22 annotated bibliographies AMA style
rjmapstt
Theimpactofphysicalactivityonbodyweightandfatgainsduringthefirst3yearsofcollege..pdf
The impact of physical activity on body weight and fat gains during the first 3 years of college
Sareen S. Gropper a *, Frank H. Newell
a , Ali Zaremba-Morgan
b , Margaret K. Keiley
b ,
B. Douglas White a , Kevin W. Huggins
a , Karla P. Simmons
c , Lenda Jo Connell
c and
Pamela V. Ulrich c
a Department of Nutrition, Dietetics, and Hospitality Management, Auburn University, 101 Poultry Science Bldg., Auburn, AL 36849, USA;
b Department of Human Development and Family Studies,
Auburn University, 203 Spidle Hall, Auburn, AL 36849, USA; c Department of Consumer Affairs,
Auburn University, 308 Spidle Hall, Auburn, AL 36849, USA
Over two-thirds of students gain weight and body fat during college, especially during the freshman year. This study examined whether participation in physical activity during the first 3 years of college was associated with favorable changes in body weight and percent body fat. Participants included 535 college students (345 females, 190 males). Height and weight (assessed by standard techniques) and body fat (assessed by bioelectrical impedance analysis) were obtained at the beginning of fall semester and at the end of spring semester of each year; in addition during the first 2 years, assessments were also conducted at the end of fall semester for a total of eight assessments between 2007 and 2010. Physical activity participation was self-reported using a questionnaire that included a subset of questions from the National College Health Risk Behavior Survey. While both males and females exhibited significant increases in weight and percent body fat over the 3-year period; for the females, participation in strength training was associated with loss of weight and percent body fat. The results of this study emphasize the benefits of physical activity, especially strength training, for college females as a means to reduce or prevent body weight and fat gains during the first 3 years of college.
Keywords: body fat gains; college students; strength training
Introduction
Young adults entering college are at risk for obesity. About two-thirds of college students
gain weight during their freshman year, and, while few gain the notorious ‘freshman 15
(referring to pounds),’ a 1.8 to 2.3-kg (4 – 5 lb) weight gain is frequently reported
(Hajhosseini et al. 2006, Edmonds et al. 2008, Kasparek et al. 2008, Gropper et al. 2009,
Mifsud et al. 2009, Pullman et al. 2009). Coupled with weight gain is a decline in physical
activity, which begins as students transition from high school and persist throughout
college (Caspersen et al. 2000, Huang et al. 2003, Bray and Born 2004, Buckworth and
Nigg 2004, Kasparek et al. 2008, Mestek et al. 2008, Racette et al. 2008, Pullman et al.
2009, Wengreen and Moncur 2009, Gropper et al. 2011).
Several factors associated with gains in weight and/or body mass index (BMI) during
college have been identified including snacking behaviors, dieting strategies, all-you-can-
eat dining, alcohol consumption, and physical inactivity. Physical inactivity is especially
ISSN 1463-5240 print/ISSN 2164-9545 online
q 2012 Institute of Health Promotion and Education
http://dx.doi.org/10.1080/14635240.2012.724190
http://www.tandfonline.com
*Corresponding author. Email: [email protected]
International Journal of Health Promotion and Education
Vol. 50, No. 6, November 2012, 296–310
D ow
nl oa
de d
by [
E P
- IP
S W
IC H
], [
A m
an da
L ev
y] a
t 08
:0 3
05 M
ar ch
2 01
3
detrimental to health, increasing the risk for stroke, hyperlipidemias, type 2 diabetes, and
hypertension (Katzmarzyk 2006, U.S. Dept of Health and Human Services 2008). Physical
activity enables muscular work and energy expenditure to maintain a healthy body
composition and decrease obesity risk. Both a healthy BMI and percentage body fat are
important; problems, including cardiometabolic abnormalities, are associated with
‘normal weight obesity’ (normal BMI – unhealthy body fat percentage) (Romero-Corral
et al. 2010). Yet, while physical activity promotes health, about one-third of young adults
in the United States are not engaged in sufficient physical activity (Schoenborn and Adams
2010). This study’s purpose was to determine whether physical activity participation
during the first 3 years of college was associated with favorable changes in body weight
and percent body fat.
Methods
Subjects
At the beginning of fall semester 2007 and 2008, two cohorts of incoming freshmen at
Auburn University, aged 17 – 19 years, unmarried, without children, and without a
diagnosed eating disorder, were recruited from introductory level courses. The sample
(n ¼ 535; 345 females, 190 males; mean age 18 years, SD ^ 0.4) was representative of
the university’s incoming freshman classes (Auburn University Office of Institutional
Research and Assessment 2007 and 2008). Around 62% of the freshmen were from
Alabama; the rest were from other U.S. states. The racial composition was Caucasian
(84%), African-American (7%), Hispanic (3%), Asian (2.5%), and other (3%). The
Institutional Review Board for the Use of Human Subjects in Research approved this
study. Informed consents were obtained from subjects prior to participation.
Study design and measures
Measures were collected from participants at eight different time points over their first 3
years (2007 – 2010) of college, including September, freshman year (Time 1 ¼ 0 months);
December, freshman year (Time 2 ¼ 2.7 months); April, freshman year (Time 3 ¼ 7.4
months); September, sophomore year (Time 4 ¼ 11.8 months); December, sophomore year
(Time 5 ¼ 14.8 months); April, sophomore year (Time 6 ¼ 19.5 months); September,
junior year (Time 7 ¼ 23.9 months); and April, junior year (Time 8 ¼ 31.3 months).
