The prevalence and determining factors of sleep disorders vary by gender in the Egyptian public officials: a large cross-sectional study

Background: Several studies have investigated sleep disorders in the general population. However, a few studies were conducted among the working population and none among the Egyptian workers. We aimed to assess the prevalence of sleep disorders and determine their associated factors. In a sample of 3143 Egyptian public officials, we collected, via a self-administered questionnaire, cross-sectional data on demographics, social, occupational, behavio-ral, and medical factors associated with sleep. The Jenkins sleep evaluation (JSEQ) questionnaire assessed four sleep disorders, and the logistic and linear regression models determined the associated factors. Results: The prevalence of sleep disorders was 26.2% and 14.5% in female and male Egyptian public officials. Common factors associated with sleep disorders in men and women were age, BMI, job demands and control, and history of chronic disease and depression. However, other associated factors varied by gender. Urban residence (aOR = 1.54 (1.10–2.17), overtime work (aOR = 1.60 (1.16–2.23), and low perceived family social support (aOR = 1.63 (1.06–2.52) in males, while higher income (aOR = 3.64 (1.71–7.73) in females were positively associated with sleep disorders. Higher education levels were associated positively in females and inversely in males with sleep disorders. These factors pre-dicted 28% and 39% of the total variance in the JSEQ score in females and males, respectively. Moreover, the associated factors varied by the type of sleep disorder. Conclusions: The determinants of sleep disorders varied by gender. Worksite interventions to improve job control and income and reduce chronic physical and mental disease risk are recommended to enhance sleep quality.

Considering the health hazards, poor productivity, and impaired safety of the workers (Pilcher and Morris 2020) with sleep disorders, determining the prevalence and correlates of these disorders among the working population is of utmost importance. However, a few studies were conducted among working people. The evidence came mainly from American (Yong et al. 2017) and Asian (Thach et al. 2020) workers whose characteristics and work environment vary significantly from those of other working populations.
In Egypt, the workforce exceeded 27 million adults in 2020 (The World Bank 2021). Unfortunately, there has been no study to determine the prevalence and correlates of poor sleep disorders in this large and vital sector of Egyptian society. The previous Egyptian studies investigated sleep disorders in small samples of students (Ismail 2017;Salama 2017) or chronically ill patients (Ibrahim and Wegdan 2011;Abd Elsadek et al. 2019;Mansour et al. 2020). Therefore, we aimed to estimate the prevalence and correlates of sleep disorders in Egyptian public officials using a large Egyptian sample of a joint bilateral project between Egypt and Japan ).

Study design and population
The design was a cross-sectional study. The staff members of the Public Health Department, Faculty of Medicine, Minia University, Egypt, collected data from Egyptian public officials from October 2019 to January 2020 to find differences between Egyptian and Japanese public officials regarding the work-family conflictrelated health outcomes. Approval to the protocol of this joint project between Egypt and Japan was given by the Faculty of Medicine Research Ethics Committee (FMREC), Minia University (approval # 194:4/2019). Details of the Egyptian survey were published previously (Eshak et al. 2021;Abdelrehim et al. 2022;Eshak and Abdelrahman 2022).
The Egyptian team surveyed public officials in their workplaces by a self-administered questionnaire, including the Jenkins Sleep Evaluation Questionnaire (JSEQ) (Jenkins et al. 1988). The questionnaire was distributed to the public officials after explaining the aim of the study and obtaining written consent. The response rate was around 65%, as 3249 questionnaires were returned out of 5000 distributed questionnaires. The eligible number of questionnaires for further analysis was 3143, as 106 were excluded due to missing information on JSEQ.

