© 2006 American Psychological Association. Published by Blackwell Publishing on behalf of the American Psychological Association.
All rights reserved. For permissions, please email: firstname.lastname@example.org
Blackwell Publishing LtdOxford, UKCPSPClinical Psychology: Science and Practice0969-5893© 2006 American Psychological Association. Published by Blackwell Publishing on behalf of the American Psychological Association.All right reserved. For permission, please email: email@example.com Original Article
EXERCISE FOR MENTAL HEALTH META-ANALYSISCLINICAL PSYCHOLOGY: SCIENCE and PRACTICE
Exercise Interventions for Mental Health: A Quantitative
and Qualitative Review
Georgia Stathopoulou and Mark B. Powers, Boston University
Angela C. Berry and Jasper A. J. Smits, Southern Methodist University
Michael W. Otto, Boston University
Associations between exercise and mental well-being
have been documented repeatedly over the last two
decades. More recently, there has been application of
exercise interventions to clinical populations diagnosed
with depression, anxiety, and eating disorders with
evidence of substantial benefit. Nonetheless, attention
to the efficacy of exercise interventions in clinical settings
has been notably absent in the psychosocial treatment
literature, as have been calls for the integration of these
methods within the clinical practice of psychologists.
In this article, we provide a quantitative and qualitative
review of these efficacy studies in clinical samples and
discuss the potential mechanism of action of exercise
interventions, with attention to both biological and
psychosocial processes. The meta-analysis of 11 treat-
ment outcome studies of individuals with depression
yielded a very large combined effect size for the advan-
tage of exercise over control conditions:
g = 1.39 (95%
CI: .89–1.88), corresponding to a d = 1.42 (95% CI:
.92–1.93). Based on these findings, we encourage clinicians
to consider the role of adjunctive exercise interventions
in their clinical practice and we discuss issues concerning
Key words: anxiety, depression, eating disorders,
exercise, meta-analysis, treatment. [Clin Psychol Sci Prac
Numerous population-based studies provide evidence for
an association between aerobic exercise and mental health.
Some of the evidence comes from large-scale studies of the
general population (for review, see Biddle, 2000). For
example, a recent population study in Finland suggests
that individuals who exercised at least twice to three
times a week experienced significantly less depression,
anger, stress, and cynical distrust than those who exercised
less frequently or not at all (Hassmén et al., 2000). Likewise,
a cross-sectional study by Schmitz, Kruse, and Kugler (2004)
examined the relationship between health-related quality
of life and physical activity among 7,124 participants in
the German National Health Interview and Examination
Survey (GHS) from 1997 to 1999. The results indicated that
higher levels of physical activity were associated with higher
health-related quality of life among individuals with
affective, anxiety, and substance dependence disorders.
Along with large-scale studies examining exercise and
mental health in the general population, a recent popu-
lation study focusing on clinical conditions in the United
States showed a negative association between regular physical
activity and major depression as well as anxiety disorders
(Goodwin, 2003). Despite the consistency of the findings
between general and clinical populations, correlational
analyses leave open the possibility that the apparent health-
promoting effects of exercise may simply reflect the effects
of mood disturbances on physical activity levels. Fortunately,
experimental studies have helped confirm that programmed
aerobic exercise has consistent effects on promoting mental
well-being in nonclinical populations. Meta-analytic reviews
indicate effect sizes in the range of small to medium
.36) or greater according to Cohen’s standards for the
advantage of exercise over relevant control conditions
Address correspondence to Michael W. Otto, Center for Anxiety
and Related Disorders, Boston University, 648 Beacon St.,
Floor 6, Boston, MA 02114. E-mail: firstname.lastname@example.org.
CLINICAL PSYCHOLOGY: SCIENCE AND PRACTICE•V13 N2, SUMMER 2006180
for state or trait anxiety symptoms (Long & van Stavel,
1995; see also Petruzzello, Landers, Hatfield, Kubitz, &
Salazar, 1991). Likewise, a meta-analysis of 30 experimental
and quasi-experimental studies on the effect of exercise on
depressed mood in nonclinical and clinical populations
indicated an overall mean effect size of ES = .72 (Craft
& Landers, 1998). A large effect size (
to 1.08) was found relative to waitlist control conditions,
with evidence of approximately equal efficacy for treatment
comparison conditions (
= .05; 95% CI: .10 to .21).
Similarly, Lawlor and Hopker (2001) focused more directly
on clinical samples and provided a meta-analysis of 14
randomized trials of exercise interventions for depression
symptoms and found that exercise was significantly more
effective than no treatment (ES = 1.10; 95% CI: .60 to
1.50) and similarly efficacious as psychotherapy (stand-
ardized mean difference = .3; 95% CI:
sizable amount of evidence supporting the benefits of
exercise for mental health in both nonclinical and, more
importantly, clinical populations points toward the further
need to evaluate existing interventions and understand
their mechanism of action for improving mental health.
In the last several years, there have been a number of
additional randomized controlled trials (RCTs) of exer-
cise interventions. Given these advances, one purpose of
this article was to update and refine the Lawlor and
Hopker (2001) meta-analysis by including four recent
RCTs of depression and deleting studies that did not tar-
get clinical levels of depression, were not published in
peer-reviewed journals, or did not provide a nonactive
comparison condition. A second objective was to critically
review the range of applications and the potential mech-
anism of action of exercise interventions in the treatment
of mental disorders. This qualitative review focuses on the
use of exercise interventions in mental disorders in addi-
tion to depression, in order to provide a perspective on
the breadth of efficacy for these interventions. Informa-
tion from both the qualitative and quantitative reviews is
discussed in the context of clinical application of exercise
interventions for depression and other mental disorders.
= .77; 95% CI: .47
0.1 to .7). The
META-ANALYSIS OF CLINICAL STUDIES
Selection of Studies
We selected randomized controlled clinical trials, which
examined the efficacy of exercise treatments for affective
disorders by performing a comprehensive search strategy
in three steps. First, we conducted a computer-based
PsycINFO (1974 to 2005) search from its inception to
the present. The search was performed using
the primary term in combination with the following
mental health, anxiety, panic, generalized
social anxiety, phobia, specific phobia,
posttraumatic stress disorder
sion, bipolar, substance abuse
we performed a computer-based MEDLINE search
(from 1966 to 2005) combined with a Cochrane Central
Register of Controlled Trials search (first quarter of
2005) using similar terms. We limited both the Psyc
INFO and the MEDLINE searches to studies published
in the English language and also to the ones conducted
with human subjects. Third, we cross-referenced our
search results with the references from reviews and
Only studies with a nonactive comparison condition
(e.g., waitlist or placebo treatment, low-level exercise and
health education) were included in the meta-analysis,
and because most studies have targeted affective disorders,
we provide a meta-analytic summary only for these dis-
orders. We also excluded from this summary studies that
target depressive symptoms secondary to other clinical
disorders, such as substance abuse (Palmer, Palmer,
Michiels, & Thigpen, 1995). Of the 17 studies that met
inclusion criteria, three were not included in this meta-
analysis because we were not able to get the necessary
statistical data (Martinsen, Hoffart, & Solberg, 1989;
Martinsen, Medhus, & Sandvik, 1985; Mather et al.,
2002), and two studies were not included because they
only reported the findings at follow-up (Babyak et al.,
2000; Singh, Clements, & Fiatarone Singh, 2001). We
also excluded the study by McCann and Holmes (1984)
as it targeted very low levels of depressive symptoms
(BDI < 11). Trials that did not meet the inclusion criteria
(e.g., studies without posttreatment data, studies that
compared exercise to active treatments, studies that
examined the effects of exercise for eating or substance
use disorders) are reviewed separately in succeeding text
as part of our qualitative analysis.
mood disorder, depres-
Effect Size Calculation
Between-group effect sizes for each study were computed
(Rosenthal, 1991). When the necessary
EXERCISE FOR MENTAL HEALTH META-ANALYSIS•STATHOPOULOU ET AL.181
data were available, all effect sizes were calculated
directly using the following formula:
is the mean of the treatment group,
mean of the comparison group, and
standard deviation. If these data were not provided,
estimated using conversion equations for significance
; see Rosenthal, 1991). All effect sizes
were corrected for small sample sizes according to
Hedges and Olin (1985). Hedges’
puted from Cohen’s
with the following formula:
is the pooled
may also be com-
Therefore, a smaller sample size reduces the estimated
effect size. These controlled effect sizes may then be
conservatively interpreted with Cohen’s (1988) conven-
tion of small (.2), medium (.5), and large (.8) effects. The
overall mean effect size for all of the studies com-
bined was computed using the following formula:
is the weight for each study
is the effect size for each study. For purposes of
comparisons between studies (where alternative effect
size metrics were used), we provide both Hedges’
in the text.
