Comorbid Insomnia and Obstructive Sleep Apnea in Military Personnel: Correlation With Polysomnographic Variables

Transcription

1 MILITARY MEDICINE, 179, 3:294, 2014 Comorbid Insomnia and Obstructive Sleep Apnea in Military Personnel: Correlation With Polysomnographic Variables LTC Vincent Mysliwiec, MC USA*; Panagiotis Matsangas, PhD ; Tristin Baxter, AAS*; Leigh McGraw, PhD*; MAJ Nici E. Bothwell, MC USA*; Bernard J. Roth, MD* ABSTRACT Objectives: Military personnel undergoing polysomnography are typically diagnosed only with obstructive sleep apnea (OSA). Comorbid insomnia with OSA is a well-established, underappreciated diagnosis. We sought to determine if military personnel with mild OSA met clinical criteria for insomnia and if there was a pattern of polysomnogram (PSG) variables that identified insomnia in these patients. Methods: Retrospective chart review of military personnel with mild OSA; cluster analysis to describe PSG variables. Results: 206 personnel assessed, predominately male (96.6%), mean age 36.5 ± 8.14 years, body mass index 30.2 ± 3.66 kg/m 2 and apnea hypopnea index of 8.44 ± 2.92 per hour; 167 (81.1%) met criteria for insomnia. Cluster analysis identified a group of patients (N = 52) with PSG variables of increased wakefulness after sleep onset 77.3 minutes (27.7) (p < 0.001) and decreased sleep efficiency 82.6% (5.82) (p < 0.001) consistent with insomnia. Patients in this group were more likely to meet criteria for insomnia with an odds ratio 5.27 (1.20, 23.1), (p = 0.009). Conclusions: The majority of military personnel with mild OSA meet criteria for insomnia. Roughly one-third of these patients can be identified by a pattern of PSG variables. Recognizing and treating both comorbid insomnia and OSA could improve clinical outcomes. INTRODUCTION Comorbid insomnia with obstructive sleep apnea (OSA) is a well-recognized yet underappreciated clinical entity. 1,2 Depending on the criteria used for insomnia, the reported prevalence of this disorder ranges from 22% to as high as 55% and is more common in females. 2 8 Patients with insomnia may receive medical therapy alone when they also have OSA and thus not receive adequate therapy for their sleep disorders. 2,9 OSA may be underdiagnosed since patients who present with insomnia symptoms often do not receive a polysomnogram (PSG) as part of their evaluation for insomnia. 10 Conversely, there is objective data for the diagnosis of OSA, by virtue of the apnea hypopnea index (AHI), whereas insomnia remains a clinical diagnosis. Military personnel frequently report sleep disturbances with a prevalence as high as 80%. 11,12 Their sleep is short in duration with previous reports of 6 hours or less sleep per night. 13,14 The etiology of their sleep disturbances may result from service-related illnesses of anxiety, depression, posttraumatic stress disorder (PTSD) 15 and mild traumatic brain injury (TBI) 16 all of which are associated with insomnia. Recent reports of sleep disorders in military personnel have shown a high rate of OSA. 17,18 Our study of a large cohort of military personnel undergoing PSG reported the most common diagnoses of OSA in 51.2% and insomnia in 30%. 15 This protocol assigned only one major sleep disorder *Department of Pulmonary, Sleep Medicine, Critical Care, Madigan Army Medical Center, 9040A Fitzsimmons Avenue, Tacoma, WA Operations Research Department, Naval Postgraduate School, 1411 Cunningham Road, Monterey, CA The opinions and assertions in this article are those of the authors and do not necessarily represent those of the Department of the Army, the Department of Defense, or the U.S. Government. doi: /MILMED-D to each patient, which was OSA if the AHI was greater than or equal to 5. In a subsequent observational study, we identified a high rate of comorbid insomnia and OSA (38.2%) in a select group of military personnel (110 subjects) who were referred for sleep difficulties. 