Prevalence of Sleep Disturbances in Latin American Populations and Its Association with Their Socioeconomic Status—A Systematic Review and a Meta-Analysis

Background: The worldwide increase in the prevalence and incidence of sleep disturbances represents a major public health issue. Among multiple determinants affecting sleep health, an individual’s socioeconomic status (SES) is the most ignored and underestimated throughout the literature. No systematic review on the relation between SES and sleep health has been previously conducted in Latin America. Methods: PRISMA guidelines were used. Results: Twenty articles were included in the final sample (all cross-sectional studies), and twelve among them were rated as fair or poor quality. Among these studies, 80.0% (n = 16) were performed in Brazil, 10.0% (n = 2) were performed in Peru, 5.0% (n = 1) were performed in Chile, and 5.0% (n = 1) were multicentric (11 countries). The combined total number of participants was N = 128.455, comprising 3.7% (n = 4693) children, 16.0% (n = 20,586) adolescents, and 80.3% (n = 103,176) adults. The results show the following: (1) The sleep outcomes analyzed were sleep duration, sleep quality/sleep disturbance, insomnia, excessive daytime sleepiness (EDS), obstructive sleep apnea (OSA)/sleep-disordered breathing (SDB) symptoms, and bruxism. (2) The most used determinants were income, education level, employment status/occupation, wealth/assets, and composite indices. (3) Higher SES was associated with shorter sleep duration. (4) Lower SES was associated with a decrease in sleep quality, less frequent snoring, more prevalent EDS, and sleep bruxism. (5) Lower education was associated with insomnia. (6) Higher education was associated with more sleep bruxism. (7) The pooled prevalence using a meta-analysis of the random effects model was 24.73% (95%CI, 19.98–30.19), with high heterogeneity (I2 = 100%). (8) The prevalence of sleep disturbances decreased with high education (OR, 0.83; 95%CI, [0.69–0.99]; I2 = 79%), while it increased with low income (OR, 1.26; 95%CI, [1.12–1.42]; I2 = 59%), unemployment (OR, 2.84; 95%CI, [2.14–3.76]; I2 = 0%), and being a housewife (OR, 1.72; 95%CI, [1.19–2.48]; I2 = 55%). Discussion: This meta-analysis shows that lower SES (education, income, and work) was associated with sleep disturbances in Latin America. Therefore, sleep disturbance management should be addressed with a multidimensional approach, and a significant investment in targeted public health programs to reduce sleep disparities and support research should be made by the government before the situation becomes uncontrollable.


Introduction
Promoting a healthy sleep is an ongoing project which never ends.Sleep management, with our modern lifestyle, is a complex public health outcome requiring multidimensional interventions at the economic level, the populational level, and the societal level [1,2].Mental health is also highly dependent on sleep, which has a recognized impact on several brains functions [3] and global health status, as well as stress [4].In addition, sleep has a significant influence on mental health due to its relationship with people's socioeconomic status (SES) and its connection with multiple biological systems involved in neurological disorders, such as anxiety [1,[5][6][7][8][9][10].In others words, sleep disturbances are mental disorders resulting from complex socioecological and economic interactions between the brain, the society in which we live, global health, and SES [7,11,12].Thus, sleep health inequalities represent a mental health outcome similar to the public health issues previously reported for cardiovascular, mental health, and metabolic diseases [11].The empirical literature in Western countries seems to validate the hypothesis that low-SES individuals reported more sleep disturbances than high-SES people [13][14][15][16][17][18].Similar evidence exists in Asia [19,20] and Oceania [21,22].Thus, sleep health disparity could be a public health issue in other areas, like Africa and Latin America.Few studies have documented sleep health disparities in South and Central America due to a wide range of determinants, such as employment, income, education, occupation, and social position [11,12,23].An exhaustive evaluation of the public health literature related to Latin America revealed that no systematic review on the relation between SES and sleep health has been previously conducted in Latin America.It is important to analyze if trends related to the influence of SES determinants of mental health on sleep health observed in Western countries are following the same patterns as in Latin America.The goals of this systematic review are to (1) document the global prevalence of sleep disturbances in Latin American populations and (2) document the influence of SES on different types of sleep disturbances.

