The Evaluation and Diagnosis of “Insomnia” in Relation to Sleep Disturbance Prevalence and Impact in Early-Treated HIV-Infected Persons

To the Editor—We read with great interest the article entitled “Prevalence and Factors Associated With Sleep Disturbances Among Early-Treated HIV-Infected Persons,” recently published in Clinical Infectious Diseases, by Crum-Cianflone et al [1]. The study investigates the impact of sleep on various domains of daytime function in human immunodeficiency virus (HIV)–positive individuals, a well-recognized but understudied relationship. This study was a controlled study with a relatively large sample size. Compared to seronegative controls, HIV-positive individuals reported a high prevalence of insomnia, which was strongly associated with reduction in several quality-of-life domains. However, we would like to raise some issues with the methodological tools and interpretative assumptions presented in the study.

First of all, the authors established the diagnosis of insomnia solely on the basis of the Pittsburgh Sleep Quality Index (PSQI). The PSQI is a validated screening tool that identifies “good” and “poor” sleepers based on their responses to 7 categorical topics related to their sleep over the past 1 month [2]. In the Crum-Cianflone et al study, all subjects qualifying as poor sleepers based on the PSQI were then categorically labeled with the diagnosis of insomnia. However, according to the diagnostic criteria of the International Classification of Sleep Disorders, second edition (ICSD-2) [3], the diagnosis of primary insomnia includes the presence of sleep quality complaints (as defined by difficulty initiating, maintaining, or acquiring restorative sleep) that is also accompanied by associated functional complaints for at least 3 months. Secondly, the PSQI has been recognized as a validated screening tool for identifying poor sleepers and not as a diagnostic tool for identifying a specific sleep disorder etiology. Thus, it is sensitive and yet nonspecific for identifying patients who may suffer from a variety of sleep disorders including insomnia, sleep apnea, restless legs syndrome, circadian rhythm disturbances, and even behavioral insufficient sleep syndrome.

Other studies evaluating the presence and impact of insomnia in the HIV population have reported a prevalence ranging from 30% to 73% [4]. We believe that the large range in estimated prevalence among HIV-positive individuals may partially result from the variety of tools utilized to assess insomnia especially if these tools do not necessarily reflect the formal ICSD-2 diagnostic criteria established by the American Academy of Sleep Medicine. To underscore this point, we conducted an institutional review board–approved pilot study formally evaluating 25 HIV-infected subjects currently stable on combination antiretroviral therapy. All of the subjects underwent a formal clinical sleep assessment with a board-certified sleep specialist that included a validated clinical sleep interview (Structured Interview for Sleep Disorders [SIS-D]) [5] along with an overnight polysomnogram. The study included 13 black males (52%), 8 black females (32%), and 4 white males (16%); a representative sample in line with the demographic distribution infected with HIV according to the current Centers for Disease Control and Prevention reports [6]. In our study, HIV-positive subjects had a variety of sleep disorders. Eighty percent of the subjects fulfilled the criteria for at least 1 sleep disorder with some subjects fulfilling criteria for as many as 5 sleep disorders concurrently (See Table 1).

Table 1.

Distribution of Sleep Disorders Based on International Classification of Sleep Disorders, Second Edition Clinical Criteria and Polysomnography^a

Sex	Insomnia	Sleep Apnea Syndromea	Circadian Rhythm Disorder	Restless Legs Syndrome	Periodic Limb Movementsa of Sleepa
Female	6 (24%)	5 (20%)	1 (4%)	1 (4%)	0 (0%)
Male	7 (28%)	14 (24%)	1 (4%)	1 (4%)	5 (20%)

All data are No. (% of total).

^a The table will not sum to 100% as multiple disorders were present in the case of some subjects.

Future studies investigating the prevalence and impact of sleep disruption in HIV-positive cohorts should utilize more specific and objective diagnostic sleep disorder tools (eg, polysomnography, actigraphy, electronic sleep diaries) and validated clinical sleep assessments (eg, SIS-D [5], Morningness-Eveningness Questionnaire [7]). Moreover, validated functional questionnaires specifically related to sleep disorders (Functional Outcomes of Sleep Questionnaire, Insomnia Severity Index, Fatigue Severity Scale) would help to provide more direct correlation with sleep complaints and functional impact in this cohort as well.

As stated earlier, we applaud the Crum-Cianflone et al research team for conducting their study, which ultimately helps to draw additional attention to this prominent issue. However, in order to enhance our understanding of the unique relationship between sleep and overall function among HIV-positive individuals, future research should use more specific diagnostic tools to more accurately characterize the sleep complaints experienced in this cohort. Enhancing our knowledge of this relationship could help provide the foundation for the development of sleep disorder–specific practice guidelines and ultimately improvement in the quality of life, function, and disease management for this patient population.