Percent body fat
Bioelectrical impedance analysis (BodyStat, BioVant Systems, Detroit, MI, USA) was
used to measure percent body fat at all eight time points. BodyStat has been validated for
accuracy against other body composition assessment methods (Fuller et al. 1994, Ghosh
et al. 1997, Benton and Swan 2007). Since hydration status affects measurement accuracy,
participants were instructed not to drink caffeine or alcohol or engage in strenuous
exercise for at least 12 h, and not to eat for 2 – 4 h before assessments (National Institutes of
Health 1996). All assessments were conducted between 8 and 11 am.
Weight and height
Weight and height were assessed using a digital scale with an attached height rod
(Healthometer, model 500KL; Pelstar, Bridgeview, IL, USA). The scale’s accuracy was
International Journal of Health Promotion and Education 297
D ow
nl oa
de d
by [
E P
- IP
S W
IC H
], [
A m
an da
L ev
y] a
t 08
:0 3
05 M
ar ch
2 01
3
verified with external weights. Subjects were asked to wear similar lightweight clothing
at assessments and to remove shoes, belts, outerwear jackets/sweaters, and items from
their pockets.
Activity
Physical activity habits were assessed using a questionnaire that included questions from the
National College Health Risk Behavior Survey (Centers for Disease Control and Prevention
1995). These questions included (1) How many days per week do you participate in vigorous
physical activity? (Vigorous activities are those that cause you to sweat and breathe hard.);
(2) How many days per week do you participate in moderate physical activity? (Moderate
activities include activities such as walking or bicycling. Be sure to include walking or
biking to class, if applicable.); (3) How many days per week do you participate in
strengthening exercises? (Strengthening exercises include activities such as push-ups,
sit-ups, and weight lifting.) Immediately following each question regarding vigorous,
moderate, and strengthening exercises, participants were asked the number of minutes per
day spent in each activity. For both moderate and vigorous activity and strength training, the
number of days of subject participation was multiplied by the minutes of participation per
day and divided by 7, to obtain the hours/day spent in each of three activities.
Statistical analysis plan
Linear growth models were fitted, one for change in percent body fat and the other for
change in weight (Singer and Willett 2003). The observed variables for the growth model
were the measures of percent body fat at eight time points (described previously). Time
was centered at September, freshman year and the latent slope regression parameters were
fixed at 0, 2.7, 7.4, 11.8, 14.8, 19.5, 23.9, and 31.3 months from the beginning of freshman
year in September. The latent intercept is interpreted as the mean percent body fat value
when participants began college as freshmen. A positive slope indicates increase in percent
body fat over time, whereas a negative slope indicates decrease in percent body fat.
Predictors for change in percent body fat over time were collected in September, freshman
year (Time 0). The above analysis plan was also used to examine a separate series of
growth models for weight change over the same eight time points and its predictors
(gender, activity, strength training).
Results
Univariate and bivariate analyses
Means, standard deviations, and correlations for the four predictors (gender, moderate
activity, vigorous activity, and strength training) and outcomes (percent body fat: eight
time points; weight: eight time points) are shown in Tables 1 and 2. These variables were
normally distributed and their bivariate relationships with each other were linear. The
average weight, BMI, and percent body fat at Time 1 were 65 kg or 143 lbs (SD ^ 28),
22.6 kg/m 2
(SD ^ 3.5), and 19.6% (SD ^ 7.5), respectively.
All multivariate models were fit with Mplus, which allowed inclusion of respondents
with missing data by using full information maximum likelihood (FIML) estimation
(Little and Rubin 1987, Muthén and Muthén 2003). In FIML estimation with missing data,
observations are sorted into missing data patterns, and each parameter was estimated using
298 S.S. Gropper et al.
D ow
nl oa
de d
by [
E P
- IP
S W
IC H
], [
A m
an da
L ev
y] a
t 08
:0 3
05 M
ar ch
2 01
3
T a b
le 1
. M
e a n
s, st
a n
d a rd
d e v
ia ti
o n
s, a n
d c o
rr e la
ti o
n s
a m
o n
g p
e rc
e n
t b
o d
y fa
t a n
d p
h y
si c a l
a c ti
v it
y v
a ri
a b
le s
(N ¼
5 3
5 ).
1 2
3 4
5 6
7 8
9 1
0 1
1 1
2
1 .
T 1
% B
F a t
– 2
. T
2 %
B F
a t
0 .9
6 d
– 3
. T
3 %
B F
a t
0 .9
5 d
0 .9
6 d
– 4
. T
4 %
B F
a t
0 .9
4 d
0 .9
5 d
0 .9
6 d
– 5
. T
5 %
B F
a t
0 .9
4 d
0 .9
5 d
0 .9
6 d
0 .9
7 d
– 6
. T
6 %
B F
a t
0 .9
2 d
0 .9
3 d
0 .9
4 d
0 .9
5 d
0 .9
7 d
– 7
. T
7 %
B F
a t
0 .9
3 d
0 .9
4 d
0 .9
5 d
0 .9
6 d
0 .9
6 d
0 .9
7 d
– 8
. T
8 %
B F
a t
0 .9
0 d
0 .9
1 d
0 .9
0 d
0 .9
0 d
0 .9
1 d
0 .9
1 d
0 .9
2 d
– 9
. T
1 V
ig A
c t
2 0
.3 0
d 2
0 .3
0 d
2 0
.2 9
d 2
0 .2
8 d
2 0
.2 9
d 2
0 .2
7 d
2 0
.3 6
d 2
0 .3
3 d
– 1
0 .
