Abstract
Colour obsessions and extreme affective responses to colour have been anecdotally reported in individuals with autistic spectrum disorder (ASD), while controlled studies have reported alterations in colour perception and memory (e.g. Franklin et al. 2008, 2010; Heaton et al., 2008). We examined colour preference and colour-emotion associations in ASD and typically developed (TD) populations, and related these to standard measurements of colour discrimination in the same individuals. Methods: Age- and IQ-matched ASD (n = 40) and TD (n = 28) young adults underwent a series of standardised tests (Neitz and Farnsworth-Munsell (F-M) tests of color discrimination; Autism Spectrum Quotient (AQ)) followed by custom-made tests of colour preference (rapid-response paired-comparison task with simple colour-block stimuli) and colour-emotion association (rapid-response association task pairing colour-blocks with cartoon emotional expressions). Results: Colour preferences in the ASD group are not significantly different from TD, and sex differences in colour preference are similar across the two groups. Colour-emotion associations are also similar across the two groups, although there are significant differences with respect to dependence on saturation and lightness, two components of colour perception. Colour discrimination is reduced in the ASD group, contrary to reports of superior performance in other sensory domains. The main support for this conclusion comes from performance on the Farnsworth-Munsell color discrimination test, which requires hue sorting as well as discrimination. Significantly, we find IQ to be a predictor of performance on the F-M discrimination test, but only in the ASD population. AQ is not a predictor of performance on color discrimination tests. F-M test performance correlates negatively with the strength of color preference, but this correlation is significant only for the TD group. We conclude that altered colour discrimination in ASD is likely not to explain the occurrence of colour obsessions, and measured deficits in chromatic discrimination may be related to generic performance skills.