International Journal of Behavioral Nutrition and Physical Activity Psychosocial Factors Underlying Physical Activity

Background: Given the increasing importance of obesity in China, prevention interventions encouraging physical activity by middle school students are needed. The purpose of this study is to illustrate how a rapid elicitation method can be used to identify salient consequences, referents, and circumstances about physical activity as perceived by middle school students and to provide suggestions for interventions and quantitative research.


Background
Overweight and obesity are becoming important public health problems in China among both adults and youth. A nationally representative cross-sectional study reported the age-standardized prevalence of overweight is 28.9% among Chinese adults [1]. The rate of overweight and obesity observed in a study of urban school children increased from almost 8% in 1991 to more than 12% six years later [2]. The National Student Constitution Survey (NSCS) in 1985, 1991, 1995, and 2000 reported the prevalence of obesity among in-school students has increased alarmingly in the metropolitan areas of China over the past 15 years [3].
It is logical to link the increasing obesity rates in developing countries with a progressive introduction of factors associated with obesity in developed societies, such as sedentary lifestyle and consumption of high fat and fast foods [4]. A cross-sectional study conducted in 9 provinces of China reported that in-school activities and active commuting represent the most important forms of moderate or vigorous physical activity of the Chinese youth [5]. According to the report on Global School-based Student Health Survey, only 21% of Chinese students living in Beijing between 13 and 15 years of age are engaging in physical activities for at least 60 minutes per day outside physical education class. In this study, physical activity was defined as participating in physical activity which increases the heart rate and makes the person get out of breath some of the time; participants were asked not to count time in physical education or gym class [6]. A review of factors influencing obesity in the United States indicated that adequate participation in physical activity in childhood may be of critical importance and that physical activity habits developed early in life may persist into adulthood [7]. Additionally, there is an emerging body of evidence recognizing the association between physical activity and improved academic outcomes as well as the association between physical activity and reduced incidence of depression, anxiety and fatigue [8][9][10][11][12][13]. In the US, a number of successful prevention programs have focused on middle schools presumably because middle school students are more cognitively mature and have more control over physical activity options than elementary school students [14]. Therefore, to prevent and control the epidemic of childhood obesity in China, effective prevention programs addressing physical activity are needed and middle school is a good place to start.
Experience has shown behavioral interventions based on an empirical understanding of psychosocial factors underlying people's decisions are more likely to be effective. Stated another way, research identifying the factors associated with the target behavior among representatives of the priority group is an important prerequisite to effective interventions. Meta-analytic reviews [15,16] have demonstrated that the Theory of Planned Behavior [17,18] can be used to understand the psychosocial determinants of physical activity behaviors and, therefore, can provide a foundation for intervention design to increase physical activity.
Briefly, the Theory of Planned Behavior (TPB) proposes that the immediate determinants of volitional behavior are a person's intention to engage in the behavior and perceived behavioral control [17,18]. Intention is, in turn, determined by a weighted combination of attitude towards the act, subjective norm, and perceived behavioral control. And, finally, attitude towards the act, subjective norm, and perceived behavioral control are determined by sets of salient behavioral, normative, and control beliefs. According to the TPB, only salient beliefs operate as psychosocial determinants. Therefore, a critical step in the application of the TPB is a salient belief elicitation study of the target behavior with members of the population of interest [15]. The purpose of an elicitation study is to determine the top-of-the-mind or salient consequences, referents, and circumstances that form the belief structure underlying the intention. It is essential that these salient beliefs be elicited for each new behavior and for each new priority group. To determine the modal salient beliefs, Ajzen and Fishbein [17,18] recommend that researchers: (a) conduct an elicitation study with representatives of the priority group using open-ended questions to identify salient consequences, referents, and circumstances; (b) perform a content analysis to rankorder the beliefs; and (c) select the 5 to 10 most frequently mentioned items as the salient set.
While the TPB has been supported in a number of domains, including physical activity, and the TPB clearly calls for a salient belief elicitation, additional attention to the belief elicitation step is warranted. The explanatory power of the TPB may be compromised by conceptual and methodological concerns [19] and a main methodological concern for understanding exercise intention and behavior is the limited use of elicitation studies [20]. In fact, a recent review of 150 TPB studies of physical activity showed that only 47 of these studies had conducted an elicitation when applying the theory. Several recent studies have focused explicitly on the elicitation method. Two studies have examined the effect of question order and question wording [21,22]; another study compared individual elicited to modal salient normative beliefs [23]. And, some studies conducted to support intervention design have focused on the results of the elicitation phase [24,25].
While there is considerable evidence that the Theory of Planned Behavior can be used to understand a variety of behaviors, including physical activity, there is less evidence as to the psychosocial factors underlying physical activity among middle school students. Only two of the TPB studies that elicited salient beliefs were conducted with middle school students and none identified the beliefs among Chinese [20]. A study in England of adolescents between 9 and 11 years of age reported salient consequences (i.e., feel healthy and better; fun and enjoyment; make friends; increased risk of injury; and too much effort) as well as social referents (i.e., parents; grandparents; other family members; friends; and teachers) [26]. A study of US adolescents between 14 and 19 years of age found several consequences (i.e., stay in shape; feel healthy; good about self; control weight and diet; increase energy; tired; and pain), normative beliefs (i.e., parents, siblings, friends, coach, and teacher), and control beliefs (i.e., lacking time; money; support; motivation; exercise facts; and access to equipment or activity programs) [27]. Motl and colleagues identified factors in study of American students based on social cognitive theory (i.e., cope with stress; fun; make new friends; keep in good shape; be more attractive; give me more energy; make me better in other sports; and make me hot and sweaty) [28].
To apply TPB to understand the psychosocial determinants of regular physical activity behavior among the Chinese middle school students, qualitative research is a necessary first step. The purpose of this study is to identify the salient beliefs underlying the decision to participate in physical activity held by this particular priority group of middle school students. Specifically, the study will identify the salient consequences of the behavior, the salient referents or social groups, and the salient circumstances or conditions of participating in at least 60 minutes of physical activity every day as perceived by middle schools students living in Beijing, China. The results of this rapid elicitation method can suggest features of interventions as well as provide input to the design of close-ended items for larger scale theory-based quantitative studies.

