In this study, we used SS OCTA to investigate the retinal microvasculature and the FAZ area in WMHs and healthy control participants. Our results showed significantly decreased VD in both the ICP and the DCP and an enlarged FAZ area in WMH participants compared with age-matched healthy control participants. These differences suggested that the FAZ area and the VD in the ICP and DCP may provide an earlier marker in screening for WMHs and the preclinical phase of AD.
Compared with healthy control participants, we found significantly decreased VD in both the ICP and the DCP of the retina in WMH participants. Peng and colleagues  compared the microvascular changes in WMH participants with different degrees of lesions and indicated that microvascular impairment was associated with the disease cascade in WMHs. Mutlu and colleagues  have also reported an association between retinal neurodegeneration and cerebral atrophy, indicating the role retinal OCT plays in providing information on neurodegeneration in the brain.
However, the mechanism of decreased microcirculation in WMHs is still unknown and needs further study. Previous studies have suggested that the changes in the retina reflect the similar pathological changes occurring in the brain , since the retina shares many similarities with the brain, including embryological origin, microvasculature and neuronal projections . It is known that retinal vascular networks are composed of several layers that differ in location, composition and function. The SVP includes flow from the ILM to the middle of the IPL and mainly supplies the ganglion cell layer. It is a network and contains both large and small vessels . The deep vascular plexus (DVP) has two subcomponents: the ICP, containing a slab extending from the IPL to the INL; and the DCP, extending from the INL to the OPL. The ICP and DCP are mainly composed of smaller vessels and capillaries . Since the SVP mainly contains larger vessels and the DVP (ICP and DCP) is related to the function of capillary microcirculation, in our study, the significant decreases in the ICP and DCP may indicate the early appearance of microcirculation dysfunction in the retina in WMH participants. Thus, as a previous report suggested that WMHs may originate from chronic small vessel ischaemia , our study showed that what is seen in the retinal microcirculation may be a reflection of the changes in cerebral microcirculation. Additionally, morphologic substrate, induced by impaired diffusion, was found through thickened vessel walls in WMH participants . Since those changes likely start with smaller vessels, it is of great possibility that those substrates may be present in the small vessels of the ICP and DCP, which may result in early vessel dropout and microvascular dysfunction.
Another possible explanation for decreased VD in the ICP and DCP may be alterations in oxygen. Previous retinal oxygen measurement experiments in animal models have demonstrated that for the inner retina, the dominant oxygen consumers may be located in the plexiform layers (IPL and OPL), probably in mitochondria-rich synapses [25, 26], indicating the need for a highly oxygenated blood supply. Notably, according to Hagag and colleagues , in healthy subjects, when compared with other capillary layers of retina, the plexiform layers may maximally constrict under the exposure to hyperoxia, which indicated that the plexiform layers may be influenced greatly under the hypoxic status. Hence, the impaired VD in the ICP and DCP may also be a reflection of the mitochondrial dysfunction and high oxygen consumption found in the brains with WMHs.
We found a significant enlargement of the FAZ area in WMH participants, which has rarely been reported in previous studies. The FAZ refers to the capillary-free area centred on the macula. In this study, the FAZ area was automatically measured in the inner retina, which extends from 5 μm above the interface ILM to 25 μm beneath the interface INL/OPL. Quantifying FAZ parameters has been a useful tool in evaluating macular microcirulation . Since a prior study showed a significant age effect on the FAZ area , we selected age-matched participants as controls. The mechanism underlying the enlarged FAZ area in WMH participants remains unknown, but microcirculation dysfunction has been reported to be a cause. Another possible explanation may be neurodegenerative changes. Partial loss of myelin, axons, and oligodendroglion cells in the brain is a common finding seen in WMH participants. The enlargement of the FAZ area in the inner retina may be secondary to both the impairment of neurons and glia in the retina and their effects on the decrease in flow density. Third, perivascular tissue changes, considered the prevailing morphologic substrate , may also be a potential explanation. According to Ma KC and colleagues , perivascular oedema in the retina, due to intermittent disruption of the blood-brain barrier, could also lead to damage in VD, which results in enlargement of the FAZ area. Hence, the enlarged FAZ area may indicate the impairment of microcirculation in the brain.
Previous studies demonstrated that retinal microvascular density might be associated with axial length , sex  and refractive errors . However, some remain disputable. Wen and colleagues  found that the axial length was negatively associated with superficial parafoveal vessel density but not correlated with VD around the optic nerve head and the FAZ area, while some previous studies indicated that the axial length was negatively correlated with the FAZ area . Milani and colleagues  reported that eyes with high myopia were negatively related to superficial VD and positively associated with blood flow of the outer retina. Hence, we adjusted for these parameters during the evaluation.
We acknowledge several limitations in this study. The first is the limited sample size; however, we have used. Second, limited by our study design, we were not able to determine whether these results are related to individual differences. Therefore, we could not be able to determine whether our findings could be translatable to all WMH individuals. However, we designed strict recruitment and made adjustments for multiple comparisons. Further studies with larger sample sizes and longer follow-ups are needed. Third, all the participants were Chinese, which may make our findings hard to generalize to other ethnic groups. The fourth limitation is that we did not measure the microstructural volume of WMH participants. Moreover, most of our WMH participants were in the early stage of the disease; hence, we did not group people according to the severity of disease. Further studies with more detailed neuroimaging data and patients in different disease stages are warranted to enhance our findings and provide in-depth meaning.
There are also some strengths in our study. First, we used SS OCTA for evaluation. The unprecedented scanning speed, imaging depth and sensitivity of SS OCTA improved the accuracy of images. Second, we detected both VD and the FAZ area in WMH participants, which has rarely been reported before and may provide more possible imaging targets for the early diagnosis of WMHs.