Associations of Smartphone Addiction, Chronotype, Sleep Quality, and Risk Of Eating Disorders Among University Students: A Cross-Sectional Study

Purpose With smartphone addiction’s (SA) increasing prevalence among young adults, there is a growing concern about its adverse effects on distorted dietary and lifestyle behaviors, including disturbed sleeping and eating patterns. We aim to study the prevalence of SA risk, poor sleeping quality, evening chronotype, and eating disorders (ED) risk among university students in the UAE. We also aim to assess the associations between them, emphasizing the one between ED and SA risks. In this cross-sectional study, a self-administered online questionnaire was disseminated using the convenience sampling technique. Pittsburgh Sleep Quality Index, Morningness-Eveningness Questionnaire, Eating Attitude Test-26 items, and Smartphone Addiction Scale-Short Version were used to measure sleep quality, chronotype, ED risk, and SA risk, respectively. Descriptive and analytical statistics were applied, and P< 0.05 was considered for statistical signicance.


Introduction
Nowadays, smartphones are an indispensable part of a person's life [1]; as they have become an integral part of facilitating daily life. In the United Arab Emirates (UAE), 92% of the population uses smartphones [2], yet excessive usage and dependency can turn problematic with compulsive overuse and be labeled as smartphone addiction (SA) [3][4][5]. With SA's increasing prevalence [6], there is also a growing concern about its adverse effects and possible associations with other prevalent distorted dietary and lifestyle behaviors, such as poor sleeping [7], disturbed eating patterns [8,9], and an eveningness preference [10].
College students are known to have poor sleep patterns, which could be due to their hectic university schedules; however, it was recently suggested to be also associated with increased smartphone usage and its resulting sleepimpairing effects [7]. In addition, the endogenous circadian clock, also known as chronotype, is impacted by the different technology devices' usage (such as mobile phones and computer screens) [10,11]. Three forms of chronotype differ in behavioral, biological, and psychological variables: 1) morningness -those who wake up early in the morning and are exhausted in the early evening hours; 2) eveningness -those who sleep late at night and wake up late the next day; and 3) intermediate -'neither type'. The evening chronotype has been associated with extensive smartphone usage [10,12].
Poor sleep quality and evening chronotype may be also associated with eating disorders (ED) [13][14][15]. ED represents abnormal or disturbed eating habits, which include anorexia nervosa, bulimia, and binge eating [16]. The prevalence of ED among college students in the UAE is 24.6% [17], and 10.5% among medical students according to a global meta-analysis [18]. Several studies have shown that ED-diagnosed individuals score higher on insomnia and sleep disturbances scales [13,14], and are more likely to categorize as evening chronotype, compared to controls [15].
Due to similar associated factors with ED and SA, including poor sleep quality, evening chronotype, impulsivity [19], comfort-seeking related to emotional dysregulation, and disturbed self-body image (secondary to social media in uence) [9], a possible relationship between these two risks can hence be suspected [8], on which the literature remains limited. With a rising prevalence of SA and ED in the Middle East, further evidence is needed to explore the associations between these two distortions. Thus, we aimed to investigate the prevalence of SA, poor sleep quality, evening chronotype, and ED among university students in the UAE; and explore the associations between them, with an emphasis on addressing the relationship between SA and ED risks.

Materials And Methods
Study design, sampling, and data collection A cross-sectional survey was conducted between March 9-25, 2021, on university students from the University of Sharjah (UoS), a national university in the UAE with an ethnically diverse student population. A self-administered online questionnaire was sent by e-mail and/or promoted on social media platforms. Inclusion criteria embraced all Arabic and English speaking university students, who were registered during the academic year of 2020-2021 Spring semester. The Research Ethics Committee at the University of Sharjah, Sharjah, UAE (REC-21-02-24-01-S), approved the study and it was performed in line with the principles of the Declaration of Helsinki.

