SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS AND PROCEDURES
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGMENTS
  8. DISCLOSURE
  9. REFERENCES

Several country-specific and global projections of the future obesity prevalence have been conducted. However, these projections are obtained by extrapolating past prevalence of obesity or distributions of body weight. More accurate would be to base estimates on the most recent measures of weight change. Using measures of overweight and obesity incidence from a national, longitudinal study, we estimated the future obesity prevalence in Australian adults. Participants were adults aged ≥25 years in 2000 participating in the Australian Diabetes, Obesity, and Lifestyle (AusDiab) study (baseline 2000, follow-up 2005). In this population, approximately one-fifth of those with normal weight or overweight progressed to a higher weight category within 5 years. Between 2000 and 2025, the adult prevalence of normal weight was estimated to decrease from 40.6 to 28.1% and the prevalence of obesity to increase from 20.5 to 33.9%. By the time, those people aged 25–29 in 2000 reach 60–64 years, 22.1% will be normal weight, and 42.4% will be obese. On average, normal-weight females aged 25–29 years in 2000 will live another 56.2 years: 26.6 years with normal weight, 15.6 years with overweight, and 14.0 years with obesity. Normal-weight males aged 25–29 years in 2000 will live another 51.5 years: 21.6 years with normal weight, 21.1 years with overweight, and 8.8 years with obesity. If the rates of weight gain observed in the first 5 years of this decade are maintained, our findings suggest that normal-weight adults will constitute less than a third of the population by 2025, and the obesity prevalence will have increased by 65%.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS AND PROCEDURES
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGMENTS
  8. DISCLOSURE
  9. REFERENCES

The prevalence of obesity among Australian adults is progressively increasing, doubling between 1980 and 2000 (1,2). Similar increases are evident worldwide (3), primarily driven by an increase in incidence. A recent US study reported a two- to threefold increase in incidence of overweight and obesity between 1950 and 2000 (4). Such increases are problematic considering the association between obesity and chronic disease, and the attendant burden on the health-care system. Understanding the impact of current trends on the future prevalence of obesity is essential for determining both the nature and size of public health interventions required to reduce or maintain current levels of obesity.

Several country-specific and global projections of the future prevalence of obesity have been conducted (5,6,7,8,9,10,11,12). For example, projections for England suggest that by 2025, 47% of males and 36% of females will be obese, an increase of 96 and 50%, respectively over 2004 levels (9). However, these studies predict the future burden of obesity by extrapolating past prevalence of obesity or distributions of body weight. This can be problematic, considering that the prevalence of obesity at any point in time is a result of decades of differing rates of moving between normal weight and overweight, and overweight and obese. These rates reflect historic environments and lifestyles, and differ from those of more recent times. The recent reported slowing in the rate of increase in BMI in Australia (2) and elsewhere (13,14,15,16,17) presents a serious potential limitation to methods of projection based on past prevalence. A more accurate method would be to base future prevalence estimates on the most recent measures of weight change.

Using measures of overweight and obesity incidence from the Australian Diabetes, Obesity, and Lifestyle (AusDiab) study, we aimed to estimate both the future burden of obesity in Australia and the lifetime risk of overweight and obesity for a cohort of people aged 25–29 years in 2000.

METHODS AND PROCEDURES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS AND PROCEDURES
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGMENTS
  8. DISCLOSURE
  9. REFERENCES

Study population

AusDiab is a national, population-based survey of 11,247 Australian adults aged ≥25 years at baseline (1999–2000). From the 17,129 eligible households, 20,347 adults completed a household interview, and 11,247 (55%) had a biomedical examination, giving an overall response rate of 37% (18). In 2004–2005, all participants (n = 11,247) were invited to a follow-up examination. Those who refused further contact (n =128), were deceased (n = 310), had moved overseas or into a high-care nursing facility, or had a terminal illness (n = 21) were considered ineligible. Of the 10,788 participants eligible for follow-up, 6,400 (59%) presented for the biomedical examination and/or blood tests (19). The full study methodology has been reported previously (18,19). For the current analysis, participants missing values for height or weight were excluded, leaving 11,067 participants at baseline and 6,350 at follow-up.

The study was approved by ethics committees of the International Diabetes Institute and Monash University. Approval for matching the cohort to the National Death Index was granted by the Australian Institute of Health and Welfare. All participants provided informed consent.

