatives.
30
Nutrition and Cancer 2001

Compared with controls, cases had similar age distribution
and body mass index but reported higher levels of education
(OR = 1.71, 95% CI = 1.33–2.20 for >11 yr compared with
<7 yr) and energy intake and more often had a history of
ovarian or breast cancer in first-degree relatives. With refer-
ence to hormonal and reproductive factors, cases reported
less frequent use of oral contraceptives and less frequently
had three or more children (OR = 0.59, 95% CI = 0.45–0.78
compared with nulliparae). These differences were ac-
counted for in the analysis.
Table 2 shows the distribution of cases and controls ac-
cording to intake of coffee and other methylxantine-
containing beverages and the corresponding ORs. All the
ORs were close to unity. Compared with coffee nondrinkers
or drinkers of <1 cup/day, the multivariate OR was 0.93 for
women drinking ³4 cups/day, with no trend in risk with
dose. Similarly, women drinking cappuccino had an OR of
1.06 compared with nondrinkers (<1 cup/mo). Only 4–6% of
cases and controls reported consumption of decaffeinated
coffee, mostly 1 cup/day, and the OR for drinkers vs. non-
drinkers was 0.64. Among women drinking tea, no apparent
association with ovarian cancer risk was found (OR = 0.90).
Table 3 shows the distribution of ovarian cancer cases
and the comparison group, with corresponding multivariate
ORs, according to various measures of alcohol drinking hab-
its. Considering total alcohol intake, compared with never
drinkers, the OR was 1.02 for <12 g/day, 1.29 for 12 to <24
g/day, 1.04 for 24 to <36 g/day, and 1.09 for ³36 g/day. No
trend in risk with dose was observed. The separate analysis
Table 3. Distribution of 1,031 Cases of Ovarian Cancer
and 2,411 Controls and Corresponding ORs With 95%
CIs According to Alcohol Intake: Italy (1992–99)
Cases
Controls
OR (95% CI)a
Total alcoholb,c
Never drinkers
<12 g/day
12 to <24 g/day
24 to <36 g/day
³36 g/day
c2 for trend
Wineb,c
288
268
218
180
77
830
550
422
427
177
1d
1.02 (0.80–1.30)
1.29 (1.00–1.67)
1.04 (0.80–1.36)
1.09 (0.76–1.57)
0.68 (p = 0.409)
Never drinkers
£13 g/day
322
242
219
181
67
908
493
415
432
159
1d
1.06 (0.83–1.35)
1.37 (1.06–1.77)
1.07 (0.82–1.39)
1.03 (0.70–1.50)
0.75 (p = 0.386)
>13 to 26 g/day
>26 to 39 g/day
>39 g/day
c2 for trend
Beerc,e
Never drinkers
Drinkers
Amaric,e
Never drinkers
Drinkers
Grappac,e
Never drinkers
Drinkers
Spiritsc,e
713
318
1,882
527
1d
1.15 (0.94–1.41)
920
111
2,215
194
1d
1.29 (0.96–1.73)
947
84
2,256
154
1d
1.20 (0.86–1.67)
Never drinkers
Drinkers
964
67
2,274
136
1d
1.20 (0.84–1.72)
a: Estimates from multiple logistic regression equations, including terms
for study center, year of interview, age, education, parity, age at meno-
pause, oral contraceptive use, family history of ovarian/breast cancer,
body mass index, and energy intake (without alcohol calories).
b: Amounts in grams correspond to approximately <1, 1 to <2, 2 to <3,
and ³3 drinks/day.
c: Sum does not add up to the total, because some values are missing.
d: Reference category.
e: Drinkers were subjects consuming ³1 drink/mo (corresponding to ~10
g/mo of ethanol).
Table 2. Distribution of 1,031 Cases of Ovarian Cancer
and 2,411 Controls and Corresponding ORs With 95%
CIs According to Coffee, Cappuccino, Decaffeinated
Coffee, and Tea Intake: Italy (1992–99)a
Cases
Controls
OR (95% CI)b
Coffee
<1 cup/day
for different sources of alcohol intake was also considered.
