Content uploaded by Geoffrey C Kabat
Author content
All content in this area was uploaded by Geoffrey C Kabat
Content may be subject to copyright.
Comparison
of
Smoking
Habits
of
Blacks
and
Whites
in
a
Case-Control
Study
Geoffry
C.
Kabat,
PhD,
Alfredo
Morabia,
MD,
PhD,
and
Emst
L.
Wynder,
MD
Intrducion
Incidence
rates
of
tobacco-related
cancers,
such
as
cancers
of
the
lung,
esophagus,
larynx,
and
oral
cavity,
are
higher
in
Black
than
in
White
Americans.1
In
addition,
Blacks
are
more
often
current
smokers
but,
on
average,
smoke
fewer
cigarettes
per
day
than
Whites.2-8
Blacks
also
tend
to
smoke
higher
tar,
higher
nic-
otine,
and
more
mentholated
cigarettes
than
Whites.5'9'10
Understanding
the
causes
of
racial
differences
in
smoking
habits
has
major
public
health
implications
since
it
deter-
mines
the
best
strategy
for
smoking
ces-
sation.
Racial
differences
are
apparently
not
attnbutable
to
socioeconomic
factors2
but,
to
our
knowledge,
the
role
of
smok-
ing-related
factors
such
as
tar
level,
men-
tholation,
inhalation,
degree
of
habitua-
tion,
butt
length,
and
quitting
has
not
been
examined
heretofore.
Metwds
Data
presented
here
derive
from
a
previously
descnibed
hospital-based
case-
control
study
of
tobacco-related
can-
cers.11
Briefly,
case
patients
with
cancers
of
the
lung,
larynx,
oral
cavity,
esopha-
gus,
bladder,
pancreas,
and
kidney
and
first
occurrences
of
nonfatal
myocardial
infarction
were
interviewed
in
28
collab-
orating
hospitals
in
eight
US
cities.
At
least
one
control
patient
with
a
condition
not
known
to
be
related
to
smoking
or
alcohol
was
interviewed
within
2
months
following
the
case
patient's
interview.
Control
patients
were
matched
to
case
pa-
tients
on
sex,
age
(+/-
5
years),
race,
and
hospital.
A
total
of 23
011
case
and
control
pa-
tients,
of
whom
9%
were
Black,
were
in-
terviewed
between
1980
and
1990.
After
removal
of
exclusively
pipe
and/or
cigar
smokers
(n
=
903),
5301
never
smokers
(8%
Black),
9252
current
smokers
(11%
Black),
and
7555
former
smokers
(6%
Black)
remained
for
analysis.
The
distribution
of
diagnoses
was
similar
between
Black
and
White
case
pa-
tients.
The
most
common
case
diagnoses
in
men
were
lung
cancer
(37%
of
White
and
Black
case
patients),
myocardial
in-
farction
(21%
of
Whites,
18%
of
Blacks),
oral
cancer
(13%,
17%),
and
larynx
cancer
(7%,
8%).
Among
control
patients,
Blacks
had
a
smaller
proportion
of
cancers
than
Whites,
because
Blacks
are
underrepre-
sented
in
collaborating
hospitals
with
al-
most
exclusively
cancer
patients.
The
study
questionnaire
contained
detailed
items
on
smoking
history,
includ-
ing
age
at
initiation;
tobacco
products
(cig-
arettes,
cigars,
pipes,
chewing
tobacco,
snuff);
up
to
four
cigarette
brands,
speci-
fying
whether
filter
or
mentholated;
years
of
smoking
each
brand;
and,
for
ex-smok-
ers,
the
number
of
years
since
quitting.
Information
was
also
available
on
inhala-
tion
(not
at
all,
slightly,
moderately,
deep-
ly),
how
much
of
the
cigarette
the
subject
usually
smoked,
and
how
long
after
wak-
ing
the
subject
generally
smoked
his
or
her
first
cigarette
of
the
day
(0
through
15
min-
utes,
16
through
30
minutes,
31
through
59
minutes,
60+
minutes).
A
smoker
was
defined
as
one
who
had
smoked
at
least
one
cigarette
per
day
for
at
least
1
year.
A
smoker
who
had
smoked
within
a
year
of
diagnosis
was
a
current
smoker,
otherwise
an
ex-smoker.