19 With this study, we sought to determine if comorbid insomnia and OSA was similarly prevalent in our larger cohort of military personnel with a PSG diagnosis of mild OSA. Once OSA, defined by an AHI ³ 5 on the PSG report, is diagnosed, many clinicians treat this disorder alone. Insomnia is only considered if the response to therapy is inadequate. However, characteristic PSG variables consistent with insomnia could suggest the diagnosis of comorbid insomnia and OSA before a trial of OSA therapy. 20 These variables are sleep onset latency (SOL) ³ 31 minutes, wakefulness after sleep onset (WASO) ³ 31 minutes, and sleep efficiency (SE) < 85%, although such quantitative criteria are not typically incorporated into PSG diagnostic interpretations. Determining PSG characteristics of patients with comorbid insomnia and OSA may improve recognition of this often underdiagnosed clinical entity. Better recognition may facilitate earlier diagnosis and treatment, likely resulting in improved clinical outcomes. We hypothesized that there is a high incidence of insomnia in military personnel diagnosed with mild OSA. We also sought to identify PSG variables that could help identify patients with comorbid insomnia and OSA as opposed to OSA alone using cluster analysis. METHODS Study Participants This retrospective study was approved by the Institutional Review Board at Madigan Army Medical Center in Tacoma, Washington. From January 1, 2010 to December 31, 2010, 294

2 207 U.S. Army, Air Force, and Navy personnel underwent a PSG and were diagnosed with mild OSA using criteria of AHI ³ 5and<15. One record was missing PSG variables that were analyzed in the cluster analysis and was excluded from the analysis for a total sample size of 206. This group of patients with mild OSA was included in our initial study on sleep disorders in military personnel which assessed the relationship between sleep disorders and service-related illnesses. 15 Polysomnography The overnight PSG was conducted at our medical facility using standardized techniques (Polysmith 5.0, Neurotronics, Gainesville, Florida) with 16 channels including: electrooculogram, electroencephalogram, electrocardiogram, electromyogram (submental, bilateral tibial), air flow measurements using both oronasal thermal sensors and nasal air pressure transducers, tracheal sounds via microphone, rib cage and abdominal movement by inductance plethysmography using thoracoabdominal belts, and continuous pulse oximetry. Polysomnograms were scored by a registered polysomnography technician in accordance with American Academy of Sleep Medicine standards. 21 Our laboratory uses the American Academy of Sleep Medicine alternative scoring method for hypopneas requiring a ³50% decrease in the nasal pressure signal excursion from baseline lasting 10 seconds or more with either a ³3% desaturation or an arousal. 22 Study Variables Biometric parameters of age, gender and body mass index (BMI) and service-related illnesses of anxiety, depression, PTSD, and mild TBI were obtained through medical record review using the Armed Forces Health Longitudinal Technology Application system, the military electronic medical health record (EMR). Active medications of antidepressants, antipsychotics, benzodiazepines, nonbenzodiazepine receptor agonists (NBDRAs), narcotics, and prazosin were obtained from the reported medications the night of the PSG. Data elements were recorded in a deidentified database before statistical analysis. PSG variables used in the analysis included SOL, rapid eye movement (REM) latency, total sleep time (TST), SE, sleep stages (stage N1, stage N2, stage N3, stage R), WASO, arousal index, AHI, and maximal desaturation. Self-reported variables collected included home sleep duration, PSG sleep duration, and Epworth Sleepiness Scale (ESS). Diagnosis of Insomnia Each patient s EMR and sleep questionnaire were assessed for self-reported symptoms of difficulty initiating sleep, maintaining sleep, waking too early, or sleep that is chronically nonrestorative or poor in quality with at least one of the associated symptoms of daytime impairment that are part of the diagnostic criteria of insomnia in accordance with International Classification of Sleep Disorders, 2nd Edition (ICSD-2) criteria. 