Obstructive sleep apnea syndrome: complaints and housing characteristics in a population in the United States

ABSTRACT CONTEXT AND OBJECTIVE: Obstructive sleep apnea syndrome (OSAS) is one of the leading causes of morbidity and mortality in adults. Early detection of the disorder and discovery of risk factors through standardized questionnaires will lead to reduction of the OSAS burden. The main aim of this study was to estimate the prevalence of OSAS symptoms and examine their association with housing characteristics. DESIGN AND SETTING: Cross-sectional study at a medical school. METHODS: Demographic, housing and body measurement data on 5,545 individuals aged 16 years and over of various races were selected from the National Health and Nutrition Examination Survey. We analyzed the probability of OSAS based on habitual snoring combined with daytime sleepiness and/or witnessed apnea. Univariate and multiple linear regression were used. RESULTS: 9.8% of the men and 6.9% of the women reported symptoms suggestive of OSAS (habitual snoring, daytime sleepiness and/or apnea). The following prevalences of symptoms were found among males and females respectively: frequent snoring 35.1%, 22.3%, excessive daytime sleepiness 6.4%, 3.4% and frequent apnea 14.9%, 20.6%. Using multiple linear regression, OSAS symptoms were correlated with gender, age, body mass index (BMI), marital status and education. Regarding housing characteristics, mildew or musty smell and pets in the environment were associated with a high probability of OSAS. CONCLUSION: OSAS symptoms were more prevalent than in developing countries. The environment was an important risk factor, but environmental factors are easier to control and manage than other variables like BMI or socioeconomic status.


INTRODUCTION
Obstructive sleep apnea syndrome (OSAS) is a considerable issue in public health.It is a highly prevalent disorder among middleaged adults 1 and is independently associated with certain risk factors.[4][5] Untreated OSAS can cause road accidents, 1 loss of work productivity, 5 occupational injuries 1 and even sudden death. 4Therefore, this disorder may lead to a huge multilateral problem unless proper control and management is implemented.The prevalence of obstructive sleep apnea among adults in the Western world ranges from 3% to 28%, 5 and in the United States of America (USA), the prevalence is currently estimated to be 5% to 10%. 6,7e prevalence of OSAS in some studies based on various questionnaires has been reported as follows: based on the Berlin Criteria, in New Zealand (2009) 2.8%, 2 Iran (2011) 4.9% 8 and USA (2006) 26%; 5 based on a self-reported questionnaire, in France (2007) 3.5% 1 and Hong Kong (2001) 2.1%; 9 and based on the Epworth Sleepiness Scale (ESS), in Nigeria (2008) 1.2% 10 and India (2004) 3.6%. 11It is noteworthy that in some regions only 10% of the population has been adequately screened for appropriate diagnosis. 4Discrepancies in observed prevalence and underreporting may be due to non-standardized definitions and variation between diagnostic methods. 1 OSAS screening, diagnosis and treatment entails some challenges.Polysomnography or respiratory polygraphy is a precise method for diagnosing OSAS, but this method has its disadvantages, such as expensiveness, inaccessibility and difficulty to perform.Thus, in some cases, these disorders are not diagnosed and only a few cases are properly treated. 1 Another diagnostic method is to screen by means of a questionnaire based on three symptoms: reported habitual snoring, daytime sleepiness and witnessed apnea.
Early detection of OSAS not only reduces the morbidity risk but also leads to significant reduction in the cost of care for other conditions. 2,10][12][13] Information on environmental factors affecting this outcome is unavailable.Housing characteristics are an important environmental variable; adverse conditions are preventable and can be dealt with cost-effectively, and thus may be a determinant in decreasing the burden of disease.

OBJECTIVE
The purpose of this study was to assess the prevalence of symp- NHANES was conducted in all 50 states of the USA.
NHANES data were not obtained using a simple random sample.Rather, a complex, multistage, probability sampling design was used to select participants such that they would be representative of the civilian of the civilian, non-institutionalized US population.
Oversampling of certain population subgroups was done to increase the reliability and precision of the health status indicator estimates for these groups.The NCHS used four questionnaires: demographic variables, housing characteristics, body measurements and OSAS.
After the datasets for demographic variables, housing characteristics, OSAS and body measurements had been merged, inconsistencies relating to 594 sequence numbers led these individuals to be excluded from the study.Thus, 5,545 individuals aged 16 years and over were selected, from several racial groups in the United States: Mexican American, Other Hispanic, Non-Hispanic White, Non-Hispanic Black and Other Race Including Multiracial.

Demographic variables
The population for the demographics questionnaire was interviewed directly in the subjects' homes and a proxy was provided for individuals who could not answer the questions themselves.
The variables selected for this evaluation included age, gender, marital status, education status, pregnancy status, household size and family size.

