Abstract
Background. Physical activity provides fundamental health benefits for children and youth. The aim of the study was to explore the possibility of conducting an empowerment-inspired intervention and examine the impact of the intervention in promoting moderate and vigorous physical activity (MVPA) among adolescents. Method. A nonrandomized trial with a concurrent control group was carried out. Physical activity data were collected before and after the intervention with daily questions by short message service. Self-efficacy, social support, and attitude were also measured before and after the intervention since they were possible mediators. Results. The intervention was created by the students, the researchers, and the teachers using an empowerment-based approach. Students in the intervention group (n = 21) increased their MVPA on average by 4.9 (SD = 28.9) minutes per day, and students in the control group (n = 25) reduced their MVPA on average by 25.4 (SD = 23.0) minutes per day (p = .000). Conclusions. The intervention might have contributed to a promotion of physical activity among students in the intervention group. The most valuable contribution this study provides is the knowledge that it is possible to develop and conduct an empowerment-inspired intervention to promote adolescent physical activity.
Keywords
Introduction
Physical activity (PA) provides fundamental health benefits for children and youth (World Health Organization [WHO], 2010), including positive effects on the musculoskeletal system and cardiovascular health (Janssen & LeBlanc, 2010), as well as self-image (Goldfield et al., 2011). Health might even be a critical partner for optimum education (Rothon et al., 2009), and studies have found associations between adolescent PA and academic performance (Fedewa & Ahn, 2011). According to WHO (2010), the recommendation for health-enhancing PA for adolescents is to be physically active for a total of at least 60 minutes daily. The activity should include both moderate and intense activity but can be divided into several sessions during the day (WHO, 2010). Boys are more physically active than girls in all age-groups, and according to Health Behaviour in School-Aged Children (HBSC), among Swedish adolescents aged 15 years only 13% of the boys and 9% of the girls achieve the recommended levels of PA 7 days per week (Folkhälsoinstitutet, 2011). There is evidence that the tendency to adopt sedentary behavior may increase through adolescence (Dumith, Gigante, Domingues, & Kohl, 2011), and there has been a decline in cardiovascular fitness in Swedish 16-year-olds between 1987 and 2007 (Ekblom, Ekblom Bak, & Ekblom, 2011).
School-based interventions that include multiple elements such as teacher training, changes in curriculum, assistance in behavior change, increased health education, and involvement of parents have a positive effect on children and adolescents’ PA during school hours, and in some cases after school as well (Swedish Council on Health Technology Assessment [SBU], 2006). However, the SBU notes that insufficient scientific data prevent conclusive determination of the impact that these methods have on children’s and adolescents’ PA levels, and they ask for further research into which methods are most effective for different age-groups, as well as for studies that use modern technology to maintain behavior change (SBU, 2006). Unfortunately, fewer than 30% of schools implement evidence-based interventions, and this might be due to a disconnect between research and the needs of end users in the real school world (Leatherdale, Manske, Wong, & Cameron, 2009).
Previous PA interventions have had only a small effect on adolescents (Metcalf, Henley, & Wilkin, 2012), and to succeed, interventions need to use theory to guide their interventions (LaPlante & Peng, 2011). Social cognitive theory (SCT) is one of the most frequently used health behavior theories. Cognitive processes such as self-efficacy and goal setting presumably influence PA levels, and the degree of self-efficacy that an individual possesses directly affects the ability to change (Bandura, 2004). Self-efficacy can partially mediate the effect of interventions for (Haerens et al., 2008). Strategies for increasing self-efficacy include setting progressive goals, behavioral contracting, and reinforcement (Bandura, 2004).
Adolescence is characterized by a shift to independent decision making that is strongly influenced by peers and technology (Gibbons & Naylor, 2007). Adolescents use a considerable amount of information and communication technology in their everyday life, and supporting health promotion with information and communication technology is a promising approach to use for adolescents (Tercyak, Abraham, Graham, Wilson, & Walker, 2009). Short messaging services (SMS) are considered to be cost- and time efficient, accessible, and convenient and have shown promising results in increasing PA among inactive adolescents (Sirriyeh, Lawton, & Ward, 2010). Students who wanted to increase their PA level suggested that reminders via SMS and social support from friends and parents could facilitate health-promoting behavior (Lindqvist, Kostenius, & Gard, 2012). The results found by Williams and Mummery (2011) confirms that parents and home environments play significant roles in adolescent PA. The intervention in this study is created from the ideas of the students themselves, as well as the theoretical base of SCT. Empowerment and the formation of partnerships with adolescents offer promising avenues for those who encounter the challenge of promoting PA among adolescents (Lindqvist et al., 2012). Empowerment is a multilevel construct, including both individual influence over one’s life as well as participation in group activities and/or activities in society (Rappaport, 1987). Raeburn and Rootman (1998) highlighted five key components of empowerment: control, competence, confidence, contributing, and participating. The aim of the study was to explore the possibility of conducting an empowerment-inspired intervention and examine the impact of the intervention in promoting moderate and vigorous PA (MVPA) among adolescents.
