ERIC EJ853645: The Association of Body Mass Index, Perceived Body Mass Index, and Predictors of Eating Disorders among a Sample of College Students

6 Pages·2003·0.19 MB·English

by  ERIC

The Association of Body Mass Index, Perceived Body
Mass Index, and Predictors of Eating Disorders Among
a Sample of College Students
Timothy J. Bungum and Stacy Lofton
ABSTRACT
Rates of eating disorders have increased recently and are a public health concern especially among college students.
Understanding variables that are associated with eating disorders could be helpful in preventing them. Participants
(N=525) were students from a large southwestern university. It was hypothesized that a desire for an underweight
body mass index (BMI) would be predicted by one’s ability to accurately identify one’s current BMI while control-
ling for variables previously shown to be associated with eating behaviors, which included body dissatisfaction;
drive for thinness; social physique anxiety; bulimia tendencies; dietary restraint; age; race; Greek membership; and
gender were included in the regression model. Women scored higher than men on measures of social physique
anxiety, body dissatisfaction, drive for thinness, and dietary restraint, although more then twice the proportion of
men (49%) as compared with women (22%) were overweight or obese. Women were more accurate in correctly
identifying their own BMI category. However, incorrectly identifying one’s BMI was the only significant predictor of
desiring an underweight BMI among women. No independent variables predicted the desire for an underweight
BMI among men. Future research should assess the preventive and predictive power of accurately assessing one’s
BMI with eating behaviors.
In recent years the incidence of eating Latzer & Schatz, 1999), and accordingly, shift toward a thinner body size standard
disorders has increased (Garfinkel, 1995; gender has been identified as the most in- (Streigal-Moore, McAvay, & Rodin, 1986).
Latzer & Schatz, 1999; Shisslak, Crago, & fluential factor associated with the preva- Despite this, over the last decade the per-
Estes, 1995), and they are now the third lence of eating disorders (Cox, Lantz, & centage of overweight and obese adults age
most common chronic illness among Mayhew, 1997), as both clinical and epide- 20 years or older has increased to 54.9%
women in the United States (Steiner-Adair miological studies support a gender ratio
& Purcell, 1996). It is estimated that 5 mil- of approximately 10 female cases to 1 male
lion Americans, which is 1.7% of the esti- eating disorder case (Anderson & Holman, Timothy J. Bungum is with the University of
mated population of 290,000,000 (U.S. 1997). The most striking prevalence of dis- Nevada-Las Vegas, 4505 Maryland Parkway,
Census Bureau, 2002) are affected by eat- ordered eating behaviors is among college- Box 45050, Las Vegas, NV 89154-3050;
ing disorders annually (Becker, Grinspoon, aged women (Hesse-Biber, 1992). [email protected]. Stacy
Kilbanski, & Herzog, 1999). Adolescent girls Obesity is a highly stigmatized condition Lofton, MS, RD, is with Providence-St.
and young women are at the highest risk because American culture values thinness. Vincent’s, 9205 SW Barnes Road, Portland, OR
for eating disorders (Becker et al., 1999; This value judgment has led to a continual 97225.
American Journal of Health Education — November/December 2003, Volume 34, No. 6 337
Timothy J. Bungum and Stacy Lofton
(National Heart Lung and Blood Institute and women view their bodies differently. and bulimia (k=7). Dietary restraint items
[NHLBI], 1998). This increase has widened Women tend to overestimate their body (k=10) were gleaned from the Drive for
the gap between the “thin-ideal” and size, whereas men are less critical of their Thinness Scale (Herman, 1978) and the
individual’s actual body weights, resulting bodies and tend to perceive their size and Social Physique Anxiety Scale (Hart et al.,
in a continual pursuit of thinness and the shape as normal (Hart et al., 1989). There- 1989), which is comprised of 12 items. Par-
potential for an unceasing drive for thin- fore, it would be interesting and potentially ticipants responded to forced-choice item
ness, which can potentially lead to disor- valuable to study whether men and women scales by rating whether each item applied
dered eating (Stice, 1994). who accurately assess their own body mass always, usually, often, sometimes, rarely, or
Most evidence suggests that eating dis- index (BMI) differ in regard to desiring never. Scale scores were the summation of
orders are linked to multiple factors that an underweight BMI from those who inac- all item scores for that scale. Three-week
include female gender, age, weight, and race, curately assess their BMI, while controlling test–retest reliabilities on a sample of 70
and sociocultural factors such as body dis- for drive for thinness; body dissatisfaction; nonclinical subjects were above .81 for the
satisfaction, drive for thinness, social phy- bulimia; social physique anxiety; dietary dietary restraint, body dissatisfaction, and
sique anxiety, and dietary restraint. Another restraint; age; and Greek membership. bulimia scales (Williamson, Anderson,
sociocultural factor that is related to eating NHLBI uses the following classifications for Jackman, & Jackson, 1995).