Exercise interventions can be characterized in terms of
frequency, duration, intensity, and total energy expenditure,
and each one of these dimensions must be considered in
the definition of exercise characteristics. One study in our
meta-analysis failed to provide a precise description of the
exercise interventions (Veale et al., 1992). Among studies
providing this information, the frequency of exercise varied
from twice (Klein et al., 1985) to four times a week (Doyne
et al., 1987), duration from 20 (Fremont & Craighead,
1987) to 45 min (Bosscher, 1993; Klein et al., 1985), and
the intensity from being unspecified (Klein et al., 1985) to
70–85% of the maximum heart rate (Bosscher, 1993).
Treatment Effect Size
Figure 1 displays a forest plot of the bias corrected
) between-group (controlled) effect sizes and
95% confidence intervals for each study. Using a random
effects analysis, we obtained a mean overall effect size
= 1.39 (95% CI: .89 to 1.88), corresponding to a
= 1.42 (95% CI: .92 to 1.93), indicating a very large
effect for exercise interventions relative to control condi-
tions. This random effects combined effect size estimate
is labeled “Random combined” in Figure 1. Across studies,
the most common control condition was no treatment/
waitlist, accounting for five (Armstrong & Edwards, 2003;
Doyne et al., 1987; McNeil, LeBlanc, & Joyner, 1991;
Pinchasov, Shurgaja, Grischin, & Putilov, 2000; Veale
et al., 1992) out of 11 studies (see Table 1). For three
studies, lower level exercise was used as the control con-
dition (Bosscher, 1993; Dunn, Trivedi, Kampert, Clark, &
Chambliss 2005; Sexton, Maere, & Dahl, 1989), and of
the remaining studies, one used meditation/relaxation
(Klein et al., 1985), one used health education (Singh,
Clements, & Fiatarone, 1997), and one used treatment as
usual (Singh et al., 2005) as control conditions. Effect
sizes tended to be larger when compared to waitlist/
no-treatment conditions (mean
compared to the other control conditions (mean
= 6), consistent with results from meta-analyses
of other treatments (e.g., Gould, Otto, & Pollack, 1995).
= 5) as
Three studies included in the meta-analysis (Armstrong
& Edwards, 2003; Doyne et al., 1987; Pinchasov et al.,
2000) did not provide dropout rates. The mean dropout
rate for the exercise intervention in the eight remaining
clinical trials in Table 1 was 19.9%. This rate was identical
to the control condition and compares well to meta-
analytic reviews of antidepressant treatments of depres-
sion. For example, in their meta-analysis of depression
studies in primary care, MacGillivray et al. (2003)
reported an overall dropout rate of 20.7% for selective
Effect sizes for studies of depression.
= Σ Σ
w g w
CLINICAL PSYCHOLOGY: SCIENCE AND PRACTICE•V13 N2, SUMMER 2006182
serotonin reuptake inhibitors (SSRIs) and 27.9% for
tricyclic antidepressants (TCAs). Slightly higher rates,
27% for SSRIs and 31.4% for tricyclics, were reported
for a combined sample of 10,553 patients being treated
in controlled trials for depression, regardless of setting.
Associations Between Sample Size, Effect Size,
and Publication Year
There was no significant relationship between sample
size and effect size. There was also no significant relation-
ship between publication year and sample size. However,
there was a significant association between publication
year and effect size (
< .043); recent studies were associated with higher
effect sizes (see Figure 1).
= 5.43, d.f. = 1, 11,
Publication Bias (The File Drawer Problem)
A number of investigators have drawn attention to the
potential discrepancy between the number of trials com-
pleted and the number of trials published (Bakan, 1967;
McNemar, 1960; Smart, 1964; Sterling, 1959). If studies
are not published because the results were not significant,
a meta-analysis of published studies may overestimate the
overall effect size. Rosenthal (1991) and others have
called this confound “The File Drawer Problem.” A con-
servative method of addressing this problem is to assume
that the effect sizes of all current or future unpublished
studies are equal to 0 and compute the number of such
studies it would require to reduce the overall effect size
to a nonsignificant level (Rosenthal & Rubin, 1988).
Using the guidelines put forth by Rosenthal (1991),
we determined the findings robust if the required
number of studies to reduce the overall effects size to a
nonsignificant level exceeded 70. Analyses revealed that
it would require 367 current or future unpublished stud-
ies with an effect size of 0 to bring the overall effect size
of exercise for depression within the nonsignificant
range, indicating that findings to date are robust.
Recent studies on the dose–response effects of physical
activity on outcomes of depression (Dunn et al., 2005,
Singh et al., 2005) showed that high-intensity exercise
provides superior results compared to lower intensity
exercise. A study conducted by Dunn et al. (2005) showed
that the public health recommended dose of aerobic
exercise (total energy expenditure of 17.5 kcal/kg/week)
conducted five times a week is more effective in reduc-
ing depression when compared to low-dose exercise (total
energy expenditure 7.0 kcal/kg/week) conducted three
times a week, or flexibility exercise conducted three times
a week. Therefore, the Public Health dose/five-times-a-
week condition provided the greatest response rate
(44% response rate versus 23% of the control group);
Exercise interventions in the treatment of depression (RCTs with nonactive control condition)
Dunn et al. (2005)
Singh et al. (2005)
Pinchasov et al. (2000)
Singh et al. (1997)
Veale et al. (1992)
McNeil et al. (1991)
Sexton et al. (1989)
Care as usual
Older depressed adults
mixed depression (47%),
anxiety (39%), and
other diagnoses (14%)
n = 80
n = 60
n = 19
n = 18
n = 32
n = 24
n = 83
n = 30
n = 53
Doyne et al. (1987)
Klein et al. (1985)
n = 40
n = 74
*Bias corrected (Hedges).
BAI = Beck Anxiety Inventory; PD = panic disorder; PHD = Public Health dose; HRSD = Hamilton Rating Scale for Depression; HIGH PRT = high-intensity
progressive resistance training; EPDS = Edinburgh Postnatal Depression Score; PRT = Progressive Resistance Treatment; BDI = Beck Depression Inventory;
SDS = Self-Rating Depression Scale; SCL = Symptom Checklist Revised.
EXERCISE FOR MENTAL HEALTH META-ANALYSIS•STATHOPOULOU ET AL.183
whereas the Public Health dose/three-times-a-week
condition provided the greatest remission rate (41%
remission rate versus 15% of the control group). Singh
et al. (2005) found similar results in a dose–response
study of weight-training exercise for the treatment of
clinical depression in older adults (> 60 years). The study
showed that high-intensity progressive resistance training
(PRT) (80% maximum load) is more effective in treating
depression in the above population than lower intensity
PRT (20% maximum load). As evidenced by greater
response rates, high-intensity outperformed low-intensity
training, which did not differ significantly from usual
care (61%, 29%, and 21%, respectively). High-intensity
PRT was also associated with more improvement in the
vitality quality of life scale and higher improvement in
sleep quality than the lower intensity PRT.
Aerobic Versus Anaerobic Exercise
Only few studies have compared the effects of aerobic
exercise to the effects of anaerobic exercise for the treat-
ment of depression (e.g., Doyne et al., 1987; Martinsen
et al., 1989). Doyne et al. (1987) randomly assigned 40
depressed women with major depressive disorder to one
of three conditions: (a) running at 80% of estimated
maximum heart rate, (b) weight lifting at less than 50%
of estimated maximum heart rate, or (c) waitlist control.
Participants in the exercise programs were instructed to
exercise four times a week for eight weeks. Results at
posttreatment and follow-up assessment (one-, seven-,
and 12-month follow-up) revealed that the two exercise
conditions achieved comparable improvements in fitness
Martinsen et al. (1989) randomly assigned 90 partici-
pants suffering from depression to either group-based
aerobic exercise training at 70% of the maximum aerobic
capacity or group-based nonaerobic exercise training
consisting of low-intensity muscle strength training. Both
programs comprised 24 one-hour sessions, which were
completed across an eight-week period. Despite greater
improvements in physical fitness observed among partici-
pants assigned to the aerobic exercise group, there were no
significant differences in the rate of pre- to posttreatment
improvement in depression between the two conditions.
These preliminary findings indicate that anaerobic
exercise may be equally effective in reducing depressive
symptoms compared to aerobic activity and suggest that
resistance training may be an alternative for patients for
whom aerobic activity may be inappropriate or for those
who do not have the initial motivation for aerobic activ-
ity. In addition, these results suggest that the decrease in
depressive symptoms associated with exercise may not be
due to increases in fitness. Further research on exercise
training parameters (i.e., exercise type, intensity, frequency)
will advance knowledge about the mechanism by which
exercise exerts its effect on emotional well-being, and
will guide clinicians in providing optimal prescriptions
for exercise training.