Housing characteristics
One study participant in each family responded for the entire family and these responses were released for all members of the same family.The housing characteristics provided family-level interview data on the type of home, number of apartments in the building, age of home, number of rooms in home, time lived in home, whether home was owned or rented, water source and treatment, and allergy component-related questions about the presence of furry animals.

Sleeping characteristics
This section included questions on sleep habits and disorders.
A subscale of eight questions, relating to general productivity, from the Functional Outcomes of Sleep Questionnaire, was also included. 14Variables pertaining to OSAS were selected in order to analyze the probability of OSAS based on habitual snoring (often or almost every night), combined with daytime sleepiness (often or almost always) and/or witnessed apnea (often or almost every night, as confirmed by self-reports).

BODY MASS INDEx (BMI)
All survey participants were eligible for the body measurement component.There were no medical, safety or other exclusions for the body measurement protocol.The body measurement data were collected by trained health technicians, who were accompanied by a recorder during each body measurement examination.The health technicians used their discretion to obtain as many measurements as practical for individuals who were using a wheelchair.Body weight data for individuals who had had limb amputations and also those from pregnant women were excluded from the analysis.Height was measured using a Seca electronic stadiometer, in an upright standing position, with head and heels against the stadiometer before taking the height, unless this position was anatomically impossible.Before the measurement, the participants took a deep breath and held it while the headboard was positioned.If the individuals were unable to stand with the head and heels against the stadiometer, the trunk needed to be vertical above the waist and the arms and shoulders needed to be relaxed. 14ight was measured on a Toledo digital scale, in pounds with automatic conversion to kilograms.The participants were weighed in their underwear.They were instructed to stand still at the center of the scale platform facing the recorder, with their hands at their sides, and to look straight ahead.After they had been properly positioned and the digital readout was stable, the recorder clicked on the capture button on the screen.The weight, length and height measurements were entered directly into the computer system by clicking on the "Get" buttons. 14In accordance with the World Health Organization (WHO) definitions, BMI (kg/m 2 ) less than 18.5 was considered to be underweight, BMI greater than 25 was considered to be overweight and greater than 30 was considered to be obese. 15

Analysis
After assessment to check that the data presented normal distribution, all the continuous variables were summarized in terms of mean ± standard deviation (SD) and categorical variables were expressed as percentages.All significance tests were two-sided, and P values less than 0.05 were considered to be statistically significant.Univariate and multiple linear regression were used for the analysis.Variables showing associations with P values less than 0.10 in univariate analyses were considered to be candidate risk factors to be used in multiple analysis.
The prevalence of symptoms associated with OSAS according to house characteristics is presented in Table 2.
In order to investigate variables relating to OSAS, univariate and multiple linear regression tests were performed.Variables with more than two categories were analyzed in the form of a dummy variable.These variables included race, education status, marital status and type of house.
In the univariate linear regression, BMI, age, gender, race, education status, marital status, type of house, age of home, rented or owned home, source of water, presence of mildew or musty smell, living or spending time with pets and ratio of population density in the home were variables that had significant relationships with OSAS.
The details the of analysis were as follows: the probability of OSAS increased linearly with increasing age and BMI respectively (ß = 0.087, P < 0.0001; ß = 0.026, P < 0.0001); the probability among the women was less than among the men (ß = -0.026,P = 0.0008); the probability among the Mexicans was greater than among other races (ß = 0.43, P < 0.0001); the probability was less among Non-Hispanic Blacks (ß = -0.43,P < 0.0001).
From examining the associations shown between the home environment and the disorder, the following results could be inferred: living in a mobile home or trailer produced a higher probability of OSAS than shown by other types of home (ß = 0.79, P = 0.005).In addition, a direct linear relationship was found between duration of living in a house and probability of the disease.
Homes with a well as the water source (rather than from a water supply company) and those with mildew or a musty smell had higher probability of disorders (ß = 0.79, P = 0.005) and (ß = -0.34,P = 0.003).
Living with pets was also associated with greater possibility of OSAS (ß = -0.33,P < 0.0001), but population density presented an inverse relationship with OSAS (ß = -0.63,P < 0.0001).Th e variables  Mexican American n = 1248 Co-factors Name of variable 37.9 (16.9) 39.5 (  of number of apartments in the building, home age and having cockroaches in the home did not show any association with OSAS.
In multiple linear regression analysis, the variables of age, gender, BMI, education status, marital status, mildew or musty smell, and animals living or spending time in the home presented significant relationships with OSAS.The test results are shown in Table 3.

