SEARCH

SEARCH BY CITATION

Keywords:

  • Adolescence;
  • community;
  • intervention;
  • obesity

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Conflict of Interest Statement
  9. Acknowledgements
  10. References

It's Your Move!’ was a 3-year intervention study implemented in secondary schools in Australia as part of the Pacific Obesity Prevention In Communities Project. This paper reports the outcome results of anthropometric indices and relevant obesity-related behaviours. The interventions focused on building the capacity of families, schools and communities to promote healthy eating and physical activity. Baseline response rates and follow-up rates were 53% and 69% respectively for the intervention group (n = 5 schools) and 47% and 66% respectively for the comparison group (n = 7 schools). Statistically significant relative reductions in the intervention versus comparison group were observed: weight (−0.74 kg, P < 0.04), and standardized body mass index (−0.07, P < 0.03), and non-significant reductions in prevalence of overweight and obesity (0.75 odds ratio, P = 0.12) and body mass index (−0.22, P = 0.06). Obesity-related behavioural variables showed mixed results with no pattern of positive intervention outcomes. In conclusion, this is the first study to show that long-term, community-based interventions using a capacity-building approach can prevent unhealthy weight gain in adolescents. Obesity prevention efforts in this important transitional stage of life can be successful and these findings need to be translated to scale for a national effort to reverse the epidemic in children and adolescents.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Conflict of Interest Statement
  9. Acknowledgements
  10. References

Obesity is a major public health issue (1,2) and the prevalence of overweight/obese children and adolescents has increased over time, both in Australia and internationally (3,4). There was no change in overweight/obesity among young girls in Australia from 1969–1985, but there was an increase of 35% among young boys (10 to 15 years) (5). Between 1985 and 1995, the prevalence of overweight/obesity almost doubled from 11.8% to 21.5% among girls (7 to 15 years) and 10.7% to 20.0% among boys (7 to 15 years) (5). Further rises were recorded in 2007–2008 for children aged 5–17 years with proportions of overweight/obesity for girls at 24% and boys at 26% (6). Furthermore, 61% of Australian adults are either overweight or obese (7). Given that adolescent obesity is a robust predictor of adult obesity (8) and its attendant negative health consequences (7), these rapid increases provide a clear mandate for child- and adolescent-focused interventions in Australia.

Interventions targeting nutrition and physical activity behavioural changes need to be broad in scope and encompass not only individual factors but also environmental and socio-cultural determinants. Priority needs to be given to multi-strategy, multi-setting prevention efforts, particularly among children and adolescents (9–14). While there have been obesity prevention interventions among younger children (15–19), there are few such initiatives targeting adolescents. Of those, most are short-duration and/or focused on one sex, and have had disappointing results; none included a wide age range of adolescents. Frenn and colleagues (20) conducted an effective adolescent intervention, which resulted in decreased fat consumption and increased exercise duration. However, the intervention was short, the sample small and no positive anthropometric outcomes have been published. Killen et al. (21) recorded decreases in mean body mass index (BMI) and skin-folds in their adolescent sample (n = 1447) after a 7-week education-only intervention. Finally, in a female-only, multi-strategy intervention which included physical education class participation, individual counselling, peer get-togethers and some parent involvement, Neumark-Sztainer et al. (22) reported no significant anthropometric outcomes but some improvements in sedentary and dietary behaviours and in self-image. Adolescence is the last period during which young people are still a ‘captive audience’ in a school setting and it is a time of significant growth and establishment of adult patterns of behaviour. Consequently, effective innovative approaches to obesity prevention that are flexible, cost-effective, equitable and sustainable are urgently needed, and comprehensive community-wide interventions are one promising approach (12,23).

Two community-based, multi-strategy obesity prevention projects in Australia that focused on empowering communities by increasing their capacity to recognize and address the obesity issue have reported positive anthropometric outcomes (18,19). The Romp & Chomp intervention targeting children under 5 years resulted in significantly lower mean weight, BMI and lower prevalence of overweight/obesity as well as some positive behavioural differences (18). Be Active Eat Well, which was aimed at primary school children, reported lower increases in body weight, waist, waist/height ratio and standardized BMI (BMI-z) in the intervention group (19). While these findings support the efficacy of a community-based approach to obesity prevention in younger children, such an approach has not been previously tested among adolescents.

The It's Your Move! (IYM) project, the Australian arm of the wider Pacific Obesity Prevention In Communities (OPIC) Project, sought to fill this gap with a 3-year intervention from 2006 to 2008. The project was developed to test the effectiveness and economic efficiency of a multi-focused, multi-site, community-based intervention to reduce adolescent overweight and obesity by building community capacity to promote healthy eating and physical activity. The objective of the current paper is to report the anthropometric and dietary and physical activity behavioural outcomes of the IYM project.

Methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Conflict of Interest Statement
  9. Acknowledgements
  10. References

Intervention design

Detailed descriptions of the intervention design and implementation methods specific to IYM are provided elsewhere (23–25) and briefly following. The IYM programme targeted secondary students aged 12–18 years and their families. There were 10 objectives, each comprising a variety of strategies delivered in schools through school project officers and student ambassadors (i) capacity building among school project officers and student ambassadors (workshop and training opportunities); (ii) increasing awareness of project messages (project logo and branding and development of social marketing materials); (iii) evaluation; (iv) promoting water, reducing soft drink consumption (distribution of refillable water bottles, water polices, installation of new water fountains, removal of soft drinks from vending machines); (v) promoting healthy breakfasts (social marketing and breakfast programmes); (vi) increasing fruit and vegetable consumption (soup days, juice days, vegetable gardens and social marketing); (vii) increasing healthiness of school food (traffic light system for food sold by school canteens, recipe books provided to canteens, healthy eating days, parent information canteen staff training); (viii) promoting active transport to/from school (riding to school programme, lunchtime and other walking groups); (ix) increasing participation in organized sports and other active recreation (professional development for physical education teachers, lunchtime activities, education sessions and sports-related excursions); (x) promoting acceptance of healthy body size and shape (education strategies and social marketing).

Study design and participants

The study design was quasi-experimental, using a longitudinal cohort follow-up. Baseline measurements were collected from 2005 to 2006 and follow-up from 2006 to 2008 as students left school. The intervention and comparison sites were in the Barwon-South West Region of Victoria (population 350 109), which covers the south-west coast of Victoria and includes the regional centre Geelong (population 205 929 in 2006). The intervention sample was selected as all five secondary schools (three government, one Catholic and one independent) in the East Geelong and Bellarine Peninsula regions of Geelong. The comparison sample was a stratified random sample of schools (n = 7; four government, one Catholic and two Christian) from the Barwon-South West Region of Victoria. All the areas from which the intervention and comparison samples were drawn were ranked less than the state's average on an index of relative socioeconomic disadvantage, meaning they were areas of relative disadvantage (26).

All students and their parents gave written consent to their participation in the project. The IYM project had ethics approval from the Deakin University Human Research Ethics Committee (EC 37-2004) and was registered as a trial (ACTRN #12607000257460).

Measures

Demographic information was collected via paper questionnaires, while electronic Personal Diary Assistants were used for the administration of a knowledge, attitudes and behaviours survey. The survey consisted of questions focusing on nutrition patterns, physical activity and leisure time behaviours, quality of life, perceptions of and attitudes about body size, family and home environment, school environment and neighbourhood environment (27) (for all reports see: http://www.deakin.edu.au/hmnbs/who-obesity/reports-and-grants/reports.php). Health-related quality of life was measured using two instruments: the Assessment of Quality of Life instrument (AQoL-6D) developed by Hawthorne and Richardson (28) in Australia and the Pediatric Quality of Life Inventory 4.0 (generic module for 13- to 18-year-olds) (PedsQL) developed by Varni and colleagues (29,30).

Anthropometric data (height, weight, bioimpedance) were collected by trained research staff using standardized protocols (25). Briefly, students were measured using a portable stadiometer (Surgical and Medical PE87) for height to the nearest 0.1 cm and a TANITA Body Composition Analyser (Model BC 418, Wedderburn Australia) for body weight. Other anthropometric measures included: body composition (percentage body fat), BMI (weight in kg/[height in m]2) and BMI-z score, calculated using the World Health Organization (WHO) Reference 2007 (31). The WHO Reference 2007 age-specific BMI cut-offs were also used to classify children's weight status as either healthy weight or overweight/obese and percentage body fat was derived from bioelectric impedance measures and equations validated for multi-ethnic adolescent populations (32).

Statistical analysis

Analyses were conducted using stata release 11.0 (StataCorp., College Station, TX, USA, 2009). Demographic data were analysed using descriptive statistics, independent groups t-tests or, where applicable, chi-squared tests. Differences between follow-up (participants who were measured twice) and non-follow-up (those who were measured once) were tested with t-tests or chi-squared tests where applicable and any significant differences were entered into a logistic regression model for further testing. Difference from baseline to follow-up in the prevalence of overweight/obesity between study conditions and between schools, while controlling for duration between measurements, was tested for significance using Newcombe's paired differences (33). In all cases, P < 0.05 was considered statistically significant. Differences in follow-up anthropometry and quality of life were determined by separate regression models with group (intervention or comparison) entered into the model with the following covariates: baseline variable, age at follow-up, height at follow-up (weight), gender and duration between measurements. Differences in follow-up weight status and behaviours (categorical measures) were also determined by separate regression models with group (intervention or comparison) entered into the model with the following covariates: baseline variable, age at follow-up, gender and duration between measurements. All regression models accounted for the clustering of data by school. All variables were checked for missing and out-of-range values and cases with outlying (>3 SD from mean) values on the anthropometric variables baseline or follow-up were removed from relevant analyses (34). Multivariate outliers were identified using Hadi's method (35) and excluded from the relevant analyses. Seven cases were removed from the fat percentage analyses because of values of less than 5%. Intraclass correlations (ICCs) between school and anthropometric measures were calculated using one-way anova at baseline.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Conflict of Interest Statement
  9. Acknowledgements
  10. References

