See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/10749826

Cannabis use and anxiety in daily life: A naturalistic investigation in a non-

clinical population

Article  in  Psychiatry Research · June 2003

DOI: 10.1016/S0165-1781(03)00052-0 · Source: PubMed

CITATIONS READS

92 196

5 authors, including:

Marie Tournier Joel Swendsen

Centre Hospitalier Charles PERRENS French National Centre for Scientific Research
99 PUBLICATIONS   1,505 CITATIONS    207 PUBLICATIONS   13,467 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Clozapine Covid View project

All content following this page was uploaded by Helene Verdoux on 15 January 2019.

The user has requested enhancement of the downloaded file.

 Psychiatry Research 118 (2003) 1–8

Cannabis use and anxiety in daily life: a naturalistic investigation

in a non-clinical population
Marie Tourniera, Frederic
´ ´ Sorbaraa, Claire Gindrea, Joel D. Swendsenb, Helene
´ ` Verdouxa,*
a
Department of Psychiatry, University Victor Segalen Bordeaux 2, Bordeaux, France
b
Laboratory of Clinical Psychology and Psychopathology, University Victor Segalen Bordeaux 2, Bordeaux, France

Received 28 May 2002; received in revised form 22 January 2003; accepted 19 February 2003

Abstract

The study’s objective was to investigate in a non-clinical population the association between cannabis use and
anxiety in daily life using the Experience Sampling Method (ESM). Seventy-nine subjects with high or low levels
of cannabis use were selected among a sample of 685 undergraduate university students. ESM was used to collect
information on cannabis use and state-anxiety in daily life. DSM-IV diagnoses were assessed using a structured
clinical interview. Statistical analyses were performed using multilevel linear random regression models. There was
no significant association between the level of state anxiety and cannabis use in daily life. However, a diagnosis of
agoraphobia was significantly associated with increased likelihood of cannabis use, independent of state anxiety and
other confounding factors. No evidence was found for an anxiolytic or anxiogenic effect of cannabis in daily life.
This finding does not support the hypothesis that subjects with high levels of anxiety use cannabis as a means of
self-medication. The association between agoraphobia and cannabis use in daily life may be explained by anticipatory
anxiety secondary to previous cannabis-induced panic-like symptoms.
Q 2003 Elsevier Science Ireland Ltd. All rights reserved.

Keywords: Cannabis; State-anxiety; Anxiety disorder; Experience Sampling Method

1. Introduction gested that cannabis use could promote anxious

Cannabis use has dramatically increased in ado-
lescents and young adults over the last decades
(Webb et al., 1996; Ogborne and Smart, 2000),
and consequently the mental health risks associated
with this substance have emerged as an important
public health issue. Among the potential risks
raised in the literature, several studies have sug-
*Corresponding author. Tel. q33-556-56-17-32; fax q33-
556-56-35-46.
E-mail address: helene.verdoux@ipso.u-bordeaux2.fr
(H. Verdoux).
symptoms in vulnerable subjects (e.g. Thomas,
1993; Hall and Solowij, 1998). Understanding the
association between cannabis use and anxiety is
therefore of considerable etiologic interest, in par-
ticular considering the high prevalence of anxiety
and substance use disorders in the general popu-
lation (Kessler et al., 1994; Regier et al., 1998).
Panic attacks constitute the most frequent acute
anxiety syndrome associated with cannabis use
(Tunving, 1987; Thomas, 1993; Hall and Solowij,
1998), and 20–30% of consumers present with
acute and brief anxiety reactions after smoking

