Lifestyle Elements for Improving Mental and Physical Health in Japanese University Students: Subjective Sleep Quality is a Common Key Factor

Abstract

This study aimed to reveal the key lifestyle elements that improve physical and mental health in university students by focusing on physical activity, nutrition, and sleep. This cross-sectional study was conducted between October 2021 and December 2021. The participants were 290 first-year students (mean age, 18.63 ± .63 years; age range, 18 to 23; 198 female). The outcomes were daily step counts measured using accelerometers, dietary intake by nutrient category, sleep duration, subjective sleep quality, exercise frequency and duration by exercise type, screen time, depression level, and subjective fatigue by body part. Depression and subjective eye fatigue represent mental and physical health outcomes. Subjective sleep quality predicted depression (β = −1.22, P < .001) and eye fatigue (β = −.23, P < .01) in the path analysis. Participants with higher subjective sleep quality performed more frequent aerobic exercise (P < .01), longer session times of physical relaxation exercise (P < .05), and shorter screen time (P < .05). Subjective sleep quality could be a key factor for high mental and physical health. Furthermore, performing aerobic and relaxation exercises and reducing screen time are important for improving the subjective sleep quality.

“Path analysis revealed that BMI and sleep duration influenced depression levels, and subjective sleep quality was an important common factor for physical and mental health.”

Introduction

In aging societies with an average lifespan of >80 years, extending healthy life expectancy is an urgent issue. Lifestyle significantly affects mental and physical health, and physical activity, nutrition, and sleep are the three most important elements of a lifestyle. Improving lifestyles among university students is important to establish subsequent lifestyles. ¹

Modern issues are caused by lifestyle changes owing to scientific and technological developments, the promotion of digitalization, and unforeseen outbreaks of infectious diseases in each element of lifestyle, such as physical activity, nutrition, and sleep. Several studies have reported that physical inactivity has become a major problem owing to lifestyle changes. The introduction of online classes in many universities during the coronavirus disease COVID-19 pandemic may lead to a decrease in physical activity.^2-4 Regarding nutrition, deficiencies in fruits and vegetables have been reported in university students.^5,6 Regarding sleep, college students reported sleep deprivation.^7-9 University students also reported different sleep patterns on weekdays and weekends. ¹⁰ Therefore, it is necessary to understand the basic status of these three major lifestyle factors for future health.

Furthermore, to address the basic issues of the three major lifestyle elements, it is important to improve current physical and mental health issues in college students. To identify interventions necessary for health improvement, it is necessary to comprehensively examine the key factors that affect the outcomes of physical and mental health. Depression is a main concern among university students.^11,12 Regarding physical health outcomes, increasing screen time strengthens eye fatigue and decreases work efficiency, which is a major issue in the digital society.^13-17 Therefore, we focused on depression as a representative mental health outcome and subjective eye fatigue as a representative physical outcome.

Taken together, understanding the actual lifestyle of university students and identifying issues to improve their physical and mental health are important for a healthy lifestyle throughout their lives. This study aimed to investigate the reality of the three lifestyle elements (physical activity, nutrition, and sleep) of university students for future health, and comprehensively explore the key factors of lifestyle elements to increase current physical and mental health. Our working hypothesis is that university students’ lifestyles show lower physical activity, especially in online classes, unbalanced nutrients such as a lack of vegetables and fruits, excessive snack consumption, and lower sleep quality with shorter sleep duration on weekdays. For physical and mental health, it is hypothesized that those with high physical and mental health will have higher physical activity, more frequent and longer performance in each of the different types of exercise, balanced nutrition, higher quality of sleep, and shorter screen times.

Methods

Participants

All participants were recruited from the student population in the Department of Pharmaceutical Sciences, Nursing, Social Management, Clinical Psychology, Physical Therapy, Occupational Therapy, and Speech-language-Hearing Therapy of the Health Sciences University of Hokkaido through an announcement in a late-semester exercise science class for 1st-grade students. All participants were native Japanese speakers. All the experimental protocols were approved by the Institutional Ethics Committee of the Health Sciences University of Hokkaido (21N019019). The experiment was conducted according to the protocols and guidelines of the latest version of the Declaration of Helsinki. The research content was explained in class, and ethics, consent, and permission for participation were obtained using an Excel file format. To ensure students' free choice of refusal to participate in the research, we explained that the consent decision was based on complete free will, that there was no disadvantage in not consenting, and that it would not affect their class performance. The total number of potential participants registered in the class was 434. Of these, 337 responded to the survey during the specified period and on the specified form. Of these, 314 agreed to participate, and 23 did not. Among those who provided consent (N = 314), 24 participants were excluded from all analyses because they had a blank column (N = 18) and filled values above the upper limit (N = 6). As a result, we analyzed the data of 290 participants (mean age, 18.63 ± .63 years; 198 female). The demographic characteristics of the participants are presented in Table 1.