The IYM intervention was delivered to all students in the intervention schools. The overall baseline response rate was about 50% and approximately two-thirds of those students were followed-up (Fig. 1). The baseline characteristics of the adolescents who were followed-up were similar between the intervention and comparison groups and the mean duration between measurements was the same (Table 1). The adolescents who were followed-up were different from those who were not followed-up on some demographic variables (Table 1). Logistic regression, with participation completion as the dependent variable, showed that, independently, those followed-up were more likely to be younger, female, with lower BMI-z scores and be from the intervention group, χ2(4,2998) = 56.83, P < 0.001.

image

Figure 1. Flow diagram of participants.

Download figure to PowerPoint

Table 1.  Baseline and follow-up characteristics of the participants
 InterventionComparisonTotal
BaselineFollow-upBaselineFollow-upBaselineFollow-upNon-follow-up
  1. Bolding indicates significance P < 0.05.

  2. AQoL, Assessment of Quality of Life, scale 0–1; BMI, body mass index; BMI-z, standardized body mass index; PedsQL, Pediatric Quality of Life, scale 0–100; SD, standard deviation.

n 1276 778 2054986
Male (%) 57.8 46.4 53.561.6
Age, years (SD)14.5 (1.40)16.8 (1.17)14.7 (1.45)16.9 (1.08)14.6 (1.42)16.8 (1.14)14.8 (1.25)
Height, cm (SD)164.2 (10.0)170.6 (9.2)164.6 (9.5)170.1 (8.7)164.3 (9.8)170.4 (9.0)165.9 (9.3)
Weight, kg (SD)58.5 (13.2)67.1 (14.0)58.8 (12.7)67.0 (12.7)58.6 (13.0)67.1 (13.5)61.7 (13.9)
BMI, kg m−2 (SD)21.6 (3.8)23.0 (4.1)21.6 (3.5)23.1 (3.6)21.5 (3.7)23.1 (3.9)22.2 (4.0)
BMI-z score (SD)0.52 (1.05)0.50 (1.05)0.48 (1.01)0.53 (0.95)0.50 (1.04)0.51 (1.01)0.65 (1.05)
Fat, % (SD)29.0 (9.3)27.4 (10.3)29.1 (9.2)28.8 (10.4)29.0 (9.2)28.0 (10.4)28.3 (9.7)
Normal weight (%)69.872.772.971.771.172.165.5
Overweight/obese (%)30.227.327.128.328.927.934.5
AQoL0.86 (0.18)0.85 (0.20)0.87 (0.16)0.85 (0.18)0.86 (0.17)0.85 (0.19)0.82 (0.21)
PedsQL78.8 (11.0)79.0 (10.9)79.3 (10.2)78.8 (10.0)79.0 (10.7)78.9 (10.6)76.8 (12.7)
Years between measures (SD) 2.3 (0.74) 2.3 (0.55) 2.3 (0.68) 

Figure 2 shows the changes in prevalence of overweight and obesity (controlled for duration) in each of the schools and for the intervention and comparison groups. Three of the five intervention schools had a significant decrease in prevalence of overweight/obesity, while two intervention schools and five of the seven comparison schools did not change. The intervention group also showed a decrease, while the comparison group did not. Table 2 describes the follow-up outcomes of all anthropometric and quality-of-life indicators. The full models which adjusted for baseline variable, duration, age, gender, height (for weight analyses) and clustering by school showed that the intervention group gained significantly less weight (740 g) and less BMI-z (0.08 units) than students in the comparison group. The fully adjusted models were not statistically significant for BMI, body fat percentage, proportion of overweight/obesity or quality of life. ICCs for anthropometric measures at baseline were all low; ranging from 0.006 to 0.023.

image

Figure 2. Change in prevalence of overweight/obese in intervention and comparison groups shown as individual schools (left panel) and combined (right panel: intervention [−3.5; CI −5.69, −1.28]; comparison [2.3; CI −0.44, 4.97]). Point estimates and 95% confidence intervals (controlled for duration between measures) are shown as filled markers for the intervention schools and unfilled markers for the comparison schools.

Download figure to PowerPoint

Table 2.  Unadjusted and adjusted* differences in outcome measures between comparison (reference) and intervention groups
MeasureICC (baseline)Difference (unadjusted)Difference (adjusted)Robust SEP
  • Bolding indicates significance P < 0.05.

  • *

    Adjusted for baseline variable, age at follow-up, gender and time between measurements (and height at follow-up in the weight regression) and clustering by school.

  • ****

    There are no ICCs to report for these variables as we only report them for continuous anthropometric measures.

