Sleep Breath DOI 10.1007/s11325-013-0911-7
ORIGINAL ARTICLE
Factors influencing the response of psychological symptoms to continuous positive airway pressure therapy Alicia Carissimi & Denis Martinez & Lenise J. Kim & Cintia Z. Fiori
Received: 16 June 2013 / Revised: 3 October 2013 / Accepted: 4 November 2013 # Springer-Verlag Berlin Heidelberg 2013
Abstract Purpose To examine whether in obstructive sleep apnea (OSA) patients the degree of sleepiness, slow-wave sleep (SWS) loss, and hypoxia influence the response of psychological symptoms to continuous positive airway pressure (CPAP) therapy. Methods A prospectively planned cohort was assessed. Participants underwent full overnight polysomnography. All answered the Symptom Checklist-90-Revised (SCL-90) and the Epworth sleepiness scale (ESS). Only cases with an apneahypopnea index ≤60 events/h were 24 invited to participate. In an interview by telephone, after a follow-up period between 2 months and 1 year, patients informed whether they were in treatment with CPAP or remained voluntarily untreated. Subjects who accepted to participate answered for the second time the SCL-90 and the ESS. The baseline variables of interest were: (a) score of the ESS, (b) duration of SWS, and (c) time with arterial oxygen saturation below 90 %. The outcomes were the change in SCL-90 scores in all dimensions and indices of the questionnaire. Results A number of 73 patients, mostly men, were included. In uncontrolled analyses, CPAP-treated patients showed significant improvement at follow-up in 10 of the 13 SCL-90 A. Carissimi : D. Martinez Graduate Program in Medical Sciences, Universidade Federal do Rio Grande do Sul, Porto Alegre, Rio Grande do Sul, Brazil A. Carissimi (*) : D. Martinez : L. J. Kim : C. Z. Fiori Cardiology Division, Hospital de Clínicas de Porto Alegre (HCPA), Ramiro Barcelos, 2350, CEP 90035-003 Porto Alegre, Rio Grande do Sul, Brazil e-mail: [email protected] D. Martinez : C. Z. Fiori Graduate Program in Cardiology and Cardiovascular Sciences, Universidade Federal do Rio Grande do Sul, Rio Grande do Sul, Brazil
scores. Comparing with the control group, only six scores were improved. Baseline sleepiness was the best predictor of SCL-90 improvement after CPAP treatment in the univariate analyses, and the only significant predictor of improvement in a multivariate regression model. Conclusions Sleepiness may supersede other factors that influence psychological improvement in CPAP-treated patients with severe OSA. Keywords Obstructive sleep apnea . Continuous positive airway pressure . Mental disorders . Hypoxia . Sleepiness
Introduction One diagnosis of mental disorder has been identified in one quarter of all patients with sleep disorders, and two diagnoses in 10 % of these patients [1]. Obstructive sleep apnea (OSA) affects up to 32.8 % of the population [2]. The pathophysiological changes caused by OSA, mainly sleep fragmentation and intermittent hypoxia [3], lead to impairment of memory, learning, and communication, and to psychosocial alterations [4–7]. The Symptom Checklist-90-Revised (SCL-90) is a validated instrument for the screening of psychological symptoms [8]. More severe OSA has been associated with higher SCL90 scores of somatization, depression, anxiety, obsession– compulsion, and hostility [9]. Continuous positive airway pressure (CPAP) is an effective treatment for OSA. Sleep fragmentation and intermittent hypoxia are reversed, as are several other disease markers [10], including sleepiness and fatigue [11]. Improvement in mood or psychological symptoms, however, varies widely. Improvement of mood has been reported in uncontrolled studies of CPAP treatment [12–15], in which regression
Sleep Breath
toward the mean cannot be ruled out. In controlled studies, conversely, no improvement has been observed [16–21], except in a few analyses [22–24]. Studies of mechanisms that may influence the psychological symptoms of OSA are emerging. Bardwell and colleagues [25] reported factors such as deep sleep, rapid eye movement (REM) sleep, and hypoxemia as possible influences on psychological symptoms in patients with OSA. Bardwell et al. [26] also showed significant improvement in scores of an abridged version of the SCL-90 after oxygen administration, indicating a role of brain hypoxia in psychological symptoms. They concluded, “hypoxemia may play a stronger role than sleep disruption” on psychological distress. Sleepiness is absent in many OSA patients, even among the most severe cases [27]. The fact that many patients with extremely severe OSA remain without sleepiness and conserve normal amounts of deep sleep is so intriguing that warrants further