Consumption of wine, which accounts for 95% of the total
alcohol consumption in this population, closely resembled
the total alcohol drinking pattern, with OR of 1.06 for £13
g/day (corresponding to <1 drink/day), 1.37 for >13 to 26
g/day, 1.07 for >26 to 39 g/day, and 1.03 for >39 g/day and
no trend in risk with dose. Compared with never-drinkers,
the ORs for drinking beer, amari, grappa, and spirits were
1.15, 1.29, 1.20, and 1.20, respectively. Among beer drink-
ers, no trend in risk was seen with increasing consumption,
since the OR was 1.25 (95% CI = 0.99–1.57) for occasional
drinkers and 0.98 (95% CI = 0.72–1.34) for regular drinkers
(³1 drink/day).
188
244
282
162
155
440
525
651
413
382
1c
1 to <2 cups/day
2 to <3 cups/day
3 to <4 cups/day
³4 cups/day
c2 for trend
Cappuccinod
Nondrinkers
Drinkers
Decaffeinated coffeed
Nondrinkers
Drinkers
Tead
Nondrinkers
Drinkers
1.12 (0.85–1.48)
1.13 (0.86–1.47)
0.86 (0.64–1.16)
0.93 (0.69–1.27)
1.32 (p = 0.251)
792
239
1,961
450
1c
1.06 (0.85–1.32)
984
47
2,268
143
1c
0.64 (0.42–0.96)
509
522
1,193
1,218
1c
0.90 (0.75–1.08)
The risk of ovarian cancer in separate strata of selected
covariates is shown in Table 4 for coffee intake and in Table
5 for total alcohol intake. There was no significant consistent
heterogeneity with ovarian cancer risk for either coffee or al-
cohol intake in any subgroup of age at diagnosis, education,
parity, oral contraceptive use, family history of ovarian/
breast cancer, body mass index, and calorie intake.
a: Abbreviations are as follows: OR, odds ratio; CI, confidence interval.
b: Estimates from multiple logistic regression equations, including terms
for study center, year of interview, age, education, parity, age at meno-
pause, oral contraceptive use, family history of ovarian/breast cancer,
body mass index, and total energy intake.
c: Reference category.
d: Drinkers were subjects consuming ³1 cup/mo.
Vol. 39, No. 1
31

Table 4. ORs and 95% CIs of Ovarian Cancer According to Coffee Intake in Strata of Selected Covariates: Italy
(1992–99)a
Coffee Intakeb
1 to <2 cups/day
2 to <3 cups/day
3 to <4 cups/day
³4 cups/day
c2 for Trend
Age
<50 yr
50–59 yr
³60 yr
1.64 (0.94–2.85)
0.95 (0.56–1.64)
1.05 (0.71–1.57)
1.15 (0.68–1.94)
1.21 (0.71–2.04)
1.16 (0.78–1.72)
0.74 (0.42–1.30)
1.12 (0.65–1.96)
0.70 (0.43–1.14)
1.05 (0.61–1.80)
0.89 (0.50–1.57)
0.95 (0.54–1.65)
1.23 (p = 0.267)
0.01 (p = 0.931)
0.65 (p = 0.421)
Education
<7 yr
7–11 yr
>11 yr
1.17 (0.82–1.68)
1.16 (0.64–2.10)
1.04 (0.53–2.05)
1.14 (0.80–1.64)
1.15 (0.65–2.02)
1.11 (0.58–2.11)
0.96 (0.64–1.43)
0.78 (0.42–1.45)
0.75 (0.37–1.54)
1.01 (0.66–1.55)
0.95 (0.51–1.80)
0.75 (0.37–1.51)
0.13 (p = 0.719)
0.62 (p = 0.431)
1.09 (p = 0.296)
Parous
No
Yes
1.49 (0.77–2.88)
1.09 (0.80–1.47)
1.03 (0.53–2.01)
1.16 (0.87–1.56)
0.70 (0.33–1.50)
0.88 (0.64–1.22)
0.65 (0.29–1.45)
0.97 (0.69–1.35)
3.03 (p = 0.082)
0.51 (p = 0.477)
Oral contraceptive use
Never
Ever
1.07 (0.80–1.43)
2.39 (0.97–5.93)
1.11 (0.84–1.48)
1.50 (0.64–3.47)
0.81 (0.59–1.12)
1.39 (0.58–3.29)