  • Percentage points.

  • Odds ratio.

  • AQoL, Assessment of Quality of Life, scale 0–1; BMI, body mass index; BMI-z, standardized body mass index; ICC, intraclass correlation; PedsQL, Pediatric Quality of Life Inventory, scale 0–100; SE, standard error.

All
Weight, kg0.0230.080.740.320.04
BMI0.018−0.18−0.220.100.06
BMI-z0.007−0.080.070.030.03
Body fat percentage0.006−1.36−0.230.400.58
Proportion overweight/obese****−4.280.750.140.12
AQoL****0.005−0.0010.010.93
PedsQL****0.610.090.440.85

Frequencies of response and adjusted differences in outcomes for the behavioural and body image measures for students in the intervention and comparison groups are presented in Table 3. Overall, there were few differences between students in the intervention and comparison groups for the dietary and activity behaviours and body image perceptions. Specifically, there were no improvements in breakfast consumption, home lunches, fruit or vegetable consumption, limiting soft drinks, cordials or snack foods, or school time activity for the intervention students over the comparison students. However, the proportion of students in the intervention group who actively travelled to school increased over the intervention period, while the proportion in the comparison group decreased (P = 0.01). It was surprising that the proportion of students who were active after school decreased in both the intervention and comparison groups, and that this decrease was greater for the intervention group. Overall, the adolescents reported increased computer use and decreased television use, but the comparison group had more favourable changes than the intervention group.

Table 3.  Unadjusted proportions (95% confidence interval) of behavioural measures at baseline and follow-up and adjusted* odds ratios (ORs) and P-values for behavioural outcomes at follow-up for comparison (reference) and intervention groups
 InterventionComparisonAdjusted ORP
BaselineFollow-upBaselineFollow-up
  • *

    Adjusted for baseline variable, age at follow-up, gender, time between measurements and clustering by school.

  • Among students living within 30-min walk from school; maximum trips were 10 per week to or from school.

  • Bolded values indicates significance P < 0.05.

Eating and diet      
 Breakfast before school (% 4–5 d)88.7 (86.9, 90.5)87.8 (85.8, 89.7)92.5 (90.5, 94.5)87.9 (85.4, 90.4)1.090.70
 Source of school lunch (% from home)88.9 (86.9, 90.4)86.0 (84.1, 87.9)85.1 (82.6, 87.6)82.8 (80.2, 88.5)1.150.46
 Fruit (≥2 servings per day)58.0 (55.2, 60.7)53.7 (50.9, 56.5)56.9 (53.3, 60.4)56.3 (52.8, 60.0)0.900.51
 Vegetable (≥2 servings per day)76.3 (73.9, 78.7)75.2 (72.8, 77.6)81.4 (78.6, 84.2)81.4 (78.6, 84.2)0.770.14
 Sugar-sweetened soft drink (≥3 d)48.3 (45.5, 51.1)44.4 (41.7, 47.2)38.0 (34.5, 41.5)35.2 (31.8, 38.6)1.210.17
 Fruit drink/cordial (≥3 d)62.5 (59.8, 65.1)55.6 (52.9, 58.4)63.5 (60.0, 66.9)57.4 (53.9, 60.9)0.870.36
 Snack food from takeaway shop or milk bar after school (≥3 d)21.3 (19.0, 23.6)26.9 (24.5, 29.4)26.0 (22.9, 29.1)27.6 (24.4, 30.8)1.010.93
Activities      
 Walk/cycle to school (≥5 times per week)61.1 (55.7, 66.5)64.6 (59.3, 69.8)55.8 (50.4, 61.1)53.6 (48.2, 59.0)1.490.01
 Recess (% mostly inactive)16.2 (13.9, 18.6)33.5 (30.5, 36.6)22.3 (19.0, 25.7)33.5 (30.5, 36.6)0.950.92
 Lunch time (% mostly inactive)12.9 (10.7, 15.1)31.5 (28.5, 34.5)18.7 (15.5, 21.8)34.2 (30.4, 38.0)1.160.57
 Active after school (3–5 d)55.7 (52.9, 58.4)48.4 (45.6, 51.1)60.6 (57.1, 64.0)54.2 (50.6, 57.8)0.750.01
 Average time watching TV, videos, DVDs per day (% ≤2 h)72.7 (69.8, 75.6)73.5 (70.7, 76.4)79.0 (75.7, 82.2)82.3 (79.2, 85.3)0.670.001
 Average time playing video games, electronic games or using computer (not for homework) per day (% ≤1 h)68.2 (65.3, 71.1)54.8 (51.7, 57.9)76.7 (73.0, 80.4)69.0 (64.9, 73.0)0.600.001
Body image      
 Describing weight as about right59.8 (57.1, 62.6)63.7 (63.0, 69.7)59.3 (55.8, 62.8)66.3 (63.0, 69.7)−0.150.25
 Trying to lose weight30.8 (28.2, 33.3)30.8 (28.2, 33.3)31.4 (28.1, 34.8)31.1 (27.8, 34.3)0.120.29

