Prevalence of overweight and obesity in Australian children and
adolescents: reassessment of 1985 and 1995 data against new standard
Anthea M Magarey, Lynne A Daniels and T John C Boulton
MJA 2001; 174: 561-564
For editorial comment, see Baur; see also Eckersley
More articles on Paediatrics
Objective: To review the prevalence of overweight
and obesity in Australian children and adolescents in two national
samples, 10 years apart, using the new standard international
definitions of the International Obesity Task Force Childhood
Obesity Working Group.|
Design: Body mass index (BMI) cut-off points defining
overweight and obesity were applied to the individual BMI values in
the two cross-sectional samples.
Setting: Australian community.
Participants: 8492 schoolchildren aged 7-15 years
(Australian Health and Fitness Survey, 1985) and 2962 children aged
2-18 years (National Nutrition Survey, 1995).
Main outcome measure: Prevalence of overweight and
Results: In the 1985 sample, 9.3% of boys and 10.6% of girls
were overweight and a further 1.4% of boys and 1.2% of girls were obese.
In the 1995 sample, overall 15.0% of boys (varied with age from 10.4% to
20.0%) and 15.8% of girls (varied with age from 14.5% to 17.2%) were
overweight, and a further 4.5% of boys (2.4%-6.8%) and 5.3% of girls
(4.2%-6.3%) were obese. The prevalence of overweight and obesity in
the 1995 sample peaked at 12-15 years in boys and 7-11 years in girls. In
schoolchildren aged 7-15 years, the rates represent a relative risk
of overweight in 1995 compared with 1985 of 1.79 (95% CI, 1.59-2.00)
and of obesity of 3.28 (95% CI, 2.51-4.29). Compared with previous
estimates from these samples, the revised prevalence data are
slightly higher for the 1985 data and considerably higher for the 1995
Conclusion: The secular trend of increasing overweight
and obesity in the decade from 1985 and the high prevalence rates in
Australian children and adolescents are a major public health
The rapidly increasing prevalence of obesity in young people is a
major global public health concern,1 not only because of the
health consequences in childhood and adolescence,1-4 but also
because of the greater risk of obesity in adulthood.3 To develop and
evaluate appropriate public health policies, the trends in
childhood obesity need to be monitored.
While, in adults, the body mass index (BMI) cut-offs internationally
accepted as definitions of overweight and obesity (25
and 30 kg/m2, respectively) are based on
increased risks of morbidity and mortality,5 no such outcome-based
definition exists for children. Definitions for children and
adolescents have generally been statistical, with percentile
cut-offs based on variably representative and sized national
anthropometric data sets, making it difficult to adequately monitor
secular changes in body size and prevalence of obesity. Furthermore,
there has been no agreement on either the most appropriate adiposity
index or the cut-off points. Thus, it has not been possible to
interpret and compare prevalence and intervention studies,
nationally or internationally.
In 1997, the International Obesity Task Force (IOTF) Childhood
Obesity Working Group accepted BMI as a consistent, pragmatic (ie,
reasonable although imperfect) index of adiposity in children and
adolescents.6 The IOTF has recently
published cut-off points to define overweight and obesity in
children and adolescents based on adult overweight and obesity BMI
ranges.7 While this remains a
statistical definition and can not be quantified in terms of current
or future morbidity, it allows analysis of national data to determine
point prevalence and identify secular trends, and enables
meaningful international comparisons. For example, a secular trend
in body fatness, estimated by BMI, from 1985 to 1997 was found in
children aged 7-12 years in New South Wales and Victoria, but the lack
of accepted standards meant that no estimates were made of the
prevalence of overweight and obesity.8,9
Using these new international definitions, we determined the
prevalence of overweight and obesity in Australian children and
adolescents aged 2-18 years in two nationally representative
samples 10 years apart — the Australian Health and Fitness Survey of
1985 (AHFS85)10 and the National
Nutrition Survey of 1995 (NNS95).11 In previous estimates of
the prevalence of overweight using these samples, any secular trend
could not be identified because different definitions were
The data collection procedures for the two surveys are described in
Box 1. The original data published from the AHFS85 survey included
percentiles for height and weight only. Subsequently published were
BMI percentiles by age and sex, and the prevalence of overweight in the
age group 12-14 years (but not in those 7-11 years of age).12 Summary
statistics of prevalence of overweight and obesity in NNS95
participants have also been published.13Box 2 shows the prevalence
of overweight in these two surveys.
