Mental health impairment in underweight women: do body dissatisfaction and eating-disordered behavior play a role?

Abstract

We sought to evaluate the hypothesis that mental health impairment in underweight women, where this occurs, is due to an association between low body weight and elevated levels of body dissatisfaction and/or eating-disordered behaviour.
Subgroups of underweight and normal-weight women recruited from a large, general population sample were compared on measures of body dissatisfaction, eating-disordered behaviour and mental health.
Underweight women had significantly greater impairment in mental health than normal-weight women, even after controlling for between-group differences in demographic characteristics and physical health. However, there was no evidence that higher levels of body dissatisfaction or eating-disordered behaviour accounted for this difference. Rather, underweight women had significantly lower levels of body dissatisfaction and eating-disordered behaviour than normal-weight women.
The findings suggest that mental health impairment in underweight women, where this occurs, is unlikely to be due to higher levels of body dissatisfaction or eating-disordered behaviour. Rather, lower levels of body dissatisfaction and eating-disordered behaviour among underweight women may counterbalance, to some extent, impairment due to other factors.

Full text

Page 1

RESEARCH ARTICLEOpen Access
Mental health impairment in underweight
women: do body dissatisfaction and eating-
disordered behavior play a role?
Jonathan Mond1*, Bryan Rodgers2, Phillipa Hay3and Cathy Owen4
Abstract
Background: We sought to evaluate the hypothesis that mental health impairment in underweight women, where
this occurs, is due to an association between low body weight and elevated levels of body dissatisfaction and/or
eating-disordered behaviour.
Methods: Subgroups of underweight and normal-weight women recruited from a large, general population
sample were compared on measures of body dissatisfaction, eating-disordered behaviour and mental health.
Results: Underweight women had significantly greater impairment in mental health than normal-weight women,
even after controlling for between-group differences in demographic characteristics and physical health. However,
there was no evidence that higher levels of body dissatisfaction or eating-disordered behaviour accounted for this
difference. Rather, underweight women had significantly lower levels of body dissatisfaction and eating-disordered
behaviour than normal-weight women.
Conclusions: The findings suggest that mental health impairment in underweight women, where this occurs, is
unlikely to be due to higher levels of body dissatisfaction or eating-disordered behaviour. Rather, lower levels of
body dissatisfaction and eating-disordered behaviour among underweight women may counterbalance, to some
extent, impairment due to other factors.
Keywords: Body dissatisfaction, eating-disordered behaviour, mental health, underweight, women
Background
In the past decade, there has been considerable interest
in the question of whether, and how, obesity might
affect individuals’ mental health, both the occurrence of
specific psychopathology [1-7] and the perceived effects
of psychopathology on psycho-social functioning [8-14].
Conclusions in this regard have been found to vary
according to the study design, the characteristics of the
population sampled and the measures of mental health
employed [5-7,15-17]. However, one finding that has
been consistently reported is that obesity is more likely
to be associated with mental health impairment in
women than in men [1-3,5-7,10,13-18]. Further, evi-
dence suggests that this disparity reflects, at least in
part, the fact that the prevalence of body dissatisfaction
and eating-disordered behaviour is higher - and the
adverse effect of these variables on mental health greater
- in overweight women than in overweight men [19,20].
Interestingly, the association between body weight and
mental health impairment in women does not appear to
be confined to overweight. Rather, findings from those
studies that have considered associations with mental
health across the full spectrum of body weight suggest
the presence of U-shaped relationships, such that both
underweight and overweight are associated with poorer
mental health in women [3,6,7,14,21]. In other studies,
only underweight has been found to be associated with
mental health impairment [22,23]. However, the evi-
dence base is small and, as with the literature relating to
obesity and mental health, associations have been found
to depend on the characteristics of the study population,
the measures of mental health employed and whether or
not the influence of potential covariates, such as
* Correspondence: Jonathan.Mond@anu.edu.au
1School of Sociology, Australian National University, Canberra, Australia
Full list of author information is available at the end of the article
Mond et al. BMC Public Health 2011, 11:547
http://www.biomedcentral.com/1471-2458/11/547
© 2011 Mond et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons
Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in
any medium, provided the original work is properly cited.