0165-1781/03/$ - see front matter Q 2003 Elsevier Science Ireland Ltd. All rights reserved.
doi:10.1016/S0165-1781(03)00052-0
2 M. Tournier et al. / Psychiatry Research 118 (2003) 1–8
cannabis (Hollister, 1986; Thomas, 1996). Other
studies have suggested that cannabis use may be
associated with long-lasting anxious symptomatol-
ogy. In a study of persons who regularly used
cannabis for at least 10 years, Reilly et al. (1998)
found that 21% of these subjects had high levels
of state anxiety, and several case reports have
described cannabis-induced agoraphobia (Moran,
1986) and panic disorder (Deas et al., 2000; Langs
et al., 1997).
Few epidemiological investigations have exam-
ined the comorbidity of anxiety disorders and
cannabis use in the general population. Using data
from the Australian National Survey of Mental
Health and Well-Being, Degenhardt et al. (2001)
have reported an association between cannabis use
over the past year and increased prevalence of
anxiety disorders. Among individuals with canna-
bis dependence, 17% had at least one anxiety
disorder compared to 5% of non-users. However,
this association disappeared after adjustment for
demographic characteristics, personality disorder
and use of others drugs. Although these studies
suggest that an association may exist between
cannabis use and high levels of trait or state
anxiety, they are not informative concerning the
direction of this association (Compton et al., 2000;
McGee et al., 2000; Mueser et al., 1998). That is,
it is currently unclear if anxiety is best conceptu-
alized as a consequence of cannabis use in vulner-
able subjects, or conversely, if anxiety disorders
may favor cannabis use. An additional possibility
is that both cannabis use and anxiety disorders
may be independently induced by a shared risk
factor, such as specific pre-existing personality
traits.
The inability of the existing literature to differ-
entiate between the diverse potential explanations
for the association of cannabis and anxiety is due
in part to the fact that their relation is likely to be
restrained to a brief time period (such as a few
hours). As such, standard assessment techniques
(used to compute ‘average’ anxiety levels, diag-
nostic status, or cannabis use over longer time
intervals) are unable to sufficiently investigate the
rapid interaction of these variables. Studies exam-
ining fluctuations of anxiety levels and their rela-
tion to cannabis use should therefore benefit from
the application of data collection techniques more
capable of capturing the brief life cycle of this
association, such as the Experience Sampling
Method (ESM) (Csikszentmihalyi and Larson,
1987; Swendsen and Norman, 1998; Swendsen et
al., 2000).
The objectives of the present study using ESM
in a population of students were as follows: (1) to
assess rapid changes in state anxiety levels in
response to cannabis use in daily life; (2) to
determine whether subjects with a diagnosis of
anxiety disorder were more likely to use cannabis
over the ESM follow-up period.
2. Method
2.1. Subjects
2.1.1. Baseline screening
The method has been described in detail in
previous work (Verdoux et al., 2002). Undergrad-
uate university students were invited to participate
in a study of daily life behavior and experiences.
After receiving a full description of the study,
subjects provided written informed consent to par-
ticipate. A standardized self-report questionnaire
was used to collect information on demographic
characteristics and substance use. Subjects were
asked to specify the frequency of use over the last
month (ranging from 1: never in the past 30 days
to 7: several times a day) concerning diverse
psychoactive substances including cannabis and
alcohol. Psychosis proneness was assessed using
the Community Assessment of Psychic Experienc-
es (CAPE) (Stefanis et al., 2002; Verdoux et al.,
2002).
2.1.2. Selection of the ESM group
The baseline sample included all students
attending an information meeting on course organ-
ization at the beginning of the university year. Of
the 685 subjects invited to participate in the survey,
649 fully completed the self-report screening ques-
tionnaire. The sample included 586 females and
63 males, consistent with the skewed gender dis-
tribution of students in psychology. The 649 sub-
jects had a mean age of 20 (S.D. 3) years; most
