Four-Year Accelerometry Outcomes from a Cluster Randomized Whole of Systems Trial of Prevention Strategies for Childhood Obesity

Abstract

Background:

We investigated the effect of the Whole of Systems Trial of Prevention Strategies for Childhood Obesity (WHO STOPS) intervention on children's objectively measured physical activity and sedentary time (ST).

Methods:

We conducted a cluster randomized controlled trial with children in grades 4 (∼9–10 years old) and grade 6 (∼11–12 years old) from 10 communities in the Great South Coast region of Victoria, Australia. Communities were randomly allocated (1:1) to receive the WHO STOPS intervention in 2015. WHO STOPS was a whole of community systems-based approach to preventing childhood obesity. Outcome data were collected using a repeat cross-sectional design in 2015, 2017, and 2019. Children were asked to wear a hip-mounted accelerometer for 7 days. Age-specific Axis 1 activity counts were converted into duration (minutes/day) spent engaged in moderate-to-vigorous physical activity (MVPA), light physical activity (LPA), and ST. Linear mixed regression models were fitted to estimate the effects of the intervention on the three activity outcomes across the study period.

Results:

Analyses were based on valid accelerometer data from 1406 children (intervention n = 745; control n = 661). Results for MVPA, LPA, and ST were nonsignificant. Between 2015 and 2017, there were positive, but nonsignificant, changes in mean MVPA favoring intervention boys [3.7 minutes/day; 95% confidence interval (CI): −5.7 to 13.1] and girls (5.5 minutes/day; 95% CI: −1.5 to 12.6). By 2019, these effects had attenuated.

Conclusions:

Although the WHO STOPS intervention did not significantly change activity levels, the magnitudes of the effects on MVPA suggest that further research with whole-of-community interventions in larger samples would be worthwhile.

Clinical trial registration: Australian New Zealand Clinical Trials Registry (ANZCTR.org.au) identifier 12616000980437.

Background

The global problem of rising of childhood obesity rates^1,2 is due, in part, to increasing physical inactivity.³ Engaging in physical activity during childhood has positive health benefits (most notably, favorable cardiometabolic outcomes⁴) and can determine patterns of physical activity into adulthood,⁵ with consequential reductions in preventable pathology and premature death.^6,7 Trends toward higher levels of physical inactivity,^8,9 poorer physical fitness,¹⁰ and higher rates of overweight and obesity^1,2 among children across the world suggest that many are not enjoying the health benefits of being active. Childhood interventions are needed to avert the considerable health^4,6 and economic¹¹ costs of physical inactivity across the lifespan.

Whole-of-community or community-based interventions (CBIs) involve using community engagement processes and implementing multiple strategies at the population level to improve health outcomes.¹² CBIs with a focus on preventing excessive weight gain in children and adolescents typically target both nutrition and physical activity, and have resulted in a modest reduction in BMI z-scores (mean difference = −0.07, 95% uncertainty interval: −0.13 to −0.01, nine trials).¹³ Less is known, however, about the impact of these interventions on children's physical activity levels,^14–16 not least because the use of objective measures of physical activity in evaluations is rare.¹⁷ In one exception—the 2-year, community-based APPLE (A Pilot Program for Lifestyle and Exercise) Project—there was a significant intervention effect for accelerometer counts in favor of children in the intervention condition at 1 year, but there was no difference between the intervention and control conditions at 2 years.¹⁷ Insufficient wear time to obtain reliable estimates¹⁸ may be a limiting factor in interpreting these findings [accelerometers were worn for 1 (74%) or 2 (26%) days at baseline and for 2 (68%) or 5 (32%) days at 1 and 2 years].¹⁷

It is also unclear whether CBIs are equally effective for girls and boys. Girls are generally less physically active than boys.^19–22 Among children 5–17 years of age, for example, pedometer data from the Australian 2011–2012 National Nutrition and Physical Activity Survey showed that 8% of girls compared with 25% of boys were meeting the recommended 12,000 steps per day.²³ A range of factors influence girls less favorably than boys (e.g., parental support for physical activity, sport club participation,²⁰ gendered stereotypes around girls and physical activity,²⁴ lower self-perceived competence in physical education²⁰). The gender gap in physical activity levels and the different influences on the participation of girls and boys mean that intervention effects need to be examined by gender.

The Whole of Systems Trial of Prevention Strategies for Childhood Obesity (WHO STOPS) was planned as a stepped-wedge cluster randomized trial of a whole of community systems-based approach to preventing childhood obesity in the Great South Coast region of Victoria, Australia.²⁵ We have previously reported that the intervention was successful in reducing takeaway food consumption and improving health-related quality of life among primary school children.²⁶ In this study, we present findings on the effect of the intervention on objectively measured physical activity of a subsample of primary school children from the main trial.

Methods

The reporting of this study followed the Consolidated Standards of Reporting Trials extension to cluster randomized trials guidelines.²⁷

Design

A detailed description of the WHO STOPS trial method is available from the published protocol.²⁵ Following the baseline measurements (April to June 2015), 10 communities were randomized to begin the intervention in late 2015 (Step 1, 5 communities) and 2017 (Step 2, 5 communities). Step 1 communities were engaged, and the intervention design was maintained across 4 years as per the protocol. Due to challenges (including severe bushfire, higher-than-usual staff turnover in partner organizations, and unexpected shifts in priorities in partner organizations), the intervention was not able to be initiated within Step 2 communities until 2019 (a 2-year delay). We refer to Step 1 communities as the intervention condition and Step 2 communities as the control condition.