Table 1.

Demographics of Participants.

Measure	All
Sample size	290 (198 female, 68.2% of all students)
Age (year)	18.63 (.63) (age range 18 to 23)
Height (cm)	163.16 (7.97)
Weight (kg)	55.41 (10.02)
BMI (kg/m2)	20.71 (2.62)

Values are shown as the mean (SD). Note: BMI, body mass index.

Research Procedure

This study was conducted between October 2021 and December 2021. Students were not under lockdown during the study period, but there were impacts from COVID-19, with recommendations to refrain from leaving the house and the introduction of online classes. The measurement method was explained in class and consisted of 7 days of measurement and a questionnaire. Accelerometers (MC-500, Yamasa) were lent to the students and were worn for seven days.

Measurements

The consent of the subjects to participate in the study and their age, height, and weight were recorded. Body mass index (BMI) was calculated using the following formula: weight (kg)/height (m).² Daily step count was measured using an activity meter (MC-500, Yamasa). The device was worn with a strap around the neck and placed in the breast pocket, inside clothing, or in the front pocket of the pants. The subjects were asked to wear the device for as long as possible during waking hours, except during bathing and sleeping. The amount of intake for each meal (breakfast, lunch, and dinner) for each nutrient category (grain dishes/vegetable dishes/fish and meat dishes/milk and milk products/fruits/snacks) were scored as 0, “not at all”; .5, “little”; 1, “small bowl”; 2, “bowl.” Participants were asked about their average and typical lifestyle patterns during the last six months, and screen time (total time spent using TV, games, smartphones, and computer devices) per day was calculated on weekdays and weekends. Items included bedtime, wake time, sleep duration, and subjective sleep quality (“Feeling refreshed after awakening”) on an 11-point scale (0, “Not at all”; 5, “Neutral” to 10, “Extremely”), exercise frequency (time/week), exercise duration (min/week) for aerobic exercise (AE), resistance exercise (RE), and physical relaxation (PR), including muscle stretching and self-massage. The total time (min/week) for AE, RE, and PR was calculated as the frequency of exercise and multiple exercise duration. Depression levels were obtained as mental health outcomes by using the BDI-2 questionnaire. The subjective fatigue of the eyes, shoulders, back, and legs was rated on an 11-point scale. Subjective eye fatigue was used as a representative item for physical health outcomes.

Outlines of Statistical Analysis

We performed two analyses to test our hypotheses. First, we compared the average of all participants between conditions (see below “Average of all participants and the examination of each category”). Second, path analysis was conducted (see below). Third, sub-group analyses were performed based on subjective sleep quality (see below “Sub-group analysis”). SPSS version 22 (SPSS, Inc., USA) and R software were used for all the statistical analyses. Statistical significance was set at P < .05.

Average of All Participants and Examination of Each Category

First, the daily averages were calculated for daily step count, AE time, RE time, PR time, dietary intake by nutrient category, sleep duration, subjective sleep quality, and screen time. Daily step counts were compared between face-to-face classes, online classes, and weekends to examine each category. In the class category evaluation of the daily step count, only students who had online classes were included (N = 207 participants, 129 females; mean age, 18.67 ± .67 years). Dietary intake was compared by nutrient category (grain dishes/vegetable dishes/fish and meat dishes/milk and milk products/fruits/snacks) using a one-way ANOVA. Sleep duration and subjective sleep quality were assessed using paired t-tests between the weekdays and weekends.

Path Analysis

Path analysis was performed using the “lavaan” package in the R software to address a possible causal relationship. Model fit was evaluated using χ² goodness of fit, root mean square error of approximation (RMSEA), and comparative fit index (CFI). The criteria used were CFI ≥ .90, RMSEA < .05, and χ² ≥ .05. The variables were BMI, sex, age, physical activity (daily step count), AE total time, RE total time, PR total time, intake of grain, vegetable, fish and meat, milk and milk products, fruits, snacks, sleep duration, subjective sleep quality, and screen time.