of them (ns619, 95.7%) were single. Nearly one
 M. Tournier et al. / Psychiatry Research 118 (2003) 1–8
out of three subjects (ns194, 29.9%) had used
cannabis over the last month (once in the past
month, ns46; two or three timesymonth, ns46;
once a week, ns26; 2-3 timesyweek, ns33; once
a day, ns22; more than once a day, ns21). The
median (interquartile range, IQR) CAPE positive
score was 29 (26–33).
In order to maximize the probability of observ-
ing sufficient variance in cannabis use in daily
life, a stratification procedure involving cannabis
(tetrahydrocannabinol; THC) consumption was
used to create a sample for the ESM phase. THC
consumption over the last month was categorized
into ‘low THC’ (no use over the past month, ns
455, 70.1%) vs. ‘high THC’ (more than once a
week, ns76, 11.7%). Respondents were also
selected on the basis of responses to the self-report
questionnaire of psychosis-proneness (PP), a var-
iable not examined by the present study but con-
trolled for in all analyses (see Verdoux et al.,
2003). Within each THC group equal numbers of
subjects with ‘low PP’ (0–27), ‘medium PP’ (28–
33), or ‘high PP’ (34–76) were randomly selected.
Since the baseline sample included less than 10%
males, a higher proportion of male subjects were
sampled within each THCyPP group in order to
increase the sex ratio (30% of males). Research
psychologists blind to the selection criteria for
each individual telephoned subjects selected
according to this stratification method, and those
agreeing to participate in the other phases of the
study received financial compensation (775).
Of the 88 subjects invited to participate in the
ESM phase of the study, seven declined to partic-
ipate and two were excluded at the completion of
the study due to deviations from the established
procedures. There were no significant differences
with regard to demographic and clinical variables
between these subjects and those included in the
ESM phase.
2.2. ESM procedure
ESM is an ambulatory self-assessment method
designed to collect information on subjective expe-
rience occurring over time in naturalistic settings
(Csikszentmihalyi and Larson, 1987; Swendsen
and Norman, 1998; Swendsen et al., 2000).
3
Responding to randomly programmed signals from
portable electronic devices, subjects are asked to
describe their present experience by answering a
brief questionnaire several times a day over con-
secutive days. During an ESM training session,
subjects were instructed on how to complete each
item of the ESM form and to complete the form
at each signal of a multi-alarm wristwatch. Sub-
jects were then studied in their daily living envi-
ronment. Over 7 consecutive days, the watch
emitted five alarm signals per day at randomized
moments within each of the following time per-
iods: 08.00–11:00 h; 11.00–14.00 h; 14.00–17.00
h; 17.00–20.00 h; and 20.00–23.00 h.
The ESM form collected information concerning
substance use for the period between the current
and previous signals (corresponding on average to
the previous 3 h), as well as state anxiety at the
moment of the signal. Substance use was explored
by the question: ‘Over the last period, did you use
any substances?’ (YesyNo), followed by an open
question, ‘if, yes, which substance(s) did you use?’
The level of state anxiety was assessed on the
experience sampling form using a seven-point
Likert scale that asked participants to evaluate
their level of anxiety at that moment, ranging from
1 (not at all anxious) to 7 (extremely anxious).
2.3. Clinical interviews
At the end of the ESM phase, the subjects were
interviewed using the Mini International Neuropsy-
chiatric Interview (MINI, 4.4 version) by two
research psychiatrists blind to the risk status of
subjects (cannabis use history) as well as with
regard to their ESM data. The MINI is a short
diagnostic interview designed to be used in non-
clinical populations (Lecrubier et al., 1997). This
instrument was designed to assess current and
lifetime diagnoses of psychiatric disorders accord-
ing to DSM-IV criteria (American Psychiatric
Association, 1994), including diagnoses of anxiety
disorders and substance use.
2.4. Statistical method
Statistical analyses were conducted using
4 M. Tournier et al. / Psychiatry Research 118 (2003) 1–8
STATA software (STATA, 2001). Multilevel linear
random regression models were used to estimate
the effect of the independent variable (cannabis