Inclusion and Exclusion Criteria
Empirical studies were defined as those of any design (cross-sectional, retrospective, or longitudinal) assessing humans of any gender, race/ethnicity, and age from the general population of any country from Latin America.The article should include an objective or a subjective measure of SES, such as income (monthly personal income, monthly family income, per capita income, and annual household income), educational level, wealth, profession/occupation, employment status, and perceived SES or self-reported SES.Proxy measures of SES, such as neighborhood and social class, were also included.Every sleep component, like sleep duration and sleep quality; and every sleep disturbance, like insomnia, excessive daytime sleepiness (EDS), obstructive sleep apnea (OSA), sleep-disordered breathing (SDB) symptoms, and bruxism, were considered dependent variables.For articles with samples composed of children and adolescents, perceived family SES measures, such as parental education, parental occupation, or annual household income, were used instead.Studies were excluded based on the following criteria: (1) they were interventional trials, every type of reviews (narrative, overview, systematic, umbrella, and meta-analyses), case series, case reports, conference series, or any writing without original research (editorial, commentary, letter to editors); (2) they were articles that did not provide statistical significance in cases where either SES or sleep was evaluated as a covariate or mediator; (3) the full text was not accessible; (4) the authors/researchers recruited participants with medical conditions at the beginning of study (for example, people with medications, including sleep pills; people with cancers; people with neurodegenerative diseases, etc.); and (5) articles that were not written in English, French, Portuguese, Arabic, or Spanish (the languages of the authors).

Inclusion and Exclusion Criteria
Empirical studies were defined as those of any design (cross-sectional, retrospective, or longitudinal) assessing humans of any gender, race/ethnicity, and age from the general population of any country from Latin America.The article should include an objective or a subjective measure of SES, such as income (monthly personal income, monthly family income, per capita income, and annual household income), educational level, wealth, profession/occupation, employment status, and perceived SES or self-reported SES.Proxy measures of SES, such as neighborhood and social class, were also included.Every sleep component, like sleep duration and sleep quality; and every sleep disturbance, like insomnia, excessive daytime sleepiness (EDS), obstructive sleep apnea (OSA), sleep-disordered breathing (SDB) symptoms, and bruxism, were considered dependent variables.For articles with samples composed of children and adolescents, perceived family SES measures, such as parental education, parental occupation, or annual household income, were used instead.Studies were excluded based on the following criteria: (1) they were interventional trials, every type of reviews (narrative, overview, systematic, umbrella, and metaanalyses), case series, case reports, conference series, or any writing without original research (editorial, commentary, letter to editors); (2) they were articles that did not provide statistical significance in cases where either SES or sleep was evaluated as a covariate or mediator; (3) the full text was not accessible; (4) the authors/researchers recruited participants with medical conditions at the beginning of study (for example, people with medications, including sleep pills; people with cancers; people with neurodegenerative diseases, etc.); and (5) articles that were not written in English, French, Portuguese, Arabic, or Spanish (the languages of the authors).

Selection of Evidence and Data Extraction
Two reviewers (FAES and SZ) independently reviewed the titles and abstracts of the studies identified by the search strategy and determined eligibility for inclusion, while disagreements were resolved by consensus with a third external reviewer.For studies that passed the initial screening, the entire text was collected, and the screening process was repeated by the same co-authors to conclude with the final articles included, validated by

Selection of Evidence and Data Extraction
Two reviewers (FAES and SZ) independently reviewed the titles and abstracts of the studies identified by the search strategy and determined eligibility for inclusion, while disagreements were resolved by consensus with a third external reviewer.For studies that passed the initial screening, the entire text was collected, and the screening process was repeated by the same co-authors to conclude with the final articles included, validated by the third reviewer.Then, four reviewers (FTS, RQR, MC, and FAES) extracted from each report the following study characteristics: population, % women, age, sample size, SES measures, relevant statistics, interaction or mediation, sleep measures, conclusions/main effects, statistical methods, and results' significance.

Quality Rating of Studies
The National Institute of Health's (NIH) Quality Assessment Tool for Observational Cohort and Cross-Sectional Studies was used to rate the quality of the included studies [25].This quality rating tool was used to analyze fourteen quality criteria, asking an equal number of questions about study objectives, population, exposures, outcomes, followup rates, and statistical analysis.SES was considered the exposure variable, and sleep measures the outcome variable, respectively.Overall quality ratings were calculated by taking the proportion of positive ratings on the sum of applicable criteria.Studies with a <50% positive rating were judged as poor quality, ≥65% as good quality, and the rest as fair quality [25].