The dependent variable (sleep disorders)
The JSEQ is a scale of items that assesses the types and intensity of common insomnia symptoms in the general community, i.e., delay of sleep start, difficulty continuing the sleep, waking up too early, or feeling tired after the usual amount of sleep (Jenkins et al. 1988). Hence, four common disorders were assessed regarding sleep initiation, maintaining, restoring, and being torn out after a usual amount of sleep. Frequency responses to these four sleep disturbances within the past 30 days were assessed on a 6-point Likert scale: 1 = not at all, 2 = 1-3 days, 3 = 4-7 days, 4 = 8-14 days, 5 = 15-21 days, and 6 = ≥ 22 days. The American Psychiatric Association indicated sleep disorders if the individual experienced any sleep disturbance for ≥ 3 nights per week within the preceding four weeks (American Psychiatric Association 2000). Therefore, we considered public officials who reported ≥ 15 days' experience of any of the four sleep disturbances to have sleep disorders (Eshak 2019).

Associated factors
The participants were asked to report their age, sex, area of residence, education, occupation, income, marital status, working status of the spouse, family structure, the level of social support provided by the family, number of family members and minors (< 14 years), working hours per day, shift work, extra jobs or overtime work, and levels of job demands, control, and social support at work. Participants were classified as never, former and current smokers of cigarettes and/or Shisha. The physical activity of the recruited public officials was estimated in units of the metabolic equivalent of energy expenditure (METs). The weight and height of the public officials were measured in standard procedures, and the body mass index (BMI) was calculated as weight in kilograms divided by the height in square meters. We also collected information on the history of physician-diagnosed chronic physical diseases (diabetes, hypertension, hypercholesterolemia, cardiovascular disease, and cancer) and mental disease (depression).

Statistical analysis
We used SAS software version 9.4 (SAS Institute Inc, Cary, NC, USA) for the analyses, and two-tailed p-values of less than 0.05 indicate statistical significance. We presented, gender-specifically, the median (interquartile range) and proportions of the public officials' characteristics. The Mann-Whitney and the Chi-square tests tested the statistical significance of the differences in these factors between public officials with and without sleep disorders. In addition, we described the gender-specific frequency scores for the four types of sleep disorders. The logistic regression models calculated the univariate and multivariate odds ratios (ORs) and 95% confidence intervals (CIs) of having sleep disorders. The stepwise multiple linear regression estimated the slope and the standardized beta of the total score of the JSEQ as a dependent variable according to standardized unit changes of the studied factors. Further, we conducted subgroup analyses to test the factors associated with each type of the four sleep disorders. Since multiple comparisons can lead to a higher potential for type I error, the subgroup analyses should be interpreted as exploratory analyses.

Results
The prevalence of sleep disorders among 3143 Egyptian public officials was 20.5% overall, 26.2% in women, and 14.5% in men. Compared to those without sleep disorders, men and women with sleep disorders were younger, more likely to reside in urban areas, to work an extra job, and to have higher BMI, higher job demands, lower job control, lower social support, and histories of chronic physical and mental diseases. Higher proportions of university education, high income, professional occupation, and night shift work, but a lower proportion of employed spouses were observed among female public officials with sleep disorders. On the other hand, higher proportions of single and living alone males or with few family members were observed among male public officials with sleep disorders (Table 1). Waking up tired after the usual amount of sleep for 15 days or more during the previous month was the most common sleep disorder reported by both genders; 17.8% in women and 7.6% in men. Waking up too early and having trouble falling asleep again was the second most common sleep disorder in women, 8.8%; while waking up several times per night but did not have difficulty falling asleep again ranked second among men, 5.7% (Table 2). Table 3 shows gender differences in the univariate and multivariate associations of sleep disorders with several correlates. The odds of having sleep disorders increased in women and decreased in men educated to a university level or more in comparison to those maximally educated to a high school level; adjusted OR (95% CI) was 1.69 (1.03-2.76) in women and 0.59 (0.36-0.97) in men. Women's odds of having sleep disorders increased in the highest income category in reference to the lowest; adjusted OR = 3.64 (1.71-7.73). In men, the odds increased with urban residence [adjusted OR = 1.54 (1.10-2.17)], working an extra job [adjusted OR = 1.60 (1.16-2.23)], and low perceived family social support [adjusted OR = 1.63 (1.06-2.52)]. However, universal risk factors of sleep disorders in both genders were high BMI, high job demands, low job control, and histories of physical and mental diseases.
Moreover, the odds of having sleep disorders decreased with age in both genders. The factors mentioned above contributed to 28% and 39% of the total variance of the JSEQ score in women and men, respectively, as shown in the linear regression analyses of Tables 4 and 5.
We reported the gender-specific variabilities in the associations of the significant factors with each type of sleep disorder in Additional file 1: Tables 1, 2, and 3. Factors that determined women's difficulties in falling asleep were job demands and histories of physical and mental diseases, while men's were age, residence, shift work, and job demands. Women's interrupted sleep was determined by age, income, BMI, the number of minors in the family, job demands, and histories of physical and mental diseases. In contrast, men's interrupted sleep was associated with BMI, job control, and histories of physical and mental illnesses. Waking up too early in women was associated with BMI, the number of minors in the family, job demands and control, and histories of physical and mental diseases. This disorder in men was associated with BMI, working an extra job, and histories of physical and mental illnesses. The most common sleep disorder, waking up tired after the usual amount of sleep, was correlated with almost all the factors except the BMI and the number of minors in the family in women, and the BMI, job demands, and job control in men.

