Article which may be of interest:

http://www.medscape.com/viewarticle/573862

Depressive Symptoms and Physical Activity in Adolescent Girls

Carolyn C. Johnson; David M. Murray; John P. Elder; Jared B. Jobe; Andrea L.
Dunn; Martha Kubik; Carolyn Voorhees; Kenneth Schachter

Med Sci Sports Exerc. 2008;40(5):818-826. ©2008 American College of Sports
Medicine

Abstract and Introduction

Abstract

Purpose: To evaluate the relationship between depressive symptoms and
physical activity in a geographically and ethnically diverse sample of
sixth-grade adolescent girls.
Methods: The Trial of Activity for Adolescent Girls (TAAG) baseline
measurement included a random sample (N = 1721) of sixth-grade girls in 36
schools at six field sites. Measurements were accelerometry and the 3-d
Physical Activity Recall (3DPAR) for physical activity, and the Center for
Epidemiological Studies-Depression scale (CES-D) for depressive symptoms.
Results: Girls with complete data (N = 1397), mean age 12 yr, had an average
CES-D score of 14.7 (SD = 9.25) and engaged in an average of about 460 min
of sedentary activity, < 24 min of moderate to vigorous physical activity
(MVPA), and < 6 min of vigorous physical activity (VPA) in an 18-h day.
Thirty-minute segments of MVPA ranged in number from 3.9 to 1.2, and METS
for these segments ranged from > 3.0 to > 6.5. Mixed-model regression
indicated no relationship between depressive symptoms and physical activity;
however, a significant but modest inverse relationship between sedentary
activity and depressive symptoms was observed.
Conclusion: A sufficient sample size, standardized procedures, and validated
instruments characterized this study; however, a relationship between
depressive symptoms and physical activity was not observed for sixth-grade
girls from diverse geographic locations. The average CES-D score was lower
than is considered clinically meaningful for either adolescents or adults,
and MET-minutes of sedentary activity were high. This combination of data
may be different from other studies and could have contributed to the
unexpected finding. This unexpected finding is informative, however, because
it shows the need for additional research that includes a wider range of
possible combinations of data, especially with young adolescent girls.
Introduction

A physically inactive lifestyle is associated with the development of
obesity, chronic physical disorders, and certain mental and emotional
conditions.[37] Despite evidence that physical activity has multiple health
benefits and is associated with reductions in morbidity and mortality,
physical inactivity among the U. S. population continues to be
widespread.[37] National studies have shown that almost two thirds of U.S.
adults are either sedentary or irregularly active,[19,23] and children's
levels of physical activity, although initially higher than adults, decline
as children grow into adolescence and young adulthood.[37] Indeed, by
middle-school age, more than one third of adolescents are not reaching
recommended physical activity levels, and just a little over one third of
adolescents participate in daily physical activity programs.[37] Multiple
studies have reported that the decrease in adolescent physical activity is
much more pronounced for girls than for boys,[29,37] and that the proportion
of adolescent boys versus adolescent girls who are consistently active is
about two to one.[23]

Major depressive disorder is one of the most common mental disorders among
adults in industrialized countries.[2] Major depressive disorder and
expressions of depressive symptomatology can begin during adolescence and
are associated with morbidity and mortality in the adolescent
population.[28] Depression in its variant forms is now occurring earlier in
life than it did in the past,[28] and, surprisingly, the annual rate of
depression among adolescents and young adults is almost twice that of older
adults.[40] A recent study by Kessler et al. (2005) found that half of all
lifetime cases of mental disorders, including depression, start at about 14
yr of age.[15]

The above studies suggest that adolescence is a developmental period when
physical activity levels decrease, especially for girls, and depressive
symptoms begin or increase, especially for girls.[1] Both adult and
adolescent females have a higher probability than males of depressive
symptom expression,[1] with adolescent girls being twice as likely as
adolescent boys to develop depression, a gender ratio similar to that for
physical inactivity.

Although not all studies agree, both cross-sectional and longitudinal
population-based studies have shown a link between depression or depressive
symptoms and physical activity levels.[4,5,13,29] For example, attendance in
physical education (PE) class has been inversely related to feelings of
sadness.[4] Calfas and Taylor (1994) reviewed adolescent physical activity
studies and concluded that physical activity was consistently related to
improvements in depressive symptoms, with an effect size of 0.38 for
depressive symptoms.[5] In a review of 102 studies of correlates of physical
activity for children and adolescents, Sallis et al. (2000) indicated that,
although beneficial effects of physical activity in youth are less well
documented than in adults, positive effects have been demonstrated in youth
on both physical and psychological health, that is, depression.[29] Again,
child and adolescent males were consistently more physically active than
females. Of the 54 adolescent studies reviewed by Sallis et al.,[29] the
majority indicated that depression was inversely correlated with physical
activity. Physical activity measurement was almost always self-report, with
only six studies using objective measures such as pedometers or
accelerometers.

