Prevalence of Nightmare Disorder in Psychiatric Outpatients

Nightmares can cause great suffering as they disturb sleep, affect daily functioning and overall well-being. They are associated with psychopathology like anxiety disorders, more specifically posttraumatic stress disorder (PTSD), psychosis, personality pathology, suicide risk, and substance abuse [1,2]. Two to five percent of the general adult population suffers from one or more nightmares a week [3,4]. One study of psychiatric patients with more severe psychopathology showed a prevalence rate of 39.8% of ‘bad dreams that disturbed sleep' [5]. However, this study did not specify the psychopathology nor describe how the bad dreams where measured.Studies concerning prevalence of nightmares in psychiatric populations that did specify psychopathology have reported on subsamples such as: PTSD 50-70% [1,2], depression 17.5% [6], insomnia 18.3% [7], schizophrenia 16.7% [6], and borderline personality disorder 49% [8]. These studies suggest a high prevalence of nightmares in a psychiatric population, regardless of the primary diagnosis. However, no study reported prevalence rates of nightmares across all psychiatric disorders. Moreover, the studies cannot be compared because of the different assessment methods that were used.In this study, we addressed the problem of different methods by assessing all consecutive patients admitted to a specialist mental health care trust situated in Almere (The Netherlands) over the course of a year and a half. Patients were referred by their family doctor. From October 2008 until April 2010, all patients (age 18-75 years) who enrolled for outpatient treatment (n = 962) were asked to complete a questionnaire regarding nightmares as part of the standard assessment procedure.Of the 962 patients, 91 patients could not participate for reasons such as acute psychiatric crisis or insufficient mastery of the Dutch language. Of the eligible 871 patients, 29 refused to participate in the research, 297 did not show up on the appointment for the standard assessment, and 34 had too many missing data. The DSM-IV-TR classifications were assessed by a clinician in a psychiatric interview. The primary diagnoses of the remaining 511 patients (58.7%) were: mood disorder (n = 83; 16.2%), anxiety disorder (n = 109; 21.3%), PTSD (n = 30; 5.9%), personality disorder (n = 161; 31.5%), psychotic disorder (n = 13; 2.5%), or other (n = 115; 22.5%). Most patients (n = 317; 62.0%) had more than one DSM-IV-TR classification; the classification with the most severe symptoms was designated as the ‘primary diagnosis'. None of the patients had a DSM-IV-TR classification of nightmare disorder. We included PTSD as a separate group because of its known association with nightmares [1,2]. Patients with a psychotic disorder had a disproportionately low response rate (19%) and were excluded from the analyses.The final sample of 498 patients had a mean age of 36.0 ± 11.8 and consisted of 359 (72.1%) women. Of the participants, 226 (45.4%) were married or cohabiting, 272 (54.6%) were single or alone. The level of education was: n = 47 (9.4%) no education or primary education only; n = 316 (63.5%) low to intermediate secondary or vocational education; n = 123 (24.7%) high secondary, pre-university education or higher education; n = 12 (2.4%) unknown. No differences were found on any of these sociodemographic variables between respondents and nonrespondents.To measure nightmare disorder, the nightmare subscale of the SLEEP-50 [9] was used, a questionnaire designed to detect sleep disorders as listed in the DSM-IV-TR. The SLEEP-50 has a sensitivity of 0.84 and a specificity of 0.77 for nightmares in relation to clinical diagnosis [9]. Patients were considered suffering from nightmare disorder if the disturbing dreams were present at least once a week; affected their well-being; caused awakening; patients remembered the content of the nightmare; patients could orientate themselves rapidly afterwards. Criterion D of the DSM-IV-TR (‘the nightmares do not occur exclusively during the course of another mental disorder') was not taken into account as we were interested in the prevalence of nightmares as a phenomenon in a psychiatric population.In this psychiatric sample, 149 patients (29.9%) suffered from nightmare disorder (table 1). No differences were found in age between patients with (mean = 35.0 ± 11.8) or without nightmare disorder (mean = 36.5 ± 11.8; F1, 496 = 1.72, p = 0.19). Women (n = 120; 33.4%) suffered more often from nightmare disorder than men (n = 29; 20.9%; χ2 = 7.45, d.f. = 1, p < 0.01). Use of psychiatric medication was equally distributed over patients with (n = 175; 50.3%) and without (n = 81; 54.4%; χ2 = 0.69, d.f. = 1, p = 0.41) nightmare disorder. In a logistic regression analysis, gender was also the only variable significantly associated with nightmare disorder, if medication (B = -0.25; SE = 0.20; p = 0.23), age (B = -0.01; SE = 0.01; p = 0.24), and gender (B = -0.62; SE = 0.24; p = 0.01) were entered simultaneously (χ2 = 11.05, d.f. = 4, p = 0.03).The observed 29.9% of the patients in this psychiatric sample that suffered from nightmare disorder is much higher than the prevalence of 2-5% found in the general population, either measured with the SLEEP-50 [4,9] or other questionnaires [3]. The data also support the findings from previous studies that nightmares are highly prevalent in psychiatric populations [5,6,7,8]. In PTSD patients, nightmare occurrence was 2.4 times higher than in the remaining sample (66.7 against 27.7%), but nightmares were highly prevalent across all disorders.It is remarkable that at the psychiatric assessment none of the patients was diagnosed with a DSM-IV-TR classification of nightmare disorder. Nightmare disorder might be ignored in mental health care because according to DSM-IV-TR criteria nightmares can only be classified as a nightmare disorder if they do not occur exclusively during the course of another mental disorder.A shortcoming of the study is that psychiatric assessments were used to determine the primary diagnosis and diagnostic categories in the sample, instead of a structured interview. However, this is a naturalistic study and the sample is representative of real-life clinical practice. Nightmares are highly prevalent in psychiatric practice and our data suggest they are often missed in psychiatric assessment. Even though DSM-IV-TR criteria for nightmare disorder might not be met completely, patients could benefit from effective symptom-based treatments such as exposure therapy or imagery rehearsal therapy [10]. Nightmares should be assessed in all psychiatric patients regardless of their primary diagnosis.The authors are grateful to all patients participating in this study, the colleagues at the Department of Psychodiagnostics at GGz Centraal de Meregaard, the staff of GGz Centraal de Meregaard and the research department at GGz Centraal Innova. We also thank Erik de Groot, Meinte Vollema, Henk Schut, Victor Spoormaker, and Willemijn Verkaar for their contributions and comments.