Effects of a daily physical activity intervention on the health-related fitness status of primary school children: A cluster randomized controlled trial

ABSTRACT An important barrier for a nationwide implementation of a daily physical activity (PA) at primary schools is the lack of spatial and human resources. Therefore, we developed a PA intervention that can be implemented without additional spatial resources or changes in school curricula. In the intervention group, children received a daily PA unit consisting of physical education lessons and simultaneous academic content over a 9-month period. The control group received conventional (physical education) classes. Body weight, height, waist circumference and health-related fitness parameters were measured. Of 412 children (9.7 ± 0.5 years) included, 228 participated in the intervention group. In regression analysis adjusted for baseline, gender, school location, sports club membership (total only) with standardized outcome variables, the intervention group showed a reduction in waist-to-height ratio (B = 0.30, p < 0.001) and an increase in several fitness parameters (cardiorespiratory endurance: B = 0.20, p = 0.037; lower body muscle strength: B = 0.11, p = 0.041; lower body muscle endurance: B = 0.12, p = 0.027; flexibility: B = 0.19, p = 0.019) compared to the control group. Intervention effects for cardiorespiratory endurance and flexibility were more pronounced in the group of children without sports club membership. Thus, especially children with no sport club membership seem to benefit from daily PA in school (Trial registration: DRKS00025515).


Introduction
Adequate physical activity (PA) in childhood and adolescence leads to short-term as well as lifelong health benefits (Campbell et al., 2014;Janssen & Leblanc, 2010;Reiner et al., 2013;Smith et al., 2014) and has additional positive effects on academic performance (Bherer et al., 2013;Fedewa & Ahn, 2011).Especially cardiorespiratory fitness and muscular fitness in childhood and adolescence are important health markers (Ortega et al., 2008), which are associated with lower body mass index (BMI), smaller waist circumference, reduced body fat percentage, and lower prevalence of metabolic syndrome later in life (Mintjens et al., 2018).
Overall, fitness improves with higher levels of PA (Bize et al., 2007;Bouchard et al., 2012;Stodden et al., 2008).In recent years, the participation in organized sports slightly increased, however in contrast, the percentage of children who participated in unorganized sports and outdoor play in leisure time decreased (Mathisen et al., 2019;Schmidt et al., 2020).In general, just a small proportion (29%) of European children and adolescents achieve the daily PA recommendation of an average of ≥60 minutes of moderate-to-vigorous PA per day.Therefore, more than two-thirds of children are insufficiently active (Steene-Johannessen et al., 2020).Generally, children who are member of a sports club more often achieve the PA recommendation (Parrish et al., 2020) compared to children without such membership (Wiium & Säfvenbom, 2019).
The COVID-19 pandemic dramatically intensified this worrisome situation.According to a systematic review, total daily PA of children and adolescents decreased by 20% during the COVID-19 pandemic (Neville et al., 2022).
Even before the pandemic, different strategies were developed worldwide to increase PA in the overall population (Reis et al., 2016).For school-aged children, the main focus is to implement interventions at school (Adamowitsch et al., 2017;Andermo et al., 2020;Brustio et al., 2019;Chesham et al., 2019;Hoelscher et al., 2001;McKay et al., 2015;Parrish et al., 2013), and a daily PA programme is a frequently applied measure (Pawlik, 2021), which has already been tested and proven to be very effective for improving PA levels in many regional projects (Erfle & Gamble, 2015;Ericsson & Karlsson, 2014;Greier et al., 2020).However, limited structural resources, such as availability of adequate sports halls (gymnasiums), limited teacher resources and the schedule of lessons pose major obstacles for general implementation of a daily PA programme in schools (Movia et al., 2022;Thiele et al., 2011).
Therefore, we developed a concept to enable nationwide implementation of a daily PA programme in primary schools without the need of additional spatial resources and curriculum changes.The aim of this study was to evaluate the effects of this programme on anthropometrics and health-related fitness (HRF) parameters of primary school children in a cluster randomized controlled trial.Possible effect moderation by children being a member of a sports club (yes or no) was assessed in addition.

