Verification of Predictive Accuracy of Index of Lens Transparency Using a Scheimpflug Camera and 12-Year Follow-Up

Abstract

Purpose: : Lens transparency properties (LTP) form an index formulated by the author's group to describe normal age-related changes of lens transparency using intensity of backscattered light (Sasaki H, et al: Ophthalmic Res. 1999), using a EAS-1000 Scheimpflug camera (NIDEK).

Methods: : In 1996 a random sample of 1045 persons 50 years and older were examined in the Reykjavik Eye Study (RES). 126 right eyes with transparent lenses at 12 year follow-up of the RES in 2008 were examined for correlation between LTP score in both years. The average age of subjects at baseline was 54.3±4.0 years old. LTP was derived by the formula, LTP =A+C×Lac+E×Lae (A,C,E: scattered light of anterior capsule, of anterior adult nucleus, of central clear zone, Lac, Lae: optical distance between A and C, optical distance between A and E). EAS-1000 (NIDEK) was used for measurement.

Results: : There was significant linear positive correlation between LTP and age both in 1996 and 2008; LTP (1996)=11.2×age-291.1 (R2=0.537), LTP (2008)=12.5×age-396.9 (R2= 0.407). There was no significant difference between these approximations (ANCOVA, significance level 5%). The correlation between the value predicted from the expression in 1996 and the actual measured value in 2008 was LTP (2008)=1.12×LTP (1996)-69.8 (R2= 0.407, P<0.001). The actual mean value of LTP in 2008 was 433.5 and the predicted mean value was 447.9 with an error rate of -3.3%.

Conclusions: : LTP makes the prediction of longitudinal value from cross-sectional date with a high degree of accuracy possible and therefore it is useful as an index of age-related change in lens.

Keywords: cataract • clinical (human) or epidemiologic studies: biostatistics/epidemiology methodology