Participants and data collection
This cross-sectional study was conducted at six universities with nursing courses in Japan between June 2018 and October 2019. Except for a university in the Tohoku region, a northern area of Japan (n=57), all the universities were located in the Kanto region (i.e., Tokyo, Kanagawa, Chiba). Self-reported questionnaires were distributed to undergraduate nursing students during the term of normal classes, not in the practical training period. To provide feedback on the personal results to the students, their student ID was obtained. The questionnaires were distributed during class time by research collaborators at each university, and collected by means of a collection box located outside of the classroom or in an envelope after class time so that answerers could not be identified in accordance with the administrative regulation of the ethical approval committee in each university. The number of target samples and distributions were 1,462 and 1,304, respectively. We received 519 questionnaires from eligible students with a response rate of 48.4%, of which 34 of the respondents were male (7%) and 485 were female (93%). Because the number of male students is small and their chronotype differs from female students, we analyzed only the data obtained from female nursing students. Thirty-eight respondents were excluded from subsequent analyses because their chronotype could not be determined due to missing data. Finally, 447 valid data were used for analyses in the present study.
The questionnaire was written in Japanese and consisted of the following items.
1. Characteristics of nursing students and lifestyle/social activities
Students were required to answer questions on age, sex, grade, starting and closing time of class, whether they were living alone or not, commute time, exercise habits per week (≥ 4 days /2-3 days/ ≤ 1 day/ none), duration of habitual exercise, start time of exercise, engagement in part-time job (yes /no), part-time job after 10 p.m, starting and closing time of part-time job, participation in a club activity (yes/no), and starting and closing time of club activity. Frequency of exercise habit was converted into a binary outcome (0: none, 1: ≥ 4 days /2-3 days/ ≤ 1day) for analysis.
Data on sleep pattern included information on sleep duration, bedtime, awakening time on weekdays and on weekends. The Japanese version of Epworth Sleepiness Scale(35, 36) was used to assess the level of daytime sleepiness. The total sum score ranges from 0 to 24, and ≥11 is regarded as having excessive daytime sleepiness. Sleep quality was assessed with the Japanese version of Athens Insomnia Scale(37, 38), which indicates possible insomnia when one has a score above five out of the total score. In order to assess sleep-related symptoms, the self-report questionnaire included questions on difficulty initiating sleep, difficulty maintaining sleep, early morning awakening, and/or difficulty awakening.
Information regarding eating patterns included meal time, duration of meal, skipping meals, and frequency of skipping meals at breakfast, lunch, and dinner. In addition, students were required to answer questions on body weight gain after entering the university. Body mass index was calculated by self-reported height and weight.
To assess chronotype, the Japanese version of Morningness-Eveningness Questionnaire(39, 40) was used. This questionnaire consists of 19 items, and a total score below 42 is defined as evening type, 42-58 as intermediate type, and above 58 as morning type.
3 Health-related quality of life and academic performance
For evaluating health-related quality of life, the physical component summary score and mental component summary score of health-related quality of life by SF-8 Health Survey(41, 42) was used. These scores were standardized with a mean of 50.
The questionnaire included questions on frequency of absence, tardiness, and falling asleep during class (often/sometimes/none) and interference with academic achievement (yes/no). Frequency was converted into a binary outcome (0: none, 1: often/sometimes).
Representative values were shown as mean ± standard deviation. Kruskal-Wallis tests and Chi-square tests followed by residual analyses were performed for continuous variables and categorical variables, respectively. To identify the related factors with chronotype, the generalized linear mixed effect model was used. In the model, explanatory variables; age, living alone, start and closing time of class, commuting time, excise habit, part-time job and club activity were set as fixed effect, and university and grade were set as random effect. In the analysis of generalized linear mixed effect model, the degree of association was represented as adjusted odds ratio and 95% confidence interval. In these analyses, complete-case analysis was performed, but no imputation for missing data was performed because missing data were only 12% of the total. The significance level was defined as p < 0.05. Generalized linear mixed effect model was analyzed using R statistical software version 3.5.1 (R Core Team, Vienna, Austria). “lme4” package(43), and other analysis were performed by IBM SPSS Statistics 25 (IBM Corporation, Armonk, USA).
Ethical approval for this survey was granted by the Ethical Committee of the Institute of Neuropsychiatry (No.162). Six universities cooperated with permission of the ethical approval committees; two universities were approved by the ethical committee of the particular university (Teikyo University and Shonan University of Medical Sciences), and the ethical committees of the other four universities delegated ethical approval to the Institute of Neuropsychiatry. All thesubjects participated on a voluntary basis, and responses were treated anonymously. Written informed consent was obtained from all participants.