Full text of DOI:10.1038/oby.2002.71

Relationship between Body Mass Index and the
Use of Healthcare Services in Australia
Daniel D. Reidpath, David Crawford, Linda Tilgner, and Carl Gibbons
Abstract
Introduction
REIDPATH, DANIEL D., DAVID CRAWFORD, LINDA
TILGNER, AND CARL GIBBONS. Relationship between
body mass index and the use of healthcare services in
Australia. Obes Res. 2002;10:526–531.
Objective: To examine the relationship between body
mass index (BMI) and the use of medical and preventive
health services.
The prevalence of obesity in industrialized countries is
high and rising (1). In Australia, for instance, it is estimated
that more than half the population is overweight or obese
(2). The high prevalence of overweight and obesity in
Australia is paralleled in other industrialized settings such
as Canada (3), the United States (4), and Europe (5). This
“obesity epidemic” carries with it the associated problems
of the rising rates of its comorbidities (1). Obesity-related
diseases include non-insulin-dependent diabetes mellitus,
coronary heart disease, hypertension, and various cancers
including breast, cervical, ovarian, gall bladder, prostate,
and colon cancer (4,6,7). Beyond the rising incidence of
conditions associated with obesity is the broader issue of the
burden that obesity places on health services in general.
Most research examining the burden of obesity on
health services have analyzed it in terms of economic
cost (3,6–11). One approach to estimating the cost of
obesity has been to estimate the direct cost to the health-
care system, such as the approach of Segal et al. (7) who
estimated the direct costs of obesity to the Australian
healthcare system, or Colditz (9) who estimated the direct
costs to the United States healthcare system. For exam-
ple, in 1995 Colditz estimated the direct cost of obesity to
be $70 billion. An alternative approach is to consider the
economic burden of obesity factoring in the direct costs
of healthcare, as well as the indirect costs of lost pro-
ductivity (12). However, regardless of how one calculates
these costs, obesity and the conditions related to it ac-
count for substantial healthcare dollars.
Research Methods and Procedures: This study involved sec-
ondary analysis of weighted data from the Australian 1995
National Health Survey. The study was a population survey
designed to obtain national benchmark information about a
range of health-related issues. Data were available from 17,033
men and 17,174 women, Ն20 years or age. BMI, based on
self-reported weight and height, was analyzed in relation to the
use of medical services and preventive health services.
Results: A positive relationship was found between BMI
and medical service use, such as medication use, visits to
hospital accident and emergency departments (for women
only); doctor visits, visits to a hospital outpatient clinics;
and visits to other health professionals (for women only). A
negative relationship was found in women between BMI
and preventive health services. Underweight women were
found to be significantly less likely to have Papanicolaou
smear tests, breast examinations, and mammograms.
Discussion: This research shows that people who fall out-
side the healthy weight range are more likely to use a range
of medical services. Given that the BMI of industrialized
populations appears to be increasing, this has important
ramifications for health service planning and reinforces the
need for obesity prevention strategies at a population level.
Unfortunately, the dollar-cost of obesity reveals little
about the day-to-day impact of obesity on the healthcare
system. This is better captured by statistics on health-service
use. Surprisingly, the relationship between weight and
health-service use has received relatively little attention,
although there are a few exceptions. These exceptions have
been discussed in two recent reviews by Fontaine and
Bartlett (13) and Zayat et al. (14), one on the use of
preventive health services by obese patients (13) and the
other on the use of medical care by patients with obesity
(14). One case-control study based in a primary-care setting
Key words: body mass index, medical-service use, pre-
ventive screening, Australia, obesity
Submitted for publication October 15, 2001.
Accepted for publication in final form February 7, 2002.
School of Health Sciences, Deakin University, Burwood, Australia.
Address correspondence to Daniel D. Reidpath, School of Health Sciences, Deakin Univer-
sity, 221 Burwood Highway, Burwood 3125, Australia.