Page 2

physical health and chronic medical conditions, is taken
into account in the analysis [5,6,8,21,24].
In any case, the observation that mental health may be
impaired in underweight women would seem to present a
paradox. That is, if it is accepted that impairment in
women’s mental health associated with obesity is due pri-
marily to the effects of body dissatisfaction and/or eating-
disordered behaviour, then it would seem reasonable to
hypothesise that low body weight would tend to be asso-
ciated with better mental health [25,26]. An ego-syntonic
effect of low body weight in women - and the fact such an
effect would not be expected in men - might help to
explain why mental health impairment has been observed
in underweight men - but not women - in some studies
[1,16,27]. On the other hand, a tendency for underweight
women to have lower levels of body dissatisfaction and
disordered eating might be counter-balanced by the pre-
sence of a sub-group of underweight women with very
high levels of body dissatisfaction and eating disorder
symptoms, namely, those with anorexia nervosa or
variants of this disorder not meeting formal diagnostic
criteria [28].
A reading of the literature indicates that an overrepre-
sentation of individuals with high levels of body dissatis-
faction and/or eating-disordered behaviour is, in fact, the
favoured explanation for the finding of mental health
impairment in underweight women. Thus, Ali & Lind-
strom [22] noted that body image distortion seems to be
associated with underweight among young women in the
industrialised world and that “anorexia and bulimia may
be considered as the most severe and ultimate causes of
underweight among young women” (p.324). Similarly,
Ford and colleagues [9] noted that low BMI may be signif-
icantly associated with female gender and, in turn, greater
weight loss goals when dieting and that lean individuals
are likely to be a heterogeneous group that includes
“healthy persons who exercise a lot, persons with eating
disorders and clinically or subclinically sick persons”
(p.26). The putative association between low body weight
and body dissatisfaction/eating-disordered behaviour has
been invoked as an explanation of the association between
low body weight and mental health impairment in at least
four other studies in which such an association has been
observed [3,14,21,23] as well as in research conducted in
other fields [26].
Importantly, however, measures of body dissatisfaction
and/or eating disorder psychopathology were not included
in any of these studies. Hence, the hypothesised associa-
tions between body weight, body dissatisfaction/eating-
disordered behaviour and mental health impairment could
not be tested. To our knowledge, only one epidemiological
study has included some assessment of body dissatisfac-
tion and/or eating-disordered behaviour, in addition to
body weight and mental health. In a community sample of
women aged 18 to 25 years, Becker et al [21] found that
the lifetime prevalence of any mental disorder was higher
in underweight women than in normal-weight women,
even after individuals with a lifetime diagnosis of anorexia
or bulimia nervosa were excluded from the underweight
group. However, findings from this study are difficult to
interpret because the assessment of both eating disorders
and other mental health problems was confined to the
presence or absence of disorders meeting formal diagnos-
tic criteria [29,30], the number of participants meeting
these criteria was small and there was no assessment of
body dissatisfaction.
In sum, there appears to be little direct evidence to
support - or refute - the popular notion that underweight
is associated with elevated levels of body dissatisfaction
and/or eating disordered behaviour in women or that
such an association accounts for mental health impair-
ment. With this in mind, the goals of the present study
were as follows. First, we sought to confirm that mental
health is in fact impaired in underweight women, when
compared with normal-weight women. Second, we tested
the hypothesis that body dissatisfaction and/or eating dis-
ordered behaviour are greater in underweight women
than in normal-weight women. If both of these condi-
tions held, then it would be possible to test the additional
hypothesis that impairment in mental health among
underweight women is accounted for by body dissatisfac-
tion and/or eating-disordered behaviour.
Methods
Study design and participants
The research was conducted as part of the Health and
Well-Being of Female ACT Residents Study, an epidemio-
logical study of disability associated with eating-disordered
behaviour among young adult women [19,30-36]. The
study was carried out in the Australian Capital Territory
(ACT) region of Australia, a highly urbanised region that
includes the city of Canberra (population of 314,000 in
2002). Young women were chosen because of the com-
paratively high prevalence of body dissatisfaction and eat-
ing-disordered behaviour in this population [37]. All
aspects of the study design and methods were approved by
the ACT Human Research Ethics Committee.
A detailed account of the study methods can be found in
several previous publications [19,30-36]. In brief, self-
report questionnaires were initially completed by 5,255
female ACT residents aged 18 to 42 years. The question-
naire included measures of eating-disordered behaviour
(including items assessing body dissatisfaction), health-
related quality of life, subjective quality of life, general psy-
chological distress, physical activity and demographic
information. Demographic variables assessed included: age
in years; country of birth (Australia, not Australia); first
language (English, not English); marital status (married,
Mond et al. BMC Public Health 2011, 11:547
http://www.biomedcentral.com/1471-2458/11/547
Page 2 of 10

Page 3

not married); parity (no children, one or more children);
main activity (employed full-time, not employed full-time);
educational attainment (bachelor’s degree or higher quali-
fication completed/not completed); and possession of (pri-
vate) health insurance (no, yes). Body mass index (BMI)
(kg/m2) was calculated from self-reported height and
weight [38].
The sample comprised approximately 10% of the total
population of young adult women in the ACT and was
representative of this population in terms of socio-demo-
graphic characteristics [34]. Thus, most participants were
born in Australia (85.3%), had English as their first
language (91.8%) and had completed 12 or more years of
formal education (90.5%). Fifty-five per cent of participants
were married or living as married, 43.8% had one or more
children, 62.8% were employed full- or part-time, 15.6%
were full-time students and 17.5% nominated home duties
as their main activity.
The mean (SD) age of participants was 30.3 (7.2) years.
The mean (SD) (BMI) among the 4,892 (93.1%) partici-
pants who provided details of both height and weight was
24.5 (5.3) kg/m2. Reflecting the study aims, participants in
the present study were the 231 women (4.7%) who were
underweight (BMI < 18.5) and 2,976 women (60.8%) who
were normal-weight (18.5 ≥ BMI < 25.0) according to the
conventional classification [39]. Findings relating to the
associations between obesity, eating-disordered behaviour
and mental health have been reported elsewhere
[19,30,36].
Study measures
Body dissatisfaction and eating disordered behaviour
Eating-disordered behaviour was assessed using the Eating
Disorder Examination Questionnaire (EDE-Q) [40], a
widely-used, 36-item, self-report measure that focuses on
the occurrence and frequency of key eating disorder atti-
tudes and behaviours during the past 28 days. Subscale
scores - relating to dietary intake/restraint, concerns about
eating and concerns about weight or shape - and a global
score, are derived from 22 items addressing attitudinal fea-
tures [34]. Scores on each (item and) scale range from 0 to
6, with higher scores indicating higher symptom levels.
Remaining items of the EDE-Q assess the occurrence and
frequency of specific eating disorder behaviours, namely,
binge eating, self-induced vomiting, misuse of laxatives or
diuretics, extreme dietary restriction and excessive exer-
cise. These items do not contribute to subscale scores.
Two of the EDE-Q (Weight/Shape Concerns subscale)
items specifically address body dissatisfaction, namely,
“How dissatisfied have you felt about your weight” and
“How dissatisfied have you felt about your shape”. The
average of scores on these 2 items, which were highly
correlated (r = 0.89), was used as a measure of body dis-
satisfaction in the present study [41].
Whereas the EDE-Q global score provided a continuous
measure of eating disorder psychopathology, eating disor-
der “cases” were identified using an operational definition
informed by our previous research, namely, the occurrence
of extreme weight or shape concerns in conjunction any
regular eating disorder behaviour [38,30]. For binge eating,
self-induced vomiting, and purging behaviours, “regular”
was defined as “at least weekly”, whereas regular extreme
dietary restriction and excessive exercise were recognised
if these behaviours occurred, on average, 3 or more times
per week [34]. Although, in the absence of interview
assessment, participants meeting these criteria should be
viewed as “probable” rather than “true” cases, the criteria
have been found to identify individuals with high levels of
eating disorder psychopathology and functional impair-
ment [38,30].
Mental health
Health-related quality of life
Health-related quality of life was assessed with the Medical
Outcomes Study (12-item) Short-Form disability scale (SF-
12) [42]. Items of the SF-12 are summarised into two
weighted scales (Physical Component Summary scale,
PCS; Mental Component Summary scale, MCS), designed
to assess physical and mental health impairment. Each
scale is scored to have a mean of 50 and standard devia-
tion of 10 (in the US population), with lower scores indi-
cating higher levels of impairment. The SF-12 has very
good psychometric properties, including demonstrated
validity in the Australian population [42,43]. PCS items
include “Does your health now limit you in moderate
activities, such as moving a table, vacuuming or playing
golf?” and “During the past four weeks, were you limited
in the kind or work or other activities undertaken as a
result of your physical health?”, whereas MCS items
include “During the past four weeks have you accom-
plished less than you would like as a result of any emo-
tional problems?” and “During the past four weeks how
much of the time have you felt calm and peaceful"? In the
present study, the SF-12 MCS was the outcome of interest
whereas physical health, as assessed by the SF-12 PCS, was
included as a covariate. Cronbach’s alpha was 0.82 for the
total scale and 0.80 for the 6 items comprising the MCS.
Subjective quality of life
Subjective quality of life was assessed using the World
Health Organization Brief Quality of Life Assessment
Scale (WHOQOL-BREF) [44,45], a 26-item measure yield-
ing scores on each of four domains relating to the indivi-
dual’s subjective evaluation of their physical health,
environmental health, psychological health and social rela-
tionships. Items are scored on a five-point, Likert-type
scale, with scores of ‘’1’’ and “5” indicating, respectively,
extreme dissatisfaction and extreme satisfaction. Only the
Psychological Functioning (QOL-P) subscale, which can
Mond et al. BMC Public Health 2011, 11:547
http://www.biomedcentral.com/1471-2458/11/547
Page 3 of 10