Measurements

At baseline and follow-up, questionnaires, anthropometric measurements, and a fasting blood sample were collected. Vital status was assessed annually by linking the AusDiab (n = 11,247) cohort to the Australian National Death Index. The National Death Index has 93.7% sensitivity and 100% specificity for the identification of deaths (20). People unable to be matched to the National Death Index were assumed to be alive. This study included all deaths until 1 June 2005 occurring among participants with recorded height and weight measurements and whose vital status was determined at follow-up. BMI was categorized as normal weight (<24.9 kg/m2), overweight (25.0–29.9 kg/m2), and obese (≥30.0 kg/m2) (3,21). Due to small numbers of underweight participants, underweight was combined with normal weight in this analysis.

Transition probabilities

Transition probabilities between states were derived for the total population, for males and females, and by 5-year age groups. Incidence was derived from participants with baseline height and weight measurements who either had these measurements repeated at follow-up (n = 6,285) or were recorded as having died within the 5-year time period (n = 341). Small numbers of people moved from normal weight to obese (one male, seven females) and obese to normal weight (one male, five females) over the 5-year period, therefore only transitions to adjacent states were modeled (Figure 1). Transitions from obese to overweight and overweight to normal weight were also allowed. The transition probabilities include: (i) normal weight to normal weight; (ii) normal weight to overweight; (iii) normal weight to dead; (iv) overweight to normal weight; (v) overweight to overweight; (vi) overweight to obese; (vii) overweight to dead; (viii) obese to overweight; (ix) obese to obese; (x) obese to dead. People missing data at follow-up (2005) but not recorded as dead were assumed to have the same probability of transitioning between states as those not missing at follow-up. Age-specific mortality probabilities for the Australian population (22) were apportioned across the three BMI categories to reflect the distribution of mortality by BMI category in AusDiab. Due to small numbers in the 85+ years age group, the mortality rate used was taken from national mortality statistics, and assumed not to differ between BMI categories. The Lowess smoother was applied to all transition probabilities between ages 25 and 85 years to reduce the amount of “noise” in transition probabilities (23). Influential outliers were removed. A sensitivity analysis was conducted to assess the impact of the smoothing of probabilities and removal of influential outliers.

image

Figure 1. Pictorial model of the age-specific transitions between body weight states.

Download figure to PowerPoint

Life-table models

Abridged 5-year multistate life tables were constructed with three alive states (normal weight, overweight, and obese), based on the transition probabilities outlined above. In the construction of multistate life tables the transition probabilities followed the Markov assumption that the probability of moving between states depends only on the current state and is independent of all preceding states. An exponential probability of mortality was assumed within each 5-year age group. Starting prevalences for the 25–29-year age group were weighted to reflect the Australian population, based on AusDiab.

Stationary life tables

Two sets of stationary multistate life tables were created, each simulating the 5-yearly progression of a cohort of 25–29-year-olds in the year 2000 exposed to the incidence and mortality observed between 2000 and 2005 throughout their future lifetime, and followed until death. The first simulates a cohort of 25–29-year-olds with the 2000 prevalence of normal weight, overweight, and obesity, whereas the second simulates a cohort of normal weight 25–29-year-olds.

Dynamic life tables

The dynamic multistate life tables were created to analyze the future population impact of the current transition rates between the years 2000 and 2025. The dynamic models are a series of linked multistate life tables, one for each 5 years between 2000 and 2025. In these life tables, the distribution of people between the different states at a given age (a) and time (t) depends on the distribution at age (a-5) and time (t-5) and on the transition probabilities between states between (t-5) and (t). The baseline multistate life-table population, representing the age structure in Australia in the year 2000 was created by applying the age-specific population numbers for the year 2000, and the age-specific prevalence of normal weight, overweight, and obesity observed in AusDiab, to the stationary life-table structure described above. The mortality rate was assumed first, to remain constant and second, to decrease by 2.2% per year, reflecting the average decrease in all-cause mortality rates in Australia between 2000 and 2004 (22). Each 5 years a new cohort of 25–29-year-olds entered the model, based on population projections between 2008 and 2025 from the Australian Bureau of Statistics (19), and assuming the prevalence of normal weight, overweight, and obesity at age 25–29 years changed to the same extent as that of those aged 30–34 years. From these dynamic models, age-specific and total prevalence of normal weight, overweight, and obesity were derived for each 5-year period between 2000 and 2025.