E-mail: reidpath@deakin.edu.au
Copyright © 2002 NAASO
526 OBESITY RESEARCH Vol. 10 No. 6 June 2002

Body Mass Index and Health Service Utilization, Reidpath et al.
(n ϭ 194), for instance, observed higher healthcare use with
increasing BMI (15). This relationship was also observed in
a much larger study (n ϭ 17,118) based on a survey of
members of a health-service organization (16). It is note-
worthy that there are few large survey studies using repre-
sentative samples (17–19). Where the studies do exist, they
are limited geographically and often, though not exclu-
sively, limited to a subpopulation. Fontaine et al. (19), for
example, studied only women, and Brown et al. (17) limited
their study to women 45 to 49 years of age. The study by
Trakas et al. (18) is the only existent study of health-service
use based on a nationally representative sample, and this
study did not include the use of preventive health services.
Improved knowledge about the impact of obesity on
health services requires more research based on nationally
representative samples that specifically examine health-ser-
vice use. This would supplement and enhance the economic
analyses that already exist. The aim of the present study is
to examine the relationship between BMI and the use of
medical and preventive health services based on data from
the 1995 Australian National Health Survey (20).
To understand the context of the present study, one
needs to understand that Australia uses a primary-health-
care model. Visits to specialists occur generally with only
a referral from general (medical) practitioners; costs are
subsidized (sometimes up to 100%) through a universal
healthcare system (“Medicare”) that covers all citizens,
permanent residents, and some temporary residents. The
healthcare is partially funded through a national levy on
income and covers such things as essential drugs, non-
elective medical procedures in public hospitals, and visits
to medical practitioners.
grams) by the square of their self-reported height
(meters). A categorical measure of BMI was also derived
according to specifications defined by the Australian
National Health and Medical Research Council (21). The
four BMI categories were: underweight (Ͻ20 kg/m2);
normal weight (20 to 25 kg/m2); overweight (Ͼ25 to 30
kg/m2); and obese (Ͼ30 kg/m2).
Participants in the National Health Survey were asked
about their use of a range of health services in the 2 weeks
before the interview, including hospitalization, admission to
a hospital accident and emergency department, visits to
outpatient clinics (i.e., a visit to a hospital clinic that does
not involve admission, such as attendance at a pain man-
agement clinic, or attendance at a hospital clinic to receive
physical therapy), visits to doctors, and visits to other health
professionals. Participants also reported whether they had
used any medications in the last 2 weeks. Women were
additionally asked about their use of preventive health ser-
vices including mammography, manual breast examination,
and Papanicolaou (pap) smears. Responses to these latter
questions were reduced to binary outcomes according to
whether a woman had a pap smear or a breast examination
in the last 2 years. Mammography outcomes were similarly
classified, but only for women Ն50 years of age.
Age was recorded by the Australian Bureau of Statistics
categorically in 5-year blocks. Estimated continuous age
was calculated by adjusting each participant’s age to the
midpoint of his or her age block.
A measure of income was calculated, which estimated the
income of the highest-earning individual in each partici-
pant’s family. Because income figures were recorded cate-
gorically in $5000 blocks in the survey, it was necessary to
adjust income to the midpoint of the income block.
Statistical Analysis
Research Methods and Procedures
The relationship between measures of health-service use
and BMI was tested using logistic regression models. Sep-
arate models were developed using the continuous or the
categorical measure of BMI as the predictor variables. Age
and estimated income were included in all models as control
variables. We developed separate models using continuous
and categorical BMI because of some evidence suggesting
that analyses using categorical BMI (unlike continuous
BMI), based on self-reported height and weight, may have
unreasonably high levels of systematic error (22–25).
In the regression models using categorical BMI, the
normal-weight group (BMI, 20 to 25 kg/m2) was used as
the reference category. In the regression models using
continuous BMI, participants with a BMI of Ͻ20 kg/m2
were excluded, because of the anticipated nonlinear
relationship (17).
