Body Image 3 (2006) 121–129

www.elsevier.com/locate/bodyimage

The relationship of drive for muscularity to sociocultural factors,

self-esteem, physical attributes gender role, and
social comparison in middle school boys
Linda Smolak *, Jonathan A. Stein
Department of Psychology, Kenyon College, Gambier, OH 43022, USA
Received 6 July 2005; received in revised form 17 March 2006; accepted 20 March 2006

Abstract
This study examines the relationship of three sociocultural factors—media influence, peer teasing, and parent teasing/comments
and three potential moderator variables—self-esteem, social comparison, and endorsement of male strength and athleticism—to
drive for muscularity in middle school boys. There were 287 seventh and eighth grade boys who completed a questionnaire
measuring these variables as well as body mass index (BMI) and pubertal status. Results indicated that media influence and male
physical attributes endorsement were particularly important correlates of drive for muscularity. These findings have implications for
programs designed to prevent body dissatisfaction among adolescent boys.
# 2006 Elsevier Ltd. All rights reserved.

Keywords: Drive for muscularity; Gender role; Media; Peer teasing; Parent teasing/comments

Introduction Compared to women, men may also be somewhat more

A decade ago, body image researchers frequently
limited their samples to girls and women on the
assumption that boys and men were not particularly
likely to suffer from serious body dissatisfaction
problems. Since then, research has clearly established
that body dissatisfaction is indeed a problem for a
substantial percentage of boys and men but that male
body concerns differ from those of girls and women
(see, e.g., Cafri et al., 2005; McCabe & Ricciardelli,
2004b; Ricciardelli & McCabe, 2004 for reviews).
Specifically, males are considerably more likely to
want to increase their size, particularly in terms of
muscularity, while females typically want to get thinner
in order to match the culturally defined body ideals.
* Corresponding author. Tel.: +1 740 427 5374;
fax: +1 740 427 5237.
E-mail address: smolak@kenyon.edu (L. Smolak).
invested in how well their bodies function, especially in
terms of athletics and fitness, than in appearance. Thus
men may engage in more exercise and weightlifting
than dieting in attempting to influence their body shapes
relative to what women do (e.g., Grogan & Richards,
2002). These findings suggest that body image is a
gendered construct and that the meaning of body image
differs for men and women (McCreary, Sasse, Saucier,
& Dorsch, 2004; Smolak & Murnen, 2001, 2004).
Gender role, then, might be expected to be related to
body image. Indeed, McCreary et al. (2004) have
recently demonstrated that endorsement of masculine
behaviors and attitudes is related to higher levels of
drive for muscularity in college age men. Furthermore,
qualitative research indicates that even young boys
show the gendered patterns of body image suggesting
that gender role may be operative in influencing body
dissatisfaction in children and adolescents (Grogan &
Richards, 2002). However, the relationship of gender

1740-1445/$ – see front matter # 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bodyim.2006.03.002
122 L. Smolak, J.A. Stein / Body Image 3 (2006) 121–129
role to body dissatisfaction, and more specifically to the
apparently masculine-valenced drive for muscularity,
has not yet been investigated in adolescents. One goal of
the current study is to examine the relationship of
gender role endorsement, particularly as it relates to
physical attributes, to drive for muscularity among
middle school boys. Middle school is a particularly
interesting time to investigate the impact of gender role
endorsement because boys’ investment in masculinity
may intensify at this time (e.g., Galambos, Almeida, &
Petersen, 1990). In the present study, drive for
muscularity may be defined as attitudes and behaviors
related to a big, lean, and muscular body type that is
culturally sanctioned for American males (McCreary &
Sasse, 2000) and is operationalized as score on the
Drive for Muscularity Scale (DMS; McCreary & Sasse,
2000).
More importantly, the role of gender in drive for
muscularity has not been investigated. It is possible that
there is simply a direct relationship, i.e., that investment
in masculinity is positively correlated with drive for
muscularity. McCreary, Saucier, and Courtenay (2005)
found that unmitigated agency, male sex-typed behavior,
and sex specific behaviors were positively correlated
with DMS in college men. Indeed, the relationship
between gender-role behavior and DMS was relatively
strong; along with participant’s gender, gender-role
behavior accounted for over 25% of the variance in DMS.
However, it is also possible that gender role
moderates the effects of sociocultural factors. Several
studies have now suggested that sociocultural factors
such as investment in media images, peer teasing and
comments, and parental pressures may be at least
correlated with drive for muscularity or dissatisfaction
with muscularity among adolescent boys (e.g., Field
et al., 2005; Jones, 2004; Ricciardelli & McCabe, 2001,
2003; Stanford & McCabe, 2005). However, as has been
true of girls and drive for thinness (e.g., Groesz, Levine,
& Murnen, 2002; Keery, van den Berg, & Thompson,
2004), there are likely to be mediators or moderators of
these sociocultural influences. In other words, not every
boy who is exposed to these media images of
muscularity or peer comments about muscularity will
show a negative effect (i.e., body image issues).
