Sleep disordered breathing has minimal association with retinal microvascular diameters in a non-diabetic sleep clinic cohort

Abstract

Introduction Obstructive sleep apnea (OSA) may increase stroke risk; retinal arteriolar (central retinal arteriolar equivalent, CRAE) diameter narrowing and/or retinal venular (central retinal venule equivalent, CRVE) widening may predict stroke. We examined relationships between sleep disordered breathing (SDB) and CRAE and CRVE and in a diabetes-free sleep clinic cohort. Methods Patients for SDB assessment were recruited (Main Group, n = 264, age: 58.5 ± 8.9 yrs [mean ± SD]; males: 141) for in-laboratory polysomnography (standard metrics, eg apnea hypopnea index, AHI) and retinal photographs (evening and morning). A more severe SDB sub-group (n = 85) entered a 12-month cardiovascular risk factor minimisation (hypertension/hypercholesterolemia control; RFM) and continuous positive airway pressure (CPAP) intervention (RFM/CPAP Sub-Group); successfully completed by n = 66 (AHI = 32.4 [22.1–45.3] events/hour, median[IQR]). Univariate (Spearman’s correlation, t-test) and multiple linear regression models examined non-SDB and SDB associations with CRAE and CRVE measures. Results Main Group: Evening CRAE predictors were: systolic blood pressure (0.18μm decrease per mmHg, p = 0.001), age (2.47μm decrease per decade, p = 0.012), Caucasian ethnicity (4.45 μm versus non-Caucasian, p = 0.011), height (0.24 μm decrease per cm increase, p = 0.005) and smoking history (3.08 μm increase, p = 0.052). Evening CRVE predictors were: Caucasian ethnicity (11.52 μm decrease versus non-Caucasian, p>0.001), diastolic blood pressure (0.34 μm increase in CRVE per mmHg, p = 0.001), hypertension history (6.5 μm decrease, p = 0.005), and smoking history (4.6 μm increase, p = 0.034). No SDB metric (all p>0.08) predicted CRAE or CRVE measures. RFM/CPAP Sub-Group: A one-unit increase in ln(AHI+1) was associated with a 0.046μm increase in CRAE (n = 85; p = 0.029). Mean evening CRAE and CRVE values did not change across the intervention (n = 66), but evening CRVE decreased ~6.0 μm for individuals with AHI >30 events/hr. Conclusion No major SDB associations with CRAE or CRVE were identified, although the RFM/CPAP intervention reduced evening CRVE for severe OSA patients. Implications for cerebro-vascular disease risk remain uncertain. Trial registration The protocol was registered with the Australian New Zealand Clinical Trials Registry (Trial Id: ACTRN12620000694910).