T 1
M o
d A
c t
2 0
.1 5
d 2
0 .0
8 a
2 0
.1 0
b 2
0 .0
8 2
0 .0
4 2
0 .0
8 2
0 .1
3 2
0 .0
5 0
.3 3
d –
1 1
. T
1 S
T 0
.3 3
d 2
0 .3
5 d
2 0
.3 6
d 2
0 .3
3 d
2 0
.2 3
c 2
0 .3
3 d
2 0
.3 1
d 2
0 .2
6 d
0 .4
1 d
0 .1
3 c
– 1
2 .
F e m
a le
0 .7
5 d
0 .7
4 d
0 .7
3 d
0 .7
5 d
0 .6
4 d
0 .7
0 d
0 .6
3 d
0 .6
2 d
2 0
.3 4
d 2
0 .1
0 b
2 0
.4 0
d –
N 5
3 4
4 7
1 4
4 7
3 9
1 1
6 3
3 4
8 1
5 2
1 5
1 5
4 1
5 4
1 5
4 2
5 4
1 M
1 9
.3 2
1 9
.9 7
2 0
.0 9
2 0
.0 9
2 0
.7 5
2 0
.7 1
2 0
.7 3
2 1
.0 2
1 8
8 .1
4 2
3 5
.5 0
7 0
.9 8
0 .6
5 S
D 7
.5 9
7 .4
7 7
.6 1
7 .6
0 7
.6 3
7 .6
1 8
.3 7
7 .3
1 2
5 7
.0 3
2 6
2 .0
4 1
0 5
.2 4
0 .4
8
N o te
s: %
B F
a t,
p e rc
e n t
b o d y
fa t;
V ig
A c t,
v ig
o ro
u s
a c ti
v it
y ;
M o d A
c t,
m o d e ra
te a c ti
v it
y ;
S T
, st
re n g th
tr a in
in g .
G e n d e r
is c o d e d
a s
1 fo
r fe
m a le
a n d
0 fo
r m
a le
. a p ,
0 .1
0 ;
b p ,
0 .0
5 ;
c p ,
0 .0
1 ;
d p ,
0 .0
0 1
.
International Journal of Health Promotion and Education 299
D ow
nl oa
de d
by [
E P
- IP
S W
IC H
], [
A m
an da
L ev
y] a
t 08
:0 3
05 M
ar ch
2 01
3
T a b
le 2
. M
e a n
s, st
a n
d a rd
d e v
ia ti
o n
s, a n
d c o
rr e la
ti o
n s
a m
o n
g w
e ig
h t
a n
d p
h y
si c a l
a c ti
v it
y v
a ri
a b
le s
(N ¼
5 3
5 ).
1 2
3 4
5 6
7 8
9 1
0 1
1 1
2
1 .
T 1
W T
– 2
. T
2 W
T 0
.9 9
d –
3 .
T 3
W T
0 .9
8 d
0 .9
9 d
– 4
. T
4 W
T 0
.9 7
d 0
.9 8
d 0
.9 9
d –
5 .
T 5
W T
0 .9
6 d
0 .9
7 d
0 .9
8 d
0 .9
9 d
– 6
. T
6 W
T 0
.9 5
d 0
.9 6
d 0
.9 7
d 0
.9 8
d 0
.9 9
d –
7 .
T 7
W T
0 .9
7 d
0 .9
7 d
0 .9
8 d
0 .9
8 d
0 .9
9 d
0 .9
9 d
– 8
. T
8 W
T 0
.9 6
d 0
.9 6
d 0
.9 7
d 0
.9 7
d 0
.9 6
d 0
.9 6
d 0
.9 7
d –
9 .
T 1
V ig
A c t
0 .1
8 d
0 .1
9 d
0 .1
6 d
0 .1
8 d
0 .1
7 b
0 .2
0 d
0 .1
4 a
0 .1
7 b
– 1
0 .
T 1
M o
d A
c t
0 .0
2 2
0 .0
2 2
0 .0
5 2
0 .0
7 2
0 .1
3 a
2 0
.0 7
2 0
.1 1
2 0
.1 1
0 .3
3 d
– 1
1 .
T 1
S T
0 .2
3 d
0 .2
5 d
0 .2
4 d
0 .2
5 d
0 .1
8 b
0 .2
0 d
0 .1
2 0
.1 2
0 .4
1 d
0 .1
3 c
– 1
2 .
F e m
a le
2 0
.5 2
d 2
0 .5
3 d
2 0
.5 3
d 2
0 .5
5 d
2 0
.5 3
d 2
0 .5
3 d
2 0
.4 4
d 2
0 .5
2 d
2 0
.3 4
d 2
0 .1
0 b
2 0
.4 0
d –
N 5
4 1
4 7
1 4
4 7
3 9
1 1
6 3
3 4
8 1
5 2
1 5
1 5
4 1
5 4
1 5
4 2
5 4
1 M
1 4
1 .9
5 1
4 4
.1 1
1 4
5 .0
2 1
4 4
.4 0
1 4
6 .3
0 1
4 6
.2 3
1 4
9 .7
2 1
4 7
.4 8
1 8
8 .1
4 2
3 5
.5 0
7 0
.9 8
0 .6
5 S
D 2
8 .5
0 2
9 .5
8 2
9 .8
5 2
8 .9
5 3
2 .4
7 3
0 .9
9 3
6 .9
8 3
5 .6
7 2
5 7
.0 3
2 6
2 .0
4 1
0 5
.2 4
0 .4
8
N o
te s:
W T
, w
e ig
h t;
V ig
A c t,
v ig
o ro
u s
c ti
v it
y ;
M o
d A
c t,
M o
d e ra
te A
c ti
v it
y ;
S T
, st
re n
g th
tr a in
in g
. G
e n
d e r
is c o
d e d
a s
1 fo
r fe
m a le
a n
d 0
fo r
m a le
. a p ,
0 .1
0 ;
b p ,
0 .0
5 ;
c p ,
0 .0
1 ;
d p ,
0 .0
0 1
.
