Comorbidities and quality of life in Australian men and women with diagnosed and undiagnosed high-risk obstructive sleep apnea
2022; American Academy of Sleep Medicine; Volume: 18; Issue: 7 Linguagem: Inglês
10.5664/jcsm.9972
ISSN1550-9397
AutoresSowmya Krishnan, Ching Li Chai‐Coetzer, Nicole Grivell, Nicole Lovato, Sutapa Mukherjee, Andrew Vakulin, Robert Adams, Sarah Appleton,
Tópico(s)Sleep and related disorders
ResumoFree AccessScientific InvestigationsComorbidities and quality of life in Australian men and women with diagnosed and undiagnosed high-risk obstructive sleep apnea Sowmya Krishnan, MBBS, Ching Li Chai-Coetzer, MBBS, PhD, Nicole Grivell, BHealth Sci (Honours), Nicole Lovato, PhD, Sutapa Mukherjee, MBBS, PhD, Andrew Vakulin, PhD, Robert J. Adams, MBBS, MD, Sarah L. Appleton, PhD Sowmya Krishnan, MBBS Respiratory and Sleep Services, Southern Adelaide Local Health Network, Adelaide, South Australia, Australia , Ching Li Chai-Coetzer, MBBS, PhD Respiratory and Sleep Services, Southern Adelaide Local Health Network, Adelaide, South Australia, Australia Flinders Health and Medical Research Institute–Sleep Health (Adelaide Institute for Sleep Health), College of Medicine and Public Health, Flinders University, Bedford Park, South Australia, Australia , Nicole Grivell, BHealth Sci (Honours) Flinders Health and Medical Research Institute–Sleep Health (Adelaide Institute for Sleep Health), College of Medicine and Public Health, Flinders University, Bedford Park, South Australia, Australia , Nicole Lovato, PhD Flinders Health and Medical Research Institute–Sleep Health (Adelaide Institute for Sleep Health), College of Medicine and Public Health, Flinders University, Bedford Park, South Australia, Australia , Sutapa Mukherjee, MBBS, PhD Respiratory and Sleep Services, Southern Adelaide Local Health Network, Adelaide, South Australia, Australia Flinders Health and Medical Research Institute–Sleep Health (Adelaide Institute for Sleep Health), College of Medicine and Public Health, Flinders University, Bedford Park, South Australia, Australia , Andrew Vakulin, PhD Flinders Health and Medical Research Institute–Sleep Health (Adelaide Institute for Sleep Health), College of Medicine and Public Health, Flinders University, Bedford Park, South Australia, Australia , Robert J. Adams, MBBS, MD Respiratory and Sleep Services, Southern Adelaide Local Health Network, Adelaide, South Australia, Australia Flinders Health and Medical Research Institute–Sleep Health (Adelaide Institute for Sleep Health), College of Medicine and Public Health, Flinders University, Bedford Park, South Australia, Australia , Sarah L. Appleton, PhD Address correspondence to: Sarah L. Appleton, PhD, FHMRI Sleep Health (Adelaide Institute for Sleep Health), Flinders University, College of Medicine and Public Health, Mark Oliphant Building, L2, 5 Laffer Drive, Bedford Park 5042, South Australia, Australia; Tel: +61 8 74219755; Email: E-mail Address: [email protected] Flinders Health and Medical Research Institute–Sleep Health (Adelaide Institute for Sleep Health), College of Medicine and Public Health, Flinders University, Bedford Park, South Australia, Australia Published Online:July 1, 2022https://doi.org/10.5664/jcsm.9972Cited by:2SectionsAbstractEpubPDFSupplemental Material ShareShare onFacebookTwitterLinkedInRedditEmail ToolsAdd to favoritesDownload CitationsTrack Citations AboutABSTRACTStudy Objectives:In a population-based survey, we determined sex differences in health profiles and quality of life between individuals who have a confirmed diagnosis of obstructive sleep apnea (OSA) and those who are at high risk of OSA yet remain undiagnosed.Methods:An online survey of Australian adults ≥ 18 years (n = 3,818) identified participants with self-reported diagnosed