Personal characteristics that focus boys on these media,
peer, and parent influences may help explain the
individual differences. Indeed, several researchers have
found that characteristics such as social comparison,
self-esteem, or internalization of media ideals partially
mediate or moderate the effects of sociocultural
influences on either body dissatisfaction or, relatedly,
use of muscle building techniques, by adolescent boys
(e.g., Ricciardelli & McCabe, 2001; Smolak, Murnen,
& Thompson, 2005). In the case of the personal
characteristic of gender role investment, it is possible
that boys who are invested in the male gender role will
be particularly susceptible to sociocultural messages
about muscularity because those messages fundamen-
tally tell the boy how to be the ‘‘ideal’’ man. In the
present study, both the direct and moderating relation-
ships of gender to DMS are investigated. The measure
of gender role used here emphasizes physical strength
and athleticism.
The present study examined the influence of three
sociocultural factors—perceived influence of media
images, peer teasing, and parental teasing and com-
ments—on drive for muscularity in middle school
(approximately ages 11–13 years old) boys. Models of
the development of body image problems in girls have
often focused on these factors (e.g., Stice, 1994;
Thompson, Heinberg, Altabe, & Tantleff-Dunn, 1999).
Recent models of body image in boys and men have also
suggested a prominent role for these sociocultural
factors (e.g., Cafri et al., 2005; Ricciardelli & McCabe,
2004; Smolak et al., 2005). Empirical data support the
possibility that these factors are at least correlated with
body image problems in boys and men (Ricciardelli &
McCabe, 2004).
Three potential moderators of sociocultural factors
were also investigated. Two of these – self-esteem and
social comparison – have been investigated previously
(Jones, 2004; Ricciardelli & McCabe, 2001; Smolak
et al., 2005). Social comparison based on appearance
has been a particular focus because the Tripartite
Influence Model of Body Dissatisfaction of Eating
Disturbance (Thompson et al., 1999) suggests it is a
mediator. Findings have been mixed, possibly because
of the use of different measures for the dependent
variables as well as regression equations that have
employed various predictor variables. Smolak et al.
(2005) reported that social comparison partially
mediated the relationships between media influence,
peer influence, and parent teasing and use of muscle
building techniques. However, Jones (2004) reported no
such role when investigating the longitudinal relation-
ship of peer norms to body dissatisfaction (measured
with a version of the EDI-Body Dissatisfaction Scale
altered for use with boys). Ricciardelli and McCabe
(2001) reported that self-esteem moderated the effects
of media pressure to gain muscle on general body
dissatisfaction. However, self-esteem did not moderate
parental or peer pressure concerning muscularity in
affecting general body dissatisfaction. Given that
McCreary and Sasse (2000) showed a main effect of
 L. Smolak, J.A. Stein / Body Image 3 (2006) 121–129
self-esteem on muscle dissatisfaction, it may still be an
important factor. It is, therefore, included as a potential
moderator in this study.
The third potential moderator is an aspect of gender
role, male physical attributes endorsement. There are a
variety of possible components of gender role that could
be investigated; McCreary et al.’s (2005) research with
college men indicated that behaviors, attitudes, and
gender role conflict were all related to DMS. More
specifically, McCreary et al. (2005) found that
endorsement of the traditional male role, including
‘‘toughness’’, as well as investment in power and
competition were related to DMS. Mahalik et al. (2003)
reported that in a sample of college men, the ‘‘winning’’
subscale of the Conformity to Masculine Norms
Inventory was significantly correlated with DMS
(r = .33) as was the total score on their Conformity
to Masculine Norms Inventory (r = .29). Grogan and
Richards (2002) found that boys and men want bodies
that look athletic. Thus, the data seem to indicate the
importance of gender-stereotyped behaviors that indi-
cate physical strength and athletic ability in drive for
muscularity. Such behaviors formed the basis of the
gender scale used in the present study.
Thus, three sociocultural variables—perceived influ-
ence of media images, parental teasing and comments,
and peer teasing and the three potential moderators—
self-esteem, social comparison, and male physical
attributes endorsement were the focus of this study.
Since the sample is comprised of middle school boys,
we also considered the possible influence of pubertal
level and BMI. Cafri et al. (2005) suggested that both
BMI and pubertal timing might be related to
muscularity-related attitudes and behaviors. Boys
who are small for their age, as reflected by low BMI
or relatively late pubertal development, might be
expected to be dissatisfied with their bodies and to be
interested in increasing their muscle mass. On the other
hand, boys who are early maturers, and thus have
relatively higher levels of circulating testosterone,
might be expected to engage more in muscle building
because they can actually see more benefits of it. Data
addressing the relationship between pubertal status and
DMS or the use of muscle-building strategies are
actually mixed. For example, in one longitudinal study,
McCabe and Ricciardelli (2004a) found that early, but
not late, maturing boys were more likely to engage in
muscle building activities although late maturing boys
were more likely to express body dissatisfaction. In
another longitudinal study, however, Ricciardelli and
McCabe (2003) found no relationship between pubertal
timing and muscle building.