Discussion
Our cross-sectional study of Egyptian public officials showed a higher prevalence of sleep disorders among women than men. Moreover, the factors correlated with the total sleep disorders, and each specific disorder varied by gender. The family's income and the number of minors were associated with the overall and specific sleep disorders in women, while residence, working an extra job, shift work, and perceived social support from the family correlated with the men's total and specific sleep disorders. Age, job demands and control, and histories of chronic diseases and depression correlated with sleep disorders in both genders. Higher education was associated positively in women and inversely in men with the odds of having sleep disorders.
The estimated prevalence in our study is in line with those estimated in American and Asian working populations. An American study reported that the prevalence of sleep disorders was 19.2% among 6338 workers aged ≥ 18 who participated in the National Health and Nutrition Examination Survey (Yong et al. 2017). Among 464 Asian workers (64% Chinese) aged 21 years or more, the prevalence of poor sleep quality was 43.2% (Thach et al. 2020). The prevalence of poor sleep quality was 15.5% among 470 Iranian workers (Yazdi et al. 2014). However, a lower prevalence of 3.3% was indicated among 7112     (Heo et al. 2013).
There was a higher prevalence of sleep disorders in females than male Egyptian public officials. Similarly, among 1171 daytime Korean workers, the prevalence of sleep disorders was 38.1% in women and 25.8% in men (Kim et al. 2018). In a large French longitudinal study of 21,378 workers, the prevalence of sleep disorders among women increased from 25. 7% in 1990 to 29.4% in 1995, and that of men increased from 19.1% to 21.0% (Ribet and Derriennic 1999). The postulated mechanisms for the higher prevalence in women and the different correlates among both genders include the earlier and shorter females' circadian clock than males (Duffy et al. 2011) and the higher domestic burden on working women (Eshak 2019). * p value < 0.05 indicating a significant association a Adjusted ORs (95%CIs) were estimated by a logistic regression model that simultaneously contained all the significant variables in the univariate analyses (variables that did not reach the level of statistical significance in the univariate analyses were excluded from the multivariable-adjusted model)   There is growing evidence that the risk of sleep disorders increases with age (Evans et al. 2021). However, age in our study was inversely associated with the odds of having sleep disorders in both genders, matching the findings among the Korean working population (Kim et al. 2018). The younger employees might consider sleep limiting their activities; thus, maintaining a regular sleep schedule was not a priority (Kim et al. 2018). Ribet and Derriennic indicated in their large longitudinal French study that the growing prevalence of sleep disorders with age was more attributed to a reduced rate of sleep disorders disappearance than to additional incidences with age (Ribet and Derriennic 1999).
In our study, sleep disorders increased among highly educated women and decreased among highly educated men, which adds to the discrepant association between education level and sleep disorders. Sleep disorders were attributed to lower educational attainment in some studies (Thach et al. 2020) and higher educational attainment in other studies (Yong et al. 2017). Heo et al. (2013) indicated both university or above-, and junior high schooleducated Koreans had significantly more sleep disorders than high school graduates. With the possible gender discrimination at the worksite (Jayachandran 2014), highly educated female public officials may have significantly less job satisfaction than males. On the other hand, highly educated females who succeed in attaining a relatively higher employment status may have greater obligations to work, leading to long working hours and inadequate and poor sleep (Heo et al. 2013). This postulation was supported by the finding that women with a high income had significantly higher odds of having sleep disorders.
