Epidemiologic risk factors for breast cancer related to menopausal status in Indonesia

Epidemiologic Risk Factors for Breast

Cancer

Related

to

Menopausal

Status

in

Indonesia

i,""iï:ï.'JIj|,:_l;*,1,:Xôffi],",",',,îffilïii",'#îil:ffJi:ïi.',',';,ffïllffir:'i"::ï::h?rîiË:''

Esti

Soetrisno+, Eudang

Sri Roostini*

Abstrak

Untuk meneliti _faktor risiko kanker payudara yang berkairan dengan status menopause, dilakukan penelitian kasus kontrol pada rumah sakit

di

Indonesia'

Pada

300

kasus _{dan 600 kontrol yang berpasantan menurut umur dan keadaan sosial ekonomi dilalatkan} wareancard' Kasus dan kontrol dikelompokkan berdasarkan slatus menopausenya dan dianalisis terpisah. Diperoleh lemuan-temuan bermakna berikut ini. Di antara penderita kanker payudara pra menopauie ditemukan _{peningkatan risiko pada yang mengalami trauma} payudara (RR:2,62; 95vo

IK: 1,09-6,31), menggunakan kontrasepsi

oral

_{(4,96; 1,51-rc:24, mengkinsumsi susu (1,g1; 1,01-3,35} konsumsi

penuruna

minggu)'

Di

antara

iap hari)'

15 tahun

atau

lebih

g,6l

_{g4; 2,92-11,66, konsumsi tiap}

hari versus nir konsumsi). _{Penurunan risiko ditemulan pada wanita}

yang

_{5g), jumlah melahirkan hidup atau} bayi yang disusuinya tinggi (0,32; 0,13-0,76), menyusui 6 atau lebih

Abstract

a hospital-based case-control study was ls were selected, matching

for

age and tus _{and analyzed separately. Following} s detected in women with breast trauma 1-16.24), milk consumption (1.g1;

j.0l-3.35,

_{daily intake vs}

with cooked-veserabte intake (0.34;0.1s-0.77, dairy inrake vs not daity

t"i:;:l:;*';::P;"â::::îï:::Ï::::::tî:::,'::.:i::r

was _{founcl in} women having menarche at the age of 15 years or over _{(2.25; 1.35-3.76), regular menstruation after thirty years old (4.61;}

2'45-B

(0'33;

umption (5'84; 2.92-11.66, daily intake vs no intake), and a decreaseâ risk _{inwomenwhowere divorced orwidowed} d whose _{number of live birth or breast-fed children was high (0.32; 0.i3-0.76, six or more breast-fed children}

vs

no bre

Keywords i case'conlrol

sludy, premenopausal women, post menopausal women, _{breast cancer, risk factors}

Deparlment of Preventive Medicine, School of Medicine Nagoya _ Llniversity, Nagoya 466, Jupan

| _{[)epartment of Commrmitl,Med.icine, Focttlty of Medicitrc, UttiversiA} _ oflnd.onesia, Jakarla 10320, Indonesia

+ _{Departmeil of Analomic Pathologt, Faculty}

of Medicine, Iltriversity . oflndonesia,Jaka a

s _{Research Center, ine, IJniversily}

of Indortesin, ,,J t ka rla I 01 30, I ndot rcsia

tlDeportmetil _{of Public Health, Medical}

School, Nagoyt City lltriver-. sity, Nagoya 467, Japatt

' Deparlment of Surgery, Faculry of Medicine, IJtiversity of Indotresia, **Jaka rta I 01 j0, I ndoncsia

Departmcnt of Surgery, Catrcer Iustihtle Hospiral, Tol<-vo 170, Japntr

Risk

factors

of

breast cancer

(BC)

have

been

widely

discussed,

and different etiologic factors

have

_been suggested related to rnenopausal

status.l-6 BC

risk has

been

reported

to

be negâtively

associated

with

the

nurnber

of

pregnancies. This protective effect

_may

apply

only to

wornen whose

BC

was diagnosed at

50 years or

older.'