123
The hypotheses investigated in the present study
were:
1. Boys who are more invested in media images of
muscularity will show higher DMS scores (Field
et al., 2005; Smolak et al., 2005).
2. Boys who are teased by peers about their size will
have higher DMS scores (Ricciardelli & McCabe,
2004; Smolak et al., 2005).
3. Boys who are teased by their parents about their size
or whose parents comment on their size will have
higher DMS scores (Ricciardelli & McCabe, 2004;
Smolak et al., 2005).
4. Boys who have relatively low BMIs will have high
DMS scores.
5. Boys who are relatively late in pubertal development
will have high DMS scores (Cafri et al., 2005).
6. The relationship between each of the sociocultural
influences (media, peer teasing, and parent com-
ments) and DMS will be moderated by appearance
social comparison such that boys who are high on
each of the sociocultural factors and are high on
social comparison will be high on DMS (Smolak
et al., 2005; Thompson et al., 1999).
7. The relationship between each of the sociocultural
influences (media, peer teasing, and parent com-
ments) and DMS will be moderated by self-esteem
such that boys who are high on each of the
sociocultural factors and are high on self-esteem
will be high on DMS (Ricciardelli & McCabe, 2001).
8. The relationship between each of the sociocultural
influences (media, peer teasing, and parent com-
ments) and DMS will be moderated by male physical
attributes endorsement such that boys who are high
on each of the sociocultural factors and are high on
male physical attributes endorsement will be high on
DMS (Mahalik et al., 2003; McCreary et al., 2005).
Method
Participants
Participants in this study were 287 7th (n = 156) and
8th (n = 129) grade boys from two middle schools in
rural Ohio (two students did not indicate a school
grade). Ages ranged from 11 to 15 years old (M = 12.9,
SD = .72). The subject pool was overwhelmingly White
(n = 227), although there were a few African American
(n = 9), Hispanic (n = 4), Asian (n = 3), and American
Indian (n = 6) participants as well as boys who
either indicated other (n = 13) ethnicities or did not
identify (n = 26) an ethnicity. The average height was
124 L. Smolak, J.A. Stein / Body Image 3 (2006) 121–129
64.54 inches with a range of 51–82, SD = 4.34. The
average weight was 124 lbs (range: 59–252, SD = 30.54).
Average BMI was 20.87 with SD = 3.79, range =
12.44–33.7. Consistent with the demographics of
the schools, students were assumed to be primarily
working and middle class.
Measures
All measures were self-report. Only measures used
in the present study are described here although some
other questions were asked. Boys were asked to report
demographic information, weight, and height.
Drive for muscularity
Drive for muscularity was measured using the Drive
for Muscularity Scale (DMS; McCreary & Sasse, 2000,
2002). The DMS represents one’s perception that he/she
is not muscular enough, and therefore must add muscle
mass to the body. The scale is a 15-question, Likert
scale, self-report instrument, designed for both men and
women. A higher score on the DMS indicates a higher
drive for muscularity. The DMS has been shown to have
good internal consistency, face validity, convergent
validity, discriminant validity, and test–retest stability
with high school boys and adult men (Cafri & Thompson,
2004; McCreary & Sasse, 2000, 2002). McCreary et al.
(2004) have also demonstrated that, when used with
men (as opposed to with women), the DMS has two
subscales: behavioral and attitudinal. The behavioral
subscale contains the items that describe what men do
in order to obtain larger muscles or become bigger
overall (e.g., ‘‘I use protein or energy supplements’’; ‘‘I
feel guilty if I miss a weight-training session’’). The
attitudinal subscale taps a body image dimension that is
focused on muscularity (e.g., ‘‘I think that my arms are
not muscular enough’’, ‘‘ I wish I were more muscular’’).
In the current sample, alpha for the total DMS was .93.
Perceived influence of media
The Media Influence Scale for Adolescent Boys
(MISAB; Haselhuhn et al., 2001) was used to assess
perceived influence of muscular images in the media.
Questions pertained to interest in looking like the
muscle ideal as portrayed in magazines, TV, and movies
as well as to obtaining information from the media
sources as to how to build muscles. Sample questions
included ‘‘Reading magazines makes me want to have
bigger muscles’’, ‘‘I compare my muscles to the
muscles of athletes in certain magazines’’, and ‘‘I can
tell if I am muscular enough by looking at the athletes
on TV’’. Answers ranged from 1 = definitely disagree to
5 = definitely agree. Items were summed to create a
total score for each individual. A higher score indicates
greater influence from the media. Haselhuhn et al.
