Johnson JG, Cohen P, Pine DS, Klein DF, Kasen S, Brook JS. Association Between Cigarette Smoking and Anxiety Disorders During Adolescence and Early Adulthood. JAMA. 2000;284(18):2348-2351. doi:10.1001/jama.284.18.2348
Author Affiliations: Departments of Psychiatry, Columbia University and the New York State Psychiatric Institute (Drs Johnson, Cohen, Klein, and Kasen), and Community Medicine, The Mount Sinai Medical Center, New York, NY (Dr Brook); and Intramural Research Program, Program on Mood and Anxiety Disorders, National Institute of Mental Health, Bethesda, Md (Dr Pine).
Context Cigarette smoking is associated with some anxiety disorders, but the
direction of the association between smoking and specific anxiety disorders
has not been determined.
Objective To investigate the longitudinal association between cigarette smoking
and anxiety disorders among adolescents and young adults.
Design The Children in the Community Study, a prospective longitudinal investigation.
Setting and Participants Community-based sample of 688 youths (51% female) from upstate New York
interviewed in the years 1985-1986, at a mean age of 16 years, and in the
years 1991-1993, at a mean age of 22 years.
Main Outcome Measure Participant cigarette smoking and psychiatric disorders in adolescence
and early adulthood, measured by age-appropriate versions of the Diagnostic
Interview Schedule for Children.
Results Heavy cigarette smoking (≥20 cigarettes/d) during adolescence was
associated with higher risk of agoraphobia (10.3% vs 1.8%; odds ratio [OR],
6.79; 95% confidence interval [CI], 1.53-30.17), generalized anxiety disorder
(20.5% vs 3.71%; OR, 5.53; 95% CI, 1.84-16.66), and panic disorder (7.7% vs
0.6%; OR, 15.58; 95% CI, 2.31-105.14) during early adulthood after controlling
for age, sex, difficult childhood temperament; alcohol and drug use, anxiety,
and depressive disorders during adolescence; and parental smoking, educational
level, and psychopathology. Anxiety disorders during adolescence were not
significantly associated with chronic cigarette smoking during early adulthood.
Fourteen percent and 15% of participants with and without anxiety during adolescence,
respectively, smoked at least 20 cigarettes per day during early adulthood
(OR, 0.88; 95% CI, 0.36-2.14).
Conclusion Our results suggest that cigarette smoking may increase risk of certain
anxiety disorders during late adolescence and early adulthood.
Previous research has demonstrated that cigarette smoking is associated
with psychiatric disorders among adolescents and adults in the general population.1- 4 Although
research has indicated that a positive association exists between cigarette
smoking and some anxiety disorders,4- 6
currently, relatively little is known about the factors underlying this association.
Two types of hypotheses have been advanced to account for the association
between cigarette smoking and anxiety disorders. One hypothesis is that anxious
individuals are at elevated risk for smoking initiation due to factors such
as peer pressure,7 facilitation of social interaction,8 and the presumed calming effects of smoking.9 Research findings supporting this hypothesis have
indicated that adolescents with symptoms of anxiety or depression were at
higher risk for smoking initiation than asymptomatic adolescents,7 that adolescents and young adults with social fears
had an increased risk of onset of nicotine dependence,8
and that smoking in the presence of a distracting stimulus was associated
with reduced anxiety.9
Another hypothesis is that cigarette smoking contributes to the development
of anxiety disorders10 due to factors including
impaired respiration11 and the presumed anxiogenic
effects of nicotine.12,13 In support
of this hypothesis, clinical studies have indicated that cigarette smoking
preceded the onset of panic attacks among patients with panic disorder.4,10 In addition, Breslau and Klein11 have reported findings of a longitudinal epidemiological
investigation indicating that daily smoking, as well as pulmonary complaints,
were associated with the development of panic disorder among young adults.
West and Hajek13 have reported that cessation
of cigarette smoking was associated with a decline in anxiety over a 4-week
period of smoking abstinence.
Prospective epidemiologic research can investigate both types of hypotheses
by investigating whether anxiety disorders predict risk for future cigarette
smoking and/or whether chronic smoking is associated with risk for subsequent
anxiety disorders.11 Because both cigarette
smoking and anxiety disorders become prevalent during adolescence and early
adulthood, it is important to investigate the association between anxiety
disorders and cigarette smoking during adolescence and early adulthood in
a large community-based sample. We report such longitudinal findings from
the Children in the Community Study.14,15
Six hundred eighty-eight youths (51% female) were interviewed in their
homes. Participants' mean (SD) age in 1983 was 14 (3) years, in 1985-1986
was 16 (3) years, and in 1991-1993 was 22 (3) years.14,15
The participating families were a subset of 976 families, randomly sampled
on the basis of residence in upstate New York, with whom maternal interviews
had been conducted in 1975 when the mean (SD) age of the youths was 5 (3)
years. Written informed consent was obtained from participants at each assessment
after study procedures were fully explained. The study procedures were approved
by the New York State Psychiatric Institute's institutional review board.