0.86 (0.62–1.20)
1.47 (0.62–3.50)
2.19 (p = 0.139)
0.09 (p = 0.759)
Family history of ovarian/breast cancer
No
Yes
1.16 (0.87–1.54)
1.36 (0.51–3.64)
1.15 (0.87–1.51)
1.23 (0.46–3.31)
0.84 (0.62–1.15)
1.26 (0.36–4.38)
0.98 (0.71–1.35)
0.49 (0.15–1.59)
0.97 (p = 0.324)
1.27 (p = 0.260)
Body mass index
Q1 (low)
Q2
Q3
Q4 (high)
1.00 (0.58–1.74)
1.26 (0.73–2.19)
1.06 (0.57–1.96)
1.10 (0.64–1.88)
0.97 (0.57–1.64)
1.18 (0.69–2.02)
1.07 (0.59–1.95)
1.06 (0.62–1.81)
0.69 (0.38–1.24)
1.05 (0.58–1.91)
0.93 (0.49–1.80)
0.66 (0.35–1.22)
1.15 (0.60–2.20)
0.82 (0.43–1.58)
0.51 (0.26–1.01)
1.23 (0.69–2.20)
0.13 (p = 0.722)
0.54 (p = 0.462)
3.71 (p = 0.054)
0.00 (p = 0.995)
Calorie intake
Q1 (low)
Q2
Q3
Q4 (high)
1.58 (0.86–2.90)
1.26 (0.75–2.12)
0.82 (0.47–1.43)
1.02 (0.56–1.87)
2.16 (1.19–3.92)
0.95 (0.57–1.58)
0.78 (0.46–1.32)
1.19 (0.67–2.13)
1.81 (0.87–3.75)
0.70 (0.39–1.25)
0.79 (0.44–1.41)
0.73 (0.40–1.34)
1.30 (0.57–2.94)
0.74 (0.39–1.39)
0.74 (0.41–1.36)
1.07 (0.59–1.95)
1.60 (p = 0.207)
2.86 (p = 0.091)
0.86 (p = 0.355)
0.09 (p = 0.765)
a: Estimates from multiple logistic regression equations, including terms for study center, year of interview, age, education, body mass index, parity, meno-
pausal status, oral contraceptive use, family history of ovarian/breast cancer, and total energy intake, as appropriate. Reference category was <1 cup/day of
coffee in each strata of covariate. Q, quartile.
b: Values are ORs, with 95% CIs in parentheses.
The issue of alcohol and ovarian cancer is also consistent
with the absence of any causal relationship, even at high lev-
els of consumption (18). Our results give further evidence on
this lack of association, with specific reference to wine,
since the Italian population is interesting, given the rela-
tively high consumption of wine and the originality of the
pattern of drinking, mainly limited to drinking during meals.
Similarly, no meaningful association emerged with other
types of alcoholic beverages, although information is limited
because of the low prevalence of consumption of alcoholic
beverages other than wine in this population and because the
amount consumed is restricted to occasional drinking or to 1
drink/day.
Discussion
Our results, based on a large study, confirm and provide
accurate information that consumption of either coffee or al-
cohol is not associated with invasive epithelial ovarian can-
cer risk. The large number of patients included in this study
allowed a worthwhile analysis of subgroups; however, no
consistent heterogeneity across strata of selected covariates
emerged with either exposure.
Thus our findings are consistent with most epidemiologi-
cal evidence (9,11) on a general lack of meaningful relation of
ovarian cancer risk in coffee drinkers. Likewise, no meaning-
ful association emerged with decaffeinated coffee and tea in-
take, although information is limited because of the low
prevalence of the habit in this population and because the
amount consumed is restricted to occasional drinking or to 1
cup/day.
