Report of the DSM-V Sleep-Wake Disorders Work Group
November 2008
Charles F. Reynolds III, M.D.
Our Sleep-Wake Disorders Work Group is reviewing each of the approximately 25-30 DSM-IV sleep and circadian rhythm disorders (text and criteria), conducting focused literature reviews, and working with advisors who have specific expertise for a given disorder (and who have been, in most cases, among the primary contributors to the peer-reviewed literature for each disorder). We expect to complete our initial review of all disorders by December 2008. We have recently begun updating text to reflect scientific advances in related basic science, epidemiology, imaging, genetics, and interventions research. Throughout, we have tried to integrate into our conversations important aspects of diagnostic spectra and developmental perspectives. We also continue to benefit from input from liaisons from other work groups.
The work group’s plan for 2009 is to continue the process of updating text, continuing to call upon advisors who are already active collaborators in this process. We expect this process to lead to an update of diagnostic criteria and to specification of scientifically appropriate and clinically useful dimensional measures of nocturnal sleep quality and of daytime alertness/impairment. (We expect to prepare a memorandum to other workgroup chairs on this topic.)
We have identified a small number of clinically useful and scientifically valid measures of sleep quality and daytime alertness/impairment. We are working with Patient Reported Outcomes Measurement Information Systems (PROMIS) investigators who have used Item Response Theory to identify a small number of informative and practicable items that convey accurately the severity, impairment, and distress occasioned by disturbed sleep and wakefulness. We think of these measures as a “sleep thermometer,” that is, highly sensitive to sleep/wake “fever” (read: “disturbance”), but non-specific as to etiology. An analysis of PROMIS measures shows potential usefulness in creation of dimensional measures. Further pilot testing is needed, however, especially in persons with mental disorders. (We intend to convey this in our memorandum on sleep disturbance to other work group chairs.) Of existing dimensional measures, the Pittsburgh Sleep Quality Index has been the most widely used and cited over the past two decades. (It should perhaps have been named the “Pittsburgh Sleep Wake Quality Index” or the “Pittsburgh Sleep Wake Disturbance Index” to convey more accurately its content.) While its performance characteristics are well known, including sensitivity to change in sleep-wake disturbance during treatment of sleep and mental disorders, its “utility” may be limited by the number of items (21). Useful PROMIS items, as of this writing, total about 15 and probably can be reduced further.
To date, there appears to be a preliminary, growing consensus on several potential recommendations to the DSM-V Task Force. With emphasis on the words “preliminary” and “potential”, these are as follows:
1) DSM-IV “primary insomnia,” while not rare in epidemiologic studies, is not nearly as prevalent as insomnia comorbid with other mental or medical disorders.
Note: the work group recommends the term “comorbid” rather than “secondary to” in order to be consistent with the 2006 recommendations of an NIH consensus conference on insomnia, because “comorbid” carries fewer, unwarranted causal assumptions than “secondary to.” Work group members are working with advisors on these issues.
2) DSM-IV “primary hypersomnia” lumps together a rather heterogeneous group of disorders, where little scientific progress has been made since DSM-IV. Nonetheless, it should be possible to further clarify this heterogeneity in the text to make this area less confusing to clinicians. In contrast, considerable progress has occurred in the field of narcolepsy cataplexy research that will mandate updating of text and possibly updating of diagnostic criteria as well. Work group members are collaborating with advisors on these issues and tasks. One issue under discussion is whether specific polysomnographic (psg) criteria should be integrated with clinical diagnostic criteria, or whether psg measures should simply be addressed in the text. (DSM-IV did not allow the inclusion of quantitative psg criteria, but the science of sleep disorders makes the revisiting of this issue mandatory.)
3) DSM-IV “breathing-related sleep disorder” text and criteria will require updating to reflect considerable scientific (basic, epidemiologic) and clinical progress in this area. An important consideration is to have scientific accuracy while still retaining accessibility to the non-sleep expert in both specialty mental health and primary care medicine. Work group members are working with advisors to address these issues and tasks. As with narcolepsy cataplexy and other disorders of excessive sleepiness, the work group is discussing whether specific psg criteria should be integrated with clinical criteria or simply explicated in the text.
4) DSM-IV “circadian rhythm sleep disorders” rely on the default “not otherwise specified” (NOS) to a greater extent than would now be warranted by scientific advances both basic (i.e., genetic) and clinical. Thus, it seems likely that further differentiation of, for example, “advanced sleep phase syndrome” (rather than subsuming it under “NOS”) may be recommended. Work group members are working with advisors on these issues and tasks as well.
5) DSM-IV placed “restless legs syndrome” (RLS) under “dyssomnia not otherwise specified.” However, in light of considerable scientific progress in basic (genetic) and clinical research (clinical phenotype and interventions), the work group is considering the evidence base for “elevating” RLS to its own specific diagnosis rather than subsuming it under a NOS grouping. RLS provides a useful example of how progress in genetics, definition of clinical and physiological phenotype, and controlled interventions research may warrant greater diagnostic specificity in DSM-V and less reliance on “NOS”. Work group members are working with advisors on these issues and tasks.
6) The work group continues to integrate discussion of the International Classification of Sleep Disorders-II (ICSD-II) into our conversation; we plan to incorporate links to ICSD-II in the text, that is, show how DSM-IV and ICSD-II map onto each other. The same is true with respect to the ICD-9. The process of mapping is helped by the fact that many of our work group members and advisors were active contributors to the ICSD-II.
The work group continues to analyze how best to integrate categorical diagnostic and dimensional measures (e.g., severity, distress, impairment, and sensitivity to treatment associated changes). A simple approach would be to juxtapose categorical and dimensional “boxes” above the text and then use the text to justify and explain the choice of each type of measure and explicate to the non-sleep expert how and when to use the categorical measures (categorical, polysomnographic, and dimensional).