Web-based need-supportive parenting program to promote physical activity in secondary school students: a randomized controlled pilot trial

Physical activity (PA) has many known benefits for children and adolescents [1, 2]. In addition to physical adaptive health outcomes such as reduced cardio-metabolic risk factors [3], the benefits can also be psychological such as reduced anxiety [4] and cognitive such as better academic performance [5]. On the other hand, physical inactivity has a detrimental effect on health, frequently leading to overweight that is difficult to lose later in life, and thus insufficient PA in childhood can result in a higher risk of non-communicable diseases such as cardiovascular disease as an adult [6]. Although WHO recommends that children and adolescents should get an average of 60 min of moderate-to-vigorous physical activity (MVPA) per day [7], the worldwide problem is that most children and adolescents to not meet the guideline levels of PA [8]. For example, data from 146 countries indicated that in 2016 more than 80% of adolescents were insufficiently active [8]. In Estonia, where the current pilot study was conducted, the proportion of children and adolescents meeting the WHO recommendation on PA has increased comparing 2018 and 2021 (28% to 43%), but even so less than half of the children get enough MVPA daily [9, 10]. Moreover, PA levels in adolescents decline with age [11, 12]. Specifically, Rubín and colleagues demonstrated that during the transition from childhood to adolescence sedentary time increases at the expense of PA [13].

PA in childhood predicts adult PA levels – being active as a child significantly increases the probability of being active as an adult [14]. As PA patterns tend to carry forward into adulthood, it is important to develop effective interventions to promote PA in children and adolescents. Researchers have tried to identify the determinants of PA that could be modified to result in higher activity levels.

One of the prominent motivational theories frequently used in developing health interventions is the self-determination theory (SDT) [15]. Deci and Ryan differentiate between several forms of motivation ranging from extrinsic to intrinsic. According to the SDT, a person’s intrinsic motivation is higher when his/her innate psychological needs for autonomy, competence, and relatedness are satisfied. The need for autonomy indicates that a person wants to feel s/he has a choice when it comes to what is happening in his/her life; the need for competence refers to one's desire to cope with difficulties and experience success; and the need for relatedness means that a person wants to be surrounded by caring companions, to be accepted by others. For optimal performance and wellbeing, social environment, including supportive behaviour from parents, peers, teachers etc., should enable satisfying all three basic psychological needs [16]. One of the main premise of SDT is that when a person’s psychological needs are fulfilled in an activity, development of intrinsic motivation towards this activity is facilitated [17]. This suggests that to promote PA in children and adolescents, interventions to support their basic psychological needs in the PA context could be used. Earlier intervention studies to support basic psychological needs have used various approaches and the obtained effects also vary [18]. Teixeira and colleagues [19] conducted a study using the Delphi method to identify techniques to support basic psychological needs. As a result, they classified 21 motivation and behaviour change techniques (MBCTs) that are used in SDT-informed interventions and described 7 techniques for satisfying each of the three psychological needs. Overall, research based on the SDT in the health domain has demonstrated that SDT-informed interventions positively affect health behaviour [18].

Understanding the determinants of PA in adolescents has been greatly advanced by the application of the trans-contextual model (TCM) [20, 21]. TCM is a multi-theory model, integrating components from SDT, the theory of planned behaviour (TPB) and the hierarchical model of intrinsic and extrinsic motivation [22]. SDT differentiates between autonomous and controlled types of motivation towards particular behaviour (e.g., PA) and TPB states that a person’s intention, strongly supported by autonomous motivation, is the most proximal predictor of PA behaviour. The hierarchical model describes how autonomous motivation in one context influences it in another context. The integration of these theories is based on the premise that they are complementary, with SDT and TPB explaining the reasons behind motivated behaviour and the hierarchical model being the connecting element between contexts [20].

