Table 3. Linear modelling of change in test scores with volumetric and clinical data.

	Change in test score
	Delayed Story Recall			List Learning			Immediate Story Recall
Predictors	β	Standard error	p-value	β	Standard error	p-value	β	Standard error	p-value
Left Thalamus	9.37	4.80	0.06	5.57	3.76	0.15	4.97	4.72	0.30
	Adjusted model			Adjusted model
Left Thalamus	7.12	4.90	0.16	8.84	3.74	0.02
Duration of Symptoms	0.12	0.09	0.19	-0.16	0.67	0.02
Age	-0.08	0.06	0.18
Right Thalamus	9.62	5.22	0.07	4.98	4.08	0.23	4.36	5.11	0.40
	Adjusted model			Adjusted model
Right Thalamus	6.86	5.39	0.21	4.16	8.36	0.05
Duration of Symptoms	0.12	0.09	0.19	-0.16	0.07	0.03
Age	-0.08	0.06	0.19
Left Hippocampus	16.90	8.80	0.06	12.10	6.82	0.09	11.87	8.58	0.18
	Adjusted model			Adjusted model
Left Hippocampus	12.58	8.92	0.17	16.13	6.72	0.02
Duration of Symptoms	0.13	0.09	0.15	-0.15	0.07	0.03
Age	-0.07	0.06	0.23
Right Hippocampus	9.06	9.79	0.36	11.76	7.21	0.11	8.43	9.14	0.36
	Adjusted model			Adjusted model
Right Hippocampus	4.25	9.76	0.67	15.91	7.12	0.03
Duration of Symptoms	0.15	0.09	0.11	-0.15	0.07	0.03
Age	-0.08	0.06	0.19

Table to show the results of linear modelling investigating the relationship of test score change with volumetric and clinical data. The absolute test score change was used as the outcome variable. In the unadjusted model, the volume of the structure of interest was used as the predictor. In the adjusted model, the volume of the structure of interest and clinical variables, found to be of interest (see S1 Table) for the test in question were chosen as predictors. Significant p-values are highlighted in bold. Duration of symptoms was found to be a significant predictor of change in list learning score. Left thalamic and hippocampal and right hippocampal volumes were also found to be significant predictors of list learning score change in the adjusted models. Age was not a significant predictor of score change in any of the neuropsychological tests.