Exercise as an Adjunctive Treatment for Depression
In addition to its application as a monotherapy, exercise
interventions may have efficacy as part of combination
treatment strategies. Four studies have investigated the
efficacy of using exercise as adjunctive treatment to
pharmacotherapy for depression. In an early study,
Martinsen et al. (1985) found that the combination of
exercise with TCAs was not more effective than exercise
alone. In a second study utilizing a similar design,
Martinsen et al. (1989) compared aerobic to nonaerobic
forms of exercise in the treatment of 99 inpatients meet-
ing criteria for major depression, dysthymia, and depres-
sive disorder not otherwise specified (NOS). Fourteen
patients in each group were administered antidepressants
(TCAs) during the study. Patients in the combined treat-
ment (exercise plus pharmacotherapy) showed trends
toward a larger reduction in depressive symptoms compared
to patients who were treated with exercise as monotherapy,
but the difference was not statistically significant.
However, the nonrandomization of participants to
medication does not allow for definite conclusions.
Blumenthal et al. (1999) examined the effects of
pharmacotherapy with sertraline, aerobic exercise, and
the combination of these two treatments among 156
clinically depressed older adults between the ages of 50
and 77 years. Although all three groups showed statisti-
cally and clinically significant reductions in depression
after 16 weeks of treatment (with no statistically signifi-
cant difference in either Beck Depression Inventory
[BDI] or Hamilton Rating Scale for Depression [HAM-D]
scores between these groups), pharmacotherapy was
associated with the most rapid therapeutic response. At
six-month follow-up (Babyak et al., 2000), participants
in the exercise group were significantly more likely than
CLINICAL PSYCHOLOGY: SCIENCE AND PRACTICE•V13 N2, SUMMER 2006184
those in the medication group to reach partial or full
recovery status (odds ratio = 6.10,
of the follow-up period, antidepressant medications were
being used by 26% of patients originally randomized to
medication, 40% of patients randomized to combination
treatment, and 7% of patients randomized to exercise
alone. Moreover, 38% of the remitted participants in the
medication group and 31% in the combination group
had relapsed, compared to only 8% in the exercise group.
Interestingly, the results for the exercise-alone group at
follow-up were stronger than those for the combination-
treatment group, which may have reflected the impact of
medication discontinuation on symptom levels.
Mather et al. (2002) examined the efficacy of exercise
interventions as an adjunct to outpatient pharmaco-
therapy for depression. A total of 86 older adults (aged 53
to 78 years) were randomized to an exercise treatment
(weight-bearing exercise performed to music) or to a
social control group (health education talks). All patients
continued to take antidepressant medications. After 10
weeks of treatment, 55% of those who received adjunc-
tive exercise met criteria for treatment response (a 30%
decrease on the HAM-D) compared to 33% in the
In summary, across four studies of the role of adjunctive
exercise in the pharmacologic treatment of depression,
there is evidence, particularly from more recent and
better controlled trials, for strong beneficial effects. In
contrast to studies of pharmacotherapy, there has been
only limited attention to the combination of exercise
with empirically supported psychosocial approaches
(e.g., cognitive-behavioral therapy [CBT], interpersonal
therapy) for depression. Our review identified only one
study. In this study, Fremont and Craighead (1987) rand-
omized patients with mild to moderate depression to
receive either 10 sessions of cognitive therapy, 10 sessions
of cognitive therapy combined with running, or 10 ses-
sions of running only. No significant additive effects for
exercise were found relative to CBT alone, with effect
= .27) indicating only a subtle advantage for
running. The failure to achieve a noteworthy additive
effect for the addition of exercise to cognitive therapy is
reminiscent of previous research, indicating that the
combination of cognitive and behavioral strategies for
depression often has little advantage over a single modality
alone (e.g., Miller & Berman, 1983).
= .01). By the end
QUALITATIVE REVIEW OF EXERCISE INTERVENTIONS FOR
SUBSTANCE ABUSE, EATING DISORDERS, AND ANXIETY
As is evident from our quantitative review, most systematic
research pertaining to the application of exercise for the
treatment of mental health has been conducted with patients
suffering from depressive disorders. Despite the scarcity
of well-controlled studies, evidence to date suggests that
patients suffering from other forms of psychopathology
may also derive significant benefit from exercise inter-
ventions. These early reports on other clinical populations
hold important information on the effects of exercise on
a wide range of symptoms, including cravings involved
in alcohol abuse (Ermalinski, Hanson, Lubin, Thornby,
& Nahormek, 1997), specific thoughts associated with
bulimia (Sundgot-Borgen, Rosenvinge, Bahr, & Schneider,
2002), as well as anxiety-related symptoms (Broocks et al.,
1998). Examining these interventions in the context of
different disorders may help clarify potential mechanisms
of change, as well as the breadth of action of exercise effects.
Evidence underscoring the potential benefit of exercise for
alcohol abuse comes from different lines of research. First,
cross-sectional data reveal a strong inverse relationship
between fitness and risk factors of substance abuse as well
as usage patterns among young people (Collingwood,
Sunderlin, Reynolds, & Kohl, 2000). Second, quasi-
experimental studies have shown that exercise in con-
junction with inpatient alcohol rehabilitation treatment
is associated with significantly lower alcohol cravings
(Ermalinski et al., 1997) as well as lower levels of anxiety
and depression (Palmer, Vacc, & Epstein, 1988). Moreover,
follow-up abstinence rates were significantly higher among
alcoholics who participated in a fitness program in conjunction
with treatment for alcohol abuse in comparison to alcoholics
in a regular alcohol abuse treatment (Sinyor, Schwartz, Per-
onnet, Brisson, & Serganian, 1983). Finally, randomized
controlled data demonstrated that relative to waitlist, an
eight-week program of running exercise resulted in a sig-
nificant reduction in alcohol consumption among nonclinical
problem drinkers (Murphy, Pagano, & Marlatt, 1986).
Sundgot-Borgen et al. (2002) examined the efficacy of
exercise, CBT, and nutritional counseling in the treatment
EXERCISE FOR MENTAL HEALTH META-ANALYSIS•STATHOPOULOU ET AL.185
of bulimia nervosa. Sixty-four normal-weight female
bulimics (18–29 years old) were randomly assigned to
= 15), CBT (
= 17), or a waitlist control group (
showed that exercise was more effective than CBT in
reducing drive for thinness, bulimic symptoms (both binge
eating and vomiting), and body dissatisfaction. In regard
to the frequency of bulimic symptoms, the exercise group
showed greater improvement than the CBT group.
This evidence is complemented by a number of quasi-
experimental and randomized controlled studies. More
specifically, in a nonrandomized study, Calogero and
Pedrotty (2004) examined the effectiveness of an exercise
program designed to reduce exercise abuse in women
receiving residential treatment for eating disorders. One
hundred and twenty-seven eating-disordered women who
participated in the exercise program were compared to
127 eating-disordered women not participating in an
exercise program. The results showed that women in the
exercise group significantly reduced obligatory attitudes
toward exercise relative to the comparison group. Also,
anorexic women in the exercise group gained one-third
more weight than those in the comparison group. These
findings suggest that the use of an exercise intervention
that targets exercise abuse in women with eating disorders
is feasible during residential treatment and may offer
benefits without interfering with weight gain.
The incremental effect of exercise to CBT treatments
of binge eating disorder has been examined in two rand-
omized control trials (Fossati et al., 2004; Pendleton,
Goodrick, Poston, Reeves, & Foreyt, 2002). Fossati
et al. (2004) provided support for the adjunctive use of
exercise in CBT for binge eating disorder employing a
3-cell randomized design. The study compared CBT to
(a) CBT combined with information about proper
nutrition and (b) CBT combined with increased physical
activity. Although the addition of nutritional education
resulted in greater weight loss relative to CBT alone,
only the combination of physical activity training and
nutritional education facilitated the effects of CBT on
depressed mood and anxiety, reflecting an advantage of
= 1.63 relative to CBT alone.
