When Diabetes Meets Broken Sleep: Untangling True Insomnia from Hidden Apnea

Millions of people with diabetes report nights spent staring at the ceiling, and for years clinicians have lumped those complaints under a single convenient label: comorbid insomnia and sleep apnea, known in the sleep-medicine world as COMISA. A new letter to the editor published in the Journal of Clinical Sleep Medicine argues that this convenience has become a liability. Shali Xu of the Department of Nephrology and Diabetes at Gansu Provincial Hospital of Traditional Chinese Medicine, together with Xiaoxia Liu and Chunyu Liu, contends that in patients with diabetes, many cases labeled as COMISA may not represent genuine insomnia at all. Instead, they may reflect sleep fragmentation caused by obstructive sleep apnea, a breathing disorder that masquerades as insomnia when the person experiencing it cannot perceive the awakenings that repeatedly interrupt their night.

The distinction matters because the two conditions demand different treatments. True insomnia is a disorder of hyperarousal, characterized by difficulty falling or staying asleep despite adequate opportunity, and it responds best to cognitive behavioral therapy for insomnia, or CBT-I, and, in some cases, carefully selected hypnotic medications. Obstructive sleep apnea, by contrast, is a mechanical and physiological problem: the upper airway collapses during sleep, oxygen levels dip, and the brain jolts the body awake just enough to reopen the airway. Its first-line treatment is positive airway pressure therapy. If a clinician mistakes apnea-driven fragmentation for insomnia and prescribes a sedative-hypnotic, the underlying breathing disorder may worsen, since hypnotics can suppress the arousal responses and upper-airway muscle tone that protect a vulnerable airway during sleep.

The letter responds directly to a recent analysis of the TURKAPNE cohort, a nationwide Turkish sleep study led by Gürkan and colleagues, which reported that diabetes mellitus is associated with increased prevalence and severity of COMISA. That finding, published in the same journal, seemed to confirm what many clinicians suspected: diabetes and the double burden of insomnia plus apnea travel together. But Xu and colleagues raise a methodological objection that cuts to the heart of how COMISA is diagnosed. Standard insomnia questionnaires, such as the Insomnia Severity Index, ask about difficulty falling asleep, frequent awakenings, and unrefreshing sleep. A patient with severe, untreated sleep apnea will answer yes to nearly all of those questions, not because their brain has forgotten how to sleep, but because their airway keeps waking them up.

This perceptual blind spot is one of the most counterintuitive findings in modern sleep science. Most people with obstructive sleep apnea do not remember the dozens or hundreds of brief arousals that fragment their night. Each arousal lasts only a few seconds, just long enough to restore airflow, and the sleeper typically falls back asleep without forming any memory of the event. What they remember is the consequence: a morning feeling of unrefreshing sleep, daytime fatigue, and the impression that they are an insomniac. When such a patient fills out an insomnia questionnaire, the answers faithfully report the symptom but misattribute the cause. In a population with high rates of obesity and apnea, such as people with type 2 diabetes, that misattribution can inflate the apparent prevalence of true insomnia comorbidity considerably.

The technical solution the letter points toward is objective sleep measurement. Polysomnography, the gold-standard sleep study, records brain waves, eye movements, muscle tone, airflow, respiratory effort, and blood oxygen saturation across a full night. It can quantify the apnea-hypopnea index, the number of breathing interruptions per hour of sleep, and it can reveal whether reported insomnia complaints correlate with objective sleep fragmentation from respiratory events. Studies of Asian patients with comorbid insomnia and sleep apnea, such as the 2025 analysis by Hoc and Lee published in Scientific Reports, have used polysomnographic characteristics to begin separating phenotypes: patients whose insomnia complaints track with respiratory-event-related arousals versus those with genuine difficulty initiating or maintaining sleep independent of breathing disturbances.