use) on the dependent variables (state anxiety).
ESM data can be conceptualized as two-level (or
hierarchical) data, with repeated observations
(ESM signal level) being nested within a given
person (subject level). Multilevel or hierarchical
linear modeling techniques are a variant of the
more often used unilevel linear regression analy-
ses. The advantages of these methods are that the
dependency of repeated measures within the same
person is taken into account and that it can
accommodate non-informative missing values
(Golstein, 1987). The B is the fixed regression
coefficient of the predictor in the multilevel model
and can be interpreted identically to the estimate
in a unilevel linear regression analysis (change in
y with one unit change in x). We first examined
the effect of cannabis (independent variable) on
state-anxiety level (dependent variable) within the
same ESM assessment period. Since the observa-
tions from a given subject that are temporally close
may be more similar than those further apart, the
variance explained by autocorrelation was taken
into account by including the autoregression factor
in the model (STATA XTREGAR procedure).
We subsequently explored using multilevel
regression analyses (STATA XTREG procedure),
i.e. without including the autoregression factor in
the model, whether: (i) cannabis use during a
given time period in the day was associated with
an increased level of state anxiety for the next
ESM assessment that same day; and (ii) the level
of anxiety reported during a given time period was
associated with increased cannabis use for the
subsequent ESM assessment.
Lastly, multilevel logistic regression giving odds
ratios (ORs) and 95% confidence intervals
(95%CI) was used to explore whether subjects
with a MINI diagnosis of anxiety disorder were
more likely to use cannabis over the ESM assess-
ment period.
All the models were adjusted a priori for gender,
age, and psychosis proneness scores, which were
simultaneously forced into the models.
3. Results
3.1. Subjects
The 79 subjects (24 My55 F) included in the
ESM phase had a mean age of 22.1 years (S.D.
5.3). Most of them were single (ns73, 92.4%).
One third (ns23, 29.1%) presented with at least
one current MINI anxiety disorder wpanic disorder
(ns4, 5%), agoraphobia (ns12, 15.2%), social
phobia (ns9, 11.4%), obsessive-compulsive dis-
order (ns5, 6.3%), generalized anxiety (ns7,
8.9%), and posttraumatic stress disorder (ns1,
1.3%)x. Eight (10.1%) subjects presented with at
least two anxiety disorders.
Of the 41 subjects identified as ‘high cannabis
users’ by the self-report questionnaire, 30 (73.2%)
fulfilled MINI criteria of cannabis abuse (ns12)
or dependence (ns18) vs. only one individual
(2.6%) among subjects identified as ‘low cannabis
users’. Only three (3.8%) subjects fulfilled the
MINI criteria of other illicit substance abusey
dependence reflecting psychostimulants (ns3) or
opiates (ns1); all three subjects also fulfilled
MINI criteria for cannabis abuseydependence.
Nearly one out of two (47.8%) subjects suffering
from at least one anxiety disorder misused
cannabis.
3.2. ESM measures
The percentage of missing data for ESM varia-
bles ranged from 7 to 10%. There were neither
large nor significant differences in the frequencies
of missing data according to demographic charac-
teristics or risk status of the sample. Out of 2765
ESM assessments, there were 2546 (92.1%) valid
ESM substance reports, including 375 (14.7%)
reports of cannabis use, 179 (7%) reports of
alcohol use, and seven reports of other drug use
(Ecstasy, ns5; cocaine, ns1; heroin, ns1).
There were 2534 (91.6%) valid reports concerning
state anxiety levels (mean 2.3, S.D. 1.6).
3.2.1. Is there a cross-sectional association
between state anxiety during the ESM assessment
period and cannabis use?
No association was found between level of state
anxiety and cannabis use during the same ESM
 M. Tournier et al. / Psychiatry Research 118 (2003) 1–8
Table 1
Cannabis use over the ESM assessment period and MINI anx-
iety disorders
Cannabis use
Adjusted OR (95%CI)
Adjusted for gender, age, psychosis proneness
No anxiety disordera b did not predict cannabis use during the next ESM
One anxiety disorder 2 (0.8; 5.1) Ps0.15
Two or more anxiety disorders 3 (1.05; 8.4) Ps0.04
OR for linear trendc 1.8 (1.1; 2.9) Ps0.03
Adjusted for gender, age, psychosis proneness, ESM level of
