Download Effect of exercise intensity on depressive symptoms in women

Mental Health and Physical Activity 2 (2009) 37–43
Contents lists available at ScienceDirect
Mental Health and Physical Activity
journal homepage: www.elsevier.com/locate/menpa
Effect of exercise intensity on depressive symptoms in women
I-Hua Chu a,1, Janet Buckworth b, *, Timothy E. Kirby c, 2, Charles F. Emery d, 3
a
Department of Family and Preventive Medicine, University of California, San Diego, 9500 Gilman Drive, La Jolla, CA 92093, USA
Health and Exercise Science, The Ohio State University, A44 PAES Building, 305 W. 17th Avenue, Columbus, OH 43210, USA
c
Health and Exercise Science, The Ohio State University, A54 PAES Building, 305 W. 17th Avenue, Columbus, OH 43210, USA
d
Department of Psychology, The Ohio State University, 145 Psychology Building, 1835 Neil Avenue, Columbus, OH 43210, USA
b
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 19 August 2008
Received in revised form
7 January 2009
Accepted 7 January 2009
Exercise has been shown to relieve depressive symptoms, yet optimal exercise intensity for treating
depression has not been established. The mechanisms that explain the antidepressant effect of exercise
also require investigation. The purpose of this study was to test (a) the effect of two different exercise
intensities prescribed for aerobic training on depressive symptoms, and (b) a previously proposed
psychological mechanism for this effect: self-efficacy. Sedentary women scoring 14 on the Beck
Depression Inventory-II (BDI-II) were randomized to one of two aerobic training groups that differed on
exercise intensity (high [65–75% MaxVO2 reserve] or low [40–55% MaxVO2 reserve]), or to a stretching
control group for 10 weeks. Main outcome variables included depressive symptoms (BDI-II) and selfefficacy (exercise self-efficacy [ESE] and depression coping self-efficacy [DCSE]), which were measured at
study entry, 5 and 10 weeks later. Participants in all groups (high, n ¼ 18; low, n ¼ 18; stretching, n ¼ 18)
had significant reductions in depressive symptoms at Week 5 (p < .001) and Week 10 (p < .001). The BDIII change scores did not differ significantly among the groups (p ¼ .066). Follow-up analyses controlling
for baseline BDI-II scores showed that the high intensity group had significantly fewer depressive
symptoms than the low intensity and stretching control groups at weeks 5 and 10 (p < .05). There was no
significant association between changes in aerobic capacity and changes in depressive symptoms
(r ¼ .099, p ¼ .491). At 10 weeks, both ESE (p ¼ .013) and DCSE (p < .001) increased significantly for the
whole sample, with no significant group difference (p ¼ .613 for ESE, p ¼ .277 for DCSE). Controlling for
baseline scores, the increase remained significant for ESE (p ¼ .005) but not for DCSE (p ¼ .629). Partial
correlations showed significant negative relationships between both types of self-efficacy and depressive
symptoms at Week 5 and Week 10 (p < .02). We concluded that both high and low intensity aerobic
exercise, as well as stretching exercise were associated with reductions in mild to moderate depressive
symptoms in initially sedentary women. Changes in depression were associated with changes in ESE and
DCSE.
Published by Elsevier Ltd.
Keywords:
Women
Depressive symptoms
Exercise training
Self-efficacy
1. Introduction
Depression is a serious mental disorder throughout the world,
and the fourth leading cause of disease burden, accounting for 4.4%
of total disability adjusted life years (Global Burden of Disease:
Ustun, Ayuso-Mateos, Chatterji, Mathers, & Murray, 2004). In the
U.S., the estimated lifetime prevalence of major depressive disorder
(MDD) is 16.2% (Kessler et al., 2003). In any given year,
* Corresponding author. Tel.: þ1 614 292 0757; fax: þ1 614 688 3432.
E-mail addresses: [email protected] (I.-H. Chu), [email protected]
(J. Buckworth), [email protected] (T.E. Kirby), [email protected] (C.F. Emery).
1
Tel.: þ1 858 457 7296.
2
Tel.: þ1 614 292 0664.
3
Tel.: þ1 614 688 3061.
1755-2966/$ – see front matter Published by Elsevier Ltd.
doi:10.1016/j.mhpa.2009.01.001
approximately 13.7 million U.S. adults suffer from a depressive
disorder (Kessler et al., 2003). Women are twice as likely as men to
suffer from depression. A national survey indicated that the lifetime
prevalence for MDD is 12.6% among women and 6.3% among men
(Riolo, Nguyen, Greden, & King, 2005). Despite the availability of
effective pharmacological and psychotherapeutic treatments for
depression, only 51.6% of adults suffering from MDD seek treatment, and only 21.7% are adequately treated (Kessler et al., 2003).
Unlike traditional treatments, which can carry social stigma, exercise is socially acceptable, has fewer side effects, and has additional
health benefits. Hence, exercise may be a more satisfactory treatment for some individuals with depression.
Involvement in structured exercise has been shown to improve
symptoms of depression (Craft & Perna, 2004; Paluska & Schwenk,
2000; Teychenne, Ball, & Salmon, 2008). Experimental and
38
I.-H. Chu et al. / Mental Health and Physical Activity 2 (2009) 37–43
meta-analytic studies also indicate that exercise training is as
effective as antidepressant medication and psychotherapy (Blumenthal et al., 1999; Craft & Landers, 1998; North, McCullagh, &
Tran, 1990). Recently, Dunn, Trivedi, Kampert, Clark, and Chambliss
(2005) reported that aerobic exercise at a dose consistent with
public health recommendations (17.5 kcal/kg/week) is an effective
treatment for MDD of mild to moderate severity. Also, when
controlling total energy expenditure, exercise frequency (3 or 5
days/week) did not moderate the effect of exercise on depression.
