The contributions in this volume, and numerous other journal articles and chapters, provide unequivocal evidence that cognitive functions decline across the adult lifespan Importantly, though, some cognitive functions are more affected than others For example, in recent work from the Betula project (Nilsson eta!, 1997), we contrasted episodic and semantic long term memory (Nyberg eta!, 2003) It was found that episodic memory performance deteriorated gradually from middle age through young-old age to old-old age By contrast, semantic memory performance increased from middle-age to young-old age, and the old-old participants performed at a level comparable to the middle aged (Figure 7 la) Furthermore, within the domam of episodic memory, increasing age was associated with a greater reduction of pei form ance on measures of recall compared to measures of recognition (Figure 7 ib) These results provide evidence that episodic long-term memory is more age-sensitive than semantic long-term memory, and that recall is more age-sensitive than recognition.
These age-related long-term memory changes can be related to patterns of data in working-memory tasks (Gick, Crajic, and Morris, 1988) Working-memory tasks differ with regard to their demand on executive processing and it has been found that age differences are small when the executive demands are low, and substantial wlten such demands are high (see Morris, Gick and Craik, 1988). Many cognitive tasks require working memory functions to a smaller or greater extent, and it has been shown that working memory capacity accounts for a considerable portion of the variance in long-term memory tasks (Hultsch, I-Iertzog, and Dixon, 1990; Park eta!., 1996). Relatedly, the relationship between age and episodic memory has been found to be mediated by proficiency of executive functioning (Troyer, Graves, and Cullum, 1994).
Oxford University Press, Oxford , 2004. 135-159 p.