In another multicell design, Pendleton et al. (2002)
randomly assigned 114 obese female binge eaters in four
groups: CBT with exercise and maintenance (CBT-EM),
CBT with exercise (CBT-E), CBT with maintenance
= 16), nutritional advice
= 16). The study
(CBT-M), and CBT only. The CBT treatment employed
in the study was based on the manualized treatment for
binge eating disorder by Telch, Agras, Rossiter, Wilfley,
& Kenardy (1990). The maintenance component con-
sisted of 12 biweekly sessions of the randomized treat-
ment over the final six months of the study. Although at
four months and 10 months of the study the CBT-EM
yielded a higher abstinence rate (e.g., four month rate:
67% versus 50% for CBT-E, 22% for CBT-M, and 41%
for CBT alone), at the end of the 16-month study
period, the CBT-E had a similarly high abstinence rate
(65%) to CBT-EM (58%), followed by the CBT-M (39%),
and 18% for the CBT-only group. Across groups, 61% of
the participants who exercised maintained abstinence
from binging at 16 months compared to 30% of the
nonexercisers. Also, exercisers had significantly greater
reductions in binge eating days compared to nonexercisers
at each follow-up point. In sum, both studies by Fossati
et al. (2004) and Pendleton et al. (2002) showed a signifi-
cant incremental effect of exercise on CBT treatments
for binge eating disorder.
Broocks et al. (1998) compared the relative efficacy of
exercise, pharmacotherapy with clomipramine, and pill
placebo in a randomized controlled study of 46 patients
with panic disorder with or without agoraphobia. Results
from this study suggest that clomipramine improved
anxiety symptoms significantly earlier than exercise.
However, by the end of 10 weeks of treatment, exercise
was as effective as clomipramine on all primary outcome
measures (e.g., Hamilton Anxiety Scale, Bandelow Panic
and Agoraphobia—Observer Rated Scale, Bandelow
Panic and Agoraphobia—Patient Rated Scale), with the
exception of the clinician-rated global impressions of
severity (Broocks et al., 1998). Both active treatments
were superior to placebo on all primary outcome
measures by the end of 10 weeks of treatment.
Additional evidence for the potential benefits of
exercise for the treatment of panic disorder comes from
studies of anxiety sensitivity, or the tendency to respond
fearfully to anxiety-related bodily sensations (McNally,
2002). Research has implicated anxiety sensitivity as a
risk factor (McNally, 2002) as well as a maintaining
factor for panic disorder (McNally, 2002; Smits, Powers,
Cho, & Telch, 2004). Findings from preliminary studies
CLINICAL PSYCHOLOGY: SCIENCE AND PRACTICE•V13 N2, SUMMER 2006186
suggest that exercise programs can indeed produce
marked changes in anxiety sensitivity (Broman-Fulks,
Berman, Rabian, & Webster, 2004; Smits, Otto, Powers,
& Utschig, 2005). As such, exercise interventions can be
conceptualized as an interoceptive (internal sensation)
exposure procedure, as it provides repeated confronta-
tion with feared bodily sensations (e.g., racing heart,
rapid breathing) in the absence of anticipated negative
consequences (Smits, Powers, Berry, & Otto, 2006). This
feature of exercise may hold benefit for the treatment of
panic disorder, but needs to be considered in the context
of broader conceptualizations of the potential mechanisms
of action of exercise.
THE MECHANISM OF ACTION OF EXERCISE INTERVENTIONS
There is little research on the mechanisms by which
exercise exerts its positive effects on mental health.
Hypothesized mechanisms of action fall into two broad
categories: physiological and psychological factors. Pro-
posed physiological mediators of the relationship
between exercise and mental health include changes in
the metabolism and availability of central neurotrans-
mitters (e.g., serotonin, endogenous opioids) and sleep
regulation. Psychological mediational hypotheses that
have received some empirical attention include changes
in coping self-efficacy and the interruption of negative
thoughts. In the next section, we discuss the preliminary
findings pertaining to these hypotheses.
Broocks and colleagues (Broocks et al., 1999, 2001,
2003) have hypothesized that the mood- and anxiety-
altering effects of physical activity may be accounted for
by a modification of serotonin (5-HT) function. Based
on animal studies showing that physical activity leads to
increased serotonin turnover (Broocks, Schweiger, &
Pirke, 1991), they posit that physical activity may lead to
an adaptive down-regulation of postsynaptic serotonin
receptors, and specifically the 5-HT2C receptors. Sup-
port for their hypothesis comes from a neuroendocrine
challenge study in which they compared marathon
runners to sedentary controls on their responses to meta-
chlorophenylpiperazine (m-CPP), a 5-HT agonist that
produces anxiogenic symptoms via 5-HT2C receptors
(Broocks et al., 1999). They found that, relative to sed-
entary participants, marathon runners showed a diminished
cortisol response to m-CPP. This reduction in cortisol
response among exercising participants suggests a reduced
hormonal reaction to m-CPP mediated by postsynaptic
5-HT2C receptors, thus suggesting that anxiolytic and
antidepressant effects of exercise may be mediated by the
down-regulation of 5-HT2C receptors.
Based on the observation that physical activity causes
a release of endogenous opioids (Morgan, 1985; Ransford,
1982), it has been hypothesized that the inhibitory
effects of beta-endorphins on the central nervous system
are in part responsible for antidepressant and anxiolytic
effects of exercise (Thoren, Floras, Hoffman, & Seals, 1990).
Indeed, several studies have shown that participants who
engage in aerobic exercise are more likely to be calmer
and less depressed than those participants who did not
engage in exercise following administration of an opiate
antagonist (Allen & Coen, 1987; Daniel, Martin, & Carter,
1992; Janal, Colt, Clark, & Glusman, 1984). Markoff,
Ryan, and Young (1982), however, failed to find this effect.
The question whether changes in central neurotransmitter
function account for the therapeutic effects of exercise
awaits further examination in clinical samples.
Sleep and Social Zeitgeber Theory
Both antidepressant therapy and exercise are associated
with improvements in the sleep cycle, a physiological
disturbance commonly observed in depression and anxi-
ety (Casper et al., 1994; Driver & Taylor, 2000; Kubitz,
Landers, Petruzzello, & Han, 1996; Tanaka & Shirakawa,
2004). Indeed, evidence suggests that certain types of
antidepressants exert their effect through a normalization
of the sleep cycle (Vogel, 1983). Collectively, these find-
ings suggest that the observed improvement in quality of
sleep following exercise may in part account for the pos-
itive mental health effects of exercise. This hypothesis is
in line with the social zeitgeber theory (Ehlers, Frank, &
Kupfer, 1988; Ehlers, Kupfer, Frank, & Monk, 1993),
which posits that activities such as exercise may improve the
regularity of daily lifestyles, which in turn improves mood.
To test this hypothesis, Singh et al. (1997) randomly
allocated depressed participants to either a 10-week
resistance exercise program or control. Participants in the
exercise condition showed significantly greater improvements
in subjective sleep quality measured by the Pittsburgh
Sleep Quality Index. Furthermore, the improvements in
sleep quality corresponded with improvements in the
EXERCISE FOR MENTAL HEALTH META-ANALYSIS•STATHOPOULOU ET AL.187
BDI. As the design of the study precludes the investiga-
tors from making inferences regarding temporality, it
remains unclear whether the improvements in sleep
quality affect depression or vice versa.
Craft (2005) proposed that engaging in an exercise pro-
gram may provide the mastery experiences needed to
promote the perceived ability to cope with depression, and
thereby, improve mood. Employing a quasi-experimental
design, Craft (2005) found that women who exercised
showed significantly higher coping self-efficacy and
lower depression scores compared to nonexercising
controls. An increase in perceived coping ability and a
decrease in anxiety have also been observed in anxious
individuals who initiated an exercise program (Steptoe,
Edwards, Moses, & Mathews, 1989). Similarly, Bodin
and Martinsen (2004) found that exercise that targeted
self-efficacy (e.g., 45 min of martial arts) corresponded
with significantly greater improvements in positive affect
and state anxiety compared to exercise that did not target
self-efficacy (e.g., 45 min of stationary bike exercise).
Instead of directly increasing a sense of coping ability,
physical exercise may alter the accessibility or intensity of
ruminations, worries, and anxiety (Bahrke & Morgan,
1978; Leith, 1994). Craft (2005) found that exercise was
associated with greater use of distraction techniques
early in treatment and a decrease in rumination through-
out the course of treatment. However, there was no sig-
nificant association between distraction and depression
in this study. A possible explanation for these results
comes from a study by Goode and Roth (1993), who
found that it is not distraction per se but the content
of the distraction techniques in which people engage
that is associated with changes in emotional well-being.