Limitations and strengths of this study are common to
other hospital-based case-control studies (28). Although this
study was not population based, cases were identified in the
major teaching and general hospitals of the area under sur-
32
Nutrition and Cancer 2001

Table 5. ORs and 95% CIs of Ovarian Cancer According to Total Alcohol Intake in Strata of Selected Covariates: Italy
(1992–99)a
Total Alcohol Intakeb
1 to <12 g/day
12 to <24 g/day
24 to <36 g/day
³36 g/day
c2 for Trend
Age
<50 yr
50–59 yr
³60 yr
1.34 (0.89–2.02)
0.97 (0.61–1.53)
0.82 (0.54–1.24)
1.42 (0.85–2.38)
1.47 (0.90–2.41)
1.14 (0.77–1.69)
1.15 (0.65–2.01)
1.35 (0.82–2.20)
0.83 (0.55–1.25)
1.69 (0.80–3.58)
0.90 (0.47–1.73)
1.08 (0.60–1.92)
1.64 (p = 0.200)
0.76 (p = 0.384)
0.01 (p = 0.931)
Education
<7 yr
7–11 yr
>11 yr
1.09 (0.77–1.54)
1.05 (0.65–1.69)
0.96 (0.56–1.64)
1.52 (1.08–2.14)
1.45 (0.83–2.53)
0.75 (0.40–1.41)
1.03 (0.72–1.45)
1.41 (0.80–2.50)
0.95 (0.44–2.03)
1.14 (0.71–1.84)
1.36 (0.67–2.78)
1.09 (0.33–3.58)
0.48 (p = 0.490)
2.19 (p = 0.139)
0.11 (p = 0.739)
Parous
No
Yes
1.01 (0.56–1.81)
1.02 (0.78–1.33)
1.08 (0.53–2.21)
1.29 (0.98–1.71)
1.61 (0.78–3.32)
0.96 (0.72–1.28)
2.11 (0.81–5.52)
0.91 (0.61–1.36)
2.96 (p = 0.086)
0.01 (p = 0.913)
Oral contraceptive use
Never
Ever
1.04 (0.80–1.35)
0.77 (0.39–1.50)
1.34 (1.02–1.77)
1.03 (0.45–2.38)
1.05 (0.79–1.40)
0.89 (0.35–2.30)
1.19 (0.80–1.76)
0.52 (0.18–1.52)
1.20 (p = 0.274)
0.67 (p = 0.414)
Family history of ovar-
ian/breast cancer
No
Yes
1.05 (0.82–1.36)
0.84 (0.35–2.03)
1.36 (1.04–1.77)
1.40 (0.43–4.58)
1.12 (0.84–1.48)
0.67 (0.26–1.68)
1.06 (0.72–1.56)
2.39 (0.64–8.90)
1.04 (p = 0.307)
0.12 (p = 0.728)
Body mass index
Q1 (low)
Q2
1.56 (0.97–2.50)
0.80 (0.49–1.30)
0.79 (0.45–1.37)
1.01 (0.62–1.64)
1.18 (0.69–2.00)
1.09 (0.64–1.83)
1.18 (0.68–2.05)
1.85 (1.09–3.16)
1.49 (0.83–2.67)
1.10 (0.65–1.88)
0.84 (0.48–1.48)
0.89 (0.52–1.53)
0.87 (0.36–2.10)
0.75 (0.35–1.58)
1.61 (0.76–3.42)
1.21 (0.62–2.34)
0.19 (p = 0.661)
0.01 (p = 0.926)
0.48 (p = 0.490)
0.26 (p = 0.609)
Q3
Q4 (high)
Calorie intake
Q1 (low)
Q2
Q3
Q4 (high)
1.34 (0.69–2.57)
0.59 (0.38–0.93)
1.14 (0.74–1.77)
1.72 (0.99–2.97)
1.42 (0.74–2.75)
1.04 (0.65–1.68)
1.55 (0.97–2.46)
1.33 (0.73–2.41)
1.03 (0.48–2.21)
0.68 (0.41–1.14)
1.21 (0.74–1.97)
1.59 (0.90–2.83)
0.56 (0.16–2.01)
1.05 (0.54–2.06)
0.82 (0.40–1.67)
2.08 (1.00–4.32)
0.02 (p = 0.886)
0.03 (p = 0.863)
0.19 (p = 0.665)
3.01 (p = 0.083)
a: Estimates from multiple logistic regression equations, including terms for study center, year of interview, age, education, body mass index, parity, meno-
pausal status, oral contraceptive use, family history of ovarian/breast cancer, and energy intake (without alcohol calories), as appropriate. Reference cate-
gory was never drinkers of total alcohol in each strata of covariate.
