Associations of quality of sleep with lifestyle factors and profile of studies among Lithuanian students

lifestyle factors among three profiles of studies regarding the anxiety about studies (P<0.0005) and subjective estimation of success in studies (χ²=27.9, P<0.0005), showing the highest anxiety and worst satisfaction among students of medicine. Conclusions. The incidence of sleep problems is high among students in Lithuania, reaching 59.4%. Medical students have worse quality of sleep and worse impact of poor sleep on the quality of life compared to students of law and economics. A significant difference was found between medical students and their peers in other profiles of studies regarding their attitudes and habits related to studies: medical students spent more time for studying, were more anxious about studies and less satisfied with the results, studied more often before going to sleep.


Introduction
A number of studies have showed a high prevalence of bad quality of sleep among university students varying from 19.17% to 57.5% and being especially high among medical students (1)(2)(3). Poor sleep has a signifi cant negative infl uence on physical and mental health (4), performance at the studies, and quality of life of students (5)(6)(7) what requires attention of doctors and pedagogues.
Quality of sleep among Lithuanian students has not been extensively studied. In a study by Proškuvienė and colleagues, fi rst-year students of Vilnius Pedagogical University were interviewed using the Pittsburgh Sleep Quality Index (PSQI), and the results showed that 58.1% of 606 students examined had sleep problems (8).
The PSQI (9) is one of the most common instruments used to evaluate subjective sleep quality and adapted for the use in Lithuania (10). The PSQI can be easily applied in different age groups including students and showed a satisfactory correlation with other clinical measures of sleep disturbances (11,12). Gellis and colleagues performed an Internet-based investigation with 220 Americans (mean age, 41.6 years) and 92 of them were poor sleepers (PSQI >5) (13). While analyzing the associations between sleep disturbance and mental health status among Japanese junior high school students, Kaneita et al. used the PSQI and 12-item General Health Questionnaire for evaluation of mental health status. The incidence of sleep disturbance during the 2 years leading to the follow-up study was 33.3%, and new onset of sleep disturbances was signifi cantly associated with new onset of poor mental health status and lasting poor mental health status (7). In the sample of 400 Hong Kong university students, 57.5% were poor sleepers according to the PSQI results, and their sleep quality was associated with sex, year of study, sleep hygiene practice, and perceived adequate sleep in the past month (2). In the study of Chinese medical students (14), the prevalence of poor-quality sleep was 19.17% and did not differ signifi cantly between genders but correlated with the year of study, worry of sleep, irregular work/ rest, worry on examination, stress, relationship with classmates, self-evaluated health condition, environments of the dormitory, and going late to bed.
Other instruments such as the Epworth Sleepiness Scale (ESS) and the Horne & Ostberg Morningness/Eveningness Questionnaire (MEQ) were used besides the PSQI for evaluation of students sleep. In the comparative studies of the PSQI and the ESS (12,15), these two scales correlated weakly with each other but segregated from each other on principal components analysis. The PSQI was more closely related to psychological symptom ratings and sleep diary measures than the ESS though none of these scales was related to objective sleep measures and cannot be used as screening measures for polysomnographic sleep abnormalities. After evaluating the stability over the past year in early middle-aged adults, the PSQI and ESS were found to be stable measures of sleep quality and sleepiness. Hirata et al. analyzed the infl uence of morningness-eveningness on depression in a medical school using the MEQ. It was found that eveningness was associated with depressive symptoms, and this association remained signifi cant after adjusting for the presence of familial depression and physical activity (16).
