Edward Pace-Schott - US grants

DESCRIPTION (provided by applicant): The proposed experiments will explore the role of sleep in generalizing memory for the extinction of conditioned fear. We recently discovered this generalization effect in an unfunded but published pilot study. We now wish to further develop this line of research within the R21 grant mechanism. Extinction training represents the neurobehavioral basis of exposure therapy, which is a first-line treatment for post-traumatic stress disorder (PTSD) as well as other anxiety disorders. Generalization extends extinction learning from specifically extinguished cues to other, similar stimuli. Such generalization is essential to successful exposure therapy because the reduction of fearful responding to specific cues, achieved in treatment, must then be extended to the multiplicity of related stimuli encountered outside the therapist's office. Our research has demonstrated that normal overnight sleep promotes generalization of extinction learning. Conditioned responding was established to two previously neutral color stimuli using an electric shock, and skin conductance response (SCR) served as the dependent measure of conditioned responding. One of the CS+'s was then extinguished (CS+E) whereas the other remained unextinguished (CS+U). After a 12-hour delay with intervening sleep, not only did SCR remain reduced following presentations of the CS+E (extinction recall) but it was also reduced following CS+U presentations (extinction generalization). However, such generalization was not seen when the delay contained only waking. The proposed project will first confirm that it is indeed sleep that produces generalization by examining time-of-day factors that potentially confound this previous observation. A 24-hour delay will now follow conditioning and extinction after which the Sleep-first group, trained in the evening, will immediately sleep then spend a normal day's waking whereas the Wake-first group, trained in the morning, will begin their 24 hours with waking. Thus, the Sleep-first group will now be tested at the time-of-day that poorer extinction generalization was previously observed and the Wake-first group at the time-of-day when generalization was best. Sleep physiological correlates of memory consolidation are widely replicated phenomena. We therefore next propose to examine polysomnographic correlates of extinction generalization. Additionally we propose to quantify phasic events in sleep that have been associated with memory phenomena including rapid-eye-movement density in REM sleep and sleep spindle density in Stage 2 non-REM sleep, and examine their association with extinction generalization. PUBLIC HEALTH RELEVANCE: The proposed work will explore relationships between sleep and learning not to fear previously threatening stimuli. Results may have clinical implications for timing of exposure therapy sessions relative to sleep in a manner that maximizes generalization of therapeutic extinction learning in PTSD and other anxiety disorders. Additionally, results may suggest behavioral interventions and pharmacological strategies that can be applied to post-therapy sleep in order to preserve or enhance those aspects of sleep most beneficial for extinction generalization.

DESCRIPTION (provided by applicant): Widely replicated studies demonstrate that sleep can enhance memory consolidation. Potential clinical applications of such findings have only begun to be explored. We have recently shown that nocturnal sleep following simulated exposure therapy for Specific Phobia (spiders) reduced both psychophysiological and self-reported reactivity when participants were re-exposed to the same and to novel spider stimuli. The proposed research will extend these findings to a more debilitating and clinically important anxiety disorder, Social Anxiety Disorder. We will examine whether post-exposure naps can be used to strengthen therapeutic extinction memories formed during exposure exercises used in the behavioral treatment of this disorder. A total of 32 individuals with diagnosed DSM-IV Social Anxiety Disorder will be enrolled in an exposure-based group treatment for this disorder. Eight successive therapy groups of 4 patients each will be offered during the 2-year funding period. The third and fourth sessions of this validated 5-week/5-session treatment will involve each participant delivering a speech on a topic individually chosen to elicit significant social anxiety Following both of these sessions, all 4 participants will go to the nearby Massachusetts General Hospital sleep laboratory where 2 will be given a 2-hour sleep opportunity with polysomnographic (PSG) monitoring and the other 2 will be similarly instrumented but undergo 2 hours of monitored quiet wakefulness (prior to session 3 participants will be randomized to the sleep or wakefulness arm). Before beginning treatment and within several days following the final treatment session, all participants will be individually assessed for social anxiety symptoms using standardized self-report instruments. At these same times, they will undergo a Trier Social Stress Test (TSST) modified for continuous psychophysiological monitoring that also includes repeated Subjective Units of Distress (SUDS) self report and sampling for salivary cortisol. In addition to laboratory PSG, ambulatory monitoring of home sleep with actigraphy and sleep diaries will take place at pre-treatment baseline and during the last 3 weeks of treatment. We hypothesize that those individuals allowed a 2-hour sleep opportunity following exposure sessions, compared to those who remained quietly awake, will show greater questionnaire- based clinical improvement as well as lesser psychophysiological and SUDS reactivity during the modified TSST. We further hypothesize that characteristics of sleep quality and architecture during naps, specifically durations of total sleep, REM sleep and slow-wave sleep as well as REM continuity, will predict clinical improvement and diminished TSST reactivity in those who napped. To help ensure that observed sleep effects are attributable to the two 2-hour sleep opportunities, we will control for actigraph and diary-measured sleep quality during treatment. Positive results will provide the first proof-of-principle for sleep augmentation of exposure therapy for a clinically significant anxiety disorder.