The families in this study generally represented families in the northeastern
United States in socioeconomic status and most demographic variables, but
they reflected the sampled region with high proportions of those taking the
1983 survey being Catholic (54%) and white (91%).14
Participating families did not differ from the remainder of the original sample
with regard to offspring temperament or maternal psychopathology, although
paternal substance abuse in 1975 was less prevalent than in the remainder
of the original sample.
Child psychiatric disorders and cigarette smoking were assessed in the
1983 and the 1985 through 1986 surveys by parental and offspring interviews
using the Diagnostic Interview Schedule for Children,16
and in the 1991 through 1993 survey through offspring interviews using a modified
and age-appropriate version of the Diagnostic Interview Schedule for Children.
Parental psychopathology was assessed with the Disorganizing Poverty Interview15 and with items adapted from The New York High Risk
Study Family Interview.17 Offspring childhood
temperament was assessed in 1975 with the Disorganizing Poverty Interview.
The respondents were interviewed separately by extensively trained and supervised
lay interviewers who were blind to the responses of the other informant. Additional
information on methods used is available elsewhere.14,15
Analyses of contingency tables were conducted to investigate bivariate
associations between adolescent smoking and anxiety disorders, assessed when
the mean age was 16 years, and early adulthood smoking and anxiety disorders,
assessed when the mean age was 22 years. Complete data sets were available
for all 688 participants. A power analysis indicated that there was sufficient
statistical power to detect an association with a modest effect size. Nevertheless,
because there were few cases with some specific anxiety disorders, these statistical
analyses were repeated using an index of threshold or subthreshold anxiety
disorders. For an individual to be diagnosed with a subthreshold anxiety disorder,
the number of symptoms of that anxiety disorder was required to be at least
2 SDs above the sample mean. Inclusion of subthreshold anxiety disorders increased
the prevalence of threshold or subthreshold anxiety disorders by 143% when
the mean age was 16 years and by 48% when the mean age was 22 years. The results
of the statistical analyses were not affected by the inclusion of subthreshold
cases. Therefore, the analyses reported herein were conducted using the standard
threshold anxiety disorder diagnoses.
Logistic regression analyses were conducted to investigate whether these
associations remained significant after controlling simultaneously for age;
sex; difficult childhood temperament; alcohol and/or drug use and anxiety
and depressive disorders during adolescence; and parental smoking, education,
andpsychopathology. Statistical analyses were conducted to investigate associations
between anxiety disorders and heavy cigarette smoking, defined as smoking
at least 20 cigarettes per day. Additional analyses were conducted to investigate
associations between anxiety disorders and daily but less frequent cigarette
smoking, defined as smoking 1 to 19 cigarettes per day.
At mean age 16 years, 39 adolescents (6%) smoked at least 20 cigarettes
per day, and 44 (6%) had anxiety disorders. At mean age 22 years, 104 young
adults (15%) smoked at least 20 cigarettes per day, and 68 (10%) had anxiety
disorders. Twenty-two participants (3%) smoked at least 20 cigarettes per
day during both adolescence and early adulthood. The other 17 participants
(2%) who smoked at least 20 cigarettes per day during adolescence continued
to smoke cigarettes at least occasionally during early adulthood. Fourteen
adolescents (2%) had anxiety disorders during adolescence and early adulthood.18 Maternal and offspring reports of daily cigarette
smoking during adolescence were significantly correlated (r = 0.71; P<.0001).
Anxiety disorders during adolescence were not significantly associated
with cigarette smoking during early adulthood. Six (14%) of the 44 adolescents
with anxiety disorders and 98 (15%) of the 644 adolescents without anxiety
disorders smoked at least 20 cigarettes per day during early adulthood (odds
ratio [OR], 0.88; 95% confidence interval [CI], 0.36-2.14). Supplemental analyses
indicated that anxiety disorders at mean age of 14 years were not significantly
associated with the onset of cigarette smoking during adolescence or early
After controlling for covariates, adolescents who smoked 20 cigarettes
or more per day were at elevated risk for agoraphobia, generalized anxiety
disorder (GAD), or panic disorder during early adulthood (Table 1). Furthermore, the quantity and frequency of cigarette smoking
during adolescence was also associated with risk for agoraphobia (adjusted
OR, 1.47; 95% CI, 1.06-2.02), GAD (adjusted OR, 1.36; 95% CI, 1.07-1.71),
and panic disorder (adjusted OR, 1.65; 95% CI, 1.05-2.62) during early adulthood.