More specifically, they reported that runners who
focused on nonassociative thoughts (those not relating
to exercising) showed less fatigue and, in some cases,
decreases in tension and anxiety, compared to runners
who focused on associative thoughts (monitoring the
body and the exercise itself). The authors suggest that the
effect could be described in terms of a stress-and-coping
perspective, where physical activity improves mood
because it evokes physiological release while allowing the
mind to engage in anxiolytic thoughts (Goode & Roth,
1993). This hypothesis receives interesting but circumstantial
support from recent neuropsychological studies. Affective
disorders have been associated with hyperactivity in select
prefrontal areas (Mayberg, 1997; Baxter et al., 1987),
and several EEG studies provide evidence that exercise
reduces prefrontal activation (for a review, see Kubitz &
Pothakos, 1997). Also, Dietrich and Sparling (2004),
using neuropsychological tests (the Wisconsin Card Sort-
ing Task, the Brief Kaufman Intelligence Test, the Paced
Auditory Serial Addition Task, and the Peabody Picture
Vocabulary Test), recently demonstrated that during
moderate exercise, prefrontal-dependent cognitive func-
tion becomes impaired while prefrontal-nondependent
cognitive processes remain intact. Accordingly, one
potential mechanism of the anxiolytic and antidepressant
effects of exercise may be through the modification of
prefrontal cognitive processes (Dietrich & Sparling, 2004),
although this accounting does not yet take into account
the complexity of frontal circuits (e.g., left versus right
hemisphere activation, activation in the orbitofrontal versus
dorsolatoral prefrontal contex) in relation to emotional
regulation and mood disturbances (Siegle, Ghinassi, &
Thase, in press).
Modification of Action Tendencies
As a therapeutic technique, exercise may also serve to
modify emotional action tendencies—one of the “fun-
damental” therapeutic strategies according to Barlow,
Allen, and Choate (2004). More specifically, exercise
involves an action (i.e., activation, approach) that is
inconsistent with the natural action tendencies associated
with depression (passivity) and anxiety (avoidance). As
Barlow et al. (2004) indicate, therapeutic packages that
emphasize the replacement of action tendencies with
actions consistent with alternative emotions have shown
efficacy for a variety of disorders, including depression,
anxiety disorders, and borderline personality disorder. In
this context, it is useful to consider the ways in which
exercise interventions are similar to behavioral activation
treatments for depression (for review, see Hopko, Lejeuz,
Ruggiero, & Eifert, 2003; Jacobson, Martell, & Dimid-
jian, 2001). These approaches emphasize the role of
depressive behaviors, such as inactivity and withdrawal,
as maladaptive coping strategies that isolate individuals
from opportunities for positive affect and propagate
depression. Accordingly, behavioral activation treatments
are focused on helping patients re-establish adaptive
CLINICAL PSYCHOLOGY: SCIENCE AND PRACTICE•V13 N2, SUMMER 2006188
positive activities, and have been shown effective for the
acute treatment of depression and relapse prevention
(e.g., Gortner, Gollan, Dobson, & Jacobson, 1998;
Jacobson et al., 1996). Treatments differ with respect to
the degree core behavioral activation interventions are
supplemented by training in other skills such as contin-
gency management, functional analyses, problem solving,
assertiveness, relaxation, or mindfulness (Hopko et al.,
2003). As such, exercise interventions represent only a
portion of behavioral activation interventions and are
limited to the promotion of adaptive action (exercise)
despite the presence of symptoms. Nonetheless, this role
may have fairly broad effects. We have already discussed
the role of exercise in reducing fears and avoidance of
anxiety-related sensations (anxiety sensitivity). This sort
of promotion of comfort with somatic sensations may
have broad effects; anxiety sensitivity has shown itself
to be a useful predictor in areas well outside the anxiety
disorders; as applied as an index of emotional/somatic
intolerance, anxiety sensitivity has been found to be a
useful predictor of the ability to tolerate withdrawal
sensations and maintain abstinence following smoking
cessation (Brown, Lejuez, Kahler, Strong, & Zvolensky,
2005). Accordingly, exercise may have a broader role in
promoting emotional/somatic tolerance in addition to
promoting adaptive action despite the presence of
SUMMARY AND CLINICAL APPLICATIONS
Research on the impact of exercise interventions on
mental health has progressed from initial efficacy studies
to well-designed clinical trials. Our meta-analysis of 11
well-controlled studies yielded a large combined effect
= 1.39 (95% CI: .89 to 1.88), indicating that
exercise can be a powerful intervention for clinical
depression. Given there is no evidence that comparison
conditions are becoming weaker over time (e.g., Walsh,
Seidman, Sysko, & Gould, 2002), the increase in con-
trolled effect sizes over time may indicate that study
designs are improving and reducing error variance,
exercise interventions themselves are growing more
sophisticated and sound over time, and/or these inter-
ventions are being applied to populations for whom they
are particularly beneficial. Results to date provide most
evidence for the treatment of depression, with these
studies enrolling patients with the mean level of
depression in the moderate range. Results from the current
meta-analysis suggest that exercise interventions could be
presented to patients as another strategy for the treat-
ment of depression, along with other empirically sup-
ported treatments (ESTs) such as CBT, interpersonal
therapy, or pharmacotherapy.
The consistent evidence for the benefits offered by
exercise interventions for depression, plus the promising
preliminary evidence for efficacy in anxiety, eating and
substance use disorders, encourages further attention to
these interventions by mental health professionals (see
also Richardson et al., 2005). However, we recognize
this is a challenging proposal on several fronts. First, the
adoption of ESTs in the United States continues to
occur at woeful levels (e.g., Goisman, Warshaw, & Keller,
1999). Second, exercise interventions differ substantially
of interventions in psychological treat-
ments. Indeed, exercise interventions may find easiest
adoption among clinicians who already utilize other
empirically supported approaches (i.e., CBT or pharma-
cotherapy), primarily because it is these approaches that
utilize specific out-of-session assignments (e.g., behavioral
assignments or pill taking, respectively) as part of treat-
ment. As we reviewed previously, application of exercise
as an adjunctive strategy to medication has been met with
success (e.g., Babyak et al., 2000; Mather et al., 2003),
while similar evidence for successful application of exercise
as an adjunctive strategy to CBT has been shown in the
treatment of binge eating disorder (Fossati et al., 2004;
Pendleton et al., 2001), but is awaiting further evaluation
in the treatment of depression. Nonetheless, the role of
exercise in the care of any one patient will need to be evalu-
ated relative to the needs and expectations of that patient.
Clinicians will need to decide whether promotion of
exercise is consistent with the patient’s goals and values,
can be offered and monitored without compromising
other core aspects of treatment, and is sequenced so that
exercise interventions can be targeted to relevant goals
(e.g., as an adjunctive strategy for stress management,
depression management, well-being enhancement, or
relapse prevention.). Moreover, it is important to
remember that exercise has a wide variety of health ben-
efits outside the realm of mental health (Hu et al., 2004;
Mokdad, Marks, & Stroup, 2004). Accordingly, exercise
should be considered a particularly broad-based inter-
vention with general health-promoting benefits.
EXERCISE FOR MENTAL HEALTH META-ANALYSIS•STATHOPOULOU ET AL.189
Third, clinicians will need to attend to providing an
adequate match between patient expectations and treat-
ment interventions, given the importance of this match
for treatment adherence (Eisenthal, Emery, Lazare, &
Udin, 1979; Grilo et al., 1998; Schulberg, Magruder, &
deGruy, 1996). Indeed, like controlled trials of pharmaco-
therapy or CBT, research on exercise interventions
has been limited to individuals who were willing to
consider this treatment (i.e., at least as part of the processes
of consent to join a randomized trial involving exercise).
In encouraging the role of exercise in clinical prac-
tice, we are not suggesting that a focus on exercise
should replace attention to the range of other treatment
strategies and psychosocial variables implicated in models
of psychiatric disorders. Instead, we are asking clinicians
to evaluate how exercise interventions might be inte-
grated into a well-formulated model of care for an
individual patient, with attention to the broader health
promotion and mood regulation that might be offered
by these interventions. It is important to remember that
previous studies have shown efficacy from simple recom-
mendations of an exercise schedule with monitoring
across sessions (e.g., Broocks et al., 1998). Alternatively,
clinicians may simply want to refer their patients to
qualified personal trainers or other exercise or sports-
medicine professionals as part of independent, adjunctive
If a clinician selects exercise as an adjunctive strategy
to other interventions, guidance on the dosing of exer-
cise is provided by the results reported by Dunn et al.
