Skip to main content
. 2012 Jan;119(1):84–89. doi: 10.1016/j.ophtha.2011.06.045

Table 6.

Univariate and Multivariate Generalized Linear Latent and Mixed Model Logistic Regression Models for Central Corneal Opacity

Variable Pupil Obscuring/Central Corneal Opacity
OR§ 95% CI P Value
Univariate analysis
 Gender
  Male 1
  Female 0.81 0.61–1.08 0.154
 Advancing age 1.52 1.39–1.67 <0.005
 Epilation status
  No epilation 1
  Unsuccessful epilation 1.18 0.92–1.52 0.200
  Successful epilation 0.57 0.40–0.81 0.002
 Entropion grade
  0 1
  1 1.54 1.12–2.12 0.008
  2 3.55 2.54–4.96 <0.005
  3 8.36 5.34–13.09 <0.005
  4 44.98 26.20–77.20 <0.005
Multivariate logistic regression model
 Advancing age 1.63 1.48–1.80 <0.005
 Unsuccessful epilation 0.87 0.66–1.13 0.29
 Successful epilation 0.61 0.43–0.88 0.007
 Entropion grade
  0 1
  1 1.68 1.22–2.31 0.002
  2 4.15 2.93–5.86 <0.005
  3 9.49 6.05–14.86 <0.005
  4 45.07 26.29–77.26 <0.005

CI = confidence interval; OR = odds ratio.

For the generalized linear latent and mixed model outcome, corneal opacity subdivided into 4 groups: (1) CO 0/1 (no or peripheral opacity), (2) CO 2a/2b (central 4 mm, not overlying visual axis), (3) CO 2c/2d (overlying visual axis, but not obscuring all central 4 mm), and (4) CO 3 (overlying whole central 4 mm). Phthisis (CO 4) excluded.

Advancing age group has a linear association with increasing corneal opacity.

Entropion grades 1–4 are compared with entropion grade 0.