This study replicated prior research showing that greater phasic HR deceleration during picture viewing was greater for subsequently remembered than forgotten stimuli (Buchanan et al., 2006
). These data suggest that a larger orienting response (as reflected in greater HR deceleration in response to a picture) is related to deeper encoding and better subsequent memory for the stimulus. We also found that individuals with higher tonic HR across the entire picture encoding epoch showed better subsequent memory for emotionally arousing pictures. Furthermore, we found that higher tonic HR predicted greater phasic HR deceleration during viewing of emotionally arousing (but not neutral) pictures. This finding suggests that individuals who were more autonomically aroused (as measured by mean HR) showed stronger orienting responses to emotionally arousing pictures (as measured by HR deceleration). In sum, those individuals with elevated HR and greater orienting responses subsequently remembered more emotionally arousing pictures than individuals with lower mean HR. Elevated tonic HR may reflect a state of physiological and emotional arousal, which is associated with increased attentional processing of emotionally salient material, reflected in enhanced phasic slowing of HR and deeper encoding of emotionally salient information.
In our data, deceleration was greatest for unpleasant pictures and least for neutral pictures. However, the difference in HR deceleration for remembered vs. forgotten pictures is stable across emotion category. These data imply that greater orienting increases the likelihood that a stimulus will be later remembered, regardless of whether the stimulus is emotionally arousing. On the other hand, the relations among tonic HR elevation, phasic HR deceleration, and subsequent memory performance were observed only for emotionally arousing pictures. Thus, autonomic arousal appears to predict the degree of orienting and subsequent memory only for emotionally arousing events. Other factors (e.g., novelty) have been shown to determine the degree of orienting to neutral stimuli (Cook & Turpin, 1997
At first glance, it may seem paradoxical that individuals with elevated tonic HR are more likely to show greater HR deceleration during processing of emotional stimuli. However, neurogenic control of HR is complex. The two-component model of attention proposed by Porges (1992)
can aid in the understanding of these relationships. This model partitions the physiological response to a stimulus into two components, reactive and sustained. The reactive portion involves a short-latency vagally-mediated heart rate deceleration occurring within one second of stimulus perception. This is followed by a longer latency period that lasts four to five seconds where heart rate response varies as a function of the stimulus. The coupling of vagal withdrawal and sympathetic activation causes this longer latency heart rate acceleration in response to important or intense stimuli whereas the response to mild stimuli involves vagal activation to cause more prolonged heart rate deceleration. The model also describes a longer duration parasympathetically-mediated response. When sustained attention is required, heart rate stabilizes and heart rate variability diminishes, which facilitates information processing. Further, the relationship between the sympathetic and parasympathetic branches of the autonomic nervous system in terms of controlling heart rate is one of “accentuated antagonism” such that sympathetic activity is partly a function of background vagal activity as recently demonstrated non-invasively in humans by Uijtdehaage and Thayer (2000)
Furthermore, autonomic activation is associated with priming of attention to stimuli promoting avoidance (Lang, Bradley, & Cuthbert, 1990
). Phasic bradycardia is the typical HR response to passively viewed emotionally arousing (particularly aversive) stimuli, and represents attentional orienting to a stimulus (Bradley et al., 2001
; Bradley & Lang, 2000
). Thus, greater tonic HR may represent a priming of attentional systems, and therefore may promote greater heart rate deceleration to emotionally arousing stimuli.
Porges’ research has examined individual differences in vagal tone and information processing (reviewed in Porges, 1992
). Their research suggests that individuals with higher baseline vagal tone show greater HR reactivity to stimulation as well as better sustained attention. While our study did not measure resting vagal tone, our data are consistent with Porges and colleagues’ data in that the individuals with the more extreme task-related HR responses (i.e., greater phasic slowing and greater tonic HR during stimulus viewing) were the individuals with the deepest encoding and best subsequent memory performance.
Past research implicates additional centrally mediated relations among arousal, orienting, and memory. For instance, the locus coeruleus-norepinephrine (LC-NE) system maintains behavioral and neuronal states that determine the level of arousal of the organism (spanning from sleep to quiet waking to arousal; Berridge & Waterhouse, 2003
). The LC-NE system also plays an important role in modulating the orienting response to salient stimuli. In addition to HR deceleration, the P300 component of the human event-related potential (ERP) serves as an index of an OR (Palomba et al., 1997
). P300-like components are also observed in monkeys. Pharmacological manipulation of LC firing as well as bilateral LC lesions alter the magnitude of these P300-like components in monkeys (Pineda, Foote, & Neville, 1987
). Thus, the LC-NE system appears to modulate these P300-like components, and may be an important modulator of the orienting response. The LC-NE system therefore appears to modulate states of arousal as well as processing of salient sensory information, which suggests that the level of arousal (even within states of attentive waking) should predict degree of orienting. Our data showing a positive relation between heightened tonic HR (reflecting a heightened state of arousal) and greater phasic HR deceleration (reflecting greater orienting) are consistent with this view.
