Table 4.

Prognostic factors for training improvement in verbal long-term memory

Study	Test for outcome assessment	Dependent variable	Prognostic factor
Multiple regression
			Age	Education	Sex	Neuropsychology	Imaging	Others
O'Hara et al. [9]	Number of words correctly recalled.	Post-test scores Pre-test and change scores were integrated in regression.	↓	↑		Gain scores following training ↑ *		Length of training (short vs. long) ↑ Reported use of mnemonic at follow-up ↑ * Type of pre-training (standard vs. comprehensive) ↓ Pre-training ↑ *
Brooks et al. [8]	Proper name recall task	Post-test scores	↑ *			Pre-test score →*		Pre-training * Length → Length of training → Pre-training →
McDougall et al. [40]	RBMT	Change score Relative gains from beginning to end of training	↑	x				Ethnical group x
Park et al. [7]	Elderly verbal learning test, delayed recall However, results are reported for “cognitive function” as outcome measure, which is not clearly defined	Change score Post-pre	→	↓*	→	Pre-test scores of neuropsychological tests (Digit Span Test, Spatial Span Test, Categorical Fluency Test, short version of Boston Naming test) →
Pesce et al. [20]	RVLT	Change score Post-pre						Change in dROMs ↓ Change in BAP ↑
Correlation analysis
Leahy, Ridout, Mushtaq et al., [25]	Autobiographical memory specificity	Change score						Independence Depression Functional limitations Memory specificity
Andrewes et al. [26]	Laboratory Prospective Memory Assessment Everyday Prospective Memory Assessment	Change score				NART → Warrington Forced Choice Recognition → RAVT →		Mattis dementia scale → Depression →
Anschutz et al. [27]	Free recall of 2 lists Recognition of 2 lists	No clear reporting.	No clear reporting.
Hill et al. [38]	Improvement in list recall	Change scores				MMSE ↑
Leahy, Ridout, and Holland, [24]	Autobiographical memory specificity.	Change scores				Baseline cognitive flexibility ↑
Group comparisons (ANOVA, t test)
McDougall et al. [40]	RBMT	Pre-test and Post-test scores calculated in an ANOVA.	x	x				Ethnicity x
O’Hara et al., [43]	List-learning test	Pre-test and Post-test scores calculated in an ANOVA.						Apolipoprotein E4 ↓
Mixed models
/

Studies are sorted according to the statistical method used for obtaining the prognostic factors

ANOVA analysis of variance, MMSE Mini Mental State Examination, NART National Adult Reading Test, RAVL Rey Auditory Verbal Learning Test, RBMT Rivermead behavioural memory test, RVLT Rey Auditory Verbal Learning Test, dROMs reactive oxygen metabolites derivative compounds, BAP antioxidant levels; ↑ the higher the prognostic factor, the higher the improvement/positive correlation; ↓ the lower the prognostic factor, the higher the improvement/negative correlation; → no direction of effect reported; * significant; x unclear reporting