Purpose
Previous studies have shown that chromatic discrimination thresholds across the life span are characterised by a dipper function that has its minimum at 20-30 years; after this age thresholds tend to rise due to age-related ocular and neuronal changes. However, it is unclear why thresholds should be elevated in the paediatric age range.<br /> In this study we tested psychophysically if the elevated discrimination thresholds of a paediatric population reflect a real anatomical and/or functional visual development; or if they are biased by the difficulty in performing the discrimination task, and the attentional resources required to execute the test.
Methods
We compared paediatric performance at two chromatic discrimination tests: the Universal Colour Discrimination Test (UCDT), and the Fansworth-Munsell 100 Hue Test (FM100HT). The UCDT used a simple 2-alternative-forced-choice task to measure the minimum saturation required to discriminate the chromatic target from its achromatic background. Saturation thresholds were measured along the protan, deutan, and tritan confusion lines. Each threshold took about 2 minutes and was repeated twice for a measure of reliability. The FM100HT required the observer to sort a large number of caps according to their hue, and on average it took about 20 minutes to complete the test. The two tests were run on the same day and in random order.<br /> We tested a population of 56 paediatric observers: 18 aged 5-6 years, 20 aged 9-10 years and 18 aged 13-14 years; all had normal colour vision, as assessed by either the Ishihara or the HRR plates. Our control group consisted of 25 adult observers aged 20-40 years; all had normal colour vision.
Results
Expectedly, we found that the mean total error scores measured with the FM100HT dramatically varied with age (Figure 1). Surprisingly, chromatic discrimination thresholds measured with the UCDT were approximately constant across age for all confusion axes. In fact, apart from a few outliers, all paediatric observers showed chromatic discrimination thresholds that fell within the normal trichromatic range (see, for example, tritan thresholds in Figure 2).
Conclusions
We found that chromatic discrimination in our paediatric population can be as good as chromatic discrimination in young adults, when assessed with a sensitive and fast colour discrimination test based on a simple task, like the UCDT.