Classroom-embedded two-phase add-on program for fall-injury prevention in rural primary schools

Abstract

Background

Falls are a leading cause of injury among Chinese children, especially in rural areas. We evaluated a classroom-embedded, two-phase add-on school program integrating Health Belief Model (HBM)-guided education with supervised in-class balance training in rural Shantou, China.

Methods

We conducted a two-school, controlled before-and-after evaluation in rural Shantou with school-level allocation. The intervention school received monthly HBM-guided education during months 1–12 and supervised, zero-equipment balance training delivered twice weekly during months 7–12; the comparison school continued usual health education.The primary outcome was the rate of fall-related injuries during the 12-month period. Secondary outcomes were the proportion of students with ≥ 1 fall-related injury and balance performance. Difference-in-differences (DID) compared pre-post changes in balance between groups with 95% CIs.

Results

We analyzed 381 students (intervention = 198; control = 183; mean age 10.07 [SD 1.39] years; 195 [51.18%] boys). Adjusted analyses showed a lower rate of fall-related injuries in the intervention versus control group (IRR 0.54, 95% CI 0.29–0.98). The proportion of students with ≥ 1 fall-related injury did not differ significantly. Balance improved more in the intervention group, the DID showed a 1.94-second increase (95%CI, 0.18 to 3.70) in eyes-closed marching-in-place time, a 0.62-second decrease (95%CI, -1.22 to -0.02) in balance-beam time, and a 0.43-point increase (95%CI, 0.003 to 0.87) in one-leg standing with eyes closed.

Conclusion

A 12-month classroom-embedded two-phase add-on program combining HBM-guided monthly education with supervised in-class balance training reduced the rate of fall-related injuries and improved balance among primary schoolchildren.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13690-026-01974-4.

Text box 1. Contributions to the literature
• This study addresses a lack of evidence on how to prevent fall injuries in children attending rural, low-resource primary schools.
• It shows how a simple, theory-informed add-on program can be woven into regular classroom time without extra staff or special equipment.
• The work offers a practical model that combines short safety lessons with balance activities that teachers can deliver themselves.
• The findings suggest a scalable approach that education and health sectors can include in school health policies to reduce childhood injuries and future health burdens.

Background

Falls are the second leading cause of unintentional injury deaths globally, resulting in an estimated 684,000 fatalities and 38.2 million disability-adjusted life years (DALYs) annually [1, 2]. Although most of these falls occur in older adults, children are another high-risk group and account for nearly 40% of fall-related DALYs worldwide [1]. Multinational data indicate that falls are a top cause of all injury-related hospital visits among children aged 5–19 years [3–6]. The incidence is consistently higher in rural than in urban children and adolescents. Although not all falls result in serious injury, they impose a substantial healthcare burden, including treatment of related injuries, hospital admissions, and rehabilitation [7–10].

Immaturity of children’s neuromuscular (and broader biological) development is an important intrinsic risk factor for falls, making children more prone to falling than adults. This immaturity is characterized by deficits in muscle strength and postural control, with many basic motor skills still maturing. Children aged 6–12 represent a sensitive/critical period for motor-skill acquisition and development [11, 12]. Therefore, school-based interventions should be designed to reduce fall risk by enhancing neuromuscular function (e.g., dynamic coordination, balance, and strength) while integrating fall-prevention knowledge and skills. Evidence indicates that balance training can improve balance in children and adolescents and is used across diverse populations, including first-grade students [13], children with cerebral palsy [14], professional athletes [15], and older adults [16–18]. However, within injury prevention, balance-training research has focused largely on older adults, athletes, and special populations, with relatively few targeted interventions aimed at preventing falls among children.

Effective behavioral interventions integrate empirical evidence on behavioral determinants with theory-driven frameworks for health behavior change. Evidence suggests that theory-based interventions are more effective than those lacking a theoretical foundation [19]. Although early injury-prevention efforts often lacked such grounding, recent work shows progress: the Health Belief Model (HBM) is now widely used in exercise-related injury-prevention programs, informing intervention design, implementation, and evaluation [20]. The HBM provides a systematic approach to understanding individuals’ health beliefs, risk perceptions, cognitions, and decision-making, which is critical for preventing injuries such as falls [21]. In balance-focused interventions, applying the HBM helps clarify the harms of falls and the value of preventive actions, thereby strengthening health beliefs and behaviors. Accordingly, using the HBM as the guiding framework may improve the effectiveness of fall-prevention strategies.