exploration. Since the evidence is scarce, one can theorize that patients with higher sleepiness scores and absence of slow-wave sleep (SWS) may have higher levels of psychological distress than the ones that remain asymptomatic. The degree of hypoxia has been shown an influential factor on mental symptoms; therefore, research needs to control for this aspect. Considering that depression/anxiety/emotional problems are among the ten most common conditions causing people to need help with daily activities or personal care [28] and that sleep apnea may affect up to one third of the population [2], it is useful to search for mechanisms linking both conditions. To test the hypothesis that sleepiness and lack of SWS participate in mental impairment, the present study recruited patients with severe OSA analyzing the effect of CPAP therapy as intervention and the change in each of the SCL-90 scores from baseline to follow-up as outcomes. The variables of interest were: (a) sleepiness assessed by the Epworth sleepiness scale (ESS), (b) total duration of SWS, and (c) time with arterial oxygen saturation (SaO2) below 90 %.
Methods Subject selection A prospectively planned cohort study at a university-affiliated sleep clinic was carried out. Patients of both genders who underwent polysomnography in the year 2010 and were diagnosed with very severe OSA (AHI≥60 events/h) were selected, aiming to maximize the proportion of symptomatic cases. Inclusion criteria were age between 18 and 75 years, and the SCL-90 questionnaire completed on the night of baseline polysomnography. Patients were approached by telephone between January and May 2011 with questions on their treatment status and willingness to participate. Individuals who
accepted to take part in the study were invited to log onto a webpage to complete the questionnaires online. Printed questionnaires were mailed to nine patients who so preferred, mostly due to lack of Internet access or computer proficiency. Patients who were currently on CPAP therapy for at least 2 months constituted the study group. Patients who reported spontaneously refraining from any therapeutic intervention, despite medical recommendations, were included as controls. The exclusion criteria were: any upper airway intervention and weight loss >2 kg after polysomnography; inability to complete the SCL-90 for any reason; any debilitating condition such as heart failure, rheumatoid arthritis, AIDS; neurologic conditions such as stroke, Alzheimer's, and Parkinson's disease, or established psychiatric/psychological diagnoses; and use of substances acting on psychological symptoms or the central and peripheral nervous system (sedatives, antidepressants, appetite suppressants, amphetamines, psychotropic drugs, muscle relaxants, and thyroxine). The local Institutional Ethics Committee approved the study protocol. A brief informed consent form was read at the telephone interview and verbal agreement on anonymous use of data was obtained from each subject. The online questionnaires were made available only after the patient had read the complete informed consent terms and clicked on the appropriate button confirming consent. Polysomnography Patients underwent baseline polysomnography during one night of sleep using standard equipment, as described elsewhere [29, 30]. Briefly, patients arrived at the sleep laboratory around 9 PM. The recording lasted from 11 PM to 7 AM. Sleep stages were identified by electroencephalogram monitoring (F4-M1, O2-M1, and C4-M1), electrooculogram (left and right eye M1), and electromyogram (submental electrodes). The recording included one-lead electrocardiography (D1 or modified V4). Inductive plethysmography of the chest and abdominal wall, nasal cannula/pressure transducer, snoring, and oxygen saturation by pulse oximetry were evaluated. Polysomnographic variables included the AHI, time with SaO2 below 90 %, total time asleep, time in SWS, time in nonREM stage 3 (N3), sleep efficiency, and sleep latency. Apnea and hypopnea events were scored using American Academy of Sleep Medicine (AASM) criteria [31]. The AHI was calculated as the sum of all apnea and hypopnea events, excluding respiratory effort-related arousals, divided by the number of hours of sleep. Questionnaires The SCL-90 was self-administered. The questionnaire is a measure of current, point-in-time, psychological symptom status. It consists of 90 items and each item is rated on a
Sleep Breath
