After obtaining approval from the university institutional review boards, the research was conducted at 3 universities in Jeddah: King Saud bin Abdulaziz University for Health Sciences, King Abdulaziz University, and Jeddah University. The study was conducted from June to December 2019. Study participants were the undergraduate health care students at the universities. Students in their internship year were excluded.

This is a cross-sectional study where undergraduate students of health care sciences were invited to participate in completing a self-administrated questionnaire about the effect of electronic devices on their sleep quality and academic performance.

Assuming that the prevalence of poor sleep is between 40% to 60% and the margin of error between 6% to 10%, the estimated sample size is approximately 600. Sample size calculation was done using PASS 2020 (NCSS LLC) software [19,20].

The data were collected through an online self-administered questionnaire (Multimedia Appendix 1) comprising 5 major sections. The first section addressed sociodemographic data, including gender, age, and educational discipline. The second section pertained to the type and pattern of electronic device use before bedtime. The grade point average (GPA) used to reflect student academic achievement performance was in the third section. The outcome measures used have good validity and reliability. To measure addiction rate, the Smartphone Addiction Scale for Adolescents–short version (SAS-SV) was used [4]. The scale comprises 5 components: daily life disturbance, withdrawal, cyberspace-oriented relationship, overuse, and tolerance. The cutoff value for males was 31 and females was 33. The values were chosen based on the Haug et al [4] study, with an area under the curve of 0.963 (0.888 to 1.000), sensitivity of 0.875, and specificity of 0.886 for males. For females, the area under the curve was 0.947 (0.887 to 1.000), sensitivity was 0.875, and specificity was 0.886. The Pittsburgh Sleep Quality Index (PSQI) was used to evaluate sleep quality [21]. The PSQI uses 7 areas of measures to differentiate between poor- and good-quality sleepers: subjective sleep quality, sleep latency, sleep duration, habitual sleep efficiency, sleep disturbances, use of sleep medication, and daytime dysfunction over the past month. The PSQI score ranges between 0 to 21, where the higher score represents poorer sleep quality. A total PSQI score of 5 or more indicates poor sleep quality. The invited participants were informed about the study purpose and their voluntary enrollment. Participant information confidentiality was assured.

P<.05 was considered significant, with a confidence interval of 95%. The collected data were analyzed and managed in SPSS (version 24.0, IBM Corp).

An electronic survey was distributed to 1000 students, and 61.80% (608/1000) of participants responded to the questionnaire. There are no missing data regarding the SAS-SV, 8.22% (50/608) of participating students did not enter their GPA, and 16.11% (98/608) of the students had missing data regarding the PSQI. The effects of missing data on the results were checked using multiple imputation, and there was no discrepancy between the results of the analysis of the original data and imputed data. In addition, pooled estimates of the original and imputed data are reported.

The median age of the cohort was 21 years with a range from 18 to 40 years. The majority were females (437/608, 71.87%), and the median GPA was 4.5 out of 5. A majority of the students were from the medicine (227/608, 37.3%) and applied medicine (226/608, 37.2%) colleges. One-third (195/608, 32.07%) of the students who completed the SAS-SV were found to be addicted to smartphones, and 62.75% (320/510) of the students who completed the PSQI were considered poor sleepers (Table 1).

Habits of electronic device use are demonstrated in Table 2. Almost all (601/608, 98.84%) of the participants use smartphones and half of them use tablets or laptops. Around 19.24% (117/608) of the participants watch TV, and 15.78% (96/608) of the group play video games. Only one participant never used electronic devices before sleep and 67 students rarely did. Three-quarters (456/608, 75.0%) of the cohort always used electronic devices at bedtime, while 18.09% (110/608) reported that they usually did. When asked about putting their devices on silent mode, 44.57% (271/608) of them always did. More than half of the cohort never or rarely wake up due to calls or email at night, and only 5.92% (36/608) said they always wake up. Sleep quality was compared across educational disciplines using a chi-square test. There was no significant difference between disciplines regarding sleep quality (P=.50). Poor sleep was similarly prevalent across disciplines ranging between 66.66% (42/63) and 48.27% (14/29; Table 3).

There was no correlation found between PSQI and GPA (r=.018, P=.70) or addiction score and GPA (r=.02, P=.60). Even when GPA, PSQI, and SAS-SV were categorized and included in a logistic regression model along with age and gender to identify GPA predictors, poor sleep and addiction to smartphones were not associated with a lower GPA. However, male gender and age over 21 years were significantly associated (Table 4).