While this study quantified the amount of exercise needed to
reduce symptoms of MDD, the effect of exercise intensity during
aerobic training was not examined. It is unknown whether exercise
intensity played a role in the antidepressant effect of exercise.
Exercise intensity is one of the essential components (i.e.,
intensity, duration, frequency, type) in exercise prescriptions.
Previous studies have indicated that reductions in depression were
not associated with differences in exercise duration, frequency, and
type (Craft & Landers, 1998; Dunn et al., 2005; North et al., 1990).
However, results from studies regarding exercise intensity and
depression have been equivocal. While some studies suggested that
both vigorous and moderate intensity exercise are equally effective
in alleviating symptoms of depression (Craft & Landers, 1998; Craft
& Perna, 2004), others reported an inverse relationship between
exercise intensity and depressive symptoms (Lampinen, Heikkinen,
& Ruoppila, 2000; Teychenne et al., 2008). In a review of studies on
the association between physical activity and depression in young
adult women, Azar, Ball, Salmon, and Cleland (2008) found that
even low levels of physical activity had a positive effect on symptoms of depression. However, dose–response effects and the
influence of exercise intensity were inconclusive because of limited
research and weaknesses in methodology. To examine the effect of
exercise intensity on depression, total energy expenditure must be
controlled in consideration of evidence that total energy expenditure is an important variable in exercise programs for depression
(e.g., Dunn et al., 2005) and to extend the majority of previous
studies that did not monitor or control for total energy expenditure.
Therefore, the purpose of this study was to examine the effect of
two different exercise intensities (high vs. low) during aerobic
training on depressive symptoms in initially sedentary depressed
women while controlling for total energy expenditure.
The mechanisms by which exercise exerts its antidepressant
effect have not been studied extensively and are not well understood (Craft, 2005). A better understanding of the mechanisms
would be helpful for future use of exercise as an adjunct or
alternative treatment of depression and more studies are needed
to examine the potential physiological and psychological mechanisms (Craft, 2005; Dunn, Trivedi, & O’Neil, 2001; Fox, 1999;
Landers & Arent, 2001). One plausible psychological mechanism
that is based on existing theories of depression and warrants
further examination is self-efficacy (Craft, 2005). Self-efficacy is
the degree to which an individual believes he or she can
successfully engage in a specific behavior in a particular situation
with known outcomes (Bandura, 1986). In the current study two
types of self-efficacy were measured, including exercise self-efficacy and depression coping self-efficacy. Exercise self-efficacy
measures one’s confidence to overcome barriers to exercise.
Depression coping self-efficacy measures one’s confidence to
engage in different activities to cope with depressive symptoms.
Research shows that depressed individuals have low self-efficacy
to cope with their depressive symptoms (deVries, Wiswell, Bulbulian, & Moritani, 1981) and mastery experience of a difficult
skill can help increase a person’s self-confidence, self-efficacy, and
ability to cope with personal problems (Bandura, 1997). Thus, it
was hypothesized that for depressed individuals, engaging in
regular exercise, which is usually perceived as a difficult task, is
likely to result in increased self-efficacy and enhanced ability to
cope with the symptoms of depression.
While self-efficacy has been found to influence exercise adoption and adherence in various populations (e.g., Dishman et al.,
2005; McAuley, Jerome, Elavsky, Marquez, & Ramsey, 2003; van
Weert et al., 2008; Wilbur, Vassalo, Chandler, McDevitt, & Miller,
2005), the relationship between exercise and self-efficacy in
depressed population has not been studied extensively and findings have been equivocal (Bodin & Martinsen, 2004; Craft, 2005;
Singh et al., 2005). In addition, no studies were found that examined the role of self-efficacy in respect to training intensity in
a depressed population. In contrast to a moderate exercise
program, successfully completing a vigorous exercise program may
lead to greater mastery experiences, more enhanced self-efficacy,
and thus greater reductions in depression. Therefore, a secondary
objective of this study was to examine the relationships between
exercise intensity, self-efficacy, and depressive symptoms.
2. Methods
2.1. Participants
Women aged 18–43 years with mild to moderate depressive
symptoms (a score of 14–28 on Beck Depression Inventory-II
[BDI-II]) were recruited from a university community via physician
referrals, posted flyers, classroom presentations, and word of
mouth. If a volunteer’s score indicated she had severe depressive
symptoms (a score above 28 on BDI-II), she could participate in this
study only with written permission from a physician or other
health care professional. To enroll in the study, participants also
had to be sedentary, which was defined as exercising less than
three times per week for less than 20 min per session, assessed
using responses to the Godin Leisure-Time Exercise Questionnaire
(Godin & Shephard, 1985). Other inclusion criteria were: having
a regular menstrual cycle during the previous six months; not
taking medication for depression or having taken current medication for at least three weeks (i.e., stabilized on current dose of
medication); and not receiving any other treatment for depression
(e.g., psychotherapy). Exclusion criteria were: suicidal ideation
reported on the BDI-II; a body mass index (BMI) less than 18.5
(underweight) or over 35 (obese); current pregnancy or nursing or
planned pregnancy during the following six months; and inability
to exercise due to physical contraindications (e.g., orthopedic
problems, heart disease). All participants completed informed
consent in compliance with the Institutional Review Board of the
university.