Analyses were performed for the total population, and for men and women, using sex-specific transition probabilities. Life tables were constructed in Microsoft Excel 2007.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS AND PROCEDURES
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGMENTS
  8. DISCLOSURE
  9. REFERENCES

Of the 11,067 Australian participants at baseline (2000) with height and weight measurements, 6,285 returned for follow-up (2005) (Table 1). There was no difference between mean age at baseline and follow-up. The unweighted baseline prevalence of normal weight, overweight, and obesity was 37.6, 40.1, and 22.3%, respectively. Over the 5-year period, 341 deaths were recorded—137 of these were in those with normal weight, 140 in those with overweight, and 68 in those with obesity.

Table 1.  Study population characteristics at baseline (1999/2000) and 5-year follow-up (2005)
inline image

The body weight trajectory of a cohort of normal weight 25–29-year-olds in the year 2000 was analyzed (Table 2). When this cohort reaches age 60–64 years, an estimated 27.5% will be normal weight, 35.6% will be overweight, 29.3% will be obese, and 7.6% will be dead. Of those alive at age 60–64 years, an estimated 29.7% will be normal weight, 38.6% will be overweight, and 31.7% will be obese.

Table 2.  Projected age-specific prevalence of normal weight, overweight, and obesity in a cohort of normal weight 25–29-year-olds using current incidence and mortality rates held constant for the life of the cohort
inline image

By the time, the cohort of 25–29-year-olds in 2000, with the 2000 prevalence of normal weight, overweight, and obesity, reach age 60–64 years, 20.2% will be normal weight, 32.5% overweight, 38.7% obese, and 8.6% dead. Of those alive at age 60–64 years, an estimated 22.1% will be normal weight, 35.5% will be overweight, and 42.4% will be obese (Table 3).

Table 3.  Projected age-specific prevalence of normal weight, overweight, and obesity in a cohort of 25–29-year-olds with initial prevalence of normal weight, overweight, and obesity as observed in the year 2000, and using current incidence and mortality rates held constant for the life of the cohort
inline image

By the time, the cohort of people aged 25–29 years in 2000 with the 2000 prevalence of normal weight, overweight, and obese reach age 60–64 years, at some time in their lives 59.6% of females and 51.8% of males will have been obese (Table 4). Over their entire lifetime, an estimated 66.3% of females and 59.4% of males will have been obese.

Table 4.  Lifetime probability of ever being obese, in a cohort of 25–29-year-olds with initial prevalence of normal weight, overweight, and obesity, as observed in the year 2000, and using current incidence and mortality rates held constant for the life of the cohort
inline image

We examined the average number of years and the percentage of their remaining lifespan that females and males will live in the normal weight, overweight, and obese states. On average, females with the 2000 distribution of BMI categories aged 25–29 years will live another 56.0 years; 19.2 years normal weight, 15.6 years overweight, and 21.2 years obese. Males aged 25–29 years will live another 51.0 years on average; 13.2 years normal weight, 21.4 years overweight, and 16.4 years obese. As a proportion of their remaining lifespan, females aged 25–29 years will live on average 34.3% normal weight, 27.8% overweight, and 37.9% obese, whereas males aged 25–29 years will live on average 25.8% normal weight, 42.0% overweight, and 32.2% obese (Figure 2).

image

Figure 2. Percentage of remaining years of life lived with normal weight, overweight, and obesity for people aged 25–29 years with initial prevalence of normal weight, overweight, and obesity as observed in the year 2000.

Download figure to PowerPoint

Between 2000 and 2025, the prevalence of normal weight in Australian adults is predicted to decrease from 40.6% (48.3% of females and 32.8% of males) to 28.1% (32.6% of females and 26.4% of males). The prevalence of obesity is predicted to increase from 20.5% (22.0% of females and 19.1% of males) to 33.9% (36.4% of females and 35.3% of males) (Figure 3). For the age groups 25–44 years, 45–64 years, and 65+ years, the prevalence of normal weight will decrease between 2000 and 2025 from 48.6, 31.8, and 36.7% to 30.2, 25.9, and 27.7% respectively; whereas the prevalence of obesity in these age groups will increase from 16.7, 26.2, and 19.6% to 28.9, 37.3, and 36.9%, respectively (Table 5). Adjusting predictions for the historic 2.2% (approximately) annual decline in mortality rates had little effect on the results.

image

Figure 3. Projected prevalence of normal weight, overweight, and obesity for the total population from 2000 to 2025 using current incidence and mortality rates held constant for the life of the cohort.