five-point scale of distress (0–4) ranging from “Not at all” to “Extremely”. The SCL-90 is scored and interpreted in terms of dimensions and global indices of distress [32]. All nine psychological symptom dimensions were analyzed: somatization, obsessive–compulsive, interpersonal sensitivity, depression, anxiety, hostility, phobic anxiety, paranoid ideation, and psychoticism [8, 33]. Seven questions that have clinical significance but are left out of the previous dimensions constitute a score of “Additional Items”. The global severity index (GSI), the positive symptom distress index, and the positive symptom total were calculated, totaling 13 SCL-90 scores. Validation procedures for the creation of a Brazilian version were conducted according with the rules of the American Psychological Association [34]. Under the supervision of the scale owners, a committee established the Brazilian Portuguese version of the SCL-90 by means of translation, backtranslation, and comparison. After a pilot test, the committee and the license owners judged the version adequate for clinical use. At baseline, when each subject arrived to undergo overnight polysomnography examination, the SCL-90 and the ESS were completed in full. After at least 2 months' followup, the subjects repeated the questionnaires online. Statistical analysis The results were expressed as mean, standard deviation, and standard error of the mean. Student's t test for independent samples was used for comparisons between untreated and CPAP-treated patients; Student's t test for paired samples was used to compare results at baseline and at follow-up. Analysis of variance was performed to determine the differences and interactions between treated and untreated groups by time and among durations of CPAP use (i.e., 2–3 months, 3–4 months, 5–6 months, >6 months) in SCL-90 scores. Parsimonious linear regression models were constructed in an attempt to predict change in SCL-90 scores, using the variables of interest (ESS, SaO2, and SWS) as regressors. Because the SCL-90 had not been validated in subjects with sleep apnea, we performed factor and reliability analyses to test its psychometric properties. Factor analysis used principal components extraction with varimax rotation and Kaiser normalization. SPSS v18 was used for all statistical analyses (SPSS, Chicago, IL). A two-tailed probability of alpha error lower than 0.05 was regarded as statistically significant.
Results As shown in Fig. 1, 189 eligible subjects that, at the telephone contact, had accepted to participate did not reply to the e-mail granting access to the consent form and to the questionnaires.
Patients in database with AHI>60/h (n=470)
Not included in the study (n=208) Not reached by telephone - 95 Upper airway intervention - 49 Weight loss >2 Kg - 34 Psychotropic medication - 15 Debilitating illness - 10 Death - 5
Elegible patients (n=262)
Declined to participate (n=189) Included in this study (n=73)
Untreated (n=27)
CPAP-treated (n=46)
Fig. 1 Flowchart of the study protocol
The reasons for this subsequent decline were not recorded. Of the 73 patients included, 46 reported having undergone treatment with CPAP for at least 2 months. Patients who rejected any treatment for OSA despite medical advice (27) constituted the control group (Fig. 1). Anthropometric, clinical, and polysomnographic data on CPAP-treated and untreated patients are displayed in Table 1. Baseline data were similar between untreated and CPAP-treated patients. Comparisons of the SCL-90 scores between CPAP-treated and untreated patients at baseline and at follow-up are shown in Table 2. CPAP-treated and untreated patients had similar SCL-90 scores at baseline. CPAP-treated patients showed reductions in ten SCL-90 scores at follow-up. On comparison with controls, CPAP-treated patients exhibited significant reductions between baseline and follow-up scores in six of the dimensions of the SCL-90 (Fig. 2). Among the CPAP users, 17 % reported using the device for less than 4 h/night and less than 4 nights/week. These patients with poor adherence to CPAP therapy were not significantly different from the compliant ones, regarding all the variables described in Tables 1 and 2. Hence, they were maintained in the analysis. Differences between baseline and follow-up in each of the SCL-90 scores by CPAP adherence, in terms of months of use, days of CPAP use per week, and hours of CPAP use per day were nonsignificant. The differences in body mass index from baseline to follow-up (time) and between CPAP groups were also nonsignificant and showed no interaction (P =0.5). Factor and reliability analyses showed that the original psychometric properties of the SCL-90 [8] were conserved in this sample. The Cronbach's alpha for the nine dimensions was 0.944 at baseline and 0.949 at follow-up. Only one component was identified on factor analysis, both at baseline and at follow-up.