Studies investigating the associations between sleep disturbances and lifestyle factors are important looking for the causes and ways to improve quality of sleep. Investigating the associations among quality of sleep, quality of life and lifestyle habits, Carney et al. (1) found that good sleepers engaged more regularly in activities with active social engagement. Increased working hours and working on weekends were the reasons of decreased quality of sleep, strain with family and peers, depression and anxiety among young students in hospitality and tourism (17). Medical studies were found to be associated with higher levels of stress symptoms. Niemi et al. explored stress symptoms among undergraduate medical students at fi ve points (fatigue, sleeping problems, anxiety, irritability, and depression) during the six-year medical training program. There was a consistent increase of stress reports throughout the medical program in both genders (18). According to the self-reported Sleep and Daytime Habits Questionnaire, sleep quality of Estonian medical students (n=413, aged 19-33 years) was associated with academic progress, leisure activity, and living conditions but was not associated with students' daily or nightly workload (3).
Studies on clinical psychology have showed that medical students experience a high level of stress because of intensive schedule, huge amount of theoretical material to be learned, pressure of responsibilities and authorities, alcohol consumption, lack of social ties and activities (18)(19)(20). Furthermore, problems with sleep are prominent among practicing physicians (11). It raises a question if medical profession or difficulties during study years have a long-lasting negative impact on quality of sleep of medical students and doctors. Even though several studies have evaluated sleep and sleep-related problems in medical students (3,5,14,21), none has compared sleep quality among different study profi les and has analyzed an impact of study profi le on quality of sleep and life. Therefore, the objective of our study was to analyze relations between the sleep quality, profi le of the studies, and lifestyle factors among students of three different study profi les (medicine, economics, and law). . The universities are situated in the two largest cities of Lithuania -Vil nius and Kaunas. Students from three different study profi les -medicine (2 faculties, 150 students), law (2 faculties, 120 students), business and economics (2 faculties, 135 students) -were studied to assess the correlations between study profi le and quality of sleep and life. Respondents were randomly selected from odd groups of the fi rst-and fourthyear studies (i.e. fi rst, third, fi fth, etc.). The mean age of fi rst-year students and fourth-year students was 19.1 years (SD, 0.83; range, 18 to 20) and 21.8 years (SD, 1.14; range, 21 to 25), respectively. There was no signifi cant variance in distribution regarding gender and year of studies among the universities and profi les of the studies.

Material and methods
Procedure. It was a cross-sectional study, carried out in the middle of the fall semester (relatively calm period of the studies). Questionnaires were handed out before lecture, and it took about 10-15 minutes to complete them. All participants answered the standardized questionnaires made of two parts: 1) the Pittsburgh Sleep Quality Index for subjective evaluation of sleep quality; 2) the questionnaire about sleep and lifestyle habits developed by the researchers.
Measures. Sleep quality. The PSQI was used to evaluate subjective quality of sleep (9). This instrument has a good internal consistency (Cronbach's alpha for this study was 0.83) and is adapted for the use in Lithuania (10). The questionnaire relates to respondent's usual sleep habits during the past month only; therefore, answers should indicate the most accurate reply for the majority of days and nights in the past month. The PSQI consists of 19 questions: 4 ordinal response questions asking to write exact time in hours and/or minutes (e.g. During the past month, how long (in minutes) has it usually taken you to fall asleep each night?), 12 questions with four response categories (e.g. During the past month, how often have you had trouble sleeping because you cannot get to sleep within 30 minutes: not during the past month, less than once a week, once or twice a week, three or more times a week), 2 questions with four response categories (e.g. During the past month, how would you rate your overall sleep quality: very good, fairly good, fairly bad, very bad), and 1 fi ve response question with fi ve parts answered according to the observations of respondent's bed partner or person sharing a room.