300 S.S. Gropper et al.
D ow
nl oa
de d
by [
E P
- IP
S W
IC H
], [
A m
an da
L ev
y] a
t 08
:0 3
05 M
ar ch
2 01
3
all available data for that particular parameter according to established guidelines (Muthén
and Muthén 2003).
Unconditional models
The first model fit was an unconditional linear growth model (with no predictors other than
time) to determine whether college students demonstrated systematic changes in percent
body fat over the first 3 years of college (eight time points). This linear model fit the data
adequately (x 2 /df ¼ 5; RMSEA ¼ 0.09, p ¼ 0.01; TLI ¼ 0.98), indicated by a x
2 /df ratio
of less than 5, root mean square error of approximate (RMSEA) less than 0.10, and a
Tucker – Lewis Index (TLI) greater than 0.90 (Wheaton et al. 1977). The change in percent
body fat was tested to see whether it was quadratic over time; it was not. At the beginning
of freshman year, students, on average, had 19.6% body fat (b0, Intercept ¼ 19.6,
p , 0.001) and increased an average of 0.06% body fat each month over the next 3 years
(b1, Slope ¼ 0.06, p , 0.001). Figure 1 illustrates the fitted trajectory of a prototypical
college-age student’s percent body fat over this period. On average, a college student has a
small but significant increase in percent body fat over the 3 years.
The second model fit was an unconditional linear growth model (a model with no
predictors other than time) to determine whether college-age students demonstrated
systematic changes in weight over 3 years (eight time points). This linear model fit the data
less well (x 2 /df ¼ 12; RMSEA ¼ 0.14, p , 0.001; TLI ¼ 0.95), but the TLI indicated it
was adequate. Change in body weight was not found to be quadratic over time. At the
beginning of freshman year, students, on average, weighed 65 kg (143 lbs) (b0 ¼ 143.2,
p , 0.001) and increased an average of 0.068 kg (0.15 lbs) each month over the next 3 years
(b1 ¼ 0.15, p , 0.001). Figure 2 illustrating the fitted trajectory of a prototypical college-
age student’s weight over this period shows that, on average, a college student has a small
but significant increase in weight from fall of freshman year until spring of junior year.
Conditional models
Significant variance existed in all of the growth parameters of the unconditional models
that could be predicted by the substantive predictors: gender, moderate and vigorous
Sept Dec April Sep Dec April Sep April
Month
10
15
20
25
30
% B
o d y
fa t
Figure 1. Change in percent body fat over the first 3 years of college.
International Journal of Health Promotion and Education 301
D ow
nl oa
de d
by [
E P
- IP
S W
IC H
], [
A m
an da
L ev
y] a
t 08
:0 3
05 M
ar ch
2 01
3
activity, and strength training. To do this, a series of nested hierarchical models was fit
testing whether physical activity (moderate – Step 1, vigorous – Step 2), strength training
(Step 3), gender (Step 4), and their interactions (Step 5) were significant in predicting
change in percent body fat or weight gain over time. Each predictor was retained if the
Dx 2
test indicated it was significant (Singer and Willett 2003). Table 3 presents the
Sept Dec April Sep Dec April Sep April
Month
125
135
145
155
165
175
W e ig
h t
Figure 2. Change in weight over the first 3 years of college.
Table 3. Model fit statistics for all the models fit (N ¼ 535).
Model x 2
df TLI a
SRMR b
RMSEA c
Dx 2
(Ddf) d
Unconditional Linear %body fat 167.6 31 0.98 0.06 0.09 Conditional Moderate activity (Step 1) 175.9 37 0.98 0.05 0.08 11.5** (2) Vigorous activity (Step 2) 180.9 43 0.98 0.05 0.08 45.2*** (2) Strength training (Step 3) 181.5 49 0.98 0.05 0.07 38.5*** (2) Female (Step 4) 190.4 55 0.98 0.04 0.07 567.5*** (2) Female £ moderate (Step 5) on slope and intercept
197.2 68 0.98 0.04 0.05 9.2* (3)
Female £ strength on slope Unconditional Linear weight 389.0 31 0.95 0.03 0.15 Conditional Vigorous activity (Step 1) 395.0 37 0.95 0.02 0.13 18.3*** (2) Strength training (Step 2) 402.7 43 0.95 0.02 0.12 18.8*** (2) Female (Step 3) 417.8 49 0.95 0.02 0.11 156.9*** (2) Female £ strength (Step 4) on slope and intercept
421.8 55 0.95 0.01 0.11 4.60 ,
(2)
, p , 0.10, *p , 0.05, **p , 0.01, ***p , 0.001.
a TLI: Tucker – Lewis Index should be greater than 0.90.
b SRMR: standardized root mean error should be #0.05.
c RMSEA: root mean square error of approximation should be #0.10.
d Testing the H0 that the added predictor is not significant in predicting growth. If the Dx
2 per change in degrees
of freedom is significant, then we can reject that H0 and say that the predictor is significant in predicting growth and we retain it in the fitted model.