There is a significant correlation between PSQI and SAS-SV (r=.2, P<.001). This significant relationship persisted even after GPA, PSQI, and SAS-SV were categorized and included in a logistic regression model along with age and gender to identify predictors of poor sleep. However, GPA, age, and gender were not significantly associated with poor sleep (Table 5).

Our findings suggested that more than half of the study cohort experiences poor sleep quality. The percentage is similar in other studies [22-25]. A recent study by Aldhawyan and colleagues [26] found that 75.4% of the first-year medical students at Imam Abdulrahman bin Faisal University had poor quality of sleep.

Sleep importance cannot be overemphasized due to its critical role in immune, hormonal, and cardiovascular systems in addition to regulating appetite and metabolism [27,28]. The poor sleep quality observed in this study could cause significant health hazards. Sleep disruption mechanisms are believed to have adverse short- and long-term general well-being consequences [29-31]. The activations of the sympathoadrenal system, sympathetic nervous system, and hypothalamic-pituitary-adrenal axis are evident by virtue of multiple observations. Increased oxygen consumption and carbon dioxide production throughout brief and extended arousals during sleep indicate increased metabolism [29].

Moreover, oversecretion of the adrenocorticotropic hormone and altered levels of epinephrine, norepinephrine, and catecholamine are noticed in chronic insomnia [30,31]. Poor sleepers tend to have a higher risk of obesity and type 2 diabetes as suppression of slow-wave sleep leads to decreased insulin sensitivity and leptin levels and increased ghrelin [32,33]. Determining reversible etiologies leading to poor quality of sleep may aid in improving sleep hygiene.

Almost all of the enrolled population uses electronic devices before bedtime to some extent. A significantly positive correlation between the SAS-SV and PSQI was observed. These results support the findings of Van den Bulck et al [34,35] where individuals using electronic devices before bedtime go to bed later and tend to be more tired during the day. Choi et al [36] conducted a study on 2336 high school students on internet overuse and its correlation with excessive daytime sleepiness as a reflection of poor sleep quality. The study found that internet addiction is strongly associated with excessive daytime sleepiness in adolescents, and the prevalence rate of excessive daytime sleepiness for internet addicts was 37.7%, in addition to a higher prevalence of insomnia, witnessed snoring, apnea, teeth grinding, and nightmares. These observations were reported in another study correlating electronic device overuse or addiction with negative consequences on students’ sleep quality and overall health [37].

There are a few suggested mechanisms through which screens affect sleep. Using electronic devices displaces time that could have been spent sleeping. Electronic devices alter bedtime behaviors as users seek more extended screen entertainment, postponing bedtime [38,39]. Furthermore, psychological stimulation from both violent and nonviolent video games increases arousal [40]. Another likely mechanism of negative impact on sleep from electronic device is exposure to the light emitted by screens at bedtime. Melatonin ideally increases in the hours before bedtime, but studies have demonstrated that screen light suppresses its levels in the blood, quelling sleep drive [41]. In addition to melatonin suppression, screen lights prolong the time to fall asleep and reduce the length of rapid eye movement sleep [42].

Minimizing electronic device use, particularly smartphones, around bedtime is expected to ensure better sleep quality. A small randomized pilot study found that sleep quality and working memory improved in participants who restricted mobile phone use for 30 minutes before bedtime [43]. Limiting the use of smartphones before bedtime could improve the quality of sleep. Restricting phone use for just a week increases lights off time by 17 minutes and sleeping time by 21 minutes [44].

Multiple studies have correlated poor academic performance with bad sleeping habits [45-47]. Interestingly, our study has shown that the PSQI score was not associated with a low GPA. However, our findings suggest that poor academic performance is mainly associated with male gender and age of over 21 years. The gender-GPA association has been described before where females had significantly higher grades than males by 6.3% [48]. Similar outcomes were described in another study where females excelled at every academic assessment tool [49]. On the other hand, GPA and age association seems to have different patterns. Sheard [49] concluded that mature‐age students scored higher GPAs compared to young undergraduates, while another study did not show an association between GPA and age [50]. Moreover, older females outperform both older males and younger females in medical school [51].

Missing data related to the PSQI and GPA probably decreased the power of the study to detect a significant association between them. The study was not designed to assess the relationship between sleep quality and outcomes like physical health, psychological well-being, and quality of life. Also, the study lacks an objective way to measure the exact duration of electronic device use. Finally, social desirability bias may be considered with self-reporting of GPA.

Using electronic devices is common among health care science students and was significantly associated with poor sleep quality. This concern should alert public and academic organizations as more attention is required to motivate students to minimize electronic device use at bedtime to ensure sound sleep hygiene and quality.