OSA (n = 460) or high-risk, undiagnosed OSA (OSA50 score ≥ 5, n = 1,015). Ever-diagnosed comorbidities, sociodemographics, and quality of life (EQ-5D-5L, Functional Outcomes of Sleep Questionnaire-10) were assessed.Results:Women were more frequently represented in the high-OSA-risk group compared with those with diagnosed OSA (55.5%, n = 563, versus 43%, n = 198; P < .001). In sex-specific logistic regression analyses, diagnosed OSA was associated with increased likelihoods of ≥ 1 cardiovascular condition (odds ratio: 3.0; 95% confidence interval: 2.0–4.5), hypertension (1.9; 1.3–2.8), gout (1.8; 1.1–2.9), and chronic obstructive pulmonary disease (3.8; 2.1–6.9) in men. In women, an association with asthma (2.0; 1.3–3.0) was seen. Diabetes, arthritis, mental health conditions (ever-diagnosed), and all EQ-5D-5L dimensions were associated with an OSA diagnosis regardless of sex, except for EQ-5D-5L anxiety/depression, which was only associated with an OSA diagnosis in women. A diagnosis of OSA was associated with sleepiness-related impairment (lowest quartile of Functional Outcomes of Sleep Questionnaire-10) in men (1.6; 1.01–2.5) and women (2.2; 1.4–3.6).Conclusions:Sex-specific health conditions may drive diagnosis of OSA; however, clinical suspicion of OSA needs to be increased in men and women. The impaired quality of life and persistent sleepiness in participants with diagnosed OSA observed at a population level requires greater clinical attention.Citation:Krishnan S, Chai-Coetzer CL, Grivell N, et al. Comorbidities and quality of life in Australian men and women with diagnosed and undiagnosed high-risk obstructive sleep apnea. J Clin Sleep Med. 2022;18(7):1757–1767.BRIEF SUMMARYCurrent Knowledge/Study Rationale: Obstructive sleep apnea (OSA) is highly prevalent and associated with significant mental and physical health impacts, yet remains undiagnosed in a large proportion of adults. Identifying factors that facilitate a diagnosis may lead to early and targeted diagnoses; however, sex-specific health correlates of a diagnosis of OSA are unknown.Study Impact: Men and women at high risk of OSA on the OSA50 questionnaire without an OSA diagnosis experienced a high burden of morbidity. Compared with participants at high risk of OSA, diagnosed participants were significantly more likely to report major morbidity and sleepiness-related impacts despite most having received a diagnosis 1–2 years prior to the survey and recommendations for treatment. These findings suggest that increased diagnostic and therapeutic efforts for OSA are required.INTRODUCTIONObstructive sleep apnea (OSA) is a condition characterized by repetitive partial or complete closure of the airway during sleep, resulting in hypoxemia, sleep fragmentation, and intrathoracic pressure swings. OSA is associated with a significant comorbid physical and mental health burden,1,2 including an increased cardiovascular and cerebrovascular risk.3 Contemporary population studies in different countries have shown that the prevalence of moderate to severe OSA ranges from 10% to 50% in men, which is at least double that of women, which ranges from 3% to 23.4%.1,4–6 These figures have led the condition to be perceived largely as a disorder of men, although OSA of any severity can be present in up to 50% of adult women.1,7Despite the high prevalence of OSA, it may remain undiagnosed in individuals who are at high risk.8–12 Furthermore, a higher index of clinical suspicion for OSA may be required in women to increase equity in health outcomes.8 Women with OSA are more likely to report atypical symptoms, such as fatigue, insomnia, and mood disturbance compared to men and experience greater impairment in quality of life.13Previous studies have explored sex differences in health