Motivation, injury prevention, and the incidence of sports injuries: A three‐wave longitudinal test of self‐determination theory

This study applied self‐determination theory (SDT) as a psychological framework to examine whether psychological need support and autonomous motivation are predictive of sports injury preventive behaviors and the incidence of sports injuries. 2042 secondary school students (mean age = 14.33, male = 44.3%) from China completed a survey of the study variables (using established scales) at three time points (baseline, 1‐month follow‐up, 3‐month follow‐up). Structural equation modeling examined the core tenets of SDT by testing if the change‐scores of the SDT variables (i.e., psychological need support from PE teachers, students' motivation, and students' behavioral adherence) between baseline and 1‐month follow‐up, were predictive of sports injury incidence assessed at 3‐month follow‐up. Our model demonstrated acceptable goodness‐of‐fit parameters (CFI = 0.97, TLI = 0.96, RMSEA = 0.03, and SRMR = 0.05). The relationships between psychological need support, autonomous motivation, and behavioral adherence were both positive and significant. These SDT variables predicted the future incidence of sports injuries following the motivational pathways of SDT. Our study provides evidence of the predictive power of SDT variables on sports injury preventive behaviors and the incidence of sports injuries: Students who perceive their PE teachers as psychological need supportive possess higher autonomous motivation and behavioral adherence towards sports injury prevention, and are also less likely to encounter sports injuries in the future.


Funding information
Building World-Class Universities (disciplines) of Renmin University of China, Grant/Award Number: KYGJC2020003 This study applied self-determination theory (SDT) as a psychological framework to examine whether psychological need support and autonomous motivation are predictive of sports injury preventive behaviors and the incidence of sports injuries. 2042 secondary school students (mean age = 14.33, male = 44.3%) from China completed a survey of the study variables (using established scales) at three time points (baseline, 1-month follow-up, 3-month follow-up). Structural equation modeling examined the core tenets of SDT by testing if the change-scores of the SDT variables (i.e., psychological need support from PE teachers, students' motivation, and students' behavioral adherence) between baseline and 1-month follow-up, were predictive of sports injury incidence assessed at 3-month followup. Our model demonstrated acceptable goodness-of-fit parameters (CFI = 0.97, TLI = 0.96, RMSEA = 0.03, and SRMR = 0.05). The relationships between psychological need support, autonomous motivation, and behavioral adherence were both positive and significant. These SDT variables predicted the future incidence of sports injuries following the motivational pathways of SDT. Our study provides evidence of the predictive power of SDT variables on sports injury preventive behaviors and the incidence of sports injuries: Students who perceive their PE teachers as psychological need supportive possess higher autonomous motivation and behavioral adherence towards sports injury prevention, and are also less likely to encounter sports injuries in the future.

K E Y W O R D S
autonomous motivation, physical activity, physical education, psychological need support, sports injury prevention, sports safety of injury prevention programmes depends upon individuals' long-term commitment to these measures, [3][4][5] and most participants do not comply with the prescribed programmes 6 due to barriers including insufficient knowledge, 7 time constraints, lack of interests, 8 and lack of facilities and resources. 9 Research showed that interventions that made specific effort to address these needs could be can be more easily accepted by coaches and smoothly implemented in organizations. 9,10 It is, therefore, important that research seek a better understanding on how the psychological factors (e.g., motivation) and psychosocial environments (e.g., support from coaches/PE teachers) could be optimized for preserving individual's adherence to injury prevention. In this study, we applied selfdetermination theory (SDT) to investigate if psychological need support and motivation predict adherence to sports injury prevention, as well as the incidence of sports injuries among physical education (PE) students.