Participants

All 70 primary schools (government, Catholic, and other independent) within the 10 communities involved in the study in South West Victoria, Australia, were invited to participate in 2015, 2017, and 2019. Students at participating schools were enrolled unless they opted out verbally or returned an opt-out form, signed by their parents or guardians, on the day of data collection. An exception to this procedure was Catholic school children who, in 2015, participated through an opt-in consent procedure; their data were not included in the analysis for 2015 because our previous research has demonstrated that opt-in recruitment methods underestimate population-level overweight/obesity by 5 percentage points.²⁸ Data of Catholic school children were included in 2017 and 2019 when an opt-out approach was subsequently approved. Children were eligible to participate in the accelerometry component of this study if they: (1) were in grades four or six; (2) were available on the day of data collection at their schools; (3) had not opted out; and (4) were willing to wear an accelerometer 7 consecutive days.

Our analyses focus on a subsample of children from whom objective physical activity data were collected. At each participating school, all children in the first class in each year level who were approached on the day for measurements (e.g., 4A and 6C) were invited to wear accelerometers. This randomization strategy was considered less disruptive than randomizing students (e.g., every second girl or boy gets an accelerometer). Furthermore, children would not feel discriminated against if there were not selected to wear the device, because the approach was seen as a whole class activity.

Intervention

The WHO STOPS intervention is a five-phase process.^25,26 The intervention had in-built flexibility such that communities could move through the phases at their own pace and tailor intervention actions toward local community context and priorities.

The first phase involved collecting baseline data and using the information derived for raising the awareness of childhood obesity and for engaging and recruiting community leaders who had interest in, and influence on, child health (e.g., leaders from government health and education departments, health services, local councils, businesses).

The second phase focused on identifying and collaborating with community leaders who had deep knowledge of the contexts of the intervention, who had the authority to initiate action, and who set the boundaries for the intervention. Through a facilitated workshop (Workshop 1), community leaders used Systems Thinking in Community Knowledge Exchange (STICKE) software²⁹ to build a causal loop diagram of the causes of childhood obesity in their community.³⁰

The third phase involved presenting the diagram back to the original group and soliciting further feedback in a second facilitated workshop (Workshop 2). The group also had the opportunity to invite additional stakeholders to the second workshop, such as representatives from retailers, schools, health organizations, and leading community groups. Workshop 2 finished with a transition to action, with initial planning for inviting the broader community to a forum for action planning.

The fourth phase focused on designing actions to prevent childhood obesity that could be implemented across the community. The causal loop diagram and an evidence brief on obesity prevention, which included case studies of previous successful interventions, informed the work of the broader group during Workshop 3. Actions implemented within each community were recorded in actions registers (e.g., a local government area constructing a new footpath to encourage active transport to and from school, junior sporting associations implementing a water-only policy, and a primary school erecting signs encouraging children to be dropped off away from the school gate to promote walking the final distance to school). A detailed list of the actions undertaken in one community has been published previously.³¹

The fifth phase involved ongoing data collection and ongoing community engagement efforts to promote further implementations of actions and to stimulate new ideas in a cycle of constructive capacity building.

At the 2017 monitoring time point, one implementation community had completed all five phases, with the remaining four implementation communities preparing for phase three. At the 2019 monitoring time point, all implementation communities were engaged in the fifth phase.

Measurements

Data collection occurred from April to June of 2015, 2017, and 2019. Moderate-to-vigorous physical activity (MVPA), light physical activity (LPA), and sedentary time (ST) were measured using ActiGraph wGT3X-BT accelerometers (ActiGraph, Pensacola, FL). Each participant was instructed to wear an accelerometer for 7 consecutive days over the right hip using an elasticized belt while awake and to remove the device for water-based activities (e.g., bathing, swimming) and sparring and high-contact sports (e.g., boxing, rugby, martial arts). On completion of the 7 days, participants were asked to return the accelerometers to their classroom teachers for researcher collection. Accelerometry data were processed using ActiLife software (version 6; ActiGraph). Nonwear time was defined as periods of ≥60 minutes consecutive zero counts, with a 1–2 minutes allowance for counts between 0 and 100.³² Wear time was calculated subtracting nonwear time from 24 hours. A valid day was defined as ≥500 minutes/day of wear time, and only participants with a minimum of 3 days of valid data were included in the analyses.³³

Reliable estimates of children's physical activity have been achieved using a minimum of 2 days of data when each day involved ≥420 minutes of wear time, with the reliability of estimates, including weekend data comparable to those with no weekend wear.¹⁸ Age-specific cut points for counts per 60-second epochs were used to examine time spent in MVPA, LPA, and ST based on vertical axis (Axis 1) data.^34,35 A binary variable indicating whether a child met the physical activity component of Australia's 24-hour movement guidelines (≥60 minutes/day of MVPA)³⁶ was defined. The component was operationalized using the average across all days method, whereby a child is deemed compliant with the guidelines if average MVPA is ≥60 minutes/day across the days, during which physical activity was measured.³⁷