Study Outcomes
The primary outcomes were (1) to determine the prevalence of sleep disturbances (sleep duration, sleep quality, insomnia, EDS, OSA/SDB symptoms, and bruxism) in Latin American populations and (2) to document the influence of SES (income, education, employment/occupation, wealth/assets, and composite indices) on different type of sleep disturbances.These were explored through an analysis of education, income, employment status, and perceived SES, if they were reported by at least two independent studies.The secondary outcomes were to (1) analyze the relationship of our findings with current public health literature and (2) promote a multidimensional approach for sleep management.

Data Analysis
A meta-analysis was performed using the meta package on R with RStudio interface (Version 4.1.3,R Core Team, 2022) to analyze the collected data.
In each study, the prevalence of sleeping problems in Latin America was obtained.Studies that did not report the prevalence were excluded.The random effects model was used with the logit transformation for obtaining the pooled results, because it produces more conservative results than fixed-effects models, regardless of the heterogenicity scores [26].The pooled prevalence estimates of sleeping problems in Latin America were presented as a percentage with 95% confidence intervals (CIs), using a forest plot.Also, a subgroup analysis was conducted for different sleep issues, cities, study quality, and age group to assess the contribution of each study to overall heterogeneity in the prevalence of sleep disturbance.
To measure the relationship between SES and sleeping problems, we extracted from the selected publications the adjusted odds ratio (aOR) with 95% confidence intervals (CIs).Their standard errors were calculated from the respective CIs.The value from each study and the corresponding standard error were transformed into their natural logarithms to stabilize the variances and to normalize the distribution.The pooled OR (and 95%CI) was estimated using a DerSimonian-Laird random effect model.In situations in which a study reported effect estimates for independent subgroups, the subgroups were treated as individual studies in the meta-analysis.
The test of the overall effect was assessed by using Z-statistics at p < 0.05.The heterogeneity among the included studies was assessed using Cochran's Q test and I 2 statistics.The thresholds 25%, 50%, and 75% were used to indicate low, moderate, and high heterogeneity, respectively [27,28].A funnel plot based on Egger's regression test was used to evaluate the publication bias [29].In all analyses, a p-value less than 0.05 was considered statistically significant.

SES was not associated with OSA
Low income was a protective factor for males (OR, 0.4; CI, 0.1-0.9),but not significant in females (p = 0.057) SES had a significant indirect effect on bruxism via sucking habits (SC = −0.08;p = 0.01).SES had a significant direct effect (SC = −0.16;p = 0.01) and the total effect on tooth wear was also significant (SC = −0.17;p = 0.00).
Lower SES was associated with more sleep bruxism The effect was mediated by sucking behavior (finger sucking, biting nails or other objects)

Descriptive Synthesis of Articles
Quantitative analyses are presented in the results section.Presented below is a descriptive analysis, which provides a deeper examination of the overall findings that were not considered in the quantitative analysis.Table 1 presents details of the individual studies included in the descriptive analysis.
In contrast, the poorest wealth index and being unemployed or not studying were associated with lower sleep percentage in another study with adults [49].One study did not find consistent associations between sleep duration and maternal education or family income in children [43].The overall quality of the selected studies was good for three studies [37,40,49] and fair for four studies [36,38,43,45].
Globally, a lower SES was associated with diminished sleep quality.Specifically, low income [32,37] and unemployment [33,46] were associated with impaired sleep quality.Two studies indicated a higher prevalence of sleep disturbance in women [33,46].More educated adults had significantly less sleep disturbances [32,35,48].In contrast, one study found that higher maternal education was associated with lower quality of sleep in students [44], but this association was not found in infants in another study [43].Additionally, psychiatric comorbidities [32,33] and alcohol and drug consumption [33,35] were also associated with sleep disturbance.

Insomnia
Concerning insomnia, two cross-sectional studies assessed its relationship with SES [30,35].One study evaluated only women [35], and one study assessed adults in general [30].The overall quality of the two studies was good for one study [35] and fair for the other [30].
In both of the aforementioned studies, insomnia was independently associated with individuals with less education [30,35].Moreover, alcohol and drug consumption was also associated with insomnia, according to another study [35].