Design
The study was designed as a parallel, cluster randomized controlled trial (intervention group vs. control group, 1:1 ratio) to evaluate the effects of a daily PA programme on anthropometrics and HRF of primary school children.

Selection of schools and participants
A list of all 39 primary schools in the urban and rural districts of Klagenfurt, Austria, was used to select the schools.A random number generator was applied to select 12 schools, stratified by rural and urban districts.The administrators of the selected schools consented to the participation of their school classes in the study.All 485 children attending third grade in spring 2021 in one of the 12 schools were invited to participate in the study.A total of 467 (96.3%) legal guardians gave their written consent for the participation of their children.Inclusion criteria were absence of physical limitations, age 8-12 years and attendance of fourth grade in September 2021.One child was excluded from study participation (due to physical inability to perform all fitness tests).Additionally, children who were absent from school on more than 25% of school days between September 2021 and June 2022 were excluded from the study analysis (n = 6).

Randomization and blinding
Using a random number generator, the 24 fourth grade school classes allocated to an intervention group (IG) and a control group (CG), stratified by district (Figure 1).Baseline assessments were carried out blinded for allocation, further blinding of participants or assessments was not possible.

Intervention programme
The aim of the intervention was to provide a daily PA lesson, to achieve the minimum recommended level of daily PA declared by the WHO (Chaput et al., 2020).All elements of the intervention were planned by physical education (PE) and sports experts and documented on more than 800 pages in written and illustrated form (for children of the 4th grade primary school -see supplements section Methods S1, S2 and S3).
The PA intervention programme consisted of two main modules (I & II) and one additional module (III): I) The PE classes for children in the IG were planned by experts, to be compliant with the Austrian school curriculum.In the current Austrian curriculum, PE classes are scheduled twice per week for 50 minutes (a usual school lesson in Austria at primary school level).In the IG, the PE classes in the schools were provided by the regular class teachers, which were supported by external coaches.Both received the weekly planned content for the PE classes (see supplements section Methods S2).
II) On the three days per week without PE classes, the children learned academic content combined with specific movement activities (in mathematics and language art) during one lesson.For example, with jumping or arm movements, values of numbers and arithmetic operations were connected.The children then performed the arithmetic tasks with the help of different movements.For the planning of these specific PAsupporting lessons, work materials were developed by external primary school teachers in collaboration with experts in PE and sports science, and these materials were provided to all class teachers (see supplements section Methods S3).
III) Once a week, PA activities were given as homework.For example, the children were instructed to go for a walk in the countryside.They were given a checklist of objects to be seen on this walk, such as a red car, a park bench, or a pedestrian crossing.If an object was spotted, the children had to tick it off on the checklist.When all objects were seen, the mission was successfully completed.The work materials for this PAsupporting homework were developed by external sports experts, and provided to all class teachers, who handed out the materials to the children.
The intervention programme was started after the baseline measurement (T1) in September 2021 and was implemented for nine months (i.e., one school year).The control group received the usual school curriculum, including PE classes twice a week for 50 minutes.

Outcome measurements
All measurements were standardized and pilot tested with trained personnel (sports scientists) and non-participating children (n = 98) of the same age group before the study began.Baseline measurements (T1) were conducted in September 2021.The second test phase (T2) was carried out in mid-June 2022 (end of intervention).All anthropometric data and fitness tests were collected barefoot and in sportswear.Baseline data (T1) were collected before randomization by four sports scientists, who were blinded for intervention allocation.Two of them also coached those subsequent PA sessions in the intervention classes, which took place in the gymnasium or on the sports field.The follow-up measurements (T2) were carried out by the same testing team.The anthropometrics, as well as the data for 6-minute run (6MR), shuttle run test (4 × 10SHR) and jumping sideways (JS) were collected by the two sports scientists not involved in the intervention, and who were thus still blinded for the allocation to IG or CG.However, data for standing long jump (SLJ), two-kilogram medicine ball throw (MB2kg), push-up test (PU), and V-Sit-and-Reach-test (VSR) were collected by the two sports scientists who were not blinded any longer.All the data from the children in a class were collected by four people during two consecutive school hours (of 50 minutes each).When performing the tests, it was important that the most important muscle groups were not tested one after the other.A detailed overview of the testing procedure can be found in the supplementary material (Methods S4).