Research based on TCM has revealed that adolescents who perceive greater autonomy support in physical education (PE) are more likely to engage in PA outside of school, due to increased autonomous motivation, social cognition beliefs (i.e., attitude, subjective norms, and perceived behavioural control), and intentions towards PA [20, 21]. Numerous intervention studies have been carried out in the school context, with PE teachers adopting need-supportive techniques [23, 24]. It has also been demonstrated that need satisfaction in PE classes and the amount of out-of-school MVPA are correlated, with the decrease of need satisfaction also decreasing PA levels [25].

The home setting has not been extensively studied so far. However, parents are also important influencers regarding children’s health behaviours. Parental support is significantly correlated to children’s PA levels [26, 27], so educating parents about need-supportive behaviours should help increase children’s PA.

“Active 1 + FUN” is an example of a family-based intervention program designed based on the tenets of SDT. The 6-month program was designed to help parents support their children’s basic psychological needs and increase children’s MVPA and co-activity with parents. The intervention was carried out using face-to-face approach and comprised of ten activity sessions. The program resulted in improving children’s fundamental moving skills, however changes in accelerometer-measured MVPA were not evident [28, 29]. The FRESH (Families Reporting Every Step to Health) study, an online family-based PA intervention program, demonstrated that children in the intervention group enjoyed participating in the study with their family members and reported more PA. Nevertheless, children’s accelerometer-measured MVPA was not significantly different between study groups after the 8-week intervention [30, 31]. Both “Active 1 + FUN” and FRESH studies had the limitation that target sample size was not reached, which might be one of the reasons for the lack of significant intervention effects [29, 32]. Furthermore, in the FRESH study randomization by county was used, which resulted in unbalanced study groups [31].

Behavioural interventions can harness digital technology to improve scalability and effectiveness [33]. Reeve and Cheon [34] have suggested that need-supportive interventions could be delivered in an online format, offering the possibility for more personalization and self-paced learning. Tilga and colleagues [35] have demonstrated that an entirely web-based intervention program for PE teachers was effective in enhancing students’ perceptions of autonomy-supportive teacher behaviours and students’ need satisfaction. The effects of the study were also evident at 15-month follow-up and in regard to intrinsic motivation even more pronounced [36]. Therefore, a web-based intervention may be an optimal and effective option to apply in future.

Parental support is significantly correlated to children’s PA levels [26, 27]. Drawing from SDT, when a child’s basic psychological needs in PA are fulfilled, it facilitates the development of intrinsic motivation towards PA [17]. Hence, educating parents about need-supportive behaviours should help increase children’s PA. In this study we tested an online need-supportive training for parents expecting a positive effect on children’s leisure-time PA. A positive, albeit statistically insignificant trend can be seen in vigorous and total MET-minutes per week in the intervention group (Table 3), so it seems the need-supportive parenting program has some positive influence on the PA of children. However, we were not able to demonstrate statistically significant intervention effect on students’ MVPA, similarly to “Active 1 + FUN” and FRESH studies that also aimed to increase children’s PA through a family-based intervention [29, 31]. Like our intervention, the FRESH study was carried out online [31].

According to SDT, satisfying the basic psychological needs enables the development of more autonomous forms of motivation [17]. As a result of the parents’ training, we expected positive changes in intrinsic motivation and lower perceptions of controlled forms of motivation in children. Consistent with our expectations, students in the intervention group scored lower in introjected regulation, an extrinsic form of motivation, compared to students in the control group. However, no differences were found in external motivation, which is an even less self-determined form of motivation. In introjected regulation, behaviour is regulated by internal pressures to feel pride or avoid guilt or shame [15]. In the context of PA this could mean going for a walk because otherwise one would feel shame for being lazy or participating in training to win a contest. Compared to introjected regulation, intrinsic motivation is more likely to be maintained over time and thus should be aimed for when attempting to increase children’s PA.