A
"heavy"
smoker
was
defined
as
a
smoker
of
more
than
a
pack
per
day
(2
21
ciga-
rettes
per
day).
Tar
level
was
categorized
into
low,
moderate,
and
high
tar
(<
10,
10
through
14.9,
2
15
mg/cigarette).
Mentho-
lation
was
categorized
into
nonmenthol
only
(those
who
had
never
smoked
a
men-
tholated
brand
for
at
least
1
year),
menthol
At
the
time
of
this
study,
all
authors
were
with
the
American
Health
Foundation,
New
York.
Alffedo
Morabia
is
now
with
the
Clinical
Epi-
demiology
Unit
of
the
University
Cantonal
Hospital,
Geneva,
Switzerland.
Requests
for
reprints
should
be
sent
to
Geoffrey
C.
Kabat,
PhD,
Division
of
Epidemi-
ology,
American
Health
Foundation,
320
East
43rd
Street,
New
York,
NY
10017.
This
paper
was
submitted
to
the
journal
April
5,
1991,
and
accepted
with
revisions
July
9,
1991.
American
Journal
of
Public
Health
1483
Public
Health
Briefs
adjust
for
confoundingvariables
and
to
as-
sess
multiplicative
interaction.
Results
The
breakdown
of
the
total
study
population
by
case-control
status,
sex,
and
race
is
displayed
in
Table
1.
As
in
national
surveys,
the
prevalence
of
cur-
rent
smoking
was
higher
in
Blacks.
Among
current
smokers
in
the
control
group,
Blacks
smoked
fewer
cigarettes
per
day,
tended
to
smoke
higher
tar
and
mentholated
cigarettes,
and
tended
to
have
smoked
for
fewer
years
compared
with
Whites
(Table
2).
Black
men
and
women
were
much
more
alike
in
terms
of
amount
smoked
than
were
White
men
and
women
(Table
2).
Among
control
patients
(not
shown
in
Table
2),
Black
men
were
5.0
(95%
con-
fidence
interval
[CI]
=
3.3-7.1)
times
more
likely,
and
Black
women
were
2.5
(95%
CI
=
1.4-6.7)
times
more
likely,
to
be
light
smokers.
For
case
patients,
the
corresponding
ORs
and
95%
CIs
were
3.0
(2.6-4.2)
for
men
and
3.0
(2.0-4.4)
for
women.
These
associations
persisted
across
strata
of
age,
education,
marital
status,
Quetelet's
Index,
menthol
use,
tar
level,
inhalation,
duration
of
smoking,
butt
length,
and
timing
of
first
cigarette.
As
shown
in
Table
3,
among
male
control
patients
the
association
of
race
with
light
smoking
was
greater
in
current
smokers
(OR
=
5.0;
95%
CI
=
3.4-7.5)
than
in
ex-smokers
(OR
=
2.5;
95%
CI
=
1.6-3.7).
The
interaction
was
statis-
tically
significant
(P
=
.01)
and
was
of
similar
magnitude
within
strata
of
age,
Quetelet
Index,
tar
level,
inhalation,
and
alcohol.
Differences
between
current
and
ex-smokers
were
even
stronger
among
more
educated
subjects
and
among
smok-
ers
of
mentholated
cigarettes.
In
these
two
subgroups,
among
current
smokers
Blacks
were
about
7
times
more
likely
than
Whites
to
be
light
smokers,
but
among
ex-smokers
Blacks
were
only
about
1.5
times
more
likely
to
be
light
smokers.
No
interaction
among
smoking
status,
level
of
smoking,
and
race
was
seen
in
female
control
patients (not
shown
in
Table
3).
1
through
14
years,
or
menthol
15
years
or
more
based
on
lifetime
smoking
history.
We
present
Black-White
compari-
sons
in
case
and
control
patients
to
dem-
onstrate
that
the
pattern
of
differences
is
not
an
artifact
of
the
study
design.
We
also
determined
whether
differences
in
smok-
ing
between
the
two
races
were
present
in
both
current
and
ex-smokers.
We
com-
puted
the
odds
ratio
(OR)
of
being
a
light
vs
a
heavy
smoker
related
to
race
(Black
vs
White)
and
used
logistic
regression
to
Diwuion
In
the
present
study,
more
Blacks
smoked;
however,
among
smokers,
Blacks
smoked
fewer
cigarettes
per
day
than
Whites.
