Excessive daytime sleepiness among rural residents in Saskatchewan

*Co-principal investigators; **Saskatchewan Rural Health Study Team members are listed in the Appendix 1Division of Respirology, Critical Care and Sleep Medicine; 2Canadian Centre for Health and Safety in Agriculture; 3Department of Community Health and Epidemiology; 4Department of Medicine, College of Medicine; University of Saskatchewan, Saskatoon Correspondence: Dr John A Gjevre, Division of Respirology, Critical Care and Sleep Medicine, University of Saskatchewan, 103 Hospital Drive, Saskatoon, Saskatchewan S7N 0W8. Telephone 306-844-1009, fax 306-844-1532, e-mail john.gjevre@usask.ca Poor sleep and excessive daytime sleepiness are common complaints, and patients frequently present to their health care providers with these concerns. Excessive daytime sleepiness has been noted to have a prevalence of 5% to 23% in the general population (1-10). While daytime sleepiness can be attributed to various underlying causes, it may be a warning sign of potential sleep-disordered breathing. There is increasing recognition that obstructive sleep apnea (OSA) is a major public health issue in the North American general population (11). Furthermore, sleep disorders are very common in general practice but symptoms are often under-recognized or under-reported (12). OSA has been acknowledged as a contributor to the development of significant comorbidities, as well as being associated with an increased rate of industrial and motor vehicle accidents (13,14). Various factors, such as age (5,7,15,16), sex (6,7,17,18), marital status (19,20), smoking (21,22), obesity (23,24), socioeconomic status (6,18,25) and medical history (23,26,27), have been associated with daytime sleepiness or sleep apnea. However, there are limited data assessing sleep apnea prevalence or reviewing predictors for, and prevalence of, excessive daytime sleepiness in rural or remote populations. Saskatchewan is a large (651,036 km2), sparsely populated (population 1,072,082) province with a strong agricultural focus located in Western Canada. Of the provincial population, approximately 35% reside outside the two major urban centres in smaller communities and rural farmstead/remote settings. Of particular relevance in an agricultural and rural population, OSA indicators have been recently reported to be associated with increased injuries in farmers (28,29). oRIGINAL ARTICLE


Excessive daytime sleepiness among rural residents in Saskatchewan
Une somnolence diurne excessive chez des habitants des régions rurales de la Saskatchewan The Epworth Sleepiness Scale (ESS) is a self-administered questionnaire that has been used to measure daytime sleepiness (30)(31)(32).The ESS has been widely used by researchers, clinicians and sleep specialists as a tool to identify and assess sleep apnea (33)(34)(35).The validity and reliability of the ESS has been evaluated (33,(36)(37)(38)(39).In an earlier pilot project, we studied the prevalence of a high ESS score in 283 rural residents (40).The objective of the present study was to broaden the evaluation of excessive daytime sleepiness prevalence and determinants in the much larger rural population participating in the Saskatchewan Rural Health Study (SRHS) (41).

METHODS baseline survey design
The SRHS was designed as a prospective cohort study conducted in two phases: baseline and follow-up.Details of study design for the baseline survey have been previously reported (41).Briefly, 39 rural municipalities (RMs) of the 298 RMs in Saskatchewan and 16 of the 145 towns (generally having a population of 500 to 5000) in Saskatchewan were selected to participate in the study.These RMs and towns were selected at random from four quadrants of the province (southeast, southwest, northeast and northwest).The local councils for most of these 32 (89%) of 36 RMs and 15 (94%) of 16 towns agreed to participate on behalf of their residents and supplied mailing addresses.The method of Dillman (42,43) was used to recruit study participants.The study population comprised 8261 individuals (men and women ≥18 years of age) living in 32 RMs and 15 towns in the study area within 4624 households.Information regarding the variables described below was collected by self-administered, mailed questionnaires based on the Population Health Framework (44,45).

ESS questionnaire
The degree of sleepiness was assessed using the ESS (Box 1).The ESS score ranges from 0 to 24.A score of 11 to 24 is considered to be abnormal and indicative of excessive daytime sleepiness (30).The primary outcome of interest was a binary variable of ESS score >10.