The association between BMI and sleep disorders is in tone with the previous literature. The elevated levels of cortisol, responsible for the desire to consume energydense foods in patients with obesity, were found to play a role in increasing the risk of nocturnal awakening and insomnia (Chan et al. 2018).
The associations between work environment (high job demands, low job control, working an extra job, and shift work) and sleep disorders found among the Egyptian public officials agree with those in other nationalities (Ribet and Derriennic 1999;Kim et al. 2018;Thach et al. 2020). Shift work can disrupt the circadian rhythm. In addition, stressful work environments could induce psychosocial stress, hormonal disturbance of melatonin, cortisol, and other hormones and may initiate/aggravate mental health problems, leading to poor sleep quality (Fernandez-Mendoza and Vgontzas 2013;Heo et al. 2013;Yazdi et al. 2014;Yong et al. 2017).
We found significant associations between physical and mental health statuses and sleep disorders. A consistent high prevalence of sleep disorders was reported in chronically ill patients (Ibrahim and Wegdan 2011;Abd Elsadek et al. 2019;Mansour et al. 2020); however, many longitudinal reports indicated sleep disorders as a risk factor for physical and mental health problems, suggesting bidirectional associations of sleep disorders and health status (Alvaro et al. 2013).
One of the current study's strengths is the subtype analyses of sleep disorders which gave a clear understanding of the correlates. For example, the number of minors in the family emerged to associate with the women's interrupted sleep and waking up too early, which is plausible considering the caregiving role of women in the family. Another example was that job demands were associated with all subtypes of sleep disorders in females. On the other hand, in men, job demands and shift work was associated with difficulties falling asleep, job control was associated with interrupted sleep, and working an extra job was associated with waking up too early and waking up tired. Some researchers pointed out that work overload and shift work could be the most significant occupational precipitating factors of insomnia (Heo et al. 2013;Pilcher & Morris 2020).
Study limitations that should be mentioned include the inability to infer causality for the observed associations between the studied factors and sleep disorders because this was a cross-sectional study. Second, the recruited participants were public officials; thus, we cannot guarantee the generalizability of the study findings to employees in the free enterprise. Third, we studied common self-reported sleep disorders of the JSEQ, which showed a high validity in Egyptians (Eshak 2019); however, JSEQ does not cover other sleep disorders such as sleep apnea syndrome, restless leg syndrome, and hypnotic medication use. Indeed, the prevalence of sleep disorders could vary by using different diagnostic tools to estimate sleep disorders. Longitudinal studies using polysomnography are recommended to confirm the observed associations.

Conclusions
There were gender differences in the prevalence of sleep disorders and the factors correlated with them in Egyptian public officials. Occupational factors were associated with sleep disorders in men and women, while the family's income and the number of minors were correlated in women more than in men. Sleep disorders could consume the employees' health and the organizations' economy; therefore, we recommend the Egyptian government improve the working environment and provide more efficient health coverage for the public officials' physical and mental health problems. Meanwhile, public officials should pay attention to their lifestyle and living arrangements to improve their sleep quality.