Subjects comprised over 600 students from a college of education in the center of Israel. The questionnaire was distributed to first-, second-, and third-year students in the science and education faculties. The students who replied were included in the study (N=367), while the students who did not reply were excluded (N=233). All 367 (100%) students replied to the pre-COVID-19 questionnaire, and 166 (45.2%) students replied to the questionnaire during the COVID-19 lockdown (postquestionnaire). Altogether, 533 questionnaires were submitted. Ordinarily, during the 4 years of the students' studies, they attend the college on campus at least 3 days a week, and they are obligated to teach at least once a week in a proscribed school during all 4 of their college study years. In their fourth year, they attend the college only once a week, and during the rest of the week, they teach in schools, under the supervision and guidance of experienced and expert teachers from those schools. This study received ethical approval from the ethics committee of the Seminar Hakibutzim College, Tel Aviv, Israel.

The questionnaire was submitted by the students using Google Forms. The questionnaire included 4 sections: sociodemographic questions, questions pertaining to the level and intensity of their weekly exercise, a well-being questionnaire, and a questionnaire regarding their quality of sleep. The Hebrew version of the exercise questionnaire was taken from the International Physical Activity Questionnaire–Short Form (IPAQ-SF) [40]. The students were asked to note how many times and the number of hours they exercise per week and the number of times and hours a week they engage in physical activity such that it causes sweat production and strenuous breathing. The questions were based on a Likert scale of 1-7, where 1 represents never and 7 represents always. The 4 exercise questions were grouped into 1 variable, and then a dichotomous variable was rebuilt and the population was divided into 2 groups by the median of 3.5. Students with a score of 0-3.49 in exercise were categorized as engaging in a lesser amount of exercise. Students with a score of 3.5-7 in exercise were categorized as engaging in a greater amount of exercise. The well-being questionnaire used was the short version of the Mental Health Inventory (MHI) developed by Veit and Ware [41] and validated by Florian and Drori [42] and included 10 questions: Cronbach α=.96. Sleep quality was measured using the Mini Sleep Questionnaire (MSQ), which included 10 questions that relate to quality of sleep using a Likert scale of 1-7 [43].

During the second and third weeks of the first semester (November 2019), the researchers entered the classrooms and explained the objectives of the study. Immediately following that, the questionnaire was distributed, and the students submitted it using Google Forms. The second questionnaire was distributed via email because at that time (at the end of the first COVID-19 lockdown, April 2020), the students studied only online and did not study on campus. Since the students answered the questionnaire of their own volition, and the questionnaire was anonymous, not all of them chose to participate; thus, only 166 (45.2%) of 367 students submitted the second questionnaire. In the second questionnaire, the students were asked to answer the questions with reference to the COVID-19 lockdown period.

An independent t test analysis was performed to measure the difference between the amount of exercise, sleep quality, and well-being among undergraduate students of education before and during the COVID-19 lockdown, and the difference between well-being and the quality of sleep by the amount of exercise engaged in by the students. The analysis of the answers to the questionnaires before and during the COVID-19 lockdown were not matched; thus, an independent t test analysis was performed. A chi-square analysis was performed to measure the difference in the amount of exercise before and during the COVID-19 lockdown. A 2-way ANOVA was conducted in order to find the main effects and interactions between the period when the questionnaire was submitted (before or during the COVID-19 lockdown), the year of study/field of study/gender, and the 3 dependent variables: amount of exercise, sleep quality, and well-being. Results were statistically significant at the .05 significance level.

The students who participated in this study are undergraduate students of education from the faculties of science and education. The average age of the students was 25.3 (SD 4.5) years. Most of the students were female (291/367, 79.3%) and single (304/367, 82.8%). In addition, 163 (44.4%) of the students were in their first year of studies and 204 (55.6%) in their second and third years of studies. Furthermore, 130 (35.4%) of the students studied in the Department of Physical Education, while the rest studied in other education disciplines, such as science, elementary school, early childhood education, and special education. All 367 (100%) students submitted the first questionnaire, while 166 (45.2%) students submitted the second one (Table 1). There was no statistical difference between the students answering the prequestionnaire and those who answered the postquestionnaire concerning their characteristics.