2.2. Study design
The study had one between-subject factor (group assignment: 3
levels) and one within-subject factor (time: 3 levels). Participants
were randomly assigned to one of two aerobic exercise groups (i.e.,
high intensity exercise or low intensity exercise), or to a stretching
exercise contact control group. The randomization was accomplished by drawing participants’ group assignments from a bag. At
study entry, participants performed maximal exercise testing using
the Balke treadmill protocol (Pollock, 2001) to determine maximal
oxygen uptake. After completing all the baseline assessments,
participants were informed of their group assignment. All participants began their training program within two weeks of their
baseline evaluation.
A 10-week training program was selected based upon results
from a meta-analysis by Craft and Landers (1998) that found exercise
interventions lasting at least 9 weeks were most effective in alleviating symptoms of depression. Over the course of 10 weeks, each
I.-H. Chu et al. / Mental Health and Physical Activity 2 (2009) 37–43
group met with the investigator once per week for supervised
exercise sessions. Participants in the aerobic groups exercised on
a treadmill for 30–40 min per supervised session, and those in the
stretching control group performed 30 min of supervised stretching
and flexibility exercise. Both aerobic groups began at an intensity of
40–55% of oxygen uptake reserve (VO2R) (Franklin, Whaley, Howley,
Balady, & Grais, 2000). After three weeks of training, the intensity
was increased to 65–75% of VO2R for the high intensity group
(Franklin et al., 2000). To make exercise more convenient and reduce
the burden and possible barrier of multiple weekly trips to the
school training facility, the aerobic groups’ participants were asked
to complete three to four exercise sessions each week on their own.
During these unsupervised sessions, participants chose their
preferred types of aerobic exercise, such as aerobic dancing, walking,
or biking. Exercise intensity and duration were prescribed by the
investigator based on the individual’s group assignment to meet an
exercise energy expenditure goal of 1000 Kcal/wk. The Compendium of Physical Activities (Ainsworth et al., 2000) was referenced to
assign the intensity for each type of exercise. Participants in the
stretching control group were instructed to maintain their physical
activity level. All participants were also given an activity diary to
complete, and diaries were reviewed weekly during supervised
sessions to monitor participants’ exercise intensity and weekly
energy expenditure. Weekly energy expenditure was calculated
from the recorded exercise intensity, duration, and frequency using
formulae provided in the ACSM’s guidelines for exercise testing and
prescription (Franklin et al., 2000).
2.3. Outcome measures
2.3.1. Depressive symptoms
Depressive symptoms were measured at study entry, Week 5 and
Week 10 using the Beck Depression Inventory-II (BDI-II) (Beck, Steer,
& Brown, 1996). The BDI-II is a self-report inventory with 21 items
assessing the behavioral and cognitive symptoms of depression. Each
item consists of four statements numbered from ‘‘0’’ to ‘‘3,’’ with
higher numbers indicating more severe depressive symptoms.
Participants were asked to circle the statement that best described
them during the prior week. The circled numbers were summed to
obtain a total score (range ¼ 0–63). Scores of 0–13 indicate minimal
depression,14–19 indicate mild depression, 20–28 indicate moderate
depression, and 29–63 indicate severe depression. This inventory has
demonstrated acceptable internal consistency with Cronbach’s
alphas ranging from .92 to .93 (Beck et al., 1996). Psychometric
analyses for both normal and psychiatric populations also have
consistently indicated that the BDI-II is a reliable and valid measure of
self-reported depression (Steer, Ball, Ranieri, & Beck, 1997).
2.3.2. Exercise self-efficacy
The Exercise Self-efficacy Questionnaire (Garcia & King, 1991)
was used to measure exercise self-efficacy (ESE) at study entry,
Week 5 and Week 10. Participants rated their confidence from 0%
(I cannot do it at all) to 100% (certain that I can do it) in their ability
to exercise under 15 different conditions (e.g., when tired; during
bad weather) over the next three months. The ratings for all items
were summed and then divided by 15 to obtain a mean score. This
scale has demonstrated acceptable internal consistency with
Cronbach’s alpha ranging from .90 to .94 (Wilcox, Sharpe, Hutto, &
Granner, 2005). Psychometric analyses in a diverse sample of men
and women indicated that the ESE questionnaire is a valid and reliable
measure for use with diverse populations (Wilcox et al., 2005).
2.3.3. Depression coping self-efficacy
Depression coping self-efficacy (DCSE) was measured at study
entry, Week 5 and Week 10 by the Depression Coping Self-Efficacy
39
Scale, a 24-item scale assessing coping self-efficacy of depressed
patients (Perraud, 2000). Each item describes a coping response
(e.g., get together with at least one very close person when I am
feeling lonely; take a bath or do some other soothing activity before
bedtime) and participants were asked to rate their confidence from
0% (not at all confident) to 100% (confident) in their ability to engage
in the coping response. Ratings for all items were summed and
divided by 24 to obtain a mean score. This scale has a test–retest
reliability of .84 and has demonstrated acceptable internal consistency reliability with a Cronbach’s alpha of .93 (Perraud, 2000). The
BDI-II was chosen as a test of concurrent validity for DCSE, and the
validity analysis showed a significant negative correlation between
DCSE and the BDI-II (r ¼ .73, p < .001) (Perraud, 2000).
2.4. Statistical analysis
Sample size calculation was based on the average effect sizes
reported in previous meta-analyses which indicated that exercise
groups decreased depression scores at least 0.50 standard deviation
more than comparison groups (e.g., control, leisure activity, or
psychotherapy groups) (Craft & Landers, 1998; North et al., 1990).
