Diabetes Linked to More Frequent and Severe Combined Insomnia and Sleep Apnea

People living with diabetes may be fighting a hidden second battle every night. A sweeping analysis of more than 12,700 adults referred for sleep testing in Türkiye has found that diabetes mellitus is independently associated with a higher prevalence of comorbid insomnia and obstructive sleep apnea, a combined sleep disorder known by the acronym COMISA. The study, drawn from the nationwide multicenter TURKAPNE cohort and published in the Journal of Clinical Sleep Medicine, also reveals that among patients with this dual disorder, those with diabetes suffer measurably worse sleep fragmentation and more severe breathing disturbances than their non-diabetic counterparts. The findings add weight to a growing body of evidence that sleep and metabolic disease are locked in a mutually reinforcing relationship, and they suggest that millions of people with diabetes may be going undiagnosed for a treatable condition that could be worsening their metabolic health.

COMISA has emerged in recent years as one of the most clinically consequential phenotypes in sleep medicine. Obstructive sleep apnea, in which the upper airway repeatedly collapses during sleep and triggers intermittent drops in blood oxygen, affects an estimated 9 to 38 percent of the general population. Insomnia, characterized by difficulty falling or staying asleep with daytime consequences, is similarly widespread. What has startled researchers is how often the two overlap: roughly 30 to 50 percent of patients with sleep apnea also report chronic insomnia symptoms, and 30 to 40 percent of chronic insomnia patients show objective evidence of sleep-disordered breathing. Crucially, the combination appears to be more than the sum of its parts. Prior studies have linked COMISA to elevated risks of cardiovascular disease and incident diabetes, exceeding the risk conferred by either disorder alone, particularly when the physiological burden is measured objectively rather than by symptoms alone.

To probe the relationship between diabetes and this dual sleep disorder, investigators turned to TURKAPNE, a prospective national registry that has enrolled adults aged 18 to 80 with suspected sleep apnea from 34 centers across Türkiye since 2017. For this cross-sectional analysis, the team included 12,715 patients who underwent full-night, in-laboratory polysomnography, the gold-standard sleep test that monitors brain waves, muscle tone, eye movements, airflow, respiratory effort, and oxygen saturation. Obstructive sleep apnea was defined as an apnea-hypopnea index of at least five events per hour, while insomnia was identified through standardized questions covering difficulty initiating sleep, sleep latency of 30 minutes or longer, trouble maintaining sleep, or use of hypnotic medications. Diabetes status was based on self-reported physician diagnosis or the use of antidiabetic medications. Participants using positive airway pressure therapy or mandibular advancement devices, and those with limited life expectancy, were excluded to keep the baseline measurements clean.

The headline numbers are striking. COMISA was present in 3,275 patients, or 25.7 percent of the cohort. Diabetes was significantly more common among those with COMISA than without it, at 22.8 percent versus 15.4 percent. Flipping the perspective, roughly one in three diabetic individuals in the cohort, 33.9 percent, met criteria for COMISA, compared with 24.1 percent of non-diabetic participants. After adjusting for age, sex, body mass index, smoking status, education level, and comorbidities, diabetes remained independently associated with COMISA, with an adjusted odds ratio of 1.17. While that figure may appear modest, in a cohort of this size and in a condition as prevalent as COMISA, it translates into a substantial population-level burden, and it held firm even after accounting for obesity, the most obvious shared risk factor.

The polysomnographic comparisons within the COMISA group tell an even more compelling story about physiological severity. Diabetic patients slept shorter overall, with a median total sleep time of 353.5 minutes versus 369.0 minutes in non-diabetics, and spent more time awake after sleep onset, 50.2 minutes versus 37.8 minutes. Their arousal index, a measure of how often the brain briefly surfaces from sleep, was 18.1 events per hour compared with 14.8. Their breathing was also markedly worse: the apnea-hypopnea index reached 30.0 events per hour versus 24.0, and the oxygen desaturation index, which counts how often blood oxygen dips each hour, hit 26.0 versus 20.0. Diabetic patients also spent more time with oxygen saturation below 90 percent and reached lower nadir saturations. Sleep architecture shifted toward lighter N1 sleep with reduced rapid eye movement sleep, a pattern associated with impaired glucose metabolism in experimental studies.