The stakes extend well beyond diagnostic bookkeeping. Large epidemiological studies have linked both sleep-disordered breathing and insomnia to cardiometabolic outcomes. The Hispanic Community Health Study/Study of Latinos, analyzed by Li and colleagues in the American Journal of Respiratory and Critical Care Medicine in 2021, found associations between sleep-disordered breathing, insomnia, and incident hypertension and diabetes. If apnea contributes to the development and worsening of diabetes through intermittent hypoxia, sympathetic activation, and impaired glucose metabolism, then misclassifying apnea as insomnia in diabetic patients does more than delay proper treatment; it obscures a potentially modifiable driver of the metabolic disease itself. Treating the apnea could, in principle, improve both sleep and glycemic control, whereas treating a phantom insomnia cannot.

The pharmacological dimension adds further urgency. A 2024 systematic review and meta-analysis by Messineo, Sands, and Labarca in the American Journal of Respiratory and Critical Care Medicine examined how hypnotic medications affect obstructive sleep apnea severity and endotypes. The evidence suggests that the effects of hypnotics on apnea are not uniform: some agents may worsen the disorder by prolonging apneas or blunting arousals, while the impact varies with the patient’s specific apnea endotype, meaning the anatomical and physiological pattern that produces their airway collapse. Prescribing a hypnotic to a diabetic patient whose insomnia symptoms are actually driven by unrecognized apnea therefore carries a real risk of aggravating the very condition that is destroying their sleep, potentially deepening both the sleep fragmentation and the metabolic consequences.

Xu and colleagues propose that studies examining COMISA in diabetes should incorporate objective respiratory monitoring before concluding that insomnia is genuinely comorbid. At minimum, they argue, the diagnostic pathway for a diabetic patient reporting chronic insomnia should include screening for sleep apnea, whether through validated questionnaires such as the STOP-BANG, home sleep apnea testing, or full polysomnography in complex cases. Only after excluding or adequately treating significant sleep-disordered breathing can a clinician confidently diagnose a true insomnia comorbidity and initiate CBT-I or hypnotic therapy on solid ground. This sequencing, they suggest, would sharpen both research definitions and everyday clinical practice, replacing a symptom-based label with a mechanism-based diagnosis.

The letter, which was received on 3 July 2026, accepted on 9 July, and published on 12 August 2026 as article 138 in volume 22 of the journal, was supported by the Gansu Provincial Natural Science Foundation under grant number 25JRRA266. The authors, who work across the departments of nephrology and diabetes, geriatrics, and rehabilitation medicine at Gansu Provincial Hospital of Traditional Chinese Medicine in Lanzhou, declare no competing interests. Their argument is deliberately modest in form but consequential in scope: it does not overturn the TURKAPNE finding, but it reframes it, asking whether the excess COMISA observed in diabetes reflects a true double sleep disorder or an apnea epidemic wearing an insomnia mask.

For the rapidly growing global population of people living with diabetes, the practical message is clear. Chronic insomnia complaints in this group should trigger a search for breathing-related sleep disruption before a diagnosis of comorbid insomnia is accepted and sedatives are prescribed. The tools to make that distinction already exist, from low-cost home sleep testing to full laboratory polysomnography, and the treatments for each condition are fundamentally different. As sleep researchers continue to refine the phenotypes of COMISA, the letter by Xu, Liu, and Liu serves as a reminder that in medicine, as in sleep itself, what appears on the surface is not always what is happening underneath. Getting the diagnosis right may be the first step toward a night of genuinely restorative sleep, and perhaps toward better metabolic health as well.

Subject of Research: Distinguishing true insomnia comorbidity from obstructive sleep apnea-related sleep disruption in patients with diabetes

Article Title: Clarifying COMISA in diabetes: distinguishing true insomnia comorbidity from OSA-related sleep disruption

Article References: Xu, S., Liu, X., & Liu, C. (2026). Clarifying COMISA in diabetes: distinguishing true insomnia comorbidity from OSA-related sleep disruption. Journal of Clinical Sleep Medicine, 22(1), Article 138. https://doi.org/10.1007/s44470-026-00154-z

Image Credits: AI Generated

DOI: 10.1007/s44470-026-00154-z

Keywords: COMISA, insomnia, obstructive sleep apnea, diabetes, sleep medicine, polysomnography, CBT-I, hypnotics, TURKAPNE cohort, sleep fragmentation, type 2 diabetes, sleep-disordered breathing