SRHS survey questions
Smoking status: Three types of smoking history were assessed, including current smoker (smoking in the past year) or ex-smoker (no current smoking and a history of smoking at least 20 packs), or nonsmoker (all others).body mass index: Body mass index (BMI) was calculated as weight (kg)/height (m) 2 .BMI was based on self-reported information on height and weight.Overweight and obese were defined as 25 kg/m 2 to 30 kg/m 2 , and >30 kg/m 2 , respectively.Marital status: Marital status was categorized into two groups: married, common law or living together; and widowed, divorced, separated or single.Alcohol: Alcohol consumption was captured using responses to the question "In the past 12 months have you had 5 or more drinks on one occasion?",and categorized as yes or no.Self-reported physician-diagnosed medical history: Sinus trouble, heart disease, heart attack, hardening of the arteries, high blood pressure, tuberculosis, stroke, attack of bronchitis, diabetes, chronic bronchitis, emphysema, chronic obstructive pulmonary disease, asthma and shortness of breath.Snoring: Data were collected using the questions "Do you snore?" and "If you snore, is your snoring: slightly louder than breathing?; as loud as talking?; louder than talking?; very loud -can be heard in adjacent rooms?".These questions were simplified into a new variable, loud snoring, by combining the above questions into two categories: no or slightly, and loud or very loud.Residence: Designation of residence as farm or nonfarm (including town and self-described acreage) rural dwelling was based on the question 'Where is your home located?' (farm, town, acreage).Town and acreage were combined to create a nonfarm category.Socioeconomic status: Socioeconomic status was assessed using household income adequacy, which was a derived variable with four categories based on various combinations of total household income and the number of people living in the household according the Statistics Canada definition (46), and a question concerning how much money was left over at the end of the month with the three categories: some money; just enough money; and not enough money.Education: Highest educational attainment was categorized into four groups: less than high school; completed high school; completed university; and completed other postsecondary education.

Statistical analysis
Statistical analysis was completed using SAS version 9.03 (SAS Institute, USA).Logistic regression models were used to predict the relationship between a binary ESS>10 (yes or no) and a set of explanatory variables.A multilevel logistic regression modelling approach, including a generalized estimating equation, with individuals (first level) nested within households (second level), was used to evaluate the effects of covariates of interest.This accounts for the within-subject dependencies that occur in the analysis due to multiple individuals from one household.A series of multilevel models were fitted to determine whether potential risk factors, confounders and interactive effects contribute significantly to the prevalence of ESS score >10.Based on bivariable analysis, variables with P<0.20 became candidates for a multivariable model.All variables that were statistically significant (ie, P<0.05) as well as important factors (location of residence), were retained in the final multivariable model.A parsimonious model was selected based on the QIC (Quasi-likelihood under the Independence model Criteria) goodness-of-fit statistic (48,49).The strength of associations is presented as ORs and associated 95% CIs.
Univariate binary logistic regression analysis showed that overweight (BMI 25 kg/m 2 to 30 kg/m 2 ) and obese (BMI >30 kg/m 2 ), male sex, older age, lower education level, married/common law marital status, lower income status, loud snoring and shortness of breath were associated with a risk of having a higher ESS score (Table 1).Also, 'doctor-diagnosed' cardiopulmonary morbidities (heart disease, heart attack, hardening of arteries, high blood pressure, tuberculosis, stroke, attack of bronchitis, diabetes, sinus troubles, chronic bronchitis and chronic obstructive pulmonary disease) were associated with a risk of having a higher ESS score (Table 2).
As shown in

