OSA in Psychiatric Patients: Insomnia as a Concurrent Presentation of Autonomic Dysregulation in OSA

In this issue, McCall1 has commented that their study² was incorrectly excluded from our systematic review,³ based on our assessment that polysomnography (PSG) was not used to diagnose obstructive sleep apnea (OSA) in their depressed participants, when in fact PSG had been used to diagnose OSA in their final study sample. The goal of our systematic review was to examine the prevalence of OSA diagnosed by PSG in individuals with schizophrenia and other psychotic disorders, mood disorders, and anxiety disorders.³ Ideally, the goal was to examine prevalence in psychiatric populations where all individuals included in the study had a clinical psychiatric diagnosis and were assessed for the OSA using PSG. Our rationale for excluding the McCall study,² where participants with symptoms of OSA were excluded via interview without PSG, was more complex than a simple table in the appendix could convey and was the subject of significant debate between the authors. In many ways, this demonstrates the subjective problem of integrating many studies with diverse aims into the specific aims of a systematic review. McCall is correct that a subset of the population studied in their article included patients with MDD and insomnia with OSA diagnosed by PSG, which does fit the letter of our inclusion criteria.^2,3 However, in the spirit of examining OSA prevalence in the full psychiatric population, the rationale for excluding the study was based on the two-part screening procedure. McCall et al. conducted an initial telephone interview, which was used to exclude potential participants with an existing diagnosis or strong signs of OSA without a PSG.² This is a study design that makes perfect sense for identifying ideal participants to study the risks of OSA in low-risk MDD patients with insomnia; however, we felt it was problematic, based on the aims of our study, to exclude patients with existing or probable OSA without PSG. It was our conclusion that the active exclusion of known individuals with OSA biased the study in such a way as to make the data inappropriate for inclusion in a study on global prevalence. This was a subjective decision and in hindsight, we should have erred on the side of inclusivity with our screening process and discussed the screening process as a potential source of sample bias.

We are glad that this review has opened a discussion on the importance of assessing treatment-refractory psychiatric patients for sleep disorders using PSG. McCall et al. provide strong evidence that OSA can be present in even the lowest risk MDD patients.² The findings of McCall et al. also suggest that insomnia and OSA are not necessarily independent conditions, especially in patients with autonomic dysregulation at the primary central nervous level, such as in some psychiatric patients. Different levels of autonomic nervous system activation and resultant hyperarousal and sleep fragmentation could result in varying levels of insomnia in OSA patients with psychiatric disorders. It is therefore possible that levels of insomnia in the psychiatric patient with OSA vary over time, depending upon the severity of the psychiatric disorder and level of autonomic activation. This underscores the importance of considering sleep fragmentation as a principal component of the neurological, metabolic, and biologic factors contributing to severity and treatment resistance in psychiatric disorders. We hope that this information will spur the increased use of PSG as a part of the psychiatric diagnostic toolkit.

DISCLOSURE STATEMENT

The authors have indicated no financial conflicts of interest.

CITATION

Gupta MA, Simpson FC. OSA in psychiatric patients: insomnia as a concurrent presentation of autonomic dysregulation in OSA. J Clin Sleep Med 2015;11(5):585.