We estimated that we required 14 participants in each group (total
sample size ¼ 42) at an alpha level of .05, a power of .80, and an
effect size of 0.5. Sample sizes were inflated by 30% to allow for
possible attrition.
Descriptive statistics were performed for baseline characteristics of the participants. Intent-to-treat analysis of all randomized
participants was conducted. Missing data were imputed by carrying
forward the last recorded observation. Three separate repeated
measures analyses of variance (ANOVAs) were performed to
examine between group differences over time on depressive
symptoms, exercise self-efficacy, and depression coping self-efficacy. Analysis of covariance (ANCOVA) and post hoc comparisons
controlling for baseline scores were conducted to evaluate simple
effects (i.e., whether HI, LO, or SC had higher mean scores) at Weeks
5 and 10. Within subjects repeated measures ANOVA was performed to determine at what time points the change of mean scores
occurred (i.e., if change occurred from study entry to Week 5 or
from Week 5 to Week 10). Preliminary analyses showed that BMI
and antidepressant use did not differ across conditions, p ¼ .50, and
did not predict BDI-II scores over 10 weeks, therefore, these variables were omitted from the model. Partial correlations were
calculated to examine (a) the relationship between ESE and BDI-II
and (b) the relationship between DCSE and BDI-II at Week 5 and
Week 10 controlling for baseline values of ESE and DCSE at each
time point.
The significance level (a level) was set at .05 and partial Eta
squared (h2) was used as an indicator of effect size. All data analyses
were performed with the Statistical Package for the Social Sciences
(SPSS), Version 14.0 (SPSS, Inc., Chicago, IL, 2004).
3. Results
A total of 126 women were assessed for eligibility through
telephone or e-mail prescreen (Fig. 1). The 93 eligible women were
then scheduled for a screening visit to assess depressive symptoms,
level of physical activity, BMI, and other inclusion and exclusion
criteria. Fifty-four women met the study criteria and were enrolled
in the study. The major reasons for ineligibility were having
minimal depressive symptoms, currently receiving psychotherapy,
and inability to make the time commitment. The 54 participants
were randomized to one of the two treatment conditions: high
intensity aerobic exercise (HI; n ¼ 18) vs. low intensity aerobic
exercise (LO; n ¼ 18), or to the stretching exercise control (SC;
n ¼ 18) group.
40
I.-H. Chu et al. / Mental Health and Physical Activity 2 (2009) 37–43
Assessed for eligibility
N = 126
Phone/email screen - ineligible
n = 33
Not meeting inclusion criteria (n = 16)
Refused to participate (n = 7)
Other (n = 10)
Enrollment
Phone/email screen - eligible
n = 93
Screening visit - ineligible
n = 39
Not meeting inclusion criteria (n = 25)
Refused to participate (n = 6)
Other (n = 8)
Analysis
Follow-up
Allocation
Screening visit - eligible
n = 54
Randomized
n = 54
HI
n = 18
LO
n = 18
SC
n = 18
HI
Lost to follow-up (n = 0)
Never started (n = 0)
Discontinued intervention (n = 3)
Worsened (n = 0)
Medical (n = 1)
Other (n= 2)
LO
Lost to follow-up (n = 0)
Never started (n = 0)
Discontinued intervention (n = 7)
Worsened (n = 1)
Medical (n = 1)
Other (n = 5)
SC
Lost to follow-up (n = 2)
Never started (n = 2)
Discontinued intervention (n = 2)
Worsened (n = 0)
Medical (n = 0)
Other (n = 2)
HI
Intent to treat (n = 18)
Excluded from analysis (n = 0)
LO
Intent to treat (n = 18)
Excluded from analysis (n = 0)
SC
Intent to treat (n = 18)
Excluded from analysis (n = 0)
Fig. 1. Participant flow from enrollment to analysis.
Baseline characteristics of the 54 participants are shown in Table 1.
There were no significant group differences in baseline characteristics. Twenty-one participants (39%) had been taking medication
(i.e., Effexor, Lexapro, Paxil, Wellbutrin, and Zoloft) for their depressive symptoms at study entry for more than three weeks (see Table 1).
The percentage of antidepressant use was not significantly different
among the groups (p ¼ .496). At study entry, the average BDI-II scores
were above 20 (total sample BDI-II baseline range 15–37), suggesting
that on average the participants in each group were experiencing
moderate depressive symptoms (BDI-II scores 20–28).
Of the 54 participants, 16 (30%) did not finish the 10-weeks
training program [three from HI (17%), seven from LO (39%), and six
from SC (33%)]. Drop-out rate was not significantly different among
the groups (p ¼ .315). The major reasons for drop-out were lack of
time for exercise and low compliance with assigned exercise
intensity and energy expenditure. Participants who dropped out of
the study were not significantly different from those who
completed the intervention in all baseline characteristics except for
antidepressant use. Significantly more participants in the drop-out
group were taking antidepressants at study entry (p ¼ .021).
ranged from 80% to 100%, with no significant group difference
(HI ¼ 97.3%; LO ¼ 99.1%; SC ¼ 99.2%, p ¼ .475). The average intensity
of training, including supervised and unsupervised sessions, was
72 12.9% of VO2R in HI, 45 7.1% of VO2R in LO, and 25 6.0% of
VO2R in SC, indicating successful maintenance of the intended
training intensity in the three groups. The average energy expenditure (EE) from exercise was significantly higher for HI (1007.3 kcal/
week) and LO (905.2 kcal/week) as compared to SC (215.8 kcal/week)
(p < .001). The average EE did not differ significantly between HI and
LO (p ¼ .166). The average exercise frequency (EF) was significantly
higher for HI (3.7 sessions/week) and LO (4.9 sessions/week) as
compared to SC (2.1 sessions/week) (p < .001). LO also had significantly higher EF than HI (p ¼ .002). During unsupervised sessions, the
majority of participants in both HI and LO engaged in walking,
jogging, bicycling, or exercising on an elliptical trainer. After 10 weeks
of exercise training, participants in HI exhibited a significant 7%
increase in maximal VO2 (from 32.3 to 34.6 ml/kg/min, F ¼ 11.692,
p ¼ .004, h2 ¼ .455). There was no change in maximal VO2 for LO or SC.