Page 4

be viewed as a measure of perceived satisfaction with key
aspects of emotional well-being, was considered in the
present study. Items of the QoL-P include “To what extent
do you feel your life to be meaningful"? and “How satisfied
are you with yourself?’ One of the (6) items comprising
the QOL-P, which addresses satisfaction with “bodily
appearance”, was excluded when calculating the scale
score. Cronbach alphas for the 5- and 6-item scales were,
respectively, 0.80 and 0.81.
General psychological distress
General psychological distress was assessed with the
Kessler Psychological Distress Scale (K-10), a 10-item
self-report measure designed for use in general popula-
tion surveys [46]. In Australia it is also used as an out-
come measure among individuals treated within mental
health services [47]. The frequency (during the past four
weeks) of each of 10 symptoms - relating to anxiety and
depressive mood - is measured on a scale from one to
five, such that total scores range from 10 to 50 with
lower scores indicating higher symptom levels. This cod-
ing method was employed - i.e. in preference to an alter-
native method in which lower scores indicate lower
symptoms levels [47] - so that lower scores would indi-
cate poorer mental health for all 3 mental health mea-
sures. Findings from the Australian National Survey of
Mental Health and Well-Being suggested that individuals
scoring in the extreme range (≤ 30) have a high probabil-
ity of meeting diagnostic criteria for an anxiety or affec-
tive disorder according to interview assessment [47].
Cronbach’s alpha in present study was 0.91.
Physical activity
In addition to the questions assessing the use of exercise
as a means of weight control (included in the EDE-Q),
three questions were included that assessed the fre-
quency of mild, moderate and hard exercise during the
past four weeks [35]. Based on these questions, a dichot-
omous variable was created such that participants who
reported any of the three forms of exercise on average
at least 3 times per week during the past four weeks
were considered to be regular exercisers.
Statistical analysis
Loess curves were used to examine the associations
between BMI, as a continuous variable, and each measure
of mental health - SF-12 MCS, QOL-P and K-10 - in the
total sample (n = 4,892). Loess, which stands for locally
weighted scatterplot smoothing, is a method for fitting a
curve to a scatter plot that provides a graphical represen-
tation of the relationship between two variables without
making any a priori assumptions about the form of that
relationship [48].
Independent-samples t-tests were used to compare
scores on continuous variables, namely, age, BMI, body
dissatisfaction, EDE-Q subscale scores and scores on the
SF-12 PCS and MCS, QOL-P and K-10, between
underweight and normal weight participants, whereas chi-
square tests were used to compare groups on categorical
outcomes, namely, demographic characteristics, the occur-
rence of specific eating disorder behaviours, the occur-
rence of regular physical activity and the prevalence of
probable eating disorder cases. Bivariate correlations were
calculated using the Pearson correlation coefficient. Linear
regression models [49] were used to test the hypothesis
that impairment in mental health associated with
low body weight, where this was observed, was accounted
for by body dissatisfaction and/or eating-disordered
behaviour.
A significance level of 0.05 was employed for all tests, all
tests were two-tailed and all analysis was conducted using
SPSS version 17.0. For analyses involving the SF-12, both
the standard scoring method, employing factor scores
derived by means of orthogonal rotation, and an alterna-
tive method, employing factor scores derived by means of
oblique rotation, were employed [50]. Since the main find-
ings were unchanged, only findings based on the standard
scoring method are reported.
Results
Figures 1, 2, and 3 show Loess curves of the relation-
ships between BMI and scores on the SF-12 MCS,
QOL-P, and K-10, respectively. As can be seen, both
very low and very high body weights were associated
with mental health impairment and this was the case for
all 3 measures.
As would be expected, moderate to high positive cor-
relations were observed between the different measures
Figure 1 Loess curve showing the association between body
mass index (BMI) (kg/m2) and mental health functioning, as
measured by the Medical Outcomes Study Short Form Mental
Component Summary Scale (SF-12 MCS), in the total sample (n
= 4,892) (Note: lower scores on the SF-12 MCS indicate greater
mental health impairment).
Mond et al. BMC Public Health 2011, 11:547
http://www.biomedcentral.com/1471-2458/11/547
Page 4 of 10