22 In all cases, a determination was made if the patient met, did not meet, or there were insufficient data to ascertain the diagnosis of insomnia. It is relevant to point out that nearly all military personnel with sleep disturbances undergoing a sleep medicine evaluation have a PSG because of the high-risk nature of their profession; this includes patients who predominantly have insomnia symptoms. Statistics Statistical analyses were carried out with a statistical software package (JMP Pro 9; SAS Institute; Cary, North Carolina). First, all variables were tested for normality using the Shapiro Wilks test. Based on these results, the decision was made to use nonparametric tests for data analysis. Then, study variables underwent descriptive statistical analysis to describe our population in terms of demographic characteristics of military personnel diagnosed with mild OSA, and PSG variables of military personnel with comorbid insomnia and mild OSA. Clustering analysis was used to classify patients according to their PSG variables, SOL, REM latency, TST, SE, sleep stages (stage N1, stage N2, stage N3, stage R), WASO, arousal index, and maximal desaturation. Clustering is a multivariate technique used in exploratory data analysis to classify data in a number of user-defined groups that share similar attributes (homogeneous groups). 23,24 Clustering herein was implemented using the K-Means method with three groups. Comparisons betweengroupsarebasedonwilcoxonrank-sumtest.for each comparison, we also report the corresponding effect size using Cohen s d. Effect size indicates the strength of the difference between two means without any assumptions regarding the underlying distributions. Contingency analysis was used to evaluate an insomnia diagnosis and the groups derived from clusters analysis as appropriate. Data are presented as mean (standard deviation [SD]) or as mean/ median (MD) with range. RESULTS Our population was predominately male (96.6%) with a mean age of 36.5 years, BMI of 30.2, and AHI of 8.4/hour (Table I). The highest prevalence of a service-related illness was depression (21.9%), with 24.3% having 2 or more service-related illnesses. Of the 206 mild OSA patients, 195 had adequate clinical data in their EMR and sleep questionnaires to determine if they met ICSD-2 criteria for insomnia; 167 met diagnostic criteria for insomnia and were classified as having comorbid insomnia and OSA (Table II), whereas 28 were negative and thus categorized as mild OSA alone. Group comparisons were made between patients with comorbid insomnia and OSA and mild OSA alone. Comorbid 295

3 TABLE I. Characteristics of Military Personnel Diagnosed With Mild OSA Mild OSA (N = 206) Demographics Age, Mean (SD) 36.5 (8.14) Male Sex, % (No.) 96.6 (199) BMI, Mean (SD) 30.2 (3.66) Previously Deployed, % (No.) 85.4 (176) Self-Reported Measures ESS, Mean (SD) 12.5 (5.06) Home Sleep Duration, Mean (SD) 5.7 (1.42) Sleep 5 Hours, % (No.) 47.0 (95) Medications % (No.) Prazosin 1.5 (3) Pain Meds 20.9 (43) Benzodiazepines 1.9 (4) Antipsychotics 0.5 (1) Antidepressants 16.5 (34) NBDRAs 6.8 (14) Associated Illnesses % (No.) Anxiety 18.0 (37) Depression 21.9 (45) PTSD 9.7 (20) Mild TBI 14.6 (30) 2 or More 24.3 (50) insomnia and OSA, when compared with mild OSA, showed significant differences in PSG variables to include increased WASO (p < 0.001) and SOL (p < 0.01) and decreased TST (p < 0.001), SE (p < 0.001), and stage R (p < 0.05) as shown in Table III. Evaluating service-related illnesses, all patients with anxiety (n = 37) were diagnosed with insomnia. Depression was also more frequent in patients with insomnia (1-sided Fisher s exact test, p = 0.027; Odds