Height and weight were measured to generate age- and sex-specific BMI z-scores. Trained health professionals and researchers measured each child in private booths. Children wore light clothing and removed their shoes. Height was measured to the nearest 0.1 cm using a portable stadiometer (Charder HM-200P Portstad; Charder Electronic Co. Ltd., Taichung City, Taiwan). Weight was measured to the nearest 0.1 kg using an electronic weight scale (A&D Precision Scale UC-321; A7D Medical, San Jose, CA). Height and weight were each measured twice, with a third measurement taken if there was a discrepancy of >0.5 cm for height or >0.1 kg for weight in the initial two measurements. Height and weight measurements were averaged for each child and transformed to BMI z-scores using the World Health Organization's child growth reference.³⁸

Demographic data obtained through survey were gender, birthdate, Aboriginal and/or Torres Strait Islander background, and language spoken at home. School type was recorded (government, Catholic, and other independent schools). Socioeconomic background was measured at the school level using the Index of Community Socio-Educational Advantage (ICSEA).³⁹ ICSEA is a measure of the average educational advantage of a school's student population compared with other schools. The measure is based on student factors (parental occupation and education) and school factors (geographic location and proportion of Indigenous students).³⁹ ICSEA has a median value of 1000. The values were obtained from the Australian Curriculum, Assessment, and Reporting Authority's My School website.⁴⁰

Analyses

Sample size calculations were undertaken for the main trial outcome (BMI) z-scores^25,26 and not the secondary analyses reported in this study. Given the differing participation rates in physical activity between boys and girls, the data were analyzed by gender. Data from 10 students who preferred not to state their gender, 2 with valid accelerometry data, were excluded from the analyses. Unfortunately, the sample of two children who reported nonbinary gender was insufficiently large to analyze separately. Demographic characteristics and mean BMI z-score were compared between the subsets of students with and without accelerometry data to assess the representativity of the subsample. Linear mixed models were fitted to assess the effects of the intervention on mean daily minutes of MVPA, LPA, and ST. Logistic mixed models were fitted to assess the effect of the intervention on meeting the physical activity component of Australia's 24-hour movement guidelines.

All models included condition (intervention or control), wave (2015, 2017, 2019), condition by wave interaction, ICSEA (in tertiles), school type (government, Catholic, or other independent), school rurality (inner regional or outer regional), age (continuous variable), and average daily minutes of accelerometer wear time. All models included school as a random effect to account for within-school clustering. For each outcome and gender, we reported two contrasts: difference between intervention and control conditions in mean changes between (1) baseline (2015) and 2017, and (2) baseline (2015) and 2019. All analyses were conducted in Stata version 16.1.⁴¹

Results

Valid accelerometer data were obtained from 1406 children (intervention n = 745; control n = 661); there were 420 children who received accelerometers but did not provide valid data (Supplementary Table S1). Compared with intervention children who were not given accelerometers and those without valid data, those with valid accelerometer data were significantly more likely to be girls, and significantly less likely to be grade six students, students from outer regional schools, and Aboriginal and Torres Strait Islanders, and have a significantly lower average age (Table 1). Control condition children with valid accelerometer data were significantly more likely to be grade six students and students from government schools, have a significantly lower average age, attend schools with greater ICSEA values, and have lower average BMI z-scores than those without valid accelerometer data.

Table 1.

Observed Demographic Characteristics and BMI z-Scores of 2015, 2017, and 2019 Children With (N = 1406) and Without (N = 2802) Accelerometry Data

	Accelerometry sample							Accelerometer not given^a or data invalid^b						With vs. without accelerometry data
	Intervention			Control			Intervention vs. control	Intervention			Control			Intervention	Control
	N	n	%	N	n	%	p	N	n	%	N	n	%	p	p
Girls	745	393	52.8	661	347	52.5	0.924	1153	523	45.4	1649	796	48.3	0.002	0.066
Grade six	745	330	44.3	661	286	43.3	0.698	1153	586	50.8	1649	810	49.1	0.005	0.011
Government school	745	584	78.4	661	494	74.7	0.106	1153	932	80.8	1649	1319	80.0	0.195	0.006
Outer regional school	745	133	17.9	661	80	12.1	0.003	1153	257	22.3	1649	241	14.6	0.020	0.115
Aboriginal and/or Torres Strait Islander	675	82	12.1	620	95	15.3	0.097	1057	181	17.1	1544	212	13.7	0.005	0.337
Speak language other than English at home	727	32	4.4	658	38	5.8	0.244	1121	65	5.8	1634	107	6.5	0.188	0.492

	N	Mean	SD	N	Mean	SD		N	Mean	SD	N	Mean	SD	p	p
Age (years)	745	10.7	1.1	661	10.8	1.1	0.332	1153	10.9	1.1	1649	10.9	1.1	<0.001	0.005
ICSEA	745	1000	45	661	1012	41	<0.001	1153	998	42	1649	1003	47	0.367	<0.001
BMIz	740	0.63	1.19	653	0.55	1.10	0.235	1112	0.68	1.19	1594	0.67	1.18	0.307	0.032

Reasons accelerometry not given are one of the following: not being in a class invited to receive an accelerometer, or child not consenting.

Reasons for having invalid accelerometry data are one of the following: failing to meet wear time criteria of >500 minutes/day over ≥3 days (N = 417), having missing data on a key covariate (e.g., days worn, n = 2), or having an average activity observation outside feasible bounds (n = 1).

