Body Mass Index of children and adolescent participants in a voucher program designed to incentivise participation in sport and physical activity

There has been limited population-level success in tackling overweight and obesity. The Active Kids program is a state-wide intervention that aims to increase participation in organised physical activity and sport among children and adolescents in NSW, Australia. This study profiled children and adolescents who registered for the Active Kids program by examining the prevalence of overweight and obesity across subgroups and by social disadvantage in this sample.


Abstract Background
There has been limited population-level success in tackling overweight and obesity. The Active Kids program is a state-wide intervention that aims to increase participation in organised physical activity and sport among children and adolescents in NSW, Australia.
This study profiled children and adolescents who registered for the Active Kids program by examining the prevalence of overweight and obesity across subgroups and by social disadvantage in this sample.

Methods
For participating children, each parent or carer was required to complete an online registration form with information about the child's height, weight, physical activity, sport participation, age, sex, primary language spoken at home, Aboriginality, disability status and postcode. Descriptive statistics were used to profile children and adolescents who registered in the program and multinomial regression models were used to determine which demographic characteristics were associated with an increased risk of overweight and obesity.

Results
In 2018, 671,375 parents registered a child or adolescent for the Active Kids Program.
Among these children and adolescents, the prevalence of overweight and obesity was 17.2% and 7.6%, respectively. A large number of children and adolescents who lived in the most disadvantaged area (n = 99,583; 14.8%) registered for the program. There was a clear socio-economic gradient for obesity prevalence across areas of increasing disadvantage, with children and adolescents living in the most disadvantaged area being 1.87 (95% CIs 1.82, 1.93) times more likely to be overweight or obese compared with children and adolescents living in the least disadvantaged area.

Conclusions
The Active Kids program successfully reached a substantial proportion of overweight and obese children from socially disadvantaged backgrounds, providing financial support and opportunities for these children to participate in organised sport and physical activity.
However, the program did not reach all children, and additional physical activity promotion strategies may be needed in a comprehensive approach. Nonetheless, these findings support government investment in reaching childhood overweight and obesity with large-scale programs.

Background
Overweight and obesity in childhood and adolescence are associated with adverse health consequences throughout the life course (1). As body mass index (BMI) increases, so does the prevalence of comorbid conditions, including cardiovascular disease, type 2 diabetes, and some cancers (2). Whilst many of these conditions occur in adulthood, early incidence of obesity poses immediate physical, social and mental health concerns during adolescence (3). Further, obese children and adolescents are five times more likely to be obese in adulthood than those who were not obese, representing a lifelong personal burden and long-term societal impacts (4).
Cross-government sector and multi-strategy approaches to obesity prevention are recommended by the World Health Organisation (WHO) (5). In New South Wales (NSW), Australia, the State Government selected childhood obesity in 2016 as one of 12 priorities (6), setting a target to reduce childhood overweight and obesity by 5% over 10 years. The strategy takes a broad state-wide approach, focussing on the modifiable behaviours associated with overweight and obesity such as physical activity and diet. Implementation is addressed through a cross-government Healthy Eating Active Living (HEAL) Strategy coordinated by the NSW Ministry of Health (7) as well as through initiatives from other Government agencies, such as the Department of Education (munch and move/healthy canteens/Live life well at school) and the Office of sport (Active Kids program) (8).
International success in tackling overweight and obesity through population-approaches has been limited (9), so it is important to understand the profile of those who participate in programs implemented at scale.
The Active Kids program, led by the NSW State Government's Office of Sport, is a policy initiative to increase participation in organised physical activity and sport among children in NSW. In the 2018 calendar year, all school-enrolled children in NSW were eligible to register for one $100 voucher that could be used towards registration costs for accredited sport, fitness and active recreation programs. The Active Kids Program has been implemented state-wide for one calendar year and it is important to describe the characteristics of those who participated. It is particularly important to determine whether this program has reached children and adolescents living in low SES areas, as there still a disparity in overweight and obesity (10). This information will inform policy makers and program deliverers and assist in improving this and future policy interventions on a national and international level. This study described the profile of children and adolescents who registered for the Active Kids program by examining the prevalence of overweight and obesity across subgroups and by social disadvantage.

Participants
All primary and secondary school-enrolled children and adolescents, residing in NSW with a valid Australian Medicare card were eligible for an Active Kids voucher. To register children in the program, the parent, carer or guardian (adult) was required to complete an online registration form.