Previous research has shown that need-supportive behaviours from PE teachers and peers enhance need satisfaction and lower need frustration for students in the PE context [48, 51]. We hypothesized that also children whose parents attended the need-supportive parenting training would score higher on need satisfaction and lower on need frustration compared to children in the control group. According to expectations students in the control group showed higher autonomy frustration at post-intervention and follow-up. This means that the children in the control group experienced more external pressure towards leisure-time PA and less perceived possibilities to choose the activities according to their own interests (e.g., “I feel I have to do a lot of things I don’t want to”). Therefore, it seems that parents using specific techniques to support children’s basic psychological needs for autonomy, competence, and relatedness, help relieve external pressure and support the development of more autonomous forms of motivation in students. To explicitly mention some techniques, the use of non-controlling language, providing choice and helping the child identify sources of external pressure allow the child to feel more in control, thus reducing autonomy frustration and controlled forms of motivation.

Regarding perceptions of need-supportive behaviour from parents, attitude and perceived behavioural control, no statistically relevant changes emerged. These findings are unexpected as previous research by Tilga and colleagues has demonstrated that a web-based autonomy-supportive training for teachers enhanced students’ perceptions of teachers’ autonomy-supportive behaviours [35]. As opposed to quite structured PE classes, domestic everyday life may not offer appropriate possibilities to try out new communication techniques discussing PA so regularly. In addition, it may take more time for the children to adjust to the new communication style by parents and change their perceptions accordingly. In further research, we aim to add another data collection point 5 months after the intervention to capture the possible long-term effects on children’s PA and psychological measures. Schneider and colleagues [61] conducted a study in the school setting, training PE teachers to use autonomy-supportive techniques. To point out the differences, the mean age of participating students was a bit higher than in our study (14.5 years vs 12.5 years), the need-supportive intervention was delivered by PE teachers vs parents in our program and the intervention was delivered face-to-face as opposed to our online approach. However, similarly to our study, Schneider and colleagues aimed to promote leisure-time PA of students by providing need support from social agents. They expected positive change in intrinsic motivation, other constructs from the trans-contextual model of motivation, and leisure-time PA, but were not able to obtain expected results. Schneider and colleagues argued that one possible reason for the null result was the baseline level of perceived autonomy support already being very high, leaving little room for improvement [61]. Similarly, the baseline values for perceptions of need-supportive behaviours from parents, attitude and perceived behavioural control in our study were also extremely high (on a 7-point scale > 5.90 for perceived need-support and attitude, > 5.50 for perceived behavioural control), being one possible explanation for the lack of significant positive changes. Both Schneider and colleagues’ and our study imply that changing student perceptions of need-supportive behaviour from social agents takes time, and the larger effect may come from reducing controlling behaviour (e.g., in our case the decreased autonomy frustration is the intervention group).

Inconsistent with our expectations, we found that intention and effort towards out-of-school PA were higher in the control group in the follow-up measuring occasion. One possible explanation could be seasonality regarding outdoor PA as both intention and effort are associated with actual PA. There are substantial seasonal differences in Estonia with rainy autumns, often snowy and windy winters, and moderately warm summers. It is also worth mentioning that from November to February daylight is scarce. Research has shown winter to be associated with lower levels of PA, especially among girls [62]. In our study, we had the first measuring occasion in October, the second in December, and the third in January. However, there were proportionally more boys in the intervention group than in the control group and as the boys’ PA levels should be more stable throughout the year, it is difficult to establish the effect of seasonality on our results.

The study has also some limitations. Firstly, we only used questionnaires for measurement, including collecting data about PA behaviour. The reason for using questionnaires solely is the aim to recruit participants as widely as possible. However, research has shown that in self-report measures participants frequently overestimate their PA levels as compared to accelerometer-measured data [63]. It seems self-report questionnaires to estimate PA levels could be used in older adolescents, especially in order to determine non-compliance with the recommended guidelines [64]. Objective measurement using accelerometers would add accuracy, at least partially to determine correlations between instruments. This is planned for future research.

The reasons for finding few intervention effects in post hoc tests can be attributed to the interaction effects being only borderline statistically significant, large standard deviations, the small number of participants in the study and uneven sample size in study groups. This does certainly reduce the statistical power of the study and the generalizability of the results. We found it difficult to get parents to participate in the training program, an obstacle also encountered by van Sluijs and colleagues [32]. In the pilot study, we used only school-based recruitment, approaching schools beforehand in written form and then visiting relevant classes in person. The results from our study make it clear that in further research it is necessary to employ several methods for recruitment. It is important to avoid relying only on children to communicate the aim, form, and content of the intervention to parents [30]. In further research we plan to include different channels for effective recruitment (information to children via schools, information to parents via schools’ online study systems and email lists, participating in parents meeting directly, posting in social media parenting and community groups, etc.) resulting in possible participants seeing the information on several occasions and being able to directly communicate with the research team.