A

few

studies suggest that

parity

may be associated

with

an increased

risk of BC in

yourrg wolnell.S-9

Iu

posttrrenopausal

in

-l7O

Suanki

etal.

MedJ Indones

Cipto Mangunkusumo Central

Hospital, University of

Indonesia, 135 and 163 patients were at premenopausal and postmenopausal status respectively. Two patients were at

unknown

menopausal

status.

The age

distribu-tion

of cases

and controls

is shown

in

Table

1. Postmenopausal cases were significantly younger than

their

controls

(54.9 vs. 57.4

years

old),

whereas

premenopausal

cases were nof significantly so

(37.5

vs. 38.2 years

old). Table

2 shows the mean and

stand-ard

deviation (SD)

of

height, weight

and

obesity

in-dices as

BMI

and

waist

/

hip ratio.

In postmenopausal

wornen,

controls

were

significantly taller

by

1.2

cm,

and

slightly smaller

in

waist

/

hip

ratio than

cases.

Excluding

these

differences,

no

other

significant

case-control differences were found

in

both

premenopausal and post-menopausal

study

subjects.

Table

3

shows

the

RRs

of

breast cancer

in

pre-menopausal

women

obtained

by

univariate

analysis.

Breast trauma and the

use of oral

contraceptives

sig-nificantly

increased

the risk. History

of malignant

neoplasm arnong

relatives (mother,

aunts

or

sisters) increased the

risk by 9I7o, though

the increase was

not

significant.

The

risk

increased

with

increasing

rnilk-intake aud amount of fresh

fruits

consumed;

significant

linear

trends

were

found

(chi-square

for

trend

-

4.83,

P<0.05 and 4.68, P<0.05

respectively).

Daily

intake

of

cooked vegetables

significantly

decreased

the risk.

creased in

first-degree relatives

of premenopausal

BC

patient,

but

not

in

relatives

of postmenopausal

BC

patients.r l,actation

showed protective

effect against

premenopausal

BC alone.''o

The

studies

exemplified

above

strongly support the idea that premenopausal

worneu

differ

from

postmenopausal

women

concern-ing

BC risk

factors.

In view

of

the above

differences,

we have

analyzedthe

risk

factors

in

prernenopausal

and post menopausal

wornen separately

in

this

paper.

METHODS

Methodological

details of the case

control

study design

and data

collection

were already described

in

the

preceeding

paper.

In

this

paper,

cases

and controls

were grouped

according

to

their

rnenopausal status and

were analysed

separately.

Relative

risk

(RR)

with

its

957o

confidence interyal

(CI),

computed

as the

ex-posure-odds

ratio, was

used as

a meâsure

of

the

as-sociations between potential

risk

factors and

BC.

Trends were

evaluated

by

the

Mantel extensiou

test.l2

To

account

for

age, socioeconornic class,

residence

and

other

variables

potentially

confounded,

uncondi-tional logistic

regressiou

aualysis" was

performed.

RBSULTS

[image:2.595.63.553.466.762.2]

Out

of

300 histologically diaglosed

BC

patients

ex-amined

from December

1988

to

November

l99l

at

Table l.

Age

Distribution

o[ (hses and

Controls in Premenopausal andPôstmenopausal group

Age

Premenopausal group Postmenopausal group

Cases Controls Cases Controls

Number Vo Number % Number % Numl>er %

20-29 30-39 40-49 50-59 60-69 70-79

10.4 44.4 43.7 1.5

30 r63 124

l2

I.l

49.5 5t-l

3.7

6.2 20.4 32.r 35.8 5.6

4 3L 115 ro7

t4

1.5

tt.4

42.4 39.5 5.2

l4

60 59 2

l0

55 52 58 9

Totals 135 100.0 329 100.0

t62

100.0 271 100.0

Mean age

t

S.D.

37.5

t

NS = not sigrrificant,' 0.05' P > 0,01

Vol 4, No 3, fufu - September 1995

Table

2. Mean Vatues of Height,

Weight and Obesity lndicators

Menopausal Slatus

t7t

Premenopausal group _{Postmenopausal group}

Number Means

r

S.D. Number Means

t

S.D.