disorders is bulimic tendencies (Garner, BMI: underweight (<18.5), normal (>18.5 The Drive for Thinness Scale (Herman,
Olmstead, & Polivy, 1983). Because socio- but <25), overweight (>25 but <30) or 1978) is a 10-item self-report questionnaire
cultural factors and individual perceptions obese (>30). The NHLBI lists no health assessing weight fluctuations, degrees of
are potentially changeable, they are the foci benefits associated with underweight BMIs, chronic dieting, and related attitudes to-
of this study. As stated, body dissatisfaction but there are numerous health risks corre- ward weight and eating. The Social Phy-
is a factor that could lead to the develop- lated with being overweight or obese sique Anxiety scale (Hart et al., 1989) is a
ment of an eating disorder (Wong & Huang, (NHLBI, 1998). 12-item self-report inventory and is de-
1999). Body dissatisfaction reflects the be- signed to measure levels of anxiety one
METHODS
lief that specific parts of the body associ- senses in regard to his or her own physique.
ated with shape are too large (Garner et al., Pilot Testing Subjects were asked to indicate the degree
1983). Studies investigating gender differ- Selected students attending the Univer- to which each of the 12-item statements was
ences consistently show that men and sity of North Texas were asked to complete characteristic or true of themselves. An-
women differ in their perceptions of and a pilot test of a questionnaire in classroom swers were based on a 5-point Likert-type
dissatisfaction with their bodies (Anderson settings. The pilot was conducted to detect scale with numerical values of 1 to 5 as-
& Holman, 1997; Connor-Green, Striegel- unclear directions, ambiguously worded signed accordingly. Numerical values were
Moore, & Cronan, 1994). Another factor questions, and potential problems in ad- summed to produce a total score (Hart et
that can influence eating disorder develop- ministrating the questionnaire. Fifty-one al., 1989; Cox et al., 1997). With an 8-week
ment is drive for thinness. This variable re- students enrolled in undergraduate health interval, test–retest administrations have
flects an excessive concern with dieting, pre- promotion courses completed the question- demonstrated acceptable reliability (r= .82),
occupation with weight, and determined naire. Requested information included de- and construct validity through moderate
pursuit of thinness (Garner et al., 1983). A mographic data and assessments of drive for correlations with measures that include fear
fourth factor that may predict disordered thinness, body dissatisfaction, bulimia of negative evaluation (r=.35) and body
eating is dietary restraint. This factor is de- (Garner et al., 1983), dietary restraint cathexis (r=.51) (Hart et al., 1989). Further-
fined as a person’s intent to restrict dietary (Herman, 1978), and social physique anxi- more, Petrie, Diehl, Rogers, and Johnson
intake to control body weight. An additional ety (Hart et al., 1989). Approval to conduct (1996) demonstrated construct validity and
factor shown to be associated with eating this study was granted by the university’s reliability for these scales among under-
disorders is social physique anxiety, defined Institutional Review Board. graduate women and men.
as a subtype of social anxiety that occurs as Instrumentation The questionnaire also requested demo-
a result of the prospect or presence of in- The questionnaire utilized for the for- graphic data that included date of birth;
terpersonal evaluation involving one’s phy- mal study consisted of 52 items, including gender; height; present weight; desired
sique (Hart, Leary, & Rajeski, 1989). Finally, the demographic items. Items to assess the weight; race/ethnicity; and Greek member-
bulimia is the tendency toward episodes of sociodemographic factors were derived ship. Self-reported height, present body
overeating and may be followed by the im- from the scales of three instruments. The weight, and desirable body weight were used
pulse to engage in self-induced vomiting Eating Disorder Inventory (Garner et al., to calculate actual and desirable BMIs.