Page 5

of mental health (MCS, QOL-P: 0.71; QOL-P, K-10:
0.72; MCS, K-10: 0.76). Further, body dissatisfaction was
highly correlated with overall levels of eating disorder
psychopathology as measured by the EDE-Q Global
score (r = 0.82).
Underweight women were less likely to be married
(24.2% vs 37.2%; × = 15.51, p < 0.01), less likely to have
one or more children (30.1% vs 39.9%; × = 8.52.5, p <
0.01), less likely to have completed tertiary studies
(29.5% vs 41.7%; × = 12.95, p < 0.01) and less likely to
have private health insurance (50.5% vs 60.1%; × = 7.14,
p < 0.01) than normal weight-women, whereas the
groups did not differ with respect to employment, coun-
try of birth or first language (all p > 0.05).
Comparisons between underweight and normal-
weight participants on continuous variables are shown
in Table 1. As can be seen, underweight women were
younger and had lower scores (indicating higher levels
of impairment) on the SF-12 PCS and on all 3 mea-
sures of mental health, than normal-weight women,
although differences on the SF-12 MCS did not reach
statistical significance. It is also apparent that under-
weight women had lower levels of body dissatisfaction
and lower scores on each of the EDE-Q subscales than
normal-weight women.
There were no differences between groups with
respect to the occurrence of eating disorder behaviours
(all p > 0.05), nor with respect to the prevalence of
probable eating disorder cases (underweight: 3.5%; nor-
mal-weight: 5.8%; × = 2.27, p = 0.13). However, under-
weight women were less likely to be regular exercisers
than normal-weight women (47.7% vs 56.6%; × = 5.25,
p < 0.05).
Since underweight was associated with lower, rather
than higher, levels of body dissatisfaction and eating dis-
order psychopathology, there was no basis on which to
proceed with formal tests of the hypothesis of media-
tion. Post-hoc analysis was conducted, however, in order
to determine which variables (other than body dissatis-
faction and eating-disordered behaviour) might have
accounted for the observed association between low
body weight and mental health impairment and to eluci-
date the comparative importance of different variables in
accounting for mental health impairment among under-
weight participants.
For the first analysis, hierarchical linear regression was
used to determine if differences between groups in men-
tal health impairment remained after controlling for
potential covariates, namely, those variables that differed
between groups in bivariate analysis. A dichotomous
variable indicating weight status (underweight, normal-
weight) was used in place of BMI for this analysis. A
similar method was employed for the second analysis,
except that all variables were entered simultaneously,
body dissatisfaction and eating-disordered behaviour (as
measured by the EDE-Q global score) were included
and weight status was replaced with BMI.
Results of the first analysis are summarised in Table 2.
As can be seen, the association between weight status
and scores on the K-10 remained significant (p = 0.04)
after controlling for demographic variables and
approached significance (p = 0.07) after physical health
and physical activity were added to the model. For the
QOL-P, by contrast, the inclusion of demographic
Figure 2 Loess curve showing the association between body
mass index (BMI) (kg/m2) and subjective mental health, as
measured by the WHOQOL-BREF Psychological Health subscale
(QOL-P), in the total sample (n = 4,892) (Note: lower scores on
the QOL-P indicate greater mental health impairment).
Figure 3 Loess curve showing the association between body
mass index (BMI) (kg/m2) and general psychological distress,
as measured by the Kessler Psychological Distress Scale (K-10)
in the total sample (n = 4,892) (Note: lower scores on the K-10
indicate greater mental health impairment).
Mond et al. BMC Public Health 2011, 11:547
http://www.biomedcentral.com/1471-2458/11/547
Page 5 of 10

Page 6

variables alone resulted in the initial contribution of
weight status becoming non-significant. Similarly, for
the SF-12 MCS, weight status no longer approached sig-
nificance after demographic variables were included.
Results of the second set of analyses are summarised
in Table 3. As can be seen, body dissatisfaction was the
strongest - and only strong - predictor of greater mental
health impairment and this was the case for all 3 depen-
dent variables.
Discussion
Summary of main findings
We sought to evaluate the hypothesis that mental
health impairment in underweight women, where this
occurs, is due to an association between low body
weight and elevated levels of body dissatisfaction and/
or eating-disordered behaviour. To this end, subgroups
of underweight and normal-weight women recruited
from a large, general population sample were com-
pared on measures of body dissatisfaction, eating-
disordered behaviour and mental health. There were
two main findings. First, underweight women had sig-
nificantly greater impairment in mental health than
normal-weight women, even after controlling for dif-
ferences in demographic characteristics and physical
health. Second, there was no evidence that higher
levels of body dissatisfaction or eating-disordered
behaviour accounted for this difference. Rather, under-
weight women had significantly lower levels of body
dissatisfaction and eating-disordered behaviour than
normal-weight women. There was also no evidence
that greater mental health impairment in the under-
weight group may have been due to the presence of a
small number of individuals with very high levels of
eating disorder psychopathology, namely, individuals
likely to have an eating disorder.
Table 1 Mean (SD) age, BMI and scores on measures of body dissatisfaction, eating disorder features and mental
health for underweight and normal-weight women
Underweight
(n = 231)
Normal-Weight
(n = 2976)
Mean (SD)Mean (SD)tp
d
Age
BMI (kg/m2)
Eating Disorder
Examination Subscales
Body dissatisfaction
Restraint
Eating concern
Weight/Shape concern
Global score
SF-12 PCSi
SF-12 MCSii
QOL-Piii
K-10iv
26.87 (7.31)
17.52 (0.97)
29.71 (7.25)
21.90 (1.71)
-5.72
-38.55
< 0.01
< 0.01
0.39
3.15
1.38 (1.48)
0.75 (1.29)
0.45 (0.90)
1.18 (1.24)
0.83 (1.05)
49.74 (8.76)
44.83 (10.78)
3.59 (0.63)
41.85 (6.67)
2.15 (1.79)
1.15 (1.33)
0.57 (0.88)
1.78 (1.45)
1.24 (1.12)
50.87 (7.70)
46.06 (10.49)
3.69 (0.60)
43.27 (5.83)
-6.31
-4.38
-1.98
-6.14
-5.28
-2.08
-1.67
-2.28
-3.25
< 0.01
< 0.01
< 0.05
< 0.01
< 0.01
< 0.05
0.09
< 0.05
< 0.01
0.47
0.31
0.14
0.44
0.38
0.14
0.12
0.16
0.23
i SF-12 Physical Component Summary scale.
ii SF-12 Mental Component Summary scale.
iii WHOQOL-BREF Psychological Functioning subscale.
iv Kessler Psychological Distress Scale.
Table 2 Multiple linear regression analysis of the association between weight status (underweight, normal-weight)
and each measure of mental health (SF-12 MCS, QOL-P and K-10) with and without the inclusion of additional
covariates (demographic variables; physical health and physical activity) (n = 3,207)
SF-12
MCS
QOL-PK-10
Model Covariates
Weight statusi
BpR2
BpR2
BpR2
1
2
3
1.229
0.330
0.437
0.095 0.001
0.677 0.028
0.583 0.049
0.095
0.057
0.034
0.023 0.002
0.205 0.023
0.455 0.043
1.420 0.001 0.004
0.896 0.039 0.043
0.799 0.072 0.062
Weight status, demographic variablesii
Weight status, demographic variables, physical healthiii,
physical activity
i Underweight = 0; normal-weight = 1.
ii Age, marital status parity, educational attainment, private health insurance.
iii As measured by the SF-12 PCS.
Mond et al. BMC Public Health 2011, 11:547
http://www.biomedcentral.com/1471-2458/11/547
Page 6 of 10