Sleep Breath Table 1 Polysomnographic and other characteristics of untreated and CPAP-treated patients with severe obstructive sleep apnea
Values expressed as means and standard deviations BMI body mass index in kg/m2 (weight/height2 ), CPAP continuous positive airway pressure, ESS Epworth sleepiness scale, AHI apnea–hypopnea index, SaO 2 arterial oxygen saturation *P 30 kg/m2, n (%) BMI at follow-up 30 kg/m2, n (%) CPAP time, n (%) 2–3 months 3–4 months 4–6 months >6 months Days on CPAP, n (%) 1–4 days/week 4–6 days/week >6 days/week Hours on CPAP, n (%) 1–4 h/day 4–6 h/day
4 (15.4) 22 (80.8) 33.4±5.0 1 (3.7) 6 (22.2) 20 (74.1)
8 (17.4) 37 (80.4) 33.6±4.9 1 (2.2) 13 (28.3) 32 (69.6)
– – – –
2 (4) 5 (11) 7 (15) 32 (70)
– – –
4 (9) 7 (15) 35 (76)
– –
8 (17) 16 (35)
6–8 h/day ESS, baseline ESS, follow-up Polysomnographic Parameters AHI (events/h) SaO2 nadir (%)
– 13.5±5.1 14.6±5.6
22 (48) 15.5±4.8 9.2±4.9
0.09
ORIGINAL ARTICLE
Factors influencing the response of psychological symptoms to continuous positive airway pressure therapy Alicia Carissimi & Denis Martinez & Lenise J. Kim & Cintia Z. Fiori
Received: 16 June 2013 / Revised: 3 October 2013 / Accepted: 4 November 2013 # Springer-Verlag Berlin Heidelberg 2013
Abstract Purpose To examine whether in obstructive sleep apnea (OSA) patients the degree of sleepiness, slow-wave sleep (SWS) loss, and hypoxia influence the response of psychological symptoms to continuous positive airway pressure (CPAP) therapy. Methods A prospectively planned cohort was assessed. Participants underwent full overnight polysomnography. All answered the Symptom Checklist-90-Revised (SCL-90) and the Epworth sleepiness scale (ESS). Only cases with an apneahypopnea index ≤60 events/h were 24 invited to participate. In an interview by telephone, after a follow-up period between 2 months and 1 year, patients informed whether they were in treatment with CPAP or remained voluntarily untreated. Subjects who accepted to participate answered for the second time the SCL-90 and the ESS. The baseline variables of interest were: (a) score of the ESS, (b) duration of SWS, and (c) time with arterial oxygen saturation below 90 %. The outcomes were the change in SCL-90 scores in all dimensions and indices of the questionnaire. Results A number of 73 patients, mostly men, were included. In uncontrolled analyses, CPAP-treated patients showed significant improvement at follow-up in 10 of the 13 SCL-90 A. Carissimi : D. Martinez Graduate Program in Medical Sciences, Universidade Federal do Rio Grande do Sul, Porto Alegre, Rio Grande do Sul, Brazil A. Carissimi (*) : D. Martinez : L. J. Kim : C. Z. Fiori Cardiology Division, Hospital de Clínicas de Porto Alegre (HCPA), Ramiro Barcelos, 2350, CEP 90035-003 Porto Alegre, Rio Grande do Sul, Brazil e-mail: [email protected] D. Martinez : C. Z. Fiori Graduate Program in Cardiology and Cardiovascular Sciences, Universidade Federal do Rio Grande do Sul, Rio Grande do Sul, Brazil
scores. Comparing with the control group, only six scores were improved. Baseline sleepiness was the best predictor of SCL-90 improvement after CPAP treatment in the univariate analyses, and the only significant predictor of improvement in a multivariate regression model. Conclusions Sleepiness may supersede other factors that influence psychological improvement in CPAP-treated patients with severe OSA. Keywords Obstructive sleep apnea . Continuous positive airway pressure . Mental disorders . Hypoxia . Sleepiness
