P1: JLSpljots2004.cls (03/27/2004 v1.1 LaTeX2e JOTS document class) pp1381-jots-495796 September 21, 2004 23:3FOR PROOFREADING ONLYJournal of Traumatic Stress, Vol. 17, No. 6, December 2004, pp. 477–484 (C2004)Clinical Correlates of Poor Sleep Qualityin Posttraumatic Stress DisorderAnne Germain,1,3Daniel J. Buysse,1M. Kathy Shear,1Rana Fayyad,2and Carol Austin2Sleep disturbances (SD) are a core clinical feature of PTSD. The goal of the study was to deter-mine the influence of patient-related characteristics, disorder-related characteristics, and psychiatriccomorbidity on the severity of SD in PTSD outpatients (n = 367) who were not recruited for asleep study. Increased severity of SD paralleled increasing overall PTSD severity. The severity ofSD did not differ according to gender, age groups, types of trauma, PTSD chronicity, or psychiatriccomorbidity. The severity of SD paralleled PTSD severity. Results suggest that age, gender, andpsychiatric comorbidity have minimal impact on sleep quality in this PTSD sample. The inclusion ofPTSD patients who were not specifically seeking treatment for SD reinforces the study findings.KEY WORDS: sleep; insomnia; posttraumatic stress; Pittsburgh Sleep Quality Index.Sleep disturbances (SD) are a core clinical feature ofPTSD and often complicate PTSD outcomes. In individ-uals with PTSD who seek specific treatments to alleviateSD, the severity of SD is comparable to or exceeds thelevels observed in other sleep-disordered and psychiatricsamples (Krakow, Germain, et al., 2001). Subjective SDand objectively measured sleep disruption occurring earlyafter trauma exposure predict the development of PTSD atfollow-up (Koren, Arnon, Lavie, & Klein, 2002; Mellman,Bustamante, Fins, Pigeon, & Nolan, 2002). In addition,PTSD patients who report significant SD also report moresubstance use and abuse, more severe health-related com-plaints, depression, and suicidality (Clum, Nishith, &Resick, 2001; Krakow, Artar, et al., 2000; Nishith, Resick,& Mueser, 2001; Saladin, Brady, Dansky, & Kilpatrick,1995). Interventions aimed at reducing SD are associ-ated with clinically significant improvements in overallPTSD symptoms, psychological well-being, and daytime1Department of Psychiatry, University of Pittsburgh School of Medicine,Pittsburgh, Pennsylvania.2Pfizer Inc., New York City, New York.3To whom correspondence should be addressed at Department of Psychi-atry, University of Pittsburgh School of Medicine, 3811 O’Hara Street,E-1124, Pittsburgh, Pennsylvania 15213; e-mail: [email protected] (e.g., Germain & Nielsen, 2003; Gillin et al.,2001; Krakow, Hollifield, et al., 2001; Neylan et al., 2003;Raskind et al., 2003). There is growing evidence that co-morbid sleep disorders including insomnia, nightmares,and sleep disordered breathing are frequent in a signifi-cant portion of PTSD patients (Krakow, Germain, et al.,2000; Krakow, Hollifield, et al., 2001; Krakow et al.,2002) who undergo standard diagnostic sleep evaluations.Together, these observations support the suggestion thatSD are a core clinical feature of PTSD, and raise thepossibility that adjunctive sleep-focused assessments andintervention strategies may facilitate care managementand enhance treatment response in PTSD. Thus, a com-prehensive characterization of the clinical correlates ofself-reported SD in PTSD may provide new directionsto help identify patient groups who may require formaldiagnostic sleep evaluations or who may be more likelyto benefit from sleep-focused clinical strategies.Sleep quality may be influenced by a variety ofpatient- and disorder-related characteristics such as age,gender, type of trauma, PTSD chronicity and severity,and psychiatric comorbidity. Thus, elucidating the influ-ence of these characteristics on the severity of subjectiveSD in PTSD may be an important step in the develop-ment of evaluation and intervention strategies for SD in4770894-9867/04/1200-0477/1C2004 Springer Science+Business Media, Inc.P1: JLSpljots2004.cls (03/27/2004 v1.1 LaTeX2e JOTS document class) pp1381-jots-495796 September 21, 2004 23:3478 Germain, Buysse, Shear, Fayyad, and AustinPTSD patients. Gender and age significantly influencesleep quality in both healthy and psychiatric samples(e.g., Bixler, Vgontzas, Lin, Vela-Bueno, & Kales, 2002;Buysse et al., 1991; Doi, Minowa, Uchiyama, & Okawa,2001). In general, women report more sleep complaintsthan men do, and older individuals report more sleep com-plaints that younger cohorts. Thus, age and gender mayprovide early identifiers of the need for adjunct sleep-focused interventions in PTD patients.It is plausible that disorder-related characteristics de-fined in this study as the type of trauma (e.g., sexual abuse,war exposure, motor vehicle accident, natural disaster),PTSD chronicity, and PTSD severity may also influencesleep quality in PTSD patients. To date, empirical evi-dence indicates that the severity of SD is elevated and ofcomparable magnitude in survivors of natural disasters,sexual abuse, and Holocaust, and war veterans with PTSD(e.g., Krakow, Germain, et al., 2000; Mellman, David,Kulick-Bell, Ashlock, & Nolan, 1995; Mellman, David,Kulick-Bell, Hebding, & Nolan, 1995; Rosen, Reynolds,Yeager, Houck, & Hurwitz, 1991). Poor sleep quality andsevere SD also appear to be independent of PTSD chronic-ity (e.g., Koren et al., 2002; Rosen et al., 1991; Schreuder,Kleijn, & Rooijmans, 2000). However, a direct compar-ison of sleep quality and severity of SD across traumatypes or PTSD chronicity using a validated measure ofsleep quality is not yet available. There is some evidencethat PTSD severity influences the severity of SD. In sex-ually abused women with PTSD who seek treatment forposttraumatic insomnia and nightmares, increased sever-ity of SD parallels increases in overall PTSD severity(Krakow, Hollifield, et al., 2001). The generalizability ofthese observations to men and women with PTSD whodo not specifically seek help for SD, however, remainsuncertain.Psychiatric comorbidity may also influence sleepquality in PTSD. As many as 88% of men and 79% ofwomen with a lifetime history of PTSD exhibit anotherpsychiatric disorder (Kessler, Sonnega, Bromet, Hughes,& Nelson, 1995). The most common comorbid Axis Idisorders in PTSD are substance use disorders, mooddisorders, and anxiety disorders (Creamer, Burgess, &McFarlane, 2001), which are all characterized by pooroverall sleep quality and SD (Breslau, Roth, Rosenthal,& Andreski, 1996; Ford