This
pattern
was
constant
across
subgroups
defined
by
sex,
case-
control
status,
sociodemographic
factors,
1484
American
Journal
of
Public
Health
November
1991,
Vol.
81,
No.
11
Public
Health
Brie
life-style
factors,
and
smoking
behaviors,
and
is
consistent
with
US
survey
data.24,6
Novotny
et
al.
reported
that
Blacks
were
less
likely
than
Whites
to
be
heavy
smokers,
after
adjustment
for
employ-
ment
status,
blue/white
collar
status,
ed-
ucation,
and
poverty
level.2
Our
results
are
consistent
with
the
latter
findings
but,
in
addition,
suggest
that
the
larger
propor-
tion
of
light
smokers
among
Black
current
smokers
is
not
explained
by
differences
in
smoking-related
factors.
The
agreement
of
our
results
with
National
Health
Interview
Survey
data
makes
selection
or
recall
bias
of
hospital
patients
an
unlikely
explanation.4.12
How-
ever,
purchasing
power
could
account
for
the
racial
difference
in
level
of
smoking.
Although,
in
the
present
study,
Black-
White
differences
persisted
within
strata
of
educational
level,
education
may
be
a
poor
indicator
of
socioeconomic
status
when
comparing
Blacks
and
Whites.
Another
explanation
is
that
since
the
proportion
of
quitters
is
higher
among
Whites
and
since
light
smokers
are
more
likely
to
quit,13
this
Could
lead
to
an
over-
representation
of
heavy
smokers
among
White
current
smokers.
Indeed,
there
is
such
a
phenomenon,
but
it
does
not
com-
pletely
explain
the
observed
association
of
race
and
smoking
habits
since
Black-
White
differences
persist,
even
though
at-
tenuated,
among
ex-smokers
(Table
3).
It
is
also
possible
that
Blacks
are
less
suscepti-
ble
to
becoming
habituated
to
smoking
since
they
inhale
less
(Table
2)
and
have
higher
mean
cotinine
levels
compared
with
Whites
for
a
similar
level
of
smoking.14.15
In
conclusion,
smoking-related
fac-
tors
do
not
seem
to
explain
racial
differ-
ences
in
smoking
habits.
Thus,
effective
preventive
strategies
aiming
at
smoking
cessation
may
require
a
better
under-
standing
of
cultural
factors
that
affect
smoking,
such
as
the
timing
of
the
most
"important"
cigarettes,
whether
smoking
is
engaged
in
mainly
in
the
company
of
others,
reason
for
smoking,
reason
for
smoking
mentholated
cigarettes
(where
applicable),
and
degree
of
satisfaction
from
smoking.
O
Acknowledgments
Portions
of
the
material
contained
in
this
paper
were
presented
at
the
12th
scientific
meeting
of
the
International
Epidemiological
Association,
Los
Angeles,
August
9,
1990.
This
research
was
supported
by
National
Cancer
Institute
Contract
N01-CP.05684
and
Grant
CA-32617.
We
would
like
to
acknowledge
the
valu-
able
contributions
of
the
following
cooperating
.....
...
...
a
i~~~~~~~~~~~~~~~~
W
!0
||Xl
s
B
|
=
|4
E
;s
1
,ey
5Se_
u
_R
:x
__
2s.4
E
...,,.
<28
4ee.'..:2.4'...0
........
*
.........4
.
.4
...........U
........
institutions
and
individuals:
Memorial
Hospital
(New
York,
NY),
Dr.
David
Schottenfeld
and
Dr.
Nael
Martini;
Manhattan
Veterans
Hospi-
tal
(New
York,
NY),
Dr.
Norton
Spritz;
Long
Island
Jewish
Hillside
Medical
Center
(New
York,
NY),
Dr.
Arthur
Sawitsky,
University
of
Alabama
Hospital
(Birmingham,
Ala),
Dr.
William
Bridgers;
Birmingham
Veterans
Hos-
pital
(Bimingham,
Ala),
Dr.
Herman
L.
Leh-
man;
Loyola
University
Hospital
(Chicago,
IIl),
Dr.
Walter
S.
Wood;
Hines
Veterans
Hospital
(Chicago,
Ill),
Dr.
John
Sharp;
Henry
Ford
Hospital
(Detroit,
Mich),
Robert
M.
O'Bryan,
MD,
and
Christine
Johnson,
PhD;
Hospital
of
the
University
of
Pennsylaa
(Philadelphia,
Pa),
Dr.