(2001) presented evidence of strong internal consis-
tency and convergent validity in two samples of 11–13
years old boys. In the present sample, a = .90, n = 271.
Peer teasing
Peer teasing was measured by a three-question scale.
The questions asked about teasing and comments by
‘‘peers and classmates’’ about appearance, weight, and
muscularity. The questions were part of the scale used
by Smolak et al. (2005) and were adapted from
questions developed for use with girls by Levine,
Smolak, and Hayden (1994). Boys were asked to rate
how often (1 = never to 5 = very often) friends or
classmates teased them on specific categories (appear-
ance, weight, muscularity). Items were summed to
calculate a total score for each individual. A higher
score indicates more frequent teasing. Internal con-
sistency in this sample was good, a = .75, n = 281.
Parent teasing
In order to determine the effect of parent teasing, the
Parent Comments Scale (Smolak et al., 2005) was used.
Three questions about mother teasing and comments
and identical questions about father teasing and
comments were asked. Boys rated how often (1 = never
to 5 = very often) mothers and fathers teased them about
three categories (appearance, weight, muscularity).
Items were summed to calculate a total score for each
individual. A higher score indicates more frequent
teasing. Internal consistency for this sample was good,
a = .81, n = 277.
Social comparison
Social comparison was measured using five items
from the Physical Appearance Comparison Scale
(Thompson, Heinberg, & Tantleff, 1991). The questions
were rated on a five-point scale (1 = never to 5 = very
often) assessing how frequently boys compared their
bodies to those of others in various social situations.
Scores were summed to create a total score for each
individual. A higher total score indicates a greater
inclination to engage in social comparison. This scale
has been used with middle school boys (Smolak et al.,
2005). For the present sample, a = .79, n = 277.
Self-esteem
The 10-question Rosenberg self-esteem scale was
used (Rosenberg, 1965). This scale measured the boys’
general self-esteem. The scale is widely used with
 L. Smolak, J.A. Stein / Body Image 3 (2006) 121–129
adolescents and has well-established validity and
reliability. A higher score indicates higher self-esteem.
In the present sample, a = .80, n = 255.
Male physical attributes investment scale
An eight-item scale was created for this study. Boys
were asked to rate statements concerning the impor-
tance of boys’ athletic and physical superiority (e.g., ‘‘It
is important for guys to be able to physically defend
themselves’’; ‘‘It would be embarrassing for a guy if
most girls could run faster than he could’’; ‘‘Guys
should be able to throw a ball farther than most girls
can’’) on a five-point scale (1 = strongly disagree;
5 = strongly agree). Items were summed to create a total
score with a higher score indicating greater investment
in traditional masculine gender role concerning strength
and athleticism. In the present sample, a = .85, n = 267.
Puberty
Pubertal status was assessed using the Pubertal
Development Scale (Petersen, Crockett, Richards, &
Boxer, 1988). This widely used scale has well-
established validity and reliability. Each question
(one pertaining to growth spurt, skin changes, facial
hair development, voice change) was scored on a scale
of 1 (has not yet begun) to 5 (is completed). Scores were
added together in order to calculate a total score for each
individual. A higher score indicated that the boy was
further along in the development of puberty. In the
present sample, a = .73, n = 267.
BMI
BMI was calculated based on the boys’ self-reported
height and weight. BMI = ((weight in pounds)/(height
in inches)2)(703).
Procedure
Both middle schools requested and the human
subjects review approved passive parental consent.
Fewer than six parents at both schools returned the
letters refusing participation for their children. The boys
were also told, orally and in written instructions, that
they could refuse to participate. Fewer than six boys at
both schools refused participation although many
skipped one or more individual questions on the survey,
as they were told they could do. Thus, for many of the
analyses, the entire sample of boys is not available.
Boys were tested in groups of 20–40 with researchers
and/or teachers present. Testing took place during
school hours at a time deemed appropriate by the boys’
teachers and school administrators.
125
Results
Preliminary analyses
Factor analysis of DMS
The DMS has not been used with a sample of middle
school boys before, although it has been used with older
adolescents. Therefore, a factor analysis was run to see
whether the factor structure established with a sample
consisting of high school and adult males (McCreary
et al., 2004) could be replicated. Because this was a
replication, the same factor analysis technique used by
McCreary et al. (2004) was employed. A principal
components analysis was performed with a varimax
rotation. The KMO statistic was .91, indicating
excellent sampling adequacy.
Factors were defined based on eigenvalues greater
than 1. Three factors emerged, accounting for 75.07%
of the variance. The first factor (eigenvalue = 7.88;
a = .93, n = 253) tapped primarily behavioral items (see
Table 1 for factor loadings) except for behavior related
to lifting weights. Factor two (eigenvalue = 2.37;
a = .92, n = 275) consisted of three questions about
body parts (arms, chest, legs) being more muscular and
so is attitudinal. Factor three (eigenvalue = 1.01;
a = .89, n = 257) consists of some attitudinal questions
(e.g., ‘‘I wish I were more muscular’’) and the weight
lifting question.
