Predictors and patterns of insomnia symptoms in OSA before and after PAP therapy

Accepted Manuscript Original article Predictors and Patterns of Insomnia Symptoms in OSA Before and After PAP Therapy Robert N. Glidewell, Brenna N. Renn, Emily Roby, William C. Orr PII: DOI: Reference:

S1389-9457(14)00187-7 http://dx.doi.org/10.1016/j.sleep.2014.05.001 SLEEP 2452

To appear in:

Sleep Medicine

Received Date: Revised Date: Accepted Date:

18 October 2013 3 March 2014 2 May 2014

Please cite this article as: Glidewell, R.N., Renn, B.N., Roby, E., Orr, W.C., Predictors and Patterns of Insomnia Symptoms in OSA Before and After PAP Therapy, Sleep Medicine (2014), doi: http://dx.doi.org/10.1016/j.sleep. 2014.05.001

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Predictors and Patterns of Insomnia Symptoms in OSA Before and After PAP Therapy Running Head: Predictors of Insomnia in OSA

Robert N. Glidewell, PsyD1, Brenna N. Renn, MA2, Emily Roby, PsyD1, and William C. Orr, PhD3,4,1 1. Lynn Institute for Healthcare Research, Colorado Springs, CO 2. University of Colorado, Colorado Springs, CO 3. Lynn Institute for Healthcare Research, Oklahoma City, OK 4.

University of Oklahoma Health Sciences Center

The authors have no conflict of interest related to this study. This research was performed at the Lynn Institute for Healthcare Research in Colorado Springs, CO. Financial support was provided by the Lynn Institute for Healthcare Research.

Key Words: insomnia, sleep apnea, positive air pressure (PAP), insomnia severity index (ISI), comorbidity, apnea-hypopnea index (AHI), respiratory disturbance index (RDI)

Robert N. Glidewell, PsyD*

Brenna N. Renn, MA

Emily K. Roby, PsyD

William C. Orr, PhD

Lynn Institute for Healthcare Research

University of Colorado, Colorado Springs

Lynn Institute for Healthcare Research

Lynn Institute for Healthcare Research

1625 Medical Center Point, Suite 260 Colorado Springs, CO 80907

Department of Psychology 1420 Austin Bluffs Pkwy Colorado Springs, CO 80918

1625 Medical Center Point, Suite 260 Colorado Springs, CO 80907

3555 NW 58th St., Suite 800 Oklahoma City, OK 73112

Phone: 719-373-0051 [email protected]

ABSTRACT Objective: Identify factors that predict improvement versus persistence of insomnia symptoms following treatment of obstructive sleep apnea (OSA) with positive airway pressure (PAP) therapy. Methods: Archival data from 68 PAP-treated sleep apnea patients ages 25-83 were analyzed using nonparametric tests and stepwise regression to assess the relationships between insomnia symptoms, multiple OSA variables, and PAP use over time. Results: Pretreatment insomnia symptom severity (ISS; b = -0.72, p < .001), PAP average use (b = -0.01, p = .01) and respiratory disturbance index (RDI; b = -0.02, p = .03) predict change in insomnia following PAP therapy. Forty-five percent (24/53) of subjects with moderate to severe insomnia at pretreatment reported no/mild symptoms after PAP therapy and were considered improved. Improved subjects had lower pretreatment ISS (p < .001), higher RDI (p = .01), and higher average PAP use (p < .035) than subjects with persistent insomnia. Number of medications and comorbidities was similar between improved and persistent groups. New onset of insomnia symptoms occurred in 13% (2/15) patients with no/mild pretreatment insomnia. Conclusions: Although insomnia symptom severity declines following PAP treatment, 55% of OSA patients have persistent moderate to severe symptoms despite treatment. More severe OSA is linked to higher likelihood of insomnia improvement and the effect of PAP therapy on insomnia may be mediated by OSA severity. Persistent insomnia is unrelated to medication use or comorbidities and may represent an independent, self-sustaining disorder requiring targeted intervention.

INTRODUCTION Insomnia complaints are common in patients with obstructive sleep apnea (OSA) [1-4]. Using stringent research criteria for clinically significant insomnia, Smith and colleagues found insomnia in 39% of their sample of 105 OSA patients [4]. Using a variety of insomnia definitions, other studies have also documented insomnia complaints in 42% to 68% of OSA patients [5-10]. In combination, OSA and insomnia are associated with significantly greater morbidity, functional impairment, and reduced quality of life than either condition alone. In one study, patients with OSA and comorbid insomnia had moderate to severe symptoms of depression, anxiety, and stress while OSA-only patients indicated no clinically significant symptoms in these areas [4]. In this same study, OSA patients with insomnia also reported greater dysfunctional beliefs and attitudes about sleep than OSA-only patients. Similar findings regarding dysfunctional beliefs and emotional disturbance were reported more recently by Yang and colleagues [11]. Sleep apnea patients with insomnia also have greater sleep disturbance as measured by both sleep diary and polysomnography [4,6]. The presence of insomnia symptoms is a predictor of excessive daytime sleepiness in patients with moderate to severe OSA [12]. In a sample of older adults, OSA with insomnia was associated with significantly reduced daytime function and slower reaction times compared with individuals having only insomnia or OSA [13]. There are also data to indicate that insomnia symptoms negatively affect adherence to positive airway pressure (PAP) therapy [9,14-15]. In recognition of the prevalence of insomnia in OSA patients and evidence that the presence of insomnia in these patients is associated with increased morbidity and poorer adherence to PAP therapy, there have been frequent calls in the literature for development of

multidisciplinary protocols with systematic attention to OSA and insomnia complaints [1-2,1617]. Unfortunately, there is a dearth of evidence to guide development of such protocols. Although many studies have reported insomnia symptoms following treatment of OSA, studies examining the course of insomnia symptoms in OSA patients over time are scarce [4-8, 18-20]. Three published studies have examined insomnia in OSA patients in response to both cognitive behavioral therapy for insomnia (CBT-I) and treatment of OSA (PAP, surgery, or nasal dilator strips) [21-23]. Findings of all three studies strongly suggest that insomnia in these patients is due to both psychophysiological and respiratory factors and cannot be resolved with treatment of OSA alone. Only two published studies have investigated insomnia as the primary outcome in OSA patients treated with PAP therapy [9,24]. In each case, although PAP use was associated with major reduction or resolution of insomnia symptoms, half of patients continued to have significant insomnia symptoms despite PAP use. The current study examines insomnia symptoms over time in a sample of OSA patients treated with PAP. The primary aims of the study are a) to document the rate of improvement, persistence, or onset of insomnia symptoms following treatment of OSA with PAP therapy and b) identify factors that predict improvement, persistence, or onset of insomnia symptoms in this population. Findings of this study will contribute to existing literature by helping to specify who might be expected to experience resolution of insomnia symptoms solely with treatment of OSA versus those who might require a more intensive multidisciplinary approach. It will also indirectly distinguish the contribution of respiratory disturbance to insomnia symptoms versus other possible insomnia symptom etiologies. This will support the selection and course of insomnia interventions for inclusion in multidisciplinary treatment models.