factors (including conditions and health service use) associated with diagnosed and undiagnosed OSA compared to those without the disorder.13–16 Other recent studies have determined health factors and comorbidities in relation to the level of risk of OSA (ascertained by screening questionnaires).11,12 This literature has been critical in highlighting the need for improved diagnosis and management of the condition for men and women. However, within the population of adults with OSA, our understanding of factors that facilitate a diagnosis is lacking. It is also unclear how a diagnosis of OSA improves health and quality of life for those with OSA. It is important to understand how clinical presentation can differ between the sexes, as this knowledge could lead to early and targeted diagnosis and treatment of this condition to reduce morbidity. To our knowledge, there have been no studies comparing adults diagnosed with OSA with those identified as undiagnosed, but at high risk of having OSA, with sex-specific sociodemographic, comorbidities, and quality of life comparisons.Using an online survey of Australian adults, we aimed to identify sex-specific demographic and health correlates of participants with a diagnosis of OSA compared to those identified with a screening questionnaire as being at high risk of OSA but who remain undiagnosed. We also examined if high-risk but undiagnosed OSA is associated with poorer health-related quality of life in men and women.METHODSSurvey sampling methodsA cross-sectional web-based survey was undertaken of adults (≥ 18 years) recruited from an online panel of over 500,000 Australian adults by Dynata, an international survey and market research company. The primary aim of this survey was to investigate the various diagnostic pathways experienced by the participants to receive an OSA diagnosis. The survey was conducted between September and November 2019 using a 3-stage randomization process to minimize the risk of bias to match a potential participant with a survey they are likely to be able to complete as follows.17 First, the panel was randomly sampled and participants were invited to take a survey, combined with others entering the Survey Sampling International Dynamix sampling platform after responding to online messaging. Second, a set of profiling questions was randomly selected (these were not affirmation questions), and, upon completion, participants were matched with a survey they were likely to be able to take using the third element of randomization of participants to a survey they may be eligible to undertake. Participants were invited to participate through invitations via email, telephone alerts, banners, and messaging to the Dynata membership panel community. The messages were varied, including invitations to provide an opinion and earn cash or prizes (value of this is approximately $1). To avoid self-selection bias, specific project details were not generally included in the invitation. Rather, participants were invited to "take a survey". The panel was sampled to generate a study sample that broadly matches Australian Bureau of Statistics estimates on sex, state, and regional representation. The survey was initially set up to sample from participants 18–65+ years of age; however, the sampling criteria was modified 3 days after the survey was launched to enrich the sample for high-risk and diagnosed OSA. This included imposing a limit for age of invited participants toward an older age (> 35 years) due to higher risk and prevalence of OSA in older adults. Prior to this modification, 100 participants in the age range of 18–34 years responded to the survey and were included in the analyses. Ethical approval for the conduct of the study was obtained from Social and Behavioral Research