Determinants of Health Inequalities in Iran and Saudi Arabia: A Systematic Review of the Sleep Literature

Sleep health inequalities represent an increasing public health concern. Among multiple determinants affecting sleep health, there is people’s socioeconomic status (SES), and no systematic review on the relationship between SES and sleep health has been previously conducted in Iran and Saudi Arabia. Following the Prisma protocol, ten articles were selected. Findings revealed that the combined number of participants was N = 37,455 participants, including 73.23% of children and adolescents (n = 27,670) and 26.77% of adults (n = 10,026). The smallest sample was N = 715 and the larger was N = 13,486. In all these studies, sleep variables were assessed using self-reported questionnaires. The studies conducted in Iran assessed the risk of obstructive sleep apnea (OSA), while those in Saudi Arabia were interested in sleep duration, nap time, bedtime, rise time and insomnia. The studies performed on adult populations in Iran and Saudi Arabia concluded that there is no significant association between SES determinants and sleep components in adult populations. One study in Iran found a significant association between parent’s low SES and children and adolescent insomnia; and one study in Saudi Arabia found a significant association between the father’s education and the longer sleep duration of their children. More longitudinal studies are necessary to establish a causal relationship between public health policies and sleep health inequalities. An extension of the investigation to more sleep disturbances is required to cover the entirety of sleep health inequalities in Iran and Saudi Arabia.


Introduction
Health inequalities are associated with economic and societal impact on every government, regardless of GDP and the global well-being of an entire population [1][2][3][4][5][6][7][8]. Health inequalities can be observed with the cardiovascular system [9], respiratory system [10] or sleep mechanisms [1,2]. Sleep is a multifactorial process that may be disturbed by a wide range of environmental, sociological, physiological stimuli and socioeconomic status (SES) [4,5,11]. The concept of sleep health provides a positive holistic framing of multiple sleep characteristics, including sleep duration, continuity, timing, alertness and satisfaction [4]. Sleep health inequalities represent a public health concern similar to public health issues previously reported for cardiovascular, mental health and metabolic diseases [12], while SES is a latent multidimensional concept of an individual's economic and socioecological situation, also associated to public health inequalities at different levels such as community or country [5,11]. Thus, sleep health disparity is a complex assessment of socio-ideological and theoretical constructs measured in a variety of ways, usually considering several determinants such as employment, income, education, occupation and social position [3,[11][12][13].
An extensive screening of the empirical literature revealed that no systematic review on the relationship between SES and sleep health has been previously conducted in Iran and Saudi Arabia. It is pertinent to understand whether public health inequalities in terms of sleep disturbances that are observed elsewhere are the same as in these two important countries, especially considering the crossover of different continents and different ethnocultural populations. This systematic review aims to document the determinants of sleep health inequalities in Iran and Saudi Arabia.

Characteristics of Studies Selected
Descriptive synthesis. Ten articles [14][15][16][17][18][19][20][21][22][23] were included in the final sample (Table 1). They were all cross-sectional studies and published between 2006 and 2019. Among these studies, 50% (n = 5) were performed in Saudi Arabia and 50% (n = 5) in Iran. The combined number of participants was N = 39,550 participants, composed of 69.35% children and adolescents (n = 27,429, mean age = 9.8 years old) and 30.65% adults (n = 12,121, mean age = 46.3 years old). The smallest sample was N = 715 and the largest were N = 13,486. The socioeconomic indicators used were perceived SES/composite indices, education, income, marital status and employment/occupation. In all the studies, sleep variables were reported using self-reported questionnaires. The studies conducted in Iran assessed principally the risk of obstructive sleep apnea (OSA), while those in Saudi Arabia were interested in sleep duration, nap time, bedtime, rise time and insomnia.
Sleep fragmentation. Four cross-sectional studies [18,[21][22][23] have studied the association between education (not educated, high school or less and university) with the outcome of the risk of obstructive sleep apnea disorder and insomnia. Amra et al. [18] (adults from the general population aged from 18 to 70 years old) and Seyedmehdi et al. [21] (staff employees at a hospital with mean age at 33.5 years old +/− 7.65) were rated as good quality articles and used the Berlin Questionnaire. Foroughi et al. [22] (adults from the general population with mean age at 40.88 +/− 15.4) had used the Stop Bang. It was rated as a poor-quality study. Ahmed et al. [23] used the ICD-2 instrument to assess insomnia symptoms. No significant result in the multivariate analysis was found in all three articles.
Sleep latency. One cross-sectional study [23] was performed on adults (mean age at 42.3 years +/− 15.5). It had used the ICD-2 criteria to test the association between education and having insomnia, defined by the presence of any of the following symptoms: (1) early morning awakening, (2) difficulty initiating sleep, (3) frequent awakening, (4) feeling tired and (5) sleep loss affects mood. This study was rated as fair. Overall, uneducated people had more risk of insomnia than those attended university. Those who attended high school had a lower risk of insomnia compared to those who attended university. This study concluded that insomnia is most prevalent among females and the elderly population.
Total sleep time. The cross-sectional study of Althakafi et al. [16] had been conducted using an online questionnaire. The mean age of participants was 21.8 years +/− 10.7. It was rated as having a good quality. It had studied the association between education and the duration of sleep and found no significant association. TST during weekdays was significantly more in students whose mothers had high school or higher education (8.5-8.7 h) compared to students whose mothers were illiterate or had elementary school education (8.1 ± 1.2 h and 8.2 ± 1.3 h, respectively, p < 0.001).
• TST during weekdays was significantly more in students whose mothers were working (8.6 ± 1.1 h) compared to students whose mothers were housewives (8.4 ± 1.2 h, p = 0.005 • Children whose mothers had high school education or higher had longer night-time sleep duration compared to children whose mothers had lower education.