In recent years in China, provincial and municipal government agencies, health-care institutions, and research institutes and universities have implemented child-injury prevention programs of varying scale. However, development has been uneven across regions [22]. Most programs are concentrated in economically more developed urban areas, whereas rural areas with less developed economies and heavier injury and disease burdens have seen limited activity [23]. Much of the available evidence comes from high-income regions and may not generalize to low-income settings, where the burden of childhood falls is greatest. Moreover, during the COVID-19 pandemic, the majority of schools faced significant disruptions, including campus closures and limitations on group activities, making it challenging for most programs to continue delivering interventions or evaluating their effects. To overcome these challenges, we developed and implemented a two-phase add-on intervention designed to minimize physical contact. This intervention involved continuous HBM-based education over months 1–12, delivered to students, teachers, and caregivers, followed by the addition of supervised balance training during months 7–12.

In light of these challenges, this study focuses on rural primary school students in Shantou, China, and evaluates a classroom-embedded two-phase add-on program integrating an HBM-guided education component with supervised in-class balance training, implemented during the COVID-19 period. The aim is to generate pragmatic evidence for scalable school-based fall-injury prevention under real-world constraints.

Methods

Design and participants

In rural Shantou, we first randomly selected two towns from strata with comparable socioeconomic and educational profiles. From each town, we randomly selected one public primary school with a similar enrollment size, yielding two rural schools as study sites. The two schools were then randomly assigned in a 1:1 ratio to the intervention or control condition. Within each school, we used grade-stratified cluster sampling to randomly select two intact classes per grade and invited all students in those classes to participate. Outcome assessors were blinded to school allocation and followed standardized protocols for all measurements. The study was conducted from September 2021 to August 2022. The study originally planned to include more than two schools, but COVID-19 related restrictions curtailed school access and only two schools could participate.

To reinforce the HBM component and support data collection, parents/caregivers of participating students were also invited to attend brief education sessions and complete questionnaires. Assessments were conducted at baseline and at 12-month follow-up. Eligibility criteria were as follows. Inclusion: (a) enrolled at one of the selected primary schools at baseline and expected to remain at the same school for the 12-month follow-up; (b) written parental/guardian consent; (c) ability to understand and follow instructions and to complete testing/training in Mandarin or the local dialect. Exclusion: (a) medical contraindications to physical activity as assessed by a physician (e.g., uncontrolled epilepsy, severe cardiopulmonary disease, fracture or surgery within the previous 3 months, acute vertigo/vestibular disorders, or severe spinal/lower-limb deformities precluding safe participation); (b) uncorrected severe visual or hearing impairment that could compromise safety or comprehension of instructions; (c) severe intellectual or behavioral problems that could compromise group safety. This study was approved by the Ethics Committee of Shantou University Medical College (SUMC-2021-36). Written informed consent was obtained from the parents or legal guardians of all participating students, and written consent was also obtained from the students themselves prior to participation.

Intervention programme

Delivery and implementation

We used a two-phase add-on intervention in which HBM-based education ran throughout months 1–12, with supervised classroom balance training added during months 7–12.Trained assessors conducted baseline evaluations at month 0, including demographics, fall-injury history, and standardized balance tests. Outcome assessors were blinded to group allocation. The intervention school received an HBM-based fall-prevention curriculum delivered in a blended format: on-site activities for students and teachers (pamphlets, brief lectures, video demonstrations, and supervised balance-exercise classes) were complemented by WeChat articles and short videos for caregivers. Educational sessions were provided once per month (20–30 min) across 12 months. Educational sessions were provided once per month for 20–30 min across 12 months. To allow HBM-based education during months 1–6 to strengthen students’risk perception and self-efficacy before training, balance training was scheduled for months 7–12. Balance training consisted of supervised in-class sessions delivered twice per week for approximately 30 min per session. During the intervention period, monthly follow-ups recorded fall events using a 30-day retrospective method, and adherence was assessed based on attendance records for balance training. Endline assessments at month 12 replicated baseline measures. The control school continued routine health education (e.g., prevention of common infectious and intestinal diseases) and standard school calisthenics; no structured balance-training curriculum or progression was introduced. Session timing and contact time were matched as far as feasible to minimize confounding.