Download figure to PowerPoint

Table 5.  Projected population prevalence of normal weight, overweight, and obesity from 2000 to 2025 using current incidence and mortality rates held constant for the life of the cohort
inline image

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS AND PROCEDURES
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGMENTS
  8. DISCLOSURE
  9. REFERENCES

Of the normal weight 25–29-year-olds in the year 2000, by age 60–64 years an estimated 29.3% will be obese. As a proportion of their remaining lifespan, females and males aged 25–29 years in 2000 will live 37.9 and 32.2% with obesity, respectively. Overall, between 2000 and 2025, the adult population prevalence of normal weight will decrease from 40.6 to 28.1% and the prevalence of obesity will increase from 20.5 to 33.9%.

By the time the 25–29-year-olds of 2000 reach age 60–64 years, over a third will be obese. At some time in their lives, over half will have had a BMI ≥30 kg/m2. The difference between these estimates is due to the ∼10% probability at each age of those with obesity becoming overweight.

Our projections are somewhat lower than many of those previously published (5,9,10,12), however, somewhat higher (at 72% in 2020) than the 65% prevalence of overweight predicted for Australia in 2022 based on a projection of past prevalence (24). Projections for England—currently with a similar prevalence of overweight and obesity to Australia—suggest that by 2025, 47% of males and 36% of females will be obese (9). Stewart et al. (2009) reported that by 2020 ∼45% of Americans will be obese. While the starting prevalence of obesity is higher for the United States, the difference in projections between ours and other studies may also be due to the methods used. Foresight, for example, obtained their projections for the United Kingdom by implementing nonlinear regression analysis of historical cross-sectional data of the population BMI distribution, whereas, we project the burden of obesity based on national transition probabilities observed between 2000 and 2005. This is the first time such a method has been used to obtain long-term obesity projections. Considering that the historical trend has been toward a higher incidence of overweight and obesity and a greater prevalence of obesity, our prediction may be an underestimation of future levels of overweight and obesity. However, recent evidence suggests that the rate of increase in BMI has slowed in recent years both in Australia (2) and other countries (13,14,15,16,17), presenting a serious limitation to methods based on past prevalence. If incidence rates had, indeed, slowed between 2000 and 2005 relative to earlier years, this would directly drive differences between our results, which use recent incidence data, and other results, which are influenced by historical rates of rise in the BMI distribution. We recommend incidence-based models such as this for forecasting the likely development of the obesity epidemic in the short to medium term.

Recent Australian data (2007/2008) on the prevalence of obesity suggests that since 2000 there has been a continued progression toward an increasingly greater prevalence of obesity (25). Our projected prevalence estimates of obesity for 2005 and 2010 of 24.2 and 26.9%, respectively, are similar to the observed prevalence for 2007/2008 of 24.8%.

Important limitations to our study are that BMI transition and mortality probabilities were derived from different sources. In addition, the study is subject to the potential selection bias of the AusDiab study—due to exclusion of individuals not living in private residential homes (e.g., prisoners, hospitalized patients), potential healthy volunteer bias, and the 37 and 61% response rates to the first and second surveys. Loss to follow-up may mean that incidence rates are not wholly reflective of population incidence rates (19). However in this case, an underestimation of BMI levels is most likely, as people absent at follow-up had lower levels of education (19), previously associated with higher body weight in developed countries (26). Moreover, when using multiple imputations to estimate follow-up BMI for all individuals present at baseline but not at follow-up, the prevalence of normal weight, overweight, and obesity were similar to the original data set. The starting or “input” prevalences of normal weight, overweight, and obesity for 25–29 years olds, to which age-specific incidence transition probabilities are later applied, has considerable influence over the projected prevalence results. An increase in prevalence of obesity in this group is likely, considering the upwards transitioning of all other age groups, but the age-specific transition probabilities for this first age group are unknown. Thus, for the dynamic life tables we chose to increase these starting prevalences in the years 2010–2025 to the same extent as the increase in the age group above (30–34 years). It should be acknowledged that the categories of BMI used in this analysis are a crude measure of excess body weight. Using this particular method, analysis of the various classes of obesity was not appropriate due to small numbers. Additionally, the population is not homogeneous and it is possible that there are nonsusceptible population subgroups; subgroups protected from weight gain and thus not following the projected population body weight trajectory.

As discussed, the key advantage of this study is that the future prevalence estimates are based on recent measures of incidence. However, an implication of this assumption is that no group is protected from weight gain. This may in fact not be the case, and is something that will become clearer over time. Other strengths of the study include accurately measured height and weight informing the BMI calculations, and the high quality of the mortality follow-up data.