Excessive Daytime Sleepiness
Three cross-sectional studies approached the association between excessive daytime sleepiness and SES [31,39,48].All studies assessed adults.The overall quality of the studies selected was good for one study [31], fair for another study [39] and poor for the third study [48].
Largely, excessive daytime sleepiness was associated with a lower SES in one the studies [39].Additionally, it was also significantly more prevalent in individuals with a lower family income [31] and less education [48].
3.2.5.Obstructive Sleep Apnea (OSA)/Sleep-Disordered Breathing (SDB) Symptoms Concerning OSA/SDB, two cross-sectional studies assessed their relationship with SES [34,39].Both studies evaluated adults in the general population.The overall quality was good in one study [34], and it was fair in the other study [39].
In one study, lower SES was associated with less frequent snoring.However, no significant association was found between SES and observed apneas [34].The other study did not find any association between SES and OSA [39].
Among these three studies, two studies reported that sleep bruxism was independently associated with higher SES, including higher education [41,47].However, the other study, which was conducted by Mota-Veloso et al. [42], found that SES had a significant indirect effect on bruxism via sucking habits (finger sucking and biting nails or other objects) and that a lower SES was associated with more sleep bruxism [42].
The overall pooled prevalence for sleep disturbance in Latin America was 24.73% (95% CI, 19.98-30.19;I 2 = 100%) (Figure 2A).To decide whether to include all of the articles examining sleep disturbances or not, a publication bias chart was created.The results showed that publication bias was not significant (p = 0.059) (Figure 3).

Sleep Length in Latin America
A total of two articles [40,49] reported the sleep length (Figure 2B).Where results were reported for men and women separately, they were entered into the analysis as separate studies.In the pooled analysis, sleep length was significantly higher in men with sleep disturbances than in women with sleep disturbances, with a standardized mean of 0.40 h (95% CI, 0.34-0.47;p = 0.11; I 2 = 100%).

Subgroup Analysis
Because of a high level of heterogeneity across the included studies, a subgroup analysis was performed by region (cities), age group, and quality of study in relation to the principal outcome variable.The analysis revealed that the prevalence of sleep disturbances among infants, i.e., , was greater than that of adolescents, i.e., 27.11 (95% CI 16.55-41.09),and that of adults, i.e., .Also, the pooled prevalence of sleep disturbances was higher in Brazil, 28.52 (95% , than in Peru, 15.91 (95% CI 6.17-35.27)(Table 3).

Sleep Length in Latin America
A total of two articles [40,49] reported the sleep length (Figure 2B).Where results were reported for men and women separately, they were entered into the analysis as separate studies.In the pooled analysis, sleep length was higher in men with sleep disturbances than women without any significant differences, with a standardized mean of 0.40 h (95% CI, 0.34-0.47;p = 0.11; I 2 = 100%).

Subgroup Analysis
Because of a high level of heterogeneity across the included studies, a subgroup analysis was performed by region (cities), age group, and quality of study in relation to the principal outcome variable.The analysis revealed that the prevalence of sleep disturbances among infants, i.e., , was greater than that of adolescents, i.e., 20.33 (95% CI 12.68-30.95),and that of adults, i.e., .Also, the pooled prevalence of sleep disturbances was higher in Brazil, 25.00 (95% CI 19.54-31.40),than in Peru, 15.91 (95% CI 6.17-35.27)(Table 3).

Risk Factors
A meta-analysis was possible for sleep disturbance prevalence with three SES factors (education, income, and employment status).
Separating the education analyses according to the quality of the studies did not reveal a significant subgroup effect for sleep disturbance prevalence (p = 0.70; Figure 6A).Similarly, when the education analyses were separated according to city, no significant subgroup effect for the prevalence of sleep disturbances was observed (p = 0.70; Figure 6B).

Work and Sleep Disturbances
Data on the relationship between work and sleep disturbances are shown in Figure 5C.The meta-analysis showed a significant relationship between unemployment and the prevalence of sleep disturbances (OR, 2.84; 95%CI, [2.14-3.76]),with no heterogeneity between studies (I 2 = 0%).Similarly, in the pooled analysis, being a housewife was associated with a high prevalence of sleep disturbances (OR, 1.72; 95%CI, [1.19-2.48]),with moderate heterogeneity between studies (I 2 = 55%).