Data
This study involved a secondary analysis of data from the
1995 Australian National Health Survey (20). The survey
was the most recent in a series of health surveys conducted
by the Australian Bureau of Statistics. It utilized a multi-
stage cluster sample of households in all six Australian
states and two territories. Information was obtained by
personal interviews. The total survey sample was 53,790
individuals (26,434 men and 27,356 women) representing a
response rate of ϳ97%. The present study was limited,
however, to adults Ն20 years of age (n ϭ 37,616). With the
exclusion of the 6.4% of cases with missing information
regarding height, weight, and/or income data, the working
sample was 35,207 (17,033 men and 17, 174 women).
Measures
National Health Survey participants were asked for
their height and weight. Continuous BMI was calculated
by dividing a participant’s self reported weight (kilo-
Results
The percentage of men and women in this Australian
study that reported using the medical or preventive health
OBESITY RESEARCH Vol. 10 No. 6 June 2002 527

Body Mass Index and Health Service Utilization, Reidpath et al.
Table 1. Percentage of people that reported using
medical or preventive health services
Table 2. Distribution of BMI category by sex
n (%)
n (%)
BMI category
Females
Males
Total
Men
Women
Underweight
(BMI Ͻ 20)
Acceptable
Medication use
12,135 (54.1) 17,391 (63.6)
2396 (14.0) 664 (3.9)
3060 (8.9)
Hospitalization
190 (0.7)
236 (0.9)
Accident and emergency
room visits
Outpatient clinic visits
Doctor visits
(BMI 20 to 25) 8284 (48.2) 7287 (42.8) 15,571 (45.5)
Overweight
(BMI 25 to 30) 4335 (25.2) 7042 (41.3) 11,377 (33.3)
Obese
262 (1.0)
567 (2.1)
4652 (17.6)
192 (0.7)
637 (2.3)
6211 (22.7)
Other health professional
visits
Regular pap smear tests
Regular breast examinations
Regular mammograms
(BMI Ͼ30)
2159 (12.6) 2040 (12.0) 4199 (12.3)
2272 (8.6)
3163 (11.6)
5667 (64.3)
3105 (33.1)
2347 (29.4)
BMI, body mass index.
A similar, though not identical, pattern of medical
service use was observed in women. Overweight (OR ϭ
1.32) and obese (OR ϭ 1.68) women were more likely to
have taken medication in the 2 weeks before interview
than women of normal weight. Obese women were also
more likely to have visited a hospital outpatient clinic
(OR ϭ 1.80) than were normal-weight women in the 2
weeks before interview. Compared with normal-weight
women, those who were overweight and obese were more
likely to have made a visit to a doctor (OR ϭ 1.15 and
OR ϭ 1.30, respectively) or other health professional
(OR ϭ 1.17 and OR ϭ 1.27, respectively).
services is shown in Table 1. Medication was the most
frequently used form of health service by both men (54%)
and women (64%). Visits to a doctor were also relatively
common events with ϳ18% of men and ϳ23% of women
having visited a doctor in the 2 weeks before interview.
Table 1 shows that in the 2 years before interview, large
numbers of women Ͼ50 years of age (67 to 71%) had
neither a breast examination nor a mammogram. A surpris-
ingly large number of the women sampled had not had a pap
smear in the 2 years before the interview (36%).
Self-reported height and weight data were used to esti-
mate BMI. A summary of the distribution of BMI is shown
in Table 2. It was more common for men (ϳ43%) and
women (48%) to be of normal weight than of any other
weight category. This notwithstanding, a similar proportion
of men were overweight and of normal weight. This was not
the case for women, who were half as likely to have been
overweight than of normal weight. Approximately 12% of
people were obese, and it was almost as common among
men (12%) as it was among women (ϳ13%).
The relationship between BMI and health-service use
was, in the first instance, examined by BMI categories
(underweight, normal weight, overweight, and obese) in
logistic regression, controlling for age and income. Normal
weight was treated as the reference category, and data were
analyzed separately for men and women (Table 3).