302 S.S. Gropper et al.
D ow
nl oa
de d
by [
E P
- IP
S W
IC H
], [
A m
an da
L ev
y] a
t 08
:0 3
05 M
ar ch
2 01
3
models that were fit, their fit statistics, and the Dx 2
statistics that determine retention of the
variable added. Table 4 includes the unstandardized parameter estimates and standard
errors of the mean levels of the growth parameters and effects of the predictors on them for
the final fitted models as well as the amount of variance predicted in the intercepts and
slopes; these results are the ones customarily presented for linear growth models (Singer
and Willett 2003).
As seen in Table 4, the model does quite well in predicting variance in the intercept
of change in percent body fat and fairly well with change in weight. However, it does
not do well in predicting variance in either the slope of change in percent body fat or
weight. The final fitted models do provide very useful information about the trajectories
of change in percent body fat and weight over the first 3 years of college and how that
change is related to physical activity and gender. And, indeed, all of the predictors of
percent body fat and weight are significant in the prediction of these important markers
of the health of undergraduate students (Dx 2
statistics for every model in Table 4 are
significant).
Prototypical plot illustrations of parameter estimates
The effects of the predictors in the final fitted model (Table 4) on the growth parameters
for change in percent of body fat and weight over 3 years of college can best be illustrated
by ‘identifying a prototypical individual distinguished by particular predictor values
(Singer and Willett 2003).’ Meaningful values of the predictors were selected to substitute
into the fitted final model, obtaining the estimated value for the outcome (percent body fat
or weight), and plotting those trajectories, which will give trajectories that would be
typical for individuals in the population with those characteristics. The sample was not
divided into groups to illustrate the findings; the fitted ‘true’ or ‘population’ trajectories of
college students similar to those in our sample are presented. The meaningful values
chosen for the plots of prototypical individuals were 1.5 SDs above and below the mean
for moderate and vigorous activity and strength training. For gender the values of 0 for
male and 1 for female were used.
Table 4. Unstandardized parameter estimates and standard errors (in parentheses) of the final conditional growth models for percent body fat and weight change in undergraduates over first 3 years of college (N ¼ 535).
Percent body fat Weight
Intercept Slope Intercept Slope
Mean level 12.4*** (0.5) 0.04* (0.01) 161.8*** (2.6) 0.18*** (0.5) Effect of Moderate activity 0.1 (0.5) 0.03 (0.02) Vigorous activity 20.5 (0.4) 0.00 (0.01) 20.6 (1.9) 0.02 (0.04) Strength training 20.9 (0.9) 0.00 (0.03) 6.2 (5.6) 0.01 (0.11) Female 12.3*** (0.6) 0.02 (0.02) 229.4*** (2.9) 20.03 (0.06) Female £ moderate 21.5* (0.7) 20.03 (0.02) Female £ strength 20.09
, (0.05) 25.5 (10.0) 20.31
, (0.17)
R 2 ¼ Amount of variance
in the growth parameter explained by the predictors
59.8% 3.5% 27.2% 2.2%
, p , 0.10, *p , 0.04, ***p , 0.001.
International Journal of Health Promotion and Education 303
D ow
nl oa
de d
by [
E P
- IP
S W
IC H
], [
A m
an da
L ev
y] a
t 08
:0 3
05 M
ar ch
2 01
3
Overall summary
Percent body fat (%BF)
In the final model for change in percent body fat, on average, young adult males begin lower
than females at the beginning of college (Mean Interceptmale ¼ b0 ¼ 12.4%BF, p , 0.001;
Interceptfemale ¼ 12.4 [Mean Interceptmale] þ 12.3 [bfemale] ¼ 24.7%BF) on percent of
body fat. Males and females then increase or decrease (females) in their percent body fat
over the next 3 years at different rates, depending on the prototypical male or female whose
trajectory we examine (see Table 4). On average, however, for both males and females
significant growth in percent body fat over time does exist (Mean Slope: b1 ¼ 0.04%BF,
p , 0.05). Males who are vigorously exercising on entry to college have lower percent
Sept Dec
ST-strength training
MA-moderate activity
VA-vigorous activity
April Sep Dec April Sep April
Month
10
15
20
25
30
% B
o d y
fa t
No ST, No MA
No ST, High MA
High ST, High M A
High ST, No MA
Figure 3. Change in percent body fat for males with no vigorous activity predicted by moderate and vigorous activities and strength training over the first 3 years of college.
Sept Dec April Sep Dec April Sep April
Month
10
15
20
25
30
% B
o d y
fa t
No ST, No MA
No ST, High MA
High ST, High M A
High ST, No MA
ST-strength training
MA-moderate activity
VA-vigorous activity
Figure 4. Change in percent body fat for males with high vigorous activity predicted by moderate and vigorous activities and strength training over the first 3 years of college.
304 S.S. Gropper et al.
D ow
nl oa
de d
by [
E P
- IP
S W
IC H
], [
A m
an da
L ev
y] a
t 08
:0 3
05 M
ar ch
2 01
3
body fat, on average, than those who engage in no vigorous activity at entry. But, regardless
of their level of vigorous activity at college entrance, males have similar trajectories over
time that only differ slightly by the levels of strength training and moderate activity. Thus,
all the trajectories for change in percent body fat for prototypical males are very similar (see
Figures 3 and 4). The story for female undergraduates is more complex.