| Self-determination theory
SDT posits that motivation provides the meanings that drive an individual's behaviors. 11 When an individual selfinitiates behavior, and perceives the action as coherent with their own interests, core values, and life goals, he/ she is believed to be driven by autonomous motivation. In contrary, when an individual's actions stem from external factors (such as reward or punishment), or introjection (such as with desire for others' approval), he/she is driven by controlled motivation. Individuals driven by controlled motivation are often accompanied by a feeling of being pressured or forced into behaviors; whereas autonomously motivated individuals behave with a sense of freedom and choice, and are thus more likely to persist in healthy behaviors and experience associated positive outcomes than those with controlled motivation. To promote autonomous motivation in a social environment, PE teachers can support the basic psychological needs of students by the provision of choice, explanation of the rationales behind the recommended actions, respect of personal values and feelings, care and acceptance, and encouragement. The tenets of SDT, including the relationships between psychological need support, motivation, and behaviors, have been tested in a wide range of health contexts, including occupational injury prevention 12 and sports injury rehabilitation. 13 Previous studies that applied SDT in sports injury contexts have provided supportive evidence to the tenets of the theory. [14][15][16] It was generally found that the psychological need support from significant others, such as coaches and PE teachers, was associated with a pattern of higher autonomous motivation and lower controlled motivation. This in turn linked to better beliefs, intention, and behavioral adherence in sports injury prevention 3,15,16 and other injury management-related behaviors. 13,17 The psychological pathway of SDT has been supported by evidence from cross-sectional studies and longitudinal studies among professional athletes as well as PE students. 3,15,16 In particular, a need supportive PE teacher not only facilitated students' autonomous motivation within PE lessons, but also in out-of-school contexts, and increased their behavioral adherence in both circumstances. 3 Recently, Lee and colleagues 5 applied concepts from SDT to develop an intervention to promote sports injury preventive behaviors among secondary school PE students. The randomized controlled intervention increased the psychological need support from PE teachers, as well as students' autonomous motivation and behavioral adherence to sports injury prevention. Overall, evidence suggests that SDT is a viable framework to explain the psychosocial and motivational antecedents of adherence in sports injury prevention.

| Knowledge gaps
Despite the robust evidence of SDT in the prediction of health behaviors, meta-analysis revealed that there is a lack of empirical research on the correlation between SDT and physical health outcomes. 18 This is because existing studies only examined behavioral adherence or intention as the final behavioral outcome, without including a prospective assessment of sports injury incidence. As such, these studies were unable to provide any evidence regarding the associations between SDT variables and sports injury outcomes. That is, the predictive value of SDT variables on sports injury remains unknown. It is important to address this gap in the research by including a prospective health outcome. This will reveal how the motivational pathways of SDT are linked to the incidence of injury.

| The present study
In this study, we examined the psychological pathways of SDT in the context of sports injury prevention among secondary school PE students. To address the limitations of previous studies, we included longitudinal assessments of behavioral adherence to sports injury prevention and sports injury incidence. With a three-wave longitudinal design (T1 (baseline), T2 (1-month follow-up), and T3 (3month follow-up)), we were able to test if the changes of the SDT variables (i.e., psychological need support from PE teachers, students' motivation, and students' behavioral adherence) between T1 and T2, were predictive to sports injury incidence assessed at T3.
Based on the tenets of SDT and previous literature in SDT and sports injury prevention, the following hypotheses were proposed: H1. Psychological need support from PE teachers would form a direct positive association with students' autonomous motivation of sports injury prevention (H1a), and an indirect positive association with students' behavioural adherence to sports injury prevention (H1b). H2. Students' autonomous motivation would be positively associated with their behavioural adherence of sports injury prevention. H3. Incidence of sports injury would be negatively predicted directly by students' behavioural adherence to sports injury prevention (H3a), and indirectly by students' autonomous motivation of sports injury prevention (H3b), and psychological need support from PE teachers (H3c).