Three models were tested to examine the causal relationship between the variables depending on the hypothesis. The first model included all lifestyle elements as independent variables, with depression and eye fatigue as dependent variables. The second model used independent variables that had significant and interpretable effects on depression and eye fatigue in the first model and had paths to the dependent variables. The third model was a modified version of the second model that excluded sex as a variable.

Sub-Group Analysis

The participants were divided into two groups according to their subjective sleep quality level (a common influential factor). The median score was used as a grouping criterion. The high group comprised participants who exhibited scores above the median, whereas the low group comprised those who exhibited scores below the median. Participants with scores equal to the median were excluded from the analysis. The following outcomes were compared between the high- and low-sleep quality groups: average daily step count, average daily dietary intake by nutrient category, average daily sleep duration, frequency and duration of exercise per week per exercise type, subjective fatigue per body part, and average daily screen time. The average daily step count, average daily sleep duration, and average daily screen time were analyzed using t-tests for subjective sleep quality levels (high and low). One-way ANOVA was conducted for average daily intake by nutrient category, exercise frequency, and duration per week by type of exercise.

Results

Averages of Physical Activity, Food Intake, and Sleep Habit

The average daily values for the daily step count, dietary intake per nutrient per meal, sleep duration, and subjective sleep quality are shown in Table 2.

Table 2.

Daily averages of Physical Activity, Food Intake, and Sleep Habit.

	Items	Average ± S.E	Male Average ± S.E	Female Average ± S.E
Physical activity	Daily step count (step)	5946.79 ± 2804.78	5694.14 ± 2733.99	6064.18 ± 2836.22
	AE total time (min)	6.65 ± 14.46	10.40 ± 17.11	4.91 ± 12.72
	RE total time (min)	4.07 ± 12.51	6.78 ± 15.57	2.81 ± 10.61
	PR total time (min)	3.20 ± 6.57	2.64 ± 6.57	3.46 ± 6.57
Food intake	Grain dishes (score)	3.59 ± 1.20	3.97 ± 1.15	3.41 ± 1.19
	Vegetable dishes (score)	1.96 ± .99	1.78 ± .81	2.04 ± 1.05
	Fish and meat dishes (score)	2.25 ± .91	2.42 ± 1.02	2.17 ± .85
	Milk and milk products (score)	.83 ± .68	.80 ± .55	.84 ± .73
	Fruits (score)	.60 ± .64	.45 ± .52	.67 ± .68
	Snack (score)	.78 ± .66	.60 ± .54	.86 ± .70
Sleep habit	Sleep time (hour)	7.10 ± 1.04	7.26 ± 1.15	7.02 ± .98
	Subjective sleep quality (score)	6.11 ± 1.84	6.17 ± 1.78	6.07 ± 1.87
Other	Screen time (hour)	3.41 ± 2.13	3.91 ± 2.58	3.17 ± 1.84

Comparison by Each Category

The daily step count was the highest in face-to-face classes, followed by weekends, and lowest in online classes (F (2,205) = 79.15, P < .001, η² = .27, Figure 1A). A comparison of the intake of each nutrient category showed that grain dishes had the highest intake, followed by fish and meat dishes, vegetable dishes, and milk and milk products/snack; fruit intake was the lowest (F (5,285) = 696.64, P < .001, η² = .70, Figure 1B). Sleep duration was longer on weekends than on weekdays (t (289) = 18.54, P < .001; Figure 1C), and subjective sleep quality was higher on weekends than on weekdays (t (289) = 14.41, P < .001; Figure 1D).

Figure 1.

Comparison of each category. (A) Daily step count by class category. (B) The daily amount of intake by category of nutrients. (C) Daily sleep time on weekdays and weekends. (D) Daily subjective sleep quality on weekdays and weekends. The values are represented as box plots where the bottom, middle, and top lines of the boxes are the 25th, 50th (median), and 75th percentiles, respectively, and the whiskers above and below each box indicate the most extreme point within 1.5 times the interquartile range. The points above or below the whiskers represent outliers. ***P < .001, **P < .01, *P < .05.