On average, vigorous activity for females has an effect, at the entry into college, which is
similar to that for males; those females who engage in high vigorous activity at the
beginning of college, on average, have lower percent body fat at that time. Strength training
for women has a moderating effect on their change in percent body fat (Slope Effect:
bfemale£strength ¼ 20.09%BF, p , 0.10); women who engage in strength training, regardless
of the level of their moderate or vigorous activity, have a decrease in percent body fat over
Sept Dec April Sep Dec April Sep April
Month
10
15
20
25
30
% B
o d y
fa t
No ST, No MA
No ST, High MA
High ST, High MA
High ST, No MA
ST-strength training
MA-moderate activity
Figure 5. Change in percent body fat for females with no vigorous activity predicted by moderate and vigorous activities and strength training over the first 3 years of college.
Sept Dec April Sep Dec April Sep April
Month
10
15
20
25
30
% B
o d y
fa t
No ST, No MA
No ST, High MA
High ST,High MA
High ST, No MA
ST-strength training
MA-moderate activity
Figure 6. Change in percent body fat for females with high vigorous activity predicted by moderate and vigorous activities and strength training over the first 3 years of college.
International Journal of Health Promotion and Education 305
D ow
nl oa
de d
by [
E P
- IP
S W
IC H
], [
A m
an da
L ev
y] a
t 08
:0 3
05 M
ar ch
2 01
3
time. Women who do not engage in strength training increase their percent body fat over
time. In addition, women who are already involved in moderate activity at the beginning of
college have lower percent body fat at that time (Intercept Effect: bfemale£moderate ¼ 21.5%-
BF, p , 0.05) (see Figures 5 and 6). For men, neither of these effects exists.
Weight (LB)
On average, males begin at a higher weight than females at the beginning of college (Mean
Interceptmale ¼ b0 ¼ 73.5 kg or 161.8 lbs, p , 0.001; Interceptfemale ¼ 73.5 kg or 161.8
[Mean Interceptmale] 229.4 [bfemale] ¼ 60.18 kg or 132.4 lbs). Males and females then
increase or decrease (females) their weight over the next 3 years at different rates,
Sept Dec April Sep Dec April Sep April
Month
125
135
145
155
165
175
W e ig
h t
No ST, No VA
No ST, High VA
High ST, High V A
High ST, No VA
ST-strength training
VA-vigorous activity
Figure 7. Change in weight for males predicted by vigorous activity and strength training over the first 3 years of college.
Sept Dec April Sep Dec April Sep April
Month
125
135
145
155
165
175
W e ig
h t
No ST, No VA No ST, High VA
High ST, High VA High ST, No VA
ST-strength training
VA-vigorous activity
Figure 8. Change in weight for females predicted by vigorous activity and strength training over the first 3 years of college.
306 S.S. Gropper et al.
D ow
nl oa
de d
by [
E P
- IP
S W
IC H
], [
A m
an da
L ev
y] a
t 08
:0 3
05 M
ar ch
2 01
3
depending on the prototypical male or female whose trajectory was examined (see
Table 4). On average, however, both males and females significantly increase their weight
over time (Mean slope: b1 ¼ 0.082 kg or 0.18 lbs, p , 0.001). Few differences exist for
males relating to their strength training or vigorous activity level. The majority of the
differences are at the entry into college. Males who engage in strength training have more
body weight at intercept and across time than males who do not engage in that activity (see
Figure 7). Once again, the story for female undergraduates is more complex. On average,
women who engage in strength training, although similar to women who do not at the
beginning of college, lose weight over time, regardless of their level of vigorous activity
(Slope Effect: bfemale£strength ¼ 2 0.14 kg or 2 0.31 lbs, p , 0.10; Mean Slope:
b1 ¼ 20.08 kg or 0.18 lbs; therefore effect of strength training on change in weight is
0.18mean 2 0.31slope effect ¼ 20.059 kg or 20.13 lbs per month decrease) (see Figure 8).
Discussion
Gains in weight and body fat during the freshman year of college are well documented as
are factors associated with these gains. In the few studies that have examined changes
during the sophomore year, compared to the freshman year, college students were found to
continue to gain weight although body composition changes were more favorable with
males and females gaining less fat and females also gaining more fat-free mass (Hull et al.
2007, Gropper et al. 2011). Unique to the literature is this 3-year study on college students
examining not only weight and body composition changes but also physical activity
habits. These results showed that, on average, males and females exhibited significant
increases in weight and percent body fat. However, for females, participation in strength
training was associated with loss of weight and percent body fat, while females not
engaged in strength training exhibited increased weight and percent body fat over the 3
years. Males, while having similar trajectories for percent body fat change over time,
differed slightly by the levels of strength training and moderate activity; however, males
engaged in strength training gained more weight across time than did those males who did
not engage in that activity.
The reasons for this study’s observed gender differences are not clear, but may relate, in
part, to higher anabolic hormone levels and the longer growth spurt experienced by males
versus females. Such differences promote greater body weight and fat-free mass gains
among males than females as observed in this study. Differences because of participation in
exercise also may have contributed. Males in this sample population were found to have
met vigorous-intensity physical activity and resistance training recommendations
significantly more often than females (Newell 2011), a finding similar to other studies
(Huang et al. 2003, Buckworth and Nigg 2004, Mestek et al. 2008). Additionally, the
observed differences may be associated with initial differences in physical fitness. Mifsud
et al. (2009) found that males entering college with a greater level of physical fitness gained
more fat than those who were less physically fit. Lastly, inter-individual variations in body
composition, especially fat mass accretion, in response to the duration and intensity of
physical activity also may be responsible (Venables et al. 2005, Barwell et al. 2009).
