K Pesudovs, DJ Coster
Department of Ophthalmology, Flinders Medical Centre and Flinders University of South Australia

Purpose. Examine the relationship between cataract severity and colour confusion.

Methods. Eighty-one subjects with cataract varying from sub-clinical to severe and no ocular comorbidity were recruited for this study. The LOCS III cataract scoring system was used to grade cataract severity and colour vision was measured with the Farnsworth D15 panel test of colour arrangement. Panel D-15 scores were converted into colour difference vectors (CDVs), Confusion index (C-index), Selectivity index (S-index) and angle by the technique of Vingrys and King-Smith. These numerical indices of colour confusion facilitate analysis of the relationship between cataract severity (LOCS III scores) and colour confusion using linear regression.

Results. The LOCS III scores and CDVs were normally distributed across a large range. Stepwise linear regression of cataract scores for posterior subcapsular (P), cortical (C), nuclear colour (NC) and nuclear opacity (NO) against the colour difference vectors, C-index and S-index found NC to be the most significant variable in both cases [S-Index = 0.54 + 0.45NC (R2=0.59), C-index = 0.21 + 0.44NC (R2=0.62)]. Despite a high correlation between NO and NC (R = 0.85), with NC as a regression factor, NO adds no further information. No association was found between the colour difference vectors and C or P scores.

Conclusions. These results extend previous work that shows nuclear colouration causes colour confusion. The severity of the colour vision defect can be predicted from NC score. None of the other indices of cataract severity were significant, suggesting that the origin of colour vision loss is nuclear colouration rather than nuclear opacity. It is proposed that the mechanism of colour vision loss is more likely to be selective wavelength absorption within the lens nucleus than a change in the adaptational state of the eye due to overall reduced retinal illumination.