Table 3.

Factor loadings from the measurement model of the multivariate latent change model of white matter FA, g_f, memory, and processing speed (reaction times are recalculated so that higher values indicate better scores)

Factor	Variable	Estimate	SE	p value
FA	Genu of the corpus callosum	0.597	0.027	<0.001
	Splenium of the corpus callosum	0.344	0.033	<0.001
	Left arcuate fasciculus	0.653	0.024	<0.001
	Right arcuate fasciculus	0.613	0.028	<0.001
	Left anterior thalamic radiation	0.645	0.026	<0.001
	Right anterior thalamic radiation	0.661	0.025	<0.001
	Left rostral cingulum	0.600	0.027	<0.001
	Right rostral cingulum	0.552	0.028	<0.001
	Left uncinate fasciculus	0.679	0.025	<0.001
	Right uncinate fasciculus	0.669	0.025	<0.001
	Left inferior longitudinal fasciculus	0.523	0.030	<0.001
	Right inferior longitudinal fasciculus	0.488	0.030	<0.001
gf	Matrix reasoning	0.648	0.022	<0.001
	Block design	0.669	0.022	<0.001
	Digit span backward	0.586	0.025	<0.001
	Letter-number sequencing	0.658	0.023	<0.001
	Digit-symbol substitution	0.306	0.052	<0.001
	Symbol search	0.378	0.047	<0.001
Speed	Digit-symbol substitution	0.527	0.049	<0.001
	Symbol search	0.429	0.045	<0.001
	Simple reaction time	0.496	0.028	<0.001
	Choice reaction time	0.795	0.023	<0.001
	Inspection time	0.522	0.026	<0.001
Memory	Logical memory	0.614	0.033	<0.001
	Spatial span	0.546	0.029	<0.001
	Verbal paired associates	0.560	0.034	<0.001

Due to the invariance assumption, factor loadings and indicator intercepts were identical at waves 1 and 2. Digit-symbol substitution and Symbol search loaded on both g_f and speed. The full model diagram is shown in Figure 2. Results taken from the model are adjusted for age at testing and sex. All estimates are standardized.