Comorbid insomnia and sleep apnea (COMISA) was associated with a markedly elevated risk of incident type 2 diabetes (HR 6.2) compared to neither disorder.
Cohort (n=1,075,720)
Do insomnia, OSA, and comorbid insomnia and OSA (COMISA) increase the risk of incident Type 2 diabetes mellitus in U.S. Veterans?
Comorbid insomnia and obstructive sleep apnea synergistically amplify the risk of incident Type 2 diabetes in young U.S. Veterans, highlighting the importance of early sleep disorder identification.
Effect estimate: HR 6.2
p-value: p=<=0.001
Abstract Introduction Type 2 diabetes mellitus (T2DM) is a major contributor to morbidity, mortality, and healthcare utilization in the U.S., including among patients served by the Veterans Health Administration (VA). Insomnia and obstructive sleep apnea (OSA), two most common sleep disorders, are each independently linked to metabolic dysfunction via glucose dysregulation. Individuals with comorbid insomnia and OSA (COMISA) may experience greater metabolic risk; yet, most prior work focused on older, community-based samples and data from young or middle-aged adults, especially Veterans, remain scarce. Importantly, the disorders often present differently in men and women and it is unknown if the sleep-T2DM also differ. Methods We conducted secondary data analyses of a cohort study of 1,075,720 U.S. Veterans (Mage: 33.8 years, 12% women) without baseline sleep disorders or T2DM. Participants were followed from first VA encounter until first T2DM diagnosis, death, last VA encounter, or administrative censoring. Inpatient/outpatient records, ICD-9/10 codes, and pharmacy data were merged. Incident T2DM was identified using laboratory values, clinical encounters, or antihyperglycemic medication use. Insomnia, OSA and COMISA were defined using ICD codes and were modeled as time-varying exposures. Cox proportional hazards models for T2DM were conducted in the entire cohort and by sex, adjusting for demographics, body mass index, smoking, hypertension, hyperlipidemia, psychiatric conditions, and substance use. Results Insomnia and OSA were each independent predictors of incident T2DM. Insomnia alone was associated with a significantly increased risk for T2DM (HR-overall: 1.36, men: 1.33; women: 1.47). OSA alone conferred a higher risk (HR 5.47 overall; 5.68 men; 3.81 women). COMISA had the greatest risk, with HRs of 6.2 overall, 6.5 in men, and 4.7 in women, compared to neither disorder. The Insomnia-x-OSA interaction was significant (p≤0.001), indicating COMISA’s metabolic risk reflects a synergistic amplification rather than the additive effects of each disorder. Conclusion Among younger Veterans, COMISA was associated with a markedly elevated risk of new-onset T2DM, substantially exceeding the risk associated with insomnia or OSA alone and observed in both men and women. These findings highlight COMISA as a key sleep phenotype and underscore the potential importance of early identification of sleep disorders as part of diabetes prevention in high-risk populations. Support (if any)
Jumaily et al. (Fri,) conducted a cohort in Type 2 diabetes mellitus (n=1,075,720). Comorbid insomnia and obstructive sleep apnea (COMISA) vs. Neither disorder was evaluated on Incident T2DM (HR 6.2, p=<=0.001). Comorbid insomnia and sleep apnea (COMISA) was associated with a markedly elevated risk of incident type 2 diabetes (HR 6.2) compared to neither disorder.