(Garner et al., 1983). 1983) supplied the items for dietary re- Subjects were also asked to describe their
Research has demonstrated that men straint (k=7), body dissatisfaction (k=9), current perceived BMIs according to the
338 American Journal of Health Education — November/December 2003, Volume 34, No. 6
Timothy J. Bungum and Stacy Lofton
four categories (e.g., underweight, normal, Mean respondent age was 22.2 (SD=3.9) noted that 3% of men were actually under-
overweight, or obese) established by the years. Racial/ethnic distribution was 64.0% weight, but 14% perceived themselves to be
NHLBI (1998). Caucasian (n=336); 19.0% African Ameri- underweight. Among women, 72% were of
Sample can (n=100); 8.2% Hispanic (n=43); 5.5% normal weight, but only 54% perceived
Asian/Pacific Islander (n=29); and 3.2% themselves as such. And 22% of women
Undergraduate students who were en-
other, including Native American (n=17). were overweight or obese, but 42% consid-
rolled in health-related fitness, weight train-
Self-reported heights and weights were used ered themselves to be in those categories.
ing, and self-defense courses were ap-
to calculate the actual and desirable BMI for Descriptive statistics for drive for thin-
proached after one of the investigators
each subject. The mean BMI for all subjects ness, body dissatisfaction, bulimia, social
contacted the instructors and requested
was 24.03 (SD=4.2), whereas the mean de- physique anxiety, and dietary restraint are
permission to survey these students. All in-
sired BMI was 22.8 (SD=3.3). Table 1 pre- presented in Table 2 by gender. Higher
structors allowed their students the oppor-
sents actual, desired, and perceived body scores indicate greater anxiety in response
tunity to be a part of the study. On the pre-
weights by gender. A plurality of men (48%) to others’ evaluation of their physique. A
arranged day, students were asked to
and a majority of women (72%) were cat- significant gender difference between group
complete the survey. No students refused to
egorized as “normal weight.” It should be means existed for all independent variables
participate, and no effort was made to in-
clude students not in attendance on the date
of survey administration. No benefits or in-
centives were offered to the students for Table 1. Comparison of Actual (ABMI), Desired (DBMI), and Perceived
their participation. Body Mass Index (PBMI) of Male and Female College Students
Data Analysis
Variable Men N (%) Women N (%) Total N (%)
Data were analyzed using the Statistical
ABMI (underweight) 7 (3) 17 (6) 24 (5)
Package of the Social Sciences (SPSS, Chi-
(normal) 119 (48) 198 (72) 317 (61)
cago, IL). Descriptive statistics were em-
(overweight) 87 (35) 46 (17) 133 (26)
ployed to describe age, race/ethnic back-
(obese) 34 (14) 14 (5) 48 (9)
ground, BMI (actual and desired), and body
perception for male and female partici- DBMI (underweight) 4 (2) 35 (13) 39 (8)
pants. Analysis of variance was performed (normal) 121 (50) 226 (83) 347 (67)
to identify statistical differences in scale (overweight) 105 (43) 9 (3) 114 (22)
scores, separately by sex, between those who (obese) 14 (6) 2 (1) 16 (3)
correctly identified their BMI, and those
PBMI (underweight) 35 (14) 11 (4) 46 (9)
who incorrectly identified their BMI on five
(normal) 136 (55) 149 (54) 285 (63)
variables: (1) drive for thinness, (2) body
(overweight) 71 (29) 108 (39) 179 (34)
dissatisfaction, (3) bulimia, (4) dietary re-
(obese) 6 (2) 8 (3) 14 (3)
straint, and (5) social physique anxiety. To
determine which means were statistically
Note: N=248 men, 277 women.
significant, a Student-Newman-Keuls post
hoc test was performed. Logistic regression
was used to identify predictors of desiring
Table 2. Mean and Standard Deviations of Variable
an underweight BMI while controlling for
Scores Categorized by Actual Body Mass Index
social physique anxiety; drive for thinness;
Category for Male and Female College Students
dietary restraint; bulimia; race; age; and
Greek membership. Alpha for all statistical Men Women Combined
procedures was set at .05. Independent Variable M (SD) M (SD) M (SD)
Social physique anxiety 30.0 (8.5) 36.8 (8.9) 33.6 (9.4)
RESULTS Bulimia 0.7 (1.6) 0.8 (1.9) 0.7 (1.7)
Body dissatisfaction 5.3 (5.8) 10.8 (7.5) 8.3 (7.2)
Respondents
Drive for thinness 4.4 (3.7) 7.2 (5.3) 5.9 (4.8)
Data were available and analyzed for 525
Dietary restraint 9.4 (5.9) 13.2 (5.8) 11.4 (6.2)
participants. The gender distribution of the
sample was fairly even; 47.2% were men
Note: N=248 men, 277 women.