Ethics Committee at Flinders University, protocol number 8435.Diagnosed OSA and high-risk (undiagnosed) OSAThe independent variable, OSA, was identified with the question "Have you ever been diagnosed or investigated for obstructive sleep apnea (OSA)?" Those answering "No" were administered the OSA50 screening questionnaire (score range of 0–10), which was developed and validated in an Australian primary care population.18 A score of ≥ 5 identified patients at high risk of having moderate to severe OSA as follows: Waist circumference greater than 102 cm (or pants waist size 40) for men; or greater than 88 cm (or a pants waist size 16/XL) for women (yes = 3).In snoring participants, has your snoring ever bothered other people? (yes = 3)Has anyone ever told you that you stop breathing during your sleep? (yes = 2)Aged 50 years or older? (yes = 2)Diagnosed health conditions and chronic disease risk factorsData were collected regarding common health conditions ("Have you ever been told by a doctor that you have …?"), including heart disease; angina or coronary artery disease; previous heart attack; stroke or transient ischemic attack; atrial fibrillation/irregular heart beat (also categorized as at least 1 cardiovascular condition); depression; anxiety or panic disorder (also categorized as at least 1 mental health disorder); diabetes; high blood pressure; arthritis, gout; asthma; nasal obstruction/hay fever/rhinitis; other lung disease, eg, chronic obstructive pulmonary disease (COPD), heartburn or reflux disease, prostate disease/bladder problems.Health-related quality of lifeThe EuroQol (European Quality of Life) 5-dimension, 5-level scale (EQ-5D-5L) is a validated, extensively used instrument (in a wide range of clinical conditions and population samples) that assesses 5 dimensions of current function (ie, today), including mobility, personal care, usual activities, pain/discomfort, depression/anxiety.19 There were too few participants reporting severe and extreme levels of impairment for robust modeling, and therefore we combined the highest 3 levels of impairment as follows: 1) no problem, 2) slight problem, and 3) moderate, severe, extreme/unable. The EQ-5D-5L has a visual analog scale scored 0–100, in response to the question "We would like to know how good or bad your health is TODAY," where "100 means the best health you can imagine" and "0 means the worst health you can imagine". EuroQol approval was obtained for the use of the instrument.The Functional Outcomes of Sleep Questionnaire (FOSQ)-10 (FOSQ-10) is the shortened version of the FOSQ-30, an OSA-specific questionnaire designed to measure the impact of daytime sleepiness on activities of daily living.20 The 10-item instrument, with scores in the range of 5–20, contains questions relating to 5 dimensions of vigilance, activity level, general productivity, intimacy, and social outcomes. No formal cut-off values have been published for the FOSQ-10. The mean FOSQ-10 score was reported as 12.5 (standard deviation [SD] = 3.2) in moderate to severe OSA compared with 17.8 (SD = 3.1) in a normal control group.20 This value of 17.8 for a normal/healthy sleeper group is not dissimilar to data published for the FOSQ-30, where the mean for a group of adults free of polysomnography-assessed sleep disorders was 17.921 and equates to 1 SD above the mean score (14.1 [SD = 3.7]) of a clinical sample with moderate to severe OSA.21CovariatesStandard sociodemographic factors included sex, age, location of residence (metropolitan or regional/rural), main language spoken at home, gross household income, and highest educational attainment. Participant area-level economic and social status was considered using the Socio-Economic Indexes for Areas Index of Relative Socioeconomic Disadvantage