Some studies, however, have not found an inverse relationship between
physical activity and depression and/or depressive symptoms.[12,17,21,22]
For example, Norris, Carroll, and Cochrane found that high-intensity, but
not moderate-intensity, physical training resulted in positive effects on
psychological well-being for adolescents.[22] Fulkerson also found that
moderate-to-vigorous physical activity (MVPA) and depressive symptoms were
negatively associated for males but not for females.[12] As Sallis et al.
(2000) have noted, these inconsistencies could be the result of studies
using mixed-sex and -age samples, measuring physical activity with possibly
unreliable and/or invalid self-report instruments, and/or other
methodological differences, including the nonmeasurement of other important
covariates that could be linked to depression or physical activity.[29]
Trial of Activity for Adolescent Girls

Because physical inactivity is an independent risk factor for coronary heart
disease,[10] and a decline in MVPA accelerates for girls during adolescence,
the National Heart, Lung, and Blood Institute (NHLBI) initiated a randomized
controlled multicenter trial, the Trial of Activity for Adolescent Girls
(TAAG).[31] The primary aim of TAAG was to design and evaluate a school- and
community-linked intervention to reduce by half the decline in physical
activity (primarily measured by Actigraph accelerometers) in middle school
girls. The intervention, based on the Social Ecological Model,[32] was
intended to affect the physical and social environment through programs in
health education and PE that link schools with community organizations to
increase opportunities for physical activity, along with promotional efforts
to enhance motivation. TAAG is a collaborative trial among six field centers
(Universities of Arizona, Maryland, Minnesota, and South Carolina; San Diego
State University; and Tulane University), the coordinating center at the
University of North Carolina, Chapel Hill, and the NHLBI. A data and safety
monitoring board provided oversight and performed an advisory role. Six
schools per field center (N = 36 schools) were randomized to either
intervention (N = 18) or control (N = 18) conditions after baseline
measurement.

In the current study, participants were sixth-grade adolescent girls
representing multiple ethnicities in six geographically diverse areas of the
United States who participated in the TAAG baseline cross-sectional
measurements prior to randomization to experimental conditions. Physical
activity was measured with both accelerometry and self-report. The objective
of the current study was to evaluate the relationship of depressive symptoms
with levels of physical activity measured by both accelerometry and
self-report in a free-living geographically and ethnically diverse sample of
sixth-grade girls.
Method
TAAG Study Design

TAAG recruitment of both schools and students followed standardized
protocols at all field sites. Schools eligible for participation in TAAG
were publicly funded, with no magnet or special populations and with a less
than 28% student dropout rate during any given year. Districts with middle
schools meeting eligibility criteria were identified. Sixty-eight schools
were contacted, 41 agreed to participate, and 36 schools, representing 19
districts, were finally selected based on inclusion criteria, one of which
was willingness to accept random assignment. Girls were recruited based on
random sampling within each school. A total of 1721 girls provided written
parental consent for a baseline participation rate of 80% of all eligible
girls. The measurement design consisted of a series of three cross-sectional
measurements, with the first providing baseline data and the last two
providing follow-up data. The data reported here represent baseline
assessment and will not be reported by intervention or control condition,
but in the aggregate. The TAAG study design is described in more detail
elsewhere.[31]
Data Collection Procedures

At baseline, the goal was to recruit 60 girls from each school to
participate in baseline measurement, which included 1) Mediators and
Moderators Student Questionnaire (MSQ), which incorporated the Center for
Epidemiological Studies-Depression scale (CES-D); 2) accelerometry for
measurement of physical activity; 3) a 3-d physical activity recall (3DPAR);
and 4) body composition (height, weight, and triceps skinfolds). All
measurement protocols were reviewed and approved by the respective human
subjects institutional review boards for each of the seven universities
engaged in the study. Written informed consent was obtained from the parents
of the participants, and the participants had the right of assent or refusal
at time of measurement. Data collection documents were prelabeled prior to
field use with either a unique identification (ID) number for each student
or a bar code representing the ID. All data were collected by TAAG staff
trained according to standardized protocols and certified for data
collection only after practice administrations met specified standards.
Periodic recertification ensured that performance standards were met
continuously.
Measurement