Method
Design
We used a nonrandomized trial with a concurrent control group (Dawson, Trapp, & Trapp, 2004).
Participants
Participants were recruited from one school in a municipality of approximately 17,000 inhabitants in the northern part of Sweden. There were eight classes of ninth graders, and the intervention group consisted of a convenience sample since we asked all students from two classes to participate. A total of 32 students (16 boys and 16 girls) were asked, and 27 (14 boys and 13 girls) agreed to participate. The control group consisted of two other classes from the same school; 34 students (15 boys and 19 girls) were asked, and 26 (7 boys and 19 girls) agreed to participate. Participant characteristics at baseline are displayed in Table 1. Two of the students in the intervention group and one of the students in the control group reached the recommendations of MVPA for adolescents at baseline.
Participant Characteristics
NOTE. BMI = body mass index; MVPA = moderate and vigorous physical activity.
Intervention
The contracts, the content of the SMSs, and the parental brochure were created by the students, the researchers, and the teachers using an empowerment-based approach. In this context, empowerment is assumed to promote MVPA through the different parts of the intervention, as depicted in Figure 1.
Theoretical Model Depicting the Mediators of the Intervention on MVPA
The students received financial compensation for the SMSs they sent during the study, approximately US$11 in the form of a movie ticket. The empowerment process concerning the contracts, the encouraging SMSs, and the parental brochure is described in the Results section.
Data Collection
Anthropometric Measures
Prior to the baseline measurements, the participants’ height and weight were measured by the school nurse. Height was measured without shoes, and weight in kilograms to the nearest decimal was measured on a newly calibrated digital scale without shoes but otherwise dressed. Body mass index was calculated (weight [kg]/height [m2]).
Measures of MVPA
Self-rated MVPA was evaluated using questions posed by SMS. The software used was the SMS Track Questionnaire (www.sms-track.com). Using text messages and mobile phones to collect frequent data has been shown to be user-friendly and to have a high response rate (Axén et al., 2012). The question used appeared in the questionnaire HBSC: “Physical activity is any activity that gets your heart beating faster and makes you breathe faster. How many minutes have you been physically active today?” and the question was sent every day at 2100 hours during 1 week pre- and postintervention. The students answered by return SMSs. If the students had not answered by 1100 hours the following day, they got a reminder offering a new chance to reply.
Measures of Self-Efficacy, Attitude, and Social Support
Since self-efficacy was a possible mediator, we used the Physical Activity Self-efficacy Scale for adolescents. The original scale has been tested and found to have satisfactory reliability and validity (Wu, Robbins, & Hsieh, 2011). The scale has been translated into Swedish according to principles of good practice for the translation and cultural adaption of patient-reported outcome measures (Wild et al., 2005). We used a 7-point Likert-type scale, and three questions used in other studies with adolescents (Trost et al., 2003) were added. Social support and attitudes were also possible mediators. To measure them we used five questions (social support) and four questions (attitudes) used in earlier studies with adolescents (Haerens et al., 2008; Lewis, Dollman, & Dale, 2007). The questionnaire began with a definition of PA from HBSC.
Physical activity is any activity that gets your heart beating faster and makes you breathe faster. Physical activity can be doing sports, various activities at school, when you play with peers, or when you are going to school. Some examples of physical activity are running, fast walking, skating or roller skating, swimming, playing soccer, cycling or dancing. (Folkhälsoinstitutet, 2011)
Questions on perceived health and parents’ and friends’ PA levels were also included. The questionnaire was completed pre- and postintervention. The questionnaire is available from the authors on request.
Ethical Considerations
Informed consent was collected from the parents who agreed to let their children participate. In addition, the authors gave verbal and written information to the students, and those who agreed to participate also signed an informed consent. The written information explained the aim of the research and stated that participation was voluntary, students could terminate their participation without needing to provide any reasons for doing so, and the results would be analyzed on a group level so that no individual could be identified. The students were divided by the teachers into pairs to promote a suitable match and to avoid leaving anyone alone. The students sent a copy of each encouraging SMS to the researchers, so we could ensure that no offensive messages were sent. The students in the control group were extended the offer to take part in the intervention after the study was completed. The research ethics committee in Umeå, Sweden, approved the study before the start of the research project (date of issue: February 11, 2011, application registration number: dnr 2010-337-31Ö).