METHODS Participants Archival data from 93 OSA patients (apnea-hypopnea index [AHI] ≥ 5 or respiratory disturbance index [RDI] ≥ 15) who initiated PAP therapy using devices with objective adherence monitoring capability between August 28, 2007 and February 4, 2010 were analyzed. Subjects were identified for inclusion regardless of age, gender, body mass index (BMI), or severity of OSA. Of these 93 patients, n = 20 were excluded because they did not have clinical follow-up within the study time period and n = 5 were excluded because they had incomplete insomnia questionnaires. A total of 68 subjects were retained for analysis. This study was reviewed by the Western Institutional Review Board's Regulatory Affairs Department (#234142) and was determined to meet conditions for exemption under 45 CFR 46.101(b)(4). Procedure All data were collected as part of the routine laboratory and clinical care of patients referred for evaluation, diagnosis and treatment of OSA. Overnight laboratory based polysomnography was used to obtain OSA diagnostic data (AHI, RDI). Self-reported demographic, medical and psychiatric status, and current prescription medications were collected via intake questionnaire at the first clinic or laboratory visit. Patients completed selfreport questionnaires regarding insomnia and daytime sleepiness following polysomnography but prior to initiation of PAP treatment (pretreatment measurement). Objective PAP adherence data was collected and symptom questionnaires were also completed at all follow-up visits. Follow-up visits were intended to occur at 7, 30, and 90 days after PAP set up. However, a large degree of variability was present in the actual timing of follow-up visits due to limitations in

clinician availability and patient availability and attendance. For the purpose of this study, the posttreatment measurement was defined as occurring between 31-60 days after PAP set up. In cases where multiple visits occurred within this day range, the earliest visit was always used for analysis. Actual follow up for this sample occurred an average of 43.1 ± 7.1 days following initiation of PAP therapy. Measures Insomnia symptomatology was measured by the three insomnia symptom severity items (onset, middle, and terminal) from the Insomnia Severity Index (ISI), which is a reliable and valid seven item self-report questionnaire for evaluation of perceived insomnia severity [25-26]. For the purpose of the following discussion, we refer to the sum of these three ISI items as the ISI subscale score. These items are scored on a five-point Likert scale (0-4) with scores of 0 or 1 indicating “None” or “Mild” symptom severity and scores of 2 to 4 indicating “moderate,” “severe,” or “very severe” symptoms. The ISI subscale has a range of possible total scores from 0 to 12. We selected a cut-off of ≥ 4 on the ISI subscale to categorize patients with clinically significant insomnia symptoms versus non-clinical symptomatology because to meet or exceed this threshold at least one insomnia severity item must be rated as moderate or higher. We chose to use the ISI subscale rather that the full ISI score because four of the seven ISI items are nonspecific to insomnia and likely to be endorsed by sleep apnea patients even when insomnia symptoms are absent. Accordingly, we were concerned that in the context of insomnia symptoms and OSA, changes in the ISI full score may not actually represent changes in insomnia symptomatology. There is evidence indicating the ISI subscale does measure a unique insomnia severity factor that is reliable in insomnia patients with medical comorbidity [26-27]

In order to measure changes in insomnia symptom severity over time we computed an ISI subscale change score by subtracting posttreatment ISI Subscale scores from pretreatment ISI Subscale scores. ISI subscale change scores have a possible range of -12 to +12 with negative scores indicating reduced insomnia symptom severity and positive scores indicating increased insomnia symptom severity following PAP treatment. The AHI and RDI were determined by laboratory-based polysomnography (PSG). A diagnostic threshold of AHI ≥ 5 or RDI ≥ 15 was used for subject selection [28]. PSG was conducted using the Alice 5 system. The following channels were monitored: frontal, central, and occipital EEG; electrooculogram; submentalis and anterior tibialis; electrocardiogram; nasal and oral airflow; thoracic and abdominal wall motion. Arterial oxygen was monitored with a pulse oximeter. Sleep and respiratory events were scored using the American Academy of Sleep Medicine scoring rules [29]. Daytime sleepiness was assessed via the Epworth Sleepiness Scale (ESS), an 8-item selfreport questionnaire that evaluates the subjective likelihood of dozing in common daily situations [30]. Items are scored on a 4-point Likert scale with 0 indicating no chance of dozing in a particular situation and 3 indicating a high chance of dozing in a particular situation. Possible scores range from 0 to 24. PAP adherence was determined by standard objective monitoring technology integrated within subjects’ PAP device. Patients were aware that their adherence was being monitored by the device and that this information would be reviewed at follow up visits. We chose average minutes of nightly PAP use as the primary measure of PAP adherence. Analysis

Statistical analysis was performed using Statistical Package for the Social Sciences version 22 for Windows. Patients were sorted into pretreatment insomnia symptom groups (“insomnia,” meaning clinically elevated insomnia symptoms versus “no-insomnia,” meaning no/minimal symptomatology) based on their pretreatment ISI subscale scores. Descriptive statistics were calculated, and a chi-square test for independence and Mann-Whitney U tests were computed to compare demographic variables (age, gender, BMI), insomnia symptoms, pretreatment ESS scores, OSA severity (i.e., AHI and RDI), PAP characteristics (average PAP use, PAP pressure), and medical and psychological covariates of these groups. Wilcoxon Signed Rank tests were used to detect change in ISI subscale scores after PAP treatment for the full sample and each pretreatment insomnia group (non-symptomatic and clinically elevated insomnia symptoms). Predictors of insomnia symptom severity following PAP therapy were examined for patients in the pretreatment insomnia group. Bivariate correlation analysis yielded a correlation matrix of all possible predictor variables and ISI change score. Possible predictor variables with r < ±.20 were discarded from the model as they are not likely to contribute to the explained variance in ISI change after PAP. These excluded variables included age (r = .01, p = .93), gender (r = -.14, p = .34), PAP pressure (r = .18, p = .21), pretreatment ESS score (r = .14, p = .33), total number of prescription medications (r = -.19, p = .21), total psychiatric diagnoses (r = .00, p = .99), and total number of medical diagnoses (r = -.07, p = .63). BMI was not entered into the regression model because it covaries with the predictors AHI and RDI (r = .38, p = .001 and r = .34, p = .004, respectively), and can be conceptualized as a risk factor for OSA rather than a direct indicator of disease state. Additionally, there is no clinical expectation that BMI would

have a meaningful effect on response to PAP therapy or directly influence insomnia symptom severity. The remaining predictor variables of AHI, RDI, average PAP use (minutes per night), and ISI subscale pretreatment score were entered into a stepwise (forward selection) regression to assess changes in insomnia symptom severity following PAP. Potential predictor variables were entered into the model if they were significant (p ≤ .05); they were only retained if, in subsequent models, they continued to show some trend (p < .10). Multicollinearity was assessed according to the variance inflation factor (VIF); no predictor variables were found to be collinear. These patients with clinically significant pretreatment insomnia symptoms were then sorted to identify subjects with persistent insomnia symptoms (posttreatment ISI subscale score ≥ 4) and those whose insomnia symptoms had improved (posttreatment ISI subscale score < 4). Non-parametric Wilcoxon Signed Rank test identified within-group differences in ISI subscale scores between pre- and posttreatment for individuals in both the improved and persistent groups. Mann-Whitney U tests analyzed differences in RDI, PAP average use, and medical covariates between patients with improved and persistent insomnia symptoms.