The beneficial changes in body composition observed with participation in strength
training in this study are consistent with the results of other studies that have directly
examined the effects of strength training on body composition. Such studies generally report
that strength training stimulates fat-free mass accretion and decreases fat mass and percent
body fat (Poehlman et al. 2002, Sillanpää et al. 2008, Hanson et al. 2009, Kemmler et al.
2010). Aerobic exercise also promotes reductions in body weight, BMI, and fat mass, and
International Journal of Health Promotion and Education 307
D ow
nl oa
de d
by [
E P
- IP
S W
IC H
], [
A m
an da
L ev
y] a
t 08
:0 3
05 M
ar ch
2 01
3
gains in fat-free mass; however, the magnitude of the effects typically varies depending
upon exercise intensity, duration, and frequency (Donnelly et al. 2003, Jakicic et al. 2003,
Slentz et al. 2004, Irving et al. 2008, Stasiulis et al. 2010).
The strengths of this study include its relatively large sample size which was followed
for 3 years and its sophisticated statistical analysis. Further, this is one of the first studies
providing an intervention strategy for females to help minimize body weight and fat gains
normally accrued during the college years. However, a limitation to this study is that it was
conducted at a public university, and thus the results may not be applicable to those not
attending a university or those attending private universities. Subject honesty was also
required and activity logs were not kept to verify self-reported activity.
Conclusion
The results of this study underscore the benefits of physical activity, especially strength
training, for college females as a means to reduce or prevent body weight and fat gains
during the first 3 years of college. Given it is this young adult group that has experienced
the greatest increases in overweight/obesity nationally (Mokdad et al. 1999), effective
strategies to minimize these gains may help to reduce the growing obesity epidemic.
Acknowledgements
This research was supported by U.S. Department of Agriculture, Alabama Agricultural Experiment Station (AAES) projects 013-020 and 07-020, and an AAES Initiative Grant. No financial disclosures were reported by the authors of this paper.
References
Auburn University Office of Institutional Research and Assessment, 2007 – 2008. Available from: https://oira.auburn.edu/ [Accessed 15 April 2010].
Barwell, N.D., et al., 2009. Individual responsiveness to exercise-induced fat loss is associated with change in resting substrate utilization. Metabolism, Clinical and Experimental, 58 (9), 1320 – 1328.
Benton, M.J. and Swan, P.D., 2007. Comparison of air displacement plethysmography and whole body bioelectrical impedance in middle aged women. Medicine and Science in Sports and Exercise, 39, S372.
Bray, S.R. and Born, H.A., 2004. Transition to university and vigorous physical activity: implication for health and psychological well-being. American Journal of College Health, 52 (4), 181 – 188.
Buckworth, J. and Nigg, C., 2004. Physical activity, exercise, and sedentary behavior in college students. American Journal of College Health, 53 (1), 28 – 34.
Caspersen, C.J., Pereira, M.A., and Curran, K.M., 2000. Changes in physical activity patterns in the United States, by sex and cross-sectional age. Medicine and Science in Sports and Exercise, 32 (9), 1601 – 1609.
Center for Disease Control and Prevention, 1995. Youth behavioral risk surveillance: National College Health Risk Behavior Survey. Available from: http://www.cdc.gov/mmwr/preview/mm wrhtml/00049859.htm#00002795.htm [Accessed 10 April 2011].
Donnelly, J.E., et al., 2003. Effects of a 16-month randomized controlled exercise trial on body weight and composition in young, overweight men and women. Archives of Internal Medicine, 163 (11), 1343 – 1350.
Edmonds, M.J., et al., 2008. Body weight and percent body fat increase during the transition from high school to university in females. Journal of the American Dietetic Association, 108 (6), 1033 – 1037.
Fuller, N.J., Sawyer, M.B., and Elia, M., 1994. Comparative evaluation of body composition methods and predictions, and calculation of density and hydration fraction of fat free mass. International Journal of Obesity, 18 (7), 503 – 512.
308 S.S. Gropper et al.
D ow
nl oa
de d
by [
E P
- IP
S W
IC H
], [
A m
an da
L ev
y] a
t 08
:0 3
05 M
ar ch
2 01
3
Ghosh, S., et al., 1997. Body composition at the bedside. European Journal of Gastroenterology and Hepatology, 9 (8), 783 – 788.
Gropper, S.S., et al., 2009. The freshman 15-A closer look. Journal of the American College of Health, 58 (3), 223 – 231.
Gropper, S.S., et al., 2011. Body composition changes during the first two years of university. Preventive Medicine, 52 (1), 20 – 22.
Hajhosseini, L., et al., 2006. Changes in body weight, body composition and resting metabolic rate (RMR) in first-year university freshmen students. Journal of the American College of Health, 25 (2), 123 – 127.
Hanson, E.D., et al., 2009. Effects of strength training on physical function: influence of power, strength, and body composition. Journal of Strength and Conditioning Research, 23 (9), 2627 – 2637.
Huang, T.T., et al., 2003. Assessing overweight, obesity, diet, and physical activity in college students. Journal of the American College of Health, 52 (2), 83 – 86.
Hull, H.R., et al., 2007. Characterization of body weight and composition changes during the sophomore year of college. BMC Women Health, 7 (17), 21 – 27.
