Table 2.

IPL and GCL Thickness Measurement for All Study Eyes and After Excluding Superpixels Where the GCL Thickness was <30 μm According to Eccentricity

	All Superpixels			Superpixels With GCL thickness >30 μm
	Healthy	Glaucoma	P*	Healthy	Glaucoma	P*
IPL
All points	34.4 ± 5.9	27.6 ± 6.1	<0.001	35.0 ± 5.7	33.2 ± 5.0	0.406
1.5° eccentricity, μm	34.1 ± 3.4	27.8 ± 5.9	<0.001	34.1 ± 3.4	34.1 ± 3.0	0.686
4.5° eccentricity, μm	39.6 ± 4.2	31.2 ± 6.5	<0.001	39.6 ± 4.2	35.2 ± 5.1	<0.001
7.5° eccentricity, μm	31.1 ± 4.9	25.2 ± 4.6	<0.001	31.6 ± 4.8	30.1 ± 3.6	0.465
GCL
All points	41.3 ± 8.4	29.5 ± 10.1	<0.001	42.8 ± 7.5	39.8 ± 6.7	0.016
1.5° eccentricity, μm	36.9 ± 4.3	26.9 ± 9.7	<0.001	38.2 ± 4.7	37.8 ± 4.6	0.380
4.5° eccentricity, μm	49.7 ± 5.6	34.6 ± 11.4	0.013	49.8 ± 5.4	42.5 ± 7.4	<0.001
7.5° eccentricity, μm	37.2 ± 6.3	26.8 ± 7.9	<0.001	38.9 ± 5.4	36.8 ± 4.6	0.492

The latter is presented to be consistent with multivariable analyses predicting IPL thickness in all eyes; 1445 of 3960 (36.4%) superpixels had GCL thickness <30 μm.

^*P values are based on multivariable linear mixed regression analyses entering variables with P < 0.15 in the model and accounting for the correlation of superpixels in each eye.