The recent article by Dumbleton et al. 1 was both interesting and intriguing. Of the many conclusions, the investigators observed a significant though small mean increase in myopia of −0.3D after extended wear of low-Dk hydrogel lenses (Etafilcon A). Intuitively, an increase in myopia could result from a steepening of corneal curvature. In reality, the authors found that mean corneal curvature, measured using an autokeratometer, remained unchanged. The autokeratometer estimates corneal curvature at a location 1.5 to 2 mm from the center of the cornea. A real change at the center of the cornea would be missed using this keratometer. The investigators offered no mathematical hypothesis to account for the change in refraction. They did suggest that changes in corneal refractive index might be responsible. Corneal optics is heterogeneous—the epithelium has an average refractive index of 1.401. The refractive index of the underlying stroma is 1.380 and the corneal posterior surface has a refractive index of 1.372. 2 The average refractive index of the stroma along the anteroposterior direction is 1.376. In addition, there is a topographic variation in thickness distribution within the epithelium and stroma. 3 If we assume that the source of the change in refractive error lies in the cornea, then four further pos sible parameters could account for the observed shift in refractive error: 1. The actual radius at the center of the refractive portion of the cornea (R1). 2. The radius of curvature at the epithelial-stromal interface (R2). 3. The radius of curvature at the posterior corneal surface (R3). 4. The refractive index of the corneal epithelium. The questions are: how much of a change needs to take place in each parameter to account for the average change in refractive error and what could be the secondary consequences of such hypothetical changes? Using a model eye incorporating the heterogeneous properties of the cornea 3 and the relevant averaged values quoted by Dumbleton et al., 1 for the average refractive error to change from −3.37D to −3.67D, one of the following must occur: 1. R1 has to decrease from 7.59 mm to 7.54 mm. 2. R2 has to increase from 7.34 mm to 8.2 mm. 3. R3 has to decrease from 5.8 mm to 5.45 mm. 4. The refractive index of the epithelium or posterior corneal surface increases by 0.001 units. 5. The refractive index at the anterior stromal surface reduces from 1.380 to 1.369. Using the standard sag formula, over a 4-mm chord diameter, the change in R1 could result from a 0.002-mm increase in central corneal thickness relative to any change in corneal thickness 2 mm from the center. R2 could increase by the computed amount if the thickness of the epithelium increased by 0.03 mm at the very center. R3 could reduce by this amount if there was a 0.025-mm circumferential increase in stromal thickness at a location 2mm from the center of the cornea. A significant change at the center of the epithelium relative to the paracentral region is unlikely. If it did occur, it would have been observed during routine slit lamp examination; therefore, a change in R2 is not a likely cause of the shift in refractive error. If topographic variations in stromal swelling response did take place, with more swelling occurring in the paracentral corneal regions, then this could have caused a sufficient change in R3. However, the paracentral swelling would have to be inwardly directed toward the retina. This is possible and could have been detected by topographic pachymetry. A drop in the refractive index of the anterior stroma could take place as a direct response to shifts in stromal hydration. If the refractive index of the anterior stroma decreased from 1.380 to 1.369, the average refractive index along the anteroposterior direction would drop from 1.376 to 1.371. Using the corneal refractive index-thickness-hydration models developed by the late Irving Fatt, 4, 5 a change in the average refractive index of the stroma could take place if the net hydration of the tissue increased from 3.73 to 4.17, which in turn would cause the stromal thickness to rise from 0.526 mm to 0.576 mm (i.e., an increase of 9.5%). Diffraction theory limits the resolving power of the keratometer to the order of 0.04 mm. 6 Therefore, R1 could reduce on average by 0.05 mm and remain undetected because the resolution of the autokeratometer was incapable of detecting such a small change in radius. Ideally, a topographic device that does not rely on the basic principles of keratometry and has the facility to gather information from within the central optical zone of the cornea should be used in association with a topographic pachymeter. Then the true source of the shifts in refraction could be isolated. Sudi Patel
Sudi Patel (Thu,) studied this question.