Data Analysis
The questionnaire was tested for face validity by both adolescents and teachers working with adolescents with satisfactory results. Factor analysis was used to test internal consistency, and the measures of self-efficacy and social support each conformed to a single-factor structure. Regarding the attitude scale, one question had to be excluded for it to form a single-factor structure. Cronbach’s α was .95 for self-efficacy, .80 for social support, and .78 for the three remaining questions on attitude. Missing responses to single items on the questionnaires varied from 1.9% to 11.3%. Missing responses to the SMS questions ranged from 3.8% to 26.4% per day, and the highest rate of missing responses to the SMS questions occurred during the weekend. By using box plot analyses, extreme outliers (defined as values exceeding the distance between median and quartile values more than 3 times) were identified. Each extreme value was carefully considered, and two answers on MVPA SMS that were judged to be incorrect were excluded. Individuals with less than four responses per week to the MVPA SMS and answers from individuals who gave notice about being sick or were absent from school due to illness were excluded. Differences in MVPA, self-efficacy, social support, and attitude before and after the intervention were calculated. The data on differences in MVPA were normally distributed, and Student’s independent sample t test was used to analyze differences in MVPA before and after the intervention, as well as differences between the intervention and control group. The influences of the independent variables of social support before, attitude before, self-efficacy before, friends’ and parents’ PA levels, and students’ MVPA levels before the intervention were analyzed through regression analyses (enter method), using the MVPA before as a dependent variable. The influences of the independent variables of differences in self-efficacy, attitude, and social support were also analyzed through regression analyses (enter method), using the difference in MVPA as a dependent variable. Statistics were computed using SPSS, Version 19.0 (SPSS Inc., Chicago, IL, 2011).
Results
The process of conducting the empowerment-inspired intervention concerning the behavioral contracts, the encouraging SMSs, and the information for parents took two plus two lessons with 3 days apart to carry out, and the process went very smoothly. The students were divided by the teachers into pairs and were asked to make a mutual written contract including a goal for PA and a promise to support each other’s PA over the course of 1 month. “Getting in shape and feel good,” “build stamina and muscle,” and “sleep better” were some examples of goals that the students chose to write in the contracts. In the contract they also agreed to send one SMS to each other once a day for 1 month to encourage PA. They were asked to talk with their partner to get a picture of what that person perceived as encouraging and at what time of the day he or she would prefer to get the message. “Remember to put in a little extra at the gym today,” “Walk the dog or something,” and “Get out of bed and on the move” were some examples of their encouraging SMSs to each other. The main headlines of the parental brochure were decided on by the entire group of students in the intervention group; examples were “Why is it good to be physically active?” “The relationship between PA and school performance,” and “How can parents support PA?” The students worked in smaller groups to create the content related to one headline per group and presented their contribution to the whole group the following day. Finally, there was an editing process to complete the brochure, and it was sent home to the parents.
On average, the students in the intervention group (n = 21) increased their MVPA per day by 4.9 (SD = 28.9) minutes, and the students in the control group (n = 25) decreased their MVPA per day by 25.4 (SD = 23.0) minutes (p = .000; see Table 2). Two of the students in the intervention group and none of the students in the control group reached the recommended levels of MVPA for adolescents after the intervention.
Difference in Minutes of MVPA per Day Before and After the Intervention for the Control Group and Intervention Group
NOTE: MVPA = MVPA = moderate and vigorous physical activity.
The results showed that social support before, attitude before, self-efficacy before, and friends’ and parents’ PA levels before intervention explained 31% of the variance in MVPA at baseline; however, self-efficacy (p = .043) was the only significant predictor of MVPA before the intervention. The model was significant (p = .015). The intervention, difference in social support, difference in attitude, and difference in self-efficacy explained 28% of the variance of the difference in MVPA, although the intervention (p = .001) was the only significant predictor of the difference in MVPA. The model was significant (p = .016). The intervention alone contributed to 24% of the variance of difference in MVPA (p = .001). The intervention and the level of MVPA before intervention explained 48% of the difference in MVPA (p = .000).
Discussion
The aim of the study was to explore the possibility of conducting an empowerment-inspired intervention and examine the impact of the intervention in promoting MVPA among adolescents. Earlier studies have shown that student-driven health-promoting activities are a fruitful avenue (Kostenius, 2013), and this is consistent with our findings. The ability and the commitment shown by the involved students are visible, for example, in the content of the encouraging SMSs and the parental brochure.