| METHOD
The dataset of this study was taken from a sub-sample of a previous study 19 that examined the temporal relationships between autonomous motivation and social cognitive constructs. Our present study is original because we preliminarily investigated the relationship between psychological need support, autonomous motivation, and behavioral adherence at a change-score level, and whether the change-scores were predictive of future incidence of sports injuries.
Our study examined unique hypotheses that could provide novel evidence regarding the predictive power of SDT on future health outcomes related to sports injuries. The original sample comprised 3952 students (mean age = 14.33, SD = 1.92; male = 44.28%) in 45 secondary schools located in Beijing, China. They were asked to complete the survey of our study at three time-points: T1 (baseline), T2 (1-month follow-up), and T3 (3-month follow-up). Ethical approval was obtained from the Institutional Review Board of The University of Hong Kong (ref: . Participants were included in the final analysis if they (1) stated that they participated in sports/ exercises in-school or during leisure-time (84.46%, n = 3338), and (2) completed the assessments during T1 and T2 (55.41%, n = 2190). The former criterion was applied to ensure that sports injury prevention was relevant to the students, and the latter to allow the calculation of the change-scores between T1 and T2, which were fundamental to testing the study hypotheses. As such, the final sample included in our analysis consisted of 2042 PE students from 40 secondary schools, mean age = 14.11, SD = 1.91; male = 43.9%, see Table 1. Informed consent forms were signed by participants and their parent or legal guardian if under the age of 18 before the start of the baseline survey to ensure that they understood the research aims and their participation rights. The attrition rate at T3 was 22.77% within our final sample. In our sample, participants spent an average of 5.13 (SD = 4.85) hours weekly in doing sports or other physical activities, for instance, running (34.76%), playing basketball (15.53%), soccer (7.76%), and badminton (4.67%). Six months prior to the study, participants experienced an average of 1.31 (SD = 3.95) sports injuries that led to suspension of sports participation for 3.69 (SD = 12.0) days. The most commonly reported sports injuries included fractures (22.00%), sprains (18.04%), and strains (13.12%).
We used an adapted short-form Health Care Climate Questionnaire to measure the students' perceived psychological need support for sports injury prevention from PE teachers. The scale was originally developed to assess the degree of perceived psychological need support from significant others for health-related behavioral changes, 20 and was adapted to measure the perceived need support for sports injury prevention from coaches and PE teachers. 3,15 We used the autonomous motivation subscale of Treatment Self-Regulation Questionnaire 21 and the Self-Reported Injury Prevention Adherence Scale 15 to evaluate students' autonomous motivation and behavioral adherence (such as stretching, physical conditioning, and adequate rest) in sports injury prevention respectively. All Chinese-translated scales adapted for sports injury context exhibited satisfactory reliability and validity in previous studies. 3 anchors of all the scales used in the current study. In addition to the key study variables, our survey also collected demographic and self-reported sports-related data, including age, gender, sports participation, and the total number of sports injuries (regardless if they occurred during PE lessons or not) sustained in the past month and past 6 months. Sports injuries are defined as any physical damage caused by physical activities, irrespective of whether the injuries required medical attention or time loss in sports. The descriptions of such injuries were further examined to ensure that the reported incidents are relevant (e.g., ankle sprain during football practice). According to the guidelines from the Ministry of Education, PE teachers in China are responsible for minimizing students' risks of sports injuries and teaching general sports injury prevention (e.g., stretching, warming up, proper equipment usage). 22 Additionally, prior research showed that when injury prevention was taught in a need-supportive manner, students' autonomous motivation could be transferred from in-school to out-ofschool context. 3 Therefore, we argue that PE teachers are important not only to students' injury preventive behaviors during PE lessons in school, but also during leisure physical activity in an out-of-school environment.
To examine the association between study variables at a change-score level, we entered the standardized residual change score of each question as indicators of SDT constructs and behavior adherence. The item-wise standardized residual change-scores of the psychological variables and behavioral adherence were calculated by regressing the follow-up response on the baseline response. 23 The fit of the hypothesized model was evaluated with structural equation modeling, which was performed in Mplus version 8.1 with a robust maximum likelihood estimation method. Mediation analysis was also employed to examine the indirect effects between SDT variables and sports injury outcomes. Missing values (0.10%-24.28% across the study variables and time-points) in the final sample caused by attrition and item-nonresponse were handled by fullinformation maximum likelihood method. The overall model fit was assessed by various goodness-of-fit indices, including the Root Mean Square Error of Approximation (RMSEA), the Comparative Fit Index (CFI), the Tucker-Lewis Index (TLI) and the Standardized Root Mean Square Residual (SRMR). The conventional cut-off values: CFI and TLI values greater than 0.90 and RMSEA and SRMR values less than 0.08, were used to evaluate whether the proposed model has an acceptable fit. 24