Development of balance exercises

The balance routine was designed to enhance postural control while prioritizing safety, simplicity, and developmental appropriateness. Design principles included progressive difficulty (from stable to more challenging tasks), symmetry when feasible, no equipment, and brief, repeatable bouts organized into 4-8-count segments (Supplementary 1). To accommodate developmental differences, two grade bands were used (≤ Grade 3 vs. ≥Grade 4). The core choreography comprised a preparatory section and a sequence of balance tasks that engaged major body segments and varied base-of-support and visual demands. For example, lateral stepping with coordinated arm raises and partial knee flexion, single-leg extensions and holds in the frontal plane, hip-flexion tasks with arms moving from abduction to overhead positions, forward/side/back leg extensions with hands on hips, combined step-knee-lift-turn sequences, running-in-place with alternating knee-lift and kick components, and a marching/closing sequence. Content validity and safety were established through an online expert panel review with three China-based sports experts, each with ≥ 5 years of relevant experience and documented professional scholarship or certification. Panel feedback informed targeted refinements, including simplification of difficult elements, grade-banded routines, structured warm-up/cool-down, in-class safety guidance with error correction, and caregiver-facing instructional videos.

Outcome variables

The survey comprised three components: basic characteristics (name, sex, class, age, height, weight), falls in the previous 12 months, and balance performance. The primary outcomes were the rate of fall-related injuries during the 12-month period. A fall is defined as an event which results in a person coming to rest inadvertently on the ground or floor or other lower level [1]. A fall-related injury was defined as an injury attributable to a fall meeting either of the following [24, 25]: (i) evaluation or treatment at a medical facility due to the fall, or (ii) absence from school or rest for ≥ 1 day because of the fall.

Secondary outcomes were the proportion of students with ≥ 1 fall-related injury and balance performance (dynamic and static balance). Dynamic balance was assessed with the eyes-closed marching-in-place test and the balance-beam test. For the marching test [26], participants stood within a 40-cm-diameter circle and, on an auditory cue, marched with eyes closed at 120 steps/min. Timing started at the cue and stopped when either foot touched the line or left the circle, and the elapsed time (s) was recorded. For the balance-beam test [27], participants started from one end of a 3-m beam, walked to the far end and returned after the cue; round-trip time (s) was recorded with eyes open and unrestricted arm movements. If a foot touched the floor, the child stepped back onto the beam at the point of descent and continued. Static balance was assessed with single-leg stance (eyes open and eyes closed) and the tandem Romberg (eyes closed). For single-leg stance [28], the duration (s) the dominant limb could be maintained before the non-dominant foot touched the ground was recorded under eyes-open and eyes-closed conditions. For the tandem Romberg [29], participants stood heel-to-toe with eyes closed and were timed until balance was lost (observable sway/step). For each static and dynamic test, two trials were performed, and the best result was retained. Outcome assessors were blinded to allocation, and all timings were recorded with a handheld stopwatch.

Adverse events

Trained assessors recorded any adverse events related to the balance-training intervention (e.g., dizziness, sprain/strain, falls, notable muscle soreness) reported by students.The control school was not queried for adverse events.

Sample size

The sample size was estimated using G*Power (version 3.1). Power calculations indicated that 194 participants (97 per group) were needed to achieve 80% power (two-sided α = 0.05) to detect a 20% decrease in fall incidence with equal allocation. Allowing for a 10% dropout rate, the target sample was 216 participants (108 per group). Ultimately, 381 students were analyzed (198 in the intervention group and 183 in the control group).

Data analysis

Baseline characteristics were summarized as mean (SD) values or frequencies (percentages) and compared using analysis of variance or the χ² test, as appropriate. For the rate of fall-related injuries, the data were counted data with overdispersion, so negative binomial regression was used instead of Poisson regression, and incidence rate ratios (IRRs) with 95% CIs were reported. The 12-month proportion of individuals with ≥ 1 fall-related injury was analyzed using modified Poisson regression with robust standard errors, yielding risk ratios (RRs) with 95% CIs. For both endpoints, we reported unadjusted estimates and estimates adjusted for age, sex, BMI, baseline fall history, participant follow-up time, and adherence. Given the two-school cluster randomization, all analyses accounted for intra-school correlation using cluster-robust standard errors. For balance function outcomes, descriptive statistics were summarized as mean (SD) values or median (IQR) values, and compared using either two-sample t-tests, paired t-tests, two-sample Wilcoxon rank-sum tests, or the Wilcoxon signed-rank test for paired samples, as appropriate. Finally, the difference-in-differences (DID) approach was applied to evaluate the impact of the balance exercise intervention on the balance ability of elementary school students. Primary injury outcomes had no missing data, and missing secondary balance outcomes were imputed using multiple imputation. All statistical analyses were performed using R 4.5.1, and statistical significance was set at a 2-sided P < 0.05.