Among men, as BMI increased so too did the use of
medical services in the 2 weeks before interview. This was
true for medication usage [overweight, odds ratio (OR) ϭ
1.13; obese OR ϭ 1.46], visits to outpatient clinics (obese,
OR ϭ 1.62) and visits to a doctor (obese, OR ϭ 1.20).
Being under weight was also significantly associated with
hospitalization (OR ϭ 2.83) and visits to outpatient clinics
(OR ϭ 2.13).
An association between BMI in women and the use of
preventive health services was only observable in under-
weight women. Underweight women were significantly less
likely than normal-weight women to have regular pap
smears (OR ϭ 0.81), regular breast examinations (OR ϭ
0.74), and, among women Ͼ50 years of age, to have a
regular mammogram (OR ϭ 0.67).
Because the BMI data analyzed here were based on
self-reported heights and weights, and the estimated BMI
category is less stable than estimated continuous BMI, the
analyses were repeated using continuous BMI. Only indi-
viduals with an estimated BMI of Ն20 were included in the
analyses (Table 4).
In men, increasing BMI was significantly associated with
medication usage (OR ϭ 1.03) and visits to a doctor (OR ϭ
1.02) in the 2 weeks before the interview. Although the ORs
were much smaller than the equivalent ORs in the compar-
ison of BMI category, it should be remembered that these
differences related to increases in single BMI units. Thus,
compared with a man who had a BMI of 23, a man who had
a BMI of 28 was 15% more likely to have taken medication.
In women, increasing BMI was significantly associated with
the use of a range of medical services. This included medica-
tion usage (OR ϭ 1.04), admission to a hospital accident and
528 OBESITY RESEARCH Vol. 10 No. 6 June 2002

Body Mass Index and Health Service Utilization, Reidpath et al.
Table 3. Tabulation of relative ORs among subjects in different BMI categories using NHS subjects’ self-reported
height and weight and controlling for age and income
Underweight
BMI (<20)
Normal* BMI Overweight
(20 to 25) BMI (25 to 30)
Obese BMI
CI
CI
(>30)
CI
Men
Medication use†
0.93
2.83†
0.78 to 1.10
1.46 to 5.51
1
1
1.13†
1.14
1.05 to 1.22
0.74 to 1.76
1.46† 1.30 to 1.64
Hospitalization†
Accident and emergency
room visits
Outpatient clinic visits†
Doctor visits†
1.05
0.56 to 2.01
0.80
2.13†
0.93
0.29 to 2.24
1.39 to 3.25
0.76 to 1.15
1
1
1
1.14
1.25
1.06
0.76 to 1.71
0.98 to 1.60
0.97 to 1.16
1.42
0.83 to 2.44
1.62† 1.17 to 2.23
1.20† 1.06 to 1.36
Other health professional
visits
1.02
0.76 to 1.37
1
1.02
0.90 to 1.16
1.07
0.89 to 1.29
Women
Medication use†
Hospitalization
Accident and emergency
room visits
Outpatient clinic visits†
Doctor visits†
1.03
1.21
0.92 to 1.14
0.75 to 1.95
1
1
1.32†
0.95
1.20 to 1.45
0.63 to 1.45
1.68† 1.48 to 1.91
1.46
1.70
0.93 to 2.30
0.94 to 3.09
0.94
1.01
1.06
0.48 to 1.84
0.73 to 1.40
0.95 to 1.18
1
1
1
1.49
1.24
1.15†
0.90 to 2.48
0.96 to 1.60
1.05 to 1.26
1.80† 1.36 to 2.37
1.30† 1.16 to 1.45
Other health professional
visits†
Regular pap smear tests†
Regular breast
1.09
0.81†
0.94 to 1.26
0.68 to 0.96
1
1
1.17†
0.90
1.04 to 1.32
0.79 to 1.03
1.27† 1.09 to 1.47
0.88
0.74 to 1.04
examinations†
Regular mammograms†
0.74†
0.67†
0.63 to 0.87
0.48 to 0.94
1
1
0.99
0.87
0.87 to 1.12
0.70 to 1.08
0.99
0.87
0.84 to 1.16
0.67 to 1.15
OR, odds ratio; BMI, body mass index; NHS, National Health Survey; CI, confidence interval.