state anxiety
No anxiety disorder *
One anxiety disorder 2 (0.8; 5) Ps0.16
Two or more anxiety disorders 3 (1.03; 8.4) Ps0.04
OR for linear trendc 1.7 (1.1; 2.9) Ps0.03
a
Panic disorderyagoraphobiaysocial phobiaygeneralized
anxietyyobsessive-compulsive disorderyposttraumatic stress
disorder.
b
Baseline category.
c
Change in odds ratio (OR) moving from one category to
the other.
period (adjusted Bsy0.03, 95%CI y0.17; 0.12,
Ps0.7). This association was not modified after
further adjustment for alcohol use during the ESM
assessment period (adjusted Bsy0.03, 95%CI y
0.14; 0.15, Ps0.9) or psychostimulants (cocaine
or Ecstasy) use (adjusted Bsy0.03, 95%CI
y0.17; 0.12, Ps0.7). However, in the model
including both alcohol use and cannabis use, a
significant negative association was found between
alcohol use and state anxiety in that subjects
reported lower levels of state anxiety in the periods
marked by alcohol use than in the periods without
alcohol use (adjusted Bsy0.22, 95%CI y0.41;
y0.03, Ps0.02).
3.2.2. Is there a longitudinal association between
state anxiety and cannabis use over the ESM
assessment?
In order to better characterize the temporal
sequence between cannabis and state anxiety, we
explored whether cannabis use was associated with
an increased or decreased level of state anxiety in
the next ESM period, or conversely, whether
change in state anxiety level was associated with
an increased or decreased likelihood of cannabis
use in the next ESM period. We did not find any
5
significant longitudinal association between can-
nabis use and state anxiety level over the ESM
assessment. Cannabis use did not predict an
increased level of state anxiety in the subsequent
ESM assessment (adjusted Bsy0.001, 95%CI y
0.17; 0.17, Ps0.9), and the level of state anxiety
period (adjusted Bsy0.09, 95%CI y0.27; 0.07,
Ps0.26).
3.2.3. Is there any association between cannabis
use over the ESM assessment period and the
existence of MINI anxiety disorders?
Compared to subjects with no MINI anxiety
disorder, subjects with at least one anxiety disorder
were two times more likely to use cannabis over
the ESM assessment (ORs2.3, 95%CI 1.04; 5.2,
Ps0.04). We further explored whether this asso-
ciation was dependent on the number of anxiety
disorders using the categorization of ‘no anxiety
disorder’ (baseline category); ‘one anxiety disor-
der’; and ‘at least two anxiety disorders’. There
was a significant linear trend in the association
between the number of anxiety disorders and
cannabis use indicating that the greater the number
of MINI anxiety disorders, the greater the likeli-
hood of cannabis use over the ESM assessment
(Table 1). However, the association between can-
nabis use and number of MINI diagnoses of
anxiety disorder was not modified after further
adjustment for the level of anxiety over the ESM
assessment, indicating that level of state anxiety
did not explain the association between anxiety
disorder and cannabis use. In order to assess
whether this association was restricted to specific
diagnoses of anxiety disorder, the role of each
specific MINI diagnostic category was examined
by entering in a same model the diagnostic
categories of social phobia, agoraphobia, panic
disorder, and generalized anxiety. Obsessive–com-
pulsive disorder and posttraumatic stress disorder
were not included due to the small number of
subjects within each category. As demonstrated by
Table 2, agoraphobia was the only anxiety disorder
significantly associated with cannabis use. Subjects
presenting with agoraphobia were three times more
likely to use cannabis over the ESM assessment
than those without agoraphobia, independently
6 M. Tournier et al. / Psychiatry Research 118 (2003) 1–8
Table 2
Cannabis over the ESM assessment period and specific anxiety
disorders
Cannabis use
Adjusted OR (95%CI)
Adjusted for gender, age, psychosis proneness
Social phobia 0.5 (0.12; 1.96) Ps0.3
Panic disorder 1.4 (0.33; 5.7) Ps0.65
Agoraphobia 3 (1.16; 7.7) Ps0.02
Generalized anxiety 2.2 (0.6; 7.8) Ps0.24
Adjusted for gender, age, psychosis proneness, ESM level of
state anxiety
Social phobia 0.5 (0.12; 1.93) Ps0.3
Panic disorder 1.4 (0.33; 5.8) Ps0.66
Agoraphobia 3 (1.15; 7.8) Ps0.02
Generalized anxiety 2.2 (0.6; 7.8) Ps0.24
from the existence of another anxiety disorder.
Again, this association was unchanged after adjust-