Page 7

Study implications
The primary implication of the present study is that men-
tal health impairment in underweight women, where this
occurs, is unlikely to be accounted for by an association
between low body weight and elevated levels of body dis-
satisfaction or eating-disordered behaviour. Rather, body
dissatisfaction and eating disordered behaviour appear to
be comparatively uncommon among underweight women.
Interestingly, however, body dissatisfaction was still
strongly predictive of poor mental health in multivariable
analysis conducted within the underweight group. Taken
together, these findings suggest not only that higher levels
of body dissatisfaction or eating-disordered behaviour
among underweight women do not account for mental
health impairment, but also that lower levels of body dis-
satisfaction and/or eating disordered behaviour among
underweight women may counterbalance, to some extent,
mental health impairment due to other factors.
Consistent with findings from other recent epidemiolo-
gical studies [7,14], the prevalence of underweight was low
among women in the present study, less than 5%. Given
current concern surrounding the high prevalence of obe-
sity in industrialised nations, research addressing the
impact of underweight on mental health has not been a
priority. Indeed, underweight individuals have often been
excluded in studies of the association between body weight
and mental health due to concerns that high levels of
impairment among underweight individuals might compli-
cate interpretation of comparisons between obese and
non-obese individuals [3,5]. Similar concerns have arisen
in research addressing the association between obesity and
mortality [24]. However, it is important to critically evalu-
ate the validity of anecdotal evidence, particularly when
there are implications for public health practice. For exam-
ple, Ali & Lindstrom [22] noted that interventions to
improve psychological health in underweight women
would need to deal with the body norms/image messages
disseminated in the popular media. The present findings
argue against this view. The findings do suggest, however,
that women who are very underweight - like those who
are very overweight - are a vulnerable group, being at
increased risk of impairment in both physical and mental
health.
We can only speculate as to why the notion that low
body weight is associated with body dissatisfaction and/or
eating-disordered behaviour is so widely accepted when
there is so little evidence to support it. There may be poor
understanding of the epidemiology of eating-disordered
behaviour among researchers not familiar with this litera-
ture, for example, low awareness of the fact that eating dis-
orders characterised by normal or above-average body
weight far outnumber those characterised by low body
weight [29]. There may also be a tendency for public
health researchers to generalise from the clinical/hospital
setting, in which individuals presenting with the combina-
tion of low body weight and extreme concerns about
weight or shape are more conspicuous [51]. In any case,
our findings suggest that there is a need to address the
misconception that low body weight is associated with
body dissatisfaction and/or eating-disordered behaviour in
unselected samples.
Study limitations and other methodological
considerations
Several limitations of the present study should be noted.
First, some potentially important covariates were not
assessed. In particular, there was no assessment of
smoking or of chronic medical conditions, both of
which may be associated with low body weight and/or
mental health impairment [6,7,22,24]. The higher levels
Table 3 Multiple linear regression analysis of variables associated with each measure of mental health (SF-12 MCS,
QOL-P and K-10) among underweight women (n = 231)
SF-12 MCS
b
Age.017
BMI-.070
Marital status.102
Parity-.016
Employment-.103
Education-.063
Country of birth.013
First language.083
Health insurance.064
Physical health-.127
Physical activity.070
Body dissatisfaction-.400
Eating disorder psychopathology-.023
QOL-P
b
-.011
-.076
.016
-.019
-.030
.024
-.095
.208
.098
.179
.085
-.513
.039
K-10
b
-.019
-.094
.052
.060
-.083
.004
-.056
.151
.153
.086
.016
-.478
.077
Covariatesppp
.872
.344
.245
.872
.197
.454
.886
.361
.387
.082
.376
.000
.817
.912
.289
.855
.842
.695
.770
.266
.018
.167
.011
.271
.000
.683
.856
.206
.552
.548
.300
.958
.529
.097
.040
.229
.844
.000
.445
Mond et al. BMC Public Health 2011, 11:547
http://www.biomedcentral.com/1471-2458/11/547
Page 7 of 10