The sample size of the pretreatment no-insomnia group was too small to allow for statistical analysis. Accordingly, only frequencies of insomnia symptom patterns and descriptions of PAP use and OSA severity are provided for patients in this group. RESULTS Sixty-eight patients were available for complete analysis. Mann-Whitney U tests revealed patients in the pretreatment insomnia and no-insomnia groups were of comparable

age, BMI, OSA severity (AHI and RDI), ESS, PAP pressure, and average PAP use. A chi-square test for independence (using Yates Continuity Correction) indicated no significant association between gender and pretreatment insomnia symptom status, χ² (1, n = 68) = .000, p = 1.00. Participant characteristics, insomnia symptom, OSA severity, and PAP characteristics for the full sample and each insomnia status group are shown in Table 1. Patients without pretreatment insomnia symptoms differed from those with insomnia symptoms on medical variables. MannWhitney U tests demonstrated that patients with pretreatment insomnia symptoms were prescribed more medication (z = -2.43, p = .02), had more medical conditions (z = -2.11, p = .04), and were more likely to have a psychiatric diagnosis (z = -2.00, p = .05). Additional detail regarding patients’ medical and psychiatric diagnoses, including sleep disorders and therapeutic classes of prescription medication, is provided in Appendix A. [Insert Table 1] Results of the Wilcoxon Signed Rank test indicated significant within-group change in insomnia subscale scores (Table 1) from pre- to posttreatment for the full sample (z = -4.61, p < .001), with improvement on the median ISI subscale score from pretreatment (Md = 5.00) to posttreatment (Md = 3.00). The subset of patients with pretreatment insomnia symptoms showed a similar decrease in ISI subscale scores (Mdpretreatment = 6.00, Mdposttreatment = 4.00), z = 5.26, p < .001. Patients without insomnia symptoms at pretreatment demonstrated no significant change in ISI subscale scores following PAP therapy, z = -1.81, p = .07, and median scores remained in the non-symptomatic range (Mdpretreatment = 2.00, Mdposttreatment = 3.00). Predictors of ISI Subscale Change Scores in Patients with Insomnia at Pretreatment

A stepwise (forward selection) regression was conducted for patients in the pretreatment insomnia group (n = 53) to assess predictors of ISI subscale change score. The overall regression model, which included three of the four predictor variables, was statistically significant, F(3, 48) = 12.32, p < .001 and predicted 40.0% (adjusted R2; R = .66) of the variance in ISI subscale change scores. The contributions of the individual predictors were as follows: Step 1 entered pretreatment ISI subscale scores, producing an R2 increment of .284, F(1, 50) = 19.83, p < .001; PAP average use was entered in Step 2, producing an R2 change of .092, F(1, 49) = 7.21, p = .01; finally, RDI was entered in Step 3, which resulted in an R2 increment of .059 and a change of F(1, 48) = 5.03, p = .03. The final model retained these three predictors, showing that ISI subscale change after PAP was associated with pretreatment insomnia symptom severity (b = -0.72), average PAP use (b = -0.01), and RDI (b = -0.02). There were no significant correlations between RDI, PAP use, or pretreatment ISI subscale scores. Patterns of ISI Subscale Change Scores and Posttreatment Insomnia Status Of the 53 patients with pretreatment insomnia symptoms, n = 24 (45.3%) were considered improved and n = 29 (54.7%) persisted with insomnia symptoms following PAP treatment. We examined average PAP use, RDI, medical covariates, and pre- and posttreatment ISI subscale scores to identify differences in these variables with regard to insomnia changestatus (Table 2). Wilcoxon Signed Rank tests of within-group differences indicated improvement in ISI subscale scores from pre- to posttreatment for individuals in both the improved insomnia symptom group (z = -4.30, p < .001) and the persistent symptom group (z = -2.75, p = .006). Mann-Whitney U tests demonstrated differences in pretreatment ISI subscale scores, such that patients in the improved group reported significantly lower ISI subscale scores at pretreatment

than patients with persistent insomnia (z = -1.97, p = .05). Patients with improved versus persistent insomnia at posttreatment differed significantly in their RDI, such that improved patients had a higher median RDI (Md = 54.00) compared to patients with persistent insomnia symptoms (Md = 34.50). Improved and persistent insomnia symptom patients also differed on average PAP use, with improved patients using PAP for longer (Md = 363.00) than patients who had persistent insomnia symptoms (Md = 263.00). The improved and persistent insomnia groups were selected based on their posttreatment ISI subscale scores resulting in an intentionally significant difference between these groups on this variable. There were no other significant between-group differences in pretreatment ISI subscale scores. [Insert Table 2] Patterns of ISI Subscale Scores in Patients without Insomnia at Pretreatment Twenty-two percent (15/68) patients in this sample had no or mild insomnia symptoms at pretreatment. Only two of these patients (3% of the total sample) reported insomnia symptoms of moderate or greater severity following PAP. The remaining 13 patients with no/mild insomnia symptoms at pretreatment continued to have no/mild insomnia symptoms following PAP therapy. DISCUSSION Although insomnia symptoms clearly improve with PAP therapy, 55% of our sample continued to have symptoms of at least moderate severity despite apparently adequate treatment of OSA with PAP. This is similar to results from two comparable published studies that reported persistent insomnia symptoms in 49% and 53% of OSA subjects treated with PAP [9,24].

Higher average PAP use was associated with greater reductions in insomnia symptom severity and subjects with persistent insomnia symptoms used PAP an average of 72 minutes less per night than those considered improved. It is reasonable to think that if subjects with persistent insomnia symptoms had used PAP at the same rate as their improved peers, they would have experienced comparable reductions in insomnia symptom severity. However, a study of insomnia symptoms before and after two years of PAP therapy found no significant differences in subject characteristics or primary outcomes between groups with a mean difference of more than three hours in average nightly PAP use durations (3.5±2.2 vs 6.8±1.2 hours) [9]. With this in mind, it seems unlikely that an additional 72 minutes of PAP use would have resulted in a 3-fold improvement for our subjects with persistent insomnia symptoms. More severe OSA was associated with greater reductions in insomnia symptom severity and subjects in the improved group had more severe OSA than subjects in the persistent insomnia symptom group. Of all prior studies that examined only pretreatment insomnia symptoms in subjects with OSA, only one, published by Lichstein and colleagues, found OSA severity and insomnia to be related [4,6-8,10,13-14,31-33]. However, consistent with the findings of the current study, studies evaluating insomnia as a treatment outcome consistently report OSA severity as a factor related to treatment response [9,24]. Lichstein and colleagues report that, at pretreatment, comorbid insomnia cannot be distinguished from insomnia symptoms due to OSA [10]. We agree with this and suggest that, as a result, distinguishing comorbid insomnia symptoms from insomnia symptoms due to OSA must be done indirectly through insomnia persistence or response in the face of evidence-based treatment. This is