DESCRIPTION (provided by applicant): Sleep aids in the consolidation of memory and also plays a key role in regulating emotions. Extinction is a form of emotional memory that promotes emotion regulation by inhibiting the expression of conditioned fear. Using a validated 2-day human protocol, we propose to compare extinction memory between persons with Primary Insomnia and normally sleeping controls to test the main hypothesis that those with insomnia will display poorer extinction memory. Conditioned fear will be established to two colors as conditioned stimuli (CS+) using a mild electric shock as an unconditioned stimulus (US) and skin conductance response (SCR) as the conditioned response (CR). A third color never paired with a shock (CS-) will confirm differential conditioning. Fear of one color (CS+E) will then be extinguished by its repeated presentation without the US. After 24 h, memory for this extinction learning as well as the conditioned fear memory for the unextinguished CS+ (CS+U) will be tested. The accompanying neural responses will be measured at each stage of this protocol (fear conditioning, extinction learning, extinction recall) using functional magnetic resonance imaging (fMRI). Sleep physiology will be measured using ambulatory polysomnography (PSG) on an acclimation night, a baseline night and on the night that will intervene between the day when conditioning and extinction are learned and when they are recalled 24 h later. Correlations of extinction memory performance with sleep physiology will be examined among sleep quality measures (e.g., sleep efficiency) as well as sleep architectural measures (e.g., sleep stage percentages) and measures specific to rapid eye movement (REM) sleep (e.g. REM continuity). It is hypothesized that extinction memory (the degree to which SCR to the CS+E remains suppressed after 24 h) will be poorer in insomniacs. Additionally, it is hypothesized that the degree to which extinction recall generalizes (i.e., the degree to which SCR to the CS+U as well as to the CS+E is reduced) will be greater in controls versus insomniacs. It is also hypothesized that, in insomniacs versus controls during extinction recall, the neural structures whose activity has been associated with extinction learning and memory (ventromedial prefrontal cortex and hippocampus) will be hypoactivated and those associated with the expression of conditioned fear (amygdala and dorsal anterior cingulate cortex) will be hyperactivated. Lastly, it is hypothesized that, in insomniacs, extinction memory will correlate positively with sleep quality as well as with the integrity of sleep architecture, especially during REM sleep, across the night intervening between extinction learning and extinction recall.

? DESCRIPTION (provided by applicant): Proposed research will examine interrelationships between hyperarousal, sleep quality and extinction memory among persons who have experienced a severe traumatic event within the past 2 years (excluding the past month). In the aftermath of such an event, sleep disruption is very common and, particularly in the case of REM sleep, can predict the later development of post-traumatic psychological symptoms. Although sleep disruption is commonly classified with waking symptoms of hyperarousal, even mild elevation of baseline arousal can disturb sleep because the circuits and neuromodulators that promote arousal overlap with those that favor wakefulness during the normal sleep-wake cycle. Disturbed sleep may contribute to post-traumatic psychopathology by interfering with sleep-dependent emotional memory processes such as the consolidation and generalization of extinction memory (the ability to remember that what once predicted danger no longer does so). Among participants in the proposed study, a range of psychological health from highly functioning to highly symptomatic will first be obtained with stratified sampling using a composite index derived from 3 trans- diagnostic assessments of psychopathology. Across participants thus selected, the degree of hyperarousal will be assessed so as to generate a continuous range of values at 2 different levels of analysis. First, a canonical self-report hyperarousal measure will be computed from 3 validated questionnaires. Second, a validated acoustic-startle paradigm will measure habituation of physiological reactivity across continuum from rapid (minimal arousal) to absent or slow habituation (maximal arousal). Participants will then complete 2 weeks of sleep diaries with wrist actigraphy followed by a validated fear conditioning and extinction protocol consisting of 2 evening sessions 24 hr. apart. Fear conditioning and extinction learning (Session 1) and extinction recall (Session 2) will be measured physiologically using skin conductance response (SCR). Concurrent activity in 2 neural circuits for the expression of conditioned fear (fear-expression network) and extinction of such fear (extinction-memory network) will be examined using fMRI scanning at 3T. Ambulatory polysomnography will measure sleep quality and architecture on an adaptation night, a baseline night and an extinction-memory consolidation night between the 2 fMRI sessions. Hypotheses are, first, that poor quality of sleep, especially of REM, will be associated with poor extinction recall expressed physiologically by maintained elevation of SCR, and neurally by less activation in extinction-memory and more activation in fear-expression networks. Second, greater hyperarousal at self-report and/or physiological levels will be associated with poorer sleep quality. And third, greater hyperarousal will be associated with less extinction recall and less activation in extinction and more in fear networks. Support for hypotheses in combination will support the possibility that trauma elevates baseline activity in arousal systems resulting in sleep disruption that, in turn, impairs consolidation of extinction memory possibly lowering thresholds for intrusion of fear memories and other post-traumatic symptoms.