b: Values are ORs, with 95% CIs in parentheses.
veillance, and the participation of cases and controls was al-
most complete. We carefully excluded from the control
group any patients admitted to the hospital for chronic con-
ditions or digestive tract diseases, which may modify pat-
terns of coffee or alcohol intake (29); thus only acute
conditions, unrelated to known or potential risk factors for
ovarian cancer or coffee or alcohol consumption, were in-
cluded in the comparison group. Moreover, the results were
consistent when separate analyses were done by major diag-
nostic categories of controls, particularly when traumatic
controls were excluded. The similar interview settings pro-
vide further reassurance against potential information bias,
there is no reason to assume different recall on the basis of
the disease status, and the potential confounding effect of
several covariates was allowed for in the analysis, but there
was no material modification of the risk estimates.
Acknowledgments and Notes
The authors thank J. Baggott and M. P. Bonifacino for editorial assis-
tance. This work was conducted with the contribution of the Italian Asso-
ciation for Cancer Research (Milan, Italy). Address correspondence to
Alessandra Tavani, Istituto di Ricerche Farmacologiche “Mario Negri,”
Via Eritrea 62, 20157 Milan, Italy. Phone: 0039-02-390 14460. FAX:
0039-02-39001916. E-mail: tavani@marionegri.it.
Submitted 20 July 2000; accepted in final form 2 November 2000.
References
1. La Vecchia, C, Franceschi, S, Decarli, A, Gentile, A, Liati, P, et al.:
Coffee drinking and the risk of epithelial ovarian cancer. Int J Cancer
33, 559–562, 1984.
2. Whittemore, AS, Wu, ML, Paffenbarger, RS, Jr, Sarles, DL, Kampert,
JB, et al.: Personal and environmental characteristics related to epithe-
lial ovarian cancer. II. Exposures to talcum powder, tobacco, alcohol,
and coffee. Am J Epidemiol 128, 1228–1240, 1988.
3. Byers, T, Marshall, J, Graham, S, Mettlin, C, and Swanson, M: A case-
control study of dietary and nondietary factors in ovarian cancer. JNCI
71, 681–686, 1983.
In conclusion, this large data set provides strong evidence
that coffee and alcohol drinking are not relevant determi-
nants of invasive epithelial ovarian cancer risk, even in strata
of major selected covariates, and that these exposures are not
of material or public health relevance.
Vol. 39, No. 1
33

4. Cramer, DW, Welch, WR, Hutchison, GB, Willett, W, and Scully, RE:
Dietary animal fat in relation to ovarian cancer risk. Obstet Gynecol
63, 833–838, 1984.
18. International Agency for Research on Cancer: Evaluation of Carcino-
genic Risks to Humans. Alcohol Drinking. Lyon, France: Int Agency
Res Cancer, 1988, vol 44. (IARC Monogr)
5. Tzonou, A, Day, NE, Trichopoulos, D, Walker, A, Saliaraki, M, et al.:
The epidemiology of ovarian cancer in Greece: a case-control study.
Eur J Cancer Clin Oncol 20, 1045–1052, 1984.
19. La Vecchia, C, Negri, E, Franceschi, S, Parazzini, F, Gentile, A, et
al.: Alcohol and epithelial ovarian cancer. J Clin Epidemiol 45,
1025–1030, 1992.
6. Miller, DR, Rosenberg, L, Kaufman, DW, Helmrich, SP, Schottenfeld,
D, et al.: Epithelial ovarian cancer and coffee drinking. Int J Epidemiol
16, 13–17, 1987.
20. Kato, I, Tominaga, S, and Terao, C: Alcohol consumption and cancers
of hormone-related organs in females. Jpn J Clin Oncol 19, 202–207,
1989.
7. Leviton, A: Methylxanthine consumption and the risk of ovarian ma-
lignancy. Cancer Lett 51, 91–101, 1990.