Another limitation of the study is the significant difference in the proportion of male and female students between the intervention and control groups. We cluster-randomized by school, not knowing beforehand how many students voluntarily enroll and what the proportion of male and female students would be. This does impact the generalizability of the results.

A possibility for improvement is related to missing values. In the pilot study we used pen-and-paper questionnaires and had randomly missing answers to questions throughout questionnaires, reducing the reliability of the obtained findings. In further research we plan to use online questionnaires as this allows marking all answers as required. Tilga and colleagues have successfully used this approach on 11–15-year-old students previously [35].

Some limitations also apply to the used questionnaire. We adapted perceived competence and perceived relatedness support scales from Standage and colleagues [45] to refer specifically to parents. The reliability of the used scales was assessed using Cronbach’s alpha coefficient and the values were between 0.873 and 0.934, which suggests reliability of the scales used. Secondly we slightly modified the BPNSNF scale by Chen and colleagues [49] that has been adapted to the PE context by Haerens and colleagues [50] and we slightly modified the adapted version to refer specifically to leisure-time PA. The reliability of the used scales was assessed using Cronbach’s alpha coefficient and the values were between 0.856 and 0.924, which suggests reliability of the scales used. Future studies with larger sample size are needed to validate the adapted versions of the Perceived Need Support Scale and BPNSNF.

Further research should be carried out to test the intervention in other domains, where messages related to PA are delivered, such as PE in schools. To achieve the highest impact on students’ PA levels, interventions in several domains (e.g., home and school settings) could be combined. This would mean creating a need-supportive social environment in several contexts for the children, thus leading to a comprehensive approach to promoting their PA levels. Such supportive environments enable satisfying all three basic psychological needs, making it possible to reach optimal performance and wellbeing for the children [16]. Potential challenges of implementing the intervention in different settings include recruiting a sufficient number of participants and students not wearing the accelerometers for 7 consecutive days during the census period (e.g., forgetting to wear the device).

Future research could be based on a newer classification of motivational behaviours by Ahmadi and colleagues set in the educational context [65]. The advantages of this classification include describing need thwarting behaviours in addition to need supporting techniques, more experts participating in expert panels, and higher criteria for reaching consensus. The inclusion of need-thwarting behaviours enables practitioners to identify which behaviours to avoid. Previous research has shown that controlling behaviour attenuates the effect of autonomy-support on students’ health-related outcomes [66]. This classification system gives clear definitions of each behaviour and provides estimates of how effective it might be for promoting motivation, thus enabling researchers and practitioners to select the most relevant techniques for the aim of the potential intervention.

The results of the study are expected to contribute to informing the development of future interventions and educational programs for parents to promote PA in children and adolescents. The online approach makes it possible to reach large audiences and enables self-paced learning for participating parents. It is vital to find effective and cost-effective ways to support PA, that can also be adapted to various countries and age groups.

This study tested an online need-supportive training for parents to enhance children's leisure-time physical activity and autonomous motivation. There was a positive trend in vigorous and total MET-minutes per week for children in the intervention group, but no statistically significant intervention effect. However, the parents' training resulted in lower perceptions of introjected regulation and autonomy frustration in children, indicating the development of more autonomous forms of motivation towards leisure-time PA in children. These findings suggest that while the parent’s need-supportive training may not have a significant impact on physical activity, it can increase children's intrinsic motivation and reduce controlled forms of motivation. Future research should aim to address the limitations of the current study, including the use of objective measures of physical activity and recruiting larger sample sizes. To improve the generalizability of the findings, future research could test the intervention in other domains beyond leisure-time PA, such as school PE lessons.