Detailed Summary of Findings
The sleep outcomes analyzed were sleep duration, sleep quality/sleep disturbance, insomnia, excessive daytime sleepiness (EDS), obstructive sleep apnea (OSA)/sleep-disordered breathing (SDB) symptoms, and bruxism.The most used determinants were income, education level, employment status/occupation, wealth/assets, and composite indices.
A higher SES was associated with lower sleep duration, and a lower SES was associated with a decrease in sleep quality, less frequent snoring, more prevalent EDS, and sleep bruxism.Lower education was associated with insomnia, and higher education was associated with more sleep bruxism.For the 20 articles included, 12 were rated as fair or poor in study quality.Therefore, a meta-analysis was performed to estimate the prevalence of sleep disturbances in Latin America and the main SES risk factors that could be associated with it.The pooled prevalence, using a meta-analysis of the random effects model, was 24.73% (95%CI, 19.98-30.19), with high heterogeneity (I 2 = 100%).The meta-analysis showed that the prevalence of sleep disturbances decreased with high education (OR, 0.83; 95%CI, [0.69-0.99];I 2 = 79%), while it increased with low income (OR, 1.26; 95%CI, [1.12-1.42];I 2 = 59%), unemployment (OR, 2.84; 95%CI, [2.14-3.76];I 2 = 0%), and being a housewife (OR, 1.72; 95%CI, [1.19-2.48];I 2 = 55%).

Relationship with Public Health Literature
Epidemiologic data continue to increase the literature about the influence of sleep on the general population's health status.Sleep plays a vital role in several body functions, as well as health disparities.The scientific community is still investigating external and environmental factors affecting sleep mechanisms, but there is still a lot that is unknown.Based on the current findings, it can be hypothesized that sleep disturbances are associated with socioeconomic status, as suggested by many other studies [17,[50][51][52][53][54][55][56][57].The fact is that the gradient of health disparity existing for some diseases, like cardiovascular illness, seems the same for sleep.Regardless of the region of the world where the investigation is conduced, sleep disparities are observed because our findings on Latin America support the previous results [17,[50][51][52][53][54][55][56][57].Our findings are consistent with those of previous studies from places other than Latin America, and they are additional arguments in favor of the establishment of a more efficient worldwide program framing sleep health management.

The Necessity of a Multidimensional Sleep Management
Sleep disturbance management should be addressed by a multidimensional approach.Recent epidemiological studies performed outside Latin America in different public health contexts reported significant associations of sleep with stress [4], work conditions [8], environment [58], and employment [17,59,60], and they also revealed latent interactions existing between government policy and public health strategies [22,54,[61][62][63] (Figure 7).Obviously, a government's economic policy influences the funding of public health programs.Similarly, SES directly influences health status regardless of the disease, as assessed through individuals' living conditions and their resulting behavioral risk factors and stress.Knowing that sleep disparities can be measured objectively and quantitatively [6,7,49], our suggestion for governments is to invest as soon as possible in preventive management programs for sleep disturbances before they become uncontrollable.It was already documented how expensive absenteeism and presenteeism due to sleep disturbances are for the economy [64,65], but diverse governments did not move forward yet with strong regulations to reduce these important losses [66,67].
nants and health outcomes.This first meta-analysis on sleep determinants in Latin America highlighted the high quality of cross-sectional studies published, as well as the lack of systematic review and longitudinal studies, similar to what has been found recently with the African population [69].To support public health strategies, randomized controlled trials and longitudinal studies are required with a broader objective related to the SESsleep health gradient, including the role of unhealthy behaviors, chronic diseases, and psychological factors [68].

Conclusions
This meta-analysis reveals that the prevalence of sleep disturbances accounts for almost a quarter of the health issues in Latin America, and they have been associated with lower SES, especially in terms of education, income, and work.Despite the growing body of literature around the importance of sleep, it seems non-significant enough for decision makers who still do not pay attention to this public health matter.Governmental programs should consider scientific evidence and could be funded to allow fast and permanent results in the near future, before reaching an uncontrollable point.In this meta-analysis, there is an inequal distribution of research because 80% of studies came from Brazil (Table 1).Even if Brazil is representative of Latin American populations, its public health's context regarding sleep management is not necessarily identical to that of its neighbors.More research should be performed in other Latin American countries to obtain an accurate overview of sleep disparities in this continent.Our suggestion for scientists is to not forget that cross-sectionals studies are often used to understand determinants of health and establish preliminary evidence [68]; however, they are useless when it is necessary to consider the correlation between theoretical determinants and health outcomes.This first meta-analysis on sleep determinants in Latin America highlighted the high quality of cross-sectional studies published, as well as the lack of systematic review and longitudinal studies, similar to what has been found recently with the African population [69].To support public health strategies, randomized controlled trials and longitudinal studies are required with a broader objective related to the SES-sleep health gradient, including the role of unhealthy behaviors, chronic diseases, and psychological factors [68].

