Marked dissociation of photopic and mesopic contrast sensitivity even in normal observers

Although contrast vision is not routinely tested, it is important: for instance, it predicts traffic incidents better than visual acuity. Mesopic contrast sensitivity (CS) testing approximates low-lighting conditions but entails dark adaptation, which can disrupt clinical routine. In receptor-specific diseases, a dissociation of photopic and mesopic sensitivity would be expected, but can photopic CS act as a surrogate measure for mesopic CS, at least for screening purposes?

Photopic and mesopic contrast sensitivities were studied in three groups: 47 normal subjects, 23 subjects with glaucoma, and three subjects with cataract. Twenty-eight of the normal subjects were additionally tested with artificial blur. Photopic contrast sensitivity was assessed with both the Freiburg Acuity and Contrast Test (FrACT) and the Mars Letter Contrast Sensitivity Charts. Mesopic contrast sensitivity, without and with glare, was measured with the Mesoptometer IIb. Coefficients of repeatability and limits of agreement were calculated for all tests.

Test-retest limits of agreement were ± 0.17 logCS for Mars, ± 0.21 logCS for FrACT, and ±0.20 logCS / ± 0.14 logCS for Mesoptometer IIb without and with glare, respectively. In terms of inter-test comparison, Mars and FrACT largely agreed, except for ceiling effects in the Mars test. While mesopic and photopic contrast sensitivities correlate significantly (r  = 0.51, p < 0.01), only 27 % of the variance is in common. In particular, subjects with high photopic results may be nearly as likely to have low as well as high mesopic results.

The photopic contrast sensitivity tests assessed here cannot serve as surrogate measures for current mesopic contrast sensitivity tests. Low photopic CS predicts low mesopic CS, but with normal photopic CS, mesopic CS can be normal or pathologic.