matched to participants' postcodes based on the 2106 Australian Census of Population and Housing,22 taking into account 16 measures of disadvantage. Current smoking was defined as smoking every day or sometimes. Body mass index (BMI; kg/m2) was calculated from self-reported height and weight and classified according to World Health Organization criteria.Treatment of OSARecommended common treatments of OSA and their uptake were assessed with the questions: 1) What treatments have you been recommended for your sleep apnea? (response options included continuous positive airways pressure [CPAP] therapy/[mask], mouthguards, throat surgery) and 2) Did you start the therapy that was recommended to you?Statistical analysisData were analyzed using IBM SPSS version 25.0 (IBM Corporation, Armonk, NY). Sex-specific differences in distribution of outcomes, including ever-diagnosed health conditions and EQ-5D-5L dimensions across OSA diagnosis status, the independent variable, were determined using Pearson χ2 tests. Differences in distribution by sex within participants with high-risk (undiagnosed) OSA and diagnosed OSA were also determined using Pearson χ2 tests. The Mann-Whitney test determined differences in distribution of FOSQ-10 scores by OSA status.The associations of ever-diagnosed health conditions (dependent variables) with a diagnosis of OSA were determined with sex-specific multivariable binary logistic regression analysis. Models for all health conditions were adjusted for age, BMI, and smoking. Models for 1 or more ever-diagnosed cardiovascular disease (CVD) condition were additionally adjusted for hypertension, diabetes, and depression. Inclusion of an interaction term in models in all participants determined the moderation of associations of OSA status by sex.Ordinal logistic regression analyses were used to determine associations of 3-level EQ-5D-5L dimensions (dependent variables) with a diagnosis of OSA with adjustment for age and BMI and additional ever-diagnosed conditions that are chronic in nature as informed by the literature and our own univariate analyses as applicable, as follows—Mobility: at least 1 cardiovascular condition, lung disease/COPD, arthritis; Personal Care: depression and or anxiety/panic disorder, lung disease/COPD, arthritis; Usual Activities: depression and or anxiety/panic disorder, lung disease/COPD, arthritis, and at least 1 cardiovascular condition; Pain/Discomfort: cancer (not skin); Depression/Anxiety: asthma, lung disease (COPD), nasal obstruction/rhinitis. For Usual Activities in men, a multinomial logistic regression was conducted because the model failed the proportional odds assumptions in ordinal logistic regression.Multiple linear regression analyses determined the sex-specific association of EQ-5D-5L visual analog scale (VAS) score with a diagnosis of OSA, with adjustment for age, BMI, and household income.FOSQ-10 scores in our sample were highly skewed, suggesting a low impact of daytime sleepiness on activities of daily living. Multivariable logistic regression analysis determined sex-specific associations of an OSA diagnosis with FOSQ-10 scores in the lowest quartile of the distribution for men ≤ 16.33 and women ≤ 15.67. These values are equivalent to the mean of a clinical population undergoing surgery for moderate or severe OSA who failed conventional therapy.23 Logistic regression models were adjusted for age, BMI, and comorbidities that may disrupt sleep, including EQ-5D-5L dimensions of pain discomfort, depression/anxiety ever diagnosed airways disease, and bladder problems.RESULTSStudy sampleAs shown in Figure 1, of 3,818 adults who commenced the survey, the sampling process generated a sample of participants with diagnosed OSA (n = 460, 12%) or probable/high risk of OSA (n = 