(n=248) and 52.8% (n=277) were women.
American Journal of Health Education — November/December 2003, Volume 34, No. 6 339
Timothy J. Bungum and Stacy Lofton
obese than women (21.8%), yet women
Table 3. Actual Body Mass Index (BMI) Versus
scored higher on the psychological scales
Perceived BMI for Men and Women
that indicate concern about their body size.
This finding suggests that the pressures
Actual BMI
women perceive in evaluating their bodies
Under Normal Overweight Obese
is more pronounced than among men, even
N N N N
though a higher proportion of men have
Perceived normal
BMIs that are in the unhealthy ranges of
BMI, men 3 83 37 4
overweight and obese. The finding also im-
Under 4 27 2 2
plies that women’s evaluation of their bod-
Overweight - 6 37 23
ies transcends health indicators and agrees
Obese - - 1 5
with findings that show women feel pres-
Perceived normal sure to maintain thin body shapes (Ander-
BMI, women 11 130 3 1 son & Holman, 1997; Rolls, Fedoroff, &
Under 6 5 - - Guthrie, 1991).
Overweight 2 55 38 9 The present data demonstrate that
Obese - - 1 - women are generally more accurate in as-
sessing their own BMI category than men,
Note: N=248 men and 277 women.
but the patterns of that inaccuracy between
men and women differs, and in fact, are
opposite. In agreement with others
except for bulimia [F (3,513) =1.161, p>.05] sification. However, women who incorrectly (Heatherton, Nichols, Mahamedi, & Keel,
and drive for thinness [F (3,508) =2.159, identified their BMI scored higher on the 1995; Sciacca, Melby, Hyner, Brown, &
p>.05]. drive for thinness scale [F(1, 263)=15.60, Femea, 1991; Wong & Huang, 1999), when
When assessing status of their body size p<.05] and social physique anxiety [F women incorrectly assessed their body
as defined by the NHLBI, women (65%, (1,263)=5.59, p<.05] than women who ac- shape, it was probable that they assigned
n=177) were more accurate in assigning curately described their body shapes. themselves to a category that was heavier
themselves to the corresponding categories Among men the regression analysis uti- than their actual weight. Men, on the other
than were men (55%, 129). Of the women lized to predict the desire for an under- hand, who incorrectly assigned themselves
who incorrectly assigned themselves a BMI, weight body revealed no significant predic- to a BMI group were apt to perceive that
80% (74/92) classified themselves as heavier tors. For women only incorrectly identifying they were lighter than was actually the case
than they were. Among men who incor- one’s BMI [F(1,263)=2.56, p<.05) predicted (Wong, Chen, Chan, Wang, & Yamamoto,
rectly assigned themselves, 85% described that same aspiration. Women who incor- 1999). This divergence in actual versus per-
their own body weight to be lighter than the rectly classified their BMIs were 2.6 (confi- ceived body weight is partially explained by
NHLBI category to which their actual BMI dence interval=1.24, 5.42) times as apt to the cultural ideal for body shape, in that
corresponded. A description of male and desire an underweight BMI as those who women are to be slender and men strong,
female actual and perceived BMIs is pre- accurately described their BMI. athletic, and muscular. The discrepancy
sented in Table 3. The data also indicate that between women’s actual body weight and
13% of women desired to be “underweight,” DISCUSSION their self-perceived body weight likely re-
whereas only 1.6% of men desired to be in In this study women differed from men flects an unrealistic view of the “ideal” fe-
that category. An additional gender differ- in that they reported statistically higher male body. This may encourage the devel-
ence in how men and women evaluate their mean scores for drive for thinness, body opment of disordered attitudes and
body size is that 49% of men were over- dissatisfaction, dietary restraint, and social behaviors toward weight, body shape, and
weight or obese, although only 31% as- physique anxiety. However, this was ex- eating. The discrepancy between men’s ac-
signed themselves to one of those groups, pected, because college-aged women are at tual body weight and their perceived weight
whereas 22% of women were overweight or higher risk for developing eating disorders remains ambiguous. However, it has been
obese, and 42% thought they were. and perceive more social pressure to be thin hypothesized that, unlike women who tend
There were no statistically significant than do men (Anderson & Holman, 1997). to perceive themselves as larger than they
differences in scores of the predictor vari- Although this finding was anticipated, it is actually are, some men perceive themselves
ables between those men who correctly and interesting, because a much higher propor- to be smaller than is the case.
incorrectly identified their actual BMI clas- tion of men (49%) were overweight or Findings from this study are also in
340 American Journal of Health Education — November/December 2003, Volume 34, No. 6
Timothy J. Bungum and Stacy Lofton
agreement with studies (Rolls et al., 1991; misperception by serving as a cue to take Bulletin, 33, 391–397.
