Diabetic retinopathy is a complication of diabetes characterized by an extremely low rate of progression. It takes several years to move from the onset of diabetes, both type 1 and type 2, to the development of retinal microaneurysms, then leading to proliferative diabetic retinopathy and vision loss. The recent demonstration that retinal microaneurysms are preceded and, possibly, caused by a subclinical neurodegeneration mainly affecting the neurovascular unit has suggested, on one hand, the possible existence of a previously unknown early neurodegenerative stage of diabetic retinopathy and, on the other, that an early “neuroprotective” treatment could end up preventing the development of the microvascular stages. This review summarizes the present situation in the field and focuses on the prevention of diabetic retinopathy, which seems, for the first time, to be within reach.

Diabetes-driven retinal neurodegeneration has recently been shown to be involved in the initial phases of diabetic retinopathy, raising the possibility of setting up a preventive strategy based on early retinal neuroprotection. To make this possible, it is crucial to identify a biomarker for early retinal neurodegeneration. To this end, in this study, we verified and confirmed that, in the Akita mouse model of diabetes, the thinning of the retinal nerve fiber layer/ganglion cell layer (the RNFL/GCL—the layer that contains the retinal ganglion cells) precedes the death of these same cells, suggesting that this dysfunction is a possible biomarker of retinal neurodegeneration. We then confirmed the validity of this assumption by starting a neuroprotective treatment (based on nerve growth factor eye drops) in concert with the first demonstration of RNFL/GCL thinning. In this way, it was possible not only to avoid the loss of retinal ganglion cells but also to prevent the subsequent development of the microvascular stage of diabetic retinopathy. In conclusion, in the case of diabetes, the thinning of the RNFL/GCL appears to be both a valid biomarker and a pharmacological target of diabetic retinopathy; it precedes the development of vascular dysfunctions and represents the ideal starting point for prevention.

A growing body of evidence indicates that bone is a target for diabetic complications. Hyperglycemia is associated with bone fragility and increased risk of fractures. In type 1 diabetes, which typically develops in children and adolescents, bone alterations already occur in the skeleton during the growth processes. In type 1 diabetic Akita mice, we analyzed with a pQCT scan the volumetric bone mineral density (vBMD) changes caused by hyperglycemia starting from 10 to 40 weeks of age. The results showed lower cortical vBMD at 20, 30 and 40 weeks of age, whereas cortical thickness and area were lower compared with control mice at all observation times from late growth (10 weeks of age) to aging (40 weeks of age). Fibroblast growth factor 23 (FGF23), synthetized mainly in osteocytes, is considered a ‘bone hormone’, that is critical to the maintenance of phosphate homeostasis. Diabetic Akita mice showed a persistent increase in the urinary calcium-to-creatinine ratio throughout the experimental period. In the presence of renal histological alterations, the urinary phosphate-to-creatinine ratio varied at different ages; its significantly lower value at 30 weeks of age, compared to wild-type mice, was associated with significantly lower plasma FGF23 levels. This study shows that in the Akita mice, the alterations in bone structure, caused by the high plasma glucose levels, are associated with a modulation of urinary calcium and phosphate excretion over time; the latter was only partially associated with the adjustment of FGF23 secretion.