CES-D. Depressive symptoms were measured with the 20-item CES-D, a
self-report measure intended for research applications in general
nonpsychiatric populations. The majority of items (N = 16) assess cognitive,
affective, behavioral, and somatic symptoms associated with depression; the
remaining four items are included to minimize biases attributable to
response sets, as well as for their intrinsic value.[26] Frequency responses
are recorded on a four-point Likert scale and range from "rarely" to "almost
always." The instrument has been well standardized and used extensively with
both adults and adolescents.[26,27] Numerous interview and
self-administration field tests have found the CES-D acceptable and
understandable by a wide range of general population respondents.[38] High
to adequate levels of internal consistency (alpha = 0.85), split-half (alpha
= 0.87), and test-retest correlations (range from 0.51 to 0.67) were found
in all groups studied.[26] Correlations of the scale with other mental
health measures and clinical ratings of depression have produced reasonable
evidence of both construct and discriminant validity (0.73-0.89).[38]
Acceptability and internal consistency have been demonstrated across age,
sex, and race (white, African American, Mexican Americans, and Hispanics),
as well as with adolescent populations.[27] Although the CES-D is not a
diagnostic instrument, a score of > 16 has been used to indicate a
"presumption of depression" with adults[21] and > 24 to indicate clinically
meaningful symptoms of depression in adolescents.[7]

MTI ActiGraph Accelerometer. The MTI ActiGraph accelerometer is an objective
measure of MVPA, the primary outcome variable for TAAG. The accelerometer
records acceleration and deceleration of movement, resulting in activity
counts, intensities of physical activity, and time of day for accelerations.
The monitor has demonstrated high reliability, validity, and sensitivity to
different intensities of physical activity, and has been used in several
studies to assess physical activity in children.[14,39] Correlations between
heart rate and activity counts have ranged from 0.64 to 0.57. Reliability
correlations over time have ranged from 0.87 to 0.42.[7,14] At baseline each
girl wore the activity monitor against the right hip bone for six complete
days except while sleeping or during any water activity (e.g., bathing,
swimming). Official data collection for the study began at 6:00 a.m. on the
day after the monitors were distributed. Downloading of data from the
ActiGraph to laptops occurred immediately upon removal of the monitor or
within the next 24 h, and data were transmitted to the Coordinating Center
according to standardized procedures using the TAAG Data Management System
(DMS).

Accelerometer readings were processed using methods similar to those
reported by Puyau et al..[25] Cut points for sedentary (< 50 counts/30 s),
light (51-1499), moderate (1500-2600), and vigorous (> 2600) activity were
based on an earlier pilot study that related accelerometer counts to MET
levels across a range of activities.[34] Half-minute counts were used
instead of full-minute counts based on the expectation that they would be
more sensitive to fluctuations in activity levels. Occasional missing
accelerometry data within a girl's 6-d record were replaced via imputation
based on the expectation maximization (EM) algorithm.[6] On average,
approximately 12 h (11%) of data per person were imputed. For the
MET-minutes of MVPA measure, counts above 1500 per half minute were
converted into metabolic equivalents (METs) using a regression equation
developed from a second pilot study for TAAG.[30] The METs were summed over
the 6:00 a.m. to midnight day to provide MET-minutes per day of MVPA, where
1 MET·min represented the metabolic equivalent of energy expended sitting at
rest for 1 min.

3DPAR. The primary objective of the 3DPAR is to measure the types and
contexts of physical activity as self-reported by participants. Immediately
on completion of accelerometry, participants were administered the 3DPAR in
a group setting that took about 45 min. For each of the past 3 d, the
students were asked to recall what activity they did for every 30-min
interval throughout the day. Codes for the girls to use in recording
"activity number," "with whom" and "where" were provided in a coding
instruction sheet. In addition to the 3-d recall, students were asked how
frequently they rode their bike to and from school and/or walked to and from
school during the previous 7 d. Forms were reviewed by TAAG staff for
completeness and clarity, and girls were asked to correct any noted
problems. Test-retest reliability of the 3DPAR among girls was 0.707 for
MVPA and 0.771 for VPA.[18] The 3DPAR has also been validated against
pedometer counts, accelerometer counts, and heart rate monitors. These
validity measures were 0.88 for step counts, 0.77 for accelerometer counts,
and 0.53-0.63 for heart rate monitors.[24]

BMI. Both height and weight were measured twice using a Shorr height board
and a Seca Model 880 weight scale. Triceps skinfolds were measured three
times with Lange skinfold calipers. Height measurements were repeated if the
difference between the two measurements was > 1 cm. Weight measurements were
repeated if the difference was >0.5 kg, and the three triceps measurements
were repeated if (maximum - minimum)/minimum was > 0.20. Girls were
evaluated in their bare feet or socks after removing all excess clothing and
any heavy accessories.