Introduction One diagnosis of mental disorder has been identified in one quarter of all patients with sleep disorders, and two diagnoses in 10 % of these patients [1]. Obstructive sleep apnea (OSA) affects up to 32.8 % of the population [2]. The pathophysiological changes caused by OSA, mainly sleep fragmentation and intermittent hypoxia [3], lead to impairment of memory, learning, and communication, and to psychosocial alterations [4–7]. The Symptom Checklist-90-Revised (SCL-90) is a validated instrument for the screening of psychological symptoms [8]. More severe OSA has been associated with higher SCL90 scores of somatization, depression, anxiety, obsession– compulsion, and hostility [9]. Continuous positive airway pressure (CPAP) is an effective treatment for OSA. Sleep fragmentation and intermittent hypoxia are reversed, as are several other disease markers [10], including sleepiness and fatigue [11]. Improvement in mood or psychological symptoms, however, varies widely. Improvement of mood has been reported in uncontrolled studies of CPAP treatment [12–15], in which regression
Sleep Breath
toward the mean cannot be ruled out. In controlled studies, conversely, no improvement has been observed [16–21], except in a few analyses [22–24]. Studies of mechanisms that may influence the psychological symptoms of OSA are emerging. Bardwell and colleagues [25] reported factors such as deep sleep, rapid eye movement (REM) sleep, and hypoxemia as possible influences on psychological symptoms in patients with OSA. Bardwell et al. [26] also showed significant improvement in scores of an abridged version of the SCL-90 after oxygen administration, indicating a role of brain hypoxia in psychological symptoms. They concluded, “hypoxemia may play a stronger role than sleep disruption” on psychological distress. Sleepiness is absent in many OSA patients, even among the most severe cases [27]. The fact that many patients with extremely severe OSA remain without sleepiness and conserve normal amounts of deep sleep is so intriguing that warrants further exploration. Since the evidence is scarce, one can theorize that patients with higher sleepiness scores and absence of slow-wave sleep (SWS) may have higher levels of psychological distress than the ones that remain asymptomatic. The degree of hypoxia has been shown an influential factor on mental symptoms; therefore, research needs to control for this aspect. Considering that depression/anxiety/emotional problems are among the ten most common conditions causing people to need help with daily activities or personal care [28] and that sleep apnea may affect up to one third of the population [2], it is useful to search for mechanisms linking both conditions. To test the hypothesis that sleepiness and lack of SWS participate in mental impairment, the present study recruited patients with severe OSA analyzing the effect of CPAP therapy as intervention and the change in each of the SCL-90 scores from baseline to follow-up as outcomes. The variables of interest were: (a) sleepiness assessed by the Epworth sleepiness scale (ESS), (b) total duration of SWS, and (c) time with arterial oxygen saturation (SaO2) below 90 %.