1,015) based on OSA symptom screening. Of diagnosed participants, 75.2% reported being diagnosed at least 1–2 years ago, and only 12.4% reported that their diagnosis occurred in the last 6 months.Figure 1: Study flow diagram.FOSQ = Functional Outcomes of Sleep Questionnaire, OSA = obstructive sleep apnea.Download FigureSample characteristicsThe sociodemographic and biomedical characteristics of the sample overall and in relation to a diagnosis of OSA are presented in Table S1 in the supplemental material. Compared to Australian Bureau of Statistics and National Health Survey estimates,24 survey participants were older (29% ≥ 65 years versus 16%) and demonstrated higher rates of overweight or obesity (81% versus 63%, consistent with a survey designed to capture participants with OSA) and showed higher levels of tertiary qualifications (29% versus 22%) and Australian place of birth (78% versus 67%), but otherwise were similar to Australian 2016 Census estimates in terms of sex, metropolitan or rural/regional residence, and extremes of annual household income (< $20,000 or > $150,000 per year).25Sociodemographic characteristics associated with a diagnosis of OSAOf those at high risk of OSA, 55.5% (n = 563) were women, while women comprised 43% of the diagnosed OSA group (n = 198, P < .001) (Table S1). Table 1 shows sociodemographic characteristics of those with diagnosed OSA compared with high-risk (undiagnosed) OSA in relation to sex. Compared to high-risk OSA, those with a diagnosis of OSA were more likely to be in the younger age group regardless of sex (P < .001) and women with a diagnosis were more likely to report higher levels of education (P = .012) and household income (P = .034). No significant associations with area-level socioeconomic disadvantage, regional/metropolitan location of residence, or speaking English at home were seen.Table 1 Sex-specific sociodemographic and biomedical characteristics of participants according to OSA status.CharacteristicMen, % (n)Women, % (n)High-Risk* OSA (n = 452)Diagnosed OSA (n = 262)χ2PHigh-Risk* OSA (n = 563)Diagnosed OSA (n = 198)χ2PAge, years 18–346.4 (29)16.8 (44)< .0012.8 (16)16.2 (32)< .001 35–5429.0 (131)35.1 (92)41.0 (231)41.4 (82) 55–6426.5 (120)17.6 (46)31.1 (175)23.7 (47) 65+38.1 (172)30.5 (80)25.0 (141)18.7 (37) < 5535.4 (160)51.9 (136)< .00143.9 (247)57.6 (114)< .001 ≥ 5564.6 (292)48.1 (126)56.1 (316)42.4 (84)Area of residence metropolitan65.9 (298)69.8 (183).28260.4 (340)67.2 (133).091Annual gross household income.919.034 < $20,0003.3 (15)3.4 (9)6.2 (35)8.1 (16) $20,001–$60,00039.4 (178)37.4 (98)41.4 (233)34.8 (69) $60,001–$100,00025.9 (117)28.2 (74)21.8 (123)24.2 (48) > $100,00023.2 (105)24.0 (63)19.4 (109)26.8 (53) Don't know/refused8.2 (37)6.9 (18)11.2 (63)6.1 (12)Highest education.075.012 High school or less32.1 (145)31.7 (83)38.5 (217)28.8 (57) Certificate, diploma, trade qualification37.6 (170)30.5 (80)37.5 (211)38.4 (76)Bachelor's degree or higher30.3 (137)37.8 (99)23.4 (132)32.8 (65)SEIFA quintiles.313.336 1 - most disadvantaged15.5 (70)15.3 (40)18.1 (102)14.2 (28) 222.8 (103)20.2 (53)21.4 (120)21.3 (42) 318.6 (84)19.8 (52)20.5 (115)21.8 (43) 417.5 (79)23.3 (61)23.3 (131)20.3 (40) 5 - least disadvantaged25.7 (116)21.4 (56)16.7 (94)22.3 (44)Language spoken at home, English93.4 (422)89.7 (235).08194.1 (530)91.4 (181).183BMI, kg/m2 (mean, SD)30.4 (6.9)31.1 (7.9).26832.1 (8.1)30.4 (8.9).019Current smoker25.0 (113)34.7 (91).00625.0 (141)26.8 (53).632*OSA-50 score of ≥ 5 identified patients at high risk of having moderate to severe OSA. Statistical difference by OSA status determined by Student's t test. BMI = body mass index, OSA = obstructive sleep apnea, SD = standard deviation, SEIFA = Socio-Economic Indexes for Areas.Ever-diagnosed chronic conditions associated with