3.1. Exercise training
As shown in Table 2, the intent-to-treat analysis for the whole
sample (N ¼ 54) revealed a significant improvement over time in
BDI-II scores across all groups (F ¼ 75.48, p < .001, h2 ¼ .597), with
no significant group by time interaction (F ¼ 2.58, p ¼ .066,
h2 ¼ .092). The main effect for group was significant (F ¼ 4.13,
p ¼ .022, h2 ¼ .139). These findings suggest that, regardless of group,
all participants had decreased depressive symptoms. Overall the
The average attendance rate for the supervised sessions ranged
from 0% to 100% for the whole sample (N ¼ 54), with no significant
difference among the groups (HI ¼ 87.2%; LO ¼ 77.2%; SC ¼ 76.1%,
p ¼ .504). For the 38 participants who completed the 10-week
training program, average attendance rate for supervised sessions
3.2. Depressive symptoms
I.-H. Chu et al. / Mental Health and Physical Activity 2 (2009) 37–43
41
Table 1
Baseline characteristics.
Characteristic
HI (n ¼ 18)
Age, y (range 18–43)
BMI, kg/m2
VO2max, ml/kg/min
BDI-II
ESE, %
DCSE, %
Antidepressants, n
LO (n ¼ 18)
SC (n ¼ 18)
Group difference
M
SD
M
SD
M
SD
p
26.4
25.0
33.1
21.6
68.2
69.4
6
7.3
4.1
5.6
6.1
15.3
9.0
26.6
23.5
34.1
22.4
58.2
70.7
9
6.3
3.7
5.5
4.9
20.7
11.9
24.6
23.8
33.6
23.4
60.2
66.8
6
4.8
4.0
6.4
4.7
19.5
13.5
.56
.50
.86
.60
.24
.59
.50
Note. BMI ¼ Body Mass Index; BDI-II ¼ Beck Depression Inventory-II; ESE ¼ Exercise Self-efficacy; DCSE ¼ Depression Coping Self-efficacy.
average BDI-II score decreased significantly from 22.5 (moderate
depressive symptoms) to 13.9 (mild depressive symptoms) at 5
weeks and continued to decrease significantly to 11.3 (minimal
depressive symptoms) at 10 weeks. Because group means appeared
different and low statistical power (0.58) was suspected for the
non-significant group by time interaction, further analyses were
conducted. Follow-up ANCOVA and post hoc comparisons
controlling for baseline BDI-II scores showed that HI had significantly fewer depressive symptoms than LO and SC at both Week 5
(HI 9.7 vs. LO 15.9, p ¼ .022; HI 9.7 vs. SC 15.9, p ¼ .037) and Week 10
(HI 6.4 vs. LO 12.7, p ¼ .034; HI 6.4 vs. SC 14.6, p ¼ .009). BDI-II
scores were not significantly different between LO and SC at either
Week 5 (p ¼ .841) or Week 10 (p ¼ .583). There was no significant
association between changes in aerobic capacity and changes in
depressive symptoms (r ¼ .099, p ¼ .491).
We analyzed the BDI-II data using the criteria for a clinically
significant reduction in depression according to Seggar, Lambert,
and Hansen (2002). Results showed that 13 participants (72%) in HI,
9 (50%) in LO, and 6 (33%) in SC met the criteria for a clinically
significant reduction in depression. These group differences did not
reach statistical significance (p ¼ .064).
Participants who were taking antidepressants at study entry did
not change the type and amount of their medication during the
course of this study, and the improvements in BDI-II over time were
similar between those who took antidepressants and those who did
not (p > .15). Thus, the effects of the medication should not have
affected the study results.
using ANCOVA to control for baseline ESE scores, with a significant
improvement from Week 5 to 10 (F ¼ 8.750, p ¼ .005, h2 ¼ .149) and
no significant group by time interaction (F ¼ 1.825, p ¼ .668,
h2 ¼ .068) or group effect (F ¼ 0.407, p ¼ .668, h2 ¼ .016). These
findings suggest that, regardless of group, all participants experienced increased exercise self-efficacy across time. Overall the
average ESE score increased significantly from 62.2% to 68.1% after
the 10-week training program.
3.3.2. Depression coping self-efficacy
The intent-to-treat analysis showed that there was a significant
improvement in DCSE over time (F ¼ 14.296, p < .001, h2 ¼ .219).
Overall, the average DCSE score increased significantly from 69.0% to
75.7% after the 10-weeks training program. There was no significant
group by time interaction (F ¼ 1.304, p ¼ .277, h2 ¼ .049) (Table 2) or
main effect for group (F ¼ .880, p ¼ .421, h2 ¼ .033). When using
ANCOVA to control for baseline DCSE scores, there was also no
significant improvement from Week 5 to 10 (F ¼ 0.236, p ¼ .629,
h2 ¼ .005) and no significant group by time interaction (F ¼ 0.502,
p ¼ .020, h2 ¼ .068) or group effect (F ¼ 1.818, p ¼ .173, h2 ¼ .068).