DISCUSSION
To the best of our knowledge, this study is the first broad study on data from the USA to contain valuable information about sleep disorders and risk factors relating to housing characteristics in the presence of important variables like weight and demographic variables.The prevalence of symptoms of OSAS was 8.3%.We compared several studies conducted among adult in different countries using questionnaire instruments.The prevalence of OSAS was 1% in Nigeria 10 (0.5% among women and 1.9% among men), 3.6% in India, 16 3.1% in Hong Kong 9 and 5% in Iran; 8 however, it was 7.4% in France 1 and 26% in the USA. 5 OSAS in the United States, like in other Western countries, is more common than in developing countries.Obesity and aging are probably crucial factors in the United States, and the rising trend of these two factors requires more attention to this context. 17,18though the prevalence of risky apnea in the present study was similar to findings in Pakistan (10-12%), 19 it was more common in Malaysia (15.2%). 20On the other hand, rates if 3.5% and 6.1% were reported in France and Turkey, respectively. 1,12vertheless, it should be noted that 90% of individuals with sleep apnea are undiagnosed. 21e prevalence of habitual snoring was similar to findings from studies conducted in the USA (46%) and Malaysia (47.3%). 5,20The prevalence observed in our study was higher than in the following other regions: Nigeria 31.6%, 10 São Paulo 31% 22 and France 22%. 1 The rate of excessive daytime sleepiness in the present study (15%) was similar to findings from France (16%) and Malaysia (14.8%). 1,20On the other hand, the reported prevalence was 6.5% in the USA (2005). 5nsistent with a study conducted in France, 1 habitual snoring and apnea were more prevalent among men (35.1 and 6.4%) than among women (22.3 and 3.4%).However, EDS was more prevalent among women (20.6 versus 14.9%).The prevalence of symptoms in the French study was consistently lower (61%, 7% and 24%). 1 OSAS was less common among women, and this was similar to other studies; for instance in France, Japan, New Zealand 1,2,4 and also in Nigeria, India, and Hong Kong. 9,10Epidemiological studies have confirmed that the gender ratio of OSAS is 2 to 3.1, 10,20,21 and this ratio was 1.42 in our study.In addition to cases of more prevalent OSAS disorders among men, these differences were statistically significant in our study and some other studies 1,2,4,8,23 although a study conducted in Iran did not reach any significant variation. 812][13]17 For example, obese individuals presented a risk of OSAS that was 10 times higher in the study by Salvador et al. 27 There was also a strong relationship between these in the presence of confounding variables.There was a direct trend between age and OSAS in the present study.These results were consistent with reports from different countries 1,5 in which some of the results were analyzed from multiple tests, like in the present study. 4,8Thus, the effect of age-related chronic diseases on OSAS needs to be investigated.
Education was another variable significantly associated with OSAS, with regard to confounding variables.It was found that high school students who never failed in the end-of-year examinations and thus never had to repeat the school year presented lower probability of OSAS.There was a direct association between education status and OSAS. 8articipants whose marital status was "separated" showed less possibility of OSAS than shown by other marital statuses.In a report from Nigeria, being married was a risk factor for OSAS. 10 Although Mexican Americans presented a high risk in univariate regression, multiple regressions did not support this result.In other studies, Asian, African-American and Hispanic racial groups presented higher risk. 13e main goal in the present study was to investigate the association between housing characteristics and OSAS.Although the home environment has been reported as a factor in people's health, 29  due to moisture and the air pressure in the home environment), these can be evaluated and thus, eventually, effective management of OSAS may be achieved.
Housing characteristics should be taken into account for public health purposes for better management of OSAS in the US, as well as in clinical decision-making relating to this syndrome.

CONCLUSION
Symptoms of OSAS were more prevalent in the USA (8.3%) than in developing countries.Moreover, the environment was an important factor for OSAS.
its role in relation to sleep status had not been particularly studied.Although the variables of type of house, population density in the home, water source and the age of the house were significant in the test, their relationship did not maintain significance in the multiple model.An odor of mildew or a musty smell and living or spending time with animals had strong relationships with OSAS both in univariate and in multiple linear regression.Our results are new in terms of exposure factors associated with OSAS.Whether the mechanisms through which housing characteristics correlate with OSAS are pathological mechanisms (e.g. through the activity of some fungi such as Aspergillum in damp environments or microorganisms found in the saliva of wool or animal), or are autoimmune mechanisms (e.g. through immune reactions against allergens such as wool and animal hair), or are physiological (e.g. through affecting concentration participants' mean age was 41.8 ± 20 years.Among this sample of individuals ≥ 16 years of age from the USA, 2,826 (51%) were women and 2,719 were men (49%) and consisted of CI: 20.95-23.36),3.4% (87 subjects; 95% CI: 3.2-3.7)and 20.6% (584 subjects; 95% CI: 19.4-21.8)among the women, respectively.

Table 1 .
Distribution of participants according to demographic characteristics

Table 2 .
Prevalence of the symptoms of obstructive sleep apnea syndrome (OSAS) according to house characteristics EDS = excessive daytime sleepiness; *Attached to one or more houses; †Total number of people in the household divided by the number of rooms.

Table 3 .
Analytical statistics on the status of obstructive sleep apnea syndrome, from multiple linear regression test