Impaired access to semantic memory for the cognition of geographic space in Alzheimer's disease.

This study explores the possibility to capitalize from a widely used semantic fluency test, in order to investigate aspects of topographical space representation, still poorly studied in neurodegenerative diseases. Twenty-six patients with mild Alzheimer's disease (AD) and 13 healthy control (CTR) subjects underwent neuropsychological assessment at baseline (T0) and about 2 years later (T1). The cities named during category verbal fluency test ("names of cities") were marked on a map, and the polygon perimeter obtained by joining the external points was computed. Mini-mental state examination (MMSE) score, number of cities named and perimeter length were compared between T0 and T1, both within-group and between groups. MMSE score and number of cities significantly differed between AD and CTR both at T0 and at T1; perimeter length differed significantly only at T1. Between T0 and T1, all the three parameters significantly decreased in AD, while they were substantially unchanged in CTR. Besides a reduction of semantic verbal fluency, there seems to be a 'restriction' of mental geographic space representation already in mild AD. These findings should be confirmed and exploited by further ad hoc investigations.