* Reference group.
† p Ͻ 0.05 in logistic regression model.
emergency department (OR ϭ 1.05), a visit to a hospital
outpatient clinic (OR ϭ 1.03), a visit to a doctor (OR ϭ 1.02),
and a visit to a non-medically trained health profes-
sional (OR ϭ 1.02). There was no significant association
in women between BMI and the use of preventive health
services when underweight women were excluded from the
analysis.
and consultations with doctors and other health profes-
sionals. Health-service use increased along with severity
of overweight, being higher in those who were obese
compared with those overweight, a finding that is con-
sistent with Quesenbury et al. (16), in their study of
health maintenance organization members. The relation-
ship between service use and BMI was not as clear
among men as it was among women in our study. Here,
increasing BMI was associated with greater medication
use and frequency of doctor visits. However, being un-
derweight was found to be associated with a greater
likelihood of hospitalization. A curvilinear relationship
was found for visits to outpatient clinics, with both
underweight and obese men having the highest rates of
outpatient clinic visits. The present findings are consis-
tent with the previous literature as it relates to women but
less so for men (15–17,19). The findings also reinforce
the need for more international data.
Discussion
This study is important because it provides nationally
representative data on the impact of obesity on medical
and preventive health services. It shows that among
women in Australia, there is a clear relationship between
BMI category and medical-service use. Increasing BMI
was associated with an increased use of medical services.
Specifically, women who were overweight had a greater
likelihood of medication use, visits to outpatient clinics,
OBESITY RESEARCH Vol. 10 No. 6 June 2002 529

Body Mass Index and Health Service Utilization, Reidpath et al.
greater use of health services. Greater use is nonetheless
a realistic possibility and needs to be monitored.
Table 4. Tabulation of ORs when considering BMI
(continuous) with health behaviors using NHS sub-
jects’ self-reported height and weight and controlling
for age and income
Unlike previous studies, the present study did not find
that obese or overweight women were any less likely to
have utilized preventive health services than were normal-
weight women. Instead, women in the underweight category
were significantly less likely to have regular screenings than
normal-weight women. Health insurance status is a poten-
tially plausible explanation for differences among the obese
participants in this study and previous studies, particularly
given that Australia had a successful universal healthcare
scheme whereas the United States, where the other signifi-
cant population-based study was conducted, did not (19).
The United States study specifically adjusted for health
insurance status; however, it left no immediately clear ex-
planation for the differences among obese women, creating
an important avenue of research inquiry.
One possible explanation for the finding that underweight
women were less likely to use preventive health services
than were normal-weight women relates to their health
status. It may be that they have serious medical conditions
that not only lead to them being underweight but that these
conditions take priority over routine preventive health
screening. This explanation is unfortunately not supported
by the medical-use data. If the illnesses of underweight
women interfered with the attendance for preventive screen-
ing, one would expect to see them over represented in the
medical service-use data, and this was not the case. An
alternative explanation is that underweight women are more
concerned with dieting and body image than other women
and, therefore, more reluctant to participant in procedures
that involve disrobing. This is, however, only speculative
and requires further investigation.