| RESULTS
Cronbach's reliability test found all the study variables at T1 and T2 had a satisfactory internal consistency (α = 0.79-0.95). Demographic and sports descriptive variables including age, gender, length of sports participation per week, and sports injury rate in 6 months during baseline were found to be correlated with the study variables and thus were added as covariates in the model and mediation analysis. Table 2 shows the zero-order correlation matrix, descriptive statistics, and internal reliability coefficients of the study variables.
The incidence of sports injury at T3 was negatively predicted by students' behavioral adherence to sports injury prevention (H3a, β = -0.07, p = 0.02). Further mediation analysis revealed that sports injury preventive behavior adherence mediated the negative effect of autonomous motivation on the incidence of sports injury, while students' autonomous motivation and behavioral adherence to sports injury prevention fully mediated the negative relationship between psychological need support from PE teachers and sports injury incidence. This indicated that the incidence of sports injury can be indirectly predicted by students' autonomous motivation (H3b, β = -0.04, p = 0.02) and psychological need support from PE teachers (H3c, β = -0.03, p = 0.02). Tables 3 details the results of the mediation analyses.

| DISCUSSION
Our findings support the motivational sequence of SDT to explain how students' sports injury preventive behaviors and subsequent injury outcomes can be predicted by psychological need support and autonomous motivation. In particular, we found significant and positive relations between (1) psychological need support and autonomous motivation; (2) autonomous motivation and preventive behavior adherence; and (3) significant and negative relation between preventive behavior and sports injury occurrence. Collectively, our findings provide initial evidence for SDT constructs in the prediction of students' injury prevention adherence and sports injuries.
Consistent with our hypothesis (H1a), psychological need support from PE teachers in sports injury prevention was significantly and positively associated with PE students' autonomous motivation in sports injury prevention. This finding is congruent with prior research which found similar relationships between SDT constructs in sports injury prevention behavior among high school students 3 and athletes. 25 PE teachers who use a need-supportive teaching style in sports injury prevention educationsuch as by providing choices, encouraging self-directed learning, and sharing student perspectives in their needs and interests-are more likely to create a self-controlled environment for students that fosters their autonomous motivation to adhere to sports injury preventive practices in the long run. 26 Students' autonomous motivation was also significantly and positively associated with their behavioral adherence to sports injury prevention. This was in line with T A B L E 2 Descriptive statistics, zero-order correlations and reliability coefficient of the variables (N = 2042).