DISCUSSION
Consistent with results from our pilot study (40), the present study highlights that a significant percentage of the rural population experienced symptoms of excessive daytime sleepiness.This finding has implications for potential increased risks for farm-related and other forms of accidents/injuries, as well as potential increased cardiovascular morbidity risk.
The prevalence rates of excessive daytime sleepiness significantly depend on the definition used.Previous studies have reported prevalence rates of 5% to 23% in adult populations (1-10).These studies have been summarized in Table 3.Our results using a definition of ESS score >10 for excessive sleepiness showed an overall prevalence of 15.9%.Thus, our data, in the total population studied with the overall excessive daytime sleepiness defined as an ESS score >10, are consistent with other epidemiological studies.However, our finding that 21% of men reported ESS scores >10 places our results at the upper range.
The conventional threshold ESS score is >10.However, it should be acknowledged that Aurora et al (49) recently reported a clear association between ESS score and average sleep latency as objectively measured from multiple sleep latency testing and identified an ESS score >13 as most effectively predicting objective sleepiness in their study population.Factors associated with excessive daytime sleepiness in the present study included age, male sex, obesity (BMI >30 kg/m 2 ), lower socioeconomic status, married or living with partner, loud snoring and 'doctor-diagnosed sinus trouble'.Increasing age, male sex and obesity are well recognized to be associated with increased risk for OSA (16,24).Lower socioeconomic status has also been clearly linked with an increased likelihood of obesity (50).Increased neck circumference and BMI are particularly are strong predictors of OSA in men (51).Increased neck circumference, BMI and several other truncal measures have also been reported to be associated with elevated apneahypopnea indexes in a recent study involving women referred for polysomnography (52).However, sex differences in OSA-associated symptoms and body fat distribution have been reported, which could impact identification or suspicion of OSA in women in a general population (53).Snoring has been also well recognized as an upper airway sign of possible OSA, and is included in validated identification questionnaires such as the Berlin Questionnaire and the STOP-BANG instrument (54,55).Nugent et al (6) reported that the strongest risk factor identified for excessive daytime sleepiness was a history of loud snoring (OR 2.62).We additionally observed 'sinus trouble' to be independently associated with excessive daytime sleepiness in our study population.This is consistent with previous reports linking OSA and sinus dysfunction (56,57).
A Greek study (19) reported increased risk of OSA in married subjects compared with individuals who were single.We observed similar results for excessive daytime sleepiness.Conversely, Teculescu et al (20) reported the prevalence of sleep-disordered symptoms were not significantly different between groups based on marital status in a French population.Our results also indicated that socioeconomic status using 'money left at end of the month' as an indicator was significantly associated with excessive daytime sleepiness.Other studies have demonstrated that socioeconomic status, as measured by education, occupation and income, was associated with excessive sleepiness (6,19,25,58).
Associations between smoking and OSA have been reported (56,59), as have alcohol use and OSA (60)(61)(62).In the present study, we did not observe associations between these factors and excessive daytime sleepiness.In addition, there have been some reports of depressive symptomatology associated with excessive daytime sleepiness (63).Because we did not include questionnaire assessments of mental health or depression in our survey instrument, we are unable to address the contribution of this possible risk factor.
We did not observe any association between location of residence (farm or nonfarm) or farming as the current occupation and excessive daytime sleepiness.In addition, there was a considerable north-south distance between the study quadrants (from the 49th to the 53rd parallels).One may speculate that this could lead to a small difference (approximately 40 min) in total daylight time, which could affect sleep patterns (64).However, analysis of the north to south cohorts did not show any significant difference for the prevalence of excessive daytime sleepiness.
Overall, we observed that the risk factors identified with excessive daytime sleepiness in our population were generally recognized as also associated with OSA in other populations.It is likely that this high prevalence of elevated ESS scores is a signal for elevated prevalence of OSA in our rural population.Although an elevated ESS score may lend weight to a suspicion of OSA, it should be acknowledged that  there is some controversy with regard to ESS accuracy.The association between subjective and objective sleepiness is not fully understood and the ESS scores should be interpreted with clinical correlation (49,65).In addition, there is only a limited relationship between the ESS and other sleep instruments, likely due to the questionnaires measuring different aspects of sleep (66).It is notable that although 70% of respondents in the present study reported 'snoring', only 21% of the men and 15.9% of total population had elevated ESS scores.This raises the concern that the prevalence of OSA may be substantially higher than the prevalence of reported excessive daytime sleepiness in this rural population.Particularly, in light of associated cardiovascular morbidity and increased risk of injury/accident linked to undiagnosed and untreated OSA, these findings highlight the need for further assessment of OSA prevalence in rural populations.
Although overnight polysomnography is the gold standard for diagnosis, timely access may be an issue particularly for rural populations.Identification questionnaires for OSA and home-based diagnostic testing (67,68) may be used for further evaluation in these populations.Increased OSA recognition, diagnosis and intervention may provide significant health care benefits to this population.
fUNDING: Funding was received from Canadian Institutes of Health Research "Saskatchewan Rural Health Study", Canadian Institutes of Health Research MOP-187209-POP-CCAA-11829.

DISCLOSURES:
The authors have no financial disclosures of conflicts of interest to declare.
How likely are you to doze off or fall asleep in the following situations, in contrast to feeling just tired?This refers to your usual way of life in recent times.Even if you haven't done some of these things recently try to work out how they would have affected you.0 = would never doze 1 = slight chance of dozing 2 = moderate chance of dozing 3 = high chance of dozing in a public place As a passenger of a car for an hour without a break Lying down to rest in the afternoon when circumstances permit Sitting and talking to someone Sitting quietly after a lunch without alcohol In a car, while stopped for a few minutes in traffic Total scoreCopyright © M.W.Johns 1990Johns -1997 L'apnée obstructive du sommeil (AOS) est un diagnostic courant en pratique clinique.La somnolence diurne excessive peut être évocatrice d'une AOS.ObJECTIfS : Évaluer la prévalence de somnolence diurne excessive, mesurée selon l'échelle de somnolence d'Epworth (ÉSE), au sein de la population d'une communauté rurale.Évaluer également les facteurs de risque potentiels d'AOS.MÉTHODOLOGIE : En 2010, dans le cadre de la Saskatchewan Rural Health Study, les chercheurs ont posté un questionnaire de base sur la santé respiratoire à 11 982 ménages de la Saskatchewan.Au total, 7 597 adultes au sein des 4 624 (42 %) ménages des répondants ont rempli le questionnaire d'ÉSE.

Table 2
, multivariate binary logistic regression analysis indicated that the risk of having a higher ESS score increased with