8. Snowdon, DA, and Phillips, RL: Coffee consumption and risk of fatal
cancers. Am J Public Health 74, 820–823, 1984.
9. International Agency for Research on Cancer: Evaluation of Carcino-
genic Risks to Humans. Coffee, Tea, Mate, Methlyxanthines and Meth-
ylglyoxal. Lyon, France: Int Agency Res Cancer, 1991, vol 51. (IARC
Monogr)
10. Stensvold, I, and Jacobsen, BK: Coffee and cancer: a prospective study
of 43,000 Norwegian men and women. Cancer Causes Control 5,
401–408, 1994.
11. Polychronopoulou, A, Tzonou, A, Hsieh, CC, Kaprinis, G, Rebelakos,
A, et al.: Reproductive variables, tobacco, ethanol, coffee and soma-
tometry as risk factors for ovarian cancer. Int J Cancer 55, 402–407,
1993.
12. La Vecchia, C, Negri, E, Franceschi, S, D’Avanzo, B, and Boyle, P:
Tea consumption and cancer risk. Nutr Cancer 17, 27–31, 1992.
13. Williams, RR, and Horm, JW: Association of cancer sites with tobacco
and alcohol consumption and socioeconomic status of patients: inter-
view study from the Third National Cancer Survey. JNCI 58, 525–547,
1977.
21. Kushi, LH, Mink, PJ, Folsom, AR, Anderson, KE, Zheng, W, et al.:
Prospective study of diet and ovarian cancer. Am J Epidemiol 149,
21–31, 1999.
22. Bonati, M, Latini, R, Galletti, F, Young, JF, Tognoni, G, et al.: Caf-
feine disposition after oral doses. Clin Pharmacol Ther 32, 98–106,
1982.
23. Franceschi, S, Negri, E, Salvini, S, Decarli, A, Ferraroni, M, et al.:
Reproducibility of an Italian food frequency questionnaire for cancer
studies: results for specific food items. Eur J Cancer 29A, 2298–2305,
1993.
24. Decarli, A, Franceschi, S, Ferraroni, M, Gnagnarella, P, Parpinel, MT,
et al.: Validation of a food-frequency questionnaire to assess dietary
intakes in cancer studies in Italy. Results for specific nutrients. Ann
Epidemiol 6, 110–118, 1996.
25. Salvini, S, Gnagnarella, P, Parpinel, MT, Boyle, P, Decarli, A, et al.:
The food composition database for an Italian food frequency question-
naire. J Food Composition Anal 9, 57–71, 1996.
26. Ferraroni, M, Decarli, A, Franceschi, S, La Vecchia, C, Enard, L, et al.:
Validity and reproducibility of alcohol consumption in Italy. Int J
Epidemiol 25, 775–782, 1996.
27. Breslow, NE, and Day, NE: Statistical Methods in Cancer Research.
The Analysis of Case-Control Studies. Lyon, France: Int Agency Res
Cancer, 1980, vol 1. (IARC Sci Publ 32)
14. West, RO: Epidemiologic study of malignancies of the ovaries. Cancer
19, 1001–1007, 1966.
15. Gwinn, ML, Webster, LA, Lee, NC, Layde, PM, and Rubin, GL: Alco-
hol consumption and ovarian cancer risk. Am J Epidemiol 123,
759–766, 1986.
16. La Vecchia, C, Decarli, A, Negri, E, Parazzini, F, Gentile, A, et al.: Di-
etary factors and the risk of epithelial ovarian cancer. JNCI 79,
663–669, 1987.
28. Mantel, N, and Haenszel, W: Statistical aspects of the analysis of data
from retrospective studies of disease. JNCI 22, 719–748, 1959.
29. Silverman, DT, Hoover, RN, Swanson, GM, and Hartge, P: The preva-
lence of coffee drinking among hospitalized and population-based con-
trol groups. JAMA 249, 1877–1880, 1983.
17. Hartge, P, Schiffman, MH, Hoover, R, McGowan, L, Lesher, L, et al.:
A case-control study of epithelial ovarian cancer. Am J Obstet Gynecol
161, 10–16, 1989.
34
Nutrition and Cancer 2001