Methods
A qualitative salient belief elicitation study was conducted as a part of cooperative activities of the Collaborative between Peking University Health Science Center and Indiana University. The primary data collection was approved by the Human Subject Committee at Peking University Health Science Center. The National Institute of Children and Adolescent Health of the Peking University facilitated access to schools through its network with the school systems. Researchers from the Department of Applied Health Science of Indiana University conducted the data analysis. The analysis of the existing data was approved by the Human Subject Committee at Indiana University.

Participants
The 155 study participants were students selected from four classes of two middle schools. In Beijing, middle school or junior high school consists of three grades: Grade 1, 2 and 3. Middle school students are from 13 to 15 years of age. Typically, three, 45-minute physical education classes are offered per week. The classes were selected following a purposive sampling plan by selecting one district in Beijing, two schools within the district, one grade within the school, and two classes within each grade. All students in the selected classes were asked to participate in the study.
More specifically, the research team contacted the School Health administrative agency of Dongcheng district, a district that is average among districts in Beijing in terms of academic grades excluding physical education for their support for this study. The research team contacted administrators from the middle schools in this district to identify study sites. Most schools in the district were not offering physical education courses during the final weeks of the semester when the data were collected. However, it was possible to identify two schools that were offering physical education classes at the time the data were collected and that were willing to participate. To facilitate ease of administration, it was decided to select two classes from Grade 1 in one school and two classes from Grade 2 in the second school. There were six Grade 1 classes in one school and eight Grade 2 classes in the second school. Doctors working in the respective school clinics were used as key informants to classify the classes in each grade into two broad groups (high and low) with respect to the average academic grades (physical education class excluded) in the grade of interest. One class ranking high and one class ranking low was selected from each school. All the students from each of the four classes were invited to participate in the study; all students who were asked agreed to participate in this study.