Path Analysis

In the first model, all the variables were lifestyle elements. For the first model, χ² = not applicable, CFI = 0.00, and RMSEA = .00. This yielded a poor fit of the data. The relative effects are shown in Figure 2A. Paths that were not significant were excluded from the subsequent models. Furthermore, the path of the positive correlation between fruit intake and eye fatigue was excluded from the next model because it was difficult to interpret. For the second model, χ² = .26, CFI = .97, RMSEA = .03. This yielded a satisfactory fit to the data. The relative effects are shown in Figure 2B. The path from sex to depression, which was not significant, was excluded from the next model. For the third model, χ² = .91, CFI = 1.00, RMSEA = .00. This yielded a good fit to the data. The relative effects were as follows (Figure 2C): BMI for depression (β = .41, P < .05), sleep time for depression (β = .77, P = .05), sleep quality for depression (β = −1.22, P < .001), and sleep quality for eye fatigue (β = −.23, P < .01). The final model showed that sleep quality was the key factor that had the greatest influence on depression and eye fatigue.

Figure 2.

Path analysis exploring key lifestyle elements for mental and physical health. (A) First model. (B) Second model. (C) Third model. ***P < .001, **P < .01, *P < .05.

Comparison by Subjective Sleep Quality Level

The high sleep quality group comprised 137 participants and the low sleep quality group comprised 118 participants. The high sleep quality group exhibited a higher frequency of AE (F (1, 253) = 11.87, P < .01, Figure 3A), longer duration of PR (F (1, 253) = 5.21, P < .05, Figure 3B), and shorter screen time (t (253) = 2.18, P < .05, Figure 3C) than the low sleep quality group. There were no significant differences in daily step count (t (253) = .71, P = .47), intake of grain dishes (F (1, 253) = .22, P = .63), vegetable dishes (F (1, 253) = 3.49, P = .06), fish and meat dishes (F (1, 253) = .20, P = .65), milk and milk products (F (1, 253) = .83, P = .36), fruits (F (1, 253) = .17, P = .67), snacks (F (1, 253) = 3.38, P = .06), and sleep duration (t (253) = 1.63, P = .10).

Figure 3.

Comparison of sub-groups according to subjective sleep quality. (A) Exercise frequency of each type of exercise according to subjective sleep quality level (SQ). (B) Exercise duration of each type of exercise by SQ. (C) Daily screen time by SQ. The values are represented as box plots where the bottom, middle, and top lines of the boxes are the 25th, 50th (median), and 75th percentiles, respectively, and the whiskers above and below each box indicate the most extreme point within 1.5 times the interquartile range. The points above or below the whiskers represent outliers. **P < .01, *P < .05.

Discussion

This study aimed to determine the actual status of the three lifestyle elements (physical activity, nutrition, and sleep) of university students, and to explore the lifestyle elements comprehensively to meet physical and mental health needs. To our knowledge, university students are physically inactive, lack vegetable intake, do not eat fruit, and do not sleep adequately on weekdays. Participants had lower physical activity levels during online classes compared to face-to-face classes and weekends. Subjective sleep quality was a common key factor for mental and physical health, and subjects with a higher subjective sleep quality performed more frequently with AE, a longer PR session, and a shorter screen time.

First, the overall trend was an average of 5946 total daily steps per person, which is well below the recommended 7000–8000 steps. ¹⁸ Increasing the daily step count decreases mortality. ¹⁹ It is important to be active between work and after school on weekdays, running errands, and exercising at home on weekends. The amount of physical activity during online classes was lower than that during face-to-face classes, which is consistent with previous studies, and may be due to the elimination of commuting to school. Interestingly, the amount of physical activity on online classes was lower than on weekends, with no commute to school. Due to the restriction that online classes must be conducted in an environment with Internet access, physical activity is believed to be even lower than on weekends, when students are free to go out and create opportunities. It is necessary to make conscious efforts to be physically active during online classes, both between and in class.

According to the Japanese Food Guide Spinning Top published by the Japanese government, the ideal order of amount of intake by nutrient category is grain dishes foods > vegetable dishes > fish and meat dishes > milk and milk products and fruits. Therefore, the current results suggest a lack of intake of vegetables and fruits. Vegetable and fruit intake decreases the mortality rate ²⁰ and is especially important in preventing cardiovascular disease. ²¹ There are several barriers to the intake of fresh diets such as vegetables and fruits. ²² Adding a vegetable dish and replacing snacks with fruits could be useful in increasing their intake.

