Background: It is difficult to quantify thresholds in most colour vision tests, and this is especially the case for tritan hues, where a strong age-related increase of threshold has been reported. With the development of computer-graphic methods it is possible to remove brightness clues caused by lens absorption. This study attempts to give normative values for colour contrast thresholds and assess the age related changes therein.
Patients and methods: 115 patients aged between 6 & 71 years were tested for central and peripheral colour contrast sensitivity. No patient had any systemic or eye disease. As a preliminary, heterochromatic flicker balance between the luminosities of the R and G and B and G phosphors was established, so that all colours subsequently generated were isoluminant for the person tested. Then, using a modified binary search technique, colour contrast thresholds were established using both 2 degree optotypes, for central vision, and a ring, 12.5 degrees in radius for peripheral vision. In the latter case, the observer had to name the position of the missing quadrant in the ring. Stimuli were presented for 200 msec at 1 Hz. Colours were modulated on protan, deutan or tritan colour axis.
Results: No correlation between age and central colour vision thresholds was observed. By contrast a significant but only minor increase of peripheral colour vision threshold was observed for the peripheral protan and tritan axis.
Discussion: The present system removes luminance clues from colour vision tests and permits both central and peripheral retina to be tested. The results are simple in that the influence of age can be neglected. The variability of threshold results is small, and it is easy to detect the relatively large changes associated with disease. Since high-quality monitors are standardised and calibrated, providing the stimulus parameters described are adhered to, the results given here for upper limits of normal may be used for other similar systems.