The complex picture of emotional effects in WM of DAT leads us to several considerations. First, it is important to further our knowledge about the relationship between emotion and WM measures. In fact, WM tasks are generally grouped in recall-based more “active” tasks such as the classical Daneman and Carpenter's Reading/Listening Span [17
], the Operation Span [18
], the active visuospatial task [37
], and recognition-based “more passive” tasks such as n
-back, recency-probe paradigms, matching-to-sample, and binding tasks to cite only a few [38
]. All of these tasks, however, differ in the type and the amount of processing required as well as in the nature of the information to be temporarily maintained.
An interesting attempt to reconcile different pattern of results with emotion, WM, and DAT would be to use different tasks and highlight the role of specific WM functions involved and specialized for affective processing. In fact, the majority of studies did not investigate the ability of actively maintaining and storing information to achieve the task goals as classical WM tasks require.
Second, the majority of studies with DAT did not carefully control for different affective dimensions of study material. For example, some studies [10
] used high- versus medium-arousal valenced pictures, while others [16
] used arousing versus nonarousing items. Still others [9
] did not mention whether arousal was controlled. In sum, it is not clear whether emotional enhancement effects in DAT patients are mainly due to arousal effects, as suggested by Nashiro and Mather [16
], valence effects, or both. From the WM studies, it is likely that WM for high arousing and negative valenced stimuli is preserved in DAT, but we need further studies to assess the role valence in WM modulation.
Third, group characteristics may be called on to explain discrepancies across studies. In fact, differences among the samples of DAT patients may also explain the mixed findings of whether DAT patients demonstrated an emotional benefit in their WM performance. These differences may be linked to different stages of deterioration in the brain regions involved in emotion and cognition processing.
Finally, the majority of studies reviewed here used affective visuospatial material. A long-term memory study by Kensinger and colleagues [39
] that used verbal material, in fact, did not detect any emotional benefit. This is in line with Mammarella et al.'s study [19
] that used a verbal WM task and did not detect an emotional effect with DAT patients. Thus, differences in the type of affective material may be relevant as well. Pictures, in fact, may create richer memory traces that may generate larger emotional enhancement effect not detectable with affective words.
A concluding thought must go to a motivational explanation of this complex picture of data as motivation towards emotional goals has been shown to be crucial to understanding the trajectory of the aging mind towards Alzheimer's disease [40
]. According to this approach, lack of motivation is considered as one of the main characteristics of dementia. Given that motivation towards meaningful emotional goals needs recruitment of cognitive resources towards emotion processing [41
], results typically obtained in WM emotional tasks with DAT may depend on the lack of attentional and WM processes motivated towards emotion processing. Importantly, this hypothesis may highlight how different levels of motivation towards emotional goals may lead to different degree of emotional effects in WM also in DAT.