BMIz, BMI z-scores; ICSEA, Index of Community Socio-Educational Advantage; SD, standard deviation.

Within the subset of children with valid accelerometry data, there were significant differences in demographic characteristics between the intervention and control condition children. The intervention condition comprised a significantly higher percentage of students from outer regional schools and students from schools with lower ICSEA values than the control condition.

The intervention effects (difference in change) were not significant for MVPA, LPA, and ST in boys and girls (Supplementary Fig. S1a–d). Mean changes in MVPA between 2015 and 2017 were greater for intervention boys [3.7 minutes/day; 95% confidence interval (CI): −5.7 to 13.1] and girls (5.5 minutes/day; 95% CI: −1.5 to 12.6), although the differences were nonsignificant (Table 2). By 2019, these differences were even less marked. The intervention effects between 2015 and 2019 for the percentages of children meeting the MVPA component of the 24-hour movement guideline were not significant, but favored control boys over intervention boys (−9.3 percentage points; 95% CI: −25.4 to 6.7) and intervention girls over control girls (10.3 percentage points; 95% CI: −8.6 to 29.2).

Table 2.

Adjusted Model Estimates for Mean Accelerometry Activity Minutes per Day and Percentage Meeting Moderate-to-Vigorous Physical Activity Guidelines for Boys (N = 666) and Girls (N = 740)

			MVPA (mean minutes/day)	LPA (mean minutes/day)	ST (mean minutes/day)	Met MVPA guidelines (%)
	Wave	N	Estimate (95% CI)	Estimate (95% CI)	Estimate (95% CI)	Estimate (95% CI)
Boys
Intervention	2015	81	77.8 (72.6 to 83.0)	163.8 (156.8 to 170.7)	529.7 (516.9 to 542.6)	78.3 (70.1 to 86.5)
	2017	139	79.2 (75.3 to 83.1)	160.0 (154.4 to 165.6)	531.4 (521.6 to 541.3)	77.2 (70.5 to 83.9)
	2019	132	74.9 (70.8 to 79.0)	155.7 (149.8 to 161.6)	541.0 (530.5 to 551.4)	64.3 (55.9 to 72.6)
Control	2015	66	82.8 (77.1 to 88.5)	158.3 (150.5 to 166.0)	528.6 (514.3 to 542.9)	81.6 (73.0 to 90.3)
	2017	125	80.5 (76.4 to 84.7)	158.0 (151.9 to 164.1)	540.0 (529.3 to 550.6)	75.0 (67.2 to 82.7)
	2019	123	77.9 (73.6 to 82.1)	159.6 (153.3 to 166.0)	537.1 (526.1 to 548.2)	76.9 (69.3 to 84.6)

	Contrast	Estimate (95% CI)	P	Estimate (95% CI)	p	Estimate (95% CI)	p	Estimate (95% CI)	p
Within condition change (Intervention)	2017 vs. 2015	1.4 (−5.0 to 7.8)	0.669	−3.8 (−11.6 to 4.0)	0.344	1.7 (−13.7 to 17.1)	0.829	−1.1 (−11.5 to 9.4)	0.839
Within condition change (Intervention)	2019 vs. 2015	−2.9 (−9.6 to 3.8)	0.392	−8.1 (−16.2 to 0.1)	0.053	11.2 (−4.9 to 27.4)	0.172	−14.0 (−25.8 to −2.3)	0.019
Within condition change (Control)	2017 vs. 2015	−2.3 (−9.5 to 4.9)	0.529	−0.3 (−9.0 to 8.5)	0.954	11.4 (−5.8 to 28.6)	0.195	−6.7 (−18.4 to 5.1)	0.266
Within condition change (Control)	2019 vs. 2015	−5.0 (−12.2 to 2.2)	0.177	1.4 (−7.7 to 10.4)	0.768	8.5 (−9.0 to 26.1)	0.340	−4.7 (−16.3 to 6.9)	0.430
Difference in change (Intervention − Control)	2017 vs. 2015	3.7 (−5.7 to 13.1)	0.442	−3.5 (−15.1 to 8.1)	0.553	−9.7 (−32.5 to 13.1)	0.404	5.6 (−9.8 to 21.0)	0.478
Difference in change (Intervention − Control)	2019 vs. 2015	2.0 (−7.5 to 11.6)	0.675	−9.4 (−21.4 to 2.6)	0.125	2.7 (−20.6 to 26.0)	0.821	−9.3 (−25.4 to 6.7)	0.255

			MVPA (mean minutes/day)	LPA (mean minutes/day)	ST (mean minutes/day)	Met MVPA guidelines (%)
	Wave	N	Estimate (95% CI)	Estimate (95% CI)	Estimate (95% CI)	Estimate (95% CI)
Girls
Intervention	2015	93	65.2 (61.1 to 69.3)	155.2 (149.5 to 160.8)	528.2 (520.0 to 536.4)	55.5 (44.5 to 66.6)
	2017	152	70.0 (66.6 to 73.3)	158.6 (154.3 to 163.0)	520.1 (513.8 to 526.4)	65.7 (56.9 to 74.5)
	2019	148	65.6 (62.2 to 69.0)	158.2 (153.7 to 162.6)	524.9 (518.4 to 531.4)	61.5 (52.7 to 70.3)
Control	2015	91	66.4 (62.0 to 70.7)	150.5 (144.7 to 156.3)	531.7 (523.3 to 540.1)	59.1 (47.7 to 70.5)
	2017	127	65.6 (62.0 to 69.3)	156.7 (151.9 to 161.5)	526.6 (519.6 to 533.5)	59.5 (49.8 to 69.1)
	2019	129	62.4 (58.6 to 66.2)	150.9 (145.9 to 156.0)	535.7 (528.4 to 542.9)	54.8 (44.6 to 64.9)