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Conflict of Interest Statement
  9. Acknowledgements
  10. References

This evaluation of the IYM intervention provides important evidence of positive anthropometric outcomes in adolescents. It has shown for the first time that a long-term, capacity-building approach can reduce overweight and obesity in adolescents. The relative reduction in body weight was about three-quarters of a kilogram. While this is a small weight change for the average individual, the classic work of Rose (36) showed how this can translate into substantial changes for the population (in adults, a mean weight change of 1 kg corresponds to 2% points change in obesity prevalence [BMI > 30 kg m−2](37)) Furthermore, the average annual secular weight gain in Australian adolescents between the last two national cross-sectional surveys in 1995 and 2007 was about 500 g per year (4,38). Thus, the longitudinal relative reduction in weight of 740 g over 2.3 years in the intervention group is substantial in terms of the background trajectory of the obesity epidemic.

These findings are encouraging as they indicate the effectiveness of a community-based obesity prevention intervention using a capacity-building approach for Australian adolescents. It is essential to target this age group, as adolescence is one of the most vulnerable periods for the development of overweight/obesity (2) and increased morbidity and mortality has been linked directly to adolescent obesity (39). Despite this, there are few published findings for effective multi-strategy obesity prevention interventions among adolescents (40,41) and none that included younger and older adolescents.

Despite the positive anthropometric outcomes, there were few observed consistent positive behavioural outcomes in self-reported dietary and activity measures in the IYM project. The one positive significant finding that was consistent was the increased use of active transport among adolescents in the intervention group compared to the comparison group. However, this finding must be interpreted with caution, as the analysis included only those adolescents who lived within a 30-min walking distance of the school and many of the adolescents, because of distance from school, used bus transport. Other school-based interventions also found mixed results for behavioural outcomes. Gortmaker et al. (16) reported a decrease in prevalence of obesity among girls concurrent with a reported decrease in television viewing and an increase in fruit and vegetable consumption. Similarly, Haerens et al. (40) reported significantly less increases in BMI and BMI-z among girls along with an increase in physical activity and a positive intervention effect on fat intake. In contrast, Spiegel and Foulk (42) found significant positive shifts in BMI among the intervention group and no corresponding shifts among the comparison group, but there were no notable changes in physical activity or dietary behaviours. Hence, the findings are still mixed and may be explained by the lack of sensitivity of the data collection instruments used. Perhaps, a mixed methodology using objective measurements such as pedometers in addition to self-report would lead to more precise measures. Another possible explanation for reported lack of behavioural change could have been the selection of behaviours for intervention targeting (and thus measurement). For example, about 90% of adolescents reportedly ate breakfast on a regular basis and around the same proportion brought their lunch from home and these figures may indicate a ceiling effect. Furthermore, vegetables were consumed on a regular basis both at baseline and follow-up by this cohort of adolescents and so, the quality of food may not have been the issue but the quantities (unmeasured) may have been. Probably, the most plausible explanation is that (i) behavioural changes did occur to create differences in energy balance; (ii) multiple, small and variable behaviour changes occurred as a result of the multi-strategy, long-term intervention and (iii) these changes were below the sensitivity of the measures used. Self-reported behavioural data have its limitations, although the level of recall and social desirability bias would be expected to be similar between groups.

Schools are important settings for obesity prevention, because young people spend a lot of time at school, thus presenting a unique opportunity for influencing positive nutritional and activity behaviours. All the intervention strategies within IYM were based on the 10 overarching objectives in the action plan, but there was flexibility across schools in development and implementation to ensure relevance and ownership. This resulted in some heterogeneity of intervention activities and their delivery and may have resulted in different levels of intervention dose. In turn, this heterogeneity may account for the differences in changes in prevalence of overweight/obesity between the intervention schools with only three of the five intervention schools recording decreases in prevalence of overweight/obesity from baseline to follow-up.

One of the strengths of this evaluation was that it included measures to ensure that there were no adverse effects of IYM on increased risk of disordered eating. We found that quality of life was similar over time and between groups. Likewise, there was no increase in the proportion of students trying to lose weight over time.

The major strengths of the IYM study were its: capacity-building approach allowing for fidelity of process but local flexibility of content and delivery; partnership structures between intervention and evaluation teams; long duration; hard anthropometric outcomes; and large sample sizes. There were also some limitations to consider when interpreting the results. The population from which the sample was drawn is not representative of the wider Australian population, particularly in terms of cultural diversity, or of adolescent populations outside of Australia. The only measure available of socioeconomic position of participants was an area-level measure of relative socioeconomic advantage or disadvantage. As using area-level indicators as individual-level indicators results in misclassification errors (43), these were not included in the analysis, but the intervention and comparison groups had similar scores for socioeconomic disadvantage.