Responses to all 19 questions were summed up into 7 components of sleep index: subjective sleep quality (1st component), sleep latency (2nd component), sleep duration (3rd component), habitual sleep effi ciency (4th component), sleep disturbances (5th component), use of sleeping medication (6th component), and daytime dysfunction (7th component). Every component can be rated from 0 to 3 points, where 3 points always indicate serious problems in the evaluated component. The sums of all 7 components were summed up again producing the PSQI, which refl ects the degree of sleep disturbances in general. It can range from 0 to 21. If the PSQI value is higher than 5, it represents a person with severe disturbances in 2 components of sleep or moderate disturbances in 3 or more components. In our study, the PSQI value of >5 was considered as "poor quality of sleep" and persons with such scores as "poor sleepers," while PSQI of ≤5 was considered as "good quality of sleep" and persons with such scores "good sleepers." Lifestyle habits. Lifestyle habits and their impact on sleep were evaluated by the structured questionnaire developed by the researchers. It consisted of 10 Likert scale-type questions as follows: a) respondent's usual activities during the past month (time distribution between studies, work, social activities, leisure time, and other activities); b) self-estimation of success in studies, work, social activities, leisure time, and other activities; c) self-estimation of physical health; d) self-estimation of emotional status; e) self-estimation of quality of sexual life; f) somatic morbidity and use of medications, mental morbidity and use of psychotropic medications; g) usual activities before going to bed; h) main reasons causing tiredness, and i) main reasons causing anxiety.
The Self-Evaluation Index (SEI) was determined summing up the scores of the answers to the questions regarding self-estimation, which ranged from 0 to 21 with the higher value representing worse self-evaluation. The Anxiety Index (AI) was calculated summing up the scores of the answers to the questions regarding causes of anxiety, which ranged from 0 to 12 with the higher value representing the lower incidence of anxiety.
Impact of sleep on the quality of life. Four multiplechoice questions were used to evaluate the impact of disturbances of sleep on four main areas of activities and health, which are usually related to the quality of life of a person (22). Respondents were asked to estimate the quality of work and studies (e.g. ability to do mental and physical work), emotional status (e.g. changes of mood), physical health (e.g. nausea, somnolence, and headache), quality of sexual life (e.g. changes of sexual activity and attraction) after a night of poor sleep.
The Sleep Impact on the Quality of Life Index (SILQI) was derived from all these questions with the values ranging from 0 to 15. The higher value of SILQI represents the increase in physical and emotional symptoms caused by poor sleep or the stronger negative impact on the quality of life.
Statistical analysis. Descriptive statistics was represented by percentage for qualitative variables and mean, standard deviation, maximum and minimum values for quantitative variables. Chi-square (χ²) criterion was used to estimate the difference in the distribution of the qualitative variables. Pearson correlation α coeffi cient was used to analyze correlations between quantitative variables. Student's t test was used to compare means between two quantitative variables, and Fisher's and Bonferroni criteria were used to compare means among three quantitative variables. Statistical signifi cance was set at P<0.05. Statistical package SPSS 15.0 was used for coding and analyzing the data.

Results
General characteristics of sleep. During the past month, the average time of going to bed for the students in this study was 00:35 AM ± 1 h 58 min (range from 8 PM to 3 AM) and average time of awakening was 7:43 AM ± 1 h 32 min (range from 4:30 AM to 1 PM). Average length of time to fall asleep was 20.9±17.2 min (range from 0 to 120 min), and the average length of actual sleep per night was 7.2±1.4 h (range from 3.5 to 12 h). Distribution of these time variables among three study profi les is shown in Table 1. Signifi cant differences were found comparing the students of different profi les regarding the average length of actual sleep per night, average time of awakening, and daytime dysfunction. Medical students woke up earlier, had shorter average length of sleep, and greater daytime dysfunction than their peers from other profi les of studies.
No signifi cant differences or correlations with sleep or quality-of-life characteristics were found regarding gender and the year of studies except greater daytime dysfunction among women as compared to men (t=-2.6, P=0.009, independent sample t test) and shorter average length of actual sleep per night among the fourth-year students as compared to the fi rst-year students (t=-2.2, P=0.003, independent sample t test).
Medications for sleep were used by 5.9% of the respondents with no signifi cant differences among genders, year or profi le of studies.