MSQ. The MSQ was developed by a TAAG working group for the purpose of
measuring mediators, moderators, and secondary outcomes as specified by the
TAAG theoretical model. Variables included in the MSQ were initially pilot
tested for timing, readability, and adequacy of questions. Another pilot
study tested the revised version of the MSQ using confirmatory factor
analyses, Cronbach alpha, test-retest reliability, and intraclass
correlations.[3] Because the CES-D is a published instrument with adequate
psychometric data for adolescent girls, it was not included in pilot
testing. MSQ variables measured at TAAG baseline that could be associated
with or mediate physical activity levels and/or depressive symptoms and
selected for this study were: demographic data (7 items); enjoyment of PE (1
item with Likert scale response from disagree a lot to agree a lot);
enjoyment of physical activity (7 items with Likert scale responses ranging
from disagree a lot to agree a lot); participation in sports teams and/or
lessons in and outside of school (checklist of 15 and 18 items,
respectively); social support for physical activity (4 items for friends and
5 items for family with Likert scale responses ranging from never to every
day); home alone without adult supervision (2 frequency items); barriers to
physical activity (10 items with Likert scale responses ranging from never
to very often); and change strategies for physical activity (9 items with
5-point Likert scale responses from never to very often). Internal
consistency coefficients for enjoyment, social support, barriers and change
strategies ranged from 0.46 (barriers) to 0.86 (enjoyment). Factor validity
for these variables ranged from 0.96 (social support) to 0.99 (enjoyment and
change strategies). Confirmatory factor analyses indicated that the
theoretical relations among the items from each variable were similar across
racial groups.
Statistical Analysis Methods

Simple descriptive statistical methods were used to characterize the sample
with regard to the independent and dependent variables. Descriptive
statistics were calculated by site, as well as for the overall sample, to
provide some indication of variability among the sites. Methods included
sample means with standard deviations and frequency distributions.

Mixed-model linear regression analyses were used to evaluate the
relationship between physical activity and CES-D score modeled as a
continuous variable, with separate analyses for each of nine measures of
physical activity (five based on accelerometry, four based on self-report).
We fit a crude model first, predicting CES-D from physical activity. Second,
we added seven demographic variables: race, mother's education, father's
education, household structure, participation in the free or reduced lunch
program, mother's employment status, and father's employment status.
Nonsignificant terms were removed until all remaining terms were significant
at P < 0.10. The third model added the following variables: enjoyment of
physical education, enjoyment of physical activity, body mass index,
participation in sports programs in school, participation in sports programs
out of school, participation in physical activity lessons, social support
for physical activity, barriers to physical activity, time spent home alone
after school, and strategies to change physical activity. Again,
nonsignificant terms were removed until all remaining terms reached a P
value of < 0.10.

Mixed-model logistic regression analyses were used to evaluate the
relationship between physical activity and the likelihood of having a CES-D
score of 24 or higher. The model-building process for the logistic models
followed the same pattern as for the linear models.

For both models, site and school within site were included in all analyses
as random effects to account for correlation among observations taken on
girls who lived in the same city and among girls who attended the same
school. SAS version 8.2 was used for all analyses.
Results
Sample Characteristics

A total of 1721 sixth-grade girls participated in the TAAG baseline
measurement. Only the girls who completed measurement of the main variables
of interest, the CES-D, the 3DPAR, and accelerometry, were selected for this
study (N = 1397). The sociodemographic descriptors of this sample are
presented in Table 1 . Overall, 46.7% of participants were white, with an
average age of 12.0 yr. Average ages were fairly consistent across sites
ranging from 11.7 yr at the Maryland site to 12.2 yr at the Louisiana site.
A measure of socioeconomic status for the school was the percentage of girls
on free or reduced lunch, which overall was 41.1%, with the Louisiana site
having the largest percentage of girls on free or reduced lunch (75.4%). The
majority of girls (72.2%) lived with both parents, and both mothers and
fathers were generally well educated, with 34.1% of mothers and 30.8% of
fathers having a college degree and/or education beyond college. A near
majority of mothers (48.4%) and a majority of fathers (69.0%) worked
full-time. Two thirds of the girls reported either very little or no time
home alone; however, more than one third were home alone at least one or
more hours a day.
Depressive Symptomatology

