Adiposity Indices Derived From Dual-Energy X-Ray Absorptiometry Are Associated With Retinal Microvascular Parameters

Objective: Evidence showed correlations between BMI and microvascular remodelling assessed with subcutaneous biopsies. Using a non-invasive technique, we aimed to compare microvascular parameters between non-obese and obese patients and to assess their association with dual-energy x-ray absorptiometry (DXA) indices and with anthropometric measurements. Design and method: We included men with hypogonadism aged 18 – 85 years who underwent a DXA and a complete medical assessment in the Endocrinology unit. We performed a non-invasive evaluation of retinal arteriolar morphology through adaptive optics (Imagine Eyes, Orsay, France). Patients with cancer, autoimmune diseases and HIV were excluded. Obesity was defined according to percentage body fat (BF(%)) => 30%. We verified the association between retinal microvascular parameters and anthropometric, DXA adiposity and lean indices using linear regressions. Results: We recruited 43 men. The mean age was 55.67 years, 60.5% was hypertensive and 46.5% was obese. Compared to non-obese, the obese population had a higher arteriolar wall thickness (WT) (12.87 μm vs 11.98 μm, p = 0.025) and wall cross-sectional area (WCSA) (4558.46 μm2 vs 4006.76 μm2, p = 0.022) (figure). WCSA and WT resulted directly associated with waist-to-hip ratio (WHR) and waist-to-height ratio (WHtR); both parameters’ association with body mass index (BMI) did not reach statistical significance (p=0.052 and p=0.053). Adiposity indices-wise, BF (%), Fat mass/height2 and Estimated visceral adiposity mass were directly associated with WCSA and WT; all these DXA parameters were also directly correlated with systolic blood pressure and pulse pressure. Wall-to-lumen ratio (WLR) was directly associated with global adiposity distribution indices like Android to ginoyd ratio and % Fat trunk/% Fat legs. All associations were significant after adjusting for age and blood pressure. There were no associations with DXA lean indices. Conclusions: Obese participants had a higher WCSA and WT, showing hypertrophic remodelling. WHR, WHtR and actual adiposity indices from DXA displayed stronger associations with retinal parameters compared to BMI, which may not fully capture such vascular impairments. In conclusion, the alterations of retinal microcirculation could be differently influenced by the extension and distribution of adipose tissue.