ratio (OR) = 4.23 [ ]). Next we assessed the association between service-related illnesses and PSG variables consistent with insomnia (Table IV). Each service-related illness (anxiety, depression, PTSD, and mild TBI) was evaluated for a positive or negative screen in each of the 3 PSG variables consistent with insomnia, SOL ³ 31 minutes, WASO ³ 31 minutes, and SE < 85%. These results show that patients with a WASO ³ 31 minutes diagnosed with mild OSA are more likely to have a diagnosis of anxiety (OR 2.49 [ ]). TABLE II. Insomnia Symptoms in Patients With Comorbid Insomnia and OSA Diagnosis Comorbid Insomnia and OSA, N = 167, % (No.) Difficulty Initiating Sleep 71.3 (119) Difficulty Maintaining Sleep 82.6 (138) Waking Up too Early 0.6 (1) Chronically Nonrestorative/Poor Quality Sleep 93.4 (156) Chronically Nonrestorative/Poor Quality Sleep 4.2 (7) in Isolation Cluster Analysis Comparing Malahanobi s distances against critical c 2 values at p < level, we did not identify any multivariate outliers. Military personnel diagnosed with mild OSA (N = 206) wereclassifiedintothreeclustersbasedonpsgvariables (Table V). Cluster 1 was marked by a mildly prolonged SOL, an elevated WASO, and decreased SE (consistent with mild initiation and maintenance insomnia), Cluster 2 was marked by a very prolonged SOL as well as an elevated WASO with marked decrease in SE (consistent with severe sleep initiation and sleep maintenance insomnia), and Cluster 3 had relatively normal variables except for those associated with OSA. Significant differences existed between Cluster 1 and Cluster 3; Cluster 2 was too small for significant analyses. The AHI in patients in Clusters 1 and 3 were similar. However, patients in Cluster 1 had significantly decreased TST and SE at 6.3 (0.7) hours and 82.6% (5.8%) respectively, with an elevated WASO at 77.3 (27.7) minutes. Although the BMIs did not differ between clusters, individuals in Cluster 1 were significantly older at 41.2 (8.3) years than those in Cluster 3 at 34.8 (7.5) years. Furthermore, the occurrence of anxiety was less frequent in patients in Cluster 3 compared to Cluster 1 (2-sided Fisher s exact test, P = 0.059; OR [ ]). Lastly, we assessed the association between the diagnosis of insomnia and the clusters derived from the application of the 3-group clustering method. This analysis focused on Cluster 1 (n=52) and Cluster 3 (n=150). Contingency analysis showed that patients with a clinical diagnosis of insomnia are more likely to be included in Cluster 1, the mild OSA cluster with PSG variables consistent with insomnia (1-sided Fisher s exact test, p = 0.009; OR 5.27 [ ]). DISCUSSION The results from our study are unique as it shows that military personnel diagnosed with mild OSA have a high prevalence of insomnia, 81.1%. This is higher than most civilian studies on comorbid insomnia and OSA 4 6 but similar to a study by Krakow et al where 40 of 44 crime victims with PTSD were diagnosed with insomnia and sleep-disordered breathing. 25 It is possible that the higher prevalence in military personnel is attributable to a number of factors to include the persistence of maladaptive sleep practices from combat deployments, sleep deprivation, and fragmentation, which is known to exacerbate sleep disordered breathing 26,27 and comorbid service-related disorders (anxiety, depression, PTSD, and mild TBI). 15,16,28 The interaction between insomnia and OSA promotes a greater severity of illness, both in terms of sleep disorders as well as increased medical and psychiatric morbidity. 29 In nonmilitary populations, anxiety, stress, and depression were all elevated in comorbid insomnia and OSA patients as compared to OSA alone. 6 This is consistent with the clinical characteristics of our cohort, noting that 24.3% 296