Instrument
The behavior selected for study was "participating in at least 60 minutes of physical activity every day." Following the WHO definition [29], physical activity was defined as "any activity that increases your heart rate and makes you get out of breath some of the time" and physical activity in physical education class was excluded. The semi-structured self-completion questionnaire began with a cover sheet that described the purpose of the study. Next came close-ended questions that were derived from the Global School-based Health Survey [29] and that assessed demographic characteristics, height, and weight, level of physical activity and inactivity, and information about physical education classes. Finally, several open-ended questions ( Table 1) were used to identify salient consequences, social referents, and circumstances of participating in physical activity.

Procedure
The self-completion instrument was administered during physical education class by two members of the research team who were not associated with the school. Neither the classroom teacher nor the physical education teacher was present during data collection. The facilitators introduced the purpose and value of the study. They emphasized that data collection was anonymous; that there were no right or wrong answers; and that participation was completely voluntary and would have no effect on their grades. The semi-structured questionnaire was completed within 30 minutes. Simple probing was used with all the students, including, for example, "who would care that you always do physical activity every day," "who would disagree with you doing physical activity," and "what are good things if you do physical activity everyday." A group discussion was conducted at the end to get feedback on the students' view of the method.

Content analysis
The purpose of the content analysis was to identify categories of positive and negative outcomes or consequences of performing the behavior, of individuals or social groups who serve as social referents, and of easy and difficult circumstances to perform the behavior. Therefore the content analysis was conducted by two researchers who were familiar with the Theory of Planned Behavior and who could abstract the theoretical constructs from the responses. One of the coders was also a native speaker of Mandarin studying for a PhD in Health Behavior in the United States. The responses to open-ended questions were translated from Chinese to English by the bilingual researcher and then entered verbatim into a word processing file. Similar responses were grouped together to form major categories of responses for each question. These groups of responses were reviewed by two researchers to create a final set of coding categories and to phrase the categories in terms of the appropriate theoretical construct. The bilingual researcher checked with the initial responses in Chinese to ensure cultural and linguistic accuracy. The data were coded into these categories and entered into an SPSS file.

Descriptive analysis
The file of open-ended data was integrated with the file of close-ended data. Students who indicated they had participated in at least 60 minutes of physical activity for four or more days in the past week were classified as doers; those who reported three or few days were classified as nondoers. Cross tabulation analyses were conducted to compare the percentage mentioning each category of salient consequences, social referents, and circumstances separately among doers and non-doers. Chi square analyses revealed no significant differences between doers and non-doers. Therefore the results are presented as the percent mentioning each category of responses over the entire sample of students.