	Contrast	Estimate (95% CI)	P	Estimate (95% CI)	p	Estimate (95% CI)	p	Estimate (95% CI)	p
Within condition change (Intervention)	2017 vs. 2015	4.8 (−0.1 to 9.6)	0.054	3.5 (−3.6 to 10.5)	0.339	−8.1 (−18.4 to 2.2)	0.124	10.2 (−2.4 to 22.7)	0.112
Within condition change (Intervention)	2019 vs. 2015	0.4 (−4.6 to 5.3)	0.880	3.0 (−4.3 to 10.2)	0.421	−3.3 (−13.8 to 7.2)	0.541	6.0 (−6.7 to 18.7)	0.358
Within condition change (Control)	2017 vs. 2015	−0.7 (−6.0 to 4.5)	0.780	6.2 (−1.4 to 13.8)	0.107	−5.1 (−16.1 to 5.9)	0.362	0.3 (−13.0 to 13.7)	0.960
Within condition change (Control)	2019 vs. 2015	−4.0 (−9.5 to 1.6)	0.162	0.5 (−7.4 to 8.3)	0.909	4.0 (−7.5 to 15.4)	0.497	−4.3 (−18.6 to 10.0)	0.553
Difference in change (Intervention − Control)	2017 vs. 2015	5.5 (−1.5 to 12.6)	0.125	−2.8 (−12.9 to 7.3)	0.591	−3.0 (−17.6 to 11.7)	0.693	9.8 (−8.2 to 27.9)	0.286
Difference in change (Intervention − Control)	2019 vs. 2015	4.3 (−3.0 to 11.7)	0.247	2.5 (−7.9 to 12.9)	0.637	−7.2 (−22.4 to 7.9)	0.348	10.3 (−8.6 to 29.2)	0.287

Six linear (mean minutes/day) and two logistic (% meeting guidelines) mixed models were fitted for boys and girls separately, school as random effect. All models include wave, condition, wave × condition, daily wear time, age, ICSEA tertile, school rurality (inner regional, outer regional), and school type (government, Catholic, other independent).

CI, confidence interval; LPA, light physical activity; MVPA, moderate-to-vigorous physical activity; ST, sedentary time.

In relation to mean LPA, for boys, there were larger nonsignificant decreases in the intervention condition than in the control condition between 2015 and 2017 (−3.5 minutes/day; 95% CI: −15.1 to 8.1) and between 2015 and 2019 (−9.4 minutes/day; 95% CI: −21.4 to 2.6). For girls, there was a small nonsignificant decrease in mean LPA in the intervention condition as compared with the control condition between 2015 and 2017 (−2.8 minutes/day; 95% CI: −12.9 to 7.3) and a small nonsignificant increase between 2015 and 2019 (2.5 minutes/day; 95% CI: −7.9 to 12.9).

Mean ST increased in both intervention and control conditions between 2015 and 2017 for boys, but to a greater extent in the control condition, contributing to a nonsignificant difference in change (−9.7 minutes/day; 95% CI: −32.5 to 13.1). There were similarly nonsignificant intervention effects among girls in the intervention and control conditions between 2015 and 2019 (−7.2 minutes/day; 95% CI: −22.4 to 7.9).

Discussion

The WHO STOPS intervention did not improve children's physical activity levels nor reduce their time spent sedentary. Some of the results seem promising, however, especially for girls.

We observed a (nonsignificant) 10.3 percentage point increase in girls meeting the MVPA component of the 24-hour movement guideline³⁶ in favor of the intervention condition. This change was reflected in the (nonsignificant) changes in mean MVPA of 5.5 minutes/day between 2015 and 2017 favoring intervention girls. Four years from baseline, in 2019, girls in intervention condition had retained most of these gains.

For both boys and girls, the magnitudes of change in MVPA mirrored a meta-analysis finding of a ∼4 minutes/day increase in objectively measured MVPA (25 studies) observed for children in interventions of 4–140 weeks (median = 26 weeks).⁴² Considered against the yearly average decline in MVPA across childhood and adolescence (3–18 years) of 3.4 minutes/day,⁴³ changes in MVPA of the magnitude shown in the WHO STOPS trial may be effective for preserving physical activity participation in children. The findings from WHO STOPS suggest these effects could be sustained over years rather than weeks or months, although the findings of this study need replication with much larger samples of children.

Specific intervention activities in WHO STOPS that may have increased MVPA were: creation of active transport drop off zones 800 m from school, development of a bike bus (where students wait at stops around town to be met to and ride in a group to school), implementation of park runs in community, walking group/running club development, bike safety education classes, walk to school programs, free outdoor pool access, and environmental (e.g., foot path) modifications within specific communities. Not all communities adopted each of these actions, and further research is currently underway to examine which intervention strategies, or combination of strategies, are likely to have the most impact on children's health.