According to HBSC (Folkhälsoinstitutet, 2011), many children fail to meet the recommendations for PA, and this is consistent with our findings. The students in our intervention group increased their MVPA by 4.9 minutes per day, but the students in our control group decreased their MVPA by 25.4 minutes per day. A recent review and meta-analysis of controlled studies concerning the effectiveness of interventions aiming to increase PA in children provided evidence of only small effects, approximately 4 minutes more walking or running per day (Metcalf et al., 2012). One explanation for the decline in MVPA in the control group might be the difference in the weather between September and November in the northern part of Sweden. In September the average temperature was 8.8 °C, with no frost days and 142 sunshine hours (SMHI, 2013). In November the average temperature was 0.1 °C, with 19 frost days and 28 sunshine hours (SMHI, 2013). A review that assessed the effects of weather conditions on PA notes that the number of published studies is small, but in general the data confirm that the weather has an impact on PA (Chan & Ryan, 2009). The students in this study participated in an earlier focus group study, and the result showed that they were more physically active during the warmer season and more inclined to be engaged with their computers during winter (Lindqvist et al., 2012). One might argue that the decline in MVPA in the control group is a natural development for the season and that the intervention managed to prevent the decline and even reverse it to a small increase.
Demetriou and Höner (2012) state that future studies should place a stronger focus on the investigation of the mediators of interventions. Self-efficacy is a core concept of SCT, and previous studies have confirmed that self-efficacy mediates the causal pathway between interventions and adolescent PA levels (Haerens et al., 2008). Our results indicate that self-efficacy was relevant to the students’ MVPA, but this intervention was not significantly mediated by self-efficacy or by attitude or social support. A reasonable explanation might be that 1 month is not enough to affect the chosen mediators, as it takes time to change one’s self-efficacy and attitudes.
Methodological Considerations
The study had an explorative approach and is not randomized, there were some differences between the groups at baseline, and the duration of the intervention was only 1 month, with no long-term follow-up, so the results should be interpreted with caution. According to Fraser, Richman, Galinsky, and Day (2009), only after an intervention has been fully developed would an randomized controlled trial be considered appropriate. Furthermore, the core of this empowerment-inspired intervention originates from a qualitative study (Lindqvist et al., 2012), and it is our opinion that it was ethically responsible to invite the students who participated in that study when we put their ideas into action.
It is difficult to achieve randomization across classes within a school, since students in the intervention group might transmit relevant information to the students in the control group. Nevertheless, we chose to form a control group at the same school since we believe that the value of engaging participants from the same context outweighs the risk of contamination. In a future efficacy or effectiveness trial, it might be appropriate to minimize this risk with a cluster randomization. In our opinion, one of the advantages of this study is the fact that 53 out of the 66 students who were invited chose to participate. This might provide a wide range of both very active and less active students, but at the same time it could be an explanation for the small effect of the intervention and the large standard deviation in MVPA. When all students are invited to participate, the already active may create a ceiling effect and dilute the effect of the intervention (Naylor & McKay, 2009). The intervention was empowerment based, and we included those who were interested and motivated to participate, although future studies might focus on the less physically active.
In many questionnaires, the children are asked to recall their PA for the past week. Youth may have an activity pattern that is much more variable than that of adults, and there might be a difference in the ability of children to think abstractly and remember their PA in detail; these factors suggest that PA reports made by children are more likely to suffer from recall bias and less likely to be accurate (Chinapaw, Mokkink, Van Poppel, Van Mechelen, & Terwee, 2010). Having a predetermined and limited time to answer a question might limit the recall bias compared to other methods (Shiffman, Stone, & Hufford, 2008). Collecting data with SMS is promising in terms of low costs and minimal time consumption (Johansen & Wedderkopp, 2010), and we hope to be able to offer a valuable contribution (not yet published) to the validation of self-rated MVPA through questions posed by SMS by correlating them with measurements by accelerometers and the International Physical Activity Questionnaire. Based on the results of this study, our opinion is that collecting data by daily questions posed by SMS is a feasible method for use among adolescents.
Conclusion
We believe that it is essential to engage the end users in decision making, and the knowledge gained from this study has strengthened our belief that empowerment could be a helpful approach for those who encounter the challenge of promoting MVPA among adolescents. The students in the intervention group increased their daily MVPA by 4.9 minutes, and the students in the control group decreased their daily MVPA by 25.4 minutes, and the difference was significant. The decline in PA in the control group might be a natural development during the season. Self-efficacy was relevant to the students’ MVPA, but this intervention was not significantly mediated by self-efficacy or by attitude or social support. In our opinion, the most valuable contribution that this study provides is the knowledge that it is possible to develop and conduct an empowerment-inspired intervention to promote adolescent MVPA, although future studies of the efficacy and the experience of taking part in an intervention with this structure are needed.
Footnotes
Acknowledgements
We would like to express our gratitude to the students who participated in the study. We are also grateful for the time given by the teachers, the principal, and the parents and for the cooperation they provided.