4 TABLE III. Polysomnographic Variables and Demographic Characteristics of Mild OSA Patients With and Without Insomnia Without Insomnia (N = 28) With Insomnia (N = 167) c 2 * p Value a Cohen s d PSG Variables, Mean (SD) SOL (Minutes) 5.36 (5.41) 12.1 (15.6) 8.3 < REM latency (Minutes) 82.0 (41.6) 113 (65.3) TST (Hour) 7.72 (0.569) 7.04 (0.941) 13.5 < SE (%) 95.5 (2.65) 90.1 (8.43) 15.5 < N1 (% TST) 9.14 (5.00) 9.58 (5.00) N2 (% TST) 48.4 (9.88) 46.4 (10.8) N3 (% TST) 18.3 (9.17) 17.3 (7.86) Stage REM (% TST) 19.9 (6.28) 17.1 (5.47) 5.25 < WASO (Minutes) 21.9 (12.9) 44.3 (36.6) 13.8 < Arousal Index (Events/hr) 18.6 (7.06) 20.9 (8.79) AHI (Events/hr) 9.05 (2.77) 8.40 (2.90) Desaturation (%) 86.3 (4.04) 86.0 (3.95) Demographic Characteristics, Mean (SD) Age 34.0 (7.33) 36.9 (8.28) Male Sex, % BMI 30.7 (3.55) 30.2 (3.69) ESS 11.3 (4.94) 12.8 (5.07) Self-Reported-Polysomnogram Sleep Duration (Hour) 5.26 (1.42) 5.32 (1.68) Self-Reported-Home Sleep Duration (Hour) 6.80 (1.22) 5.61 (1.34) 16.4 < a Wilcoxon rank-sum test analysis between mild OSA without insomnia and mild OSA with insomnia. had two or more service-related illnesses and those with comorbid insomnia and OSA were more likely to have anxiety and depression. To our knowledge, previous studies have used a PSG to diagnose OSA and clinical criteria for insomnia except for Smith et al who required either a SOL or WASO greater than 30 minutes as part of the diagnostic criteria for insomnia. 6 The use of cluster analysis allowed us to identify a pattern of PSG variables, patients in Cluster 1, which even in the setting of mild OSA are consistent with the comorbid diagnosis of insomnia. The two most predictive variables for our population were a decreased SE and increased WASO. Prior studies utilizing PSG in the diagnosis of insomnia alone have reported a high specificity for a SOL (³31 minutes), WASO (³31 minutes), and SE (<85%). 20,30,31 In previous studies of comorbid insomnia and OSA that reported PSG variables, Krakow reported a SOL of 65 minutes, WASO of 63 minutes (calculated from reported variables), and SE of 75% 3 ; Smith reported a SOL of 20.5 minutes, WASO of 34.3 minutes, and SE of 60.2% 6 ; Yang reported a SOL of 17.2 minutes and SE of 83.9%. 32 In the reports by Krakow and Smith, these variables indicated patients with comorbid insomnia and OSA had more severe sleep disturbances than patients with OSA alone; whereas, in the study by Yang, there was no significant difference between patients with OSA and comorbid insomnia and OSA. Determining which patients in our cohort have maintenance insomnia as a consequence of OSA or the more complex diagnosis of comorbid insomnia and OSA cannot be definitively ascertained in this retrospective analysis. TABLE IV. Differences in Associated Illnesses Between SOL, WASO, and SE Comorbid Anxiety (N = 37) Depression (N = 45) PTSD (N = 20) mtbi (N = 30) SOL, Minutes <31 Minutes 17.6 (33) 22.9 (43) 9.57 (18) 14.4 (27) ³31 Minutes 22.2 (4) 11.1 (2) 11.1 (2) 16.7 (3) OR 1.34 ( ) 0.42 ( ) 1.18 ( ) 1.19 ( ) WASO, Minutes <31 Minutes 11.5 (12) 22.1 (23) 9.62 (10) 18.3 (19) ³31 Minutes 24.5 (25) 21.6 (22) 9.80 (10) 10.8 (11) OR 2.49 ( ) a 0.97 ( ) 1.02 ( ) 0.54 ( ) SE, % <85% 25.7 (9) 17.1 (6) 11.4 (4) 11.4 (4) ³85% 16.4 (28) 22.8 (39) 9.36 (16) 15.2 (26) OR 0.57 ( ) 1.43 ( ) 0.80 ( ) 1.39 ( ) Data are presented as % (No.). anxiety diagnosis. a OR suggests that patients with increased WASO (³31 minutes) demonstrate an increased likelihood of an 297

5 TABLE V. Polysomnographic Variables and Demographic Characteristics of Mild OSA Cluster-Derived Groups Cluster 1 (N = 52) Cluster 2 (N = 4) Cluster 3 (N = 150) c 2 p Value a Cohen s d PSG Variables, Mean (SD) SOL (Minutes) 16.1 (14.5) 46.0 (45.6) 8.61 (10.8) 15.6 < REM latency (Minutes) 140 (87.5) 198 (103) 96.2 (41.9) 4.97 < TST (Hour) 6.28 (0.698) 3.83 (0.735) 7.53 (0.526) 89.6 < SE (%) 82.6 (5.82) 56.8 (4.32) 94.7 (2.82) 112 < N1 (% TST) 12.9 (6.14) 14.3 (4.73) 8.23 (3.77) 25.4 < N2 (% TST) 40.2 (8.19) 24.1 (6.88) 49.6 (9.72) 35.1 < N3 (% TST) 15.7 (7.54) 11.6 (9.28) 18.1 (8.16) Stage REM (% TST) 14.3 (5.44) 6.80 (1.18) 18.8 (5.05) 27.8 < WASO (Minutes) 77.3 (27.7) 177 (41.3) 24.9 (13.5) 103 < Arousal Index (Events/hr) 24.9 (8.92) 33.5 (14.3) 18.6 (7.34) 20.5 < AHI (Events/hr) 8.67 (3.78) 10.6 (1.52) 8.30 (2.76) Desaturation (%) 86.4 (3.87) 85.0 (2.82) 85.7 (4.39) Demographic Characteristics, Mean (SD) Age 41.2 (8.30) 41.8 (2.50) 34.8 (7.50) 21.9 < Male Sex, % OR = 4.08 ( ) BMI 30.0 (3.15) 32.6 (4.18) 30.2 (3.81) ESS 12.0 (5.64) 12.8 (7.93) 12.7 (4.79) Self-Reported-Polysomnogram Sleep Duration (Hour) 4.80 (1.66) 4.67 (1.16) 5.57 (1.61) 8.86 < Self-Reported-Home Sleep Duration (Hour) 5.59 (1.39) 5.63 (1.38) 5.86 (1.38) a Wilcoxon rank-sum test analysis between Cluster 1 and Cluster 3. Notably, maintenance insomnia, whereby sleep-disordered breathing results in awakenings and insomnia symptoms, is viewed differently than initiation insomnia which is potentially a distinct disorder. 4,33 Integrating the patient s comorbid disorders, symptoms, and PSG variables could help in determining the difference between OSA that causes insomnia symptoms or the complex insomnia model postulated by Krakow. 34 There were significant differences in Cluster 1 versus Cluster 3 in SOL and WASO, consistent with difficulties initiating and maintaining sleep; regarding servicerelated illnesses, those in Cluster 3 were less likely to have a diagnosis of anxiety. Thus, the military personnel in Cluster 1 likely manifested complex insomnia. The patients in Cluster 3, with relatively normal PSG parameters, likely had less severe insomnia or one of the following forms of insomnia: psychophysiologic insomnia, where their sleep improved in the sleep lab; paradoxical insomnia in that they perceive sleeping little or none at all when in fact their sleep by PSG is relatively normal; or insomnia as a result of maladaptive sleep practices, 26 which in military personnel include excessive caffeine intake, prolonged work hours, and the near universal behavior of clock watching during their sleep period. 35 The importance of recognizing insomnia in military personnel with OSA is suggested by studies that show improved compliance with continuous positive airway pressure (CPAP) after receiving sedative hypnotics as well as improved quality of PSG after sedative pretreatment in military populations. 36,37 It is currently unknown what is optimal treatment in patients with comorbid insomnia and OSA. Sedatives alone are likely not appropriate treatment. 9,38 CPAP alone, without addressing the insomnia symptoms, will also likely fail. 39 A combined cognitive behavioral therapy and CPAP approach has shown benefit, but not in a randomized or concurrent fashion. 40 Clearly, the first step is improved diagnostic accuracy for this complex condition which to date has relied on integrating clinical and polysomnographic established diagnoses. There are a number of limitations to our study which merit discussion. Because of the retrospective nature of our study, we relied on review of clinical records to determine the diagnosis of insomnia. It is possible that this resulted in an elevated prevalence of insomnia diagnoses; further, we were unable to determine if the insomnia was secondary to servicerelated illnesses and thus could not determine the actual prevalence of comorbid insomnia and OSA in accordance with the research diagnostic criteria for insomnia. 41 The ICSD-2 criteria for insomnia include sleep that is chronically nonrestorative or poor in quality which could be a symptom of their OSA alone and not due to comorbid insomnia. Notably, there were only 7 patients (Table II) who solely had this criterion for their diagnosis; removing them from our analysis would not detract from our conclusion that there is a high rate of comorbid insomnia in military personnel with mild OSA. The severity of SDB was determined solely by the patient s AHI and did not account for respiratory event related arousals. It is possible that the increased arousal index in Cluster 1 was attributable to respiratory events that were not scored; however, as the actual increase was only 6 arousals per hour, the patients in Cluster 1 would still be classified with mild OSA. We only assessed patients with mild OSA. This severity of disease was chosen as it is the most frequent level of disease seen in a relatively young active duty 298

