Emotional memories are not all created equal: evidence for selective memory enhancement

Abstract

Human brain imaging studies have shown that greater amygdala activation to emotional relative to neutral events leads to enhanced episodic memory. Other studies have shown that fearful faces also elicit greater amygdala activation relative to neutral faces. To the extent that amygdala recruitment is sufficient to enhance recollection, these separate lines of evidence predict that recognition memory should be greater for fearful relative to neutral faces. Experiment 1 demonstrated enhanced memory for emotionally negative relative to neutral scenes; however, fearful faces were not subject to enhanced recognition across a variety of delays (15 min to 2 wk). Experiment 2 demonstrated that enhanced delayed recognition for emotional scenes was associated with increased sympathetic autonomic arousal, indexed by the galvanic skin response, relative to fearful faces. These results suggest that while amygdala activation may be necessary, it alone is insufficient to enhance episodic memory formation. It is proposed that a sufficient level of systemic arousal is required to alter memory consolidation resulting in enhanced recollection of emotional events.

Figure 1.

(A) Pre-experimental ratings of valence and intensity (i.e., arousal) of emotional experience during viewing faces and scenes. Valence is along the y-axis and arousal the x-axis. Values are presented in z-transformed units. Arrows denote the relative distance between fear and neutral faces (solid line) and negative and neutral scenes (dotted line) according to valence and arousal. Negative scenes used in the experiment were rated as both more arousing and negative. (B) Experimental ratings of emotional intensity experienced during encoding of faces and scenes. Negative scenes were reported as resulting in the most intense emotional response. Negative faces = fear expressions.

Figure 2.

Recognition memory for scenes. Proportion corrected (A) “remember” and (B) “familiar” responses for negative and neutral scenes at different test delays.

Figure 3.

Recognition memory for faces. Proportion corrected (A) “remember” and (B) “familiar” responses for fear and neutral faces at different test delays.

Figure 4.

Comparison of degree of forgetting for scenes and faces from day 1 (15 min) to 2 wk. Proportion corrected recognition difference score (2 wk minus day 1) for scenes and faces.

Figure 5.

Comparison of sympathetic arousal and recognition memory. (A) Galvanic skin responses and (B) Proportion corrected “remember” responses for faces and scenes in Experiment 2.