The 3 main research questions asked were whether there was any difference between the 2 periods (1) in the amount of exercise engaged in, (2) in their well-being, and (3) in their quality of sleep. An independent t test analysis was conducted, and the results showed that the students engaged in more exercise during the lockdown compared to the beginning of the same school year (mean 4.27 [SD 1.4] vs mean 3.4 [SD 1.7], P<.001). They also had better quality of sleep (mean 5.34 [SD 0.92] vs mean 5.12 [SD 0.46], P=.02) and better well-being (mean 3.79 [SD 0.62] vs mean 3.67 [SD 0.59], P=.02); see Figure 1.

When analyzing the amount of exercise the students engaged in before and during the COVID-19 lockdown, 150 (40.9%) of the 367 students engaged in a greater amount of exercise before the COVID-19 lockdown. However, during the COVID-19 lockdown, the number of students engaging in a greater amount of exercise increased to 102 (61.4%) of 166 students (Table 2). The chi-square analysis showed a significant difference (χ²₁=19.41, P<.001). The φ correlation showed low (0.2) but significant (P<.001) strength. An independent t test analysis was performed to compare the quality of sleep among the students who engaged in a greater amount of exercise compared to those who engaged in a lesser amount of exercise during the COVID-19 lockdown. Those who engaged in a greater amount of exercise had better sleep quality (mean 5.57 [SD 0.84] vs mean 4.96 [SD 0.93], P<.001). No difference was found in their well-being (Table 3).

A 2-way ANOVA was conducted to compare the main effects of the interaction of the time when the questionnaire was submitted (before or during the COVID-19 lockdown) and the year of study/field of study/gender on the amount of exercise, sleep quality, and well-being (Tables 4-6).

As shown in Table 4, the effect of the 2 independent variables (time when the questionnaire was submitted and the year of study) on the amount of exercise, sleep quality, and well-being was studied. Results showed effects that were statistically significant at the 0.05 significance level only for the time when the questionnaire was submitted. Regarding the amount of exercise, the main effect of the time when the questionnaire was submitted yielded an F_{(1, 527)} ratio of 33.0 (P<.001), indicating a significant difference between the amount of exercise the students engaged in before the COVID-19 lockdown (mean 3.4 [SD 1.7]) and during the COVID-19 lockdown (mean 4.28 [SD 1.46]). No effect was found of the year of study, and no interaction was found between the time when the questionnaire was submitted and the year of study. When examining the quality of sleep, the main effect of the time when the questionnaire was submitted yielded an F_{(1, 527)} ratio of 6.17 (P=.01), indicating a significant difference between the students’ quality of sleep before the COVID-19 lockdown (mean 5.12 [SD 0.96]) and during the COVID-19 lockdown (mean 5.35 [SD 0.91]). No effect was found of the year of study, and no interaction was found between the time when the questionnaire was submitted and the year of study. The same trend was found regarding well-being. The main effect of the time when the questionnaire was submitted yielded an F_{(1, 527)} ratio of 5.79 (P=.02), indicating a significant difference between the students’ well-being before the COVID-19 lockdown (mean 3.67 [SD 0.59]) and during the COVID-19 lockdown (mean 3.8 [SD 0.61]). No effect was found of the year of study, and no interaction was found between the time when the questionnaire was submitted and the year of study.