Paul
Stolley;
Jefferson
Medical
College
and
Thomas
Jefferson
University
Hospital
(Philadelphia,
Pa),
Dr.
J.
E.
Colberg;
Alleghe-
ny
General
Hospital
(Pittsburgh,
Pa),
Dr.
Stan-
ley
A.
Briller;
University
of
Pittsburgh
Eye
and
Ear
Hospital
(Pittsburgh,
Pa),
Dr.
Lewis
H.
Kuller;
Pittsburgh
Veterans
Hospital
(Pitts-
burgh,
Pa),
Dr.
Eugene
N.
Meyers;
Moffitt
Hospital
and
University
of
California
at
San
Francisco
and
County
Hospital
(San
Fran-
cisco,
Calif),
Dr.
Nicholas
Petrakds;
St.
Luke's
Hospital
(San
Francisco,
Calf),
Dr.
Richard
A.
Bohannan;
Georgia
Baptist
(Atlanta,
Ga),
Dr.
A.
H.
Letton;
and
Emory
University
Clinic
(Atlanta,
Ga),
Dr.
K.
Mansour.
We
would
also
like
to
thank
Dr.
Ken
Resnicow
for
thoughtful
comments
on
the
manuscript,
Pao
Huan
Chung
for
programming
assistance,
and
Laurie
Marmelstein
for
preparation
of
the
manuscript.
References
1.
Ries
LAG,
Hankey
BF,
Edwards
BK.
Cancer
Statistics
Review:
1973-87.
Be-
thesda,
Md:
National
Cancer
Institute,
1990.
NIH
Publ.
No.
90-2789.
2.
Novotny
TE,
Warner
KE,
Kendrick
JS,
Remington
PL.
Smoking
by
Blacks
and
Whites:
socioeconomic
and
demographic
differences.
Am
J
Public
Health.
1988;78:1187-1189.
3.
US
Department
of
Health
Education
and
Welfare.
Report
of
the
Secretary's
Task
Force
on
Black
and
Minrity
Healh
Vol-
unew.
Washington,
DC:
US
Departmentof
Health
and
Human
Servies;
1985.
4.
US
Department
of
Health
and
Human
Services.
Health
Prom
on
and
Disease
Peeion
United
States,
1985.
Washing-
ton,
DC:
Public
Health
Service;
1988.
DHHSPublicationNo.
(PHS)88-1591.
Na-
tional
Health
Survey,
series
10,
No.
163.
5.
US
Department
of
Health
and
Human
Services.
Reducing
the
Health
Conse-
quences
of
Smoking:
Twenty-five
Years
of
Progress.
A
Repot
of
the
Swgeon
Gen-
erL
Washington,
DC:
US
Department
of
Health
and
Human
Services,
Public
Health
Service,
Center
for
Chronic
Disease
Pre-
vention
and
Health
Promotion,
Office
of
Smoldng
and
Health;
1989.
DHHS
Publi-
cation
No.
(CDC)
89-8411.
6.
Marcus
AC,
Shopland
DR,
Crane
LA,
Lynn
WR.
Prevalence
of
cigarette
smoking
November
1991,
Vol.
81,
No.
11
American
Journal
of
Public
Health
1485
Pubfic
Health
Briefs
in
the
United
States:
estimates
from
the
1985
Current
Population
Survey.
J
Natl
Cancer
Inst.
1989;81:409-414.
7.
Orleans
CT,
Schoenbach
VJ,
Salmon
MA,
et
al.
A
survey
of
smoking
and
quitting
among
Black
Americans.
Am
J
Public
Healt
1989;79:176-181.
8.
Stellman
SD,
Garfinkel
L.
Smoking
habits
and
tar
levels
in
a
new
American
Cancer
Society
prospective
study
of
1.2
million
men
and
women.
J
Natl
Cancer
Inst.
1986;76:1057-1063.
9.
Cummings
KM,
GiovinoG,
MendicinoAJ.
Cigarette
advertising
and
Black-White
dif-
ferences
in
brand
preference.
Public
Health
Rep.
1987;102:698-701.
10.
Sidney
S,
Tekawa
I,
Friedman
GD.
Men-
tholated
cigarette
use
among
multiphasic
examinees,
1979-86.
Am
JPublic
Health.