Anthropometrics
Weight (kg) was measured to the nearest 0.1 kg using a Bosch PPW4202/01 body scale.Height (cm) was measured to the nearest 0.1 cm using a SECA 213 stadiometer, and waist circumference (cm) was measured to the nearest 0.1 cm using a GIMA 27,343 body tape measure.The crude BMI (body weight in kg divided by height squared in metres) and the waist-to-height ratio (WtHR -waist circumference in cm divided by height in cm) were calculated.

6-minute run (6MR)
The children were instructed to run continuously for 6 minutes a maximum possible distance around a rectangle (6 m × 18 m).The test was completed by seven children simultaneously, there was one scoring attempt and the running performance was measured to the nearest metre.

Standing long jump (SLJ)
From a starting line, the children had to jump with both legs as far as possible.The minimum distance between the starting line and the contact of the child's heels with the ground was measured with a tape measure to the nearest cm.The children had three scoring attempts, and only the best score was used.

Two-kilogram medicine ball throw (MB2kg)
From a starting line, the children had to throw a 2 kg medicine ball, which was held with both hands and touched the chest, as far forward as possible.The minimum distance between the starting line and the ball's contact with the ground was measured to the nearest cm with a tape measure.The best of two scoring attempts was used.

Jumping sideways (JS)
Children were instructed to jump sideways over a wooden stick (width 4 cm, height 2 cm) as many times as possible within 15 seconds.The jumps had to be performed using both feet at the same time and the children had two scoring attempts.The sum of correct jumps from both attempts was used.

Push-ups (PU)
The children started lying in prone position on the floor and the hands touched on the buttocks.After the start command, the hand contact was released, the hands touched the floor next to the shoulders and pushed the body up into a push-up position as straight extended as possible.Then the child released one hand from the floor and touched the other hand with it.The rest of the body remained straight and extended, only the hands and feet were in contact with the floor.The hand was returned to the position on the floor and the elbows were bent until the body was again in prone position.The hands touched the buttocks again in the prone position, then a push-up was completed.The knees did not make contact with the ground during the whole test.The children had 40 seconds to complete as many correctly performed push-ups as possible.One scoring attempt was done and the number of correctly performed push-ups was used.

V-sit-and-reach (VSR)
The VSR (Jarnig et al., 2022) was selected to measure flexibility, a simple variation of the classic Sit-and-Reach-Test being carried out without additional costs.A tape measure and marking tape were needed to perform it.A heel line was marked with the tape and the tape measure was fixed to the floor.The children sat down on the floor with their legs spread 30-40 cm apart, their heels were placed on the heel line, then the children placed one hand on top of the other and slowly stretched forward as far as possible.The distance between the heel line and the maximum reached position, which could be held with the fingertips for two seconds, was noted.Children had two scoring attempts; the longest distance reached was used.To be able to use reference values of the classic Sit and Stretch test, 15 cm were added to the scoring attempt.

× 10 shuttle run (× 10 SHR)
A shuttle run test was performed for assessing the children's action speed.A starting and turning line was marked on the floor 10 m apart using a marking tape.Three easy-to-grasp pieces of foam (S1, S2, S3) were needed.S1 and S3 were placed behind the turning line and S2 in front of the starting line.Children had to run from the starting line crossing the turning line, pick up S1, run back crossing the starting line, and put S1 down.Then they picked up S2, ran again crossing the turning line, put down S2, picked up S3, and ran with it crossing the starting line.The children were instructed to run as fast as possible.Two scoring attempts were made, and the time was measured with a stopwatch to the nearest 0.01 seconds.Participants had two attempts, and the fastest run was used.