Data for analysis of the AHFS85 sample were supplied on disc by the
Australian Council on Health, Physical Education and Recreation and
those for the NNS95 were provided by the Australian Bureau of
Statistics as a confidentialised unit record file.
New international standard definitions
The new international standard definitions for overweight and
obesity published by the International Obesity Task Force were based
on six large, nationally representative cross-sectional growth
studies from birth to 25 years (Brazil 1989, Great Britain 1979-1993,
Hong Kong 1993, the Netherlands 1980, Singapore 1993, and the United
States 1963-1980).7 For each of these, centile
curves of BMI were drawn that at age 18 years passed through the BMI
cut-off points of 25 kg/m2 and 30 kg/m2 for adult
overweight and obesity, respectively. The resulting curves were
averaged to provide age- and sex-specific BMI cut-off points to
define overweight (cut-off 1) and obesity (cut-off 2) in half-year
intervals from 2 to 18 years.7
Re-estimate of overweight and obesity prevalence
We applied the new international cut-off points to each child in the
two data sets to re-estimate the prevalence of overweight and
obesity. For each age, the cut-off point at the mid-year value was
applied (eg, for those aged 7 years, we used the cut-off at 7.5 years).
For data with one-year age groups, as in these two samples, this will
give an essentially unbiased estimate.7 We grouped children in the
AHFS85 into two age bands and those in the NNS95 into five age bands to
facilitate comparison between the two studies.
Analyses were performed using the Stata statistical
(RR) of overweight (≥ cut-off 1) or obesity (≥ cut-off 2)
within years (controlling for sex) and between years (controlling
for age and sex) was determined using "Epitab - tables for
epidemiologists"14 and reported as RR (95%
For each age group, the proportion of boys and girls classified,
according to the new definitions, as overweight (BMI ≥ cut-off 1
and < cut-off 2) and obese (BMI ≥ cut-off 2) in the AHFS85 and
NNS95 samples are listed in Box 3.
In the AHFS85 sample:
In the age group 7-15
years, 9.3% of boys and 10.6% of girls were overweight and a
further 1.4% of boys and 1.2% of girls were obese.
In the two age groups 7-11 years and 12-15 years, the relative
risks of overweight or obesity did not differ between boys and girls
and between age groups.
In the NNS95 sample:
In the age group 2-18 years,
15.0% of boys and 15.8% of girls were overweight and a further 4.5% of
boys and 5.3% of girls were obese.
In the age group 7-15 years, 15.3% of boys and 16.0% of girls
were overweight and a further 4.7% of boys and 5.5% of girls were obese.
Overall, there was no difference between boys and girls in
relative risk of overweight (RR, 1.08; 95% CI, 0.94-1.24).
In the age group 7-11 years, the relative risk of overweight
was significantly greater for girls than boys (RR, 1.53; 95%
CI, 1.16-2.02; P = 0.002).
Overall or within any age group, there was no difference in
relative risk of obesity between boys and girls.
Data from NNS95 v data from AHFS85:
The relative risk of overweight was 1.79 (95% CI, 1.59-2.00; P
< 0.001), independent of sex and after controlling for age.
The relative risk of obesity was 3.28 (95% CI, 2.51-4.29; P
< 0.001) and independent of sex and age.
We used the new international cut-off points to estimate the
prevalence of overweight and obesity in two nationally
representative samples of Australian children from 1985 and 1995,
respectively.7 Depending on age, 13%-26%
and 19%-23% of Australian boys and girls, respectively, aged 2-18
years are overweight or obese, with prevalence peaking at 12-15 years
in boys and 7-11 years in girls. The prevalence of obesity is 2%-7% in
boys and 4%-6% in girls.
