The potential association between obstructive sleep apnea and diabetic retinopathy in severe obesity-the role of hypoxemia.

BACKGROUND

Obstructive sleep apnea (OSA) is common in obese patients with type 2 diabetes mellitus (DM) and may contribute to diabetic microvascular complications.

METHODS

To investigate the association between OSA, hypoxemia during sleep, and diabetic retinal complications in severe obesity. This was a prospective observational study of 93 obese patients mean (SD) age: 52(10) years; mean (SD) body mass index (BMI): 47.3(8.3) kg/m(2)) with DM undergoing retinal screening and respiratory monitoring during sleep. OSA was defined as apnea-hypopnea index (AHI) of ≥15 events/hour, resulting in two groups (OSA+ vs. OSA-).

RESULTS

Forty-six patients were OSA+: median (95% CI) AHI = 37(23-74)/hour and 47 were OSA-ve (AHI = 7(4-11)/hour). Both groups were similar for ethnicity, BMI, cardiovascular co-morbidities, diabetes duration, HbA1c, and insulin treatment (p>0.05). The OSA+ group was significantly more hypoxemic. There was no significant difference between OSA+ and OSA- groups for the presence of retinopathy (39% vs. 38%). More OSA+ subjects had maculopathy (22% vs. 13%), but this did not reach statistical significance. Logistic regression analyses showed that AHI was not significantly associated with the presence of retinopathy or maculopathy (p>0.05). Whilst minimum oxygen saturation was not significantly associated with retinopathy, it was an independent predictor for the presence of maculopathy OR = 0.79 (95% CI: 0.65-0.95; p<0.05), after adjustment.

CONCLUSIONS

The presence of OSA, as determined by AHI, was not associated with diabetic retinal complications. In contrast, severity of hypoxemia during sleep (minimum oxygen saturations) may be an important factor. The importance of hypoxia in the development of retinal complications in patients with OSA remains unclear and further studies assessing the pathogenesis of hypoxemia in patients with OSA and diabetic retinal disease are warranted.