Epidemiological studies show that preexisting insomnia increases the odds of incident anxiety disorders and vice versa. We have suggested that preexisting sleep disturbance can disrupt sleep-dependent emotion regulatory circuits, such as those supporting consolidation of fear extinction memory, leading, in turn, to greater daytime anxiety symptoms. Exacerbation of both sleep disruption and anxiety symptoms can then occur via a positive feedback (?vicious cycle?) mechanism. The proposed research will test this 3- component mutual-reinforcement model in Generalized Anxiety Disorder (GAD), the anxiety disorder most commonly comorbid with Insomnia Disorder (ID). In a sample of individuals with GAD having varying degrees of sleep disturbance, we will measure associations between each pair of the above elements ? 1) sleep quality, 2) extinction memory and its associated neural substrates and 3) GAD symptomatology ? to determine whether such associations support this positive feedback model. In a recent fMRI study comparing individuals with Primary Insomnia (PI) to age and sex-matched good sleeping healthy controls (HC) and individuals with GAD, we have identified two factors that might predispose persons with ID to develop GAD. First, using resting state functional connectivity, we found that individuals with PI showed connectivity between the left amygdala and the rostral anterior cingulate cortex (a region believed to regulate amygdala activity) that was both significantly less than HC and intermediate between that of HC and individuals with GAD. Thus emotion regulation may be weakened in PI thereby increasing vulnerability to GAD. Second, using a validated 2-day fear conditioning and extinction protocol, we identified delayed engagement of extinction-related structures in PI relative to HC. In HC, activation of extinction-related structures (ventromedial prefrontal cortex and hippocampus) occurred during extinction learning, but did not do so in PI until extinction recall 24 h later. Research will examine 50 individuals with GAD half of whom can be diagnosed with comorbid ID and half of whom will not meet criteria for ID. Following psychiatric and sleep-disorders screening interviews, participants will complete 14 days of longitudinal actigraph and sleep-diary monitoring, an online battery of anxiety and sleep quality assessments, three nights of ambulatory polysomnography and two fMRI scanning sessions. fMRI will consist of both resting state scans and the above fear conditioning and extinction protocol with simultaneous measurement of skin conductance. The three components of the positive feedback model will be analyzed from both a categorical (GAD with vs. without co-morbid ID) and a dimensional perspective that accords with the NIMH Research Domain Criteria approach. General hypotheses tested will be 1) that co-morbid insomnia in GAD worsens behavioral and neural aspects of fear extinction as well as resting state-connectivity between fear expression and emotion regulatory structures, and 2) that poor sleep quality, poor extinction memory and greater GAD symptomatology show the predicted mutual exacerbation predicted by the positive feedback model.

Trauma-related nightmares (TRNs) are a hallmark re-experiencing symptom of Posttraumatic Stress Disorder (PTSD) and a severe, re-traumatizing source of distress to those with this condition. Proposed research will investigate whether such nightmares might serve as targets for imaginal exposure during prolonged exposure therapy (PE), a first-line treatment for PTSD. Learning and memory of fear extinction are the neurocognitive underpinnings of PE and increased physiological arousal during PE can aid in extinction learning. Thus, PE using TRNs might enhance treatment outcomes because of the greater immediacy and salience of nightmares relative to the more temporally distant traumatic event. Our research group has developed standardized script- driven imagery (SDI) procedures whereby the degree of psychophysiological arousal induced by recollection of the traumatic event that precipitated posttraumatic symptoms (index trauma) can be assessed. Our team has also developed a novel ambulatory device, the NINscan, that can record skin conductance (SC), facial electromyography (EMG) and electrocardiography (ECG) along with simultaneous functional near-infrared spectroscopic (fNIRS) imaging of the lateral prefrontal cortex (LPFC). The LPFC encompasses regions that activate and/or deactivate during SDI in persons with PTSD and includes the right inferior frontal gyrus (rIFG), an area that plays an important role in inhibiting unwanted behavior and cognition and shows reduced inhibitory performance in PTSD. Following screenings that include a structured clinical evaluation establishing PTSD diagnosis and self-report of a least 2 TRNs per week, qualified participants will record an account of their index trauma. They will then complete 2 weeks of sleep diaries and wrist actigraphy with 4 nights of ambulatory polysomnography (PSG), during which they will audio-record, on a time-stamped digital recorder, a detailed nightmare report following any nightmare-induced awakening. A TRN report having sufficient length, clarity and resemblance to the index trauma will be selected and, along with the index-trauma report, will be re- recorded as a script for SDI. Only participants who generated a TRN suitable for SDI (N=40) will then undergo 2 SDI sessions on a single day while wearing the NINscan. One script will be derived from their index trauma and the other from their TRN, the order of which will be counterbalanced across participants. Psychophysiological response and fNIRS LPFC activation will be compared between scripts using 2 primary outcome measures: (1) a validated canonical psychophysiological variable combining SC, EMG and ECG, and (2) script-related activation of the rIFC. We hypothesize that both activations will be greater for the TRN than for the index-trauma script. Exploratory aims will include (1) examination of the sleep stage and electroencephalographic characteristics of sleep preceding awakenings when a nightmare occurred on a night with PSG, and (2) combined PSG and fNIRS recording of trauma-related nightmares using the NINscan in the sleep laboratory in 5 individuals who recorded the most frequent nightmares during home monitoring.