Previous estimates of the prevalence of overweight and obesity in the
AHFS85 used the 85th percentile of BMI for the age group 12-14 years in
the United States National Health and Nutrition Examination Survey
II (NHANES II; 1976-1980) to define overweight (Box 1).12 In the NNS95
sample, children aged 2-8 years were classified as high weight for
height if their standard deviation (Z) score of weight for
height was greater than + 2, based on World Health Organization
reference values.11 For older subjects, the
85th and 95th BMI percentiles derived from NHANES I data of
1971-1974,15 which are lower than those
from the later NHANES II data, were used to define "at risk of
overweight" and overweight, respectively. The reason for the use of
the older NHANES data for the later 1995 survey is not clear. These
differences in reference populations and approaches to classifying
participants as overweight or obese clearly limit comparison of
prevalence rates and assessment of trends.
Our revised estimates of prevalence of overweight in the 1985 and 1995
samples, respectively, are slightly higher and considerably higher
than those previously reported.12,13 The risks of both
overweight and obesity in 1995 are significantly higher than in 1985
and the increase in prevalence is higher than previously reported.
Methodological differences between the two surveys must be taken
into consideration when making these comparisons. Although the
effect of sampling on the trend is unknown, the different method of
determining age in the 1985 sample may result in a very few subjects
being misclassified as acceptable weight and slightly fewer being
misclassified as overweight. The possible net underestimate of
prevalence of overweight would be extremely small and could not
account for the large increase in prevalence of overweight from 1985
Furthermore, our prevalence rates are high by international
standards. The revised prevalence figures for 1985 for both
overweight and obesity fall at about the mid-point of the range of the
international reference populations at age 18 years.7 However,
prevalence figures from the 1995 sample indicate rates for both
overweight and obesity that are higher than all the international
reference populations and considerably higher than those of 1985.
Comparison with recently published trends from 1974 to 1994 in
British children aged 4-11 years16 indicates that more
Australian boys aged 7-11 years were overweight in 1985 and 1995 than
British boys, although the decade increase was similar. For girls,
prevalence of overweight in Australia in 1985 was slightly higher
than in Britain, but considerably higher in 1995 (ie, the decade
increase was greater in Australian girls). Similarly, prevalence of
overweight in 1994 in Australian children aged 4-6 years was
considerably higher than in British children (boys, 13.4% v 6.4%;
girls, 19.5% v 9.2%).
The new standard definitions for childhood overweight and obesity
are not directly related to measurable outcomes in morbidity and
mortality in adulthood, although dyslipidaemia, elevated blood
pressure and insulin resistance are associated with these levels of
adiposity in children.2-4 Overweight and obesity in
childhood and adolescence increase the risk of overweight in
adulthood, but these new BMI centile curves give no information on the
degree of tracking within the different levels of adiposity (ie, the
extent to which individuals maintain their level of adiposity).
Longitudinal studies are urgently needed to assess the change in BMI
percentiles during growth, and the long term health outcomes
associated with these defined levels of adiposity, so that
overweight and obesity in childhood can be defined by biological
The secular trend towards increasing overweight and obesity in the
decade from the mid-1980s and the high prevalence figures described
here are of major public health concern and should stimulate urgent
action: treatment of those already overweight and obese, prevention
of progression of overweight to obesity, and prevention of
overweight.18 The International
Obesity Task Force proposes three levels at which prevention should
- Universal/public health prevention directed
at the whole population;
- Selective prevention directed at subgroups at increased risk of
developing obesity; and
- Targeted prevention directed at high risk individuals who are
overweight but not yet obese.1
The National Health and Medical Research Council (NHMRC) report
Acting on Australia's weight places particular emphasis on
strategies, at the population level, to change the macroenvironment
of food supply and opportunities for physical activity.18 The global
epidemic of increasing overweight and obesity in both adults and
children is attributed to changing lifestyle.1 Partly because
of methodological difficulties in assessment,19 no data exist
in Australia on either the activity or inactivity levels of children
and adolescents in the past 15 years, nor how these may have changed.