consistent with results from the few studies examining treatment of both insomnia and sleep disordered breathing using parallel or sequential treatment methodology [21-23] The documented relationships between insomnia, PAP use, and OSA severity suggest that, once a minimally adequate level of PAP use is achieved, the magnitude of benefit from PAP therapy on insomnia symptoms may have less to do with duration of PAP use and more to do with the severity of OSA. From this perspective, the more severe the OSA, the greater the likelihood insomnia symptoms are directly related to respiratory disturbance, and the more likely they are to use and respond to PAP therapy. This interpretation has good face validity and supports a) conceptualization of insomnia as partially secondary to respiratory disturbance and b) that OSA severity may moderate the effect of PAP therapy on insomnia symptoms. Higher pretreatment insomnia severity was associated with greater reductions in insomnia symptom severity following PAP therapy. We had some concern that this association may represent a regression toward the mean with subjects having more severe insomnia symptoms at pretreatment inherently also having greater likelihood and opportunity for improvement. However, despite having significantly less severe pretreatment insomnia symptoms, subjects in the improved group experienced three times greater improvement in insomnia symptom severity following PAP therapy than subjects with persistent insomnia symptoms. This suggests that regression toward the mean had a negligible influence on reported changes in insomnia symptom severity in this study. Subjects in the persistent insomnia group had more severe insomnia symptoms at pretreatment than those in the improved group. There is some evidence that higher frequency of insomnia symptoms at pretreatment is associated with lower PAP use [9,16]. This may

partially explain the lower average PAP use in the persistent group. However, in this study, there was no significant correlation between pretreatment insomnia symptom severity and average PAP use. In combination, the identified differences in pretreatment insomnia severity, OSA severity, and response to PAP therapy suggest that persistent and improved subjects may represent two distinct diagnostic groups. Subjects in the improved group, with more severe OSA (RDI 70% higher than in the persistent group) and virtual elimination of insomnia symptoms with PAP therapy may represent relatively pure OSA pathology. While subjects in the persistent insomnia symptom group, with more severe insomnia symptoms, less severe OSA, and minimal resolution of insomnia symptoms with PAP therapy, may represent subjects with comorbid insomnia symptoms unrelated to respiratory disturbance and requiring targeted intervention. Although our study examined symptoms and not insomnia diagnoses, there is growing support for this conceptualization of insomnia in the context of OSA [1-2,9-10,16-17]. There is evidence from treatment studies to suggest a role for psychophysiological factors as a probable cause for persistent insomnia symptoms in our sample. Guilleminault and colleagues found that CBT-I following surgical intervention for mild OSA resulted in significant improvements in total sleep time, rapid eye movement sleep, slow wave sleep, sleep onset latency, and wake after sleep onset after controlling for the effects of the surgical intervention on these variables [23]. Krakow and colleagues found similar independent effects of CBT-I and mixed OSA interventions on insomnia symptoms in subjects with comorbid insomnia and OSA [22]. OSA subjects with insomnia at pretreatment did have more medical and psychiatric comorbidities and used more medications than subjects without insomnia symptoms at

pretreatment. However, the absence of significant differences in these variables between improved versus persistent subjects suggests that medication use, medical, and psychiatric comorbidity likely played no significant role in persistence of insomnia symptoms in our sample. Interestingly, visual inspection of differences between these groups for individual medications and comorbidities reveals higher rates of chronic pain, opiate use, metabolic disorders, and antihypertensive/antiarrhythmic medication use in the improved group. Unfortunately the frequencies for individual medications and comorbidities are insufficient for statistical analysis of group differences. Subjects without insomnia at pretreatment rarely develop significant insomnia symptoms following initiation of PAP therapy. There was no significant change in insomnia symptom severity for subjects with no insomnia at pretreatment. Certainly, those with minimal or no insomnia symptoms at pretreatment have little room for improvement. More importantly, as a group these subjects did not report a significant increase in insomnia symptom severity following PAP therapy. In fact, only two subjects in our sample (13% of the 15 subjects with no-insomnia at pretreatment) reported onset of new insomnia symptoms of moderate or greater severity following PAP therapy. Caetano Mota and colleagues reported that 21% of the 56 individuals with OSA in their study developed new-onset insomnia following initiation of PAP [33]. A more recent study found new onset of insomnia symptoms in 22% of 294 OSA subjects after 24 months of PAP therapy [9]. Use of different insomnia symptom definitions and measures likely explain the difference in rates of new-onset insomnia between studies.

The current study is not without limitations. We did not control for other treatments for insomnia (e.g., medication, including over-the-counter sleep aids; CBT-I; self-help; etc.). Accordingly, undocumented insomnia treatments could have skewed insomnia change scores in the direction of insomnia improvements being over-attributed to treatment of OSA, but it seems unlikely that this would affect the results in any systematic fashion. Additionally, although all subjects had effective PAP pressures identified via in-laboratory titrations, effectiveness of PAP therapy in the home setting was not evaluated and it is possible that the current findings of persistent insomnia are associated with subtherapeutic treatment of OSA. One study has reported that 17% of those with OSA who were asymptomatic (e.g. no complaint of snoring or excessive daytime sleepiness) and compliant with PAP therapy for three months continued to have an AHI of greater than 10 when evaluated with 16 channel at home PSG [34]. Additionally, our reliance on a clinical population resulted in a small sample size, despite data collection over a period of approximately 30 months. Nonetheless, this study is likely representative of the populations that sleep clinicians encounter in their practices. Future studies are needed to generalize and extend these findings. In particular, replication of the regression model would more clearly ascertain the predictive role of pretreatment insomnia symptoms, PAP use, and RDI in insomnia symptom remission following PAP treatment. In summary, PAP therapy is clearly associated with a reduction in insomnia symptom severity. The effect of PAP on insomnia symptoms in OSA may be moderated by OSA severity, with those with more severe OSA being more likely to experience resolution of insomnia symptoms following PAP therapy. Over half of subjects continued to have insomnia symptoms of at least moderate severity despite PAP therapy. Persistent insomnia symptoms were not

related to medication use or medical or psychiatric comorbidity and may represent psychophysiological insomnia.

References 1. Luyster FS, Buysse DJ, Strollo PJ. Comorbid insomnia and obstructive sleep apnea: challenges for clinical practice and research. J Clin Sleep Med 2010;6:196-204. 2. Wickwire EM, Collop NA. Insomnia and sleep-related breathing disorders. Chest 2010;137:1449-63. 3. Wickwire EA. Combined insomnia and poor CPAP compliance a case study and discussion. Clinical Case Studies 2008;7:267-86. 4. Smith S, Sullivan K, Hopkins W, Douglas J. Frequency of insomnia report in patients with obstructive sleep apnoea hypopnea syndrome (OSAHS). Sleep Med 2004;5:449-56. 5. Björnsdóttir E, Janson C, Gíslason T, Sigurdsson JF, Pack AI, Gehrman P, et al. Insomnia in untreated sleep apnea patients compared to controls. J Sleep Res. Article first published online 12 October 2011: DOI: 10.1111/j.1365-2869.2011.00972.x. 6. Krakow B, Melendrez D, Ferreira E, Clark J, Warner TD, Sisley B, et al. Prevalence of insomnia symptoms in patients with sleep-disordered breathing. Chest 2001;120:1923-9. 7. Krell SB, Kapur VK. Insomnia complaints in patients evaluated for obstructive sleep apnea. Sleep Breath 2005;9:104-10. 8. Chung KF. Insomnia subtypes and their relationships to daytime sleepiness in patients with obstructive sleep apnea. Respiration 2005;72:460-5. 9. Björnsdóttir E, Janson C, Arnardóttir E, Pack A, Gíslason T, Benediktsdottir B. Symptoms of insomnia among OSA patients before and after 2 years of PAP treatment. Sleep Med 2013;14S:e74.