The CES-D was analyzed for internal consistency, and the resulting
Cronbach's alpha was 0.88. The means and standard deviations for the CES-D (
Table 2 ) are presented both for CES-D as a continuous score as well as
after categorizing the CES-D using a cut point of 24, with 0 = less than 24,
and 1 = greater than or equal to 24. The mean overall CES-D score for the
continuous presentation was 14.7 and ranged across sites from 12.2
(Minnesota) to 17.9 (Louisiana). The average proportion of participants with
scores greater than or equal to 24 was 18%, with a range of 11% to 29% for
the Minnesota and Louisiana sites, respectively.
Physical Activity

Table 2 also presents the means and standard deviations for physical
activity measured by accelerometry and self-reported with the 3DPAR.
Accelerometry data were skewed with an overall average of almost 460 min·d-1
spent in sedentary behavior and an average of about 342 min for light
physical activity. Fewer than 24 (23.8) minutes per day were spent in
moderate to vigorous physical activity (MVPA), and fewer than 6 (5.6)
minutes per day were spent in vigorous physical activity (VPA).

The 3DPAR was analyzed in 30-min segments, which were categorized according
to the MET level assigned to the intensity-weighted activity in those
segments ( Table 2 ). The criterion for MVPA was METs = 3.0, and any
segments with MET activities less than 3.0 were excluded. The average number
of segments with METs greater than 3.0 was 3.9 while only 1.2 segments had
METs greater than 6.5.
Relationship between Physical Activity and Depressive Symptoms

BMI ( Table 2 ) has been associated with physical activity.[36] Other
variables measured at TAAG baseline that could be associated with or mediate
physical activity levels and/or depressive symptom scores are presented in
Table 3 . Participation in sports teams, social support for and enjoyment of
physical activity, barriers and change strategies are all related to
physical activity. It was expected that enjoyment of physical activity would
be compromised when depressive symptoms are present. Additionally,
adolescents who are experiencing depressive symptoms would not be expected
to initiate strategies to overcome barriers or make changes to increase
physical activity.[2] Based on the theoretical and behavioral expectations
for these variables, they were used to build both the linear and logistic
regression models.

As stated previously, a mixed-model linear regression model was used to
predict CES-D (dependent variable) separately from each of nine measures of
physical activity (independent variables). Only sedentary behavior, as
measured by accelerometry, was statistically significant in the crude model
fit for each activity variable. All of the sociodemographic variables in
Table 1 were then included as additional predictors with the intention of
retaining only those variables with a P value of < 0.10. Father's education
(P = 0.07), family structure (P = 0.03), and participation in the
free/reduced lunch program (P = 0.09) were related to CES-D within the
specified P value. These three variables were retained in the model and the
additional variables were then added. Of these, the variables that remained
at P < 0.10 were enjoyment of physical education (P = 0.09), enjoyment of
physical activity (P < 0.0001), BMI (P < 0.0001), number of sports
participated in outside of school (P = 0.0002), social support (P < 0.0001),
number of barriers to physical activity (P < 0.0001), and time home alone (P
= 0.004).

Consistent with the crude model, only sedentary behavior measured by
accelerometry was significant in the adjusted models. The regression
coefficients and P values from the adjusted model are presented in Table 3 .
The data show that higher sedentary behavior was significantly, but
modestly, associated with lower CES-D scores; that is, for every additional
minute of sedentary behavior measured by accelerometry, the average CES-D
score decreased by 0.00978 points. Inasmuch as the overall mean for
sedentary behavior in this sample is 460 and the SD is 69.3, a 2-SD increase
in sedentary behavior would predict a 1.35-point decline in CES-D scores.
The overall mean BMI was 20.8 and the SD is 4.8; therefore, a 2-SD increase
in BMI would predict an increase in CES-D of 1.89. Compared with girls who
have no time home alone after school, those girls who spent more than 2
h·d-1 home alone had CES-D scores that were 2.8 units higher. The
coefficients for enjoyment of physical education and physical activity,
sports participation, support and barriers to physical activity were quite
small and, therefore, not very meaningful.