Methods Subject selection A prospectively planned cohort study at a university-affiliated sleep clinic was carried out. Patients of both genders who underwent polysomnography in the year 2010 and were diagnosed with very severe OSA (AHI≥60 events/h) were selected, aiming to maximize the proportion of symptomatic cases. Inclusion criteria were age between 18 and 75 years, and the SCL-90 questionnaire completed on the night of baseline polysomnography. Patients were approached by telephone between January and May 2011 with questions on their treatment status and willingness to participate. Individuals who
accepted to take part in the study were invited to log onto a webpage to complete the questionnaires online. Printed questionnaires were mailed to nine patients who so preferred, mostly due to lack of Internet access or computer proficiency. Patients who were currently on CPAP therapy for at least 2 months constituted the study group. Patients who reported spontaneously refraining from any therapeutic intervention, despite medical recommendations, were included as controls. The exclusion criteria were: any upper airway intervention and weight loss >2 kg after polysomnography; inability to complete the SCL-90 for any reason; any debilitating condition such as heart failure, rheumatoid arthritis, AIDS; neurologic conditions such as stroke, Alzheimer's, and Parkinson's disease, or established psychiatric/psychological diagnoses; and use of substances acting on psychological symptoms or the central and peripheral nervous system (sedatives, antidepressants, appetite suppressants, amphetamines, psychotropic drugs, muscle relaxants, and thyroxine). The local Institutional Ethics Committee approved the study protocol. A brief informed consent form was read at the telephone interview and verbal agreement on anonymous use of data was obtained from each subject. The online questionnaires were made available only after the patient had read the complete informed consent terms and clicked on the appropriate button confirming consent. Polysomnography Patients underwent baseline polysomnography during one night of sleep using standard equipment, as described elsewhere [29, 30]. Briefly, patients arrived at the sleep laboratory around 9 PM. The recording lasted from 11 PM to 7 AM. Sleep stages were identified by electroencephalogram monitoring (F4-M1, O2-M1, and C4-M1), electrooculogram (left and right eye M1), and electromyogram (submental electrodes). The recording included one-lead electrocardiography (D1 or modified V4). Inductive plethysmography of the chest and abdominal wall, nasal cannula/pressure transducer, snoring, and oxygen saturation by pulse oximetry were evaluated. Polysomnographic variables included the AHI, time with SaO2 below 90 %, total time asleep, time in SWS, time in nonREM stage 3 (N3), sleep efficiency, and sleep latency. Apnea and hypopnea events were scored using American Academy of Sleep Medicine (AASM) criteria [31]. The AHI was calculated as the sum of all apnea and hypopnea events, excluding respiratory effort-related arousals, divided by the number of hours of sleep. Questionnaires The SCL-90 was self-administered. The questionnaire is a measure of current, point-in-time, psychological symptom status. It consists of 90 items and each item is rated on a
Sleep Breath
five-point scale of distress (0–4) ranging from “Not at all” to “Extremely”. The SCL-90 is scored and interpreted in terms of dimensions and global indices of distress [32]. All nine psychological symptom dimensions were analyzed: somatization, obsessive–compulsive, interpersonal sensitivity, depression, anxiety, hostility, phobic anxiety, paranoid ideation, and psychoticism [8, 33]. Seven questions that have clinical significance but are left out of the previous dimensions constitute a score of “Additional Items”. The global severity index (GSI), the positive symptom distress index, and the positive symptom total were calculated, totaling 13 SCL-90 scores. Validation procedures for the creation of a Brazilian version were conducted according with the rules of the American Psychological Association [34]. Under the supervision of the scale owners, a committee established the Brazilian Portuguese version of the SCL-90 by means of translation, backtranslation, and comparison. After a pilot test, the committee and the license owners judged the version adequate for clinical use. At baseline, when each subject arrived to undergo overnight polysomnography examination, the SCL-90 and the ESS were completed in full. After at least 2 months' followup, the subjects repeated the questionnaires online. Statistical analysis The results were expressed as mean, standard deviation, and standard error of the mean. Student's t test for independent samples was used for comparisons between untreated and CPAP-treated patients; Student's t test for paired samples was used to compare results at baseline and at follow-up. Analysis of variance was performed to determine the differences and interactions between treated and untreated groups by time and among durations of CPAP use (i.e., 2–3 months, 3–4 months, 5–6 months, >6 months) in SCL-90 scores. Parsimonious linear regression models were constructed in an attempt to predict change in SCL-90 scores, using the variables of interest (ESS, SaO2, and SWS) as regressors. Because the SCL-90 had not been validated in subjects with sleep apnea, we performed factor and reliability analyses to test its psychometric properties. Factor analysis used principal components extraction with varimax rotation and Kaiser normalization. SPSS v18 was used for all statistical analyses (SPSS, Chicago, IL). A two-tailed probability of alpha error lower than 0.05 was regarded as statistically significant.