10. Lichstein KL, Justin Thomas S, Woosley JA, Geyer JD. Co-occurring insomnia and obstructive sleep apnea. Sleep Med 2013;14:824-829. 11. Yang CM, Liao YS, Lin CM, Chou SL, Wang EN. Psychological and behavioral factors in patients with comorbid obstructive sleep apnea and insomnia. J Psychosom Res 2011;70:35561. 12. Kapur VK, Baldwin CM, Resnick HE, Gottlieb DJ, Nieto FJ. Sleepiness in patients with moderate to severe sleep-disordered breathing. Sleep 2005;28:472-7. 13. Gooneratne NS, Gehrman PR, Nkwuo JE, Bellamy SL, Schutte-Rodin S, Dinges DF, et al. Consequences of comorbid insomnia symptoms and sleep-related breathing disorder in elderly subjects. Arch Intern Med 2006;166:1732-8. 14. Pieh C, Bach M, Popp R, Jara C, Crönlein T, Hajak G, et al. Insomnia symptoms influence CPAP compliance. Sleep Breath. Article first published online 4 February 2012: DOI: 10.1007/s11325-012-0655-9. 15. Wickwire EM, Smith MT, Birnbaum S, Collop NA. Sleep maintenance insomnia complaints predict poor CPAP adherence: A clinical case series. Sleep Med 2010;11:772-6. 16. Ong, JC, Crisostomo MI. The more the merrier? Working towards multidisciplinary management of obstructive sleep apnea and comorbid insomnia. J Clin Psychol 2013;69:106677. 17. Glidewell RN. Comorbid insomnia and sleep disordered breathing. Curr Treat Option Ne 2013;15:692-703. 18. Kribbs NB, Pack AI, Kline LR, Smith PL, Schwartz AR, Schubert NM, et al. Objective measurement of patterns of nasal CPAP use by patients with obstructive sleep apnea. Am Rev

Respir Dis 1993;147:887-95. 19. Weaver TE, Kribbs NB, Pack AI, Kline LR, Chugh DK, Maislin G, et al. Night-to-night variability in CPAP use over the first three months of treatment. Sleep 1997;20(4):278-283. 20. Aloia MS, Arnedt JT, Stanchina M, Millman RP. How early in treatment is PAP adherence established? Revisiting night-to-night variability. Behav Sleep Med 2007;5:229-40. 21. Guilleminault C, Palombini L, Poyares D, Chowdhuri S. Chronic insomnia, premenopausal women and sleep disordered breathing: part 2. Comparison of nondrug treatment trials in normal breathing and UARS post-menopausal women complaining of chronic insomnia. J Psychosom Res 2002;53:617-23. 22. Krakow B, Melendrez D, Lee S, Warner T, Clark J, and Sklar D. Refractory insomnia and sleep-disordered breathing: A pilot study. Sleep Breath 2004;8:15-29. 23. Guilleminault C, Davis K, Huynh NT. Prospective randomized study of patients with insomnia and mild sleep disordered breathing. Sleep 2008;31:1527-33. 24. Nguyên XL, Rakotonanahary D, Chaskalovic J, Fleury B. Insomnia related to sleep apnoea: Effect of long-term auto-adjusting positive airway pressure treatment. Eur Respir J 2013;41:593-600. 25. Morin CM. Insomnia: Psychological assessment and management. New York: The Guilford Press, 1993. 26. Bastien C. Validation of the Insomnia Severity Index as an outcome measure for insomnia research. Sleep Med 2001;2:297-307. 27. Savard MH, Savard J, Simard S, Ivers H. Empirical validation of the insomnia severity index in cancer patients. Psycho-Oncology 2005;14:429-441.

28. American Academy of Sleep Medicine Task Force. Sleep-related breathing disorders in adults: Recommendations for syndrome definition and measurement techniques in clinical research. The Report of an American Academy of Sleep Medicine Task Force. Sleep. 1999; 22:667-89. 29. Iber C, Ancoli-Isreal S, Chesson A, et al. The AASM manual for the scoring of sleep and associated events: Rules, terminology, and technical specifications. Westchester, Illinois: American Academy of Sleep Medicine; 2007. 30. Johns M. A new method for measuring daytime sleepiness: The Epworth sleepiness scale. Sleep 1991;14:540-545. 31. Hagen C, Patel A, McCall WV. Prevalence of insomnia symptoms in sleep laboratory patients with and without sleep apnea. Psychiatry Res 2009;170:276-7. 32. Bixler E, Vgontzas A, Lin H, Vela-Bueno A, Kales A. Insomnia in central Pennsylvania. J Psychosom Res 2002;53:589-92. 33. Caetano Mota P, Morais Cardoso S, Drummond M, Santos AC, Almeida J, Winck JC, et al. Prevalence of new-onset insomnia in patients with obstructive sleep apnoea syndrome treated with nocturnal ventilatory support. Rev Port Pneumol 2012;18:15-21. 34. Baltzan MA, Kassissia I, Elkholi O, Palayew M, Dabrusin R, Wolkove N, et al. Prevalence of persistent sleep apnea in patients treated with continuous positive airway pressure. Sleep 2006;29:557-63.

Legend Table 1. Demographic, OSA severity, PAP characteristics, medical and psychological status, and insomnia symptoms for the full sample and pretreatment insomnia symptom status groups. Data are presented as mean (SD). aPretreatment ISI Subscale score <4; bPretreatment ISI Subscale score ≥ 4; c Non-parametric test for characteristics of pretreatment insomnia symptomatology versus non-symptomatic groups; dGroup selection variable, intentionally significant between-group difference, p<.001; #Significant change in ISI Subscale score from preto posttreatment, p<.001; OSA, obstructive sleep apnea; PAP, positive airway pressure; Rx, Prescription; ISI, Insomnia Severity Index. Table 2. All data are presented as mean (SD). aPosttreatment occurred between 31 and 60 days post PAP setup; bGroup selection variable, intentionally significant between-group difference, p < .001; cMinutes per night; †Statistically significant between-group difference, p < .05; * Statistically significant within-group difference pre- to posttreatment, p < .05; ISI, Insomnia Severity Index; PAP, positive airway pressure; RDI, respiratory disturbance index; Rx, Prescription. Table 3. Frequency of prescribed medications, medical diagnoses, and psychiatric diagnoses for the full sample and pre- and posttreatment insomnia status groups. Data are presented as number (% of sample).