The response rate was approximately 50%, which was similar to other community-based, child obesity prevention interventions conducted in Australia (19,44) and elsewhere (45,46). However, the prevalence of overweight and obese adolescents in this sample was similar to the wider population of young Australians (6). It remains possible, however, that overweight and obese children were under-represented, making our estimates of prevalence on the low side. Non-participation in the evaluation by some obese children could also have influenced the apparent effectiveness of the intervention in either direction (i.e. obese adolescents may respond more or less to the intervention compared to other adolescents).

There were important differences between students who were followed-up and not followed-up and this may have resulted in response bias. More students who were not followed-up were male and from the comparison group. The comparison schools were generally in more rural settings than the intervention schools, and students, particularly male students, in more rural and remote areas tend to leave school at an earlier age than their urban counterparts (47). These facts may explain why there was a differential attrition rate between conditions and genders. Of potential concern for external validity are the higher BMI-z scores of the students who dropped out. Other obesity interventions also reported that participants with higher BMI or BMI-z scores were more likely to drop out of the study (16). Future interventions could benefit from more engagement with those adolescents who are more likely to drop out.

The issue of cluster effects is central to the analyses of these quasi-experimental studies. ICCs measure the relatedness of students within each school on the outcome measure (48). The baseline ICCs for the main anthropometric variables (Table 2) were very low; hence, it can be concluded that the students were very similar within and between schools. While these correlations were low, the fact that there were few schools meant there were major design effects such that cluster adjustment resulted in greatly widened confidence intervals around the outcome estimates (48). Whole-of-community programmes, such as IYM, where the intervention is across a whole area containing several schools, do not lend themselves to cluster-randomized trial designs because of the costs of intervening in 10–20 whole communities with a similar number of comparison sites. Because of this, IYM could not be powered (49) to ensure that adjustment for clustering by school gave a true result (i.e. a true positive result could easily have been masked by the widened confidence intervals created by the cluster-adjusted analysis). Nevertheless, the results presented in this paper were adjusted for cluster effects and two of the anthropometric measures (weight and BMI-z) remained statistically significant despite this adjustment, giving confidence that a true positive effect did occur.

Conclusion

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Conflict of Interest Statement
  9. Acknowledgements
  10. References

This paper reports on a successful obesity prevention programme in adolescents, which resulted in 740 g less weight gain compared to a comparison population. This is one of the first studies worldwide to demonstrate success in reducing unhealthy weight gain within this age group. The intervention itself took a complex, multifaceted approach based on capacity building within a community-based intervention. These promising results indicate that obesity prevention efforts, even in this challenging and transitional stage of childhood, can make an important impact. The success of this novel approach provides a new avenue for interventions to prevent obesity among adolescents and points to the potential for community-based, capacity-building interventions in preventing adolescent obesity and its subsequent complications.

Conflict of Interest Statement

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Conflict of Interest Statement
  9. Acknowledgements
  10. References

L. Millar, M. Malakellis, P. Kremer, H. Mavoa, M. Moodie and B. A. Swinburn's institutions have received grants from National Health and Medical Research Council. Support was provided to cover costs of travel to New Zealand and to Investigator's meetings. The authors were employed by Deakin University.

N. Robertson's institution has received funds from National Health and Medical Research Council. The author was employed by Deakin University.

A. de Silva-Sanigorski and M. P. McCabe's institutions received grants from the National Health and Medical Research Council. The authors were employed by Deakin University.

In addition, McCabe's costs of travelling to New Zealand and Investigator's meetings were covered.

L. Mathews' institution has received grants from the Department of Human Services, Victoria. Support was provided to cover the cost of travel by the National Health and Medical Research Council & VICHEALTH. Fees for participation in review activities such as data monitoring boards, statistical analysis, end point committees, and the like, were also covered. The author is currently employed.

J. Utter's institution has received a grant from the Health Research Council of New Zealand.

G. Roberts and C. Bell declared no conflict of interests.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Conflict of Interest Statement
  9. Acknowledgements
  10. References