(2005). Nonetheless, as evident from our meta-analysis,
benefit has been provided by exercise interventions at
other dosing levels. Detailed information on the pre-
scription of exercise is provided by the American
College of Sports Medicine (2005). In addition, Meyer
and Broocks (2000) have provided guidelines for exercise
testing and prescription for patients with psychiatric
disorders. In brief, their guidelines for implementing
exercise training include the following. First, screening is
important to ensure that patients are free from cardio-
vascular and acute infectious diseases. Second, selection of
a training mode must take into account gender, age, exer-
cise experiences, and location-specific exercise circum-
stances. Third, workout sessions of moderate to high
intensity should last up to 30 min. Such sessions should
be completed twice to four times a week if improvement
of endurance capacity is one of the program aims. Fourth,
patients should use portable heart rate devices to moni-
tor exercise intensity during these sessions. Fifth, to
increase program compliance, intense sessions during the
first week of the program may be substituted with easier
forms of physical activity, such as walking. Sixth, pro-
grams should be at least four weeks in length in order for
patients to adopt an exercise training habit. Meyer and
Broocks (2000) also recommend posttraining evaluation of
fitness, and evaluation of clinical improvement after 10 to
12 weeks, with a subsequent decision about the appro-
priateness of continuation of the exercise program and
establishment of appropriate home-based exercise supports.
In addition to these guidelines, we recommend medical
clearance for exercise with evaluation of the general,
cardiovascular, and joint fitness of individuals relative to
the dosing of exercise to be selected. This is likely to be
an area of easy collaboration with primary care and other
physicians, given evidence that the majority of physicians
consider physical activity as an intervention for anxiety
and depression (60% and 85%, respectively, according to a
survey of 1,750 physicians; Dishman, 1986).
When considering exercise as an intervention, clini-
cians will need to plan for the maintenance of these
health behaviors. The available evidence suggests that
adherence to short-term prescribed exercise is approxi-
mately 70% (Fentem, 1994), but drops to 50% within six
months (Dishman, 1991, 1994). Because many physi-
cians may not be familiar with specific aerobic exercise
recommendations, and therefore may fail to offer spe-
cific exercise prescriptions (Williford, Barfield, Lazenby,
& Olson, 1992), the prescription of a structured and
focused program of exercise may be most helpful. Indeed,
in a review of the literature on activity counseling,
Eakin, Glasgow, and Riley (2000) concluded that success
rates may be enhanced when practitioners provide writ-
ten materials that not only indicate treatment dose, but
also suggestions for potentially anticipated barriers.
Corroborating evidence comes from studies that have
found that the inclusion of strategies such as prestudy
educational programs, ongoing motivational support,
and making exercise more convenient facilitate the
effects of activity counseling (Gorin et al., 2005; Jakicic,
Winters, Lang, & Wing, 1999).
In summary, an increasing number of outcome studies
provide evidence for the efficacy of exercise interventions
CLINICAL PSYCHOLOGY: SCIENCE AND PRACTICE•V13 N2, SUMMER 2006190
for clinical samples. This evidence joins a wealth of pre-
vious studies documenting the positive mood effects of
exercise. Given this evidence, we join other recent efforts
(Richardson et al., 2005) to encourage mental health
clinicians to consider adding exercise interventions to their
armamentarium of strategies to help patients in distress.
Allen, M. E., & Coen, D. (1987). Naloxone blocking of
running-induced mood changes.
American College of Sports Medicine. (2005).
lines for exercise testing and prescription
Lippincott Williams & Wilkins.
Armstrong, K., & Edwards, H. (2003). The effects of exercise
and social support on mothers reporting depressive
symptoms: A pilot randomized controlled trial.
Journal of Mental Health Nursing
Babyak, M., Blumenthal, J. A., Herman, S., Khatri, P.,
Doraiswamy, M., Moore, K., et al. (2000). Exercise treatment
for major depression: Maintenance of therapeutic benefit at
Bakan, D. (1967).
San Francisco: Jossey-Bass.
Bahrke, M. S., & Morgan, W. P. (1978). Anxiety reduction
following exercise and meditation.
Barlow, D. H., Allen, L. B., & Choate, M. L. (2004). Toward a
unified treatment for emotional disorders.
Baxter, L. R., Phelps, M. E., Mazziota, J. C., Guze, B. H.,
Schwartz, J. M., & Selin, C. E. (1987). Local celebral
glucose metabolic rates in obsessive-compulsive disorder: A
comparison with rates in unipolar depression and normal
Archives of General Psychiatry
Biddle, S. J. H. (2000). Emotion, mood and physical activity.
In S. J. H. Biddle, K. R. Fox, & S. H. Boutcher (Eds.),
Physical activity and psychological well-being
Blumenthal, J. A., Babyak, M. A., Moore, K. A., Craighead,
W. E., Herman, S., Khatri, P., et al. (1999). Effects of exercise
training on older patients with major depression.
Bodin, T., & Martinsen, E. W. (2004). Mood and self-efficacy
during acute exercise in clinical depression: A randomized,
Journal of Sport & Exercise Psychology
Bosscher, R. J. (1993). Running and mixed physical exercise
with depressed psychiatric patients. International Journal
of Sport Psychology, 24, 170–184.
Annals of Sports Medicine
(6th ed.). Philadelphia:
Cognitive Therapy and
Broman-Fulks, J. J., Berman, M. E., Rabian, B. A., &
Webster, M. J. (2004). Effects of aerobic exercise on anxiety
sensitivity. Behaviour Research and Therapy, 42, 125–136.
Broocks, A., Bandelow, B., Pekrun, G., George, A., Meyer, T.,
Bartman, U., et al. (1998). Comparison of aerobic exercise,
clomipramine, and placebo in the treatment of panic
disorder. American Journal of Psychiatry, 155, 603–609.
Broocks, A., Meyer, T., George, A., Hillmer-Vogel, U.,
Meyer, D., Bandelow, B., et al. (1999). Decreased neuro-
endocrine responses to meta-chlorophenylpiperazine (m-CPP)
but normal responses to ipsapirone in marathon runners.
Neuropsychopharmacology, 20(2), 150–161.
Broocks, A., Meyer, T., Gleiter, C. H., Hillmer-Vogel, U.,
George, A., Bartmann, U., et al. (2001). Effect of aerobic
exercise on behavioral and neuroendocrine responses to
meta-chlorophenylpiperazine and to ipsapirone in untrained
healthy subjects. Psychopharmacology, 155, 234–241.
Broocks, A., Meyer, T., Opitz, M., Bartmann, U., Hillmer-
Vogel, U., George, A., et al. (2003). 5-HT1A responsivity
in patients with panic disorder before and after treatment
with aerobic exercise, clomipramine or placebo. European
Neuropsychopharmacology, 13, 153–164.
Broocks, A., Schweiger, U., & Pirke, K. M. (1991). The
influence of semistarvation-induced hyperactivity on hypotha-
lamic serotonin metabolism. Physiology and Behavior, 130,
Brown, R. A., Lejuez, C. W., Kahler, C. W., Strong, D. R.,
Zvolensky, M. J. (2005). Distress tolerance and early
smoking lapse. Clinical Psychology Review, 25, 713–733.
Calogero, R. M., & Pedrotty, K. N. (2004). The practice and
process of healthy exercise: An investigation of the treat-
ment of exercise abuse in women with eating disorders.
Eating Disorders: The Journal of Treatment & Prevention, 12,
Casper, R. C., Katz, M. M., Bowden, C. L., Davis, J. M.,
Koslow, S. H., & Hanin, I. (1994). The pattern of physical
symptom changes in major depressive disorder following
treatment with amitriptyline or imipramine. Journal of
Affective Disorders, 31(3), 151–164.
Cohen, J. (1988). Statistical power analysis for the behavioral sci-
ences. San Diego, CA: Academic Press.
Collingwood, T. R., Sunderlin, J., Reynolds, R., & Kohl, H. W.
(2000). Physical training as a substance abuse prevention
intervention for youth. Journal of Drug Education, 30, 435–
Craft, L. L. (2005). Exercise and clinical depression: Examining
two psychological mechanisms. Psychology of Sport and Exercise,
Craft, L. L., & Landers, D. M. (1998). The effects of exercise
on clinical depression and depression resulting from mental
EXERCISE FOR MENTAL HEALTH META-ANALYSIS• STATHOPOULOU ET AL.191
illness: A meta-regression analysis. Journal of Sport & Exercise
Psychology, 20, 339–357.
Daniel, M., Martin, A. D., & Carter, J. (1992). Opiate recep-
tor blockade by naltrexone and mood state after acute phys-
ical activity. British Journal of Sports Medicine, 26(2), 111–
Dietrich, A., & Sparling, P. B. (2004). Endurance exercise
selectively impairs prefrontal-dependent cognition. Brain
and Cognition, 55, 516–524.
Dishman, R. (1991). Increasing and maintaining exercise and
physical activity. Behavioural Therapy, 22, 345–378.
