Abstract
Purpose:
To compare the historical emmetropic ocular biometry data reported by Zeeman (Graefes Arch. Ophthal. 1911;78:93-128) and Sorsby (Spec Rep Ser Med Res Counc 1957: 293) with data from contemporary emmetropes published by Atchison et al. (J Vis. 2008;8:29, 1-20).
Methods:
Zeeman and Sorsby published the ocular biometry for Caucasian subjects using similar methods: retinoscopy for ocular refraction, Javal keratometry for the corneal radius of curvature, a slit lamp method for the anterior chamber depth and lens thickness, and phakometry for the lens radii of curvature. In order to make sure that the contemporary data are comparable with the historical data, it is important that compatible biometry methods are used. Hence Placido topography and ultrasound biometry were used (both of which have been validated against the historical methods in the literature), as well as autorefraction and phakometry. For all three datasets the axial length was derived using a method proposed by Sorsby. As this method does not take the aging of the crystalline lens into account, only subjects aged between 18 and 38 were considered. This left 25 emmetropes from Zeeman (refraction within ±1 D), 38 from Sorsby and 20 from Atchison et al. for analysis. Comparisons were done by ANOVA, a t test after Fisher r-to-z transformation to compare correlations, and Flury Hierarchy and random skewers to compare covariance matrices.
Results:
After Šidák correction to account for alpha inflation in case of multiple comparisons, the posterior lens radius was found to be significantly flatter for Atchison et al. (-6.88 ± 0.96 mm) than for Zeeman (-6.02 ± 0.53 mm; P = 0.001), and the anterior chamber depth was significantly shallower for Sorsby (3.55 ± 0.24 mm) than for Zeeman (3.76 ± 0.29 mm; P = 0.004). No significant differences were seen in the correlations between the parameters of the three datasets. For the covariance matrices, considerable and significant differences were seen between the Atchison et al. data on the one side and the Zeeman and Sorsby data on the other side.
Conclusions:
The average biometry and correlation between biometric parameters of emmetropic subjects has remained similar throughout the last hundred years. The significant differences between the studies for the biometry and the covariance matrices may reflect differences in methods or instrument calibration.
Keywords: 459 clinical (human) or epidemiologic studies: biostatistics/epidemiology methodology •
630 optical properties •
676 refraction