Several factors, however, suggest that the increased prevalence of
overweight and obesity in Australian children may be attributed to
decreasing activity, increasing inactivity and increasing food
- a study of fitness levels in Australian
children from 1985 to 1997 reported that these have declined,
suggesting a decrease in physical activity;20
- it is generally recognised that the time spent in sedentary
activities, such as television watching and playing computer games,
has increased in the past decade;
- a comparison of the dietary intakes of children aged 10-15 years from
the national surveys of 198521 and 199513 showed that
mean daily energy intake has increased by 0.5-2.0 MJ in the 10-year
- a comparison of the energy intake of children aged 4-8 years in 1995
with that of Adelaide children in the early 1980s suggested that
energy intake in this age group has also increased by 0.5-1.5
Health professionals have a major role to play in promoting, to all
family members, the benefits of a healthy diet, an increase in
habitual physical activity and a decrease in inactivity.23-25 Such
strategies have major implications for maintenance of a healthy
weight, and the prevention of overweight, obesity, and future
cardiovascular disease risk.
We wish to thank Adrian Esterman, biostatistician, Paul Hakendorf,
clinical epidemiologist, and Sonia Anderton, statistical
consultant, Flinders University, for their advice on analysis.
(Received 22 Sep 2000, accepted 15 Feb 2001)
- World Health Organization (WHO). Obesity: preventing and
managing the global epidemic. Report of a WHO consultation on
obesity, Geneva, 3-5 June, 1997. Geneva: WHO, 1998.
Dietz WH. Health consequences of obesity in youth: childhood
predictors of adult disease. Pediatrics 1998; 101 (3 Suppl
Power C, Lake JK, Cole TJ. Measurements of long-term health risks of
child and adolescent fatness [review]. Int J Obes Relat Metab
Disord 1997; 21: 507-526.
Guo SS, Roche AF, Chumlea WC, et al. The predicitve value of
childhood body mass index values for overweight at age 35 y. Am J
Clin Nutr 1994; 59: 810-819.
Manson JE, Willett WC, Stampfer MJ. Body weight and mortality among
women. N Engl J Med 1995; 333: 677-685.
Bellizzi MC, Dietz WH. Workshop on childhood obesity: summary of
the discussion [review]. Am J Clin Nutr 1999; 70: 173S-175S.
Cole TJ, Bellizzi MC, Flegal KM, Dietz WH. Establishing a standard
definition for child overweight and obesity worldwide:
international survey. BMJ 2000; 320: 1240-1243.
Lynch J, Wang XL, Wilcken DEL. Body mass index in Australian
children: recent changes and relevance of ethnicity. Arch Dis
Child 2000; 82: 16-20.
Lazarus R, Wake M, Hesketh K, Waters E. Change in body mass index in
Australian primary school children, 1985-1997. Int J Obes Relat
Metab Disord 2000; 24: 679-684.
Pyke JE. Austalian Health and Fitness Survey 1985. Parkside, SA:
The Australian Council for Health, Physical Education and
Recreation Inc, 1987.
McLennan W, Podger A. National nutrition survey user's guide
1995. Canberra: AGPS, 1998.
Harvey PWJ, Althaus M-M. The distribution of body mass index in
Australian children aged 7-15 years. Aust J Nutr Diet 1993;
McLennan W, Podger A. National nutrition survey, 1995. Nutrient
intakes and physical measurements. Canberra: AGPS, 1998.
Stata statistical software, release 7.0. College Station,
Texas: Stata Corporation, 2001.
Must A, Dallal GE. Reference data for obesity: 85th and 95th
percentiles of body mass index (wt/ht2) and triceps skinfold
thickness. Am J Clin Nutr 1991; 53: 839-846.
Chinn S, Rona RJ. Prevalence and trends in overweight and obesity
in three cross sectional studies of British children, 1974-94.
BMJ 2001; 322: 24-26.