Page 8

of mental health impairment observed in underweight
women might also have been due to the presence of a
small number of individuals with very high symptom
levels, namely, those with anxiety, affective, substance
use or other mental disorders [6,7]. Interview assess-
ment would be required to test this hypothesis. Our
goal was to test the hypothesis that body dissatisfaction/
eating-disordered behaviour mediates the association
between low body weight and mental health impair-
ment, rather than to examine factors associated with
impairment.
Second, approximately 40% of individuals approached
to participate in the study chose not to return a com-
pleted questionnaire and individuals with anorexia or
variants of anorexia may be over-represented in this
subgroup [52]. To the extent that a bias of this kind
occurred, both the extent of mental health impairment
in the underweight group and the role of body dissatis-
faction/eating disorder psychopathology in accounting
for this impairment may have been underestimated.
Individuals with other mental disorders may also have
been over-represented among non-respondents [53].
However, these observations do not change the fact
that, in the present study, greater mental health impair-
ment was observed among underweight women despite
these women having lower levels of body dissatisfaction
and eating disorder psychopathology than normal-
weight women.
Third, the present findings necessarily apply to under-
weight defined as a BMI of < 18.5 kg/m2. Although this
criterion is widely accepted, it is nevertheless arbitrary
and different findings may have been observed had a
more or less stringent operation definition of low body
weight been employed [1,22]. In addition, BMI was cal-
culated based on self-reported height and weight in the
present study. However, we found very good agreement
between BMI based on self-reported height and weight
and BMI derived from actual (measured) height and
weight in pilot work [38].
Fourth, the present findings necessarily apply to
younger women from an urbanised, affluent region. This
population was appropriate for an initial study because
the hypothesis that impairment in mental health asso-
ciated with low body weight is due to body dissatisfaction
and/or eating-disordered behaviour has been proposed
primarily in relation to young women from industrialised
nations [22,26]. As suggested previously, it may make
more sense to consider the role of body dissatisfaction in
relation to mental health impairment in underweight
men, given that underweight males are more likely to be
dissatisfied with their bodies than normal-weight males
and given that the prevalence of body dissatisfaction and
its impact on mental health may be increasing in males
[1,20,27,54].
Some comment is warranted concerning the treat-
ment of body dissatisfaction and eating-disordered
behaviour as distinct constructs. The key distinction
between individuals with extreme weight or shape con-
cerns and individuals with eating disorders is the regu-
lar occurrence of one or more eating disorder (i.e.
binge eating or extreme weight-control) behaviours.
Since extreme weight or shape concerns in the absence
of eating disorder behaviours are more common than
the combination of concerns and behaviours, it is not
surprising that body dissatisfaction emerged as the
stronger predictor of impairment among underweight
participants. But it needs to be remembered that there
is extensive overlap between these constructs in gen-
eral population samples [41].
Finally, since this was a cross-sectional study, the
usual caveats concerning the direction of any observed
associations apply [4,55]. The available evidence from
longitudinal studies suggests that associations between
body dissatisfaction/eating disordered behaviour and
mental health impairment are likely to be bidirectional
[56-58]. Notable strengths of the present research were
the recruitment of a large, general population sample of
women, comprehensive assessment of eating-disordered
behaviour and the inclusion of three different measures
of mental health.
Conclusions
To conclude, the findings of the present study suggest
that mental health impairment in underweight women,
where this occurs, is unlikely to be due to higher levels
of body dissatisfaction or eating-disordered behaviour.
Rather, lower levels of body dissatisfaction and eating-
disordered behaviour among underweight women may
counterbalance, to some extent, impairment due to
other factors. The findings also suggest that women
who are very underweight are a vulnerable group, being
at increased risk of impairment in both physical and
mental health.
Ethics approval
The research was conducted with the approval of the
ACT Human Research Ethics Committee.
Acknowledgements
The Health and Well-Being of Female ACT Residents Study was funded by
The Canberra Hospital Private Practice Fund, ACT Health and Community
Care and ACT Mental Health. Dr Mond is supported by a National Health
and Medical Research Council Sidney Sax Fellowship.
Author details
1School of Sociology, Australian National University, Canberra, Australia.
2Australian Demographic & Social Research Institute, Australian National
University, Canberra, Australia.3School of Medicine, University of Western
Sydney, Campbelltown, Australia.4Rural Clinical School, Medical School,
Australian National University, Canberra, Australia.
Mond et al. BMC Public Health 2011, 11:547
http://www.biomedcentral.com/1471-2458/11/547
Page 8 of 10