A logistic regression model using CES-D scores greater than or equal to 24
was developed using the same model-building approach as for the linear
regression model. The results from the logistic regression model were
similar to those from the linear regression model, indicating no
relationship between physical activity and CES-D greater than or equal to
24.
Discussion

The purpose of this study was to evaluate the relationship between
depressive symptoms and levels of physical activity. The sixth-grade girls
randomly recruited from the 36 TAAG schools showed comparable levels of
depressive symptoms with a sample of approximately 1900 seventh-grade girls
who participated in the TEENS Study,[20] with a mean CES-D score of 14.7 in
TAAG and 14.8 in TEENS. The mean score for TAAG, however, was slightly
higher than that for the AddHealth Study in which the average CES-D was
12.2. This slightly lower average may be reflective of a very large sample
in the AddHealth study of 13,568, which included both boys and girls, and
symptoms in males are generally lower than in females.[28]

Some 18% of the TAAG girls had a CES-D score greater than 24. This is twice
as high as the 9% found in the AddHealth Study, but not as great as the 29%
observed in another sample of girls 16-18 yr of age.[11] These data are
alarming because there is increasing evidence that even subsyndromal
symptoms of depression are associated with significant psychosocial
impairment and increased likelihood of developing major depressive
disorder.[16]

In this sample of sixth-grade girls, no association between depressive
symptoms and physical activity levels was observed. The methods used in this
study were rigorous, and state-of-the-science sampling, measurement, and
related methods were used. The CES-D is a widely used and respected
instrument for measuring depressive symptoms and has been shown to be
reliable and valid in a variety of population groups, including adolescents.
Physical activity levels were measured using both accelerometry and
self-report, and neither of these methods indicated a relationship between
physical activity and depressive symptoms. The information that emerged from
these data, however, was that physical activity levels for these sixth-grade
girls were well below public health recommendations of 60 min of MVPA daily.
On average, this sample of girls engaged in less than 24 min of
moderate-intensity physical activity daily and less than 6 min·d-1 of
vigorous physical activity.

Other studies have reported similar results. After participants in their
study initiated a physical activity course, Mack et al. observed that mood
state scores did not change, with participants remaining "fairly happy
throughout the length of the course".[17] There may be a higher probability
of finding an inverse relationship between physical activity and depression
in clinical samples than in population-based samples, although, as mentioned
previously, other population-based studies have found such a
relationship.[29]

It should be noted that the present sample of girls was young (mean age 12
yr), and two possibilities could be associated with such a young age. First,
it is possible that they generally did not have symptoms associated with
depression, or, secondly, had difficulty identifying and documenting the
emotions, feelings, and symptoms listed in the CES-D.

In this study, the inverse relationship noted between depressive symptoms
and sedentary behavior was unexpected; however, this is most likely a
statistical anomaly given that the size of the correlation coefficient was
quite small and the sample size was large. Additionally, baseline values
showed that sedentary behavior was high and generally CES-D scores were not.

A variable that was significantly predictive of increases in CES-D scores
was BMI. The overall BMI in this study was 20.8. When there is no
statistical relationship between physical activity and depressive symptoms
and a population-based measurement of BMI is not low, as is the case in this
study, other possible factors may be involved, such as negative self-image
or dissatisfaction with one's weight or size.[8] Some studies have even
evaluated weight-teasing and shame among adolescents as predictive of
negative mood states,[9] as well as failed attempts to weigh less.[33,35]

Another variable that was significantly predictive of CES-D scores was being
home alone more than 2 h each day after school. Some studies have indicated
that individuals who engage in more social interactions, and who enjoy those
interactions, do not experience the levels of depression or depressive
symptoms of those who are less socially active.[23] Adolescents who are home
alone most evenings obviously have limited opportunities to be socially
interactive with others.

The absence of a significant association between depressive symptoms and
physical activity in this and other studies is indicative of the complex
nature of that relationship, its possible dependence on other variables
(some known and some yet to be examined), and the sociodemographic factors
within specific subgroups. Within some subgroups, the relationship may
simply not exist. Additionally, the combination of data in this study, low
physical activity levels and high BMI, may not have existed in studies
reporting an inverse relationship between physical activity and depressive
symptoms. All of the possibilities discussed result in the conclusion that
additional research on this issue is warranted that includes a wider range
of combinations of data, as well as measurement of other factors that could
be involved when depressive symptoms are present, specifically with girls at
middle school ages.

What else is published in Medicine & Science in Sports & Exercise®? Visit
www.acsm-msse.org.

Table 1. Demographic Characteristics of Sixth-grade Girls Participating in
Trial of Activity for Adolescent Girls Baseline Measurement by Site


Table 1: Demographic Characteristics of Sixth-grade Girls Participating
in Trial of Activity for Adolescent Girls Baseline Measurement by Site


Table 2. Means and Standard Deviations for Measurements of Physical Activity
(3DPAR and CSA) and Depressive Symptoms (CES-D) by Site


Table 2: Means and Standard Deviations for Measurements of Physical
Activity (3DPAR and CSA) and Depressive Symptoms (CES-D) by Site