Sample size
By using G*Power (Faul et al., 2009) we calculated that a total sample size of 423 children was required for detecting an effect size d of 0.30, with a power of 80% and alpha of 5% in a twotailed test, assuming a 20% drop out (more details are provided in the supplementary materials -Methods S6).

Moderation
Children with sports club memberships have a greater amount of average weekly PA outside of the school settings (Schmidt et al., 2020) and mostly show healthier/better health status/HRF scores (Basterfield et al., 2021;Hanssen-Doose et al., 2021;Jarnig et al., 2021;Moeijes et al., 2019;Vella et al., 2014).When completing the consent form, legal guardians provided information whether their child was member of a sports club (yes/no)at baseline.We assessed whether sports club membership moderated the effects of the intervention.

Anthropometric data
Standard deviation scores (SDS) for BMI and WtHR were calculated using the LMS method (Cole & Green, 1992) based on recent international age-and sex-specific references.To calculate SDS for BMI, we used the internationally used reference values of the International Obesity Taskforce (BMI IOTF SDS) (Cole & Lobstein, 2012).
Since there are no national reference values of the WtHR available, we used international reference values from Poland from 2016 for the SDS calculation (WtHR SDS) (Nawarycz et al., 2016).Whereas lower raw values of WtHR were expressed with higher positive SDS.

Fitness parameters
To transfer the results (raw scores) of the fitness tests into established age-and sex-specific reference values, SDS and traditional z-scores were created.Since no Austrian national reference values were available for this age group, international reference values were used.For 6MR, SLJ, and MB2kg, the most recent German values of the Düsseldorf Model (Stemper et al., 2020(Stemper et al., ) (collected 2011(Stemper et al., -2018) ) were used, for JS and 4 × 10SHR, Portuguese reference values (Rodrigues et al., 2019) from the Motor Competence Assessment of 2019 were used, and for VSR, Greek (Tambalis et al., 2016) standard values from the year 2015 were used.For these international references, calculations were performed based on the LMS method (Cole & Green, 1992).For the PU, German values (Bös, 2016) of the German Motor Test from 2016 were used and z-scores were calculated by a traditional z-score standardization.

Statistical analysis
Continuous variables are reported as means (M) and standard deviations (SD) and categorical variables as absolute values (n) and percentages (%) for descriptive statistics.No imputation of the data was performed.
To determine effects of the intervention on health status (BMI IOTF SDS, WtHR) and parameters of fitness (6MR SDS, SLJ SDS, MB2kg SDS, JS SDS, PU z-score, VSR SDS and 4 × 10SHR SDS), multilevel analyses were performed with a two-level structure: individual and class.The influence of the class level was assessed, whereas a significant Wald z-test and intraclass correlation coefficients > 0.05 were deemed as significant influence of class level and to be included in the analyses.Dependent variables were SDS values at follow-up with intervention (IG = 1 and CG = 0) as the independent variable.Models were adjusted for gender, school location (urban, rural), sports club membership (yes, no) and respective baseline SDS values.
To assess possible effect moderation of sports club membership, an additional predictor (intervention × sports club membership) was added to the model.In addition, the Cohen effect size (d) was calculated for the unadjusted comparison of delta values (difference Sept.21 to Jun.22) between the IG and the CG (as well as separately for the subgroups sports club membership yes or no).Thresholds for small, moderate, and large effects were defined as 0.20, 0.50, and 0.80, respectively.
All tests were two-tailed, with a p-value <0.05 considered statistically significant.All statistical calculations were performed using SPSS Version 28 (IBM Corp. Released 2021.IBM SPSS Statistics for Windows, Armonk, NY: IBM Corp).