Page 9

Authors’ contributions
JM was responsible for the design and conduct of the research as well as
data processing, data analysis and manuscript preparation. BR, PH and CO
contributed to the design and conduct of the research and to critical
revision of an earlier version of the manuscript. BR contributed to data
analysis and interpretation. All authors read and approved the final
manuscript.
Competing interests
The authors declare that they have no competing interests.
Received: 8 February 2011 Accepted: 10 July 2011
Published: 10 July 2011
References
1. Carpenter KM, Hasin DS, Allison DB, Faith MS: Relationships between
obesity and DSM-IV Major Depressive Disorder, suicide ideation, and
suicide attempts: results from a general population study. Am J Public
Health 2000, 90:251-257.
2. Onyike CU, Crum RM, Lee HB, Lyketsos CG, Eaton WW: Is obesity
associated with major depression? Results from the Third National
Health and Nutrition Examination Survey. Am J Epidemiol 2003,
158:1139-1147.
3.Jorm AF, Korten AE, Christensen H, Jacomb PA, Rodgers B, Parslow RA:
Association of obesity with anxiety, depression and emotional well-
being: a community survey. Aust N Z J Public Health 2003, 27:434-440.
4. Simon GE, Von Korff M, Saunders K, Miglioretti DL, Crane PK, van Belle G,
Kessler RC: Association between obesity and psychiatric disorders in the
US adult population. Arch Gen Psychiatry 2006, 63:824-830.
5.Scott KM, Bruffaerts R, Simon GE, Alonso J, Angermeyer M, de Girolamo G,
Demyttenaere K, Gasquet I, Haro JM, Karam E, Kessler RC, Levinson D,
Medina Mora ME, Oakley Browne M, Ormel JH, Posada Villa J, Uda H, Von
Korff M: Obesity and mental disorders in the general population: results
from the world mental health surveys. Int J Obes 2008, 32:192-200.
6.Zhao G, Ford ES, Dhingra S, Li C, Strine TW, Mokdad AH: Depression and
anxiety among US adults: associations with body mass index. Int J Obes
2009, 33:257-266.
7. Chen Y, Jiang Y, Mao Y: Association between obesity and depression in
Canadians. J Women’s Health 2009, 18:1687-1692.
8.Doll HA, Peterson SEK, Stewart-Brown SL: Obesity and physical and
emotional well-being: Associations between body mass index, chronic
illness, and the physical and mental components of the SF-36
questionnaire. Obes Res 2000, 8:160-170.
9. Ford ES, Moriarty DG, Zack MM, Mokdad AH, Chapman DP: Self-reported
body mass index and health-related quality of life: findings from the
Behavioral Risk Factor Surveillance System. Obes Res 2001, 9:21-31.
10.Larsson U, Karlsson J, Sullivan M: Impact of overweight and obesity on
health-related quality of life - a Swedish population study. Int J Obes
2002, 26:417-424.
11.Heo M, Allison DB, Faith MS, Zhu S, Fontaine KR: Obesity and quality of
life: mediating effects of pain and comorbidites. Obes Res 2003,
11:209-216.
12. Jia J, Lubetkin RI: The impact of obesity on health-related quality-of-
life in the general adult US population. J Public Health 2005,
27:156-164.
13. Mond JM, Baune BT: Overweight, medical comorbidity and health-related
quality of life in a community sample of women and men. Obesity 2009,
17:1627-1634.
14. Renzaho A, Wooden M, Houng B: Associations between body mass index
and health-related quality of life among Australian adults. Qual Life Res
2010, 19:515-520.
15.Heo M, Pietrobelli A, Fontaine KR, Sirey JA, Faith MS: Depressive mood and
obesity in U.S. adults: comparison and moderation by sex, age, and
race. Int J Obes 2006, 30:513-519.
16.McLaren L, Beck CA, Patten SB, Fick GH, Adair CE: The relationship
between body mass index and mental health: A population-based study
of the effects of the definition of mental health. Soc Psychiatry Psychiatr
Epidemiol 2008, 43:63-71.
17. Herva A, Laitinen J, Miettunen J, Veijola J, Karvonen JT, Laksy K,
Joukamaa M: Obesity and depression: results from the longitudinal
Northern Finland 1966 Birth Cohort Study. Int J Obes 2006, 30:520-527.
18. Desai RA, Manley M, Desai MM, Potenza MN: Gender Differences in the
Association Between Body Mass Index and Psychopathology. CNS Spectr
2009, 14:372-383.
Mond JM, Rodgers B, Hay PJ, Owen C, Baune BT: Obesity and impairment
in psycho-social functioning: the mediating role of eating-disordered
behavior. Obesity 2007, 15:2769-2779.
Mond JM, van den Berg P, Boutelle K, Neumark-Sztainer D, Hannan PJ:
Obesity, body dissatisfaction, and psycho-social functioning in early and
late adolescence: findings from the Project EAT Study. J Adolesc Health
2011, 48:373-378.
Becker ES, Margraf J, Turke V, Soeder U, Neumer S: Obesity and mental
illness in a representative sample of young women. Int J Obes 2001,
25(Supp 1):5-9.
Ali SM, Lindstrom M: Socioeconomic, psychosocial, behavioural, and
psychological determinants of BMI among young women: differing
patterns for underweight and overweight/obesity. Eur J Public Health
2005, 16:324-330.
Molarius A, Berglund K, Eriksson C, Eriksson HG, Lindén-Boström M,
Nordström E, Persson C, Sahlqvist L, Starrin B, Ydreborg B: Mental health
symptoms in relation to socio-economic conditions and lifestyle factors
- a population-based study in Sweden. BMC Public Health 2009, 9:302.
Kelly SJ, Lilley JM, Leonardi-Bee J: Associations of morbidity in the
underweight. Eur J Clin Nutr 2010, 64:475-482.
Mond JM, Robertson-Smith G, Vitere A: Stigma and eating disorders: is
there evidence of negative attitudes towards individuals suffering from
anorexia nervosa? J Mental Health 2006, 15:519-532.
Smeesters D, Mussweiler T, Mandel N: The effects of thin and heavy
media images on overweight and underweight consumers: Social
comparison processes and behavioral implications. J Consum Res 2010,
36:930-949.
Kostanski M, Fisher A, Gullone E: Current conceptualisation of body image
dissatisfaction: have we got it wrong? J Child Psychol Psychiat 2004,
45:1317-1325.
American Psychiatric Association: Diagnostic and statistical manual of mental
disorders. 4 edition. (DSM-IV). Washington, DC: Author; 1994.
Mond JM, Rodgers B, Hay PJ, Korten A, Owen C, Beumont PJV: Disability
associated with community cases of commonly occurring eating
disorders. Aust N Z J Public Health 2004, 28:246-251.
Mond JM, Hay PJ, Rodgers B, Owen C: Comparing the health burden of
overweight and eating-disordered behavior in young adult women.
J Women’s Health 2009, 18:1081-1089.
Mond JM, Hay PJ, Rodgers B, Owen C, Crosby R, Mitchell JE: Use of
extreme weight control behaviors with and without binge eating in a
community sample of women: Implications for the classification of
bulimic-type eating disorders. Int J Eat Disord 2006, 39:294-302.
Mond JM, Hay PJ, Rodgers B, Owen C, Mitchell JE: Correlates of the use of
purging and non-purging methods of weight control in a community
sample of women. Aust N Z J Psychiatry 2006, 40:136-142.
Mond JM, Hay PJ, Rodgers B, Owen C: Recurrent binge eating with and
without the “undue influence of weight or shape on self-evaluation”:
Implications for the diagnosis of binge eating disorder. Behav Res Ther
2006, 45:929-938.
Mond JM, Hay PJ, Rodgers B, Owen C: Eating Disorder Examination
Questionnaire (EDE-Q): Norms for young adult women. Behav Res Ther
2006, 44:53-62.
Mond JM, Hay PJ, Rodgers B, Owen C: An update on the definition of
“excessive exercise” in eating disorders research. Int J Eat Disord 2006,
39:147-153.
Darby A, Hay PJ, Mond JM, Rodgers B, Owen C: Disordered eating
behaviours and cognitions in young women with obesity: relationship
with psychological status. Int J Obes 2007, 31:876-82.
Hay PJ, Mond JM, Buttner P, Darby A: Eating disorder behaviors are
increasing: findings from two sequential community surveys in South
Australia. PLoS ONE 2007, 3:e1541.
Mond JM, Hay PJ, Rodgers B, Owen C, Beumont PJV: Validity of the Eating
Disorder Examination Questionnaire (EDE-Q) in screening for eating
disorders in community samples. Behav Res Ther 2004, 42:551-567.
National Heart Lung, Blood Institute in cooperation with the National
Institute of Diabetes and Digestive and Kidney Diseases: Clinical guidelines
on the identification, evaluation, and treatment of overweight and
obesity in adults. Washington, DC: Natl Inst Health; 1998.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
Mond et al. BMC Public Health 2011, 11:547
http://www.biomedcentral.com/1471-2458/11/547
Page 9 of 10