Height

(cm)

Cases

Controls

weight

(kg)

Cases

Controls

Maximum weight (kg)

Cases Controls Waist (cm) Chses Controls Hip (cm) Cases C-ontrols

BMI (kg/m/m)

Cases

Controls Waist

/

hip ratio

Cases Controls t33 324 134 325 94 223 133 323 132 323 t33 325 132 321

153.4

t

5.1

153.4

t

5.0

49.6

r

8.8

50.0

t

6.7

52.8

t

7.8

52.5

*

7.9

70.0

t

7.9

69.8

x, 7.5

91.3

*.9.7

91.5

r

9.1

21.0

*. 3.3

21.2

x. 2.7

0.77

=

0.09

O.77

r

0.09

*

5.2

t

5.4

50.8

r

10.6

51.5

*,8.2

54.8

r

10.9

55.5 t

8.6

74.2

t

ll.5

72.9

t

9.2

94.1

r

13.3

95.0 t

10.6

2t.9 27.9 NS NS NS NS NS NS NS NS NS NS NS NS 161 269 161 270 119

2tl

t60 269

l6l

270 160 269 160 269 152.3 r53.5

*

4.4

t

3.2

O.79

r

0.09

O.77

*

0.09 NS = not significant,' 0.05 > P > 0.01,

Table 3. Distribution of Risk Factors in Premenopausal Group

Factors Cases Controls R.R. 95Vo C.l. Trend

Breast trauma Never Ever

Pill use

No Yes

Malignancy among relatives No

Yes

Milk

Never

Less than once/week Once-twice/week 3-4 times/week

Daily

Fresh fruits

Never - less than once /weél( Once-twiceÂveek

3.4 tintes/week Daily

(looked vegetable Not daily Daily 120

l5

126 9

to'l

l9

38 26

2l

l2

38 23

l3

28 7t

l8

tt7

300

l3

324 5 290 2't 98 87 59 32 53 86 47 40 156 14 315 1.00 2.88 1.00 4.63 1.00 1.91 r.00 0.77 0.92 0.97 1.85 1.00 1.03 2.62 1.70 1.00 o.29

4.83

I

Reference (1.37

-6.07)

Reference (1.67

-

12.84)

Reference (1.03

-

3.54)

Relerence (o.43

-

t.37)

(0.4e

-

l.7l)

(0.45

-2.07)

(1.06

-3.23)

Reference (0.48

-2.23)

(1.36

-

5.05)

(1.0o

-

2.sl)

[image:3.595.54.551.96.719.2]

172

Suzaki

et al.

Table 4. Distribution of Risk Factors in Postmenopausal Group

Med J Indones

Factors Cases Controls R.R. 9s% c.r. Trend

Marital status

Manied

Separated or widowed Unmarried

Height (cm) <150

150 - 159

> 160

Waist / hip ratio

< 0.70

o;lo

- 0.79 0.80 - 0.89

È 0.90 Menarche

< 15 years old > 15 years old

86

68.

9 36 1t3 L4 28 57 53

))

7L 175 33 130 r44 t9 25 44 55 36 29 45 56

3l

148

l5

54 24 20

l5

50 95 173 3 50 t77 44 65 101 84 19 92 95 105 r65 263 8 22 35 150 62 27 33 153 57 259 12 138 63 27 9 33 1.00 0.43 3.3r 1.00 0.89 0.44 1.00

l.3l

r.46 2.69 r.00 2.36 1.00 2.5r 1.00 4.34 r.00

t.1l

0.32 o.52 1.00 t.27 0.34 0.51 1.00 2.t9 1.00 o.97 1.89 4.26 3.87 Reference

(0.2e

-

0.6s)

(0.e3

-

11.84)

Reference

(0.s4

-

1.4s)

(o.21

-

O.ez)

Reference

(0.76

-

2.27)

(0.84

-

2.s6)

(r.27

-

s.67)

Reference

(1.60

-

3.s0)

Reference

(t.60

-

3.92)

Reference

(r.e1

-

e.s7)

Re[erence

(0.s4

-

2.28)

(0.17

-

0.61)