Data collection
The Active Kids program registration form was designed by a multi-sector steering group to ensure the information collected included all relevant socio-demographic information.
The registration form data obtained for this study includes the child's date of birth (validated by Medicare), sex, primary language spoken at home, Aboriginality, disability status, postcode, adult reported height and weight of the child and 7 day recall of physical activity participation (11). Once registration data was submitted, the Active Kids voucher, valued up to AUD $100, was emailed to the adult. The Human Research Ethics Committee at University of Sydney granted approval for the evaluation of Active Kids (Project number: 2017/946).

Body Mass Index
Height and Weight was reported by the parent or carer of each child and adolescent at the point of voucher registration. This method of collecting height and weight is consistent with the NSW state-based surveillance (i.e. the NSW Population Health Survey). BMI was calculated as weight divided by height, squared (i.e., kg/m 2 ). Each child was categorised as thin, healthy weight, overweight or obese using the International Obesity Task Force (IOTF) definitions (13). The IOTF definitions provide a standard international definition for childhood overweight and obesity.

Physical activity
Meeting physical activity guidelines was assessed using a single item reported by the parent or carer (11). The item asked, "In a typical week, how many days was the child physically active for at least 60 minutes?" There is evidence that this is a valid and reliable measure of physical activity in adolescents (11).

Sport participation
Sport participation was determined using a single item reported by the parent or carer.
The item asked "Approximately, how many organised sessions of sport or physical activity has the child participated in, in total, outside of school hours, during the last 12 months?" Demographic characteristics Demographic characteristics included age, sex, primary language spoken at home, Aboriginality, disability status, SES, and remoteness. SES was determined using postcode of residence and categorised using the Socio-Economic Index for Area (SEIFA), specifically the Index of Relative Socio-Economic Disadvantage (14), which ranks regions in Australia according to relative socioeconomic disadvantage. Postcode-based SEIFA percentiles were converted into quartiles ranging from 1 (most disadvantaged area) to 4 (least disadvantaged area). Location was assessed using postcode of residence and categorised using the Accessibility and Remoteness Index of Australia (ARIA+). ARIA + groups areas on the basis of relative access to services into major city, inner regional, outer regional or remote (15).

Data analysis
Frequencies and proportions for demographic characteristics were calculated for height and weight and body mass index data from children and adolescents. Chi-squared tests were conducted to determine whether there were significant differences between those providing and not providing height and weight data. Multinomial regression models were conducted to determine which demographic characteristics were associated with an increased risk of overweight and obesity and examined the interactions between SES and each demographic characteristic. All analyses were performed in SAS Enterprise Guide 9.4 (SAS Institute, Cary, NC, USA).

Results
Of the initial 671,375 parents who registered their children or adolescents for the Active Kids program in 2018, 306,450 (45.7%) provided height and weight data. Of these 1.1% were outside the accepted International Obesity Taskforce range (13) and were excluded from the main analyses. Children who had valid height and weight data were significantly (p < 0.001) more likely to be older, boys, speak a language other than English at home, live in an area of higher SES, live in a major city, meet the physical activity guidelines, not identify as Aboriginal or Torres Strait Islander and not have a disability (Table 1).  Table 2). The majority spoke English at home (92%), were non-Indigenous (95%) and did not have a disability (97%). One in five participants met physical activity guidelines and less than half of the participants (43%) played sport at least twice a week. The overall sample prevalence of overweight and obesity was 17.2% and 7.6%, respectively.

Correlates of overweight or obesity
The odds of overweight and obesity for children and adolescents registered in the Active Kids program are displayed in Table 3. There was an inverse relationship between SES and overweight or obesity, with Active Kids participants living in the most disadvantaged area being 1.91 (95% CIs 1.88, 1.95) times more likely to be overweight or obese compared with children and adolescents living in the least disadvantaged area.