Dishman, R. (1994). Advances in exercise adherence. Champaign,
IL: Human Kinetics Publishers.
Dishman, R. K. (1986). Mental health. In V. Seefeldt (Ed.),
Mental Health (pp. 303–341). Reston, VA: American
Alliance for Health, Physical Education, Recreation and
Doyne, E. J., Ossip-Klein, D. J., Bowman, E. D., Osborn, K. M.,
McDougall-Willson, I. B., & Neimeyer, R. A. (1987).
Running versus weight-lifting in the treatment of
depression. Journal of Consulting and Clinical Psychology, 55,
Driver, H. S., & Taylor, S. R. (2000). Exercise and sleep. Sleep
Medicine Reviews, 4(4), 387–402.
Dunn, A. L., Trivedi, M. H., Kampert, J. B., Clark, C. G., &
Chambliss, H. O. (2005). Exercise treatment for depression:
Efficacy and dose response. American Journal of Preventive
Medicine, 28, 1–8.
Eakin, E. G., Glasgow, R. E., & Riley, K. M. (2000). Review
of primary care-based physical activity intervention studies:
Effectiveness and implications for practice and future
research. Journal of Family Practice, 49, 158–168.
Ehlers, C. L., Frank, E., & Kupfer, D. J. (1988). Social zeitge-
bers and biological rhythms. Archives of General Psychiatry,
Ehlers, C. L., Kupfer, D. J., Frank, E., & Monk, T. H. (1993).
Biological rhythms and depression: The role of zeitgebers
and zeitstorers. Depression, 1, 285–293.
Eisenthal, S., Emery, R., Lazare, A., & Udin, H. (1979).
Adherence and the negotiated approach to patienthood.
Archives of General Psychiatry, 36, 393–398.
Ermalinski, R., Hanson, P. G., Lubin, B., Thornby, J. I., &
Nahormek, P. A. (1997). Impact of a body-mind treatment
component on alcoholic inpatients. Journal of Psychosocial
Nursing and Mental Health Services, 35(7), 39–45.
Fentem, P. H. (1994). Benefits of exercise in health and disease.
British Medical Journal, 308, 1291–1295.
Fossati, M., Amatti, D., Painot, M., Reiner, C., Haeni, C., &
Golay, A. (2004). Cognitive-behavioral therapy with simul-
taneous nutritional and physical activity education in
obese patients with binge eating disorder. Eating & Weight
Disorders, 9, 134–138.
Fremont, J., & Craighead, L. W. (1987). Aerobic exercise and
cognitive therapy in the treatment of dysphoric moods.
Cognitive Therapy Research, 11, 241–251.
Goisman, R. M., Warshaw, M. G., & Keller, M. B. (1999).
Psychosocial treatment prescriptions for generalized anxiety
disorder, panic disorder, and social phobia, 1991–1996.
American Journal of Psychiatry, 156, 1819–1821.
Goode, K. T., & Roth, D. L. (1993). Factor analysis of
cognitions during running: Association with mood change.
Journal of Sport & Exercise Psychology, 15, 375–389.
Goodwin, R. D. (2003). Association between physical activity
and mental disorders among adults in the United States.
Preventive Medicine, 36, 698–703.
Gorin, A., Phelan, S., Tate, D., Sherwood, N., Jeffery, R., &
Wing, R. (2005). Involving support partners in obesity
treatment. Journal of Consulting and Clinical Psychology, 73,
Gortner, E. T., Gollan, J. K., Dobson, K. S., & Jacobson,
N. S. (1998). Cognitive–behavioral treatment for depression:
Relapse prevention. Journal of Consulting and Clinical
Psychology, 66, 377–384.
Gould, R. A., Otto, M. W., & Pollack, M. H. (1995). A
meta-analysis of treatment outcome for panic disorder.
Clinical Psychology Review, 15, 819–844.
Grilo, C. M., Money, R., Barlow, D. H., Goddard, A. W.,
Gorma, J. M., Hofmann, S. G., et al. (1998). Pretreatment
patient factors predicting attrition from a multicenter
randomized controlled treatment study for panic disorder.
Comprehensive Psychiatry, 39, 323–332.
Hedges, L. V., & Olkin, I. (1985). Statistical methods for meta-
analysis. Orlando, FL: Academic Press.
Hopko, D. R., Lejuez, C. W., Ruggiero, K. J., & Eifert, G. H.
(2003). Contemporary behavioral activation treatments
for depression: Procedures, principles, and progress. Clinical
Psychology Review, 23(5), 699–717.
Hu, F. B., Willett, W. C., Li, T., Stampfer, M. J., Colditz, G. A.,
& Manson, J. E. (2004). Adiposity as compared with
physical activity in predicting mortality among women.
New England Journal of Medicine, 351, 2694–703.
Jacobson, N. S., Dobson, K. S., Truax, P. A., Addis, M. E.,
Koerner, K., Gollan, J. K., et al. (1996). A component
analysis of cognitive-behavioral treatment for depression.
Journal of Consulting and Clinical Psychology, 64, 295–304.
Jacobson, N. S., Martell, C. R., & Dimidjian, S. (2001). Behav-
ioral activation treatment for depression: Returning to con-
textual roots. Clinical Psychology: Science and Practice, 8, 255–270.
Jakicic, J. M., Winters, C., Lang, W., & Wing, R. R. (1999).
Effects of intermittent exercise and use of home exercise
CLINICAL PSYCHOLOGY: SCIENCE AND PRACTICE• V13 N2, SUMMER 2006192
equipment on adherence, weight loss, and fitness in over-
weight women: A randomized trial. Journal of the American
Medical Association, 282, 1554–1560.
Janal, M. N., Colt, E. W., Clark, W. C., & Glusman, M.
(1984). Pain sensitivity, mood and plasma endocrine levels
in man following long-distance running: Effects of
naloxone. Pain, 19(1), 13–25.
Klein, M. H., Greist, J. H., Gurman, A. S., Neimeyer, R. A.,
Lesser, D. P., Bushnell, N. J., et al. (1985). A comparative
outcome study group psychotherapy vs. exercise treatments
for depression. International Journal of Mental Health, 13,
Kubitz, K. A., Landers, D. M., Petruzzello, S. J., & Han, M.
(1996). The effects of acute and chronic exercise on sleep: A
meta-analytic review. Sports Medicine, 21, 277–291.
Kubitz, K. A., & Pothakos, K. (1997). Does aerobic exercise
decrease brain activation? Journal of Sport & Exercise Psychol-
ogy, 19, 291–301.
Lawlor, D. A., & Hopker, S. W. (2001). The effectiveness of
exercise as an intervention in the management of depres-
sion: Systematic review and meta-regression analysis of
randomized controlled trials. British Medical Journal, 322,
Leith, L. M. (1994). Foundations of exercise and mental health.
Morgantown, WV: Fitness Information Technology.
Long, B. C., & van Stavel, R. (1995). Effects of exercise training
on anxiety: A meta-analysis. Journal of Applied Sport Psychology,
MacGillivray, S., Arroll, B., Hatcher, S., Ogston, S., Reid, I.,
Sullivan, F., et al. (2003). Efficacy and tolerability of selec-
tive serotonin reuptake inhibitors compared with tricyclic
antidepressants in depression treated in primary care:
Systematic review and meta-analysis. British Medical Journal,
Markoff, R. A., Ryan, P., & Young, T. (1982). Endorphin and
mood changes in long-distance running. Medicine & Science
in Sports & Exercise, 14, 11–15.
Martinsen, E. W., Hoffart, A., & Solberg, O. (1989). Com-
paring aerobic with nonaerobic forms of exercise in the
treatment of clinical depression: A randomized trial.
Comprehensive Psychiatry, 30, 324–331.
Martinsen, E. W., Medhus, A., & Sandvik, L. (1985). Effects
of aerobic exercise on depression: A controlled study. British
Medical Journal (Clin Res ed.), 291(6488), 109.
Mather, A. S., Rodriguez, C., Guthrie, M. F., McHarg, A. M.,
Reid, I. C., & McMurdo, M. E. (2002). Effects of
exercise on depressive symptoms in older adults with poorly
responsive depressive disorder: Randomised controlled trial.
British Journal of Psychiatry, 180, 411–415.
Mayberg, H. S. (1997). Limbic-cortical dysregulation: A
proposed model of depression. Journal of Neuropsychiatry and
Clinical Neuroscience, 9, 471–481.
McCann, I. L., & Holmes, D. S. (1984). Influence of aerobic
exercise on depression. Journal of Personality and Social
Psychology, 46, 1142–1147.
McNally, R. J. (2002). Anxiety sensitivity and panic disorder.