6 population and it is unclear if their clinical presentation resulted from the combined effects of insomnia, short sleep duration, or other service-related disorders. The population studied is entirely active duty military personnel with relatively uniform clinical characteristics that may differ from civilian patients and thus the PSG characteristics of the clusters may not be applicable to civilian populations. In conclusion, our study provides further insight into the diagnosis of comorbid insomnia and OSA, which is increasingly recognized in civilians and military personnel. Insomnia remains a clinical diagnosis, although there are quantifiable PSG variables that can aid clinicians in having a higher suspicion for comorbid insomnia and OSA. Given the difficulties in making this diagnosis, a combined approach utilizing both clinical criteria as well as PSG variables could help differentiate those who have a complex disease process from OSA with insomnia symptoms. Further studies are required to better determine optimal diagnostic criteria and treatment for comorbid insomnia and OSA. ACKNOWLEDGMENTS The authors would like to thank Angela Mysliwiec, MD, Madigan Army Medical Center, for editing assistance and review of the manuscript. Dr. Angela Mysliwiec did not receive compensation for her contributions. This study was supported, in part, by grant no from the Center for Neuroscience and Regenerative Medicine. REFERENCES 1. Guilleminault C, Eldridge FL, Dement WC: Insomnia with sleep apnea: a new syndrome. Science 1973; 181(4102): Al-Jawder SE, Bahammam AS: Comorbid insomnia in sleep-related breathing disorders: an under-recognized association. Sleep Breath 2012; 16(2): Krakow B, Melendrez D, Ferreira E, et al: Prevalence of insomnia symptoms in patients with sleep-disordered breathing. Chest 2001; 120(6): Chung KF: Insomnia subtypes and their relationships to daytime sleepiness in patients with obstructive sleep apnea. Respiration 2005; 72(5): Krell SB, Kapur VK: Insomnia complaints in patients evaluated for obstructive sleep apnea. Sleep Breath 2005; 9(3): 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(5): Wickwire EM, Collop NA: Insomnia and sleep-related breathing disorders. Chest 2010; 137(6): Lavie P: Insomnia: more questions need answers. J Sleep Res 2010; 19(1 Pt 1): Krakow B, Ulibarri VA, Romero E: Persistent insomnia in chronic hypnotic users presenting to a sleep medical center: a retrospective chart review of 137 consecutive patients. J Nerv Ment Dis 2010; 198(10): Littner M, Hirshkowitz M, Kramer M, et al: Practice parameters for using polysomnography to evaluate insomnia: an update. Sleep 2003; 26(6): Hoge CW, McGurk D, Thomas JL, Cox AL, Engel CC, Castro CA: Mild traumatic brain injury in U.S. Soldiers returning from Iraq. N Engl J Med 2008; 358(5): Hoge CW, Terhakopian A, Castro CA, Messer SC, Engel CC: Association of posttraumatic stress disorder with somatic symptoms, health care visits, and absenteeism among Iraq war veterans. Am J Psychiatry 2007; 164(1): Luxton DD, Greenburg D, Ryan J, Niven A, Wheeler G, Mysliwiec V: Prevalence and impact of short sleep duration in redeployed OIF soldiers. Sleep 2011; 34(9): Seelig AD, Jacobson IG, Smith B, et al: Sleep patterns before, during, and after deployment to Iraq and Afghanistan. Sleep 2010; 33(12): Mysliwiec V, McGraw L, Pierce R, Smith P, Trapp B, Roth BJ: Sleep disorders and associated medical comorbidities in active duty military personnel. Sleep 2013; 36(2): Collen J, Orr N, Lettieri CJ, Carter K, Holley AB: Sleep disturbances among soldiers with combat related traumatic brain injury. Chest 2012; 142(3): Lettieri