ment for state anxiety over the ESM period,
indicating that the level of state anxiety did not
explain the association between MINI diagnosis of
agoraphobia and cannabis use.
3.2.4. Is there any association between alcohol use
and anxiety disorders?
In order to assess the specificity of the associa-
tion between MINI anxiety disorders and ESM
cannabis use, we explored the association between
MINI anxiety disorder and ESM alcohol use. The
likelihood of alcohol use over the ESM assessment
was not significantly different between subjects
with no MINI anxiety disorder and subjects with
at least one anxiety disorder (ORs0.95, 95%CI
0.4; 2.3, Ps0.9). Furthermore, there was no sig-
nificant linear trend between the number of anxiety
disorders and alcohol use as assessed over the
ESM assessment (ORs1.36, 95%CI 0.87; 2.1,
Ps0.17).
4. Discussion
The present findings argue that cannabis use
does not modify the level of state anxiety in daily
life, and similarly, that the likelihood of using
cannabis is not influenced by state anxiety. Sub-
jects with a diagnosis of anxiety disorders assessed
through diagnostic interview, however, were more
likely to use cannabis over the naturalistic assess-
ment, independently from their level of state anx-
iety. In particular, a significant association was
found between cannabis use over the ESM assess-
ment period and a diagnosis of agoraphobia in that
such subjects presenting with this diagnosis were
three times more likely to use cannabis over the
ESM assessment, independently from the existence
of other anxiety disorders.
4.1. Methodological limitations
We have little reason to suspect a selection bias
in this sample, since the rate of participation in
the study was satisfactory, with only 5% incom-
plete questionnaires at the baseline screening and
less than 10% refusals to participate in the ESM
phase. As a study sample, students may differ from
subjects from the general population in regard to
several characteristics, including the prevalence of
substance use disorders or anxious symptomatolo-
gy. However, it has been shown that these subjects
did not significantly differ from young subjects
from the general population regarding risk factors
for psychiatric disorders (Vrendenburg et al.,
1993). Moreover, it is unlikely that differences in
prevalence would have modified the direction and
the strength of the association between cannabis
use and anxiety. Indeed, there is little reason to
suspect that the status of ‘‘student’’ is associated
with a different impact of cannabis on anxiety
level compared to non-student status.
Another consequence of the sampling procedure
is that the students included in the ESM phase
were selected according to level of psychosis
proneness. However, this potential confound was
controlled by the fact that all analyses were a
priori adjusted for this variable. As a result of this
selection procedure, the ESM sample was charac-
terized by a relatively high prevalence of anxiety
disorders.
The potential underreporting of cannabis use
remains a possible bias. However, there is little
stigmatization of cannabis use in the student pop-
ulation (due to the widespread use of this sub-
stance), and the prevalence in the whole student
population was similar to that reported in com-
parable samples. In addition, this reporting bias, if
 M. Tournier et al. / Psychiatry Research 118 (2003) 1–8
any, should have attenuated the strength of the
association between cannabis and anxiety if sub-
jects who underreported cannabis use were also
more likely to underestimate their level of state
anxiety.
As we used a unimethod approach to measure
anxiety, we cannot exclude that different findings
would have been obtained if we had used measures
of changes across multiple systems (biological,
cognitive, and behavioral).
4.2. Interpretation of findings
The first hypothesis tested in the present study
was to assess changes in state anxiety in response
to cannabis use in daily life. The fact that no
significant cross-sectional or sequential associa-
tions were found between cannabis use and the
level of state anxiety could be interpreted as a
general lack of association between cannabis and
anxious symptomatology. While individuals may
indeed mention relaxation, stress and anxiety
reduction as their motivation for using cannabis
(Boys et al., 1999), our findings do not support