Biological Psychiatry, 52, 938–946.
McNeil, J. K., LeBlanc, E. M., & Joyner, M. (1991). The
effect of exercise on depressive symptoms in the moderately
depressed elderly. Psychology and Aging, 6, 487–488.
McNemar, Q. (1960). At random: Sense and nonsense. Ameri-
can Psychologist, 15, 295–300.
Meyer, T., & Broocks, A. (2000). Therapeutic impact of exer-
cise on psychiatric diseases: Guidelines for exercise testing
and prescription. Sports Medicine, 30, 269–279.
Miller, R. C., & Berman, J. S. (1983). The efficacy of cog-
nitive behavior therapies: A quantitative review of the
research evidence. Psychological Bulletin, 94(1), 39–53.
Mokdad, A. H., Marks, J. S., & Stroup, D. F. (2004). Actual
causes of death in the United States, 2000. Journal of the
American Medical Association, 291(10), 1238–1245.
Morgan, W. P. (1985). Affective beneficence of vigorous
physical activity. Medicine & Science in Sports & Exercise, 17,
Murphy, T. J., Pagano, R. R., & Marlatt, G. A. (1986). Life-
style modification with heavy alcohol drinkers: Effects of
aerobic exercise and meditation. Addictive Behaviors, 11,
Palmer, J., Palmer, L. K., Michiels, K., & Thigpen, B. (1995).
Effects of type of exercise on depression in recovering
substance abusers. Perceptual and Motor Skills, 80, 523–530.
Palmer, J., Vacc, N., & Epstein, J. (1988). Adult inpatient alco-
holics: Physical exercise as a treatment intervention. Journal
of Studies on Alcohol, 49, 418–421.
Pendleton, V. R., Goodrick, G. K., Poston, W. S. C., Reeves,
R. S., & Foreyt, J. P. (2002). Exercise augments the effects
of cognitive-behavioral therapy in the treatment of binge
eating. International Journal of Eating Disorders, 31, 172–184.
Petruzzello, S. J., Landers, D. M., Hatfield, B. D., Kubitz, K. A.,
& Salazar, W. (1991). A meta-analysis on the anxiety-
reducing effects of acute and chronic exercise. Sports Medi-
cine, 11(3), 143–182.
Pinchasov, B. B., Shurgaja, A. M., Grischin, O. V., & Putilov,
A. A. (2000). Mood and energy regulation in seasonal and
non-seasonal depression before and after midday treatment
with physical exercise or bright light. Psychiatry Research,
Ransford, C. P. (1982). A role for amines in the antidepressant
effect of exercise: A review. Medicine & Science in Sports &
Exercise, 4(1), 1–10.
EXERCISE FOR MENTAL HEALTH META-ANALYSIS• STATHOPOULOU ET AL.193
Richardson, C. R., Faulkner, G., McDevitt, J., Skrinar, G. S.,
Hutchinson, D. S., & Piette, J. D. (2005). Integrating
physical activity into mental health services for persons with
serious mental illness. Psychiatric Services, 56(3), 324–331.
Rosenthal, R. (1991). Meta-analytic procedures for social
research. London: Sage.
Rosenthal, R., & Rubin, D. B. (1988). Comment: Assump-
tions and procedures in the file drawer problem. Statistical
Science, 3, 120–125.
Schmitz, N., Kruse, J., & Kugler, J. (2004). The association
between physical exercises and health-related quality of
life in subjects with mental disorders: Results from a cross-
sectional survey. Preventive Medicine, 39(6), 1200–1207.
Schulberg, H., Magruder, K., & deGruy, F. (1996). Major
depression in primary medical care practice: Research
trends and future priorities. General Hospital Psychiatry, 18,
Sexton, H., Maere, A., & Dahl, N. H. (1989). Exercise intensity
and reduction of neurotic symptoms: A controlled follow-
up study. Acta Psychiatrica Scandinavica, 80(3), 231–235.
Siegle, G. J., Ghinassi, F., & Thase, M. E. (in press). Neurobe-
havioral therapies in the 21st century: Summary of an
emerging field and an extended example of Cognitive Control
Training for depression. Cognitive Therapy and Research.
Singh, N. A., Clements, K. M., & Fiatarone, M. A. (1997).
A randomized controlled trial of progressive resistance
training in depressed elders. Journals of Gerontology. Series A:
Biological Sciences & Medical Sciences, 52A(1), M27–M35.
Singh, N. A., Clements, K. M., & Fiatarone Singh, M. (2001).
The efficacy of exercise as a long-term antidepressant in
elderly subjects: A randomized controlled trial. Journals of
Gerontology. Series A: Biological Sciences & Medical Sciences,
Singh, N. A., Stavrinos, T. M., Scarbek, Y., Galambos, G.,
Liber, C. L., & Fiatarone Singh, M. (2005). A randomized
controlled trial of high versus low intensity weight training
versus general practitioner care for clinical depression in
older adults. Journal of Gerontology. Series A: Biological Sciences
& Medical Sciences, 60A, 768–776.
Sinyor, D., Schwartz, S. G., Peronnet, F., Brisson, G., &
Serganian, P. (1983). Aerobic fitness level and reactivity to
psychosocial stress: Physiological, biochemical, and subjective
measures. Psychosomatic Medicine, 45, 205–217.
Smart, R. G. (1964). The importance of negative results in
psychological research. Canadian Psychologist, 5A, 225–232.
Smits, J. A. J., Otto, M. W., Powers, M. B., & Utschig, A.
(2005). A waitlist-controlled evaluation of the utility of an aerobic
exercise program for the treatment of panic disorder: Work in
progress. Poster presented at the annual meeting of the Anx-
iety Disorders Association of America, Seattle, Washington.
Smits, J. A. J., Powers, M. B., Cho, Y., & Telch, M. J. (2004).
Mechanism of change in cognitive-behavioral treatment of
panic disorder: Evidence for the fear of fear mediational
hypothesis. Journal of Consulting and Clinical Psychology, 72,
Smits, J. A. J., Powers, M. B., Berry, A. C., & Otto, M. W.
(2006). Translating empirically supported interventions
into accessible strategies: An analysis of the potential utility of
aerobic exercise for the treatment of panic. Manuscript
submitted for publication.
Steptoe, A., Edwards, A., Moses, J., & Mathews, A. (1989).
The effects of exercise training on mood and perceived
coping ability in anxious adults from the general popula-
tion. Journal of Psychosomatic Research, 33(5), 537–547.
Sterling, T. D. (1959). Publication decisions and their possible
effects of inferences drawn from tests of significance—or vice
versa. Journal of the American Statistical Association, 54, 30–34.
Sundgot-Borgen, J., Rosenvinge, J. H., Bahr, R., & Schnei-
der, L. S. (2002). The effect of exercise, cognitive therapy,
and nutritional counseling in treating bulimia nervosa.
Medicine & Science in Sports & Exercise, 34, 190–195.
Tanaka, H., & Shirakawa, S. (2004). Sleep health, lifestyle and
mental health in the Japanese elderly: Ensuring sleep to
promote a healthy brain and mind. Journal of Psychosomatic
Research, 56(5), 465–477.
Telch, C. R., Agras, W. S., Rossiter, E. M., Wilfley, D., &
Kenardy, J. (1990). Group cognitive-behavioral therapy for
the nonpurging bulimic: An initial evaluation. Journal of
Consulting and Clinical Psychology, 58, 629–635.
Thoren, P., Floras, J. S., Hoffmann, P., & Seals, D. R. (1990).
Endorphins and exercise: Physiological mechanisms and
clinical implications. Medicine & Science in Sports & Exercise,
Veale, D., Le Fevre, K., Pantelis, C., De Souza, V., Mann, A.,
& Sargeant, A. (1992). Aerobic exercise in the adjunctive
treatment of depression: A randomized controlled trial.
Journal of the Royal Society of Medicine, 85, 541–544.
Vogel, G. W. (1983). Evidence for REM sleep deprivation as
the mechanism of action of antidepressant drugs. Progress in
Neuropsychopharmacology & Biological Psychiatry, 7, 343–349.
Walsh, B. T., Seidman, S. N., Sysko, R., & Gould, M. (2002).
Placebo response in studies of major depression: Variable,
substantial, and growing. Journal of the American Medical
Association, 287, 1840–1847.
Williford, H. N., Barfield, B. R., Lazenby, R. B., & Olson, M. S.
(1992). A survey of physicians’ attitudes and practices related
to exercise promotion. Preventive Medicine, 21, 630–636.
Received August 23, 2005; revised January 9, 2006; accepted
February 13, 2006.