Noradrenergic activation of the basolateral nucleus of the amygdala may partly determine positive relations among arousal, orienting to emotional stimuli, and emotional enhancement of memory. Amygdala activation during memory formation underlies the superiority of memory for emotional information (Cahill, Haier, Fallon, Alkire, Tang, Keator, Wu, & McGaugh, 1996
; Canli, Zhao, Brewer, Gabrielli, & Cahill, 2000
). In particular, noradrenergic activation in the basolateral nucleus of the amygdala is necessary for memory enhancement associated with emotionally provocative events and stimuli (Liang, Juler, & McGaugh, 1986
; McGaugh, 2000
). In sum, norepinephrine plays a crucial role in the emotional enhancement of memory (McGaugh, 2000
), and contributes to long-term alterations in synaptic strength, gene transcription, and other processes necessary for learning (Berridge & Waterhouse, 2003
Furthermore, the amygdala plays a role in central regulation of HR, as projections from the central nucleus of the amygdala innervate the nucleus tractus solitarius (Brownley, Hurwitz, & Schneiderman, 2000
). Noradrenergic activation of the amygdala may be partly responsible both for alterations in HR and heightened memorability of emotional information. Future research examining central nervous system activation is needed to further test whether amygdala activation mediates the relations between HR and memory.
Most likely, HR alterations that accompany emotionally laden events do not play a major causal role in the greater memorability (Öhman, Hamm, & Hugdahl, 2000
). HR and other peripheral alterations likely reflect
(rather than cause) central nervous system processes involved in the strength of encoding stimuli and events. Indeed, manipulation of central but not peripheral adrenergic systems is necessary for enhancement of memory for emotional material (van Stegeren et al., 1998
However, various data suggest that central mechanisms may not be sufficient, and that autonomic arousal is necessary for emotional enhancement of memory. Cardiovascular responses may play a causal role in cortical sensitivity to stimuli via afferent connections from the heart to the brain (Lacey, 1967
; see Öhman et al., 2000
for review). For instance, an intact nucleus tractus solitarius (NTS), which can be viewed as an integration center for afferent autonomic signals, is necessary for arousal-related memory improvement (Miyashita & Williams, 2003
; Williams & McGaugh, 1993
). Stimulation of afferents to the NTS during memory formation improves retention (Clark, Naritoku, Smith, Browning, & Jensen, 1999
; Clark, Smith, Hassert, Browning, Naritoku, & Jensen, 1998
). Thus, it appears that peripheral activation plays an important contributory role in emotional enhancement of memory, and is not merely a proxy measure for central mechanisms (see also Anderson, Yamaguchi, Grabski, & Lacka, 2006
). Pharmacological blockade of parasympathetic, sympathetic, and dual blockade of parasympathetic and sympathetic influences on cardiac variation (e.g., Berntson, Cacioppo, & Quigley, 1994
) could be used to test whether blocking stimulus-related variation in HR alters the differential memorability of emotionally laden vs. neutral stimuli.
Research has shown that alterations in arousal and emotion-related information processing are associated with the symptomatology of depression and anxiety. For instance, Shalev, Sahar, Freedman, Peri, Glick, Brandes, Orr, and Pitman (1998)
showed that trauma victims with higher heart rates in the hospital emergency department following a traumatic incident are more likely to develop PTSD than trauma survivors with lower emergency room heart rates. Shalev and colleagues (1998)
also showed that higher emergency department heart rates predicted a higher rate of intrusive memories of the traumatic incident at 4 months following the event. Furthermore, alterations in emotion-related information processing and heightened memory for negative information is often observed in depression (Mineka, Rafaeli, & Yovel, 2003
; Shestyuk, Deldin, Brand, & Deveney, 2005
). Measuring HR deceleration and other electrophysiological indices of information processing (e.g., ERPs) during processing of negative vs. neutral stimuli may serve as an important physiological index of alterations in emotion-related information processing in future research investigating PTSD, depression, and other forms of psychopathology. In addition, electrophysiological measures during cognitive processing of emotional material may serve as important indicators of treatment progress in cognitive behavioral therapy. Future research on clinical samples should further investigate both central and peripheral responses to emotionally evocative material and their relationships to memory and symptomatology.
Future research must also examine potential sex differences in the association between HR and emotional memory. Because the current study was conducted on men only, the results presented here may not be generalizable to women. Women often show greater enhancement of memory for emotional information than men (Canli, Desmond, Zhao, & Gabrieli, 2002
). In addition, women generally show greater resting HRs than men (Taneja, Windhagen Mahnert, Passman, Goldberger, & Kadish, 2001
). It should be determined whether women’s higher resting HRs are related to priming of attentional systems, and enhanced orienting and/or emotional memory.
In the sample presented here, HR and cortisol responses to the stressor were not related to memory performance and do not appear to alter the relations between HR during encoding and subsequent memory. However, because all subjects participated in the stressor immediately following encoding the results presented herein may depend on exposure to emotional arousal and stress immediately following encoding. Future research, in which exposure to a stressor is experimentally manipulated, is needed in order to determine whether an emotionally arousing stressor immediately following encoding alters relations between cardiac activity during encoding and subsequent memory.
Because emotional pictures are more memorable and cause greater HR deceleration than neutral pictures, we predicted that the relations between cardiac activity and memory enhancement would be most pronounced for emotionally-laden compared to neutral pictures. We found that elevated mean HR across the entire encoding epoch predicted better memory performance for emotionally arousing pictures. Phasic HR deceleration during viewing of individual pictures was greater for subsequently remembered than forgotten pictures across all three emotion categories. Tonic HR and phasic HR deceleration were associated, but only for emotionally arousing pictures, such that those individuals with greater tonic HR also showed greater HR decelerations during viewing of emotionally arousing pictures. Results suggest that tonic elevations in HR are associated both with greater orienting and heightened memory for emotionally arousing stimuli. These data imply that greater orienting predicts the memorability of an event, and that autonomic arousal may be associated with enhancement of the orienting response and depth of encoding during an emotionally arousing event. Future research is needed to examine the central nervous system mediators of these effects as well as their ramifications for treatment of emotion-related information processing alterations in psychopathology.