Conclusions
This meta-analysis reveals that the prevalence of sleep disturbances accounts for almost a quarter of the health issues in Latin America, and they have been associated with lower SES, especially in terms of education, income, and work.Despite the growing body of literature around the importance of sleep, it seems non-significant enough for decision makers who still do not pay attention to this public health matter.Governmental programs should consider scientific evidence and could be funded to allow fast and permanent results in the near future, before reaching an uncontrollable point.

Figure 1 .
Figure 1.PRISMA flowchart of study selection process.

Figure 2 .
Figure 2. Forest plot showing the primary outcomes in 31 cross sectional studies from 20 pu reports in Latin America: (A) prevalence of sleep disturbances and (B) sleep length (hours)

Figure 2 .
Figure 2. Forest plot showing the primary outcomes in 31 cross sectional studies from 20 published reports in Latin America: (A) prevalence of sleep disturbances and (B) sleep length (hours).

Figure 3 .
Figure 3. Funnel plot for meta-analysis of the prevalence of sleep disturbances in Latin Americ Egger's test: p = 0.0598.

Figure 3 .
Figure 3. Funnel plot for meta-analysis of the prevalence of sleep disturbances in Latin America.Egger's test: p = 0.0598.

Figure 4 .
Figure 4. Subgroup analysis on the prevalence of sleep disturbances in Latin America by type of sleep disturbance.The black dot point is the estimate, and the horizontal line is the 95% CI for prevalence plotted for each study.The black diamond at the bottom of each type of sleep disturbance is the estimated average prevalence.CI: confidence interval.

Figure 4 .
Figure 4. Subgroup analysis on the prevalence of sleep disturbances in Latin America by type of sleep disturbance.The black dot point is the estimate, and the horizontal line is the 95% CI for prevalence plotted for each study.The black diamond at the bottom of each type of sleep disturbance is the estimated average prevalence.CI: confidence interval.

Figure 5 .
Figure 5. (A) Forest plot for education and sleep disturbance prevalence (compared to the reference group).(B) Forest plot for income and sleep disturbance prevalence (compared to the reference group).(C) Forest plot for employment status and sleep disturbance prevalence (compared to the reference group).Box sizes reflect the weights of studies included in the meta-analysis, horizontal

Figure 5 .
Figure 5. (A) Forest plot for education and sleep disturbance prevalence (compared to the reference group).(B) Forest plot for income and sleep disturbance prevalence (compared to the reference group).(C) Forest plot for employment status and sleep disturbance prevalence (compared to the reference group).Box sizes reflect the weights of studies included in the meta-analysis, horizontal lines are the 95% CIs, and the summary ORs are represented by the diamond.OR, odds ratio; CI, confidence interval.

Figure 6 .
Figure 6.Subgroup analyses to explore sources of heterogeneity in risk factors of sleep disturbances.(A) Forest plot demonstrating that higher education was associated with sleep disturbance prevalence by quality of the study (good vs. fair/poor).(B) Forest plot demonstrating that higher education was associated with sleep disturbance prevalence by city (Brazil vs. multicentric).

Figure 6 .
Figure 6.Subgroup analyses to explore sources of heterogeneity in risk factors of sleep disturbances.(A) Forest plot demonstrating that higher education was associated with sleep disturbance prevalence by quality of the study (good vs. fair/poor).(B) Forest plot demonstrating that higher education was associated with sleep disturbance prevalence by city (Brazil vs. multicentric).

Figure 7 .
Figure 7.The Socioeconomic and Environmental Model of Health (SEEMOH).

Figure 7 .
Figure 7.The Socioeconomic and Environmental Model of Health (SEEMOH).

Table 1 .
Characteristics of included studies investigating determinants of sleep health in Latin America.

Table 2 .
Quality rating of the included studies, using the NIH quality assessment tool.

Table 3 .
Subgroup analyses of the prevalence of sleep disturbances in Latin America.