OR
CI
Men
Medication use*
Hospitalization
1.03 1.02 to 1.04
1.01 0.95 to 1.06
Accident and emergency room
visits
Outpatient clinic visits
Doctor visits*
1.03 0.98 to 1.08
1.04 1.01 to 1.07
1.02 1.00 to 1.03
1.01 0.99 to 1.02
Other health professional visits
Women
Medication use*
Hospitalization
1.04 1.03 to 1.05
1.03 0.99 to 1.07
Accident and emergency room
visits*
Outpatient clinic visits*
Doctor visits*
Other health professional visits*
Regular pap smear tests
Regular breast examinations
Regular mammograms
1.05 1.01 to 1.10
1.03 1.01 to 1.06
1.02 1.01 to 1.03
1.02 1.01 to 1.03
0.99 0.98 to 1.00
1.00 0.98 to 1.01
0.99 0.97 to 1.01
Individuals with an underweight BMI were excluded from this
analysis.
OR, odds ratio; BMI, body mass index; NHS, National Health
Survey; CI, confidence interval.
* p Ͻ 0.05 in logistic regression model.
There are three limitations to the findings discussed here.
First, and perhaps most importantly, the analyses are based
exclusively on Australian data. The degree to which the
results bear on other healthcare systems will depend 1) on
the general pattern of health-service use shown in the pop-
ulations in question and 2) on where the country lies in the
epidemiological transition. The results will, therefore, have
more relevance to those countries with similar health pro-
files to Australia and similar background health-service use
patterns. That our results have generally agreed with the
results from other international studies bolsters the obser-
vation that increases in BMI are associated with increasing
health-service use. However, it would be useful to have
more population based data on this relationship from other
countries. A second limitation relates to the study design.
The analyses are based on data from a cross-sectional sur-
vey, a fact that imposes limitations on inferences of causal
relationships. A third limitation relates to the fact that the
height and weight data are self-reported (2,23,25). The
“gold standard” is measured height and weight data, which
is, unfortunately, usually impractical for large population-
In the analysis of the data using a measure of contin-
uous BMI, the statistically significant ORs were small, in
the range of 1.02 to 1.05. Individuals more familiar with
the larger ORs typical of clinical research may be in-
clined to dismiss these results as statistically significant
but of little practical interest. It must be noted, however,
that these data relate to populations, and in an environ-
ment of increasing weight, the implications for healthcare
use are of real public-health significance. An increase of
one BMI unit in the population of men Ն20 years of age
represents, for example, a potential 4% increase in the
use of medication and a 5% increase in the use of
emergency medical services. Thus, even these small in-
creases in the odds of health services use may, in popu-
lations like Australia that are gaining in weight, have an
enormous impact on health-service delivery. This pessi-
mistic possibility needs to be tempered with the realiza-
tion that the relationship may not be a causal one, and
increasing population weight need not necessitate a
530 OBESITY RESEARCH Vol. 10 No. 6 June 2002

Body Mass Index and Health Service Utilization, Reidpath et al.
8. Seidell JC. The impact of obesity on health status: some
implications for health care costs. Int J Obes Relat Metab
Disord. 1995;19(Suppl 6):S13–6.
9. Colditz GA. Economic costs of obesity and inactivity. Med
Sci Sports Exerc. 1999;31:S663–7.
10. Wolf AM, Colditz GA. The cost of obesity: the US perspec-
tive. Pharmacoecon. 1994;5(Suppl 1):34–7.
11. Seidell JC, Deerenberg I. Obesity in Europe: prevalence and
consequences for use of medical care. Pharmacoecon. 1994;
5(Suppl 1):38–44.
12. Wolf AM, Colditz GA. Current estimates of the economic
cost of obesity in the United States. Obes Res. 1998;6:97–106.
13. Fontaine KR, Bartlett SJ. Access and use of medical care
among obese persons. Obes Res. 2000;8:403–6.
14. Zayat EN, Fontaine KR, Cheskin LJ. Use of preventive
health care services by patients with obesity. Obes Res. 1999;
7:223–6.
15. Sansone RA, Sansone LA, Wiederman MW. The relation-
ship between obesity and medical utilisation among women in
a primary care setting. Int J Eating Disord. 1998;23:161–167.
16. Quesenbury CP, Caan B, Jacobson A. Obesity, health
services use, and health care costs among members of a
health maintenance organization. Arch Intern Med. 1998;
158:466–72.