Pittsburgh Sleep Quality Index. The PSQI (the sum of 7 components of sleep quality) was calculated for 387 (95.6%) participants. The PSQI values varied from 1 to 17 with a mean value of 6.2±2.5. Female respondents scored higher on the PSQI than males (6.3 vs. 5.7, respectively; t=-1.9, P=0.048, independent sample t test). No signifi cant differ-ence was found in the PSQI scores regarding the years of studies or city of studying. Although there was no signifi cant difference among universities, VU MF students had the highest mean PSQI score (7.3±2.9) and VU KFH students the lowest mean PSQI score (5.6±2.3) in absolute numbers. A statistically signifi cant difference in the mean PSQI score was found among three profi les of studies (F=4.8, P=0.009). The highest mean PSQI score showing the greatest disturbances of sleep was found among the students of medicine (m 1 =6.56), lower mean score was among the law students (m 2 =6.26), and the lowest mean score among the economics students (m 3 =5.65). After multiple comparisons with Bonferroni criterion, a signifi cant difference was found between medical and economics students (P<0.01).
More than half (n=230, 59.4%) of the students scored higher than 5 on the PSQI, which allowed suspecting sleep disorders. The frequency of sleep disorders was higher among females than males (61.5% and 53.5%, respectively), but this difference was not statistically signifi cant (χ²=2.02, P=0.1). There were no signifi cant differences regarding the year of studies and university. Signifi cant difference in the frequency of poor sleepers was found regard-

F=4.4 P<0.0005
*Chi-square (χ²) criterion to estimate the difference in the distribution of the qualitative variables; **Fisher's (F) criterion to compare means among three quantitative variables.  Table 2). Students of medicine and law spent more time studying and less time working or having leisure time than their peers studying economics.
The SEI values varied from 0 to 15 with a mean value of 6.7±2.6. There was no signifi cant difference in the SEI regarding gender or profi le of studies. A statistically signifi cant difference among three profi les of studies was found just regarding subjective estimation of success in the studies (χ²=27.9, P<0.0005), showing the worst satisfaction with the results among the students of medicine. Self-estimation of physical health, emotional status, and quality of sexual life did not vary signifi cantly regarding the profi le of studies.
The AI values varied from 0 to 12 with a mean value of 3.9±2.4. There was no signifi cant difference in the AI regarding gender or profi le of studies. Studies (55.1%), relations with a partner (16.6%) and family members (10.6%) were the most frequent reasons for anxiety mentioned by respondents. A statistically signifi cant difference was found among three profi les of studies regarding the anxiety about the studies (χ²=30, P<0.0005), showing that students of medicine were most anxious about their studies.
Impact of poor sleep on the quality of life. The most frequent complaints about a negative infl uence of poor night sleep on academic and work performance included inability to concentrate (68.6%) and problems performing mental work (68.1%). Mood changes (57.5%) and bad mood (35.8%) were the most frequent complaints about consequences of poor night sleep on emotional status. Somnolence (86.7%) and general asthenia (47.2%) were the consequences of poor night sleep on physical health and decreased sexual attraction and activity were consequences on the quality of sexual life (44.9% and 27.2%, respectively). All these symptoms were more prevalent among female respondents than in males but the difference was not statistically signifi cant.
The SILQI values varied from 0 to 13 with a mean value of 6.2±2.5. There was a statistically signifi cant difference in the SILQI among three study profi les (F=4.9, P=0.008) with higher means among medical students. After multiple comparisons with the Bonferroni criterion, a signifi cant difference was found comparing medical and economics students (P=0.01), and medical and law students (P=0.04).
Correlations between quality of sleep, lifestyle habits, and subjective evaluation of quality of life. Correlations between quality of sleep measured with the PSQI, lifestyle habits (time planning, everyday activities before going to bed), and subjective evaluation of quality of life were analyzed. There were no signifi cant correlations between the amount of time given for work, social activities, leisure time and quality of sleep. A signifi cant correlation was found between the PSQI and activities before going to bed showing that the students studying before going to sleep had worse quality of sleep (Pearson correlation, α=0.18; P<0.01). Similarly, a signifi cant correlation was found between the PSQI and amount of time spent for studying (α=0.14, P<0.01), showing that the students spending more time for studying had worse quality of sleep.