Table 3. Predicting CES-D Score from Sedentary Behavior and Other Factors


Table 3: Predicting CES-D Score from Sedentary Behavior and Other
Factors




References

1. Allison SL, Roeger G, Martin J. Gender differences in the relationship
between depression and suicidal ideation in young adolescents. Aust N Z J
Psychiatry. 2001;34:498-503.
2. American Psychiatric Association. Diagnostic and Statistical Manual of
Mental Disorders. 4th ed. Washington (DC): American Psychiatric Association;
1994. p. 327-49.
3. Birnbaum AS, Evenson KR, Motl RW, et al. Scale development for
perceived school climate for girls' physical activity. Am J Health Behav.
2005;29:250-7.
4. Brosnahan J, Steffen LM, Lytle L, Patterson J, Boostrom A. The
relation between physial activity and mental health among Hispanic and
non-Hispanic white adolescents. Arch Pediatr Adolesc Med. 2004;158:818-23.
5. Calfas JJ, Taylor WC. Effects of physical activity on psychological
variables in adolescents. Pediatr Exerc Sci. 1994;6:406-23.
6. Catellier JD, Murray DM, Hannan PJ, et al. Imputation of missing data
when measuring physical activity by accelerometry. Obes Res. 2003;11:A32.
7. Chabrol H, Montovany A, Chouicha K, Duconge E. Study of the CES-D on a
sample of 1,953 adolescent students. Encephale. 2002;28:429-32.
8. Durkin SJ, Paxton SJ. Predictors of vulnerability to reduced body
image satisfaction and psychological wellbeing in response to exposure to
idealized female media images in adolescent girls. J Psychosom Res.
2002;53:995-1005.
9. Eisenberg ME, Neumark-Sztainer D, Wall M. Weight-teasing and emotional
well-being in adolescents: longitudinal findings from Project EAT. J Adolesc
Health. 2006;38:675-83.
10. Fletcher GF, Balady G, Blair SN, et al. Statement on exercise:
benefits and recommendations for physical activity programs for all
Americans. A statement for health professionals by the Committee on Exercise
and Cardiac Rehabilitation of the Council on Clinical Cardiology, American
Heart Association. Circulation. 1996;94:857-62.
11. Franko DL, Striegel-Moore RH, Thompson D, Schreiber GB, Daniels SR.
Does adolescent depression predict obesity in black and white young adult
women? Psychosom Med. 2005;35:1505-13.
12. Fulkerson JA, Sherwood NE, Perry CL, Neumark-Sztainer D, Story M.
Depressive symptoms and adolescent eating and health behaviors: a
multifaceted view in a population-based sample. Prev Med. 2004.;38:865-75.
13. Haarasilta L, Marttunen MJ, Kaprio JA, Aro IM. Correlates of
depression in a representative nationwide sample of adolescents (15-19
years) and young adults (20-24 years). Eur J Public Health. 2004;14:280-5.
14. Janz KF, Witt J, Mahoney LT. The stability of children's physical
activity as measured by accelerometry and self-report. Med Sci Sports Exerc.
1995;27(9):1326-32.
15. Kessler RC, Berglund P, Demler O, Jin R, Merikanga KR, Walters EE.
Prevalence and treatment of mental disorders, 1990 to 2003. N Engl J Med.
2005;352:2515-23.
16. Lewinsohn PM, Solomon A, Seeley JR, Zeiss A. Clinical implications of
"subthreshold" depressive symptoms. J Abnorm Psychol. 2000;109(2):345-51.
17. Mack MG, Huddleston S, Dutler KE, Bian W. Mood state changes of
students enrolled in physical activity classes. Percept Mot Skills.
2000;90:911-4.
18. McMurray RG, Ring KB, Treuth MS, et al. Comparison of two approaches
to structured physical activity surveys for adolescents. Med Sci Sports
Exerc. 2004;36(12):2135-43.
19. Mokdad AH, Serdula MK, Dietz WH, Bowman BA, Marks JS, Koplan JP. The
spread of the obesity epidemic in the United States, 1991-1998. JAMA.
1999;282:1519-22.
20. Motl RW, Birnbaum AS, Kubik MY, Dishman RK. Naturally occurring
changes in physical activity are inversely related to depressive symptoms
during early adolescence. Psychosom Med. 2004;66:336-42.
21. Myers JK, Weissman MM. Use of a self-report symptom scale to detect
depression in a community sample. Am J Psychiatry. 1980;137:1081-4.
22. Norris R, Carroll D, Cochrane R. The effects of physical activity and
exercise training on psychological stress and well-being in an adolescent