Table 1. Demographic, OSA severity, PAP characteristics, medical and psychological status, and insomnia symptoms for the full sample and pretreatment insomnia symptom status groups. Full Sample (N = 68)

Pretreatment No-Insomnia Symptomsa (n = 15)

Pretreatment Insomnia Symptomsb (n = 53)

Z (p)c

32.4

33.3

32.1

--- (1.00)

Age (years)

47.5 (12.4)

46.7 (12.9)

47.8 (12.3)

-0.10 (.92)

Body Mass Index

32.2 (7.3)

32.9 (6.1)

32.0 (7.7)

-0.52 (.60)

Apnea Hypopnea Index

34.7 (32.2)

31.2 (32.6)

35.7 (32.3)

-0.82 (.41)

Respiratory Disturbance Index

47.2 (29.6)

43.1 (30.1)

48.4 (29.6)

-0.89 (.38)

Epworth Sleepiness Scale

12.1 (4.9)

10.60 (4.2)

12.6 (5.0)

-0.55 (.58)

286.2 (132.1)

244.5 (178.5)

297.9 (115.1)

-1.25 (.21)

PAP pressure

10.8 (2.9)

10.4 (2.0)

10.9 (3.1)

-0.42 (.68)

Total Number of Rx Medications

3.7 (3.80)

1.62 (2.3)

4.28 (3.9)

-2.43 (.02)

Total Number of Medical Conditions

2.9 (1.5)

2.1 (1.1)

3.1 (1.5)

-2.11 (.04)

Total Number of Psychiatric Diagnoses

0.3 (.53)

.1 (.3)

.4 (.6)

-2.00 (.05)

Pretreatment ISI Subscale Score

5.4 (2.8)

1.7 (1.3)

6.4 (2.2)

-.59 (<.001)d

Posttreatment ISI Subscale Score

3.5 (2.4)

2.7 (2.2)

3.7 (2.4)

-1.60 (.11)

-1.9 (3.1)#

1.1 (2.5)

-2.7 (2.8)#

-4.47 (<.001)

Gender, Female (%)

Average PAP Use (min./night)

ISI Subscale Change Score

Data are presented as mean (SD). aPretreatment ISI subscale score <4; bPretreatment ISI subscale score ≥ 4; c Non-parametric d

test for characteristics of pretreatment insomnia symptomatology versus non-symptomatic groups; Group selection variable, intentionally significant between-group difference, p < .001; #Significant change in ISI subscale score from pre- to posttreatment, p < .001; OSA, obstructive sleep apnea; PAP, positive airway pressure.; Rx, Prescription; ISI, Insomnia Severity Index.

Table 2. ISI Subscale scores, Average PAP Use, RDI, and Medical Covariates for Persistent versus Improved Insomnia Groups. Improved

Persistent

Insomnia

Insomnia

Symptoms

Symptoms

(n = 24)

(n = 29)

Pretreatment ISI Subscalea

5.8 (1.8)†

6.9 (2.3)

-1.97 (.05) †

Posttreatment ISI Subscalea,b

1.5 (1.2)*

5.6 (1.4)*

-6.28 (<.001)†

337.1 (119.9)

265.5 (102.0)

-2.40 (.02) †

RDI

57.9 (29.1)

40.1 (28.1)

-2.45 (.01) †

Total Number of Rx Medications

4.79 (4.5)

4.3 (3.7)

-0.25 (.80)

Total Number of Medical Diagnoses

3.33 (1.7)

3.0 (1.4)

-0.57 (.57)

Total Number of Psychiatric Diagnoses

0.3 (0.6)

0.4 (0.6)

-0.21 (.84)

Average PAP Usec

z (p)

All data are presented as mean (SD). aPosttreatment occurred between 31 and 60 days post PAP setup; b

Group selection variable, intentionally significant between-group difference, p < .001; cMinutes per night;

†

Statistically significant between-group difference, p < .05; * Statistically significant within-group difference

pre- to posttreatment, p < .05; ISI, Insomnia Severity Index; PAP, positive airway pressure; RDI, respiratory disturbance index; Rx, Prescription.

Appendix A

Table 3. Frequency of prescribed medications, medical diagnoses, and psychiatric diagnoses for the full sample and pre- and posttreatment insomnia status groups. Full Sample (N = 68)

Medical Diagnoses Heart disease

Pretreatment No-Insomnia Symptoms (n = 15)

Pretreatment Insomnia Symptoms (n = 53)

Improved Persistent (n = 24) (n = 29) n (%) 26 (49.1%) 13 (54.2%) 13 (44.8%)

n (%) 27 (39.7%)

n (%) 1 (6.7%)

Pulmonary disease

7 (10.3%)

0 (0%)

7 (13.2%) 4 (16.7%) 3 (10.3%)

Gastrointestinal disorder

22 (32.4%)

3 (20.0%)

19 (35.8%) 10 (41.7%) 9 (31.0%)

Neurologic disorder

2 (2.9%)

1 (6.7%)

1 (1.9%) 0 (0%) 1 (3.4%)

Head trauma

4 (5.9%)

2 (13.3%)

2 (3.8%) 2 (8.3%) 0 (0%)

10 (14.7%)

2 (13.3%)

8 (15.1%) 7 (29.2%) 1 (3.4%)

Headache disorder

5 (7.4%)

2 (13.3%)

3 (5.7%) 1 (4.2%) 2 (6.9%)

Endocrine disorder

9 (13.2%)

0 (0%)

9 (17.0%) 4 (16.7%) 5 (17.2%)

Metabolic disorder

8 (11.8%)

1 (6.7%)

7 (13.2%) 5 (20.8%) 2 (6.9%)

Kidney disease

4 (5.9%)

0 (0%)

4 (7.5%) 1 (4.2%) 3 (10.3%)

Autoimmune disorder

3 (4.4%)

0 (0%)

3 (5.7%) 2 (8.3%) 1 (3.4%)

68 (100%)

15 (100%)

53 (100%) 24 (100%) 29 (100%)

Chronic pain disorder

Obstructive sleep apnea

Full Sample (N = 68)

Medical Diagnoses Periodic limb movement disorder

Pretreatment No-Insomnia Symptoms (n = 15)

Pretreatment Insomnia Symptoms (n = 53)

Improved Persistent (n = 24) (n = 29) n (%) 11 (20.8%) 5 (20.8%) 6 (20.7%)

n (%) 15 (22.1%)

n (%) 4 (26.7%)

3 (4.4%)

1 (6.6%)

2 (3.8%) 2 (8.3%) 0 (0%)

Psychiatric Diagnoses Unipolar depressive disorder

8 (11.8%)

1 (6.7%)

7 (13.2%) 3 (12.5%) 4 (13.8%)

Bipolar depressive disorder

4 (5.9%)

1 (6.7%)

3 (5.7%) 2 (8.3%) 1 (3.4%)

Anxiety disorder

3 (4.4%)

0 (0%)

3 (5.7%) 2 (8.3%) 1 (3.4%)

0 (0%)

0 (0%)

0 (0%)

Restless leg syndrome

Psychotic disorder

0 (0%)

0 (0%)

Cognitive disorder

5 (7.4%)

0 (0%)

5 (9.4%) 1 (4.2%) 4 (13.8%)

Prescription Medication Antidepressant

22 (32.4%)

4 (26.7%)

18 (34.0%) 8 (33.3%) 10 (34.5%)

Anxiolytic

4 (5.9%)

0 (0%)

4 (7.5%) 2 (8.3%) 2 (6.9%)

Antipsychotic

5 (7.4%)

1 (6.7%)

4 (7.5%) 3 (12.5%) 1 (3.4%)

Hypnotic/sedative

8 (11.8%)

4 (26.7%)

4 (7.5%) 2 (8.3%) 2 (6.9%)

Antihypertensive/antiarrhythmic

24 (35.3%)

1 (6.7%)