There were statistically signifi cant associations between quality of sleep and subjective evaluation of quality of life: poor sleepers were more pessimistic about their achievements in the university

Discussion
Quality of sleep is interrelated with both emotional and physical health -somatic and mental diseases decrease quality of sleep and vise versa poor sleep worsens emotional and physical condition (5)(6)(7)23). Because of that, sleep can be considered an important indicator of health of a person and his/ her ability to cope successfully with everyday stress. Impact of these factors on emotional and physical health of medical students has not been properly examined in Lithuania. In this study, medical students were compared with two other "prestigious" and highly demanding profi les of studies -law and economics -intending to evaluate the impact of lifestyle factors excluding diffi culties of studies.
Results of the study showed that sleep problems were rather prevalent among students: 59.4% of all students scored higher than 5 on the PSQI that allowed suspecting them as having signifi cant problems with sleep. Students of medicine had the highest incidence of the problems with sleep. These results are consistent with the fi ndings in other countries (2,5,14). Greater disturbances of sleep among medical students were found in general evaluation of sleep quality measured by the PSQI and in the main characteristics of sleep: time of awakening, length of sleep, and daytime dysfunction caused by poor sleep. Probably that was the main reason why poor sleep had the highest negative impact on the quality of life for the students of medicine. A negative impact of poor sleep was observed on the quality of studies, emotional and physical health showing the same risks related to poor sleep as noted in other studies (6,24). It shows the risks for health and professional performance associated with sleep problems, which are common among medical students and require interventions to decrease these risks and their negative consequences.
The main factor discriminating medical students from their peers in other universities was their attitude to studies. Students of medicine were spending more time studying (F=4.9, P<0.0005), were more often anxious about their studies (χ²=30.3, P<0.0005), less satisfi ed with their results at studies (χ²=27.9, P<0.0005), were studying more often before going to sleep (χ²=73.2, P<0.0005). This "submerge into studies" was not successfully counterbalanced with the leisure time as among students of economics.
It seems not probable that worse subjective evaluation of sleep quality could be explained by generally more pessimistic attitudes of medical students as their self-estimation in other aspects did not differ signifi cantly from their peers. There were no signifi cant differences in medical students also in respect to anxiety levels because of other problems than studies (interpersonal relationships, work, fi nances), or incidence of physical or mental disorders. Further studies would be important to clarify if these differences are infl uenced by the nature of studies, the amount of material to be studied, students-teachers relationships or other factors.
Our study did not fi nd a signifi cant decrease in subjective evaluation of quality of sleep and health between younger and older medical students, which was reported in some other studies (3,14,18). It could be due to the fact that we did not include students of the last two years of studies. Assessment of the cumulating negative effect during the course of studies and the role of gender would require larger sample with inclusion of the students of all 6 years of studies.
Comparisons among the profi les of studies revealed that the students of economics had the best quality of sleep. Students of law evaluated quality of their sleep better than students of medicine but it was worse than among the students of economics. In most aspects, differences between the students of medicine and law were smaller than between students of law and economics. An important and signifi cant difference was found regarding the time spent for daily activities. Students of law were not different from medical students, and students of economics spent signifi cantly more time for leisure activities and work than their peers. Probably more balanced priorities in life allowed students of economics to remain less anxious and more satisfi ed with the results of their studies.

Conclusions
Incidence of sleep problems is high among Lithuanian students, reaching 59.4%. Medical students had worse quality of sleep and worse impact of poor sleep on quality of life compared with students of law and economics. A signifi cant difference was found between medical students and their peers in other profi les of studies regarding their attitudes and habits related to the studies; medical students spent more time studying, were more anxious about studies and less satisfi ed with the results, studied more often before going to sleep. It allows hypothesizing that attitudes and habits related to studies may have a negative impact on sleep quality and health in medical students. These fi ndings suggest that medical students have to receive more knowledge about sleep hygiene, effective skills of coping with stress and sleep improvement, psychological support to improve satisfaction with their results of studies.