Table 1 shows the factor loadings. A question was
considered to load significantly on a factor if its rotated
loading exceeded .60.
This three-factor solution differs from the two-factor
solution reported by McCreary et al. (2004). It appears
that, at least in the present sample, weight lifting may be
a distinct endeavor from other forms of muscle building.
This makes some sense intuitively since weight lifting
may be commonly associated with normal training for
sports whereas some of the other methods (e.g., protein
shakes, steroids) may seem more extreme or be less
accessible to middle school boys.
However, there are two problems with the three-
factor model. First, the third factor barely meets the
criteria for factor status, although it is theoretically
interesting. Secondly, and most importantly, there is a
lack of independence of several of the items across the
factors. If we use the criteria that an item must load at
least .60 but cannot load .30 on another factor to be
included on a factor, six items will not be permitted to
load on a factor. This means that just over one-third of
the items will be eliminated, a procedure that will
substantially affect the interpretability of factors two
and three in particular. Therefore, only the total DMS
126 L. Smolak, J.A. Stein / Body Image 3 (2006) 121–129

Table 1
Rotated factor loadings for Drive for Muscularity Scalea
Scale item 1 2 3
1. I wish I were more muscular .10 .50 .62
2. I lift weights to build more muscle .22 .11 .81
3. I use protein or energy supplements .84 .13 .18
4. I drink weight gain or protein shakes .87 .16 .12
5. I try to consume as many calories as I can in a day .78 .21 .17
6. I feel guilty if I miss a weight-training session .76 .08 .32
7. I think I would feel more confident if I had more muscle mass .25 .53 .64
8. Other people think I work out with weights too often .77 .07 .38
9. I think I would look better if I gained 10 pounds .36 .30 .68
10. I think about taking anabolic steroids .80 .32 -.02
11. I think I would feel stronger if I gained a little more muscle mass .18 .54 .65
12. I think that my weight-training schedule interferes with other aspects of my life .79 .18 .21
13. I think that my arms are not muscular enough .17 .89 .27
14. I think that my chest is not muscular enough .17 .90 .21
15. I think that my legs are not muscular enough .30 .82 .27
a
Drive for Muscularity Scale (McCreary & Sasse, 2000).

score was used in the regression analysis although the
descriptive data and correlations for the factors are
reported.
Descriptive statistics and correlations
There were no school or grade differences on any of
the major variables so data were combined for all
analyses. Means and standard deviations as well as
intercorrelations among all variables are shown in
Table 2. In general, all variables are intercorrelated.
However, there are two important exceptions. First,
BMI is not correlated with the DMS total. Furthermore,
because it was possible that there was a curvilinear
relationship between BMI and the DMS scores,
regressions were performed in which the total DMS
score was regressed on BMI and the squared value of
BMI (quadratic term). This regression was not
significant. Therefore, BMI was not included in any
additional analyses. Similarly, pubertal status was not
related to the total DMS score and was not included in
the regression.
Regression analyses
The major hypotheses of this study were tested in
a regression analysis in which the total DMS score
was regressed on the following variables: the three
major sociocultural influences of perceived media
influence, peer teasing, and parent teasing; the potential

Table 2
Means, standard deviations, and intercorrelations
Variable Mean SD 1 2 3 4 5 6 7 8 9 10 11
1. DMS total 35.59 18.29 –
2. DMS-behavior 13.29 9.36 .85 *** –
3. DMS-attitude 8.25 5.11 .79 *** .46*** –
4. DMS-weightlifting 14.83 7.42 .87 *** .56*** .72 *** –
5. Parent teasing 7.91 3.16 .25 *** .12 .28 *** .25*** –
6. Peer teasing 4.95 2.32 .18 ** .04 .24 *** .19** .49 *** –
7. Media investment 17.79 7.63 .48 *** .28*** .40 *** .59*** .27 *** .25 *** –
8. Social comparison 8.97 3.74 .42 *** .23*** .36 *** .46*** .42 *** .45 *** .55*** –
9. Self-esteem 53.73 10.91 .17 * .15* .19 ** .13 .24 *** .30 *** .17** .28 *** –
10. Gender role 25.99 6.93 .31 *** .17** .26 *** .38*** .10 .12 .41*** .32 *** .09 –
11. Puberty 9.63 2.49 .09 .03 .02 .14* .10 .10 .12* .15 * .12 .18 ** –
12. BMI 20.86 3.80 .02 .03 .05 .06 .01 .14 * .12 .09 .12 .04 .27 ***
n’s range from 220 to 282.
*
p < .05.
**
p < .01.