Description of participants
Almost all of the participants (98%) were between 13 and 15 years of age; about half were male and about half were female (Table 2). Thus, the sample was similar to middle school students in Beijing and China with respect to age and gender [6]. Only 14.1% of the middle school students in this study performed at least 60 minutes of physical activities every single day (Table 3). Approximately 86% of this sample of middle school students watched TV two hours or less per day ( Table 3). The mean BMI of students was 19.3 ± 3.7. Almost all the students (97.4%) attended the three physical education classes offered by the school every week. Tables 4 and 5 present the salient consequences of participating in physical activity perceived by these middle school students. Comparing these tables reveals that these students saw more positive outcomes than negative ones. The most frequently mentioned advantage (i.e., "will strengthen my body") was mentioned by over half (58.7%) of the students. Many of the perceived conse-    quences involved physical health outcomes. In fact, the three most frequently advantages of physical activity (i.e., "will strengthen my body", "will improve health", and "will get more exercise") can be viewed as physical health consequences. However, these middle school students also mentioned outcomes beyond physical health (e.g., "will relax me," "will improve my physical education grades," and "will give me energy to study"). The most frequently mentioned disadvantage, "will take too much time," was mentioned by 40.6% of the students. In addition, 24.5% of the students specifically mentioned that participating in physical activity "will take time away from studies" and 36.8% of the students mentioned that participating in physical activity "will make me tired." Salient referents Table 6 shows the individuals and groups mentioned when these students were asked who approved and disapproved of their engaging in physical activity. Clearly, most of the salient referents for this behavior were family members, including parents, mothers, fathers, and grandparents. Teachers and classmates were also mentioned by these middle school students. Doctors and other health professionals were not very frequently mentioned. Parents were the most frequently mentioned approving group and disapproving group and were mentioned more frequently as approving (42.6%) than as disapproving (27.7%). It is interesting to note that when a specific parent was mentioned, mothers were mentioned more frequently than fathers. Tables 7 and 8 present the circumstances which students believed made participating in physical activity at least 60 minutes every day easier and more difficult, respectively. Comparing the two tables, it can be seen that many of the circumstances (e.g., assignments, time, and weather) were mentioned both as circumstances that make physical activity easy and as circumstances that make it difficult. The most frequently reported facilitator of physical activity, "having fewer assignments", was mentioned by 27.7% of the students. The most frequently mentioned barrier, "having too many assignments", was mentioned by about half (48.4%) of the students. Additionally, "time" was the second most frequently mentioned circumstance; 14.2% of the students mentioned having more time as a facilitator and 18.8% mentioned not having enough time as a barrier. These data also suggest that "having fun activities," "having others to participate with," "approval from others" and "making facilities more available" operate as facilitating and hindering circumstances.