(o.zs

-

1.03)

Reference

(0.64

-

2.s3)

(0.1e

-

0.62)

(0.26

-

1.00)

Reference

(r.oo

-

4.72)

Reference

(O.ss

-

r.7r)

(o.ee

-

3.63)

(1.8s

-

e.8r) (2.2e

-

6.ss)

4.30

Regularity of menstruation over 30 years

Irregular Regular Menopausal process

Natural Induced

Number of live birth None

r,2

3-5

>6

Number of lactation None

r,2

3-5

:6

Smoking habit Nonsmoker Current ex-smoker

Mitk

Never

Less than once / week Once-twice / week 3-4 times, week Almost daily

5.53

18.317

18.51I

26.07

Ï

28.83 1

32.201

'0.05

>P>().01,

_0.01

>P>

_0.001,10.001 [image:4.595.54.545.89.713.2]

menopausal

women

detected

by

univariate

analysis.

Compared

to married women,

separated

or

widowed

women

were

at

a

smaller risk, but

unmarried wolnen

were at a greater risk. When the marital

status

was

dichotomized

into unmarried

and ever married,

then

unmarried women showed

a

RR

of

5.22

(95o/o

CI

1.58-17.21). Height

was

inversely

associated

with BC

(P<0.05),

while

the larger the waist

/

hip ratio,

the

higher the

risk (P<0.05).

When compared to those

with

menarche

at

less

than

15 years

old,

those

with

menar-che

at

15 years

old or

more

had a

higher RR. Regular

menstruation

after 30

years

old

and

induced

rnenopause increased the

risk.

Since the nurnber

of

live

birth

and that

of lactation

showed

very similar

distribu-tion,

they showed

sirnilar RRs.

When

cornpared

with

nulliparous

womell,

those

with

one or

two births

had a

RRof

t.n

(0.54-2.28),3-5

births 0.32(0.17-0.61)

and 6

or more

births

0.52 (0.25-1.03); chi-square

for

trend

being 26.07 (P<0.001). Similarly when

compared

to women

with

no

breast-fed

child,

those

with

one or

two

breast-fed

children

had

a RR

of

1.27

(0.64-2.53),3-5

(0.26-1.00);

chi-square

for

trend being

28.83

(P<0.001). Compared

to

nonsmokers,

current or

ex-smoker experienced

a

higher

risk.

Postmenopausal

women who drunk

milk

less than once per

week

had a

RR of

0.97 (0.55-1.71), once

to twice

per

week

1.89

(0.99-3.63),

3-4

times per

week 4.26 (1.85-9.81),

al-most

daily 3.87

(2.29-6.55),

when

compared

to

non-drinker

of

milk;

linear trend being

significant

(P<0.001).

After

being adjusted

for

age,

socioeconomic

class,

residence and

other variables

potentially

confounded,

by

unconditional

logistic

regression

analysis,

the

results (Table 5,6) were

quite similar to

those before

adjustrnent. Association

between

BC

with

malignancy

among relatives (premenopausal),

height, waist

/

hip

ratio, unmarried

status,

smoking habit

and menopausal

process (postmenopausal) became insignificant or

marginally significant.

The number

of live

births

was

not included

in

the

logistic

regression analysis,

since

its distribution

was

very similar

to that

of

the

number

of

breast-fed

children.

Table 5. Adjusted Relative Risks

o[

Breast Cancer in Unconditional LogisticAnalysis in Premenopausal Group

Factors R.R 95% C..t.

Breast trauma

Never Ever

Pill

use

No Yes

Mal igna ncy among rel atives

No Yes

Mitk

Never

Less than once/week Once-twice/week 3-4 times/week Daily

Fresb fruits

Never - less than once/week Once-twice/ week

3-4 times/week Daily

Cboked vegetable Not daily Daily

Reference

(1.0e

-

6.31)

Reference

(t.sI

-

16.24)

Reference

(0.e8

-

4.02)

Reference

(o.42

-

t.s4)

(0.s0

-

l.es)

(0.42

-

2.23)

(1.01

-

3.3s)

Reference

(0.2e

-

r.67) (1.16

-

s.0s)

(0.74

-

2.66)

Reference

(0.ls

-

0.77)

1.00

2.62

1.00

4.96

1.00 1.99

1.00 0.81

0.9s 0.97

l.8l

1.00

0.70 2.42

1.40

1.00

[image:5.595.47.543.414.742.2]

1,74

Suzuki

et al.

