97th DOG Annual Meeting 1999



C. Erb1-3, A. Ulrich1, M. Adler2, E. Zrenner2, J. Flammer1

Aim: To study the influence of different illumination levels on the results of the hue-discrimination test, Roth 28-Hue desaturated.

Subjects and Methods: The cap-sorting test, Roth 28-Hue desaturated, was used under 4 different illumination levels in 4 groups from a total test cohort of 172 ocularly and generally healthy subjects: 4800 lux (Group 1:76 subjects; mean age ± SD: 30±9 years), 2000 lux (Group 2: 35 subjects; mean age ± SD: 31±5 years), 1000 lux (Group 3: 34 subjects; mean age ± SD: 31±8 years), and 100 lux (Group 4: 26 subjects; mean age ± SD: 30±7 years). The illumination levels were determined with the Minolta Chroma Meter XY-1. Illumination was provided by fluorescent light sources with a balanced spectral distribution and a colour-rendering index of 93 (Osram L36W/12 LDL Daylight).

Results: No differences between the illumination level of 4800 lux (average error sum=median ± MAD; 42±18) and 2000 lux (average error sum: 40±22) could be detected. However, examination with 1000 lux showed a small increase in the average error sum (52±21), while examination with 100 lux showed a statistically significant increase in the average error sum (78±20) [Wilcoxon, p<0.001] but without a preferred disturbance of any particular colour axis.

Discussion: These differences in error score demonstrate that the results of hue-discrimination tests can be influenced by the illumination level. This has to be considered when results of different colour tests are compared. As a standard, we recommend an illumination level of 2000 lux.

1University Eye Clinic, Mittlere Strase 91, CH-4012 Basel, Switzerland
2University Eye Clinic, Schleichstrasse 12, D-72076 Tübingen, Germany
3Eye Clinic, Medizinische Hochschule Hannover, Carl-Neuberg-Str. 1, D - 30625 Hannover