Abstract
Purpose :
The strikingly dichotomised responses to the “dress” photograph allow exploration of processes underlying colour perception. In a previous twin study, with a limited sample size (184 pairs), we suggested that environmental factors may play a significant role. In the present study, with >1300 participants, we sought to quantify relative genetic and environmental contributions with greater precision.
Methods :
Adult volunteers from the TwinsUK cohort were shown a standard image of the photograph, and asked to make a forced choice between “white and gold” (WG) or “blue and black” (BB); these are the commonest reported colour combinations. An age-adjusted liability threshold twin model was used to estimate relative genetic and environmental contributions to twin resemblance. Maximum likelihood structural equation models, to compare variance and covariance in twin pairs, were constructed using the OpenMx (http://openmx.psyc.virginia.edu) package in R.
Results :
After exclusion of unpaired twins, and twins of unknown zygosity, responses from 1314 individuals (433 monozygotic and 222 dizygotic twin pairs) were analysed. Mean (SD) age was 53.2 (16.4) years. 85% were female. 29.8% chose BB. Individuals choosing BB were significantly younger than those choosing WG (mean (SD) ages were 49.8 (17.9) and 54.6 (15.4) respectively, p<0.0001). Using the AE (additive genetic factors, unique environmental factors) model, age-adjusted heritability was estimated to be 41.4% (95% CI, 27.1 - 54.3%). Thus, the estimated contribution of unique environmental factors was 58.6% (95% CI, 45.7 - 72.9%).
Conclusions :
Our findings confirm a significant non-genetic contribution to the variance in perception of the colours of this image, that might equal or be greater than the contribution from genetic factors. This indicates that our colour perception is not solely genetically determined (e.g. by opsin sequences), but that environmental factors (e.g. prior colour experience or learning) play a significant role also (and maybe more important that genetic factors in some circumstances).
This abstract was presented at the 2019 ARVO Annual Meeting, held in Vancouver, Canada, April 28 - May 2, 2019.