Questionnaire Design
Pittsburgh Sleep Quality Index (PSQI), Morningness-Eveningness Questionnaire (MEQ), Eating Attitude Test -26 Items (EAT-26), Smartphone Addiction Scale -Short Version (SAS-SV), were used as assessment tools, alongside a sociodemographic questionnaire. These questionnaires were compiled into one questionnaire using Google Forms, and consisted of 6 sections: rst, research description, objectives, and request for the informed consent; second to fth for PSQI, MEQ, EAT-26, and SAS-SV, respectively; and nally, the sixth section included sociodemographic  [20,21]. Its questions are based on a 4-point Likert score and are grouped into seven component scores, scoring from 0 to 3. The nal sum of components' scores can range from 0 to 21, with a score >5 indicating poor sleep quality and a score ≤5 for good sleep quality. The MEQ is also a validated 19-items questionnaire, which measures chronotype or circadian peak time [22]. The score ranges from 16- Council (GCC) countries, 38.2% were non-Arabs, and 12.5% were Arabs from non-GCC countries. One-third of students were from colleges of health sciences and medicine (37.1%) and two-thirds from other colleges (62.9%).
Concerning the academic level, 56% were junior students (freshmen to year 2) and 44% were seniors, and 65.6% of students had a cumulative grade point average (CGPA) >3 (very good-excellent). The majority of participants reported a monthly income of <5,000 Arab Emirates Dirham (AED) (84.6%), identi ed as non-smokers (94.2%,) and were living with their families (94.6%). Half of the participants (50%) had a body mass index (BMI) within the normal range, 23.6% were overweight, 10.7% were underweight, and 14.7% were obese. Regarding the sleeping quality and chronotype preference, 71% reported poor sleep quality, with 33.9% belonging to the evening chronotype. Close to two-fourth of participants (37.9%) had a risk of ED and 56.2% presented a risk of SA.  Associations between ED risk, with chronotype and sleep quality, are shown in Table 3. The ED risk was independent of the chronotype category. However, there was a signi cant association between ED risk and sleep quality (p=0.016), as those with ED risk were more likely to report poor sleep quality (80.4%) compared to those without ED risk (65.3%).   Table 5 illustrates linear regression analysis using EAT-26 score as the dependent variable. The EAT score was positively associated with sleep quality score (B=0.87; 95% CI: 0.49-1.26; p<0.001) and negatively associated with SAS score (B= -0.13; 95%CI: -0.24--0.02; p=0.018), even after adjustment for age and sex. The sleep quality score seems to be the best predictor for the EAT score.

Discussion
The current study identi ed a high prevalence of poor sleep quality, and SA and ED risk among university students in the UAE.
According to our results, students who are female, in health sciences or medical-related colleges, or who smoke tend to have poorer sleep quality. Moreover, singles and smokers scored more as an eveningness chronotype.
Meanwhile, higher ED risk was reported among UAE and GCC nationals, junior students, those with lower CGPA, and obese students. Furthermore, females, singles, and Muslims faced higher SA risk.
SA's prevalence among our university students aligns with the rates reported by similar studies on university students in Lebanon (49%) [27] and KSA (82.5%) [3]. Smartphones' multi-functionality, ease of use, portability, social media applications could be attributed as potential drivers for their compulsive habitual use [28].
Additionally, students may use their smartphones as means of escapism to cope with academic/psychosocial stress [29].
Some of our participants' sociodemographic characteristics were associated with SA risk. With regard to marital status, SA risk was the highest in singles, followed by married, and divorced or widowed students. A possible explanation can be that marriage and having a family impose additional responsibilities compared to the single life, which allows less time to do activities alone, like navigating the internet. A study conducted on Nigerian students presented similar ndings, in which being single increased the prevalence of SA by an odds of 5.809 in comparison with those married [30].
This study's reported high poor sleep quality rate closely corresponds with a previous study done upon college students in KSA [31]. Similarly, the higher prevalence of poor sleep quality in females aligns with the aforementioned study [31] and can be related to women's higher reports of anxiety disorders and stress [32,33]. Poor sleep quality can be also attributed to other factors, including psychological distress and depression in university students [34], and excessive caffeine usage [35][36][37] ─ factors that were not assessed in our study. Although our participants' predominant chronotype was intermediate, a higher percentage of evening compared to morning chronotypes was reported (33.9% vs. 9.4%, respectively). Our participants' evening chronotype prevalence is higher compared to similar studies conducted on college students in Brazil (27%) and the USA 20.8%) [38,39].
The trend of eveningness among college students could be similarly explained by their crowded schedules and needs to nd distraction-free environments to complete their academic tasks. Additionally, the prevalent stimulant beverages (such as coffee and energy drinks) consumption can also prolong their circadian rhythm and shift it towards eveningness [36,37].
This study's nding of poorer sleep quality relating to higher SA aligns with the literature [7,40], and can be explained by the fact that excessive smartphone usage before sleep is possibly impairing the normal sleeping pattern and sleep quality [41]. Similar results were reported in a KSA study conducted on university students [31].
Likewise, we found an association between chronotype and SA risk, which is evident by a higher representation of the evening chronotype in those with SA risk compared to those with no SA risk. This can be explained by melatonin's level disruption by smartphone's blue light [11,42], inducing later bedtimes or shifting chronotype towards eveningness [10]. Conversely, eveningness been has associated with high impulsivity and hypothetically may increase addiction proneness [10,43].
Concerning ED risk, the majority of participants reported normal eating behaviors, and the rate of ED was similar to those documented in other countries, such as 29.4% in female university students in Saudi Arabia [44] and 23% among medical students in Pakistan [45]. Perceived stress by adjustments to a new environment, hectic college schedules, and getting accustomed to independent living for those living in dorms or without families can account for the risk of developing ED among university students [46]. Dysregulated circadian rhythm of appetite-regulating and food intake hormones can play an important role in ED behavior [47]. Conversely, nocturnal purging/ bingeeating might impair sleep quality by affecting sleep onset time and/or blood electrolytes before sleep [13], shifting the circadian rhythm preference towards eveningness [15]. Nonetheless, the risk of ED was found independent of the chronotype category similar to Kandeger et al.'s study [48]; however, this nding did not match that of Natale et al. study [15]. Unlike Natale et al, we have used EAT-26 test, which is only a screening tool and not a diagnostic tool for ED, thereby possibly altering our result's dissociation accuracy and causing this discrepancy. ED risk was however correlated with sleep quality, which was also reported in studies on adolescent and college students [49,50]. An additional explanation is that poor sleep quality and sleep deprivation induce a higher rate of impulsivity, which plays a role in ED (such as night purging and/or binging) [19,51].
We speculated that increased SA was associated with higher ED risk, mediated by factors, like impaired sleep quality and evening chronotype. Contrary to our hypothesis, we did not nd any link between SA and ED risks. In contrast, Fatma et al. (2020) found an association between the two and suggested that it was mediated by social media-driven body dissatisfaction [8]. Compared to the aforementioned study, we used different versions of EAT and SAS tests and different ethnicities were studied. Given our usage of non-diagnostic but rather screening tools of ED risk (EAT-26) and SA (SAS-SV), further well-designed studies with proper diagnoses are needed to investigate the association between ED and SA.
The synchrony found among these four main outcomes in this study (evening chronotype, poor sleep quality, and SA and ED risks) implies that components of healthier lifestyle behaviors are clustering among one group of university students, while unhealthy behaviors are clustering among another group. This is part of the fact that many of the lifestyle behaviors are interrelated, and maybe co-exist in one group of people with shared characteristics, and may result in combined, similar health effects on that particular group of people [52].
An awareness program on the circadian typology and its potential impact on daytime functioning will help students in better planning their study schedules for optimal daytime performance [53]. Students could also bene t from tips on minimizing SA and improving sleep quality by practicing sound sleep hygiene, such as maintaining regular sleep and wake time [54], ensuring a quiet and comfortable environment [54], and reducing smartphone usage at bedtime [55].

