Background: We sought to determine whether the change in cortical excitability secondary to deafferentation in patients with Charles Bonnet Syndrome (CBS) who hallucinate in a predominant color or combination of colors is related to an alteration in color contrast thresholds and whether the change is specific to the color of the hallucination.
Methods: We prospectively categorized each patient's hallucinations using the Institute of Psychiatry Visual Hallucinations Interview. We measured color contrast thresholds with a computerized test designed to assess red-green and blue-yellow color confusion axes against a background of luminance noise. We calculated the ratio of red-green threshold to blue-yellow threshold (R-G/B-Y ratio) for each patient. Because central vision was impaired in all patients, we used a sectoral annular stimulus that projected to the retina at 12.5 degrees eccentricity.
Results: There were 10 patients with age-related macular degeneration and CBS who were hallucinating in a predominant color or combination of colors at the time of recruitment. Patients hallucinating in red, green, or a combination of red and green had R-G/B-Y ratios of less than 1.0 (n = 5). Patients hallucinating in blue, yellow, or a combination of blue and yellow had R-G/B-Y ratios of greater than 1.0 (n = 2). Patients hallucinating in purple had ratios between the red-green and blue-yellow hallucinators (n = 2). The 1 patient hallucinating in white had the lowest thresholds for red-green and blue-yellow confusion axes. Comparing the R-G/B-Y ratios for the "red/green hallucinators" and "blue/yellow hallucinators" returned a significant result with Fisher's exact test (P = 0.047, n = 7).
Conclusions: Deafferentation and secondary cortical hyperexcitability in CBS have a correlate in psychophysical threshold. This change in sensitivity relates specifically to the hallucinated color axis rather than across all colors. This is the first published evidence for cerebral hyperexcitability leading to a decrease in color contrast thresholds.