A comparison of weekday and weekend lifestyle patterns revealed lack of sleep on weekdays. This result is consistent with previous studies, and differences in weekday and weekend sleep patterns negatively impact body composition, mental health, and academic performance. ²³ The early start of school on weekdays, heavy homework, and increased extracurricular and social activities may have contributed to the lack of sleep on weekdays. ²⁴ The delayed circadian rhythm and increased eveningness during adolescence may also relate to insufficient sleep on weekday. ²⁵ The clock gene types, which are closely related to morning and night patterns, affect waking time and physical activity during the day, but only on weekends. ²⁶ Eveningness may result in accumulation of sleep debt due to socially required early waking times on weekdays and a tendency to recover on weekends. An increase in weekend sleep duration may reflect a lack of sleep on weekdays. The accumulation of sleep debt is a concern since sleep deprivation accumulated over five weekdays cannot be compensated for by a longer sleep on a two-day weekend. Accumulation of sleep debt is known to cause cognitive decline, ²⁷ and there are concerns about a decline in physical and mental performance. Therefore, it is necessary to compensate for the lack of sleep on weekdays.

Path analysis revealed that BMI and sleep duration influenced depression levels, and subjective sleep quality was an important common factor for physical and mental health. A high BMI could increase the risk of depression, which has been reported^28,29 and is consistent with our current results. Subjective sleep restfulness has begun to attract attention as a predictor of mortality risk. ³⁰ Sleep quality can influence mental and physical outcomes such as depression, ³¹ academic performance, ³² and fatigue.^33,34 Although sleep duration is also an important indicator, sleep quality is more useful as a subjective measure to predict health because subjective sleep duration often does not coincide with actual sleep duration, excluding bedtime. Although traditional questionnaires for subjective sleep quality, such as the PSQI, consist of multiple questions, ³⁵ the wakefulness measure used in the current study predicts depression and the level of eye fatigue in only one simple question, “How much do you feel refreshed upon awakening,” making it a useful measure that is easy to use in daily life.

Further sub-group comparisons by subjective sleep quality score suggested that exercise participation and reduced screen time are important for improving subjective sleep quality. Physical activity improves sleep quality. ³⁶ In the current study, participants with high subjective sleep quality exhibited a higher frequency of AE and a longer duration of PR, but the daily step count and RE were not significantly different. A previous study also showed that AE^37-39 and yoga interventions, including components of PR, ⁴⁰ improved sleep quality, which is consistent with the current results. Not only total physical activity, but also the type, frequency, and duration of exercise could be important factors for health outcomes. An association between longer screen time and lower sleep quality has also been reported in previous studies. ⁴¹ A previous study reported that low physical activity and long screen times could increase depression and poor sleep quality ⁴² supporting our results. Taken together, to improve subjective sleep quality, which is a common key factor for a higher state of physical and mental health outcomes, performing AE frequently and PR longer and having shorter screen time could be important.

This study has several limitations. First, it is necessary to use objective measurements of the nutritional and sleep habits. Second, lifestyle patterns, including sleep habits, were measured over six months, not daily; therefore, daily measurements are required in future studies. Third, depression among university students, one of the primary outcomes, is known to be influenced not only by lifestyle but also by a variety of other factors, including psychological (e.g., loneliness), academic (e.g., workload pressure, subjects), biological (e.g., sex), social and financial factors. ⁴³ The level of depression can fluctuate depending on the students’ progression and academic status. Longitudinal studies from before entering to graduation that consider a wide range of factors are needed in future research. Fourth, because cross-sectional studies do not reveal causal relationships, experimental studies that manipulate the variables extracted in this study are required to clarify causal relationships.

Conclusion

These results revealed that the amount of physical activity was lower than the recommended line, vegetable and fruit intake was insufficient, and sleep debt could accumulate on weekdays in university students. In particular, online classes were found to reduce physical activity compared with weekends. The subjective quality of sleep is a key factor in a high state of physical and mental health outcomes, and may be ideal for improving lifestyle, focusing on AE and PR, and reducing screen time. This study provides fundamental knowledge for university students to achieve a high state of physical and mental well-being by acquiring an ideal lifestyle in a digital society.

ORCID iD

Takemune Fukuie https://orcid.org/0000-0002-1375-2522