If nothing occurs to affect 2000–2005 transition rates over time, these projections of the obesity burden will be realized over coming decades. However, as the trend over recent decades has been toward increasingly obesogenic environments and a higher prevalence of obesity (26), our projections could easily be surpassed. In contrast, if obesity prevention strategies are able to reduce the current rates of weight gain, the rate of increase in the prevalence of obesity will also decrease (27).

These findings suggest that new effective interventions to reduce current levels of weight gain are essential if we are to avoid the predicted burden of obesity. It is likely that such effective interventions will involve high-level policy and legislative changes to alter the obesogenic environments in which we live and render healthy eating and appropriate levels of physical activity an easy lifestyle to follow (26,28).

In conclusion, our findings suggest a 70% increase in the prevalence of obesity amongst Australian adults over coming decades; and a high probability of future obesity among contemporary normal-weight young adults. We recommend continued development of projection models based on recent measures of population weight gain to form the basis for the prioritization, evaluation, and choice of obesity prevention strategies.

ACKNOWLEDGMENTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS AND PROCEDURES
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGMENTS
  8. DISCLOSURE
  9. REFERENCES

H.L.W. was supported by a National Health and Medical Research Council and National Heart Foundation scholarship, and a National Health and Medical Research Council grant (no. 465130). C.E.S. and H.R.M. were supported by a National Health and Medical Research Council grant (no. 465130). A.P. was supported by a VicHealth fellowship. We thank Alison Beauchamp for editing the final manuscript. The AusDiab study co-coordinated by the Baker IDI Heart and Diabetes Institute, gratefully acknowledges the generous support given by: National Health and Medical Research Council (NHMRC grant 233200), Australian Government Department of Health and Ageing. Abbott Australasia Pty Ltd, Alphapharm Pty Ltd, AstraZeneca, Bristol-Myers Squibb, City Health Centre-Diabetes Service-Canberra, Department of Health and Community Services—Northern Territory, Department of Health and Human Services—Tasmania, Department of Health—New South Wales, Department of Health—Western Australia, Department of Health—South Australia, Department of Human Services—Victoria, Diabetes Australia, Diabetes Australia Northern Territory, Eli Lilly Australia, Estate of the Late Edward Wilson, GlaxoSmithKline, Jack Brockhoff Foundation, Janssen-Cilag, Kidney Health Australia, Marian & FH Flack Trust, Menzies Research Institute, Merck Sharp & Dohme, Novartis Pharmaceuticals, Novo Nordisk Pharmaceuticals, Pfizer Pty Ltd, Pratt Foundation, Queensland Health, Roche Diagnostics Australia, Royal Prince Alfred Hospital, Sydney, Sanofi Aventis, Sanofi Synthelabo. Also, for their invaluable contribution to the set-up and field activities of AusDiab, we are enormously grateful to A. Allman, B. Atkins, S. Bennett, A. Bonney, S. Chadban, M. de Courten, M. Dalton, D. Dunstan, T. Dwyer, H. Jahangir, D. Jolley, D. McCarty, A. Meehan, N. Meinig, S. Murray, K. O'Dea, K. Polkinghorne, P. Phillips, C. Reid, A. Stewart, R. Tapp, H. Taylor, T. Whalen, F. Wilson, and P. Zimmet.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS AND PROCEDURES
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGMENTS
  8. DISCLOSURE
  9. REFERENCES
  • 1
    Cameron AJ, Welborn TA, Zimmet PZ et al. Overweight and obesity in Australia: the 1999–2000 Australian Diabetes, Obesity and Lifestyle Study (AusDiab). Med J Aust 2003;178:427432.
  • 2
    Walls HL, Wolfe R, Haby MM et al. Trends in body mass index in urban Australian Adults, 1980–2000. Public Health Nutrition 2009;22:18.
  • 3
    WHO. Obesity: Preventing and Managing the Global Epidemic. World Health Organization: Geneva, 1998.
  • 4
    Parikh NI, Pencina MJ, Wang TJ et al. Increasing trends in incidence of overweight and obesity over 5 decades. Am J Med 2007;120:242250.
