The dispute began with a paper from the TURKAPNE Study Group, led by Canan Gündüz Gürkan of Süreyyapaşa Chest Diseases Research and Training Hospital in Istanbul and Yüksel Peker, a pulmonary medicine specialist whose affiliations span Koç University in Istanbul, the University of Gothenburg, Lund University, and the University of Pittsburgh. Drawing on a nationwide Turkish cohort of adults evaluated for sleep-disordered breathing, the group reported that diabetes mellitus was associated with both an increased prevalence and an increased severity of a condition known by the acronym COMISA: co-morbid insomnia and sleep apnea. The finding landed with force because COMISA sits at the intersection of two of the most common chronic sleep disorders, and because patients who carry both conditions simultaneously are known to fare worse than those with either alone.

COMISA is not a niche concern. Epidemiological surveys across multiple countries suggest that a substantial fraction of patients presenting at sleep clinics with obstructive sleep apnea also report chronic insomnia symptoms, and conversely that many insomnia patients show objective evidence of upper airway obstruction during sleep. The clinical consequences of this overlap are well documented in the review literature: people with COMISA tend to report greater daytime sleepiness, more fatigue, worse quality of life, higher rates of depression and anxiety, and greater healthcare utilization than patients with a single disorder. Treatment is also trickier, because continuous positive airway pressure, the standard therapy for sleep apnea, can be poorly tolerated by patients whose insomnia makes it difficult to sleep while wearing a mask, and hypnotic medications used for insomnia can suppress arousal responses that protect apnea patients from prolonged airway collapse.

Into this landscape stepped Xu, Liu, and Liu, whose letter to the journal bore a title that framed the entire controversy: Clarifying COMISA in diabetes: distinguishing true insomnia comorbidity from OSA-related sleep disruption. Their argument, as the title suggests, was methodological and conceptual. They contended that the TURKAPNE analysis may have conflated two fundamentally different phenomena. On one side stands genuine insomnia disorder, a condition defined by persistent difficulty initiating or maintaining sleep despite adequate opportunity, accompanied by daytime impairment, and rooted in the hyperarousal mechanisms that modern insomnia research has mapped with increasing precision. On the other side stands something quite different: the fragmented, unrefreshing sleep that results when the upper airway repeatedly collapses during the night, triggering micro-awakenings that patients may perceive as insomnia.

This distinction matters because the two conditions demand different treatments and carry different prognoses. A patient whose sleep complaints are driven entirely by untreated apnea may find their insomnia-like symptoms dissolve once positive airway pressure therapy stabilizes their breathing. A patient with true co-morbid insomnia, by contrast, will typically continue to struggle with sleep even after their apnea is controlled, and will need cognitive behavioral therapy for insomnia or pharmacological treatment in addition to their breathing therapy. If researchers label apnea-related sleep fragmentation as insomnia, they inflate the apparent prevalence of COMISA, overestimate its burden in populations such as people with diabetes, and risk sending clinicians down the wrong therapeutic path.

The TURKAPNE investigators responded in a reply published on 13 August 2026, and their response defends the clinical reality of the association they reported. Peker and Gürkan, writing on behalf of the TURKAPNE Study Group, stand by the central claim of their original paper: that in their nationwide cohort, diabetes mellitus was associated with increased prevalence and severity of COMISA. The reply, published as a letter in volume 22 of the Journal of Clinical Sleep Medicine, represents the group’s formal answer to the critique, and it underscores how seriously the original authors take the challenge to their interpretation. The exchange is compact, but the stakes it illuminates are anything but.

Why would diabetes be expected to sit at the center of this storm in the first place? The physiological connections between type 2 diabetes and disordered sleep run in both directions. Obesity, the dominant risk factor for obstructive sleep apnea, is also the dominant driver of insulin resistance, so the two conditions travel together at the population level. But the links go deeper than shared adiposity. Intermittent hypoxia and sleep fragmentation from apnea activate sympathetic nervous system outflow and stress hormone release, worsen glucose tolerance, and promote insulin resistance, creating a plausible causal pathway from breathing disorder to metabolic disease. Conversely, diabetic autonomic neuropathy can destabilize the neural control of upper airway muscles and breathing rhythm, and diabetes-related nocturia and peripheral neuropathy can fragment sleep independently of any breathing problem. Insomnia, for its part, has been repeatedly linked to impaired glucose metabolism in large cohort studies, with chronic short sleep and poor sleep quality both associated with elevated diabetes risk.