Table 6. Adjusted Relative Risks

of

Breast Cancer in Unconditional l-ogistic Analysis in Postmenopausal Group

Factors R.R. 95Vo C.l.

MedJ

Indones

Marital status Manied

Separated or widowed Unmarried

Height (cm) < 150 150 - 159

:

160 Waist / hip ratio

< 0.70 0.70 - 0.79 0.80 - 0.89 = 0.90 Menarcbe

< 15 years old

:

15 years old

Reference

(0.18

-

0.s8)

(o.31

-23.63)

Reference

(0.40

-

1.36)

(0.17

-

1.13)

Reference

(0.48

-

2.04)

(0.20

-

1.00)

(o.3r

-

2.t4)

Reference

(r.3s

-

3.76)

Reference

(2.4s

-

8.6',t)

Reference

(o.84

-

7.74)

Reference

(o.48

-

2.72)

(0.13

-

0.63)

(0.13

-

0.76)

Reference

(0.76

-

7.0e)

Reference

(0.5s

-

2.33)

(0.e7

-

s.ze)

(2.63

-24.13)

(2.e2

-

rr.66)

Regularity of menstruation over 3O years Irregular

Regular

Menopausal process Natural

lnduced Number of lactation

None

l-2

3-5

>6

Smoking habit Nonsmoker C\rrent/ex-smoker

Milk

Never

Less than once

/

week Once-twice / week 3-4 times / week Almost daily

1.00 0.33 2.71

1.00 0.74 o.43

1.00 0.99 o.44 0.81

1.00 2.25

1.00 4.61

r.00

2.54

1.00 1.15 0.29 o.32

1.00 2.33

r.00

t.t3

2.26 7.96 5.84

DISCUSSION

In

the present

study, we found

the association between

premenopausal

BC

and breasl

trauma, oral

contracep-tive

use,

high intake of rnilk

and

fresh

fruits,

and

low

intake

of

cooked

vegetable.

Benign

b_reast diseases have been

reported

as

BC risk

factoi,11'14

but

there have been

nô

reports

on

the

relationship between breast trauma and

BC

risk.

Therefore,

studies

are required

to coufirm

such

as-sociation.

Many

reports

uoted

that

tbe use

of

oral

contraceptives

did not affect BC

,irk,15-17

while

our study

revealed the

contrary.

This

rnight

be due to the srnall number

of

pill

users

in

our series, thus

our

finding

might possibly

be

obtained

by

chance.

In

the

present

study,

milk

intake

had

positive

association

with BC risk

and

sig-nificant linear

trend was

observed among both

premenopausal and postmenopausal

group. Milk

might

represent

dietarv fat which

has

lons

been

in-crimiuateà

for

8C.18-2( Protective effect

if

cooked

vegetables

might

be

due

to

the vitamins, such

as

vitamin-

{

and beta-carotene, though inconsistent

r"sults20-22

were

reported. The

assoéiation

between

fresh fruits intake and

premenopausal

BC might

be

related

to

the

level

of total calory

intake."

Associations between postmenopausal

BC

risk with

[image:6.595.49.547.96.528.2]

live

birth or

breast-fed

children

and

milk

consumption

were noted

in

this study.

Our study showed that

unmarried

women had 5.2 times

higher risk

than

married women.

Unmarried women

had"a

higher

risk,

although

some reports

did

not prove

it.'''*

The

reason

why

separated

or

widowed women

were at a lower risk of

postmenopausal

BC

was

un-clear.

Menarche

at

older

age

was

associated

with

post-menopaFBal

BC

in

this study, but this

is not a

common

finding.l

1'25

Regular menstruation bqfore

_menopause is

known

to be a

risk

factor of

BC ,26'21

undour

finding

support

this.