Interactions between socioeconomic status and demographic subgroups
Interactions were tested between SES and all demographic characteristics (i.e., age, sex, primary language spoken at home, Aboriginality, disability status, and remoteness).
Significant interactions were found between SES and age, Aboriginality and location (Table   4).
Within each age group, the predicted probability of overweight or obesity decreased as socioeconomic disadvantage increased. For example, the predicted probability of overweight or obesity for 9-11 year olds living in the most disadvantaged area was 39.4% compared with 23.6% living in the least disadvantaged area. Across all socioeconomic areas, children (4-11 year olds) had a significantly higher predicted probability of overweight or obesity compared with adolescents (12-18 year olds).
Within each SEIFA quartile, the predicted probability of overweight or obesity was Within location categories, the probability of overweight or obesity increased as socioeconomic disadvantaged increased. In the most disadvantaged area, children and adolescents living in major cities had a significantly higher probability of overweight or obesity (38.6% in major cities, compared with 31.3% in inner regional and 32.9% in outer regional and remote). Note. Interactions between socio-economic status and sex, primary language spoken at home, and disability status were not significant. While there was no difference in the prevalence of overweight and obesity between children living in different locations (i.e., major cities, regional and remote areas), the interaction between location and SES did show clear patterns. Within each SEIFA quartile, children living in major cities had higher levels of overweight or obesity. The greatest disparity for children living in major cities was seen in the most disadvantaged area,

Discussion
where the prevalence of overweight or obesity was 38%, compared with 30% for children living in inner regional areas, and 33% for children living in outer regional area. Children living in disadvantaged major cities in NSW could face a number of unique barriers to living a healthy lifestyle. Some of these barriers include land use issues (e.g., high access to energy-dense, nutrient poor foods and lack of supermarkets that supply fresh healthy food), infrastructure and maintenance issues (e.g., lack of sidewalks and street lighting), and social environment issues (e.g., high crime rates and the fear of crime) (22,23).
Overall, the patterns of overweight and obesity across subgroups were as expected and consistent with previous research. However, one exception to this was age 24 − 2 , as obesity declined with age among Active Kids registrants. The prevalence of obesity in 15- 18 year olds was half that observed among 4-8 year olds. This is initially counter-intuitive as other population studies show increases in obesity through adolescence (24). This is likely due to the differential self-selection effects in Active Kids sample. Young children are less selected, as all groups and all weight ranges participate in sport, especially swimming lessons. However, participation in physical activity and sport is known to decline with age (25), our older adolescent sub-group is likely to be represented by those adolescents who maintain sport and physical activity, and show selection effects in fewer of them being overweight or obese. There is potential to increase the representativeness of the mid-older adolescents by more focused targeting of Active Kids to those aged 15 years and older, especially focusing on those who have dropped out of organised sport.
The major strength of the Active Kids program and this study is the large sample of children and adolescents who participated in the program and provided height and weight data. This large sample (n = 239,433) allowed the examination of subgroup estimates of overweight and obesity with precision. There were also some limitations to this study which must be considered when interpreting the findings. First, as the Active Kids data was cross-sectional, no determination of causality between variables (e.g., physical activity) and overweight or obesity is implied. Second, this study could have underestimated the prevalence of overweight or obesity due to reporting bias by parents, social desirability bias and selection bias. Parents may mis-specify weight in adolescents and children where the true BMI values are in the upper end of the BMI distribution (26,27). However, parent-reported height and weight data is widely used in surveillance systems, such as the NSW SPANS (24), and has been shown to provide relatively accurate overweight and obesity estimates (insert Skinner reference). Selection bias may have occurred differentially, as this sample consisted of children and adolescents who registered for an Active Kids voucher, and especially older overweight or obese adolescents are less likely to participate in sport less likely to register for the voucher (28).

Conclusions
The Active Kids program successfully reached a substantial proportion of overweight and obese children across the NSW state population, including many from socially disadvantaged backgrounds (i.e., live in low SES areas and/or Aboriginal and Torres Strait Islander status). The program provided financial support and opportunities for these children to participate in organised physical activity and organised sport. As such, it contributes to the population health strategy in NSW around childhood obesity prevention (6). However, there is still potential to improve the reach of the program by further targeting socially disadvantaged groups, and older adolescents. The clear socio-economic gradient for obesity prevalence was maintained across age and other sub-groups, and warrants specific programmatic efforts. Continued evaluation of the Active Kids program provides policy-relevant information to guide future implementation of this program to increase physical activity and sport participation in children and adolescents on a large scale. No. 2017/946). By registering for the Active Kids program, the parent/guardian or carers provides consent for the Child's data to be used to evaluate the program.

Consent for publication
Yes.

Availability of data and material
The data that support the findings of this study are available from NSW Government, Office of Sport but are not publicly available.

Competing interests
The authors declare that they have no competing interests

Funding
The Active Kids program and evaluation was funded by the NSW Government, Office of Sport.
The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Authors' contributions KBO, BB, BCF, AB and LJR contributed to the study design. KBO conducted the data analysis and drafted the paper. KBO, BB, BCF, AB and LJR critically revised the paper and approved this version of the manuscript.