3.4. Correlation analysis
After controlling for baseline values, partial correlations
revealed significant negative correlations between exercise selfefficacy and depressive symptoms at Week 5 (r ¼ .348, p ¼ .011)
and Week 10 (r ¼ .492, p < .001). Partial correlations also revealed
significant negative correlations between depression coping selfefficacy and depressive symptoms at Week 5 (r ¼ .487, p < .001)
and Week 10 (r ¼ .652, p < .001).
3.3. Self-efficacy
3.3.1. Exercise self-efficacy
The intent-to-treat analysis showed that there was a significant
improvement in ESE over time (F ¼ 5.062, p ¼ .013, h2 ¼ .090) but
no significant group by time interaction (F ¼ .622, p ¼ .613,
h2 ¼ .024) (Table 2). The main effect for group was also not significant (F ¼ 1.383, p ¼ .260, h2 ¼ .051). Similar results were found
4. Discussion
A number of exercise intervention studies have demonstrated
a reduction in depressive symptoms following exercise programs of
various modes and durations (Blumenthal et al., 1999; Craft, 2005;
Table 2
Mean psychological values by group at study entry, Week 5, and Week 10.
Measure
BDI-II pre
BDI-II w5
BDI-II w10
ESE pre
ESE w5
ESE w10
DCSE pre
DCSE w5
DCSE w10
HI (n ¼ 18)
LO (n ¼ 18)
SC (n ¼ 18)
Time
M
SD
M
SD
M
SD
21.6
9.7
6.4
68.2
67.2
74.4
69.4
74.7
79.7
6.1
4.4
4.0
15.3
17.3
15.2
9.0
10.6
10.4
22.4
15.9
12.7
58.2
62.7
66.3
70.7
70.9
74.7
4.9
9.8
10.4
20.7
20.1
21.5
11.9
14.1
16.5
23.4
15.9
14.6
60.2
61.3
63.5
66.8
70.2
72.7
4.7
8.0
9.2
19.5
18.3
16.7
13.5
9.9
9.5
Group time
F
p
F
p
75.47
<.001
2.58
.07
.62
.61
1.30
.28
5.06
.01
14.29
<.001
Note. BDI-II ¼ Beck Depression Inventory-II; ESE ¼ Exercise Self-efficacy; DCSE ¼ Depression Coping Self-efficacy. F ratio and p values from repeated measures analysis of
variance model.
42
I.-H. Chu et al. / Mental Health and Physical Activity 2 (2009) 37–43
Dimeo, Bauer, Varahram, Proest, & Halter, 2001; Dunn et al., 2005).
In our study, we found that both high and low intensity aerobic
exercise groups and stretching control group had significant
improvements in depressive symptoms at both 5 weeks and 10
weeks. All participants began the study with moderate depressive
symptoms, and after 10-weeks of aerobic exercise training or
stretching, their BDI-II scores indicated only minimal depressive
symptoms. There was no significant difference in the reduction in
BDI-II scores among the groups. These results suggest that both
high and low intensity aerobic exercise, as well as stretching
exercise were associated with reductions in mild to moderate
depressive symptoms. However, when comparing BDI-II scores
between groups at each time point, participants in HI reported
significantly fewer depressive symptoms than those in LO and SC at
both weeks 5 and 10 (p < .05). The LO and SC, on the other hand, did
not differ significantly at either time point. These results suggest
that a high intensity (65–75% of VO2R) aerobic exercise program
may be more effective in reducing symptoms of depression. Yet,
since the overall reductions in depression did not differ significantly among the groups, whether high intensity exercise is
necessary for improving symptoms of depression requires further
investigation. We found that changes in aerobic capacity were not
significantly associated with changes in depressive symptoms. This
finding mirrors results of previous studies suggesting that fitness
gain is not required for reductions in depression (Craft & Landers,
1998; North et al., 1990; Veale et al., 1992).
The reductions in depression found in this study may have been
a result of placebo effects since the SC group also showed significant
improvements. In other words, participant’s expectation, social
contact, and personal attention given by the investigator may have
partially contributed to the improved depressive symptoms for all
participants after exercise training. Furthermore, stretching exercise,
although used as a placebo control, may actually be an effective
activity treatment for depressive symptoms. In our study, we introduced a yoga-based stretching exercise. Recently, research conducted
on yoga and depression has suggested that yoga practice can be an
effective treatment for depressive symptoms (Elavsky & McAuley,
2007; John, Sharma, Sharma, & Kankane, 2007; Lavey et al., 2005). The
beneficial effect of yoga on depression may explain the significant
improvement in depressive symptoms seen in our stretching control
group. This may help explain the absence of a difference in BDI-II
change scores between the aerobic groups and SC. For future
randomized control trials, we would recommend a different type of
activity (e.g., lecture meetings) or a wait-list control rather than
stretching exercise for the control condition to examine the effects of
training intensity and mediating variables on symptoms of depression.