1989;79:1415-1416.
11.
Wynder
EL,
Steilman
SD.
Comparative
epidemiology
of
tobacco-related
cancers.
Cancer
Res.
1977;37:4608-4627.
12.
Sterling
TD,
Weinkam
JJ.
Comparison
of
smoking-related
risk
factors
among
Black
and
White
males.
Am
J
Ind
Med.
1989;15:319-333.
13.
Kabat
GC,
Wynder
EL.
Detenninants
of
quitting
smoking.
Am
J
Public
Health.
1987;77:1301-1305.
14.
Wagenknecht
L,
Cutter
G,
Smoak
C,
et
al.
Black/White
differences
in
cotinine
levels
among
smokers
in
the
United
States.
In.
Rand
MJ,
Thurau
K
(eds).
Proceed-
ings
of
Sateflite
Symposium
of
the
Tenth
Intemational
Congress
of
Phannacology:
The
Pharmacology
of
Nicotine.
Oxford:
IRL
Press.
ICSU
Symposium
series,
Vol.
9,
1988.
15.
WagenknechtLE,
Cutter
GR,
HaleyNJ,
et
al.
Racial
differences
in
serum
cotinine
lev-
els
among
smokers
in
the
Coronary
Artery
Risk
Development
in
(Young)
Adults
study.
Am
JPublic
Health.
1990;80:1053-
1056.
A
5-Year
Follow-up
Study
of
117
Battered
Women
Bo
Bergman,
MD,
DrMed
Sci,
and
Bo
Bnsmar,
MD,
DrMed
Sci
'.''..Ste.S
S
:'-
{
^
:'-
S'-'_'............
2-''''-
f'.'...'.'.'{.S
S'..':':,.'..,t,'.'sf:
'~.............
e.
'.
s
..'..
t
.
'
5
' '
5
. . .
'
.
},}-}}.i.
,,,,,~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.. , ,- _ . ' '..'. '.....
S~~~~~~~~~~~~~~~~~~~~~~~-
------------S
'S~~~~~~~~~~~~~~~~.......
sSSs
!
--
-S
s--ss--:sss
s~~~---------
s...S::.sS
s'f'
e'
s^.<s!ss
!s !ss_sS-
~~~~~~~~~~~~~~~~~~~~~~~~~~~~... ..............fi_f. ''',f
z..ie,¢,'.'
,
°¢
f
tf
stestE~~~~~~~~~~~~~~~~~~~~~~~~~~~.......
s,},S,ffisSs,aa
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~..s.-.<f
>.R
.
.._.S. ,-'..'
g~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.g..
5. 5.t..S sss- .f..
fss~~~~~~~~~~~~~~~~~~~.....
RE..S...........s:
f:-sS
Sa,:-:s::si
:f::-:::
S
f::
::::
sfefSs-st
atf
s
sss
s.
S
R.S*e
fSass;
.iss
2
Bs
a'
''
Sf
...:......
S
,:S,
- s-,,
,e-,,:,s
-
a
,
_
_
!
B ; f , ,
-
,
~~~~~~~~~~~~~~.
.--........
:,.,.,0.,,..,,,-,,.,
R
'
'~ ~ ~ ~ ~ ~ ~ ~
...............
.e
........
..
a...S'-.i
,e~~~~~~~~~~~~~~~~~~~~~~~.
.,-S
f
S,,fi.S,S:f.......
2$..
..
..
..ss
s.
..
...............f:
:5:t
.......................
------------fS::
::f::-:::-2::S:S
s:s:--:::.:::sS:
f
gs~~~~~~~~~~~~~~~~~~~~~~.
...'!"."""-''
.........:s
fs
a
-s
~~~~~~~~~~~~~~~~~~~...................
:::;::':'5::'':':
:a:.
::':
ss:
::sf::::':'::':S:'
:-
f
:"'
:'':Ss'
::
::
.:::::::-
:;:s:::
::-:
--:-:::::.::
:::::'.-::::::.''::f.::
s
f,
t f
ss
sf
5-
S'
"s5''2'''
--
Sfff
tf
-ss
2;"
5
*-~~~...............
............::::::
.......::::::::..:.2:::::
:....:::.
:s:s:
:
:
S::
::::::>
-:::::::
S:
S: :S
S
55
:S::::
:S
S:::S
:S5
:sS
2
f:::::::::::::::.f.S:.::..S.:.f.:.