Dwyer T, Blizzard CI. Defining obesity in children by biological
endpoint rather than population distribution. Int J Obes Relat
Metab Disord 1996; 20: 472-480.
National Health and Medical Research Council. Acting on
Australia's weight: a strategic plan for the prevention of
overweight and obesity. Canberra: NHMRC, 1997.
Goran MI. Measurement issues related to studies of childhood
obesity: assessment of body composition, body fat distribution,
physical activity and food intake. Pediatrics 1998; 101 (3
Suppl March): 505-518.
Dollman J, Olds T, Norton K, Stuart D. The evolution of fitness and
fatness in 10-11 year old Australian children: changes in
distributional characteristics between 1985 and 1997. Pediatr
Exer Sci 1999; 11: 108-121.
Department of Community Services and Health. National dietary
survey of schoolchildren (aged 10-15 years), 1985: no 2 — nutrient
intakes. Canberra: AGPS, 1989.
Magarey A, Boulton TJC. Food intake during childhood:
percentiles of food energy, macronutrient and selected
micronutrients from infancy to eight years of age. Med J Aust
1987; 147: 124-127.
National Health and Medical Research Council. Dietary
Guidelines for Australians; Canberra: NHMRC, 1992.
National Health and Medical Research Council. Dietary
guidelines for children and adolescents. Canberra: NHMRC, 1995.
Smith A, Kellett E, Schmerlaib Y. The Australian guide to healthy
eating. Canberra: Commonwealth Department of Health and Family
Department of Public Health, Flinders University of South
Australia, Adelaide, SA.
Anthea M Magarey, PhD, DipNutDiet, National Health and
Medical Research Council (NHMRC) Research Fellow;
Daniels, PhD, DipNutDiet, Associate Professor.
Women's and Children's Health, The Nepean Hospital, Penrith; and
Western Clinical School, University of Sydney, NSW.
T John C Boulton, MD, FRACP, Clinical Professor of
Reprints will not be available from the authors.
Correspondence: Dr A
M Magarey, Nutrition Unit, Department of Public Health, Flinders
University of South Australia, Bedford Park, SA 5034.
Readers may print a single copy for personal use. No further
reproduction or distribution of the articles
should proceed without the permission of the publisher. For
permission, contact the
Australasian Medical Publishing Company.
Journalists are welcome to write news stories based on what they read here, but should acknowledge their source as "an article published on the Internet by The Medical Journal of Australia <http://www.mja.com.au>".
© 2001 Medical Journal of Australia.
1: Data collection procedures for
the 1985 Australian Health and Fitness Survey (AHFS85) and the 1995 National
Nutrition Survey (NNS95)
Australian Health and Fitness
Survey 1985 (n = 8492)
The AHFS85 was conducted by the Australian Council on Health, Physical
Education and Recreation from May to October 1985 inclusive.10
A representative sample of Australian schoolchildren aged 7-15 years was
selected using a two-stage probability sampling design. Firstly, 109 schools
were selected randomly from a list of all primary and secondary schools
with 10 or more students enrolled in each age and sex category. The probability
of selecting each school was proportional to the number of children enrolled
at that school. All States and Territories were included and a good geographical
distribution was achieved. Secondly, within selected schools, 15 students
in each age and sex category were sampled from enrolment lists using a
systematic selection process. The age of participants was determined in
whole years as at 30 June 1985. Measurements were made by trained observers
using standardised procedures.
National Nutrition Survey 1995
(n = 3007)
The NNS95 was a joint project of the Australian Bureau of Statistics
and the Commonwealth Department of Health and data were collected from
February 1995 to March 1996.11 The NNS sample
was a systematically selected subsample of private dwellings of the National
Health Survey (NHS), conducted on a multistage area sample. Up to two
eligible people per household (in urban areas) and three per household
(in rural areas) were randomly selected from the subsample. Weighting
factors were applied to adjust for non-respondents. All States and Territories
and urban, rural and remote areas were represented. Age was determined
as a whole year at the most recent birthday. Measurements were made by
trained observers using standardised protocols.
|Back to text|
|Back to text|
|Back to text|