Regardless of group assignment, all participants had significantly increased exercise self-efficacy after 10 weeks of training,
but no change in depression coping self-efficacy when controlling
for baseline scores. However, we found significant negative relationships between both types of self-efficacy and depressive
symptoms at Weeks 5 and 10, indicating that higher levels of selfefficacy were associated with lower levels of depression. These
results were similar to findings in previous studies on self-efficacy
mechanisms (Bodin & Martinsen, 2004; Craft, 2005). Although this
was not evidence of self-efficacy mediating the beneficial effects of
exercise on depression, these results provide some support for
further exploring both exercise self-efficacy and depression coping
self-efficacy as potential mechanisms for the antidepressant effects
of exercise. Future research should further examine these mechanisms using mediation analyses. In addition, an exercise self-efficacy instrument that addresses intensity, duration, and frequency
would provide greater insight into the relationship between
different training programs and changes in depression than the
general exercise self-efficacy questionnaire applied in this study.
There were several limitations to the study. First, participants in
the aerobic groups had one supervised exercise session and two to
four unsupervised sessions. We did not use a heart rate monitor or
a step counter to monitor physical activity outside of supervised
sessions and had to rely on self-report to measure the participants’
exercise behaviors and estimate the weekly energy expenditure
using their activity diaries. It was possible for participants to
overestimate the intensity and duration of exercise, or record extra
exercise sessions. However, the significant increase in maximal VO2
in HI suggests that participants in HI indeed exercised at their
recorded exercise intensity, duration, and frequency. In addition,
participants were instructed to record their exercise session in
detail (e.g., speed and time on treadmill, distance and time if
walking outdoors) and not to count any breaks during their exercise
sessions. These instructions reduced the likelihood of overestimation of exercise intensity and duration, and helped estimate
participants’ weekly energy expenditure.
In addition, participants were not required to have a diagnosis of
depression to enroll in this study. Therefore, findings in this study
may not reflect responses in women with clinical depression,
although research conducted with clinically depressed populations
resulted in similar reductions in BDI-II scores following involvement in an exercise program (Blumenthal et al., 1999; Craft, 2005).
Also, the sample is relatively small compared to many exercise
intervention studies, resulting in low statistical power, which may
explain the finding of non-significant interaction effect for BDI-II
scores. Based on the results of current study, we would have needed
72 participants (i.e., 24 participants per group) for adequate power
to detect an interaction effect for BDI-II scores. Therefore, it is
important for future research to replicate this study with larger
samples to further test the relationship between exercise intensity
and depression.
In summary, both high and low intensity aerobic exercise, as
well as yoga-based stretching exercise were associated with
reductions in depressive symptoms after 10 weeks of training in
sedentary depressed women. The reductions in depression did not
differ significantly among the groups. A high intensity (65–75% of
VO2R) aerobic exercise program appeared to be more effective in
reducing symptoms of depression based on between group
comparisons at Week 5 and Week 10, although there was no
significant association between changes in aerobic capacity and
changes in depressive symptoms. Further, exercise self-efficacy and
depression coping self-efficacy were inversely associated with
depressive symptoms at repeated assessments. These results
suggest the importance of evaluating self-efficacy as a mediator of
the relationship between exercise and depressive symptoms in
future research utilizing a larger sample.
References
Ainsworth, B. E., Haskell, W. L., Whitt, M. C., Irwin, M. L., Swartz, A. M., Strath, S. J., et al.
(2000). Compendium of physical activities: an update of activity codes and MET
intensities. Medicine and Science in Sports and Exercise, 32(9 Suppl.), S498–S504.
Azar, D., Ball, K., Salmon, J., & Cleland, V. J. (2008). The association between physical
activity and depressive symptoms in young women: a review. Mental Health &
Physical Activity, 1, 82–88.
Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory.
Englewood Cliffs, NJ: Prentice-Hall.
Bandura, A. (1997). Self-efficacy: The exercise of control. New York: W.H. Freeman.
Beck, A. T., Steer, R. A., & Brown, K. (1996). Beck Depression Inventory (2nd ed.). San
Antonio, TX: Harcourt Brace.
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. Archives of Internal Medicine, 159, 2349–2356.
Bodin, T., & Martinsen, E. W. (2004). Mood and self-efficacy during acute exercise in
clinical depression. A randomized, controlled study. Journal of Sport & Exercise
Psychology, 26, 623–633.
Craft, L. L. (2005). Exercise and clinical depression: examining two psychological
mechanisms. Psychology of Sport and Exercise, 6, 151–171.
I.-H. Chu et al. / Mental Health and Physical Activity 2 (2009) 37–43
Craft, L. L., & Landers, D. M. (1998). The effect of exercise on clinical depression and
depression resulting from mental illness: a meta-analysis. Journal of Sport &
Exercise Psychology, 20, 339–357.
Craft, L. L., & Perna, F. M. (2004). The benefits of exercise for the clinically depressed.
Primary Care Companion to The Journal of Clinical Psychiatry, 6, 104–111.
Dimeo, F., Bauer, M., Varahram, I., Proest, G., & Halter, U. (2001). Benefits from
aerobic exercise in patients with major depression: a pilot study. British Journal
of Sports Medicine, 35, 114–117.
Dishman, R. K., Motl, R. W., Sallis, J. F., Dunn, A. L., Birnbaum, A. S., Welk, G. J., et al.
(2005). Self-management strategies mediate self-efficacy and physical activity.
American Journal of Preventive Medicine, 29, 10–18.
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.
Dunn, A. L., Trivedi, M. H., & O’Neil, H. A. (2001). Physical activity dose–response
effects on outcomes of depression and anxiety. Medicine and Science in Sports
and Exercise, 33, s587–s597.
Elavsky, S., & McAuley, E. (2007). Physical activity and mental health outcomes
during menopause: a randomized controlled trial. Annals of Behavioral Medicine, 33, 132–142.
Fox, K. R. (1999). The influence of physical activity on mental well-being. Public
Health Nutrition, 2, 411–418.
