Effectiveness (or Not?) of cognitive rehabilitation in a person with Semantic Dementia

We tested the possible efficacy of relearning ecologically relevant information and its maintenance in Marian, a person affected by Semantic Dementia (SD). In Phase 1, 20 images, belonging to two different categories of objects (vegetables and fruit), have been chosen among the ones Marian did not recognize: in each category, 5 have been used for the 2 month training (twice a week) with explicit and implicit memory techniques while the other 5 haven't. Statistical analysis comparing trained-not trained words at the end of the rehabilitation period was not significant, even if a slightly better performance in rehabilitated words was evident. At 6 months follow-up no training gain was evident. In Phase 2, we performed a repeated baseline assessment and then we introduced a photo-vocabulary, which included 14 relatives' names. At baseline, Marian constantly denominated 1 relative only. In this second phase, Marian both followed structured rehabilitation sessions and rehearsed face-name associations at home. After two months Marian named 8/14 relatives (14/14 after phonemic cue). At 6 months follow-up, she spontaneously named 11/14 relatives. names (14/14 after phonemic cue). In the second training phase, Marian showed a clearer improvement. This better result could depend on the fact that relatives. identification was less affected than identification of different kinds of food, or on the intensiveness of training or on Marian's better motivation. The progression of the disease, however, stopped us from further treating Marian. Therefore, as shown by other Authors, a cognitive rehabilitation training based on relearning with explicit and implicit memory techniques seems to show a limited effectiveness in SD.