23 (43.4%) 13 (54.2%) 10 (34.5%)

Statin

15 (22.1%)

2 (13.3%)

13 (24.5%) 7 (29.2%) 6 (20.7%)

Thyroid agent

8 (11.8%)

0 (0%)

8 (15.1%) 2 (8.3%) 6 (20.7%)

Full Sample (N = 68)

Prescription Medication Antihyperglycemic agent

Pretreatment No-Insomnia Symptoms (n = 15)

Pretreatment Insomnia Symptoms (n = 53)

Improved Persistent (n = 24) (n = 29) n (%) 6 (11.3%) 4 (16.7%) 2 (6.9%)

n (%) 7 (10.3%)

n (%) 1 (6.7%)

Antiasthmatic

7 (10.3%)

0 (0%)

7 (13.2%) 4 (16.7%) 3 (10.3%)

Dopamine agonist

4 (5.9%)

1 (6.7%)

3 (5.7%) 2 (8.3%) 1 (3.4%)

Anticonvulsant

8 (11.8%)

1 (6.7%)

7 (13.2%) 4 (16.7%) 3 (10.3%)

Muscle relaxant

4 (5.9%)

1 (6.7%)

3 (5.7%) 2 (8.3%) 1 (3.4%)

Opiate

5 (7.4%)

1 (6.7%)

4 (7.5%) 4 (16.7%) 0 (0%)

13 (19.1%)

2 (13.3%)

11 (20.8%) 5 (20.8%) 6 (20.7%

Other headache medication

2 (2.9%)

1 (6.7%)

1 (1.9%) 1 (4.2%) 0 (0%)

Immunosuppressant

4 (5.9%)

0 (0%)

4 (7.5%) 3 (12.5%) 1 (3.4%)

0 (0%)

0 (0%)

Nonsteroidal anti-inflammatory

Acetylcholinesterase inhibitor

0 (0%) 0 (0%)

Stimulant

1 (1.5%)

1 (6.7%)

0 (0%) 0 (0%)

Hormone replacement

3 (4.4%)

Data are presented as number (% of sample)

1 (6.7%)

0 (0%) 0 (0%)

2 (3.8%) 2 (8.3%) 0 (0%)

Abbreviations AHI: Apnea-Hypopnea Index BMI: Body Mass Index CBT-I: Cognitive Behavioral Therapy for Insomnia ESS: Epworth Sleepiness Scale ISI: Insomnia Severity Index OSA: Obstructive Sleep Apnea PAP: Positive Airway Pressure PSG: Polysomnography RDI: Respiratory Disturbance Index RX: Prescription

Predictors and Patterns of Insomnia Symptoms in OSA Before and After PAP Therapy

Table 1. Demographic, OSA severity, PAP characteristics, medical and psychological status, and insomnia symptoms for the full sample and pretreatment insomnia symptom status groups. Full Sample (N = 68)

Pretreatment No-Insomnia Symptomsa (n = 15)

Pretreatment Insomnia Symptomsb (n = 53)

Z (p)c

32.4

33.3

32.1

--- (1.00)

Age (years)

47.5 (12.4)

46.7 (12.9)

47.8 (12.3)

-0.10 (.92)

Body Mass Index

32.2 (7.3)

32.9 (6.1)

32.0 (7.7)

-0.52 (.60)

Apnea Hypopnea Index

34.7 (32.2)

31.2 (32.6)

35.7 (32.3)

-0.82 (.41)

Respiratory Disturbance Index

47.2 (29.6)

43.1 (30.1)

48.4 (29.6)

-0.89 (.38)

Epworth Sleepiness Scale

12.1 (4.9)

10.60 (4.2)

12.6 (5.0)

-0.55 (.58)

286.2 (132.1)

244.5 (178.5)

297.9 (115.1)

-1.25 (.21)

PAP pressure

10.8 (2.9)

10.4 (2.0)

10.9 (3.1)

-0.42 (.68)

Total Number of Rx Medications

3.7 (3.80)

1.62 (2.3)

4.28 (3.9)

-2.43 (.02)

Total Number of Medical Conditions

2.9 (1.5)

2.1 (1.1)

3.1 (1.5)

-2.11 (.04)

Total Number of Psychiatric Diagnoses

0.3 (.53)

.1 (.3)

.4 (.6)

-2.00 (.05)

Pretreatment ISI Subscale Score

5.4 (2.8)

1.7 (1.3)

6.4 (2.2)

-.59 (<.001)d

Posttreatment ISI Subscale Score

3.5 (2.4)

2.7 (2.2)

3.7 (2.4)

-1.60 (.11)

-1.9 (3.1)#

1.1 (2.5)

-2.7 (2.8)#

-4.47 (<.001)

Gender, Female (%)

Average PAP Use (min./night)

ISI Subscale Change Score

Data are presented as mean (SD). aPretreatment ISI subscale score <4; bPretreatment ISI subscale score ≥ 4; c Non-parametric d

test for characteristics of pretreatment insomnia symptomatology versus non-symptomatic groups; Group selection variable, #

intentionally significant between-group difference, p < .001; Significant change in ISI subscale score from pre- to posttreatment, p < .001; OSA, obstructive sleep apnea; PAP, positive airway pressure.; Rx, Prescription; ISI, Insomnia Severity Index.

Legend Table 1. Demographic, OSA severity, PAP characteristics, medical and psychological status, and insomnia symptoms for the full sample and pretreatment insomnia symptom status groups. Data are presented as mean (SD). aPretreatment ISI Subscale score <4; bPretreatment ISI Subscale score ≥ 4; c Non-parametric test for characteristics of pretreatment insomnia symptomatology versus non-symptomatic groups; dGroup selection variable, intentionally significant between-group difference, p<.001; #Significant change in ISI Subscale score from preto posttreatment, p<.001; OSA, obstructive sleep apnea; PAP, positive airway pressure; Rx, Prescription; ISI, Insomnia Severity Index.

Predictors and Patterns of Insomnia Symptoms in OSA Before and After PAP Therapy

Table 2. ISI Subscale scores, Average PAP Use, RDI, and Medical Covariates for Persistent versus Improved Insomnia Groups. Improved

Persistent

Insomnia

Insomnia

Symptoms

Symptoms

(n = 24)

(n = 29)

Pretreatment ISI Subscalea

5.8 (1.8)†

6.9 (2.3)

-1.97 (.05) †

Posttreatment ISI Subscalea,b

1.5 (1.2)*

5.6 (1.4)*

-6.28 (<.001)†

337.1 (119.9)

265.5 (102.0)

-2.40 (.02) †

RDI

57.9 (29.1)

40.1 (28.1)

-2.45 (.01) †

Total Number of Rx Medications

4.79 (4.5)

4.3 (3.7)

-0.25 (.80)

Total Number of Medical Diagnoses

3.33 (1.7)

3.0 (1.4)

-0.57 (.57)

Total Number of Psychiatric Diagnoses

0.3 (0.6)

0.4 (0.6)

-0.21 (.84)

Average PAP Usec

z (p)

All data are presented as mean (SD). aPosttreatment occurred between 31 and 60 days post PAP setup; b

Group selection variable, intentionally significant between-group difference, p < .001; cMinutes per night;

†

Statistically significant between-group difference, p < .05; * Statistically significant within-group difference

pre- to posttreatment, p < .05; ISI, Insomnia Severity Index; PAP, positive airway pressure; RDI, respiratory disturbance index; Rx, Prescription.

