In this study, we investigated pRNFL thickness, peripapillary VD, and their ratio in T2DM patients without clinical DR to understand the relationship between DRN and microvascular impairment in more detail, and found that pRNFL thickness and VD were significantly reduced in the DM group, and the RNFL/VD ratio was significantly increased in the DM group compared to the control group. Patients with relatively short T2DM duration showed a higher RNFL/VD ratio than patients with a longer T2DM duration, which implies that prominent microvascular damage may precede inner retinal reduction by DRN. Additionally, this ratio was significantly related to age, T2DM duration, and axial length in the DM group, whereas none of the factors showed a significant relationship in the control group.
Previous studies reported a significant progressive loss of the RNFL and the ganglion cell layer in patients with DM6,17. Our study also showed a significantly thinner pRNFL in T2DM patients. T2DM induces activation of microglial cells which migrate to the subretinal space and release cytokines contributing to neuronal cell death9,18. Such neuronal loss of the retina could result in a reduction of the pRNFL thickness in T2DM patients even though they do not show any DR changes such as retinal hemorrhage, hard exudates, or cotton wool spots. Additionally, Vujosevic et al.8 reported a significant decrease in peripapillary VD in the DM groups compared to the control group, which was similar to the results of our study. The accumulation of glutamate and the loss of neuroprotective factors trigger vascular endothelial growth factor activation, and neuronal death and glial dysfunction lead to vasoregression, which could result in early microvascular impairment in T2DM9,10. These retinal damages are related to various visual functions7,12,13,19,20. Therefore, the status of microvasculature and pRNFL thickness should be observed carefully in T2DM patients, even if they do not show signs of clinical DR.
We tried to infer the sequence of the DRN and microvascular impairment by analyzing the RNFL/VD ratio and found that the RNFL/VD ratio was significantly higher in the DM group than in the control group. In the situation of a decrease in both pRNFL thickness and VD by T2DM, the higher ratio would mean a more prominent impairment of VD than that of pRNFL. DRN and microvascular impairment in T2DM are closely related to each other and share similar mechanisms triggered by hyperglycemia, which we mentioned above. However, the temporal relationship has yet to be clarified. The higher value of the RNFL/VD ratio in T2DM patients without clinical DR implies that prominent microvascular damage may precede inner retinal reduction by DRN. Although DRN would occur before DR-related microvascular damages including microaneurysm or retinal hemorrhage, the impairment of peripapillary microvasculature may occur before a definite inner retinal reduction.
Patients with relatively short T2DM duration showed a higher RNFL/VD ratio than patients with a longer T2DM duration. After microvascular impairment was preceded in the early stages of T2DM, inner retinal reduction due to DRN would be accompanied over time, which resulted in a lowered ratio once again. Taken together, these results support the hypothesis that the impairment of peripapillary microvasculature precedes DRN. However, there could be some bias coming from the exclusion of clinical DR. Patients with relatively longer T2DM duration are more likely to experience DR change, along with greater microvascular impairment. As we excluded such patients with DR change, there was a possibility that only those with relatively minimal microvascular damage were enrolled in the DM group with relatively longer T2DM duration. Further longitudinal studies are needed to confirm this hypothesis.
Age was negatively correlated with the RNFL/VD ratio in T2DM patients. Although it is controversial, Lee et al.18 reported that age significantly influenced changes in the pRNFL thickness over time in patients with T2DM. However, previous studies did not show a significant relationship between age and retinal microvasculature in T2DM12,13,21. In older T2DM patients, inner retinal damage by DRN seemed to be more prominent than microvascular impairment. Additionally, T2DM duration was negatively related to the RNFL/VD ratio, which would mean that the reduction in pRNFL thickness was more prominent than the microvascular impairment with increasing T2DM duration. T2DM duration is a well-known factor associated with both DRN and impairment of retinal microvasculature6,7,11,12,13,17,18. As mentioned above, inner retinal damage by DRN seemed to appear gradually and severely over time after microvascular damage which occurred in the early stage of T2DM. In other words, damage to the inner retina would become more prominent than damage to the retinal microvasculature over time in T2DM.
Axial length was positively correlated with the RNFL/VD ratio in the DM group. Bundenz et al.22 reported that for every 1 mm increase in axial length, the mean pRNFL was measured as being thinner by approximately 2.2 μm in normal subjects. Other previous studies have also reported a negative association between pRNFL thickness and axial length23,24. Meanwhile, He et al.25 reported that radial peripapillary complex VD showed a significantly negative correlation with axial length. Lee et al.13 also reported that axial length was a significant factor associated with peripapillary VD in T2DM patients. As such, both the pRNFL thickness and peripapillary VD had a negative relationship with axial length, and the VD seemed to be more affected than the pRNFL thickness by the axial length in T2DM from the results of a positive correlation between axial length and the RNFL/VD ratio. In T2DM, the retinal microvasculature would be influenced more prominently than the structure of the inner retinal layer by axial elongation. However, we did not enroll patients with axial length ≥ 26.0 mm, and did not perform the corrections of the OCTA image scale according to the axial length, which may contribute to a slight image scale difference among individuals. As such, analyses involving the axial length would be somewhat limited.
This study had several limitations. First, the retrospective nature of the study inevitably introduced some selection bias. Prospective longitudinal studies are needed to confirm our hypothesis. Second, since eyes with axial length ≥ 26.0 mm were excluded, additional studies including high myopia are needed to analyze the relationship more precisely between axial length and other parameters. The strength of this study was that we enrolled a relatively large number of patients, and enrolled OCTA images with signal strength > 8 allowing for accurate analyses. Additionally, this is the first study to infer the temporal relation of DRN and retinal microvasculature impairment.
In conclusion, the RNFL/VD ratio of T2DM patients was higher than that of normal controls, and patients with relatively short T2DM duration showed a higher ratio than patients with a longer T2DM duration. These would indicate that the impairment of peripapillary microvasculature precedes DRN and that the peripapillary microvasculature is affected more prominently than the inner retina at the early stages of T2DM. Additionally, age and T2DM duration were negatively correlated with the RNFL/VD ratio, from which it could be inferred that inner retinal damage by DRN becomes more prominent than microvascular impairment over time in T2DM. These results are expected to be helpful for physicians interpreting changes in the inner retina and OCTA parameters in T2DM patients and may also be useful for a better understanding of the pathogenesis of the early stages of the disease to find out more efficient preventive strategies for retinal damage in T2DM.