This dense web of interconnection is precisely what makes the conceptual question raised by the letter writers so difficult and so important. In a patient with diabetes, poor sleep could reflect apnea-driven fragmentation, true insomnia with its hyperarousal physiology, painful neuropathy, nocturia from poor glycemic control, depression, medications, or some tangled combination of all of these. Disentangling those threads requires careful phenotyping: polysomnography to quantify apnea severity objectively, validated insomnia instruments that probe the subjective experience of sleep and its daytime consequences, and enough clinical detail to separate apnea-related awakenings from the difficulty falling asleep and staying asleep that characterizes insomnia disorder. Large registries and cohorts such as TURKAPNE, which has previously published cross-sectional analyses of more than twelve thousand adults with moderate-to-severe obstructive sleep apnea, offer the statistical power to detect associations, but their ability to resolve fine diagnostic distinctions depends entirely on the instruments used and the definitions applied.

The broader field has been converging on a more nuanced view of COMISA in recent years. Reviews of the co-morbid insomnia and sleep apnea literature, including influential work by Sweetman and colleagues, have emphasized that the overlap is real, that it is more common than chance would predict, and that it carries additive clinical burden. At the same time, researchers have acknowledged that the boundary between the two disorders is porous, that subjective insomnia symptoms can be produced or amplified by objective sleep disruption, and that the field still lacks a universally accepted algorithm for deciding when the two conditions are genuinely co-present. Some investigators have proposed sequential treatment trials, using the response to apnea therapy as a diagnostic probe: if insomnia symptoms persist after the breathing disorder is adequately treated, that persistence itself becomes evidence for true co-morbidity. Others advocate for parallel treatment pathways that address both conditions from the outset, arguing that waiting for diagnostic clarity delays relief for patients who are suffering.

What the exchange in the Journal of Clinical Sleep Medicine ultimately demonstrates is that a single word, comorbidity, can carry enormous weight. If the TURKAPNE finding survives the methodological scrutiny, it suggests that diabetes should be treated as a red flag for combined sleep pathology, prompting clinicians who evaluate diabetic patients with sleep complaints to look for both disorders rather than stopping at the first diagnosis. If the critics are right that part of the observed association reflects apnea-related sleep disruption masquerading as insomnia, then the clinical message shifts: the priority is aggressive detection and treatment of sleep apnea in diabetic patients, with insomnia reassessed only after breathing is stabilized. Either way, the practical takeaway for patients is the same. People with diabetes who struggle with sleep should not accept fragmented nights as an inevitable feature of their disease, because the causes are identifiable, the diagnostic tools exist, and effective treatments are available for both conditions. The scientific argument over definitions will continue in the letters pages, but its resolution will shape real decisions in sleep clinics, and for the growing global population living with diabetes, those decisions could determine whether restorative sleep remains out of reach or finally comes within grasp.

Subject of Research: Co-morbid insomnia and sleep apnea (COMISA) in patients with diabetes mellitus

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

Article References: Peker, Y., Gürkan, C. G., & on behalf of the TURKAPNE Study Group (2026). Reply to “Clarifying COMISA in diabetes: distinguishing true insomnia comorbidity from OSA-related sleep disruption”. Journal of Clinical Sleep Medicine, 22(1), Article 141. https://doi.org/10.1007/s44470-026-00158-9

Image Credits: AI Generated

DOI: 10.1007/s44470-026-00158-9

Keywords: COMISA, insomnia, obstructive sleep apnea, diabetes mellitus, type 2 diabetes, sleep medicine, TURKAPNE cohort, sleep fragmentation, CPAP therapy, cognitive behavioral therapy, sleep disorders, diagnostics