***
p < 001.
 L. Smolak, J.A. Stein / Body Image 3 (2006) 121–129
moderating variables of self-esteem, social comparison,
and male physical attributes endorsement; and, to test
the moderating effects, the interactions of the three
putative moderators with the primary sociocultural
variables. Because this was the first study to examine
the impact of an aspect of gender role (male physical
attributes endorsement) on drive for muscularity among
middle school boys, we also included a quadratic term
for this variable to test for nonlinear relationships.
Furthermore, since most boys are likely to endorse some
questions of the scale, it is possible that there is a
threshold rather than a linear relationship. All variables
were centered prior to analysis to simplify the
interpretation of the interaction terms (Aiken & West,
1991). Block entry was used with all variables entered
simultaneously. Non-significant interactions and quad-
ratic terms were dropped and the analyses were run
again in order to yield the most parsimonious final
regression models.
Results of the regression analysis are shown in
Table 3. The regression for DMS total was significant, F
(8, 189) = 11.68, p < .001, adj R2 = .30. Significant
contributing variables were media influence, social
comparison, endorsement of male physical attributes,
endorsement of male physical attributes quadratic, and
the interaction of media and endorsement of male
physical attributes. Parental comments were marginally
significant ( p < .06). The interaction effect has a
negative sign. Additional analysis indicated that when
endorsement of male physical attributes is close to zero,
media investment has a large effect on DMS total score
(as indicated by the beta associated with the main effect
of media). But, as the endorsement of male physical
attributes increases, media’s effect on DMS score
decreases, though it is always positive. However, it is
important to consider the role of the significant
endorsement of male physical attributes quadratic term
Table 3
Regression for Drive for Muscularity Scale
Variable Beta
Parent teasing .14
Peer teasing .09
Media influence .28
Social comparison .17
Self-esteem .08
Male physical attributes .21
Male physical attributes squared .19
Media  male physical attributes .15
*
p < .05.
**
p < .01.
***
p < 001.
127
in the joint effect of media and endorsement of male
physical attributes on DMS total scores. Analysis of the
endorsement of male physical attributes quadratic term
showed that DMS total scores began to increase
dramatically once endorsement of male physical
attributes scores reached 20.96 (assuming that all other
predictor variables are at their mean value). This score is
5.03 points below the mean and represents an average
answer of 2.62 to each question. Thus, taken together,
the main effects, quadratic effect of endorsement of
male physical attributes, and (negative) interaction
effect of media and endorsement of male physical
attributes, indicate that DMS total scores increase in
association with media and endorsement of male
physical attributes scores but that this increase is
particularly dramatic at lower levels of the media and
endorsement of male physical attributes scores (taken
together). The rate of increase (i.e., the slope) slows as
media and endorsement of male physical attributes
scores increase. This is due to the nonlinear relationship
between endorsement of male physical attributes and
DMS total.
Discussion
The purpose of this research was to explore the
influence of sociocultural factors, specifically perceived
influence of media, peer teasing and comments, and
parental comments and teasing, as moderated by self-
esteem, social comparison, and male physical attributes
endorsement, on drive for muscularity among middle
school boys. The results indicated that sociocultural
factors, particularly investment in media did indeed
impact drive for muscularity (DMS). Furthermore, male
physical attributes endorsement was also an important
factor, both as a direct influence and as a moderator.
It is evident that body image is gendered (see, e.g.,
Smolak & Murnen, 2004 for a discussion). Yet, research
exploring the ways that gender role influences body
image are quite limited. The present research indicates
that even low to moderate endorsement of the
t importance of physical strength and athletic skill by
1.92 boys is associated with higher levels of drive for
1.18 muscularity. This is an important finding because
3.61*** researchers have developed prevention programs aimed
2.04* at gender role stereotyping (e.g., Bigler & Liben, 1992).
1.23
3.06**
Elements of these programs could be integrated into
2.78** body image and steroid use prevention programs.
2.25* Our data also underscore the importance of media in
shaping drive for muscularity. While peer and parent
teasing were correlated with DMS total in the simple
correlations, these correlations were relatively modest.
128 L. Smolak, J.A. Stein / Body Image 3 (2006) 121–129
Furthermore, neither variable was significant in the
regression equation, although parent comments
approached significance. Perceived media influence,
however, tended to have both direct and moderated
relationships to DMS total. It should be noted that this
measure approximated an internalization of media
image measure and was not a simple exposure measure
whereas the peer teasing and parent comments
measures were exposure (albeit perceived) measures.
Interestingly, neither BMI nor pubertal status was
related to drive for muscularity total score, even in the
bivariate correlations. Barker and Galambos (2003)
similarly found that BMI and pubertal status were not
risk factors for adolescent boys’ body dissatisfaction,
although they measured body dissatisfaction generally
rather than drive for muscularity specifically. In other
words, in the present study, it was not simply boys who
were ‘‘undersized’’ who were interested in building
muscles. Clearly, the cultural pressures that shape
gender role and ideal body type override size alone.