Strengths and Limitations

Key strengths of our study were the long follow-up (4 years), in comparison to other contemporary community-based childhood obesity prevention interventions, which have typically been 1–2 years in duration,⁴⁴ use of accelerometers to objectively measure physical activity, and high child participation rates (>80% using an opt-out approach, as compared with school student participation rates of 30%–60% when opt-in procedures are used⁴⁵). The main limitation of this analysis was that the study was not designed or powered to detect changes in objectively measured physical activity or ST. The number of accelerometers available to the research (n = 400 in 2015) and the prohibitive cost of purchasing more devices (>$US220 each) limited the potential size of this subsample.

We encourage future CBIs to incorporate objective measures of physical activity and ST to evaluate trial outcomes, particularly as lower cost research-grade accelerometers emerge (<$US30 each).⁴⁶ Other limitations included the variable lengths of time it took for communities to move through the phases of the intervention (meaning that the actions initiated in some communities had a longer time to influence children's behaviors than others) and differences between communities with respect to what actions were implemented (reflecting variations in community priorities, resources, and capacities to engage).

Conclusion

WHO STOPS did not have an impact on children's activity levels or their time spent being sedentary. The magnitude of change in children's MVPA was similar to that observed for children's physical activity studies, however. If these findings were replicated in larger samples of children, whole-of-community approaches to addressing child obesity are likely to be an effective way of attenuating the declines in MVPA across childhood and adolescence. More work is required to understand how to increase and sustain children's physical activity levels through community interventions.

Acknowledgment

Aspects of this article were presented at the Proceedings of the 8th International Society for Physical Activity and Health Congress in 2021.

Impact Statement

This study highlights that whole-of-community systems interventions to prevent obesity among primary school children (∼4–12 years) are likely to have favorable impacts on physical activity, however, further research with larger participant samples is required.

Ethics Approval and Consent to Participate

Full ethics clearances have been received from: Deakin University's Human Research Ethics Committee (DU-HREC) 2014-279, DU-HREC 2013-095, Deakin University's Human Ethics Advisory Group-Health (HEAG-H) HEAG-H 194_2014, HEAG-H 17 2015, HEAG-H 155_2014), the Victorian Department of Education and Training 2015_002622, 2013_002013, and the Catholic Archdiocese of Melbourne, Sale, Sandhurst, and Ballarat. An opt-out procedure was used whereby students at participating schools were enrolled unless they opted out verbally or returned an opt-out form, signed by their parents or guardians, on the day of data collection.

Availability of Data and Materials

The datasets generated during and/or analyzed during the current study are not publicly available due to data privacy issues and nature upon which the data were collected under an opt-out approach.

Footnotes

Authors' Contributions

S.A., C.S., L.O., and C.B. conceived the trial design and data collection for the whole trial. C.S., N.C., P.F., K.A.B., and L.O. monitored data collection for the whole trial. D.B. and L.O. wrote the statistical analysis plan and cleaned and analyzed the data. S.A., N.C., K.A.B., P.F., A.D.B., H.L., and C.S. supported communities to implement the trial. C.J.G. and C.S. led the writing of this article. All authors contributed to interpretation of results and drafting and revision of the article and approved the final article.

Disclaimer

The opinion and analysis in this article are those of the author(s) and are not those of the Department of Health and Human Services, the Victorian Government, the Secretary of the Department of Health and Human Services, or the Victorian Minister for Health.

Funding Information

This article reports on research funded by the Australian National Health and Medical Research Council (NHMRC) Partnership Project (GNT111411) and the Western Alliance. S.A. was supported during this time period by funding from an Australian NHMRC/Australian National Heart Foundation Career Development Fellowship (GNT1045836). C.S., N.C., K.A.B., A.D.B., C.B., and S.A. during this time period were researchers within a NHMRC Center for Research Excellence in Obesity Policy and Food Systems (GNT1041020). S.A. during this time period was also a researcher on a US National Institutes of Health grant titled Systems Science to Guide Whole-of-Community Childhood Obesity Interventions (1R01HL115485-01A1) and is an investigator on the EU-funded Confronting Obesity: Co-creating policy with youth (H2020-SFS-2016-2017). S.A. is also supported by the NHMRC Center of Research Excellence in Food Retail Environments for Health (RE358 FRESH). These funders had no role in the design of the study, collection, analyses, and interpretation of the data or involvement in writing the article.

Author Disclosure Statement

The authors declare that they have no competing interests.

Supplementary Material

References

Abarca-Gómez

, Abdeen

, Hamid

, et al. Worldwide trends in body-mass index, underweight, overweight, and obesity from 1975 to 2016: A pooled analysis of 2416 population-based measurement studies in 128.9 million children, adolescents, and adults. Lancet, 2017; 390(10113):2627–2642; doi: 10.1016/S0140-6736(17)32129-3.

, Fleming

, Robinson

, et al. Global, regional, and national prevalence of overweight and obesity in children and adults during 1980–2013: A systematic analysis for the Global Burden of Disease Study 2013. Lancet, 2014; 384(9945):766–781; doi: 10.1016/S0140-6736(14)60460-8.