Strengths and Limitations
This study tested the effects of various variables and assessed potential correlations between themselves and between a wide array of demographic characteristics, providing a solid foundation for future research opportunities. Despite our study's large sample size, our ndings cannot be generalized to all UAE students as participants were recruited only from one university in the UAE. Given the observational nature of our study, it was di cult to con rm some of the well-known associations between the different risks (of such SA and ED risks) and confer causality. Therefore, well-designed longitudinal studies are warranted to help attest to such associations. In addition, response bias may have affected the accuracy of results, as it was self-reported and memory-based questionnaire. It is also important to note that this study was conducted during the COVID-19 pandemic, which on its own represents a stressor and may have aggravated the measured risks and sleep patterns of students. Lastly, we focused on university students in the UAE; further cross-country comparative studies, which take into account other cultures and different age groups will provide more conclusive ndings.

Conclusion
In conclusion, poor sleep quality, ED risk, and SA risk were highly prevalent among our university students. We found that ED risk was independent of the chronotype category, yet it was dependent on the sleep quality. SA risk was associated with both poor sleep quality and evening chronotype. Therefore, awareness campaigns may be needed to be geared to this population to improve their sleep quality and reduce their usage time of smartphones.
What is already known on this subject?
Literature shows that poor sleep quality, evening chronotype, and risk of developing ED and/or SA are prevalent among adolescents and young adults. Poor sleep quality and evening chronotype are shown to be bi-directionally associated with ED and SA through prior studies. Lastly, literature shows that sleep quality can predict the risk of ED.
What does this study add?
The study identi ed a high prevalence of poor sleep quality, SA risk, and ED risk among university students in the UAE. The study also further studied the possible relationship between ED and SA risks, which is scarcely investigated; yet no association was found in contrast to the single prior study done on this subject. This points out the need for further research on this relationship. The associations between poor sleep quality and both risks of SA and ED, and the association between evening chronotype and SA risk were promoted which emphasizes the need for preventive public health measures.

Declarations
There is no con ict of interest or nancial disclosure to declare.

Funding:
No funding was received for conducting this study.

Con icts of interest:
Authors declare no con icts of interest.
Authors' contributions: Data of this study are available on request from the corresponding author.