Induced

menopause

was

found to

be

a

risk factor for

postrnenopausal

BC by our

univariate

analysis,

but not by logistic

regression

analysis.

No

risk difference was

detected

between natural

and

in-duced

menopuu.".28

An

inverse

linear

trend between

postmenopausal

BC

and

the number

of

live

birth

/

breast fed

children

remained

significant

after

ad.iusting

for

age and

other

factors

in

Jur

study.

tvtany striai"r7,fl,2n,'o huu"

reported that an increasing

number

of

pregnancy or

delivery

has

an

independent

protective

effect.

This

protective effect was found

rnore

consistently in older

or

postmenopausal

wolneu.

A

few

studies8i9 suggest that

parity

may be associated

with

an increased

risk

for

BC in young women.

These are

in

good

agreement

with our findings.

Postmenopausal

women

who

had

smoking

experience

had

approximately a two-fold BC risk

in

our

study,

however,

this was not

significant by logistic

regression

analysis. Most reports

suggested no clear association

between

BC

and

cigarette

smokiug

habits.31,32

CONCLUSION

Our study disclosed

some

different BC risk

factors

in

premenopausal and postmenopausal women.

We

found more risk factors

and stronger associations

in

postmenopausal than

in

premenopausal

women. Milk

intake wâs

a

common

risk

factor,

which

rnight

repre-sent

fat

intake.

Acknowledgrnents

The authors

are

grateful to

the

nurses

and the

public

health

uurses

for

excellent

care

and

data

collection

from

breast cancer cases

ald

controls. This

work

was

supported

by

the

Ministry

of

Education and Culture,

Japan,

Grants

No.

01042007 and

04042013; and was

This

collaborative study was

a

part

of

Special

Cancer Research

Project in Monbusho

lnternational Scientific

Research

Program,

with

the approval

of

the

Dean,

Faculty

of

Medicine, University

of

Indonesia,

No.

43 83 I PT 0 2.H4. FIVE/8 8.

REFERENCES

1. Clemmesen J. Carcinoma of the breast: Results from statis-tical research. BrJ Radiol I94B;2I:583-9O.

2. DeWaard F, Baanders-Van Halewijn EA, Huizihgs J. The bimodal age distribution

of

patients

with

mammary car-cinoma. Cancer 1964;17

:l4l-52.

3. Anderson

DE. A

genetic study

ol

human breast câncer. JNCI 1972;48:1029-34.

4. Craig T, Camstock GW, Geiser

GB.

Epidemiologic com-parison of breast cancer patients with early and late onset

o[

malignancy and general

population

controls.

JNCI 1974;53:1577-81.

5. Byere T, Graham S, Rzepka, Marsball J. t-actation and breast cancer. Am J Epidemiol 1985;12I:664-74.

6. McTiernan Ao Ttromas

DB.

Evidence for a protective effect o[ lactation on risk o[ breast cancer in young women. Am J

Epidemiol 1986;124:353 -8.

7. Leon DA. A prospective study of the independent etfects

of

parity and age at first birth on breast

cancer

incidence in England and Wales- Int J Cancer 19891,42:986-9I. 8. Pathak DR, Speizer FE, Willer WC, Rosner B, Lipnick RI.

Parity and breast cancer risk: possible effect on age at

diag-nosis.

lnt

J Cancer 1986;37:2l-5

9. Negri E,

ta

Vecchia

C,Bruzzi

p,Dardanoni G, Deearli A, Palli D, et

al.

Risk factors for breast cancer: period results from three Italian case-control studies.

Am J

Epidemiol 1988;128:1207

-I5

10. Swanson CA, Brinron LA, Taylor pR, Licitra

LM,

Ziegler RG, Schairer

C.

Body size and breast cancer risk assessed

in

women participating

in

the breast cancer detection demonstration project. Am J Epidemiol 1989;130:1133-41. 11. Yuan JM, Yu MC, Ross RK, Gao

yT,

Henderson BE. Risk lactors

for

breast cancer

in

Chinese women

in

Shanghai. Cbncer Res 1988;48: 1949-53.

