Corneal thickness in a population-based, cross-sectional study: the Tehran Eye Study.

PURPOSE

To determine the distribution of corneal thickness at 6 different points in an Iranian population and find any possible relationship between the thickness values and some independent variables.

METHODS

As part of the "Tehran Eye Study," 410 people (800 eyes) aged 14 years and older, residing in the first 4 municipality areas of Tehran, were selected through a stratified random cluster sampling method. The participants were examined with Orbscan II to measure the corneal thickness at the central, thinnest, and 4 peripheral points of the cornea. In addition, the relationships between central corneal thickness (CCT) values and variables of age, sex, refractive error, pupil diameter, corneal diameter, anterior chamber depth (ACD), and body mass index were assessed. The design effect was also considered, and all estimates were standardized for age and sex. The reported results pertain to the right eyes only.

RESULTS

The mean thickness (+/-SD) at the central and thinnest points was 555.6 +/- 39.9 and 550.7 +/- 40.6 microm, respectively. The minimum corneal thickness was below 500 microm in 9.6% of the participants. The superior area showed the largest thickness, and the thinnest point was most commonly in the inferotemporal quadrant. The mean "distance" and "thickness difference" between the central and thinnest points were 0.52 +/- 0.31 mm and 4.85 +/- 6.2 microm, respectively, and they showed a significant direct relationship (r = 0.708 and P < 0.001). Neither sex had a significant relationship with thickness values at different areas. Participants younger than 20 years, compared with those older than 20, had significantly thicker corneas (P = 0.001 for the central and thinnest points, P < 0.001 for peripheral areas). In the multiple regression model, age and ACD showed significant inverse correlations with CCT.

CONCLUSIONS

The mean CCT in the present study was 555.6 microm, which is relatively higher in comparison with that in other ethnic groups. In the multiple regression model, age and ACD were correlated with the CCT, but refractive errors showed no correlation.