Table V.

Data structure of the first 14 principal components

PC	Variance (%)	Subject weight controls, mean (SD)	Subject weight patients, mean (SD)	t‐test (Controls vs. Patients)	Null hypothesis (P)	Max. load	Max. cluster size	Clusters (n)*
1	7.7	−0.106 (0.099)	0.114 (0.230)	3.149	0.0045	0.0128	498	22
2	7.2	−0.008 (0.061)	0.008 (0.294)	−0.195	0.8472	0.0133	796	16
3	6.5	−0.128 (0.139)	0.138 (0.173)	4.26	0.0003	0.0129	388	20
4	5.9	−0.026 (0.109)	0.028 (0.276)	−0.643	0.5268	0.0129	522	34
5	5.5	−0.052 (0.176)	0.057 (0.225)	−1.355	0.1886	0.0139	465	29
6	5.2	0.025 (0.152)	−0.027 (0.025)	0.631	0.5342	0.0142	722	15
7	5.1	0.005 (0.180)	−0.006 (0.235)	0.130	0.8977	0.0105	478	38
8	4.8	0.035 (0.161)	−0.038 (0.245)	0.883	0.3862	0.0121	809	16
9	4.4	0.014 (0.237)	−0.015 (0.170)	0.350	0.7293	0.0132	686	31
10	4.2	0.019 (0.209)	−0.021 (0.206)	0.484	0.6329	0.0135	505	27
11	3.9	0.005 (0.162)	−0.005 (0.249)	0.114	0.9099	0.0127	292	32
12	3.7	0.026 (0.253)	−0.029 (0.140)	0.662	0.5148	0.0150	811	19
13	3.6	0.024 (0.229)	−0.026 (0.180)	0.601	0.5536	0.0107	325	54
14	3.5	−0.007 (0.255)	0.008 (0.142)	−0.176	0.8621	0.0122	329	55

Σ(PC1–PC14) = 71.3.

^*Clusters composed of voxels exceeding 99.5% or less than 0.5% of all voxel values in the PC.

PC, principal component.