12. Breslow NE, Day NE.The analysis of case-control study In: Davis

W,

editor.

Statistical methods

in

cancer research.

Vol.l.

Lyon: IARC, 1980:146-54.

13. Breslo _{WNE, Day NE. Uncondtional logistic regression for} large strata. In: Davis W, editor. Statistlcal methods in can-cer research, Vol. L Lyon: IARC, 1980:192-246.

14. Siskind V, Schofield F, Rice D, Bain C. Breast cancerand

breast feeding: results from an Australian case-control study. Am J Epidemiol L989;t3O:229-36.

15.

Miller

DR, Rosenberg

L,

Kautman DW, Stolley p, War-shauer ME, Shapiro S. Breast cancer before age 45 and oral

contraceptive use:

new findings.

Am

J

Epidemiol 1989;129:269-80.

176

Suzuki

et al.

17. UK National Case-Control Study Group. Oral contraceptive

use

and

breast cancer

risk

in

young women. Lancet 1989;l:931-82.

18.

Willet

WC, Stampfer

MJ,

Colditz GA, Rosner BA, Hen-neken SCH, Speizer FE. Dietary fat and the risk of breast

canoer. N Engl J Med 1987;316:22-8.

19. Hirohata T, Nomura

AM,

Hankin JH, Kolonel LN, Lee J.

An epidemiologic study on the association between diet and breast cancer. JNCI 1987;78:595-600.

20. Katsouyanni

Fi

Willet

WC,

Trichopoulos

D,

Boyle P,

Trichopoulo

A,

Vasilaros S, et al. Risk

of

breast cancer

among Greek wonen in relation to nutrient intake. Cancer

1988;61:181-5.

21. Rohan

TE,

McMichael

AI,

Bagharst

PA.

A

population-based case-control study

of

diet

and breast cancer in

Australia. Am J Epidemiol 1988;128:478-89.

22. Russel MJ, Ttromas BS, Bulbrook RD. A prospective study of the relationship between serum vitamins A and E and risk

of breast cancer. Br J Cancer 1988;57:213-15.

23. Iscovich JM, lscovich RB, Howe G,Shiboski S, Kaldor JM.

A case-control study ofdiet and breast cancer in Argentina. Int

J

Clancer 1989;44:770-6.

24. INtréIieman A, Thomas DB. Evidence for a protective effect

of lactation on risk of breast cancer in young woman.

Am

J Epi demiol 1986;724:,353-8.

25. Hsieh CC, Trichopoulos D, Katsouyanni K, Yuasa S. Age at

menarche, age

at

menopause, height and obesity as risk

MedJ

Indones

factors for breast cancer: associations and interactions in an

i ntemationa I case-control study. Int J Cancer 199O;46:796-800.

26. Olsson H, Landin-olsson M, Gullberg B. Retrospective

as-sessment of menstrual cycle leogth

in

patients

with

breast câncer, in patients with benign breast disease, and women without breast disease. JNCI 1983;70:17-2O.

27. Henderson BE, Ross RK, Judo

HL,

Krailo MD, Pike MC.

Do

regular ovulatory cycles increase breast cancer risk?

Cancer 1985;56:12O6-8.

28. Pa[feinbarger

Jr.

RS, Kampert

JB,

Chang

HG.

Charac-teristics that predict risk of breast cancer before and after the menopause. Am J Epidemi ol 198O;|12:258-68.

29. Eweft. M, Duffy SW. Risk of breast cancer in relation to

reproductive factors in Denmark. Br J Cancer

1988;58:99-t04.

30. Bouchardy C, Lê MG,

Hill

C. Risk factors for breast cancer

according to age at diagnosis in French case-control study.

J Clin Epidemiol 199O;43:267-75.

3L. Meara J, McPberson

K,

Roberts

M,

Jones

L,

Vessey M.

Alcohol, cigarette smoking and breast câncer. Br J C-ancer

1989;60:70-3.

32. Schechter MT, Miller AB, Howe GR, Baines CJ, Craib KJ,

Wall C. Cigarette smoking and breast cancer: case-control study of prevalent and incident cancer in Canadian National