a diagnosis of OSAIn men, self-reported, ever-diagnosed cardiovascular conditions, high blood pressure, COPD, and gout were significantly more prevalent in those with a diagnosis of OSA compared to participants at high risk of OSA but not diagnosed (Table 2). In women, ever-diagnosed asthma and a healthy body weight (based on BMI) were more prevalent in those with a diagnosis of OSA (Table 2). Ever-diagnosed diabetes, mental health conditions, arthritis, and reflux were significantly more prevalent in both male and female participants with a diagnosis of OSA compared to those at high risk of OSA but not diagnosed.Table 2 Prevalence of ever-diagnosed chronic conditions in relation to OSA diagnosis and sex and adjusted OR and 95% CI for chronic conditions associated with a diagnosis of OSA.Men, % (n)Women, % (n)P†High-Risk# OSADiagnosed OSAχ2PHigh-Risk# OSADiagnosed OSAχ2P≥ 1 Cardiovascular condition17.9 (81)40.1 (105)< .00117.1 (96)20.7‡ (41).25 OR (95% CI)1.03.0 (2.0–4.5)1.00.9 (0.6–1.6).001Cardiometabolic conditions High blood pressure42.9 (194)50.8 (133).04343.0 (242)41.9 (83).794 OR (95% CI)1.01.9 (1.3–2.8)1.01.4 (0.9–2.1).532 Diabetes14.6 (66)28.2 (74)< .00115.1 (85)21.7 (43).032 OR (95% CI)1.02.6 (1.7–3.9)1.02.2 (1.3–3.5).520BMI, kg/m2 Overweight or obese ≥ 25.084.3 (350)79.8 (182).14783.4 (398)68.5 (111)< .001 OR (95% CI)1.00.96 (0.6–1.5)1.00.5 (0.3–0.8).063Mental health conditions Depression27.4 (124)35.5 (93).02436.1* (203)50.5‡ (100)< .001 Anxiety/panic disorder15.0 (68)22.1 (58).01727.9* (157)39.9‡ (79)< .001 ≥ 1 condition30.3 (137)40.2 (110).00243.0* (242)59.1‡ (117)< .001 OR (95% CI)1.01.6 (1.1–2.3)1.02.5 (1.7–3.7).462Airways diseases Asthma15.3 (69)18.7 (49).23318.7 (105)32.8‡ (65)< .001 OR (95% CI)1.01.2 (0.8–1.8)1.02.0 (1.3–3.0).167 Lung disease, eg, COPD, emphysema4.9 (22)15.6 (41)< .0015.0 (28)8.1‡ (16).107 OR (95% CI)1.03.8 (2.1–6.9)1.01.7 (0.8–3.8).091 Nasal obstruction/hay fever/rhinitis18.6 (84)23.7 (62).10525.8* (145)31.8 (63).100 OR (95% CI)1.01.4 (0.9–2.0)1.01.3 (0.8–2.0).816Pain disorders Arthritis27.4 (124)37.8 (99).00438.0* (214)46.5 (92).037 OR (95% CI)1.02.0 (1.4–2.9)1.01.9 (1.3–2.9).709 Gout10.8 (49)17.2 (45).0165.2* (29)6.1‡ (12).626 OR (95% CI)1.01.8 (1.1–2.9)1.01.0 (0.5-2.4).363 Heartburn/reflux27.2 (123)33.2 (87).09032.7 (184)38.9 (77).114 OR (95% CI)1.01.5 (1.02–2.1)1.01.6 (1.05–2.3).913 Bladder problems14.4 (65)16.8 (44).3874.8* (27)13.1 (26)< .01 OR (95%CI)1.01.4 (0.9–2.0)1.01.2 (0.8–1.8).841All binary logistic regression models were adjusted for age, smoking, and BMI, except for reflux and bladder problems (age and BMI only); at least 1 cardiovascular condition additionally adjusted for hypertension, diabetes, and at least 1 mental health condition; overweight or obesity (age and smoking only). †P for the interaction of OSA status with sex in separate models for chronic conditions. Each comorbidity was analyzed as the dependent variable in separate models. #OSA-50 score of ≥ 5 identified patients at high risk of having moderate to severe OSA. Cardiovascular conditions included heart disease, angina or coronary artery disease, previous heart attack, stroke or transient ischemic attack (TIA), atrial fibrillation/irregular heartbeat. *P < .05 for difference in distribution of chronic condition between men and women within undiagnosed OSA. ‡P < .05 for difference in distribution of chronic condition between men and women within diagnosed OSA. BMI = body mass index, CI, confidence interval, COPD = chronic obstructive pulmonary disease, OR = odds ratio, OSA = obstructive sleep apnea.These associations persisted after adjusting for age, BMI, and smoking (at least 1 CVD condition was additionally adjusted for ever-diagnosed hypertension, depression, and diabetes). In logistic regression models in