Thus, if one is interested in ‘‘targeted’’ prevention of
drive for muscularity or muscle building techniques, the
targeted groups should be boys who are high in male
physical attributes endorsement, social comparison, or
media investment rather than boys who are small for
their age.
The factor analysis of the DMS also yielded some
interesting results. We were unable to replicate the
McCreary et al. (2004) factor analysis of the scale. This
was true in two ways. First, the present factor structure
was not as clear-cut as the McCreary et al. structure.
Many more variables seemed to potentially load on two
factors in the present data. Second, the present factor
analysis yielded three factors. Interestingly, weightlift-
ing appeared to be separate from other muscle building
techniques. Future research should examine whether
there are developmental differences in the factor
structure, and hence the meaning, of DMS scores.
This study has several limitations. First and most
importantly, this is a cross-sectional study. Despite a
theoretical model and some replication of previous
findings, these data cannot establish causal relation-
ships. Longitudinal data are needed to determine
causality; such research may need to start early in
childhood in order to avoid the confound of bi-
directional relationships between the sociocultural
factors and DMS. Second, the sample is overwhel-
mingly white and middle-class American and the small
numbers of boys from ethnic minority groups precluded
the examination of ethnic group differences. These
relationships may differ by ethnicity and social class,
particularly for groups with different gender roles,
different traditions concerning teasing, and, of course,
different body ideals. Much more research is needed on
cultural differences in the development of drive for
muscularity among boys.
Despite these limitations, this research does point to
several provocative questions for future research. It is
important that future research investigate the factor
structure of the DMS with this age group. It is possible
that the meaning of muscularity changes throughout
development. If it does, then it is also possible that the
predictors of drive for muscularity will differ across
development. The present research also underscores the
importance of not only sociocultural influences but also
moderators of those influences, particularly social
comparison and male physical attributes endorsement.
When combined with previous research, it appears that
future work should not simply look at direct influences
but should consider more complex models. Particular
attention should be paid to gender role issues, whether
they are measured with the present scale or with other
instruments. Furthermore, many of these factors are
potentially amenable to prevention efforts, especially if
one adopts the ecological approach that emphasizes the
importance of changing environmental factors such as
gender role pressures and media messages rather than
just the individual (Levine & Smolak, 2006). While
there are some steroid prevention programs available
(e.g., MacKinnon et al., 2001), the prevention of
negative body image in boys is still in its infancy.
References
Aiken, L., & West, S. (1991). Multiple regression: Testing and
interpreting interactions. Sage: Thousand Oaks, CA.
Barker, E., & Galambos, N. (2003). Body dissatisfaction of adolescent
boys and girls: Risk and resource factors. Journal of Early
Adolescence, 23, 141–165.
Bigler, R., & Liben, L. (1992). Cognitive mechanisms in children’s
gender stereotyping: Theoretical and educational implications of a
cognitive-based intervention. Child Development, 63, 1351–1363.
Cafri, G., & Thompson, J. K. (2004). Measuring male body image: A
review of the current methodology. Psychology of Men and
Masculinity, 5, 18–29.
Cafri, G., Thompson, J. K., Ricciardelli, L., McCabe, M., Smolak, L.,
& Yesalis, C. (2005). Pursuit of the muscular ideal: Physical and
psychological consequences and putative risk factors. Clinical
Psychology Review, 25, 215–239.
Field, A., Austin, B., Camargo, C., Taylor, C. B., Striegel-Moore, R.,
Loud, K., & Colditz, G. (2005). Exposure to the mass media, body
shape concerns, and use of supplements to improve weight
and shape among male and female adolescents. Pediatrics, 116,
214–220.
Galambos, N., Almeida, D., & Petersen, A. (1990). Masculinity,
femininity, and sex role attitudes in early adolescence: Exploring
gender intensification. Child Development, 61, 1904–1914.
 L. Smolak, J.A. Stein / Body Image 3 (2006) 121–129
Groesz, L., Levine, M., & Murnen, S. (2002). The effect of experi-
mental presentation of thin media images on body satisfaction: A
meta-analytic review. International Journal of Eating Disorders,
31, 1–16.
Grogan, S., & Richards, H. (2002). Body image: Focus groups with
boys and men. Men and Masculinities, 4, 219–232.
Haselhuhn, G., Thompson, J. K., Roehrig, M., Shroff, H., van den
Berg, P., Keery, H., Williams, R., Boroughs, M., Himes, S.,
Herbozo, S., & Cafri, G. (2001, November). Development of
the media influence scale for adolescent boys. Paper presented
at the Association of the Advancement of Behavior Therapy.
Jones, D. C. (2004). Body image among adolescent girls and boys: A
longitudinal study. Developmental Psychology, 40, 823–835.