Hills

, Andersen

, Byrne

. Physical activity and obesity in children. Br J Sports Med, 2011; 45(11):866–870; doi: 10.1136/bjsports-2011-090199.

Poitras

, Gray

, Borghese

, et al. Systematic review of the relationships between objectively measured physical activity and health indicators in school-aged children and youth. Appl Physiol Nutr Metab, 2016; 41(6 Suppl 3):S197–S239; doi: 10.1139/apnm-2015-0663.

Telama

, Yang

, Viikari

, et al. Physical activity from childhood to adulthood: A 21-year tracking study. Am J Prev Med, 2005; 28(3):267–273; doi: 10.1016/j.amepre.2004.12.003.

Warburton

DER

, Nicol

, Bredin

SSD

. Health benefits of physical activity: The evidence. CMAJ, 2006; 174(6):801–809; doi: 10.1503/cmaj.051351.

Lavie

, Ozemek

, Carbone

, et al. Sedentary behavior, exercise, and cardiovascular health. Circ Res, 2019; 124(5):799–815; doi: 10.1161/CIRCRESAHA.118.312669.

Gomes

, Katzmarzyk

, Hedeker

, et al. Correlates of compliance with recommended levels of physical activity in children. Sci Rep, 2017; 7(1):16507; doi: 10.1038/s41598-017-16525-9.

Mark

, Joel

, Silvia

, et al. Global Matrix 2.0: Report card grades on the physical activity of children and youth comparing 38 countries. J Phys Act Health, 2016; 13(Suppl 2):S343–S366; doi: 10.1123/jpah.2016-0594.

10.

Tomkinson

, Olds

. Secular changes in pediatric aerobic fitness test performance: The global picture. Med Sport Sci, 2007; 50:46–66; doi: 10.1159/000101075.

11.

Lee

, Adam

, Zenkov

, et al. Modeling the economic and health impact of increasing children's physical activity in the United States. Health Affairs, 2017; 36(5):902–908; doi: 10.1377/hlthaff.2016.1315.

12.

Wolfenden

, Wyse

, Nichols

, et al. A systematic review and meta-analysis of whole of community interventions to prevent excessive population weight gain. Prev Med, 2014; 62:193–200; doi: 10.1016/j.ypmed.2014.01.031.

13.

Ananthapavan

, Nguyen

, Bowe

, et al. Cost-effectiveness of community-based childhood obesity prevention interventions in Australia. Int J Obes, 2019; 43(5):1102–1112; doi: 10.1038/s41366-019-0341-0.

14.

de Silva-Sanigorski

, Bell

, Kremer

, et al. Reducing obesity in early childhood: Results from Romp & Chomp, an Australian community-wide intervention program. Am J Clin Nutr, 2010; 91(4):831–840; doi: 10.3945/ajcn.2009.28826.

15.

Folta

, Kuder

, Goldberg

, et al. Changes in diet and physical activity resulting from the Shape Up Somerville community intervention. BMC Pediatr, 2013; 13:157; doi: 10.1186/1471-2431-13-157.

16.

Utter

, Scragg

, Robinson

, et al. Evaluation of the Living 4 Life project: A youth-led, school-based obesity prevention study. Obes Rev, 2011; 12(Suppl 2):51–60; doi: 10.1111/j.1467-789X.2011.00905.x.

17.

Taylor

, McAuley

, Barbezat

, et al. APPLE Project: 2-y Findings of a community-based obesity prevention program in primary school age children. Am J Clin Nutr, 2007; 86(3):735–742; doi: 10.1093/ajcn/86.3.735.

18.

Rich

, Geraci

, Griffiths

, et al. Quality control methods in accelerometer data processing: Defining minimum wear time. PLoS One, 2013; 8(6):e67206; doi: 10.1371/journal.pone.0067206.

19.

Pearce

, Basterfield

, Mann

, et al. Early predictors of objectively measured physical activity and sedentary behaviour in 8–10 year old children: The Gateshead Millennium Study. PLoS One, 2012; 7(6):e37975; doi: 10.1371/journal.pone.0037975.

20.

Telford

, Telford

, Olive

, et al. Why are girls less physically active than boys? Findings from the LOOK longitudinal study. PLoS One, 2016; 11(3):e0150041; doi: 10.1371/journal.pone.0150041.

21.

Hallal

, Andersen

, Bull

, et al. Global physical activity levels: Surveillance progress, pitfalls, and prospects. Lancet, 2012; 380(9838):247–257; doi: 10.1016/S0140-6736(12)60646-1.

22.

Strugnell

, Turner

, Malakellis

, et al. Composition of objectively measured physical activity and sedentary behaviour participation across the school-day, influence of gender and weight status: Cross-sectional analyses among disadvantaged Victorian school children. BMJ Open, 2016; 6(9):e011478; doi: 10.1136/bmjopen-2016-011478.

23.

Australian Bureau of Statistics. Australian Health Survey: Physical Activity, 2011–2012 Financial Year 2013. Available from: https://www.abs.gov.au/statistics/health/health-conditions-and-risks/australian-health-survey-physical-activity/2011-12 [Last accessed: April 30, 2021].

24.

Schmalz

, Kerstetter

. Girlie girls and manly men: Chidren's stigma consciousness of gender in sports and physical activities. J Leis Res, 2006; 38(4):536–557; doi: 10.1080/00222216.2006.11950091.