Results

Descriptive statistics

A total of 420 students were assessed for eligibility and consented, and 405 were enrolled. At 12 months, 381 completed both assessments and were included in the primary analysis (198 intervention, 183 control); 24 were lost to follow-up (11 intervention, 13 control). Baseline characteristics were balanced between completers and those lost to follow-up. Further details are shown in Fig. 1. The mean (SD) age was 10.07 (1.32) years in the intervention group and 10.07 (1.47) years in the control group, and the proportion of boys was 53.5% and 48.6% respectively. Age, sex, and other baseline characteristics did not differ significantly between groups (all P > 0.05, Table 1).

Fig. 1.

Flow of participants through the trial

Table 1.

Participant Characteristics at Baseline

Characteristic	Total (n = 381)	Group		P value
		Intervention group (n = 198)	Control group (n = 183)
Age, mean (SD), y	10.07 (1.39)	10.07 (1.32)	10.07 (1.47)	1.000
Sex, No. (%)				0.339
Male	195 (51.18)	106 (53.54)	89 (48.63)
Female	186 (48.82)	92 (46.46)	94 (51.37)
Grade, No. (%)				0.647
Grade 3 and below	167 (43.83)	89 (44.95)	78 (42.62)
Grade 4 and above	214 (56.17)	109 (55.05)	105 (57.38)
Only-child status, No. (%)				0.542
Only child	28 (7.35)	13 (6.57)	15 (8.20)
Non-only child	353 (92.65)	185 (93.43)	168 (91.80)
Height, mean (SD), cm	136.87 (9.96)	135.99 (9.69)	137.84 (10.20)	0.070
Weight, mean (SD), kg	31.22 (8.39)	31.41 (8.72)	31.02 (8.03)	0.652
BMI, mean (SD), kg/m2	16.47 (3.00)	16.73 (3.13)	16.17 (2.83)	0.069
Fall history, No. (%)	56 (14.70)	27 (13.64)	29 (15.85)	0.543

Abbreviations: BMI Body Mass Index, SD Standard Deviation

Primary outcomes

Rate of fall-related injuries

At baseline (Pre-Test), the number of fall-related injury events was 32 in the intervention group and 34 in the control group. During follow-up (Post-Test), events totaled 17 vs. 28, respectively. Negative binomial models indicated a lower event rate in the intervention group: unadjusted IRR = 0.56 (95% CI, 0.30 to 1.01; P = 0.060) and adjusted IRR = 0.54 (95% CI, 0.29 to 0.98; P = 0.048), controlling for age, sex, BMI, baseline fall history, follow-up time, and adherence (Table 2).

Table 2.

Association of the Intervention With Fall-Related Injury Outcomes

Group	Unadjusted IRR/RR (95% CI)	P Value	Adjusted IRR/RR (95% CI)	P Value
Rate of fall-related injuries	0.56 (0.30 to 1.01)	0.060	0.54 (0.29 to 0.98)	0.048
Proportion of fall-related injuries	0.58 (0.31 to 1.07)	0.081	0.59 (0.32 to 1.10)	0.097

Analyses included variables of age, sex, BMI, baseline fall history, participant follow-up time and adherence. Rate of fall-related injuries was analyzed with negative binomial regression, and proportion of of fall-related injuries with modified Poisson regression

Abbreviations: IRR incidence rate ratio, RR relative risk, CI confidence interval

Proportion of individuals with fall-related injuries

At baseline, the proportion with ≥ 1 fall-related injury was 13.64% in the intervention group and 15.85% in the control group. At follow-up, the proportions were 7.58% and 13.11%, respectively. In modified Poisson regression, the between-group difference at follow-up was not statistically significant (unadjusted RR, 0.58; 95% CI, 0.31 to 1.07; P = 0.081; adjusted RR, 0.59; 95% CI, 0.32 to 1.10; P = 0.097), with adjustment for age, sex, BMI, baseline fall history, follow-up time, and adherence (Table 2).