  • 5
    Wang Y, Beydoun MA, Liang L et al. Will all Americans become overweight or obese? estimating the progression and cost of the US obesity epidemic. Obesity (Silver Spring) 2008;16:23232330.
  • 6
    Kelly T, Yang W, Chen CS, Reynolds K, He J. Global burden of obesity in 2005 and projections to 2030. Int J Obes (Lond) 2008;32:14311437.
  • 7
    Dal Grande E, Gill T, Taylor AW, Chittleborough C, Carter P. Obesity in South Australian adults—prevalence, projections and generational assessment over 13 years. Aust N Z J Public Health 2005;29:343348.
  • 8
    Wang YC, Colditz GA, Kuntz KM. Forecasting the obesity epidemic in the aging U.S. population. Obesity (Silver Spring) 2007;15:28552865.
  • 9
    Government Office for Science. Tackling Obesities: Future choices - Modelling Future Trends in Obesity and Their Impact on Health. Government Office for Science: Canberra, 2007.
  • 10
    Stewart ST, Cutler DM, Rosen AB. Forecasting the effects of obesity and smoking on U.S. life expectancy. N Engl J Med 2009;361:22522260.
  • 11
    Zaninotto P, Head J, Stamatakis E, Wardle H, Mindell J. Trends in obesity among adults in England from 1993 to 2004 by age and social class and projections of prevalence to 2012. J Epidemiol Community Health 2009;63:140146.
  • 12
    Ruhm C. Current and Future Prevalence of Obesity and Severe Obesity in the United States. Forum for Health Economics and Policy 2007. <http:www.nber.orgpapersw13181.pdf>.
  • 13
    Ministry of Health. Tackling the Obesity epidemic: New Zealand 1977–2003. Ministry of Health: Wellington, 2004.
  • 14
    Booth ML, Dobbins T, Okely AD, Denney-Wilson E, Hardy LL. Trends in the prevalence of overweight and obesity among young Australians, 1985, 1997, and 2004. Obesity (Silver Spring) 2007;15:10891095.
  • 15
    Kramer H, Cao G, Dugas L et al. Increasing BMI and waist circumference and prevalence of obesity among adults with Type 2 diabetes: the National Health and Nutrition Examination Surveys. J Diabetes Complicat 2010;24:368374.
  • 16
    Ministry of Health. A Portrait of Health: Key Results of the 2006/07 New Zealand Health Survey. Ministry of Health: Wellington; 2008.
  • 17
    Flegal KM, Carroll MD, Ogden CL, Curtin LR. Prevalence and trends in obesity among US adults, 1999–2008. JAMA 2010;303:235241.
  • 18
    Dunstan DW, Zimmet PZ, Welborn TA et al.; Australian Diabetes, Obesity and Lifestyle Study (AusDiab). The Australian Diabetes, Obesity and Lifestyle Study (AusDiab)–methods and response rates. Diabetes Res Clin Pract 2002;57:119129.
  • 19
    Magliano DJ, Shaw JE, Shortreed SM et al. Lifetime risk and projected population prevalence of diabetes. Diabetologia 2008;51:21792186.
  • 20
    Magliano DJ, Liew D, Pater H et al. Accuracy of the Australian National Death Index: comparison with adjudicated fatal outcomes among Australian participants in the Long-term Intervention with Pravastatin in Ischaemic Disease (LIPID) study. Aust N Z J Public Health 2003;27:649653.
  • 21
    WHO, IASO (International Association for the Study of Obesity), IOT (International Obesity TaskForce). The Asia-Pacific Perspective: Redefining Obesity and its Treatment. Sydney, 2000.
  • 22
    Australian Institute of Health and Welfare. Grim Books Version 8. 2005 <http:www.aihw.gov.au>. Accessed 1 June 2007.
  • 23
    Cleveland W, Devlin S, Grosse E. Regression by local fitting. J Econom 1988;37:87114.
  • 24
    Schneider H, Dietrich ES, Venetz WP. Trends and stabilization up to 2022 in overweight and obesity in Switzerland, comparison to France, UK, US and Australia. Int J Environ Res Public Health 2010;7:460472.
  • 25
    Australian Bureau of Statistics. National Health Survey: Summary of Results. Australian Bureau of Statistics, Canberra, 2009.
  • 26
    Walls HL, Peeters A, Loff B, Crammond BR. Why education and choice won't solve the obesity problem. Am J Public Health 2009;99:590592.
  • 27
    Backholer K, Walls HL, Magliano DJ, Peeters A. Setting population targets for measuring successful obesity prevention. Am J Public Health 2010;100:20332037.
  • 28
    Walls HL, Peeters A, Proietto J, McNeil JJ. Public health campaigns and obesity - a critique. BMC Public Health, in press.