Here, we test whether early visual and OCT rod energy-linked biomarkers indicating pathophysiology in nicotinamide nucleotide transhydrogenase (Nnt)-null 5xFAD mice also occur in Nnt-intact 5xFAD mice and whether these biomarkers can be pharmacologically treated. Four-month-old wild-type or 5xFAD C57BL/6 substrains with either a null (B6J) Nnt or intact Nnt gene (B6NTac) and 5xFAD B6J mice treated for one month with either R-carvedilol + vehicle or only vehicle (0.01% DMSO) were studied. The contrast sensitivity (CS), external limiting membrane-retinal pigment epithelium (ELM-RPE) thickness (a proxy for low pH-triggered water removal), profile shape of the hyperreflective band just posterior to the ELM (i.e., the mitochondrial configuration within photoreceptors per aspect ratio [MCP/AR]), and retinal laminar thickness were measured. Both wild-type substrains showed similar visual performance indices and dark-evoked ELM-RPE contraction. The lack of a light-dark change in B6NTac MCP/AR, unlike in B6J mice, is consistent with relatively greater mitochondrial efficiency. 5xFAD B6J mice, but not 5xFAD B6NTac mice, showed lower-than-WT CS. Light-adapted 5xFAD substrains both showed abnormal ELM-RPE contraction and greater-than-WT MCP/AR contraction. The inner retina and superior outer retina were thinner. Treating 5xFAD B6J mice with R-carvedilol + DMSO or DMSO alone corrected CS and ELM-RPE contraction but not supernormal MCP/AR contraction or laminar thinning. These results provide biomarker evidence for prodromal photoreceptor mitochondrial dysfunction/oxidative stress/oxidative damage, which is unrelated to visual performance, as well as the presence of the Nnt gene. This pathophysiology is druggable in 5xFAD mice.