':
-:S:
-
ss-:':'S:f':'sf:sf::2.::::-:::::,:::':'s
5:
::::::
::::::::::;S:
f:::::::::
...:.:....
sS:a.
f. :: :.:.:: :.
.S:.:S
:S:sS:
:::
:::
:::
::':::-
:
::
:5:
:::
:::
:..
.:::::::
::::::.::::
5S'
'''"!"""""'""
;'2""''sf'"'
i'"
'S"
" '"'
"'
5
"''"
'"
''''"
".".'''
.'.'"'
.''''."''
.'"
.'.'''
,,.-..'.'..'":."".
':''::
'::' :::'::
':'
-''''
5''S'S
S''S::S:
:::
:S
:
f
::
S
:a:f
S
:::
::::
:
::
ff:
:S.::
:::..::..:.:::..:....:
::
s'S
t' :: s' 'S R' 'S 'S
2
SS
S
SS.
SS
-.-.
S' k S SS
kS.......
S.
S..........
Intdudion
During
the
last
few
decades,
with
more
and
more
attention
being
given
to
the
problem
of
wife
battering,
various
in-
tervention
programs
aimed
at
stopping
the
battering
or
otherwise
helping
or
support-
ing
battered
wives
have
been
tried.1-3
However,
conclusive
information
con-
cerning
the
effect
of
different
intervention
programs
is
sorely
lacking.
In
this
study
the
use
of
medical
care
will
be
used
as
an
indicator
of
the
general
life
strain,
including
battering,
under
which
battered
women
live.
The
aim
of
the
study
was
to
investigate
the
use
of
somatic
and
psychiatric
hospital
care
by
battered
women
and
to
use
these
data
as
a
model
to
evaluate
the
effectiveness
of
a
wife
battering
intervention
program
in
a
medical
setting.
Subjects
and
Method
From
November
1983
to
June
1984,
there
was
an
emergency
room
project
at
the
Huddinge
Hospital,
Huddinge,
Swe-
den,
that
attempted
to
identify
and
sup-
port
battered
women
seeking
hospital
care.
All
women
in
the
emergency
room
who
reported
injuries
resulting
from
bat-
tering
were
offered
admission
to
the
hos-
pital
even
if
the
injuries
did
not
warrant
it
per
se.
The
women
were
given
supportive
counseling
by
a
social
worker
and,
if
nec-
essary,
they
talked
to
a
psychiatrist.
They
were
also
offered
continued
supportive
counseling
by
the
social
worker
after
re-
turning
home,
and,
on
these
occasions,
they
were
provided
an
opportunity
to
con-
tact
social
services,
lawyers,
and
other
sources
of
help.
During
the
8-month
period
in
ques-
tion,
a
total
of 117
battered
women
sought
help
in
the
emergency
room.
Fifty-eight
agreed
to
enter
the
treatment
program,
and
59
went
home
after
routine
care.
The
batterer
was
the
woman's
husband,
co-
habitant,
or
steady
partner
(65%);
her
ex-
husband
or
ex-cohabitant
(19%);
or
a
close
male
relative
(16%).
The
mean
age
at
the
time
of
the
battering
was
33
+
2
years
(range:
16-75).
The
battered
women
were
compared
with
a
control
group
that
was
selected
from
the
population
register
and
matched
with
the
group
of
117
battered
women
for
age
and
geographic
area
in
Stockholm.
The
battered
women
and
the
control
subjects
were
studied
regarding
consump-
tion
of
somatic
and
psychiatric
hospital
care
during
the
period
from
10
years
be-
fore
to
5
years
after
the
battering.
These
data
were
obtained
from
the
Stockholm
County
Council's
computer
files,
supple-
Bo
Bergman
is
with
the
Department
of
Psychi-
atry
and
Bo
Brismar
is
with
the
Department
of
Surgery
of
Huddinge
Hospital,
Karolinska
In-
stitute,
Huddinge,
Sweden.
Requests
for
reprints
should
be
sent
to
Dr.
Bo
Bergman,
Department
of
Psychiatry,
Karo-
linska
Institute/Huddinge
Hospital,
S-141
86
Huddinge,
Sweden.
This
paper
was
submitted
to
the
journal
May
5,
1990,
and
accepted
with
revisions
April
4,
1991.
November
1991,
Vol.
81,
No.
11