To determine if levodopa will slow the appearance of blood vessel changes in the eyes of patients with diabetes. Treatment will be started in patients with diabetes show delays in the electrical activity of the retina when measured non-invasively with a electroretinogram.

Diabetic Retinopathy (DR) is a leading cause of vision loss in the US. To detect DR, individuals with diabetes are instructed to receive annual eye exams until visible retinopathy such as hemorrhages or aneurysms appear that often take years to develop. Even so, treatment is only provided when visually threatening disease is detected. The investigators' group and others have established that in people with diabetes, retinal neuronal dysfunction precedes clinically visible retinopathy. Non-invasive recordings of retinal function using the electroretinogram (ERG) have shown dysfunction with diabetes, particularly in the oscillatory potentials (OPs) that are generated by inner retinal neurons. A fundamental gap in the knowledge of DR pathology is whether neuronal dysfunction is associated with or causal to the late stage vascular defects. The overall hypothesis is that neuronal defects precede vascular defects in DR and that treating neuronal deficits early in DR will prevent late stage vascular defects that result in vision loss.

Dopamine, a key neuromodulator in the retina, is reduced in DR. The investigators demonstrated that treating rodent models of diabetes with levodopa, a dopamine precursor, is neuroprotective for neuronal dysfunction. Importantly, the investigators also showed that in patients with diabetes and retinal dysfunction, but without retinopathy, levodopa taken for only 2 weeks restored retinal dysfunction to normal levels. Thus, the preliminary data suggest that earlier screening and treatment are possible to prevent or delay retinal dysfunction in early DR. Since current clinical management of DR is directed at more advanced stages of disease when vascular defects are present, it is critical to determine if levodopa will also prevent vascular pathology.

The investigators propose the following specific aims to investigate the link between neuronal and vascular defects in DR by using neuronal (dim flash ERG) and vascular (fundus photography and optical coherence tomography angiography) primary outcome measures:

Aim 1: Investigate whether the appearance of early neuronal dysfunction predicts late stage vascular pathology in diabetes. The investigators propose follow-up testing on a cohort of participants with diabetes, and normal or delayed OPs, from a prior clinical study to determine how many develop signs of retinal vascular defects after 3-5 years.

Aim 2: Determine whether levodopa treatment initiated at detection of retinal dysfunction will prevent retinal dysfunction and vascular defects. The investigators will conduct a randomized clinical trial with levodopa versus placebo using participants with diabetes and confirmed OP delays from two groups: 1) without retinopathy and 2) with the earliest signs of DR (microaneurysms). Patients will receive levodopa or placebo twice daily for 6- or 24-months. For the 6-month duration, testing will be done at baseline, 3 months and 6 months. Patients will then return to routine standard of care and be re-tested at 12 and 24 months. For the 24-month duration, testing will be done at baseline and every 3 months until 24 months. Participants will be carefully monitored for levodopa side effects with the assistance of a neurologist.

Determining the association between neuronal and vascular defects in DR is critical to shifting clinical practice toward early diagnostic markers, a move that could transform the way DR is monitored and treated, ultimately leading to better preservation of normal visual function.

Inclusion Criteria:

• HbA1c 8-12%
• Diabetic patients with no retinopathy as screened with teleretinal imaging
• Diabetic patients with microaneurysms as detected with fundus teleretinal screening
• ERG oscillatory potential delays in response to dim flash stimuli

Exclusion Criteria:

• Patients with pituitary tumor, psychosis, Parkinson's disease
• Patients with confounding ocular disease (visually significant cataract, glaucoma, macular degeneration, retinitis pigmentosa)
• Patients with cognitive deficits (score of 24 or less on the Montreal Cognitive
• Assessment-MOCA
• No anti-VEGF or steroid treatments within the last 12 months
• Pregnancy

Participants will be assigned a number and treatment group will be assigned according to randomization strategy by pharmacy personnel