Results
In September 2021, 431 children participated in the baseline measurements (T1).Thirteen children did not participate in the follow-up measurement in June 2022 (T2) and six children were present at school on less than 75% of all school days between T1 and T2.These 19 children were excluded from analyses, resulting in 412 children (95.6% from baseline) with complete data (Figure 1).
Of the 412 children, 228 (55.3%) participated in the intervention.Baseline characteristics of the study population are described in Table 1.Children had a mean (SD) age of 9.7 (0.5) years, 196 (47.6%) were girls, 171 (41.5%) were members of a sports club, and 257 (62.4%) lived in the urban region of Klagenfurt (Table 1).Differences in baseline values between IG and CG were found for JS SDS (p = 0.005) and VSR SDS (p = 0.033) (Table 2).

Primary outcomes
The raw and standardized scores of all outcomes at baseline and after the intervention are provided in the supplementary material (Table S1 &amp; S2).We found a statistically significant improvement in WtHR SDS (unstandardized regression coefficient (B) = 0.30, p < 0.001) in the IG compared to the CG.In contrast, no statistically significant changes were observed between the IG and CG groups in the development of BMI IOTF SDS (B = 0.01, p = 0.68) (Table 3).

Moderation by sports club membership
We performed tests for moderation effects and separate analyses for the subgroups of children without a sports club membership (n = 241) and children with a sports club membership (n = 171; see Table 3).Evidence for moderation by sports club membership was found for cardiorespiratory endurance (6MR SDS: B mod = 0.46 [95% C.I. 0.09;0.83],p = 0.016) and flexibility (VSR SDS: B mod = 0.33 [95% C.I. 0.01;0.65],p = 0.043).Among children who were not a member of a sports club, the IG improved cardiorespiratory endurance (6MR SDS: B = 0.38, p < 0.001) and flexibility (VSR SDS: B = 0.31, p = 0.005) compared to CG, whereas no differences were found among children who were a member of a sports club (Table 3).Borderline evidence for moderation were found in action speed (4 × 10SHR SDS: B mod = 0.24 [95% C.I. −0.01;0.49],p = 0.057).Among children who were not a member of a sports club, statistically significant differences were found between the IG and CG (4 × 10SHR SDS: B = 0.20, p = 0.017).No differences were found for this parameter in children who were a member of a sports club.

Cohen's (d) for unadjusted delta values
A comparison of the unadjusted change values (∆) showed a better development in the IG compared to the CG with a small effect (d = 0.244) for the WtHR.This could also be observed in both subgroups (sports club membership yes or no).Additionally, in children without sports club membership, larger improvements in cardiorespiratory endurance (d =    0.213) and flexibility (d = 0.273) were observed in the IG than in the CG with small effects (Table S3, Figure S1 &amp; S2).