17. Brown WJ, Dobson AJ, Mishra G. What is healthy weight
for middle aged women? Int J Obes. 1998;22:520–8.
18. Trakas K, Lawrence K, Shear NH. Utilization of healthcare
resources by obese Canadians. Can Med Assoc J. 1999;160:
1457–63.
based surveys. It is known, however, that self-reported
weight tends to be underestimated with greater underesti-
mation by those who weigh more. There is also a tendency
(particularly among men) to overestimate height. The un-
derestimation of weight and overestimation of height will
result in the BMI being an underestimate (more so among
those who weigh more). Thus, the ORs reported here
are also likely to be underestimates, with the association
between BMI and use of health services even stronger.
The use of continuous data in some analyses was to
“smooth out” the bias over the distribution, rather than
containing it within discrete categories. Although this does
not overcome the bias, it provides an alternative way of
approaching the data.
Although economic analyses have shown that overweight
and obesity have a significant impact on health costs, there
has been very little data showing the actual impact of
overweight and obesity on health-service use. This research
shows that people who fall outside the healthy-weight range
are more likely to use a range of medical services. Given
that the BMI of industrialized populations appears to be
increasing, this has important ramifications for health-ser-
vice planning and reinforces the need for obesity prevention
strategies at a population level.
Acknowledgments
David Crawford was supported by an Australian, Na-
tional Health and Medical Research Council/National Heart
Foundation Career Development Award.
19. Fontaine KR, Faith MS, Allison DB, Cheskin LJ. Body
weight and health care among women in the general popula-
tion. Arch Fam Med. 1998;7:381–4.
20. Australian Bureau of Statistics. 1995 National Health Sur-
vey: Data Reference Package. Canberra, Australia: Australian
Bureau of Statistics; 1995.
21. Australian National Health and Medical Research Coun-
cil. Acting on Australia’s weight: a strategy for the prevention
of overweight and obesity. Canberra, Australia: Australian
Government Publishing Service; 1997.
22. Bostrom G, Diderichsen F. Socioeconomic differentials in
misclassification of height, weight and body mass index based
on questionnaire data. Int J Epidemiol. 1997;26:860–6.
23. Nieto-Garcia FJ, Bush TL, Keyl PM. Body mass definitions
of obesity: sensitivity and specificity using self-reported
weight and height. Epidemiology. 1990;1:146–52.
24. Niedhammer I, Bugel I, Bonenfant S, Goldberg M, Leclerc
A. Validity of self-reported weight and height in the French
GAZEL cohort. Int J Obes. 2000;24:1111–8.
References
1. World Health Organization. Obesity: Preventing and Man-
aging the Global Epidemic. Geneva, Switzerland: World
Health Organization; 1997.
2. Australian Bureau of Statistics. How Australians Measure
Up. Canberra, Australia: Australian Bureau of Statistics; 1995.
3. Birmingham CL, Muller JL, Palepu A, Spinelli JJ, Anis
AH. The cost of obesity in Canada. Can Med Assoc J. 1999;
160:483–8.
4. Must A, Spadano J, Coakley EH, Field AE, Colditz G,
Dietz WH. The disease burden associated with overweight
and obesity. JAMA. 1999;282:1523–9.
5. Seidell JC. Time trends in obesity: an epidemiological per-
spective. Horm Metab Res. 1997;29:155–8.
6. Colditz GA. Economic costs of obesity. Am J Clin Nutr.
1992;55(Suppl):S503–7.
25. Stewart AW, Jackson RT, Ford MA, Beaglehole, R. Un-
derestimation of relative weight by use of self-reported height
and weight. Am J Epidemiol. 1987;125:122–6.
7. Segal L, Carter R, Zimmet P. The cost of obesity: the
Australian perspective. Pharmacoecon. 1994;5(Suppl. 1):
45–52.
OBESITY RESEARCH Vol. 10 No. 6 June 2002 531