Stice, 1994) indicating that women appear weight reduction measures such as diet and Becker, A. E., Grinspoon, S. K., Klibanski, A.,
to be more concerned with aspects of their exercise. Conversely, most women who in- & Herzog, D. B. (1999). Current concepts: Eat-
appearance, particularly their weight, accurately identified their BMI thought that ing disorders. New England Journal of Medicine,
whereas men are more likely to report that they were overweight or obese and were 340, 1092–1098.
if they are fit and exercise they feel positive actually of normal weight. A correction of Burton, W. N., Chen, C., Schultz, A. B., &
about their bodies. The desire to be fit may this erroneous belief could attenuate body Edington, D. W. (1998). The economic costs as-
reflect the desire to be heavier or the drive shape-related anxiety and perhaps reduce sociated with body mass index in a workplace.
for bulk. However, fitness or bulk is not con- the likelihood of engaging in unhealthy eat- Journal of Occupational and Environmental
gruent with obesity. The drive for bulk has ing behaviors. Medicine, 40, 786–792.
been described as the desire to gain muscle Finally, the limitations of this study need Connor-Greene, P. A., Striegel-Moore, R. H.,
(Wong et al., 1999). to be considered. Although the current find- & Cronan, S. (1994). Perceived social climate
Results from the logistic regression ings merit discussion, the cross-sectional and weight preoccupation in college women.
model predicting the desire for an under- study design allows threats to validity to Eating Disorders: The Journal of Treatment and
weight BMI also warrant discussion. The affect the results. Because of the nature of Prevention, 2, 126–134.
data from men indicate that none of the this study we are unable to determine Cox, L .M., Lantz, C. D., & Mayhew, J. L.
independent variables predicted the desire whether it is the failure to accurately de- (1997). The role of social physique anxiety and
for being below an 18.5 BMI. This finding scribe their BMI that creates the desire other variables in predicting eating behaviors
could be expected, as only four men (1.6%) for an underweight BMI or whether this in college students. International Journal of
desired an underweight BMI. Consequently, finding is tapping the possibility that those Sports Nutrition, 7, 310–317.
the statistical power to predict this phenom- with eating disorder tendencies have dis- Garfinkel, P. E., Kennedy, S. H. & Kaplan, A.
enon is small. For women the sole signifi- torted body images. Further, the notion that S. (1995). Views on classification and diagnosis
cant predictor of desiring an underweight those who desire an underweight BMI are of eating disorders. Canadian Journal of Psychia-
BMI, which 14% did desire, was incorrectly apt to engage in disordered eating behav- try, 40, 445–456.
identifying their actual BMI. Because eat- iors holds intuitive appeal, but a longitudi- Garner, D. M., Olmstead, M. P., & Polivy, J.
ing disorders are more prevalent among nal study design will be required to estab- (1983). Development and validation of a mul-
women, these findings, when viewed in iso- lish that relationship. tidimensional eating disorder inventory for an-
lation, suggest that the ability to accurately Present researchers are not suggesting orexia nervosa and bulimia. International Jour-
describe one’s BMI is not likely a predictor that eating disorders are not serious and nal of Eating Disorders, 2 (2), 15–34.
of eating behaviors, as women may be more complicated illnesses, but efforts to assist Hart, E. A., Leary, M. R., & Rejeski, W. J.
able to identify their BMIs as compared with those tending toward disordered eating in (1989). The measurement of social physique
men. However, among those women who accurately describing their own body shape anxiety. Journal of Sport & Exercise Psychology,
inaccurately identified their BMIs, there was may play a role in preventing or predicting 11, 94–104.
more than double the risk of being assigned these illnesses. Prospective studies are Heatherton, T. R., Nichols, P., Mahamedi, F.,
to the group that desires an underweight needed to determine whether simply ask- & Keel, P. (1995). Body weight, dieting, and eat-
BMI. As stated previously, the NHLBI listed ing college students to describe their own ing disorder symptoms among college students,
no health benefits associated with under- BMIs might be an inexpensive and yet ef- 1982 to 1992. American Journal of Psychiatry,
weight BMIs, but there are unhealthy con- fective screening instrument for eating dis- 152, 1623–1629.
sequences such as osteoporosis that are as- orders. Further, self-report was used to as- Herman, C. P. (1978). Restrained eating. Psy-
sociated with low body weights (National sess body weights in this study. To the extent chiatric Clinics of North America, 1, 593–607.