population. J Psychosom Res. 1992;36:55-65.
23. Paluska SA, Schwenk TL. Physical activity and mental health: current
concepts. Sports Med. 2000;29:167-80.
24. Pate RR, Ross R, Dowda M, Trost SG, Sirard J. Validation of a 3-day
physical activity recall instrument in female youth. Pediatr Exerc Sci.
2003;15:257-65.
25. Puyau MR, Adolph AL, Vohra RA, Butte NF. Validation and calibration of
physical activity monitors in children. Obes Res. 2002;10:150-7.
26. Radloff LS. The CES-D Scale: a self-report depression scale for
research in the general population. Appl Psychol Meas. 1977;1:385-401.
27. Radloff LS, Locke BZ. The community mental health assessment survey
and the CES-D scale. In: Weissman MM, Myers JK, Ross CE, editors. Community
Surveys of Psychiatric Disorders. New Brunswick (NJ): Rutgers University
Press; 1986. p. 177-189.
28. Rushton JL, Forcier M, Schectman RM. Epidemiology of depressive
symptoms in the National Longitudinal Study of Adolescent Health. J Am Acad
Child Adolesc Psychiatry. 2002;41:199-205.
29. Sallis JP, Prochaska JJ, Taylor WC. A review of correlates of physical
activity of children and adolescents. Med Sci Sports Med. 2000;32(5):963-75.
30. Schmitz KH, Treuth M, Hannan P, et al. Predicting energy expenditure
from accelerometry counts in adolescent girls. Med Sci Sports Med.
2005;37(1):155-61.
31. Stevens J, Murray DM, Catellier DJ, et al. Design of the Trial of
Activity in Adolescent Girls (TAAG). Contemp Clin Trials. 2005;26:223-33.
32. Stokols D. Establishing and maintaining healthy environments. Toward a
social ecology of health promotion. Am Psychol. 1992;47:6-22.
33. Tanofsky-Kraff M, Faden D, Yanovski SZ, Wilfley DE, Yanovski JA. The
perceived onset of dieting and loss of control eating behaviors in
overweight children. Int J Eat Disord. 2005;38:112-22.
34. Treuth MS, Almeida J, Catellier DJ, et al. Defining accelerometer
thresholds for activity intensities in adolescent girls. Med Sci Sports Med.
2004;36(7):1259-66.
35. Trost SG, Ward DW, Moorehead SM, Watson PD, Riner W, Burke JR.
Validity of the Computer Science and Applications (CSA) activity monitor in
children. Med Sci Sports Med. 1998;30(4):629-33.
36. U.S. Department of Health and Human Services. Physical Activity and
Health: A Report of the Surgeon General. Washington (DC): U.S. Department of
Health and Human Services; 1996.
37. U.S. Department of Health and Human Services. Physical Activity
Fundamental to Preventing Disease. Washington (DC): U.S. Department of
Health and Human Services; 2002.
38. Weissman MM, Sholomskas D, Pottenger M, Prusoff BA, Locke BZ.
Assessing depressive symptoms in five psychiatric populations: a validation
study. Am J Epidemiol. 1977;106:203-14.
39. Weston AT, Petosa R, Pate RR. Validation of an instrument for
measurement of physical activity in youth. Med Sci Sports Med.
1997;29(1):138-43.
40. Wight RG, Sepulveda JE, Aneshensel CS. Depressive symptoms: how do
adolescents compare with adults? J Adolesc Health. 2004;34:314-23.

Acknowledgements

The TAAG investigators are grateful to all of the girls and schools who
supported, cooperated with, and participated in this study.
Funding Information

This research was supported by the National Heart, Lung, and Blood Institute
(National Institutes of Health) grants U01HL66845, U01HL66852, U01HL66853,
U01HL66855, U01HL66856, U01HL66857, and U01HL66858.
Reprint Address

Carolyn C. Johnson, Ph.D., Tulane University School of Public Health and
Tropical Medicine, 1440 Canal St., Rm 2309, New Orleans, LA, 70112; E-mail:
cjohnso5@... .

Carolyn C. Johnson,1 David M. Murray,2 John P. Elder,3 Jared B. Jobe,4
Andrea L. Dunn,5 Martha Kubik,6 Carolyn Voorhees,7 Kenneth Schachter8

1Tulane University School of Public Health & Tropical Medicine, New Orleans,
LA
2The Ohio State University, Columbus, OH
3San Diego State University, San Diego, CA
4National Heart, Lung, and Blood Institute, Bethesda, MD
5Klein Buendel, Inc., Golden, CO
6University of Minnesota, Minneapolis, MN
7University of Maryland, College Park, MD
8University of Arizona, Tucson, AZ