Table 5 presents the results of a 2-way ANOVA when the 2 independent variables this time were the time when the questionnaire was submitted and the field of study. The 2-way ANOVA showed effects on the amount of exercise that were statistically significant (P<.05). The main effect of the time when the questionnaire was submitted yielded an F_{(1, 529)} ratio of 9.798 (P=.002), indicating a significant difference between the amount of exercise the students engaged in before the COVID-19 lockdown (mean 3.4 [SD 1.7]) and during the COVID-19 lockdown (mean 4.28 [SD 1.46]). The main effect of the field of study yielded an F_{(1, 529)} ratio of 166.29 (P<.001), indicating a significant difference between the amount of exercise the physical education students engaged in (mean 4.77 [SD 1.3]) and the students from all the other fields of study (mean 2.85 [SD 1.46]). The interaction effect between the time when the questionnaire was submitted and the field of study was significant (F_{(1, 529)}=8.34, P=.004), showing an ordinal interaction between the variables, meaning a greater amount of exercise was performed during the COVID-19 lockdown and among the students of physical education (Figure 2). Regarding sleep quality, the main effect of the field of study showed effects that were statistically significant at the 0.05 significance level and yielded an F_{(1, 529)} ratio of =11.363 (P=.001), indicating a significant difference between the sleep quality of students of physical education (mean 5.37 [SD 0.87]) compared to students from all the other disciplines (mean 5.05 [SD 0.99]). Regarding sleep quality, no interaction was found between the time when the questionnaire was submitted and the field of study (F_(1,529)=0.435, P=.51).

A 2-way ANOVA was conducted for the influence of another pair of independent variables: time when the questionnaire was submitted and gender (Table 6). Regarding the amount of exercise, the main effect of the time when the questionnaire was submitted yielded an F_{(1, 529)} ratio of 8.3 (P=.004), indicating a significant difference between the amount of exercise the students engaged in before the COVID-19 lockdown (mean 3.4 [SD 1.7]) and during the COVID-19 lockdown (mean 4.28 [SD 1.46]). The main effect of gender yielded an F_{(1, 529)} ratio of 46.3 (P<.001), indicating a significant difference between the amount of exercise the male students engaged in (mean 4.77 [SD 1.49]) compared to female students (mean 3.4 [SD 1.49]). The interaction effect between the time when the questionnaire was submitted and gender was significant (F_{(1, 529)}=7.38, P=.01), showing an ordinal interaction between the variables, meaning a greater amount of exercise was performed during the COVID-19 lockdown among male students, although it is interesting to observe that female students showed a more significant change in the amount of exercise during the COVID-19 lockdown (Figure 3 and Table 6).

Undergraduate students of education showed improvement in their well-being, sleep quality, and exercise during the first COVID-19 lockdown. These results were contrary to all expectations, since the restriction enforced due to the COVID-19 pandemic was shown to have a serious impact on the population, resulting in an enormous change in lifestyle and in daily routine [44,45]. Moreover, some studies have shown that during the first COVID-19 lockdown or during quarantine, there was a decrease in well-being, sleep quality, and physical activity among most of the general population [5,7,22,46-51]. Our results can be explained by other studies that have shown that the more time people have, the more they invest in their hobbies and in exercising [52]. A study conducted on European Union citizens revealed that 73.1% exercised in their free time [53], and a study in Spain showed that more than 50% enjoyed new positive experiences during the lockdown; specifically, students of health and science reported an increase in both the number of days in which they engaged in physical activity and in the total number of minutes of physical activity they engaged in per week [39].

However, there are studies showing that during the COVID-19 lockdown, people experienced more time alone, increased the use of electronic devices, and stayed indoors more, which all led to a reduction in sleep quality and a decrease in their well-being [4,11,19-22,50,54,55]. This study does not contradict these findings but complements them with additional aspects of life that have not been tested yet, such as the possible benefit of physical activity on well-being and sleep quality in a time when lifestyle changes may cause sudden distress. The students who participated in this study are studying toward earning a teaching certificate in an Israeli education college in Tel Aviv. Typically, they commute between 1 and 3 hours every day to get to college (the college has no dorms) or to the schools where they teach. After school hours, many of them need to go out to work to support themselves. The first COVID-19 lockdown forced the students to stay at home; discontinue working, teaching in the schools, and studying on campus; and switch to distance learning or distance teaching. Although this situation increased many sitting hours, it also led to more free hours because of less travel time and less or no work (88% reported working less than normal times; see Multimedia Appendix 1). The statistical correlation found between exercising more and sleeping better might be explained by the free time the students gained because of less or no work and no commuting, while not having additional personal or family responsibilities. Doing more exercise and sleeping better may have improved their well-being. These results are in line with previous works showing that free-time activities play an important role in subjective well-being [56] and that high levels of free-time exercise are associated with benefits for psychological well-being [57].