Eating Disorders Association, 2002). that the self-reports were inaccurate, this Hesse-Biber, S. (1992). Report on a panel
A potential implication of the present would be a potential threat to the internal longitudinal study of college women’s eating
findings for health educators is a need to validity of the findings. However, others patterns and eating disorders: Noncontinuum
correct the mistaken perceptions that some have concluded that self-reporting weight versus continuum measures. Health Care for
men and women have about their BMIs. is a feasible method of data collection (Bur- Women International, 13, 375–391.
That most men who inaccurately identified ton, Chen, Schultz, & Edington, 1998). Latzer, Y., & Shatz, S. (1999). Comprehen-
their BMIs were overweight or obese and sive community prevention of disturbed atti-
thought that they were in the normal range REFERENCES tudes to weight control: A three-level interven-
suggests that overweight or obese men who Anderson, A. E., & Holman, J. E. (1997). tion program. Eating Disorders: The Journal of
believe their weight is normal could ben- Males with eating disorders: Challenges for Treatment and Prevention, 7 (1), 3–31.
efit from interventions that rectify this treatment and research. Psychopharmacology National Eating Disorders Association.
American Journal of Health Education — November/December 2003, Volume 34, No. 6 341
Timothy J. Bungum and Stacy Lofton
(2002). Anorexia nervosa. Retrieved December 159–168. EST2002-01.php.
12, 2002, from www.Nationaleatingdisorders. Shisslak, C. M., Crago, M., & Estes, L. S. Williamson, D. A., Anderson, D. A., Jackman,
org/p.asp?WebPage_ID=286&Profile_ID (1995). The spectrum of eating disturbances. L. P., & Jackson, S. R. (1995). Assessment of eat-
=41142. International Journal of Eating Disorders, 18, ing disordered thoughts, feelings and behaviors.
National Heart Lung and Blood Institute. 209–219. In D.B. Allison (Ed.), Handbook of assessment
(1998). Guidelines on overweight and obesity. Steiner-Adair, K., & Purcell, A. (1996). Ap- methods for eating behaviors and weight-related
Retrieved March 12, 2002, from www. nhlbi.nih. proaches to mainstreaming eating disorders problems (pp. 348–352). Thousand Oaks,
gov/guidelines/ob_gdlns.pdf. prevention. Eating Disorders: The Journal of CA: Sage.
Petrie, T. A., Diehl, N., Roger, R. L., & Treatment and Prevention, 4, 294–309. Wong, Y., Chen, S., Chan, Y., Wang, M., &
Johnson, C.L. (1996). The social physique anxi- Stice, E. (1994). Review of the evidence for Yamamoto, S. (1999). Weight satisfaction
ety scale: Reliability and construct validity. Jour- a sociocultural model of bulimia nervosa and and dieting practices among college males. Jour-
nal of Sport & Exercise Psychology, 18, 420–425. an exploration of the mechanisms of action. nal of the American College of Nutrition, 18,
Rolls, B. J., Fedoroff, I. C., & Guthrie, J. F. Clinical Psychology Review, 14, 633–661. 223–228.
(1991). Gender differences in eating behavior Striegel-Moore, R., McAvay, G., & Rodin, J. Wong, Y., & Huang, Y. (1999). Obesity con-
and body weight regulation. Health Psychology, (1986). Psychological and behavioral correlates cerns, weight satisfaction and characteristics of
10, 133–142. of feeling fat in women. International Journal of female dieters: A study on female Taiwanese
Sciacca, J. P., Melby, C. L., Hyner, G. C., Eating Disorders, 5, 935–947. college students. Journal of American College of
Brown, A. C., & Femea, P. L. (1991). Body mass U.S. Census Bureau. (n.d.). State population Nutrition, 18, 194–200.
index and perceived weight status in young estimates. Retrieved March 27, 2003, from
adults. Journal of Community Health, 16, eire.census.gov/popest/data/states/tables/ST-
342 American Journal of Health Education — November/December 2003, Volume 34, No. 6