Irving, B.A., et al., 2008. Effect of exercise training intensity on abdominal visceral fat and body composition. Medicine and Science in Sports and Exercise, 40 (11), 1863 – 1872.
Jakicic, J.M., et al., 2003. Effect of exercise duration and intensity on weight loss in overweight, sedentary women. A randomized trial. Journal of the American Medical Association, 290 (10), 1323 – 1330.
Kasparek, D.G., et al., 2008. Selected health behaviors that influence college freshman weight change. Journal of the American College of Health, 56 (4), 437 – 444.
Katzmarzyk, P.T., 2006. Physical activity status and chronic diseases. In: L.A. Kaminsky, ed. ACSM’s resource manual for guidelines for exercise testing and prescription. 5th ed. Philadelphia, PA: Lippincott Williams & Wilkins, 122 – 135.
Kemmler, W., et al., 2010. Exercise, body composition, and functional ability: a randomized control trial. American Journal of Preventive Medicine, 38 (3), 279 – 287.
Little, R.J.A. and Rubin, D.B., 1987. Statistical analysis with missing data. New York: Wiley. Mestek, M.L., Plaisance, E., and Grandjean, P., 2008. The relationship between pedometer-
determined and self-reported physical activity and body composition variables in college-aged men and women. Journal of the American College of Health, 57 (1), 39 – 44.
Mifsud, G., Duval, K., and Doucet, É., 2009. Low body fat and high cardiorespiratory fitness at the onset of the freshman year may not protect against weight gain. British Journal of Nutrition, 101 (9), 1406 – 1412.
Mokdad, A.H., et al., 1999. The spread of the obesity epidemic in the United States, 1991-1998. Journal of the American Medical Association, 282 (16), 1519 – 1522.
Muthén, L.K. and Muthén, B.O., 2003. Mplus Version 6 short courses: growth modeling with latent variables using Mplus. Available from: http://www.statmodel.com [Accessed 30 September 2010].
National Institutes of Health, 1996. Bioelectrical impedance analysis in body composition measurements. National Institute of Health Technology Assessment Conference. American College of Clinical Nutrition, 64 (suppl 1), 524S – 532S.
Newell, F.H., 2011. Physical activity and body composition changes during the first two years of college. Master’s Thesis. Auburn University, AL.
Poehlman, E.T., et al., 2002. Effects of endurance and resistance training on total daily energy expenditure on young women: a controlled randomized trial. Journal of Clinical Endocrinology and Metabolism, 87 (3), 1004 – 1009.
Pullman, A.W., et al., 2009. Effect of the transition from high school to university on anthropometric and lifestyle variables in males. Applied Physiology and Nutrition Metabolism, 34 (2), 162 – 171.
Racette, S.B., et al., 2008. Changes in weight and health behaviors from freshman through senior year of college. Journal of Nutrition Education and Behavior, 40 (1), 39 – 42.
Romero-Corral, A., et al., 2010. Normal weight obesity: a risk factor for cardiometabolic dysregulation and cardiovascular mortality. European Heart Journal, 31 (6), 737 – 746.
Schoenborn, A. and Adams, P.F., 2010. Health behaviors of adults: United States, 2005 – 2007. National Center for Health Statistics. Vital Health Statistics, 10 (245), 1 – 132.
Sillanpää, E., et al., 2008. Body composition and fitness during strength and/or endurance training in older men. Medicine and Science in Sports and Exercise, 40 (5), 950 – 958.
International Journal of Health Promotion and Education 309
D ow
nl oa
de d
by [
E P
- IP
S W
IC H
], [
A m
an da
L ev
y] a
t 08
:0 3
05 M
ar ch
2 01
3
Singer, J.D. and Willett, J.B., 2003. Applied longitudinal data analysis: modeling change and event occurrence. New York: Oxford University Press, 60.
Slentz, C.A., et al., 2004. Effects of the amount of exercise on body weight, body composition, and measures of central obesity. STRRIDE: a randomized control study. Archives of Internal Medicine, 164 (1), 31 – 39.
Stasiulis, A., et al., 2010. Aerobic exercise-induced changes in body composition and blood lipids in young women. Medicina, 46 (2), 129 – 134.
U.S. Department of Health and Human Services, 2008. Physical activity guidelines for Americans. Hyattsville, MD: US Department of Health and Human Services. Available from: http://www. health.gov/paguidelines [Accessed 11 January 2011].
Venables, M.C., Achten, J., and Jeukendrup, A.E., 2005. Determinants of fat oxidation during exercise in healthy men and women: a cross-sectional study. Journal of Applied Physiology, 98 (1), 160 – 167.
Wengreen, H.J. and Moncur, C., 2009. Change in diet, physical activity, and body weight among young adults during the transition from high school to college. BMC Nutrition Journal, 8, 32 – 39.
Wheaton, B., Muthen, B., Alwin, D., and Summers, G., 1977. Assessing reliability and stability in panel models. In: D. Heise, ed. Sociological methodology. San Francisco: Jossey-Bass, 84 – 136.
310 S.S. Gropper et al.
D ow
nl oa
de d
by [
E P
- IP
S W
IC H
], [
A m
an da
L ev
y] a
t 08
:0 3
05 M
ar ch
2 01
3
Copyright of International Journal of Health Promotion & Education is the property of Institute of Health
Promotion & Education and its content may not be copied or emailed to multiple sites or posted to a listserv
without the copyright holder's express written permission. However, users may print, download, or email
articles for individual use.