the hypothesis that most subjects with high levels
of anxiety use cannabis to self-medicate anxious
symptoms (Stewart et al., 1997). Indeed, there are
many tenable explanations for cannabis use which
include conviviality, recreational purposes, social
enhancement or affiliation to a peer group (Webb
et al., 1996; Stewart et al., 1997; Swift et al.,
2000), any of which could have masked its pos-
sible association with anxiety. However, a signifi-
cant association was nonetheless found in this
sample concerning state anxiety and alcohol use
(a substance also consumed for heterogeneous
reasons; Swendsen et al., 2000). We therefore have
no reason to suspect that the diversity of motiva-
tions for cannabis use should prevent detection of
a specific association with anxiety, should it exist.
The second hypothesis tested in the present
study was to assess whether subjects with a diag-
nosis of anxiety disorder were more likely to use
cannabis over the ESM follow-up period. We
found a diagnosis of anxiety disorders was asso-
ciated with an increased likelihood to use cannabis
over the naturalistic assessment. The lack of anx-
iolytic or anxiogenic effects of cannabis use in
7
daily life is in apparent contrast to its association
with anxiety disorders. This latter association was
independent of the level of state anxiety, and thus
cannot be explained by the fact that subjects with
anxiety disorders preferentially used cannabis
when they were anxious. One explanation for this
finding is supported by investigations demonstrat-
ing that panic-like symptoms are the most frequent
adverse effects of cannabis use, particularly among
inexperienced users (Tunving, 1987; Thomas,
1993; Hall and Solowij, 1998). Cannabis-induced
panic attacks may provoke an anticipation anxiety
and promote panic disorder among subjects with a
high level of anxiety sensitivity (Barlow, 1988;
Langs et al., 1997; Deas et al., 2000). In addition,
the lack of a significant association between panic
disorder and cannabis use in the present study may
therefore be explained by a deterrent effect of full
panic attacks, as has been observed for caffeine
(Boulenger et al., 1984). A speculative hypothesis
regarding the association of cannabis use and
agoraphobia concerns the emergence of anticipa-
tory anxiety secondary to previous cannabis-
induced panic-like symptoms. Moran (1986)
reported six cases of subjects with a cannabis-
induced experience of depersonalization who
developed anticipatory anxiety, then agoraphobia.
Subjects presenting with acute depersonalization
may not associate the subsequent occurrence of
agoraphobic symptoms to cannabis use, and may
thus continue to use this drug despite the devel-
opment of agoraphobia. However, the interpreta-
tion of these findings is limited by the fact that
the temporal relationship between cannabis use
and agoraphobia has not been explored in this
study. Future investigations of this association may
benefit from integrating experience sampling meth-
odology into prospective investigations with fol-
low-ups over longer durations.
Acknowledgments
We thank Olivier Grondin, Mathilde Husky, and
Nadia Chakroun for their help in the organization
of the survey and in data entry.
 8 M. Tournier et al. / Psychiatry Research 118 (2003) 1–8
References
American Psychiatric Association, 1994. Diagnostic and Sta-
tistical Manual of Mental Disorders (DSM-IV). Fourth ed.
American Psychiatric Association, Washington, DC.
Barlow, D., 1988. Anxiety and Its Disorder. The Nature and
Treatment of Anxiety and Panic. Guilford Press, New York.
Boulenger, J.P., Uhde, T.W., Wolff, E.A., Post, R.M., 1984.
Increased sensitivity to caffeine in patients with panic
disorders. Preliminary evidence. Archives of General Psy-
chiatry 41, 1067–1071.
Boys, A., Marsden, J., Griffiths, P., Fountain, J., Stillwell, G.,
Strang, J., 1999. Substance use among young people: the
relationship between perceived functions and intentions.
Addiction 94, 1043–1050.
Compton, W.M., Cottler, L.B., Phelps, D.L., Ben Abdallah,
A., Spitznagel, E.L., 2000. Psychiatric disorders among drug
dependent subjects: are they primary or secondary? Ameri-
can Journal on Addictions 9, 126–134.
Csikszentmihalyi, M., Larson, R., 1987. Validity and reliability
of the Experience-Sampling Method. Journal of Nervous
and Mental Disease 175, 526–536.
Deas, D., Gerding, L., Hazy, J., 2000. Marijuana and panic