Secondary outcomes

Dynamic balance tests

At follow-up, dynamic balance improved in the intervention group. Eyes-closed marching-in-place time increased from 7.57 s at baseline to 10.50 secondsat follow-up (P < 0.001). Balance-beam traversal time decreased from 7.59 s to 6.89 s (P < 0.001). Changes in the control group were not statistically significant for either test (P = 0.098 and P = 0.150). The DID favored the intervention for both outcomes: eyes-closed marching-in-place time increased by 1.94 s (95%CI: 0.18 to 3.70, P = 0.031), and balance-beam traversal time decreased by 0.62 s (95% CI: −1.22 to − 0.02, P = 0.044) (Table 3).

Table 3.

Comparison of Balance Ability Between the Intervention Group and the Control Group Before and After Intervention

Outcomes	Intervention Group			Control Group			DID (95%CI)	P value for DID
	Pre-Test	Post-Test	P value	Pre-Test	Post-Test	P value
Dynamic Balance Tests
Closed-eye marching in place, median (IQR)	7.57(5.80 to 10.54)	10.50(7.47 to 15.15)	< 0.001***	7.86(5.68 to 10.87)	8.61(6.51 to 11.84)	0.098	1.94(0.18 to 3.70)	0.031*
Balance beam test, median (IQR)	7.59(6.42 to 9.30)	6.89(5.89 to 8.32)	< 0.001***	7.52(6.43 to 9.28)	7.46(6.23 to 8.92)	0.150	-0.62(-1.22 to -0.02)	0.044*
Static Balance Tests
Standing on one foot with eyes open, mean (SD)	4.85 (0.10)	4.87 (0.10)	0.862	4.87 (0.10)	4.95 (0.10)	0.526	-0.06(-0.46 to 0.33)	0.759
Standing on one foot with eyes closed, mean (SD)	3.31 (0.10)	3.68 (0.12)	0.009**	3.33 (0.11)	3.27 (0.11)	0.598	0.43(0.003 to 0.87)	0.049*
Tandem Romberg, mean (SD)	5.09 (0.08)	5.18 (0.09)	0.408	5.05 (0.09)	5.14 (0.08)	0.321	0.01(-0.33 to 0.35)	0.961

Abbreviations: DID Difference in Difference, CI confidence interval

^*: P < 0.05, ^**: P < 0.01, ^***: P < 0.001

Static balance tests

For eyes-closed single-leg stance, the intervention group improved from 3.31 at baseline to 3.68 at follow-up (P = 0.009), whereas the control group showed no significant change (P = 0.598). The corresponding DID indicated an increase of 0.43 points (95% CI: 0.003 to 0.87, P = 0.049). No significant between-group effects were observed by DID for eyes-open single-leg stance or the tandem Romberg test (P = 0.759 and P = 0.961) (Table 2).

Adverse events

Other than falls, no adverse events related to the intervention occurred.

Discussion

This controlled before-after study evaluated a combined HBM-based education plus balance training program in rural primary schools, delivered during the COVID-19 period. The intervention was associated with a lower rate of fall-related injuries at follow-up than control (IRR = 0.54). All dynamic balance measures showed significant improvement, whereas improvements in static balance were mainly observed in the eyes-closed single-leg stance. Overall, these findings suggest that a theory-guided school program is feasible in rural settings and may help reduce fall risk indicators in primary schoolchildren.

This study implemented a combined falls-prevention education program based on the HBM and balance exercises for rural primary school students. The intervention led to a significant reduction in both the rate and proportion of fall-related injuries over the past year. Using HBM as a theoretical foundation, the program enhanced participants’ awareness of fall risks, established proper health beliefs, and encouraged fall-prevention behaviors through balance training. Improving cognitive awareness is crucial for fostering health-conscious beliefs, which in turn motivate the adoption of effective fall-prevention behaviors [30, 31]. These findings suggest that the combination of HBM-based education and balance training not only improved knowledge and attitudes toward fall prevention but also significantly reduced fall incidents by enhancing balance.