Discussion
In the present study, we investigated the efficacy of a daily PA intervention programme over one school year in primary schools on anthropometrics and fitness.We found that the intervention led to improvements, in HRF parameters (cardiorespiratory endurance, muscle strength, flexibility) compared to the control group.Some effects (for cardiorespiratory endurance and flexibility) were more pronounced in children who were not a member of sports club.This moderation effect underlines the importance of PA in the school setting.
Especially for children who are not in a sports club, providing a daily PA intervention is of utmost importance.These findings are in line with the results of a study from Australia, which demonstrated positive effects of sports club membership on health-related parameters (Telford et al., 2016), and are supported by previous reports on the need to expand organized sports (Logan et al., 2019).Children with a sport club membership and sufficient extracurricular PA engagement might benefit from other aspects, such as respect for rules and fair play, teamwork, motivation, confidence, knowledge, and understanding (Gould & Carson, 2008;Opstoel et al., 2020;Whitehead et al., 2018).The intervention goal was certainly not new; different strategies have been used around the world to increase the amount of PA in school settings (García-Hermoso et al., 2020) and different types of interventions have been used to achieve this goal (Watson et al., 2017).In primary schools, PE is often taught by the class teacher, which might result in a weakening of the priority given to this subject compared to other school subjects (Johns and Dimmock 1999).Therefore, the approach taken in many intervention programmes is to employ external sports coaches to teach the PE classes (Jansen et al., 2008).Previously, as in our intervention the concept of teaching academic content in combination with movement and PA was used, in particular in mathematics and language learning, and led to an improvement in the academic performance of the participating children (Mullender-Wijnsma et al., 2016;Petrigna et al., 2022).Other intervention designs prefer to organize additional sports activities in the school setting outside of class time (Jansen et al., 2008), adding an additional daily PA lesson to the curriculum (Dwyer et al., 1983;Ericsson & Karlsson, 2014;Greier et al., 2020), spending the daily morning break at school with physical activities (Reilly et al., 2016), or implementing high-intensity PA programmes during short breaks in academic lessons (Ma et al., 2015).
However, for a nationwide implementation of the abovementioned concepts, many barriers must be overcome: an external source of funding is necessary, additional time, spatial and human resources would have to be provided.Additionally, it may take considerable time and effort, to successfully plan and teach academic content with PA, this may cause significant reservation or even rejection in teachers (Lomsdal et al., 2022).
In contrast, our concept does not require any changes in the school's spatial resources or curricula structure, and movement during lessons is an integral part of the academic content of maths and language arts and therefore does not interrupt the flow of children's learning.
All elements of the intervention were planned by external experts and documented in more than 800 pages in written and illustrated form, and thus a nationwide implementation in the fourth grade of primary school would be possible immediately, since PE classes could also be carried out by the classroom teachers, using the precisely planned content.

Strengths and limitations
A major strength is that the one school year long programme was evaluated in a 9-month-long randomized controlled trial, with measurement of anthropometric data and all fitness assessments performed by trained research assistants using validated measures.
However, some limitations should be noted.We did not collect data on other covariates such as nutrition or mental health.Another limitation of our study is the lack of data on the frequency and intensity of club training for those children who were members of sports clubs.Furthermore, we did not use accelerometers or other methods to measure the daily PA of the study participants, and we did not assess the PA intensity during the intervention.However, given the increase of fitness in the intervention group compared to the control group, the entire intervention apparently was intensive and frequent enough to positively influence this important parameter.The success of combining academic learning content with movement and PA depends strongly on the motivation level of the classroom teachers (Mullender-Wijnsma et al., 2015), this aspect was not assessed in our study.

Conclusion
The daily PA intervention for primary schools showed positive effects on important health-related parameters, especially for children who were no members in a sports club.PA may be combined with academic contents, resulting in a positive effect for both.The intervention programme applied in our study can be performed everywhere without significant additional costs and is therefore a candidate even for nationwide programmes.

Ethical approval and consent to participate
This study was conducted according to the guidelines of the Declaration of Helsinki and approved by the Research Ethics Committee of the University of Graz, Styria, Austria (GZ.39/23/63 ex 2018/19).All participants gave their consent to participate in the study.

Figure 1 .
Figure 1.Flow diagram.s = number of primary schools; c = number of fourth grade primary school classes; n = number of children.
The study was carried out in the greater Klagenfurt area, Austria, between September 2021 and June 2022, registered in the German Clinical Trials Registry (DRKS00025515) and approved by the Research Ethics Committee of the University of Graz, Styria, Austria (GZ.39/23/63 ex 2018/19).

Table 1 .
Baseline characteristics of participants included in the analyses, by intervention group.

Table 2 .
Baseline values of raw and standardized anthropometrics and fitness data by group (intervention vs control).

Table 3 .
Multilevel regression assessing the intervention effects on health status and fitness parameters, for all children and for subgroups with and without sports club membership.