Patients with diabetes mellitus and comorbid obstructive sleep apnea (OSA) may be at greater risk of developing diabetic retinopathy (DR) in part due to common underlying factors, such as inflammation, according to a report published in Clinical Ophthalmology. Moreover, continuing stress on small retinal vessels can further bring about diabetic macular edema (DME).

The study evaluates the effect of sleep metrics — apnea-hypopnea index, sleep efficiency, and oxygen saturation —  on DME, as well as how individuals with OSA respond to 12 months of anti-vascular endothelial growth factor (anti-VEGF) therapy and continuous positive airway pressure (CPAP). Investigators evaluated patients with type 1 or type 2 diabetes, enrolled at an ophthalmology clinic in Ontario, Canada from March 2017 to November 2019. OSA cases were moderate or severe, with apnea-hypopnea index of ≥15, and DME was described as central subfield retinal thickness >275 µm, with the existence of intraretinal or subretinal fluid.

Baseline polysomnography of 74 eyes (49 participants), mean age 64±8.4 years revealed that OSA incidence was significantly higher in those with DME (P <.05), and average minimum SaO2 was considerably lower in individuals with OSA in the presence of DME (P <.05). Patients with DME tended to have somewhat greater apnea-hypoxia index (P =.22) and apneas (P =.15), but decreased hypopnea (P =.20), compared with those without DME. Also significantly impacted in the DME-positive group was sleep efficiency (P =.07) and mean oxygen saturation (P =.09). “We found minimum SaO2 to be the only significant clinical variable of DME in patients with OSA, in agreement with other studies finding minimum SaO2 to be associated with DR,” researchers explain. Thus, SaO2 may be a sensitive indicating factor for hypoxia in DR and DME.

Continue Reading

Although logMAR visual acuity and mean central retinal thickness (CRT) of patients with DME who were not experiencing OSA was significantly better after 12 months, those with both did not meaningfully improve. The analysis notes that OSA may have reduced effectiveness of the aflibercept therapy for DME — serum VEGF was detected at a higher level in sleep-disrupted DR participants, and additional inflammatory cytokines potentially had an impact on these individuals. DR severity and OSA severity were not correlated, based on structural data or sleep metrics.

Patients with OSA were provided with CPAP, although those using the device for 4 hours or more on 70% of nights was adhered to by a somewhat low proportion of participants: 20% of individuals with DME, and 36% without. After 12 months, DME-positive CPAP users displayed a significantly reduced DR severity score compared with those who underused this tool (P =.042).

Of the 49 participants, 95% presented with type 2 diabetes. Due to 23 dropping out before undergoing polysomnography, the sample was relatively small, limiting the investigation. Although there was no association between DME and OSA after controlling for multiple variables, this may have simply reflected a stronger influence of age and BMI on OSA in this particular cohort. Low compliance rate with CPAP also reduced the power of resulting data. 


Kaba Q, Tai F, Al-Awadi A, Somani S. Examining the Relationship Between Diabetic Macular Edema, and Obstructive Sleep Apnea. Clin Ophthalmol. 2022;16(4):1215–1223. doi:10.2147/OPTH.S354087