F I G U R E 1
Path estimates for the self-determination theory model of sports injury prevention. The paths associated with the control variables (i.e., age, gender, sports participation per week and incidence of sports injury in 6 months reported during baseline) are omitted. *p < 0.05, two-tailed. **p < 0.001, two-tailed.
our hypothesis (H2a) and prior research findings, in which elite athletes with a heightened self-determined form of motivation (ratio of autonomous and controlled motivation) engaged in sports injury prevention measures with more frequency and effort. 15 Our findings also partly echo studies which found that autonomous motivation was indirectly related to the sports injury preventive behavior via social cognitive constructs such as attitude, subjective norms, perceived behavioral control, and intention, 3 which represent more proximal psychological antecedents to actions. Additionally, autonomous motivation served as a significant mediator between psychological need support and behavioral adherence of sports injury prevention, and demonstrated the importance of supportive significant others in the injury prevention context. In other words, PE teachers' psychological need support indirectly encourages students to adopt sports injury prevention behaviors by increasing their autonomous motivation. This full mediation aligns with SDT and provides a plausible mechanism to target students' autonomous motivation to promote sports injury prevention. Unsustainable adherence to sports injury prevention observed in previous research could be partially accounted by the inadequate support provided by coaches/PE instructors, who might have forgone sports injury prevention due to time constraints. 27 On the other hand, when PE teachers were trained to adopt a need-supportive teaching style, students were more autonomously motivated and engaged in injury prevention more enthusiastically. 5 Inclusion of staff training in sports injury prevention intervention might potentially increase students' autonomous motivation in the long term.
Last but not least, the number of sports injuries in the future was negatively predicted directly by students' behavioral adherence to sports injury prevention (H3a), and indirectly by students' autonomous motivation of sports injury prevention (H3b), and psychological need support from PE teachers (H3c). The SDT model explained 21.6% variance in sports injuries occurrences. Our results align with various studies showing that sports injury prevention programmes could minimally to moderately reduce sports injuries in participants with higher adherence to the programme. 28,29 More importantly, we provided initial evidence for the predictive value of SDT constructs in sports injury reduction. Our findings may imply that sports injury prevention should not only aim at providing proper guidelines and education to PE teachers to enable them to create a need-supportive environment (see Emery et al. 30 ). Although SDT concepts have been incorporated into physical activity promotion during PE lessons, 31,32 they are rarely applied in the promotion of sports injury prevention. In most Chinese high schools support for sports injury prevention is sparse, and taught mostly through classroom-based lectures. 33 This might be the reason why some students tend to perform sports injury prevention with controlled motivation (e.g. complying with the safety regulation or compulsory injury prevention routine) and amotivation (e.g. following warm-up instructions without reasons). 34 As coaches' perception of the efficacy of injury prevention programs is the main determinant for program uptake, 27 our findings supporting the linkage between psychological need support provided by PE teachers and significant health benefits might serve as an evidence for convincing schools to implement SDTbased training and prevention programmes. Having said that, such programmes will need to cater to not only students' needs, but also the teachers' needs to ensure their success in long-term implementation. Interventions that provide ongoing support and respect teachers' time, such as app-based daily reminders, 5 will likely to be more well perceived by the schools.
Our current study has a few limitations. Firstly, our final sample only consisted of participants who completed the assessments at T1 and T2; this may be selective and potentially introduce sampling bias. A separate analysis was conducted with the imputed data from the full dataset after missing values were handled by multiple imputation. This method allowed change-score calculation before model evaluation and was shown to be quite robust even in datasets with a notable proportion of missing values. 35 The presented results were highly consistent with the model results of imputed data. Note: Need Support = psychological need support; Sports Injury = Incidence of sports injury at T3. *p < 0.05, two-tailed. **p < 0.01, two-tailed. ***p < 0.001, two-tailed.
T A B L E 3 Results from the mediation analysis of the self-determination theory model in sport injury prevention.
Secondly, our study consisted of only self-reported measures, which are subject to various response biases such as response order effects 25 and recall bias. 36 Further research should consider the use of more objective data collection methods, such as teacher-reported/medicalreported injury prevention adherence 28 and injury incidence 29 or independent observation methods. 31 Finally, given the correlational nature of our study, we could not imply the absolute causal relationships between SDT, preventive behavior, and injury variables. Further investigation should consider the use of interventional or randomized control trial designs to better evaluate the effect of improved psychological need support on the changes in sports injuries.

| Perspective
Our study applied SDT to sport injury prevention behaviors and sport injury occurrence in secondary school students. Results demonstrated that PE teachers' psychological need support and students' autonomous motivation can explain behavioral adherence to sports injury prevention, and also the future occurrence of sports injuries. When PE teachers promote sports injury prevention in a psychological need supportive manner, students are more likely to endorse higher autonomous motivation and adherence to sports injury prevention and consequently, sustain fewer sports injuries. Therefore, SDT appears to be a robust framework that could be useful for the development of a theory-driven sports injury prevention programme for PE students. The current model extends the prior findings from Lee et al. 3 by including a prospective injury outcome, which further consolidates the evidence for promoting the use of SDT-based sports injury prevention education among school teachers. The findings from the present study is also encouraging for future interventional study to include injury-related outcomes for measuring real-world implication.