the sample not stratified by sex, interaction terms (OSA diagnosis × sex) indicated that associations of an OSA diagnosis with 1 or more CVD conditions were significantly stronger in men and that sex is likely to influence relationships with COPD (in males) and healthy weight (in females). The prevalence of individual CVD conditions in relation to OSA diagnosis and sex is shown in Table S2 in the supplemental material.In univariate analyses in participants with diagnosed OSA, women were significantly less likely to report ever being diagnosed with CVD conditions, COPD, or gout, but more likely to report a diagnosis of asthma, depression, and anxiety and be of healthy weight (BMI < 25 kg/m2) than men.Similarly, in participants with high-risk (undiagnosed) OSA, women reported significantly less gout and more ever-diagnosed mental health conditions, arthritis, and nasal symptoms than their male counterparts.EQ-5D-5L current health status dimensions associated with a diagnosis of OSAIn both men and women, a diagnosis of OSA was significantly associated with a report of current moderate to extreme problems/unable in all 5 EQ-5D-5L dimensions compared to those at high risk of OSA (Table 3). This finding persisted in covariate-adjusted models with 1 exception where the association of diagnosed OSA with currently feeling anxious or depressed was attenuated in men.Table 3 EQ-5D health-related quality of life in relation to diagnosis of OSA and sex, and adjusted OR and 95% CI for increasing functional impairment associated with a diagnosis of OSA.Men, % (n)Women, % (n)High-Risk# OSADiagnosed OSAPHigh-Risk# OSADiagnosed OSAPMobility< .001< .001 No problem65.7 (297)43.5 (114)59.3 (334)44.9 (89) Slight problem21.2 (96)30.5 (80)24.5 (138)25.3 (50) Moderate, severe difficulty, unable13.1 (59)26.0 (68)16.2 (91)29.8 (59) Model 1 OR (95% CI)1.02.4 (1.7–3.3)1.01.9 (1.3–2.8) Model 2 OR (95% CI)1.01.9 (1.3–2.6)1.01.7 (1.1–2.4)Personal care< .001< .001 No problem88.9 (402)70.2 (184)89.3 (503)73.2 (145) Slight problem8.0 (36)18.7 (49)7.6 (43)17.7 (35) Moderate, severe difficulty, unable3.1 (14)11.1 (29)3.0 (17)9.1 (18) Model 1 OR (95% CI)1.02.6 (1.7–4.1)1.02.3 (1.4–3.8) Model 2 OR (95% CI)1.02.1 (1.3–3.3)1.01.8 (1.1–3.1)Usual activities< .001< .001 No problem70.6 (319)52.7 (138)65.5 (369)49.5 (98) Slight problem22.1 (100)25.6 (67)23.1 (130)23.7 (47) Moderate, severe difficulty, unable7.3 (33)21.8 (57)11.4 (64)26.8 (53) Model 1 OR (95% CI)1.03.3 (2.0–5.6)†1.02.4 (1.7–3.5) Model 2 OR (95% CI)1.02.0 (1.1–3.6)†1.02.0 (1.4–3.0)Pain/discomfort.026*‡< .001 No problem33.0 (149)26.3 (69)23.6 (133)23.2 (46) Slight problem41.4 (187)38.9 (102)45.8 (258)30.3 (60) Moderate-extreme problem25.7 (116)34.7 (91)30.6 (172)46.5 (92) Model 1 OR (95% CI)1.01.4 (1.03–1.9)1.01.6 (1.1–2.2) Model 2 OR (95% CI)1.01.4 (1.04–2.0)1.01.5 (1.1–2.2)Anxiety/depression.005*‡< .001 No problem56.6 (256)46.6 (122)46.4 (261)27.3 (54) Slight problem24.3 (110)24.4 (64)30.2 (170)38.4 (76) Moderate-extreme problem19.0 (86)29.0 (76)23.4 (132)34.3 (68) Model 1 OR (95% CI)1.01.2 (0.9–1.7)1.01.9 (1.3–2.6) Model 2 OR (95% CI)1.01.1 (0.8–1.5)1.01.8 (1.2–2.5)FOSQ-10 quartiles< .001< .001 1 - most impaired17.1 (77)32.7 (85)19.4 (109)42.4 (84) 229.0 (130)28.8 (75)25.1 (141)22.7 (45) 323.6 (106)17.3 (45)25.8 (145)17.7 (35) 4 - least impaired30.3 (136)21.2 (55)29.7 (167)17.2 (34) Model 1 OR** (95% CI)1.01.8 (1.2–2.6)1.02.7 (1.8–4.1) Model 2 OR** (95% CI)1.01.6 (1.01–2.5)1.02.2 (1.4–3.6)Ordinal logistic regression except for usual activities in men that did not meet assumptions for proportional odds (test of parallel lines P < .05). Model 1 adjusted for age and BMI. Model 2 adjusted for age and BMI and ever-diagnosed conditions as follows: Mobility additionally adjusted for at least 1 cardiov
Referência(s)