Results As shown in Fig. 1, 189 eligible subjects that, at the telephone contact, had accepted to participate did not reply to the e-mail granting access to the consent form and to the questionnaires.
Patients in database with AHI>60/h (n=470)
Not included in the study (n=208) Not reached by telephone - 95 Upper airway intervention - 49 Weight loss >2 Kg - 34 Psychotropic medication - 15 Debilitating illness - 10 Death - 5
Elegible patients (n=262)
Declined to participate (n=189) Included in this study (n=73)
Untreated (n=27)
CPAP-treated (n=46)
Fig. 1 Flowchart of the study protocol
The reasons for this subsequent decline were not recorded. Of the 73 patients included, 46 reported having undergone treatment with CPAP for at least 2 months. Patients who rejected any treatment for OSA despite medical advice (27) constituted the control group (Fig. 1). Anthropometric, clinical, and polysomnographic data on CPAP-treated and untreated patients are displayed in Table 1. Baseline data were similar between untreated and CPAP-treated patients. Comparisons of the SCL-90 scores between CPAP-treated and untreated patients at baseline and at follow-up are shown in Table 2. CPAP-treated and untreated patients had similar SCL-90 scores at baseline. CPAP-treated patients showed reductions in ten SCL-90 scores at follow-up. On comparison with controls, CPAP-treated patients exhibited significant reductions between baseline and follow-up scores in six of the dimensions of the SCL-90 (Fig. 2). Among the CPAP users, 17 % reported using the device for less than 4 h/night and less than 4 nights/week. These patients with poor adherence to CPAP therapy were not significantly different from the compliant ones, regarding all the variables described in Tables 1 and 2. Hence, they were maintained in the analysis. Differences between baseline and follow-up in each of the SCL-90 scores by CPAP adherence, in terms of months of use, days of CPAP use per week, and hours of CPAP use per day were nonsignificant. The differences in body mass index from baseline to follow-up (time) and between CPAP groups were also nonsignificant and showed no interaction (P =0.5). Factor and reliability analyses showed that the original psychometric properties of the SCL-90 [8] were conserved in this sample. The Cronbach's alpha for the nine dimensions was 0.944 at baseline and 0.949 at follow-up. Only one component was identified on factor analysis, both at baseline and at follow-up.
Sleep Breath Table 1 Polysomnographic and other characteristics of untreated and CPAP-treated patients with severe obstructive sleep apnea
Values expressed as means and standard deviations BMI body mass index in kg/m2 (weight/height2 ), CPAP continuous positive airway pressure, ESS Epworth sleepiness scale, AHI apnea–hypopnea index, SaO 2 arterial oxygen saturation *P 30 kg/m2, n (%) BMI at follow-up 30 kg/m2, n (%) CPAP time, n (%) 2–3 months 3–4 months 4–6 months >6 months Days on CPAP, n (%) 1–4 days/week 4–6 days/week >6 days/week Hours on CPAP, n (%) 1–4 h/day 4–6 h/day
4 (15.4) 22 (80.8) 33.4±5.0 1 (3.7) 6 (22.2) 20 (74.1)
8 (17.4) 37 (80.4) 33.6±4.9 1 (2.2) 13 (28.3) 32 (69.6)
– – – –
2 (4) 5 (11) 7 (15) 32 (70)
– – –
4 (9) 7 (15) 35 (76)
– –
8 (17) 16 (35)
6–8 h/day ESS, baseline ESS, follow-up Polysomnographic Parameters AHI (events/h) SaO2 nadir (%)
– 13.5±5.1 14.6±5.6
22 (48) 15.5±4.8 9.2±4.9
0.09