Studies have demonstrated that physical activity can improve well-being regardless of gender, socioeconomic status, and health status [8,58-61]. In addition, physical activity (especially high levels) can improve sleep quality [24,25,62,63]. Furthermore, studies have shown that the amount and extent of physical activity are positively associated with the researched population's well-being. During the COVID-19 lockdown period, the student population was at high risk for developing mental health problems, indicating a decrease in their well-being [48,49]. Here, it was shown for the first time that there is not only a correlation between an increase in the intensity and the amount of the students' exercise (102/166 [61.4%] of the students engaged in a greater amount of physical activity) but also a correlation between the students who engaged in a greater amount of exercise and their well-being and sleep quality.

Transitional times in life, such as going to college and being a freshman, include leaving home, gaining more independence, and adhering to less structured schedules, and other stressful conditions can affect sleep patterns and well-being [64-66]. Thus, there was expected to be a difference between first-year and second- and third-year students with regard to maintaining a healthy lifestyle. Here, no interaction was found between the time the questionnaire was submitted and the year of study regarding the amount of exercise, sleep quality, and well-being, emphasizing that all students, regardless of what year of study they are in, showed better sleep quality and well-being and engaged in more exercise during the lockdown compared to the previous period. The study by Romero-Blanco et al [39] showed that first- and second-year students, but not final-year students, who increased the amount of physical activity during COVID-19 accounted for the difference in the exercise time by the fact that third-year students are required to take on a large number of written assignments, hence increasing their sitting time. In contrast, in this study, the undergraduate students of education had a similar curriculum during the first 3 years of studies; thus, no difference was found between the years of study with respect to physical activity.

Finally, an ordinal interaction was found between the time when the questionnaire was submitted and gender regarding the amount of exercise, that is, even though all students exercised more during the COVID-19 lockdown, male students engaged in a greater amount of exercise before and during the lockdown compared to female students, although female students had a more significant change during the COVID-19 lockdown. This is in line with other studies showing that male students, in general, engage in more sports and more physical activity because of the positive sensation they experience from exercise, stimulation, and enjoyment compared to female students [67]. In addition, male students mention enjoyment, challenges, social recognition, affiliation, competition, and strength as motivating factors for exercise, whereas female students mention preventing poor health and maintaining good health, weight management, and a pleasing appearance [68].

This study had several limitations. First, the study was conducted on a small and specific population most of whom are single and without children; therefore, our findings cannot be generalized to all undergraduate student populations, particularly those who are married and with children. Second, there was a difference between the students’ responses to the first and second questionnaires because the pre- COVID-19 questionnaire was given to the students by their teacher during 1 of their classes and they were given time to answer and submit during class. During the COVID-19 lockdown, all student-teacher communication was only online. Hence, the second questionnaire was sent via email, and not all students who answered the pre-COVID-19 lockdown questionnaire answered the second one. Although the number of respondents decreased, our findings were statistically significant. Third, it should be noted that there might be other factors that can affect well-being, sleep quality, and the amount of exercise that were not examined, such as health status, economic status, personality type, social status, and even living area (city vs rural locality). Finally, this effect was explored during the first lockdown only and lacks information regarding whether there could be a long-term, during-COVID-19 effect or whether the effect would similar during the second COVID-19 lockdown. This should be studied in continued research.

In conclusion, although this study was conducted on a specific population of undergraduate students of education, and thus it is not possible to draw conclusions about the entire population, we showed a statistical correlation between the tested variables, indicating a beneficial effect on sleep quality, well-being, and amount of exercise in this population during the COVID-19 lockdown. These results can encourage people to understand that there is another positive aspect to crisis situations, such as lockdowns, and one can use such situations to improve the quality of life. The results can also inspire policymakers to propose programs focusing on promoting physical activity, and emphasizing sleep quality and well-being, to maintain a healthy lifestyle during lockdowns.