CJ, Eliasson AH, Andrada T, Khramtsov A, Raphaelson M, Kristo DA: Obstructive sleep apnea syndrome: are we missing an at-risk population? J Clin Sleep Med 2005; 1(4): Gellis LA, Gehrman PR, Mavandadi S, Oslin DW: Predictors of sleep disturbances in Operation Iraqi Freedom/Operation Enduring Freedom veterans reporting a trauma. Mil Med 2010; 175(8): Mysliwiec V, Gill J, Lee H, et al: Sleep disorders in US military personnel: a high rate of comorbid insomnia and obstructive sleep apnea. Chest 2013; 144(2): Okun ML, Kravitz HM, Sowers MF, Moul DE, Buysse DJ, Hall M: Psychometric evaluation of the Insomnia Symptom Questionnaire: a self-report measure to identify chronic insomnia. J Clin Sleep Med 2009; 5(1): Iber C, Ancoli-Israel S., Chesson AL Jr, Quan SF: The AASM Manual for the Scoring of Sleep and Associated Events. Westchester, IL, American Academy of Sleep Medicine, American Academy of Sleep Medicine: International Classification of Sleep Disorders: Diagnostic and Coding Manual, Ed 2. Westchester, IL, American Academy of Sleep Medicine, Hauri PJ: A cluster analysis of insomnia. Sleep 1983; 6(4): Tan PN, Steinbach M, Kumar V: Cluster Analysis: Basic Concepts and Algorithms. Introduction to Data Mining, Ed 1, pp Boston, MA, Pearson Addison-Wesley, Krakow B, Melendrez D, Pedersen B, et al: Complex insomnia: insomnia and sleep-disordered breathing in a consecutive series of crime victims with nightmares and PTSD. Biol Psychiatry 2001; 49(11): Miller NL, Shattuck LG, Matsangas P: Sleep and fatigue issues in continuous operations: a survey of U.S. Army officers. Behav Sleep Med 2011; 9(1): Peterson AL, Goodie JL, Satterfield WA, Brim WL: Sleep disturbance during military deployment. Mil Med 2008; 173(3): Hoge CW, Castro CA, Messer SC, McGurk D, Cotting DI, Koffman RL: Combat duty in Iraq and Afghanistan, mental health problems, and barriers to care. N Engl J Med 2004; 351(1): Bjornsdottir E, Janson C, Gislason T, et al: Insomnia in untreated sleep apnea patients compared to controls. J Sleep Res 2012; 21(2): Lichstein KL, Durrence HH, Taylor DJ, Bush AJ, Riedel BW: Quantitative criteria for insomnia. Behav Res Ther 2003; 41(4): Lineberger MD, Carney CE, Edinger JD, Means MK: Defining insomnia: quantitative criteria for insomnia severity and frequency. Sleep 2006; 29(4): 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(4): Ong JC, Crisostomo MI: The more the merrier? Working towards multidisciplinary management of obstructive sleep apnea and comorbid insomnia. J Clin Psychol 2013; 69(10): Krakow B, Melendrez D, Warner TD, Dorin R, Harper R, Hollifield M: To breathe, perchance to sleep: sleep-disordered breathing and chronic insomnia among trauma survivors. Sleep Breath 2002; 6(4):

7 35. Centers for Disease Control and Prevention: Energy drink consumption and its association with sleep problems among U.S. service members on a combat deployment Afghanistan, MMWR Morb Mortal Wkly Rep 2012; 61(44): Lettieri CJ, Quast TN, Eliasson AH, Andrada T: Eszopiclone improves overnight polysomnography and continuous positive airway pressure titration: a prospective, randomized, placebo-controlled trial. Sleep 2008; 31(9): Lettieri CJ, Shah AA, Holley AB, Kelly WF, Chang AS, Roop SA: Effects of a short course of eszopiclone on continuous positive airway pressure adherence: a randomized trial. Ann Intern Med 2009; 151(10): Mysliwiec V, Roth B: Pharmacotherapy refractory insomnia in soldiers with traumatic brain injury. Chest 2013; 143(2): Wickwire EM, Smith MT, Birnbaum S, Collop NA: Sleep maintenance insomnia complaints predict poor CPAP adherence: a clinical case series. Sleep Med 2010; 11(8): Krakow B, Melendrez D, Lee SA, Warner TD, Clark JO, Sklar D: Refractory insomnia and sleep-disordered breathing: a pilot study. Sleep Breath 2004; 8(1): Edinger JD, Bonnet MH, Bootzin RR, et al: Derivation of research diagnostic criteria for insomnia: report of an American Academy of Sleep Medicine Work Group. Sleep 2004; 27(8):