Discussion
While a number of the psychosocial factors underlying the decision to engage in physical activity identified among these middle school students from Beijing seem to be similar to those found with students in England and the   [26][27][28], there are some clear differences.
The physical health advantages of feeling healthy, staying in shape, and controlling weight were perceived by the students in the studies from England, the United States, and China. Similarly, the disadvantages of being tired or getting injured were identified in common. Two of the studies, this study and the study of adolescents in the United States [27], elicited easy and difficult circumstances. In both, students mentioned time, motivation, support, and access to facilities as barriers and facilitators. However, several beliefs about the relationship between participating in physical activity and academics seemed to be unique to this sample of middle school students from Beijing. More specifically, "having energy to study" and "improving my grades in physical education" were mentioned as advantages; "taking time away from studying" was mentioned as a disadvantage; and "having fewer assignments" and "having too many assignments" were mentioned as easy and difficult circumstances. In terms of salient social referents, family members were important for these students as well as for the samples of students from England [26] and the US [27]. Not surprising given the one child policy in China, siblings were not mentioned as people who would approve or disapprove in this study of middle school students from Beijing. While more rigorous quantitative research with larger samples is necessary to explore these unique beliefs, these qualitative findings suggests that some of the underlying beliefs found in other populations might not be applicable to the Beijing students and confirms the recommendation to conduct salient belief elicitations in the specific priority group of interest as a first step in formative research to design interventions.
This type of rapid elicitation method to identify the salient consequences, referents, and circumstances of the behavior to be changed with the members of the population of interest can serve two major purposes [24]. Elicitation results are essential to designing close-ended items for theory-based instruments to be used in larger scale quantitative formative research to identify these psychosocial determinants. Results in Chinese are available on request. Following the recommendations of the TPB [17,18], the salient consequences, referents, and circumstances presented in Tables 3 through 8 can be used to generate measures of behavioral beliefs, normative beliefs, and control beliefs. For example, two close-ended items can be created for the salient consequences presented in Tables 4 and 5: one item would assess the strength of the participant's belief that performing the behavior would lead to the consequence; and the other would assess how positive or negative the consequence is perceived to be. For the consequence, "will give me energy to study", the two items in a self-completion format would be (figure 1) In addition, the results of the qualitative study can suggest implications for interventions to increase physical activity among middle school students in Beijing and to provide program planners terminology and words in the language of the population of interest to use in these interventions. Again, while they will need to be confirmed by quantitative research with a larger sample, three related implications might be explored.
First, according to this study, these Chinese middle school students perceived not having enough time and having too many assignments as barriers to daily participation in physical activity. Clearly, these perceptions are consistent with their environment. The Chinese youth are under great pressure to perform well in school and heavy homework loads are typical [30]. Under the parental and societal pressures for academic success manifested in the Chinese education policy changes [5], the Chinese students tend to spend all the time studying after school. A study in Jiangsu province indicated adolescents from 12 to 14 years of age in junior high schools spent 10.8 hours per day on studies (in school and at home) during weekdays [31]. Participation in moderate to vigorous physical activity outside of school among Chinese adolescents is almost non-existent [5]. A study of environmental factors and physical activity [32] in Xi'an city found lack of extracurricular exercise in school and fewer sports meetings were associated with physical inactivity. While promoting the benefits of physical activity to physical health is one approach to increasing physical activity, these data suggest that the issue of physical activity and academic achievement must be also be addressed in prevention interventions. Many students perceived other benefits to participating in physical activity, including increasing energy to study, being fun, and being relaxing. To be effective for Beijing middle school students, an intervention encouraging physical activity might need to address the relationship between physical activity and academic success and achievement.
Sample Close-ended Items for Salient Consequences Figure 1 Sample Close-ended Items for Salient Consequences.
My participation in physical activity at least 60 minutes every day Will give me energy to study.
Third, these findings suggest that prevention interventions address the students' environment. Clearly, the factors of "not having enough time" and "having too many assignments" are factors that are not under the control of most middle school students. While education and communication with students will be important components of prevention interventions encouraging physical activity, to be effective, it is likely that these interventions will need to address the issue of physical activity and academic achievement not just as intrapersonal and individual factors but also as social and environmental factors. Further, given the role of parents and teachers as sources of approval, it will be important to determine the views of these groups when it comes to physical activity and academic achievement. Finally, the findings on access to facilities perceived by these students confirms a study in Xi'an city of China which found that access to community recreational facilities and concerns about safety in the neighborhood were associated with inactivity [30].
Whatever approach is taken to increasing physical activity, policy approaches need to be explored to build supportive school environment. In 1995, the World Health Organization launched the "Global School Health Initiative" which offers a vision for how to develop a school that promotes the health of its students [33]. The experience of developing Health Promoting Schools in Zhejiang province [34], Fujian province [35], Hong Kong [36] and Taiwan [37] all demonstrated the positive changes of school setting, involvement of family, positive behavior changes and learning potential among Chinese students. Therefore, a model of Health Promoting School could be a good approach to promote physical activity. The study has limitations. It is a qualitative study that is meant to provide information useful for a more rigorous and larger scale quantitative study. Thus, the data are responses to open-ended questions and the opportunity for quantitative comparisons is limited. The sample size is small and based on a convenience sample from one school district in Beijing. It examined the perceptions of physical activity behavior held by only middle school students. Thus, the results from this study are more suggestive than affirmative.

Conclusion
The results of this qualitative study suggest that student perceptions of the relationship between their participation in physical activity and their school work might be more important factors underlying their decisions to participate in physical activity than their beliefs about the health benefits. Programs to increase physical activity should address the social and environmental factors underlying these perceptions with the goal of strengthening student engagement in the physical activity and improving learning potentials. Quantitative studies with a larger and representative sample and with close-ended items based on the qualitative research are needed to more fully understand middle school students' decisions to engage in daily physical activity. In addition, given the role of the parents as a social referent and the students' perception of the connection between physical activity and academic performance, research is needed to understand the views of parents, teachers, and school administrators.