•
No association for TST and job status after controlling the cofounders.

•
No relation could be found between father's level of education or job status and student's sleep

Determinants of Sleep Health Inequalities in Iran
Four cross-sectional studies assessed the relationship between SES and subjective sleep measurement. One of these was conducted on children and adolescents (mean age = 12.47 years old) and the others with adults.
For the study conducted on children and adolescents, SES was measured by a SES score summarized by parents' education and occupation, possessing a private car, school type (public/private), home type (private/rented) and having a personal computer at home. Sleep variables were sleep duration and insomnia assessed by a non-specific questionnaire administered both to parents and children by interviewers.
For those conducted on adults, the SES measurement was the educational level. Sleep variables were risks for obstructive sleep apnea assessed by the self-reported Berlin questionnaire in two studies and the Stop Bang questionnaire in one study. One of these studies assessed, additionally, the quality of sleep summarized by difficulty falling asleep, difficulty staying asleep and waking up too early.
Overall, there was no association between SES and sleep variables. Specifically, only the study conducted in children and adolescents reported an association between low socioeconomic conditions of parents and insomnia. The association between sleep duration and SES was not assessed in this study. In adults, no association between educational level and self-reported risk of OSA was found.

Determinants of Sleep Health Inequalities in Saudi Arabia
Among the five studies conducted in Saudi Arabia, two were conducted on children, one on adolescents and two on adults. For those conducted on children, SES was measured by the educational level of parents, job status of mothers and family income. For the one conducted on adolescents, the SES measurement was self-reported perceived socioeconomic status. For the last two articles on adults, SES was measured by educational level, marital status, occupation and job status. Sleep variables were principally night sleep duration. In one study conducted on children nap duration, bedtime and rise time were also measured. For the studies conducted on adolescents and adults, sleep duration was used to characterize sleep deprivation which was defined by sleeping less than 7 h daily.
Overall, no association was found between SES and sleep duration or sleep deprivation. Only one study conducted on children reported that a high educational level of fathers was associated with longer sleep duration.

Summary of Findings
All these ten studies were cross-sectional studies including a total number of 39,550 participants, an average of 55.94% of women in Iran and 52.92% of women in Saudi Arabia. The majority of participants were children and adolescents (N = 27,670, 73.23%). The SES indicators used were perceived SES/composite indices, education, income and employment/occupation. Sleep components were reported using self-reported questionnaires. Studies conducted in Iran assessed principally the risk of obstructive sleep apnea while those in Saudi Arabia were interested in sleep duration, and additionally in nap time, bedtime, rise time and insomnia. The studies performed on adult populations in Iran and Saudi Arabia concluded that there is no significant association between SES and sleep components in adult populations. One study in Iran found a significant association between the parent's low SES and children and adolescent insomnia, and one study in Saudi Arabia found a significant association between the father's education and longer sleep duration of their children.

