BACKGROUND: Episodic memory impairment can be subdivided into two subtypes, encoding failure vs. retrieval deficit. We defined two subtypes of amnestic mild cognitive impairment (aMCI) according to recognition performance on the memory test: aMCI with encoding failure or aMCI-E and aMCI with retrieval deficit or aMCI-R. We hypothesized that compared to aMCI-R, subjects with aMCI-E are more likely to convert to Alzheimer's disease, as encoding failure suggests that medial temporal lobes are affected early in the disease process. We also investigated whether aMCI-E can be a predictor for progression to AD. METHODS: Using the Alzheimer's Disease Neuroimaging Initiative (ADNI) database, a total of 397 aMCI subjects were included. APOE genotype, cerebrospinal fluid (CSF) fluid biomarkers, and a set of neuropsychological measures were also collected. Unadjusted and adjusted odds ratios were calculated to predict the conversion to AD dementia. The Spearman's rs test was used to measure the degree of correlation between aMCI subtypes and the prognostic factors for progression to AD. RESULTS: Among the 397 subjects, 209 (52.6%) subjects were classified into aMCI-E and 188 (47.4%) into aMCI-R. One hundred two (48.8%) subjects with aMCI-E and 57 (30.3%) of those with aMCI-R progressed to AD (unadjusted odds ratio=2.19 with 95% confidence interval [CI] 1.45-3.31) over 3 years. However, when adjusted odds ratio by a logistic regression was calculated, probability value of aMCI-E disappeared with the odds of conversion by of 1.47 (95% CI 0.89-2.43). There were statistically significant correlations between aMCI-E subtype and MMSE, CDR Memory, RAVLT delayed recall, CSF biomarkers, and genotypes. CONCLUSIONS: This analysis did not show that aMCI-E is an independent prognostic factor to predict the progression to AD. However, this subtype significantly correlates with other prognostic factors for progression. This may suggest that aMCI-E might be a later stage of aMCI and aMCI-R an earlier stage which might be a better target than aMCI-E for therapeutic intervention. Further studies are needed to validate this conjecture.
Alzheimer Disease ; Apolipoproteins E ; Biomarkers ; Dementia ; Deoxycytidine ; Disease Progression ; Genotype ; Logistic Models ; Memory ; Memory, Episodic ; Mild Cognitive Impairment ; Neuroimaging ; Odds Ratio ; Temporal Lobe