Page 10

40.Fairburn CG, Beglin SJ: Assessment of eating disorders: Interview or self-
report questionnaire? Int J Eat Disord 1994, 16:363-370.
Mond JM, Hay PJ: Dissatisfaction vs over-evaluation in a community
sample of women. Int J Eat Disord .
Ware JE, Kosinski M, Keller SD: A 12-item short-form health survey:
Construction of scales and preliminary tests of reliability and validity.
Med Care 1996, 34:220-233.
Sanderson K, Andrews G: The SF-12 in the Australian population: Cross-
validation of item selection. Aust N Z J Public Health 2002, 26:343-345.
WHOQOL Group: Development of the World Health Organization
WHOQOL-BREF quality of life assessment. Psychol Med 1998, 28:551-558.
Skevington SM, Lofty M, O’Connell KA: The World Health Organization’s
WHOQOL-BREF quality of life assessment: Psychometric properties and
results of an international field trial. A report from the WHOQOL Group.
Qual Life Res 2004, 13:299-310.
Kessler RC, Andrews G, Colpe LJ, Hiripi E, Mroczek DK, Normand SLT,
Walters EE, Zaslavsky AM: Short screening scales to monitor population
prevalences and trends in non-specific psychological distress. Psychol
Med 2002, 32:959-976.
Andrews G, Slade T: Interpreting scores on the Kessler Psychological
Distress Scale (K10). Aust N Z J Public Health 2001, 25:494-497.
Cleveland WS, Devlin SJ: Locally weighted regression: An approach to
regression analysis by local fitting. J Amer Statistical Assoc 1988,
83:596-610.
Frazier PA, Tix AP, Barron KE: Testing moderator and mediator effects in
counseling psychology research. J Counseling Psychol 2004, 151:115-34.
Windsor TD, Rodgers B, Butterworth P, Anstey KJ, Jorm AF: Measuring
physical and mental health using the SF-12: Implications for community
surveys of mental health. Aust N Z J Psychiatry 2006, 40:797-803.
Munn MA: A real epidemiological study? Eur J Clin Nutr 2005, 59:621.
Mond JM, Rodgers B, Hay PJ, Owen C, Beumont PJV: Non-response bias in
a general population survey of eating-disordered behaviour. Int J Eat
Disord 2004, 36:89-98.
Henderson S, Conclusion: the central issues: Unmet Need in Psychiatry.
Edited by: Andrews G, Henderson S. Cambridge: Cambridge University
Press; 2000:422-428.
Austin SB, Haines J, Veugelers PJ: Body satisfaction and body weight:
Gender differences and sociodemographic determinants. BMC Public
Health 2009, 9:313.
Kasen S, Cohen P, Chen H, Must A: Obesity and psychopathology in
women: a three decade prospective study. Int J Obes 2008, 32:558-566.
Neumark-Sztainer D, Paxton SJ, Hannan PJ, Haines J, Story M: Does body
satisfaction matter? Five-year longitudinal associations between body
satisfaction and health behaviors in adolescent females and males.
J Adolesc Health 2006, 39:244-251.
Paxton SJ, Eisenberg ME, Neumark-Sztainer D: Prospective predictors of
body dissatisfaction in adolescent girls and boys: A five-year
longitudinal study. Dev Psychol 2006, 42:888-899.
Patton GC, Coffey C, Carlin JB, Sanci L, Sawyer S: Prognosis of adolescent
partial syndromes of eating disorder. Br J Psychiatry 2008, 192:294-299.
41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
51.
52.
53.
54.
55.
56.
57.
58.
Pre-publication history
The pre-publication history for this paper can be accessed here:
http://www.biomedcentral.com/1471-2458/11/547/prepub
doi:10.1186/1471-2458-11-547
Cite this article as: Mond et al.: Mental health impairment in
underweight women: do body dissatisfaction and eating-disordered
behavior play a role? BMC Public Health 2011 11:547.
Submit your next manuscript to BioMed Central
and take full advantage of:
• Convenient online submission
• Thorough peer review
• No space constraints or color figure charges
• Immediate publication on acceptance
• Inclusion in PubMed, CAS, Scopus and Google Scholar
• Research which is freely available for redistribution
Submit your manuscript at
www.biomedcentral.com/submit
Mond et al. BMC Public Health 2011, 11:547
http://www.biomedcentral.com/1471-2458/11/547
Page 10 of 10