Franklin, B. A., Whaley, M. H., Howley, E. T., Balady, G. J., & Grais, I. M. (Eds.). (2000).
ACSM’s guidelines for exercise testing and prescription (6th ed.). Philadelphia, PA:
Lippincott Williams & Wilkins.
Garcia, A. W., & King, A. C. (1991). Predicting long-term adherence to aerobic
exercise: a comparison of two models. Journal of Sport and Exercise Psychology,
13, 394–410.
Godin, G., & Shephard, R. J. (1985). A simple method to assess exercise behavior in
the community. Canadian Journal of Applied Sport Sciences, 10, 141–146.
John, P. J., Sharma, N., Sharma, C. M., & Kankane, A. (2007). Effectiveness of yoga
therapy in the treatment of migraine without aura: a randomized controlled
trial. Headache: The Journal of Head and Face Pain, 47, 654–661.
Kessler, R. C., Berglund, P., Demler, O., Jin, R., Koretz, D., Merikangas, K. R., et al.
(2003). The epidemiology of major depressive disorder: results from the
National Comorbidity Survey Replication (NCS-R). Journal of the American
Medical Association, 289, 3095–3105.
Lampinen, P., Heikkinen, R.-L., & Ruoppila, I. (2000). Changes in intensity of physical
exercise as predictors of depressive symptoms among older adults: an eightyear follow-up. Preventive Medicine, 30, 371–380.
Landers, D. M., & Arent, S. M. (2001). Physical activity and mental health. In
R. N. Singer, H. A. Hausenblaus, & C. M. Janelle (Eds.), Handbook of sport
psychology (2nd ed.). New York, NY: John Wiley and Sons.
Lavey, R., Sherman, T., Mueser, K. T., Osborne, D. D., Currier, M., & Wolfe, R. (2005).
The effects of yoga on mood in psychiatric inpatients. Psychiatric Rehabilitation
Journal, 28, 399–402.
43
McAuley, E., Jerome, G. J., Elavsky, S., Marquez, D. X., & Ramsey, S. N. (2003). Predicting long-term maintenance of physical activity in older adults. Preventive
Medicine, 37, 110–118.
North, T. C., McCullagh, P., & Tran, Z. V. (1990). Effect of exercise on depression.
Exercise and Sport Sciences Reviews, 18, 379–415.
Paluska, S. A., & Schwenk, T. L. (2000). Physical activity and mental health: current
concepts. Sports Medicine (Auckland, N.Z.), 29, 167–180.
Perraud, S. (2000). Development of the Depression Coping Self-Efficacy Scale
(DCSES). Archives of Psychiatric Nursing, 14, 276–284.
Pollock, K. M. (2001). Exercise in treating depression: broadening the psychotherapist’s role. [Special issue]. Psychotherapy in Practice, 57, 1289–1300.
Riolo, S. A., Nguyen, T. A., Greden, J. F., & King, C. A. (2005). Prevalence of depression
by race/ethnicity: findings from the National Health and Nutrition Examination
Survey III. American Journal of Public Health, 95, 998–1000.
Seggar, L. B., Lambert, M. J., & Hansen, N. B. (2002). Assessing clinical significance:
application to the Beck Depression Inventory. Behavior Therapy, 33, 253–269.
Singh, N. A., Stavrinos, T. M., Scarbek, Y., Galambos, G., Liber, C., & Fiatarone
Singh, M. A. (2005). A randomized controlled trial of high versus low intensity
weight training versus general practitioner care for clinical depression in older
adults. The Journals of Gerontology Series A, Biological Sciences and Medical
Sciences, 60, 768–776.
Steer, R. A., Ball, R., Ranieri, W. F., & Beck, A. T. (1997). Further evidence for the
construct validity of the Beck Depression Inventory-II with psychiatric outpatients. Psychological Reports, 80, 443–446.
Teychenne, M., Ball, K., & Salmon, J. (2008). Associations between physical activity
and depressive symptoms in women. International Journal of Behavioral Nutrition and Physical Activity, 5, 27.
Ustun, T. B., Ayuso-Mateos, J. L., Chatterji, S., Mathers, C., & Murray, C. J. L. (2004).
Global burden of depressive disorders in the year 2000. The British Journal of
Psychiatry: the Journal of Mental Science, 184, 386–392.
Veale, D., LeFevre, K., Pantelis, C., deSouza, V., Mann, A., & Sargent, A. (1992). Aerobic
exercise in the adjunctive treatment of depression: a randomized controlled
trial. Journal of the Royal Society of Medicine, 85, 541–544.
deVries, H. A., Wiswell, R. A., Bulbulian, R., & Moritani, T. (1981). Tranquilizer effect
of exercise. Acute effects of moderate aerobic exercise on spinal reflex activation level. American Journal of Physical Medicine, 60, 57–66.
van Weert, E., Hoekstra-Weebers, J. E., May, A. M., Korstjens, I., Ros, W. J., & van der
Schans, C. P. (2008). The development of an evidence-based physical selfmanagement rehabilitation programme for cancer survivors. Patient Education
and Counseling, 71, 169–190.
Wilbur, J., Vassalo, A., Chandler, P., McDevitt, J., & Miller, A. M. (2005). Midlife
women’s adherence to home-based walking during maintenance. Nursing
Research, 54, 33–40.
Wilcox, S., Sharpe, P. A., Hutto, B., & Granner, M. L. (2005). Psychometric properties
of the self-efficacy for exercise questionnaire in a diverse sample of men and
women. Journal of Physical Activity and Health, 3, 285–297.