Keery, H., van den Berg, P., & Thompson, J. K. (2004). An evaluation
of the Tripartite Influence Model of body dissatisfaction and eating
disturbance with adolescent girls. Body Image, 1, 237–251.
Levine, M. P., & Smolak, L. (2006). The prevention of eating problems
and eating disorders: Theory, research, and practice. Mahwah,
NJ: Erlbaum.
Levine, M. P., Smolak, L., & Hayden, H. (1994). The relation of
sociocultural factors to eating attitudes and behaviors among
middle school girls. Journal of Early Adolescence, 14, 471–490.
MacKinnon, D., Goldberg, L., Clarke, G., Elliot, D., Cheong, J.,
Lapin, A., Moe, E., & Krull, J. (2001). Mediating mechanisms in a
program to reduce intentions to use anabolic steroids and improve
exercise self-efficacy and dietary behavior. Prevention Science, 2,
15–28.
Mahalik, J., Locke, B., Ludlow, L., Diemer, M., Scott, R., Gottfried,
M., & Freitas, G. (2003). Development of Conformity to Mascu-
line Norms Inventory. Psychology of Men and Masculinity, 4,
3–25.
McCabe, M., & Ricciardelli, L. (2004a). A longitudinal study of
pubertal timing and extreme body change behaviors among ado-
lescent boys and girls. Adolescence, 39, 145–166.
McCabe, M., & Ricciardelli, L. (2004b). Weight and shape concerns
of boys and men. In J. K. Thompson (Ed.), Handbook of eating
disorders and obesity (pp. 606–634). New York: Wiley.
McCreary, D., & Sasse, D. (2000). An exploration of the Drive for
Muscularity in adolescent boys and girls. Journal of American
College Health, 48, 297–304.
McCreary, D., & Sasse, D. (2002). Gender differences in high school
students’ dieting behavior and their correlates. International
Journal of Men’s Health, 1, 195–213.
McCreary, D., Sasse, D., Saucier, D., & Dorsch, K. (2004). Measuring
the drive for muscularity: Factorial validity of the Drive for
129
Muscularity Scale in men and women. Psychology of Men and
Masculinity, 5, 49–58.
McCreary, D., Saucier, D., & Courtenay, W. (2005). The Drive for
Muscularity and masculinity: Testing the associations among
gender-role traits, behaviors, attitudes, and conflict. Psychology
of Men and Masculinity, 6, 83–94.
Petersen, A., Crockett, L., Richards, M., & Boxer, A. (1988). A self-
report measure of pubertal status: Reliability, validity and initial
norms. Journal of Youth and Adolescence, 17, 117–133.
Ricciardelli, L., & McCabe, M. (2001). Self-esteem and negative
affect as moderators of sociocultural influences on body
dissatisfaction, strategies to decrease weight, and strategies to
increase muscles among adolescent boys and girls. Sex Roles, 44,
189–207.
Ricciardelli, L., & McCabe, M. (2003). A longitudinal analysis of the
role of psychosocial factors in predicting body change strategies
among adolescent boys. Sex Roles, 48, 349–360.
Ricciardelli, L., & McCabe, M. (2004). A biopsychosocial model of
disordered eating and the pursuit of muscularity in adolescent
boys. Psychological Bulletin, 130, 179–205.
Rosenberg, M. (1965). Measurement of self-esteem. In M. Rosenberg
(Ed.), Society and the adolescent self-image (pp. 297–307). New
York: Princeton University Press.
Smolak, L., & Murnen, S. (2001). Gender and eating problems. In R.
Striegel-Moore, & L. Smolak (Eds.), Eating disorders: Innova-
tions in research, treatment, and prevention (pp. 91–110).
Washington, DC: American Psychological Association.
Smolak, L., & Murnen, S. (2004). A feminist approach to eating
disorders. In J. K. Thompson (Ed.), Handbook of eating disorders
and obesity (pp. 590–605). New York: Wiley.
Smolak, L., Murnen, S., & Thompson, J. K. (2005). Sociocultural
influences and muscle building in adolescent boys. Journal of Men
and Masculinity, 6, 227–239.
Stanford, J., & McCabe, M. (2005). Sociocultural influences on
adolescent boys’ body image and body change strategies. Body
Image, 2, 105–113.
Stice, E. (1994). Review of the evidence for a sociocultural model of
bulimia nervosa and an exploration of the mechanisms of action.
Clinical Psychology Review, 14, 633–661.
Thompson, J. K., Heinberg, L., Altabe, M., & Tantleff-Dunn, S.
(1999). Exacting beauty: Theory, assessment, and treatment of
body image disturbance. Washington, DC: American Psycholo-
gical Association.
Thompson, J. K., Heinberg, L., & Tantleff, S. (1991). The physical
comparison scale (PACS). The Behavior Therapist, 14, 174.