disorder. Journal of the American Academy of Child and
Adolescent Psychiatry 39, 12.
Degenhardt, L., Hall, W., Lynskey, M., 2001. The relationship
between cannabis use, depression and anxiety among Aus-
tralian adults: findings from the National Survey of Mental
Health and Well-Being. Social Psychiatry and Psychiatric
Epidemiology 36, 219–227.
Golstein, H., 1987. Multilevel Models in Educational and
Social Research. Griffin, London.
Hall, W., Solowij, N., 1998. Adverse effects of cannabis.
Lancet 352, 1611–1616.
Hollister, L.E., 1986. Health aspects of cannabis. Pharmaco-
logical Reviews 38, 1–20.
Kessler, R.C., McGonagle, K.A., Zhao, S., Nelson, C.B.,
Hughes, M., Eshleman, M.A., Wittchen, H.-U., Kendler,
K.S., 1994. Lifetime and 12-month prevalence of DSM-III-
R psychiatric disorders in the United States: results from
the National Comorbidity Survey. Archives of General
Psychiatry 51, 8–19.
Langs, G., Fabish, H., Fabish, K., Zapotoczky, H.G., 1997.
Can cannabis trigger recurrent panic attacks in susceptible
patients? European Psychiatry 12, 415–419.
Lecrubier, Y., Sheehan, D.V., Weiller, E., Amorin, P., Bonora,
I., Sheehan, K., Janavs, J., 1997. The Mini International
Neuropsychiatric Interview (MINI). A short diagnostic
structured interview: reliability and validity according to the
CIDI. European Psychiatry 12, 224–231.
McGee, R., Williams, S., Poulton, R., Moffitt, T., 2000. A
longitudinal study of cannabis use and mental health from
adolescence to early adulthood. Addiction 95, 491–503.
View publication stats
Moran, C., 1986. Depersonalization and agoraphobia associ-
ated with marijuana use. British Journal of Medical Psy-
chology 59, 187–196.
Mueser, K.T., Drake, R.E., Wallach, M.A., 1998. Dual diag-
nosis: a review of etiological theories. Addictive Behaviors
23, 717–734.
Ogborne, A.C., Smart, R.G., 2000. Cannabis users in the
general Canadian population. Substance Use and Misuse 35,
301–311.
Regier, D.A., Rae, D.S., Narrow, W.E., Kaelber, C.T., Schatz-
berg, A.F., 1998. Prevalence of anxiety disorders and their
comorbidity with mood and addictive disorders. British
Journal of Psychiatry 173 (Suppl. 34), 24–28.
Reilly, D., Didcott, P., Swift, W., Hall, W., 1998. Long-term
cannabis use: characteristics of users in an Australian rural
area. Addiction 93, 837–846.
STATA, 2001. STATA Statistical Software: Release 7.0. STA-
TA Corporation, College Station, TX.
Stefanis, N., Hanssen, M., Smyrnis, N., Avramopoulos, D.,
Evdokimidis, I., Verdoux, H., van Os, J., 2002. Evidence
that three dimensions of psychosis have a distribution in the
general population. Psychological Medicine 32, 347–358.
Stewart, S.H., Karp, J., Pihl, R.O., Peterson, R.A., 1997.
Anxiety sensitivity and self-reported reasons for drug use.
Journal of Substance Abuse 9, 223–240.
Swendsen, J.D., Norman, S., 1998. Preparing for community
violence: mood and behavioral correlates of the second
Rodney King verdicts. Journal of Traumatic Stress 11,
57–70.
Swendsen, J.D., Tennen, H., Carney, M.A., Affleck, G., Wil-
lard, A., Hromi, A., 2000. Mood and alcohol consumption:
an experience sampling test of the self-medication hypoth-
esis. Journal of Abnormal Psychology 109, 198–204.
Swift, W., Hall, W., Copeland, J., 2000. One year follow-up
of cannabis dependence among long-term users in Sydney,
Australia. Drug and Alcohol Dependence 59, 309–318.
Thomas, H., 1993. Psychiatric symptoms in cannabis users.
British Journal of Psychiatry 163, 141–149.
Thomas, H., 1996. A community survey of adverse effects of
cannabis use. Drug and Alcohol Dependence 42, 201–207.
Tunving, K., 1987. Psychiatric aspects of cannabis use in
adolescents and young adults. Pediatrician 14, 83–91.
Verdoux, H., Sorbara, F., Gindre, C., Swendsen, J., van Os, J.,
2002. Cannabis use and dimensions of psychosis in a non-
clinical population of female subjects. Schizophrenia
Research 59, 77–84.
Verdoux, H., Gindre, C., Sorbara, F., Tournier, M., Swendsen,
J.D., 2003. Effects of cannabis and psychosis vulnerability
in daily life: an experience sampling test study. Psycholog-
ical Medicine 33, 23–32.
Vrendenburg, K., Flett, G., Krames, L., 1993. Analogue versus
clinical depression. A critical reappraisal. Psychological
Bulletin 113, 327–344.
Webb, E., Ashton, C.H., Kelly, P., Kamali, F., 1996. Alcohol
and drug use in UK university students. The Lancet 348,
922–925.