25.

Allender

, Millar

, Hovmand

, et al. Whole of Systems Trial of Prevention Strategies for Childhood Obesity: WHO STOPS childhood obesity. Int J Environ Res Public Health, 2016; 13(11):1143; doi: 10.3390/ijerph13111143.

26.

Allender

, Orellana

, Crooks

, et al. Four-year behavioral, health-related quality of life, and BMI outcomes from a cluster randomized Whole of Systems Trial of Prevention Strategies for Childhood Obesity. Obesity (Silver Spring), 2021; 29(6):1022–1035; doi: 10.1002/oby.23130.

27.

Campbell

, Piaggio

, Elbourne

, et al. Consort 2010 statement: Extension to cluster randomised trials. BMJ, 2012; 345(7881):e5661; doi: 10.1136/bmj.e5661.

28.

Strugnell

, Orellana

, Hayward

, et al. Active (opt-in) consent underestimates mean BMI-z and the prevalence of overweight and obesity compared to passive (opt-out) consent. Evidence from the Healthy Together Victoria and Childhood Obesity Study. Int J Environ Res Public Health, 2018; 15(4):747; doi: 10.3390/ijerph15040747.

29.

Hayward

, Morton

, Johnstone

, et al. Tools and analytic techniques to synthesise community knowledge in CBPR using computer-mediated participatory system modelling. NPJ Digit Med, 2020; 3(1):22; doi: 10.1038/s41746-020-0230-x.

30.

Allender

, Owen

, Kuhlberg

, et al. A community based systems diagram of obesity causes. PLoS One, 2015; 10(7):e0129683; doi: 10.1371/journal.pone.0129683.

31.

Bolton

, Fraser

, Lowe

, et al. Generating change through collective impact and systems science for childhood obesity prevention: The GenR8 Change case study. PLoS One, 2022; 17(5):e0266654; doi: 10.1371/journal.pone.0266654.

32.

Troiano

, Berrigan

, Dodd

, et al. Physical activity in the United States measured by accelerometer. Med Sci Sports Exerc, 2008; 40(1):181–188; doi: 10.1249/mss.0b013e31815a51b3.

33.

Steele

, van Sluijs

EMF

, Sharp

, et al. An investigation of patterns of children's sedentary and vigorous physical activity throughout the week. Int J Behav Nutr Phys Act, 2010; 7(1):88; doi: 10.1186/1479-5868-7-88.

34.

Freedson

, Pober

, Janz

. Calibration of accelerometer output for children. Med Sci Sports Exerc, 2005; 37(11 Suppl):S523–S530; doi: 10.1249/01.mss.0000185658.28284.ba.

35.

Trost

, Loprinzi

, Moore

, et al. Comparison of accelerometer cut points for predicting activity intensity in youth. Med Sci Sports Exerc, 2011; 43(7):1360–1368; doi: 10.1249/mss.0b013e318206476e.

36.

Department of Health. Australian 24-hour movement guidelines for children (5–12 years) and young people (13–17 years) —An integration of physical activity, sedentary behaviour and sleep. Australian Government: Canberra; 2019.

37.

Olds

, Ridley

, Wake

, et al. How should activity guidelines for young people be operationalised?. Int J Behav Nutr Phys Act, 2007; 4:43; doi: 10.1186/1479-5868-4-43.

38.

de Onis

, Onyango

, Borghi

, et al. Development of a WHO growth reference for school-aged children and adolescents. Bull World Health Organ, 2007; 85(9):660–667; doi: 10.2471/blt.07.043497.

39.

Australian Curriculum, Assessment and Reporting Authority. What does the ICSEA value mean? ACARA: Sydney, Australia; 2015.

40.

Australian Curriculum, Assessment and Reporting Authority. My School^®: ACARA; 2021. Available from: https://myschool.edu.au [Last accessed: May 14, 2021].

41.

StataCorp. Stata statistical software: Release 16. StataCorp LLC: College Station, TX; 2019.

42.

Metcalf

, Henley

, Wilkin

. Effectiveness of intervention on physical activity of children: Systematic review and meta-analysis of controlled trials with objectively measured outcomes (EarlyBird 54). BMJ, 2012; 345(7876):e5888; doi: 10.1136/bmj.e5888.

43.

Farooq

, Martin

, Janssen

, et al. Longitudinal changes in moderate-to-vigorous-intensity physical activity in children and adolescents: A systematic review and meta-analysis. Obes Rev, 2020; 21(1):e12953; doi: 10.1111/obr.12953.

44.

Bleich

, Segal

, Wu

, et al. Systematic review of community-based childhood obesity prevention studies. Pediatrics, 2013; 132(1):e201–e210; doi: 10.1542/peds.2013-0886.

45.

Tigges

. Parental consent and adolescent risk behavior research. J Nurs Scholarsh, 2003; 35(3):283–289; doi: 10.1111/j.1547-5069.2003.00283.x.

46.

Barkley

, Glickman

, Fennell

, et al. The validity of the commercially-available, low-cost, wrist-worn Movband accelerometer during treadmill exercise and free-living physical activity. J Sports Sci, 2019; 37(7):735–740; doi: 10.1080/02640414.2018.1523039.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.33 MB

0.06 MB