The authors would like to thank the many people involved in the Pacific OPIC Project including co-investigators, other staff and postgraduate students, partner organizations, and especially the schools, students, parents and communities. The funding for the project was from the Victorian Department of Health, the National Health and Medical Research Council (in conjunction with the Health Research Council [New Zealand] and the Wellcome Trust [UK] as part of their innovative International Collaborative Research Grant Scheme), and AusAID.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Conflict of Interest Statement
  9. Acknowledgements
  10. References
  • 1
    Haslam DW, James WPT. Obesity. Lancet 2005; 366: 11971209.
  • 2
    Lobstein T, Baur L, Uauy R. Obesity in children and young people: a crisis in public health. Obes Rev 2004; 5: 485.
  • 3
    Wang Y, Lobstein T. Worldwide trends in childhood overweight and obesity. Int J Pediatr Obes 2006; 1: 1125.
  • 4
    Department of Health and Ageing (DoHA). Australian National Children's Nutrition and Physical Activity Survey – Key Findings. Australian Government: Canberra, 2007.
  • 5
    Magarey A, Daniels L, Boulton T. Prevalence of overweight and obesity in Australian children and adolescents: reassessment of 1985 and 1995 data against new standard international definitions. Med J Aust 2001; 174: 561564.
  • 6
    Australian Bureau of Statistics. Children who are overweight or obese. Australian Social Trends, Sep 2009. Australian Government: Canberra, 2009. [WWW document]. URL http://www.abs.gov.au/AUSSTATS/abs@.nsf/Lookup/4102.0Main+Features20Sep+2009 (accessed March 2010).
  • 7
    Australian Institute of Health and Welfare. Australia's health 2010. Australia's Health Series No. 12. AIHW: Canberra, 2010, p. 62.
  • 8
    Magarey AM, Daniels LA, Boulton TJ, Cockington RA. Predicting obesity in early adulthood from childhood and parental obesity. Int J Obes Relat Metab Disord 2003; 27: 505513.
  • 9
    Summerbell C, Waters E, Edmunds L, Kelly S, Brown T, Campbell K. Interventions for preventing obesity in children. Cochrane Database Syst Rev 2005; 3: 170.
  • 10
    de Silva-Sanigorski A, Economos C. Evidence of multi-setting approaches for obesity prevention: translation to best practice for future interventions. In: Waters E, Swinburn B, Seidell J, Uauy R (eds). Preventing Childhood Obesity: Evidence Policy and Practice. Wiley-Blackwell; BMJ Books: West Sussex, 2010, pp. 5763.
  • 11
    Economos CD, Hyatt RR, Goldberg JP, Must A, Naumova EN, Collins JJ et al. A community intervention reduces BMI z-score in children: shape up Somerville first year results. Obesity 2007; 15: 13251336.
  • 12
    Flynn MAT, McNeil DA, Maloff B, Mutasingwa D, Wu M, Ford C et al. Reducing obesity and related chronic disease risk in children and youth: a synthesis of evidence with ‘best practice’ recommendations. Obes Rev 2006; 7: 766.
  • 13
    Swinburn BA, de Silva-Sanigorski AM. Where to from here for preventing childhood obesity: an international perspective. Obesity 2010; 18: S4S7.
  • 14
    Egger G, Swinburn B. An ‘ecological’ approach to the obesity pandemic. BMJ 1997; 315: 477480.
  • 15
    Sallis JF, McKenzie TL, Alcaraz JE, Kolody B, Hovell MF, Nader PR. Project spark – effects of physical-education on adiposity in children. Prev Treat Child Obes 1993; 699: 127136.
  • 16
    Gortmaker SL, Peterson K, Wiecha J, Sobol AM, Dixit S, Fox MK et al. Reducing obesity via a school-based interdisciplinary intervention among youth: planet health. Arch Pediatr Adolesc Med 1999; 153: 409418.
  • 17
    Sahota P, Rudolf MCJ, Dixey R, Hill AJ, Barth JH, Cade J. Randomised controlled trial of primary school based intervention to reduce risk factors for obesity. BMJ 2001; 323: 15.
  • 18
    de Silva-Sanigorski AM, Bell AC, Kremer P, Nichols M, Crellin M, Smith M et al. Reducing obesity in early childhood: results from Romp & Chomp, an Australian community-wide intervention program. Am J Clin Nutr 2010; 91: 831840.
  • 19
    Sanigorski AM, Bell AC, Kremer PJ, Cuttler R, Swinburn BA. Reducing unhealthy weight gain in children through community capacity-building: results of a quasi-experimental intervention program, be active eat well. Int J Obes 2008; 32: 10601067.
  • 20
    Frenn M, Malin S, Bansal NK. Stage-based interventions for low-fat diet with middle school students. J Pediatr Nurs 2003; 18: 3645.
  • 21
    Killen JD, Telch MJ, Robinson TN, Maccoby N, Taylor CB, Farquhar JW. Cardiovascular disease risk reduction for tenth graders: a multiple-factor school-based approach. JAMA 1988; 260: 17281733.
  • 22
    Neumark-Sztainer DR, Friend SE, Flattum CF, Hannan PJ, Story MT, Bauer KW et al. New moves – preventing weight-related problems in adolescent girls: a group-randomized study. Am J Prev Med 2010; 39: 421432.
