Table 2 Estimated correlation coefficients and their 95% confidence intervals in the random intercepts and random slopes, from the primary analysis and three sensitivity analyses using bivariate linear mixed models
From: Estimating correlation between multivariate longitudinal data in the presence of heterogeneity
Models | Correlations between random intercepts | Correlations between random slopes |
|---|---|---|
Primary analyses: without adjusting heterogeneity | 0.278 (0.121, 0.420) | 0.579 (0.349, 0.810) |
Sensitivity #1: adjusting the severity of vision damage at diagnosis (MD < −5 dB vs. MD ≥ −5 dB) | 0.267 (0.090, 0.452) | 0.519 (0.285, 0.753) |
Sensitivity #2: excluding those who labelled as “Rapid Progression”. | 0.217 (0.077, 0.354) | 0.289 (0.010, 0.593) |
Sensitivity #3: estimating class-specific intercepts and slopes | 0.145 (−0.037, 0.324) | 0.106 (−0.255, 0.490) |