The retina may offer a particularly accessible window onto the vascular changes accompanying Alzheimer’s disease, but new research suggests that simply looking at the retinal vascular network may not tell the whole story.
In a multimodal imaging TVST study, researchers from Bascom Palmer Eye Institute examined whether retinal blood flow, capillary network density, and retinal tissue volume change in parallel in people with Alzheimer’s disease (AD) and mild cognitive impairment (MCI). Their central finding was a striking dissociation: retinal perfusion was reduced, while the microvascular network itself remained structurally preserved.
The study included 82 participants: 28 with AD, 21 with MCI, and 33 cognitively normal controls. Each participant underwent retinal blood flow assessment using a retinal function imager, OCT angiography to quantify vessel density and capillary morphology, and ultra-high-resolution OCT to measure retinal tissue volume. The researchers also calculated derived measures of retinal tissue perfusion and retinal capillary flow.
In the combined AD and MCI group, retinal blood flow was significantly lower than in controls, and retinal capillary flow index and retinal tissue perfusion were also reduced. By contrast, retinal vessel density, vessel length density, capillary perfusion density, vessel width, and retinal tissue volume showed no significant differences in the groups.
This distinction matters. OCTA provides a picture of the density and organization of the perfused capillary network, whereas retinal function imaging captures volumetric blood flow through larger arterioles and venules. As illustrated by the multimodal imaging workflow in the study, these techniques interrogate complementary aspects of retinal neurovascular health rather than interchangeable ones.
The study authors suggest that reduced flow in the presence of preserved capillary density could reflect impaired upstream regulation – for example, changes in arteriolar tone, endothelial function, pericyte activity, or neurovascular coupling – before measurable capillary rarefaction occurs. This raises the possibility that hemodynamic abnormalities emerge earlier than the structural vascular changes detectable with conventional OCTA.
Another notable finding was the breakdown of normal structure–function relationships. In cognitively healthy controls, retinal blood flow correlated with vessel density and vessel length density after adjustment for age. In the AD and MCI group, those relationships were absent, suggesting a disease-related disruption of retinal neurovascular homeostasis.
The study is cross-sectional and relatively modest in size, so it cannot establish whether reduced retinal perfusion precedes neurodegeneration or results from it. Longitudinal studies will be needed to answer that question.
Even so, the findings underline the case for looking beyond static vessel-density metrics. If retinal blood flow declines before the vascular network visibly changes, multimodal retinal imaging could eventually provide a more sensitive, noninvasive biomarker framework for detecting and monitoring early cognitive disease.