Relation with Current Knowledge
The novelty and the significance of this work is the epidemiological investigation of health determinants in a world region not well documented in the field of health inequities and sleep science. The current findings in this research are mitigated because they confirm an association between the parent's SES and sleep disturbances, but they reported no association between SES and sleep health for adult populations. The majority of studies investigating sleep health inequalities found, until now, an important trend in favor of an association between SES and sleep inequalities [2,3,12,[24][25][26][27][28]. Most of them seem to support the hypothesis stating that low SES is associated with an increase in sleep disturbances for children, adolescents and adults [1,10,25,[29][30][31][32], while a moderate or a high SES is associated with a decrease in sleep disturbances for the same sub-groups of the general population [2,4,[10][11][12]. One reason explaining this empirical discrepancy is the small quantity of research on sleep health inequalities. As revealed by this review, the empirical literature included peer review papers published from 2006 to 2019. It seems relatively new compared to sleep science or public health research related to sleep in other countries, where a few studies on the same subject were published almost twenty years before the first study included in this systematic review [33,34]. Another potential reason explaining the current findings is the experimental design, which is essentially cross-sectional. It is commonly admitted and demonstrated that a cross-sectional study is one of the best experimental designs to rigorously collect data and quickly release pertinent results on a subject [35]. At the same time, it is missing a lot of causal information and correlation, which is very often available with longitudinal study design [36]. It is difficult to establish the causal effects of SES on sleep disturbances in these circumstances. The third reason hypothesized is the incomplete conceptualization of SES in this public health literature. The majority of these articles employed education as a key component to assess the health inequalities of their sample. Education alone cannot capture the complexity of socioeconomic determinants of health inequalities in general and sleep health inequalities in particular [5,6,11,12,37]. It is important to analyze other indicators such as income and employment to understand how far sleep disturbances are important for public health in the countries concerned.

Recommendations for Future Research
The first recommendation is to perform longitudinal studies to capture the effect of time in the relationship between SES, public health and sleep. Life conditions are markers of health status variance in time and with changes in demography and lifestyle in certain subgroups of the general population, it might be interesting to observe how sleep patterns evolve negatively or positively for future public health policies.
The second recommendation is to use a quantitative measure of sleep components to collect unbiased data. Using actigraphy or polysomnography will add an additional layer of evidence in support of future appropriate public health programs.
The third recommendation is to extend the scope of future research by investigating other sleep disorders. The current literature in Iran and Saudi Arabia focuses on insomnia and OSA, even if a few general sleep parameters such as sleep quality and sleep duration were also explored. Sleep science is very wide and multidisciplinary. An integration of different conceptualizations and theoretical models developed by sleep scientists based on inputs from sociology, psychology or health economics are a few examples of possible actions to improve future research and clinical practice.

Limitations of This Review
There was no objective sleep measurement in all studies included. The self-assessment of sleep outcomes is very often used in cross-sectional studies, but it was recently demonstrated that the SES influence on health can be observed with quantitative methods [1,2,38]. So, future studies in this region should consider moving forward the knowledge by using an objective measure of sleep to verify trends reported by local authors. There are also a lot of sleep disorders not studied, while a few such as OSA can be linked to other sleep disorders such as excessive daytime sleepiness [8,39,40], and thus have to be studied. Finally, the heterogeneity of SES involved a deep consideration of the sociological approaches and psychological theories in biomedical research, in order to be able to accurately capture this complex concept.

Inclusion and Excusion Criteria
Empirical studies were defined as being of any design (cross-sectional, retrospective or longitudinal) that evaluated humans of any age, gender or race/ethnicity from the general populations of Iran and Saudi Arabia. The article should include an objective or a subjective measure of SES, such as education, income, assets, occupation, employment status, perceived SES or self-reported by participants. Indirect measures of SES, such as area deprivation or social class, were also included when individual data were not available. For studies examining children or adolescents, perceived family SES measures such as parental education, parental profession or annual household income were used instead. Studies were excluded based on the following criteria: (1) they were interventional trials, reviews or meta-analyses, case series or case reports, commentary, editorial or did not present original research; (2) they were not written in English, French or Arabic; (3) the full text was not accessible; (4) authors/researchers recruited participants that already presented specific conditions at the beginning of the study (for example, pregnant women or people with chronic sleep disturbances, those taking medication such as sleep aid, etc.); (5) they did not provide statistical significance in cases where either SES or sleep were evaluated as covariates or mediators.

Inclusion and Excusion Criteria
Empirical studies were defined as being of any design (cross-sectional, retrospective or longitudinal) that evaluated humans of any age, gender or race/ethnicity from the general populations of Iran and Saudi Arabia. The article should include an objective or a subjective measure of SES, such as education, income, assets, occupation, employment status, perceived SES or self-reported by participants. Indirect measures of SES, such as area deprivation or social class, were also included when individual data were not available. For studies examining children or adolescents, perceived family SES measures such as parental education, parental profession or annual household income were used instead. Studies were excluded based on the following criteria: (1) they were interventional trials, reviews or meta-analyses, case series or case reports, commentary, editorial or did not present original research; (2) they were not written in English, French or Arabic; (3) the full text was not accessible; (4) authors/researchers recruited participants that already presented specific conditions at the beginning of the study (for example, pregnant women or people with chronic sleep disturbances, those taking medication such as sleep aid, etc.); (5) they did not provide statistical significance in cases where either SES or sleep were evaluated as covariates or mediators.

Data Availability Statement:
No new data were created or analyzed in this study. Data sharing is not applicable to this article.