PURPOSE: To examine the diurnal variations in corneal autofluorescence in normal and diabetic patients. METHODS: We measured corneal autofluorescence using a fluorophotometer fitted with an anterior segment adapter. Corneal autofluorescence was measured 10 times at 3 min intervals to evaluate the reproducibility of this instrument in 4 eyes of 4 normal subjects. The diurnal variation in corneal autofluorescence was determined by measuring the fluctuations in 10 eyes in 10 normal subjects and one unoperated eye each of 10 patients with proliferative diabetic retinopathy (PDR). We performed five consecutive measurements at 1000, 1130, 1400, 1630 and 1900 hours. The mean value of five measurements, the variation range and the coefficient of variation were analysed. RESULTS: The mean coefficient of variation in the measurement using this instrument was 8.6 +/- 1.0%. In the patients with PDR, the mean corneal autofluorescence value was significantly higher (p < 0.001), the variation range was significantly wider (p < 0.001) and the coefficient of variation was significantly greater (p < 0.01) than in the normal subjects. CONCLUSIONS: The results of this study suggest that corneal autofluorescence changes over the course of a day in patients with diabetes. This may be caused by the breakdown of the blood-aqueous barrier that we reported previously.
PURPOSE: To examine the diurnal variations in corneal autofluorescence in normal and diabeticpatients. METHODS: We measured corneal autofluorescence using a fluorophotometer fitted with an anterior segment adapter. Corneal autofluorescence was measured 10 times at 3 min intervals to evaluate the reproducibility of this instrument in 4 eyes of 4 normal subjects. The diurnal variation in corneal autofluorescence was determined by measuring the fluctuations in 10 eyes in 10 normal subjects and one unoperated eye each of 10 patients with proliferative diabetic retinopathy (PDR). We performed five consecutive measurements at 1000, 1130, 1400, 1630 and 1900 hours. The mean value of five measurements, the variation range and the coefficient of variation were analysed. RESULTS: The mean coefficient of variation in the measurement using this instrument was 8.6 +/- 1.0%. In the patients with PDR, the mean corneal autofluorescence value was significantly higher (p < 0.001), the variation range was significantly wider (p < 0.001) and the coefficient of variation was significantly greater (p < 0.01) than in the normal subjects. CONCLUSIONS: The results of this study suggest that corneal autofluorescence changes over the course of a day in patients with diabetes. This may be caused by the breakdown of the blood-aqueous barrier that we reported previously.