Why would diabetes and COMISA feed into each other? The proposed mechanisms run in both directions. Sleep apnea drives metabolic dysfunction through intermittent hypoxia, sympathetic nervous system activation, oxidative stress, and systemic inflammation, all of which impair insulin sensitivity and pancreatic beta-cell function. Insomnia, meanwhile, is thought to act through chronic physiological hyperarousal and dysregulation of the hypothalamic-pituitary-adrenal axis, the body’s central stress system, which elevates cortisol and disrupts glucose homeostasis. When the two disorders coexist, these pathways may amplify one another: the hyperarousal of insomnia can intensify the hypoxic and inflammatory insult delivered by apneic events, deepening insulin resistance. Conversely, diabetes itself may sabotage sleep through nocturia, peripheral neuropathy, autonomic dysfunction, and altered thermoregulation, creating a self-perpetuating cycle in which metabolic derangement and sleep disruption each worsen the other.

One intriguing detail from the data supports this bidirectional reading. Among diabetic patients with COMISA, difficulty maintaining sleep was more prevalent than difficulty initiating sleep, at 50.8 percent versus 44.0 percent for trouble falling asleep. Sleep-maintenance insomnia, waking repeatedly in the middle of the night, is precisely the symptom pattern one would expect from nocturnal hypoglycemia risk, medication effects, and the urge to urinate that accompanies poorly controlled blood sugar. The diabetic COMISA patients were also more frequently female and more often obese, and they carried heavier loads of hypertension, coronary artery disease, obstructive airway disease, hypothyroidism, and psychiatric disorders. Beyond diabetes, the regression analysis identified female sex, psychiatric disease, obstructive airway disease, smoking, lower education level, and higher body mass index as independent correlates of COMISA, sketching a portrait of a phenotype that clusters with social and medical vulnerability.

The sex difference deserves particular attention. Women with sleep apnea are known to present atypically, complaining of insomnia, fatigue, and mood disturbance rather than the classic snoring and witnessed apneas that raise suspicion in men, a pattern that contributes to systematic underdiagnosis of sleep-disordered breathing in women. Hormonal changes around menopause, including declines in progesterone and estrogen, may simultaneously increase upper airway collapsibility and insomnia prevalence. Because diabetic patients with COMISA in this cohort were more often female, clinicians evaluating women with diabetes who report poor sleep should consider that both conditions may be present and that neither should be dismissed as mere stress or menopausal discomfort.

The clinical implications are straightforward, though implementing them will require a shift in routine practice. The authors argue that individuals with diabetes should be screened systematically for both insomnia symptoms and sleep-disordered breathing, rather than waiting for one or the other to declare itself. Early identification of COMISA opens the door to integrated treatment, pairing positive airway pressure therapy for the apnea with cognitive behavioral therapy for insomnia, the first-line behavioral treatment that avoids the pitfalls of long-term hypnotic medication. Whether treating COMISA aggressively improves glycemic control and reduces cardiometabolic risk in diabetic patients remains to be tested in intervention trials, but the severity gradient documented here makes the hypothesis plausible and urgent.

The study has limitations worth keeping in view. Diabetes and comorbidities were identified by self-report and medication use rather than laboratory confirmation, insomnia was assessed with symptom questions rather than full diagnostic criteria, and details on diabetes duration, glycemic control, and specific drug classes were unavailable. Because the design is cross-sectional, causality cannot be established; the data show association, not direction. The cohort was also exclusively Turkish, which may limit generalizability. Still, with 12,715 polysomnographically characterized patients, the statistical signal is hard to dismiss. If roughly a third of people with diabetes carry a combined sleep disorder that worsens both their nights and their metabolism, then the sleep clinic and the diabetes clinic may need to stop operating as separate worlds, and the humble question about sleep quality may become one of the cheapest diagnostic tools in metabolic medicine.

Subject of Research: The association between diabetes mellitus and comorbid insomnia and obstructive sleep apnea (COMISA)

Article Title: Diabetes mellitus is associated with increased prevalence and severity of COMISA: evidence from the nationwide TURKAPNE cohort

Article References: Gürkan, C. G., Pihtili, A., Kiyan, E., Tasbakan, M. S., Basoglu, Ö. K., Arbatli, S., Aydin, S., Cilli, A., Dursunoglu, N., Baran, B., Peker, Y., Turkapne Study Group, Ursavaş, A., Annakkaya, A. N., Aksakal, A., Balcı, A., Utkusavaş, A., Ugurlu, A. Ö., Elmalı, A. D., … Zeren Uçar, Z. (2026). Diabetes mellitus is associated with increased prevalence and severity of COMISA: evidence from the nationwide TURKAPNE cohort. Journal of Clinical Sleep Medicine, 22(1), Article 103. https://doi.org/10.1007/s44470-026-00109-4

Image Credits: AI Generated

DOI: 10.1007/s44470-026-00109-4

Keywords: diabetes mellitus, COMISA, insomnia, obstructive sleep apnea, polysomnography, sleep fragmentation, TURKAPNE cohort, insulin resistance, sleep-disordered breathing, metabolic dysfunction, intermittent hypoxia, sleep medicine