2.1 Objects Between September 16, 2023, and October 24, we employed a cluster sampling and cross-sectional survey method to select two classes each from the first, second, and third grades, as well as four classes from the first year of high school in a certain middle school in Henan. A total of 412 students were chosen as the study subjects. Inclusion criteria for the study were: ① Age between 13 and 17 years old; ② Han ethnicity; ③ Consent to participate in the survey. Exclusion criteria included: ① Those with concurrent physical illnesses; ② Those with a history of or current addiction to psychoactive substances. Ultimately, 31 students with concurrent physical illnesses or incomplete data were excluded, and the final study sample comprised 381 students.
The survey was conducted by 10 trained psychiatrists who distributed questionnaires and answered questions during the students' response process. The survey was conducted in batches based on students' study times, with each survey involving approximately 40 students from a single class. In total, 10 questionnaire surveys were conducted. A total of 412 questionnaires were distributed, with 381 valid questionnaires returned, achieving a response rate of 92.5%.
During each session of the questionnaire assessments, 10 psychiatrists and volunteers provided guidance and answered questions throughout. The Pittsburgh Sleep Quality Index (PSQI) [19], consisting of 19 self-reported questions, with a total score ranging from 0 to 21, was used. In this study, a score of ≥ 5 was defined as poor sleep quality. According to this criterion, 381 middle school students were classified as having good sleep quality (n = 199) or poor sleep quality (n = 182). The study obtained approval from the Ethics Committee of the First Affiliated Hospital of Zhengzhou University (Ethics Review Number: 2023-KY-0220-001). Prior to the survey, participants signed informed consent forms, and approval was obtained from relevant school authorities.
2.2 Tools
2.2.1 Self-Developed General Questionnaire ① Includes demographic characteristics: gender, age, years of education, only child status, personality, parents' personalities, and body mass index (BMI). ② General clinical features: including childhood trauma history and family history, with a total of 2 variables.
2.2.2 International Physical Activity Questionnaire Short Version (IPAQ-SF) [16] Used to measure physical activity in the past week, comprising 7 items, including high-intensity and moderate-intensity exercise, walking, and sitting. Each type of physical activity is assigned an intensity code in metabolic equivalent of task (MET) units. The MET values are 3.3 for walking, 4.0 for moderate-intensity exercise, and 8.0 for high-intensity exercise. Total energy expenditure is estimated by multiplying the MET score by the minutes spent on each activity per week, expressed in MET-minutes/week. Physical activity levels are categorized as low, moderate, and high, with moderate and high levels combined as the physical activity group, and low level as the non-physical activity group.
2.2.3 Hamilton Depression Rating-Scale-24-item version (HAMD) [17] This scale assesses the severity of depressive symptoms, with 24 items. The total score ranges from 0 to 74, with scores ≤ 8 indicating the absence of depressive symptoms, 9–19 indicating possible depressive symptoms, ≥ 20 indicating mild or moderate depression, and ≥ 35 indicating severe depressive symptoms. In this study, a total HAMD score ≥ 20 was considered indicative of depression.
2.2.4 Hamilton Anxiety Scale (HAMA) [18] Used to assess the severity of anxiety symptoms, with 14 items. The total score is 56, with ≤ 7 indicating the absence of anxiety symptoms, 8–13 indicating possible anxiety symptoms, ≥ 14 indicating mild anxiety, ≥ 21 indicating severe anxiety, and ≥ 29 indicating extremely severe anxiety. In the study, a total HAMA score ≥ 14 was considered indicative of anxiety.
2.2.5 Pittsburgh Sleep Quality Index (PSQI) [19] This index is used for self-assessment of sleep disorders, including 19 self-reported questions and 5 questions rated by bed partners or roommates. In this study, only the self-reported questions were utilized. These questions contribute to scores in 7 components covering subjective sleep quality, sleep onset latency, sleep duration, habitual sleep efficiency, sleep disturbances, use of sleep medications, and daytime dysfunction. Each of these factors reflects the severity of the corresponding issue (0 = none, 1 = mild, 2 = moderate, 3 = severe), constituting an ordered categorical variable. The total score is the sum of these 7 factor scores, ranging from 0 to 21. Specifically, scores of 0–5 are classified as good sleep, 6–10 as average sleep, 11–15 as poor sleep, and ≥ 16 indicates very poor sleep quality. In this study, a score of ≥ 5 was defined as poor sleep quality.
2.2.6 Morning and Evening Questionnaire 5 (MEQ-5) [20] This questionnaire is used to assess students' circadian rhythm types, comprising 5 items with a total score of 25. Based on the score, ≤ 11 is classified as evening type, and ≥ 12 as non-evening type.
2.3.Statistical Methods
This study utilized SPSS 25.0 software for data analysis. For categorical data, frequencies (n) and percentages (%) were used for representation. For metric data conforming to a normal distribution, means ± standard deviations were used for representation, while for skewed metric data, the representation used was (median, interquartile range/between the upper and lower quartiles). In comparing the demographic data, general clinical features, and the composition ratios of HAMD, HAMA, and MEQ-5 scales between the groups with good and poor sleep quality, the chi-square (χ²) test was employed. For metric data such as BMI, years of education, and age, as well as comparisons of ordered multiclass variables such as physical activity levels (low, moderate, high), subjective sleep quality, sleep onset latency, sleep duration, habitual sleep efficiency, sleep disturbances, use of sleep medications, and daytime dysfunction, the Mann-Whitney U test was used, and comparisons among three groups were conducted using the Kruskal-Wallis test. Additionally, Spearman rank correlation analysis and binary logistic stepwise regression analysis (1 = good sleep quality, 2 = poor sleep quality) were conducted to explore the factors related to sleep quality. In all statistical analyses, the significance level was set at P<0.05, indicating statistical significance for observed differences.