Legend Table 2. All data are presented as mean (SD). aPosttreatment occurred between 31 and 60 days post PAP setup; bGroup selection variable, intentionally significant between-group difference, p < .001; cMinutes per night; †Statistically significant between-group difference, p < .05; * Statistically significant within-group difference pre- to posttreatment, p < .05; ISI, Insomnia Severity Index; PAP, positive airway pressure; RDI, respiratory disturbance index; Rx, Prescription.

Appendix A

Table 3. Frequency of prescribed medications, medical diagnoses, and psychiatric diagnoses for the full sample and pre- and posttreatment insomnia status groups. Full Sample (N = 68)

Medical Diagnoses Heart disease

Pretreatment No-Insomnia Symptoms (n = 15)

Pretreatment Insomnia Symptoms (n = 53)

Improved Persistent (n = 24) (n = 29) n (%) 26 (49.1%) 13 (54.2%) 13 (44.8%)

n (%) 27 (39.7%)

n (%) 1 (6.7%)

Pulmonary disease

7 (10.3%)

0 (0%)

7 (13.2%) 4 (16.7%) 3 (10.3%)

Gastrointestinal disorder

22 (32.4%)

3 (20.0%)

19 (35.8%) 10 (41.7%) 9 (31.0%)

Neurologic disorder

2 (2.9%)

1 (6.7%)

1 (1.9%) 0 (0%) 1 (3.4%)

Head trauma

4 (5.9%)

2 (13.3%)

2 (3.8%) 2 (8.3%) 0 (0%)

10 (14.7%)

2 (13.3%)

8 (15.1%) 7 (29.2%) 1 (3.4%)

Headache disorder

5 (7.4%)

2 (13.3%)

3 (5.7%) 1 (4.2%) 2 (6.9%)

Endocrine disorder

9 (13.2%)

0 (0%)

9 (17.0%) 4 (16.7%) 5 (17.2%)

Metabolic disorder

8 (11.8%)

1 (6.7%)

7 (13.2%) 5 (20.8%) 2 (6.9%)

Kidney disease

4 (5.9%)

0 (0%)

4 (7.5%) 1 (4.2%) 3 (10.3%)

Autoimmune disorder

3 (4.4%)

0 (0%)

3 (5.7%) 2 (8.3%) 1 (3.4%)

68 (100%)

15 (100%)

53 (100%) 24 (100%) 29 (100%)

Chronic pain disorder

Obstructive sleep apnea

Full Sample (N = 68)

Medical Diagnoses Periodic limb movement disorder

Pretreatment No-Insomnia Symptoms (n = 15)

Pretreatment Insomnia Symptoms (n = 53)

Improved Persistent (n = 24) (n = 29) n (%) 11 (20.8%) 5 (20.8%) 6 (20.7%)

n (%) 15 (22.1%)

n (%) 4 (26.7%)

3 (4.4%)

1 (6.6%)

2 (3.8%) 2 (8.3%) 0 (0%)

Psychiatric Diagnoses Unipolar depressive disorder

8 (11.8%)

1 (6.7%)

7 (13.2%) 3 (12.5%) 4 (13.8%)

Bipolar depressive disorder

4 (5.9%)

1 (6.7%)

3 (5.7%) 2 (8.3%) 1 (3.4%)

Anxiety disorder

3 (4.4%)

0 (0%)

3 (5.7%) 2 (8.3%) 1 (3.4%)

0 (0%)

0 (0%)

0 (0%)

Restless leg syndrome

Psychotic disorder

0 (0%)

0 (0%)

Cognitive disorder

5 (7.4%)

0 (0%)

5 (9.4%) 1 (4.2%) 4 (13.8%)

Prescription Medication Antidepressant

22 (32.4%)

4 (26.7%)

18 (34.0%) 8 (33.3%) 10 (34.5%)

Anxiolytic

4 (5.9%)

0 (0%)

4 (7.5%) 2 (8.3%) 2 (6.9%)

Antipsychotic

5 (7.4%)

1 (6.7%)

4 (7.5%) 3 (12.5%) 1 (3.4%)

Hypnotic/sedative

8 (11.8%)

4 (26.7%)

4 (7.5%) 2 (8.3%) 2 (6.9%)

Antihypertensive/antiarrhythmic

24 (35.3%)

1 (6.7%)

23 (43.4%) 13 (54.2%) 10 (34.5%)

Statin

15 (22.1%)

2 (13.3%)

13 (24.5%) 7 (29.2%) 6 (20.7%)

Thyroid agent

8 (11.8%)

0 (0%)

8 (15.1%) 2 (8.3%) 6 (20.7%)

Full Sample (N = 68)

Prescription Medication Antihyperglycemic agent

Pretreatment No-Insomnia Symptoms (n = 15)

Pretreatment Insomnia Symptoms (n = 53)

Improved Persistent (n = 24) (n = 29) n (%) 6 (11.3%) 4 (16.7%) 2 (6.9%)

n (%) 7 (10.3%)

n (%) 1 (6.7%)

Antiasthmatic

7 (10.3%)

0 (0%)

7 (13.2%) 4 (16.7%) 3 (10.3%)

Dopamine agonist

4 (5.9%)

1 (6.7%)

3 (5.7%) 2 (8.3%) 1 (3.4%)

Anticonvulsant

8 (11.8%)

1 (6.7%)

7 (13.2%) 4 (16.7%) 3 (10.3%)

Muscle relaxant

4 (5.9%)

1 (6.7%)

3 (5.7%) 2 (8.3%) 1 (3.4%)

Opiate

5 (7.4%)

1 (6.7%)

4 (7.5%) 4 (16.7%) 0 (0%)

13 (19.1%)

2 (13.3%)

11 (20.8%) 5 (20.8%) 6 (20.7%

Other headache medication

2 (2.9%)

1 (6.7%)

1 (1.9%) 1 (4.2%) 0 (0%)

Immunosuppressant

4 (5.9%)

0 (0%)

4 (7.5%) 3 (12.5%) 1 (3.4%)

0 (0%)

0 (0%)

Nonsteroidal anti-inflammatory

Acetylcholinesterase inhibitor

0 (0%) 0 (0%)

Stimulant

1 (1.5%)

1 (6.7%)

0 (0%) 0 (0%)

Hormone replacement

3 (4.4%)

Data are presented as number (% of sample)

1 (6.7%)

0 (0%) 0 (0%)

2 (3.8%) 2 (8.3%) 0 (0%)

Legend Table 3. Frequency of prescribed medications, medical diagnoses, and psychiatric diagnoses for the full sample and pre- and posttreatment insomnia status groups. Data are presented as number (% of sample).

Predictors and Patterns of Insomnia Symptoms in OSA Before and After PAP Therapy Highlights

•

Patterns of insomnia symptoms prior to and following PAP treatment of OSA are examined.

•

PAP adherence and OSA severity are strong predictors of insomnia improvement.

•

Resolution of insomnia symptoms following PAP is more likely with higher RDI.

•

Significant insomnia symptoms persist in 55% of this sample despite PAP.