Red Blood Cell Distribution and Volume Changes at Capillary Junctions in the Living Human Retina

To characterize red blood cell (RBC) distribution and associated changes in cell size at capillary junctions in the human neural retina.

Cell-resolved blood flow across 60 capillary junctions in the retina of 6 healthy human subjects was measured, using flood-illumination adaptive optics at 200 to 400 fps. Several empirical RBC partitioning models, including a widely used model developed by Pries et al., were compared for predictive accuracy.

We provide updated Pries model coefficients to suit human retinal capillaries. We also propose a simpler linear model that predicts RBC flow distribution based solely on the share of blood flow received by a branch, without reference to the main vessel. Cell size analysis revealed that the average RBC volume in branch vessels was 5.5% lower than in the main vessel (P < 0.05). The reduction in cell volume accompanied a 2.3 µm increase in inter-cell spacing (P < 10-4) and a 3.0% decrease in hematocrit (P < 10-4). These findings support the hypothesis of a net transfer of water from RBCs to plasma as they pass through narrower capillary branches, with potential reabsorption by the time cells enter wider collecting capillaries.

This study provides the first noninvasive in vivo characterization of RBC partitioning behavior at capillary junctions in the living human retina. The subtle fluid exchange between cells and plasma may play a role in optimizing oxygen delivery to support immediate metabolic needs.