Regarding balance ability, the intervention group outperformed the control group in most metrics, particularly in dynamic balance. The central nervous system integrates inputs from the visual system, vestibular apparatus, and proprioceptive signals from muscles, tendons, and skin to activate and coordinate the muscles responsible for postural stability, thus regulating balance. Balance training, which repeatedly challenges both static and dynamic postures, enhances balance function, consistent with previous studies (e.g., Hu Jingyue [32], Li Yan [33]). Proprioceptive signals are transmitted to the cerebral cortex, where reinforced neural encoding enhances deep muscle excitability and improves intermuscular coordination, as well as balance sensitivity and coordination across muscle groups [34].

Currently, scientific evidence supports the effectiveness of balance-enhancing exercise programs in improving balance function and preventing falls across different countries and regions. However, most studies focus on older adults, with relatively limited evidence for children. Traditional Chinese fitness practices, such as Tai Chi and Baduanjin, are widely recognized for improving balance and preventing falls in older adults [35, 36], and a few studies have applied these practices to enhance balance in children. In recent years, football has been introduced into school settings, with studies suggesting that school-based football can increase lower-limb strength and improve balance in primary school children [37]. Nevertheless, challenges such as movement complexity, a shortage of qualified instructors, and limited facilities hinder the widespread implementation and long-term adherence to these programs. In contrast, balance training programs tailored to the physiology and fitness levels of the target population, featuring simple and easily learned movements, have greater potential for broad fall prevention applications. This study, designed to align with the growth and development characteristics of primary school students, implemented a balance-training intervention. Results showed that the balance routine was well accepted and learned by participants, leading to significant improvements in dynamic balance, with concurrent improvements in static balance. However, no significant differences were observed in eyes-open single-leg stance or the Romberg tests. Possible explanations include the relatively short duration of the intervention, which may not be sufficient for certain metrics to change, and evidence from systematic reviews suggesting that balance training has moderate effects on static balance and substantial effects on dynamic balance [38]. Additionally, balance training alone may not fully enhance balance and strength in children. Research has shown that resistance training can significantly improve motor skills in children and adolescents [15]. Therefore, integrating balance training with resistance training is recommended to enhance both balance and strength development in school-age children, thereby improving fall prevention effectiveness.

This study has several strengths: it developed balance training methods tailored to the developmental stage of the elementary school years and conducted interventions within a HBM framework to evaluate their effectiveness in preventing falls among rural primary schoolchildren. However, it also has limitations. First, because of COVID-19 related disruptions, the study was ultimately implemented in only two rural schools in Shantou, leading to a small number of clusters and a modest sample size. This limits precision, cluster-level inference, and generalizability. Future work should use multicenter, multi-school designs with larger samples. Second, fall-related injuries were reported by students or caregivers rather than verified with medical records, so recall and reporting bias are possible. Third, balance outcomes were measured only twice, at baseline and at 12 months. When conditions permit, additional repeated assessments are recommended in future research.

In conclusion, a classroom-embedded two-phase add-on program, consisting of continuous HBM-guided education and the later addition of supervised in-class balance training, was associated with a lower rate of fall-related injuries and improved balance among rural primary schoolchildren. This low-contact, zero-equipment model offers pragmatic guidance for school-based injury prevention and may inform preparedness for future public health emergencies.

Abbreviations

DALYs

Disability-adjusted life years

HBM

Health Belief Model

COVID-19

Coronavirus disease 2019

BMI

Body mass index

IQR

Interquartile range

ISD

Standard deviation

RR

Incidence rate ratio

RR

Relative risk

DID

Difference-in-differences

CI

Confidence interval

Authors' contributions

XC and ZY conceived and designed the study, conducted the investigation, curated and analyzed the data, and drafted and revised the manuscript. KZ and LX contributed to the investigation and data curation. LL conceived and supervised the study, obtained funding, and critically revised the manuscript. All authors read and approved the final manuscript.

Funding

This study was supported by the Guangdong Province Key Areas Special Project for Ordinary Colleges and Universities (No.2024ZDZX4022).

Data availability

Data available on request from the authors.

Declarations

Ethics approval and consent to participate

This study was approved by the Ethics Committee of Shantou University Medical College (SUMC-2021-36). The participants provided written informed consent. All methods and study procedures were carried out in accordance with relevant guidelines and regulations, including the ethical principles of the Declaration of Helsinki. Written informed consent was obtained from the parents or legal guardians of all participating students, and written consent was also obtained from the students themselves prior to participation. Participation was voluntary, and respondents could withdraw at any time without consequence.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.