  • 23
    Bell AC, Simmons A, Sanigorski AM, Kremer PJ, Swinburn BA. Preventing childhood obesity: the sentinel site for obesity prevention in Victoria, Australia. Health Promot Int 2008; 23: 328336.
  • 24
    Mathews L, Moodie M, Simmons A, Swinburn B. The process evaluation of It's Your Move!, an Australian adolescent community-based obesity prevention project. BMC Public Health 2010. doi: 10.1186/1471-2458-10-448
  • 25
    Swinburn BA, Millar L, Utter J, Kremer P, Moodie M, Mavoa H et al. The Pacific Obesity Prevention in Communities project: project overview and methods. Obes Rev 2011; 12(Suppl. 2): 311.
  • 26
    Australian Bureau of Statistics. Information Paper: Socio-Economic Indexes for Areas. Australian Government: Canberra, 2001.
  • 27
    Mathews L, Malakellis M, Sanigorski AM, Kremer P, Swinburn BA. Nutrition and Physical Activity in Children and Adolescents. Barwon-South Western Region. Sentinel Site Series. Department of Human Services (Victoria): Melbourne, 2008.
  • 28
    Hawthorne G, Richardson J, Osborne R. The Assessment of Quality of Life (AQoL) instrument: a psychometric measure of health-related quality of life. Qual Life Res 1999; 8: 209224.
  • 29
    Varni JW, Seid M, Kurtin PS. PedsQL 4.0: reliability and validity of the Pediatric Quality of Life Inventory version 4.0 generic core scales in healthy and patient populations. Med Care 2001; 39: 800812.
  • 30
    Varni JW, Seid M, Rode CA. The PedsQL: measurement model for the pediatric quality of life inventory. Med Care 1999; 37: 126139.
  • 31
    World Health Organization. WHO Reference 2007; Growth Reference Data for 5–19 Years. WHO: Geneva, 2010.
  • 32
    Sluyter JD, Schaaf D, Scragg RKR, Plank LD. Prediction of fatness by standing 8-electrode bioimpedance: a multiethnic adolescent population. Obesity 2010; 18: 183189.
  • 33
    Newcombe RG. Interval estimation for the difference between independent proportions: comparison of eleven methods. Stat Med 1998; 17: 873890.
  • 34
    Tabachnick BG, Fidell LS. Using Multivariate Statistics, 5th edn. Pearson/Allyn & Bacon: Boston, 2007.
  • 35
    Hadi AS. Identifying multiple outliers in multivariate data. J R Stat Soc Ser B Methodol 1992; 54: 761771.
  • 36
    Rose G. The Strategy of Preventive Medicine. Oxford University Press: Oxford, UK; New York, USA, 1992.
  • 37
    Rose G, Day S. The population mean predicts the number of deviant individuals. BMJ 1990; 301: 10311034.
  • 38
    Australian Bureau of Statistics. National Nutrition Survey: Selected Highlights Australia. ABS: Canberra, 1995.
  • 39
    Must A, Jacques PF, Dallal GE, Bajema CJ, Dietz WH. Long-term morbidity and mortality of overweight adolescents. N Engl J Med 1992; 327: 13501355.
  • 40
    Haerens L, Deforche B, Maes L, Cardon G, Stevens V, De Bourdeaudhuij I. Evaluation of a 2-year physical activity and healthy eating intervention in middle school children. Health Educ Res 2006; 21: 911921.
  • 41
    Rosenbaum M, Nonas C, Weil R, Horlick M, Fennoy I, Vargas I et al. School-based intervention acutely improves insulin sensitivity and decreases inflammatory markers and body fatness in junior high school students. J Clin Endocrinol Metab 2007; 92: 504508.
  • 42
    Spiegel SA, Foulk D. Reducing overweight through a multidisciplinary school-based intervention[ast]. Obesity 2006; 14: 8896.
  • 43
    Baker J, Adhikari P. Socio-Economic Indexes for Individuals and Families. Australian Bureau of Statistics: Canberra, 2007.
  • 44
    Waters E, Ashbolt R, Gibbs L, Booth M, Magarey A, Gold L et al. Double disadvantage: the influence of ethnicity over socioeconomic position on childhood overweight and obesity: findings from an inner urban population of primary school children. Int J Pediatr Obes 2008; 3: 196204.
  • 45
    Chomitz VR, McGowan RJ, Wendel JM, Williams SA, Cabral HJ, King SE et al. Healthy living Cambridge kids: a community-based participatory effort to promote healthy weight and fitness. Obesity 2010; 18: S45S53.
  • 46
    Pate RR, Ward DS, Saunders RP, Felton G, Dishman RK, Dowda M. Promotion of physical activity among high-school girls: a randomized controlled trial. Am J Public Health 2005; 95: 15821587.
  • 47
    Australian Bureau of Statistics. Australian Social Trends. Cat. No. 4102.0. Australian Bureau of Statistics: Canberra, 2003.
  • 48
    Siddiqui O, Hedeker D, Flay BR, Hu FB. Intraclass correlation estimates in a school-based smoking prevention study. Am J Epidemiol 1996; 144: 425433.
  • 49
    Swinburn B, Pryor J, McCabe MP, Carter R, de Courten M, Schaaf D et al. The Pacific OPIC Project (Obesity Prevention In Communities) – objectives and designs. Pac Health Dialog 2007; 14: 139146.