Chronic smoking during adolescence was significantly more strongly associated
with agoraphobia, GAD, and panic disorder than with obsessive-compulsive disorder
and social anxiety disorder, for which no increased risk was observed.
Overall, 12 (31%) of the 39 adolescents who smoked at least 20 cigarettes
per day had anxiety disorders during early adulthood. In comparison, 56 (9%)
of the 649 adolescents who were not chronic smokers had anxiety disorders
during early adulthood. There were 69 participants who smoked every day and
had an anxiety disorder during adolescence and/or early adulthood. Of these,
29 (42%) began smoking before they were diagnosed with an anxiety disorder,
and 13 (19%) were diagnosed with anxiety disorders before they reported daily
smoking (χ22 = 6.61; P
After controlling for covariates, heavy smoking during both adolescence
and early adulthood was associated with elevated risk for early adulthood
GAD (adjusted OR, 3.28; 95% CI, 1.42-7.61) and panic disorder (adjusted OR,
7.55; 95% CI, 1.55-36.86). Adolescents who smoked less than 20 cigarettes
per day were not at elevated risk for anxiety disorders during early adulthood
after the covariates were controlled. Among the 475 youths who did not smoke
cigarettes and who did not have anxiety disorders at mean age 14 years, heavy
smoking at mean age 16 years was associated with risk for anxiety disorders
during early adulthood (OR, 10.78; 95% CI, 1.48-78.55). Anxiety disorders
at mean age 16 years did not predict heavy cigarette smoking during early
adulthood in this subsample.
Statistically significant associations were obtained with the following
covariates: age (r687 = 0.17; P<.0001), female sex (OR, 3.51; 95% CI, 1.64-7.52), difficult childhood
temperament (OR, 2.49; 95% CI, 1.09-5.65), alcohol or drug abuse during adolescence
(OR, 9.95; 95% CI, 4.15-23.83), and depressive disorders during adolescence
(OR, 4.07; 95% CI, 1.57-10.53) were significantly associated with heavy cigarette
smoking during adolescence, and female sex (OR, 2.80; 95% CI, 1.61-4.87) and
depressive disorders during adolescence (OR, 6.88; 95% CI, 3.25-14.58) were
significantly associated with anxiety disorders during early adulthood.
Our review of the literature indicates that these are the first findings
from a community-based longitudinal study to demonstrate that heavy cigarette
smoking during adolescence is associated with increased risk for agoraphobia,
GAD, and panic disorder during early adulthood. Our findings are consistent
with research suggesting that cigarette smoking may increase risk for certain
At the same time, our findings indicate that cigarette smoking may not be
associated with risk for obsessive-compulsive disorder or social anxiety disorder.
Of considerable interest, previous research has indicated that impaired respiration
may be associated with agoraphobia,19 GAD,20 and panic disorder,19,21
but not with obsessive-compulsive disorder22
or social anxiety disorder.20 It will be of
interest for future research to investigate whether different mechanisms,
including impaired respiration,11,19- 21
and the potentially anxiogenic effects of sustained nicotine intake12,13 may underlie the associations between
cigarette smoking and agoraphobia, GAD, and panic disorder. It will also be
important to investigate possible biological or psychological vulnerability
factors that may increase risk for both cigarette smoking and certain anxiety
Although some previous studies have suggested that some types of anxiety
symptoms during adolescence may be associated with risk for initiation of
cigarette smoking7 or nicotine dependence,8 our findings suggest that adolescents with anxiety
disorders may not be at elevated risk for chronic smoking during early adulthood.
Further research will be needed to investigate whether some types of anxiety
disorders may increase risk for cigarette smoking under certain circumstances
or in specific populations.
Our findings provide health care professionals with additional evidence
regarding the harmful consequences of cigarette smoking. By providing adolescents
with information indicating that cigarette smoking may increase risk for the
onset of anxiety disorders, it may be possible to increase the effectiveness
of interventions that are designed to persuade young people to stop smoking
cigarettes and to avoid initiating cigarette use.