David C. Rubin - US grants

The distribution of autobiographical memory is described by a three component model. The first component is a retention function derived from laboratory studies of retention. The second component is a childhood memory function which is used to account for the lack of memories from the earliest years of life. The third component is a reminiscence function which plays a noticeable role only in older adutls. Existing evidence for these three components is presented. In order to further refine and test this three component model, a series of experiments is presented in which verbal and olfactory stimuli are used to prompt autobiographical memories. Data from groups of healthy 18 and 70 year olds, individual 18 and 70 year olds, and individual nursing home residents will be analyzed. Besides increasing our understanding of human memory and how it changes with age, the research will provide normative data against which changes due to general health and specific memory related disorders can be evaluated.

Description (adapted from investigator's abstract): A common anecdotal report of older adults is that they remember events that occurred when they were younger better than more recent events. When this reversal of a normal monotonic retention function is examined carefully, either by asking for life stories or by curing individual autobiographical memories with words, older adults do indeed remember more events from when they were 10 to 30 years old. This "bump" phenomenon is one of the few cognitive effects of aging that is not a decrement in performance. It has many explanations that have resisted being teased apart using standard techniques of cognitive psychology. However, by examining people who migrated at various points in their lives, changing language, culture, and environment, we propose to separate classes of explanations. A central aspect of autobiographical memory is language, yet little work has examined whether memory, or discourse of any kind, is easier to retrieve in the language in which it was encoded. Yet a few studies with young adults, clinical data from psychotherapeutic treatment, and the introspections of older bilinguals indicate that it is. By examining the memories of people who migrated and learned a second language at different times (and those who did not migrate or learned both languages simultaneously, or know only one language), we can examine such questions. Since many older adults are bilinguals who have changed their relative competence in their languages over their lifespans, and even more have made major migrations, this work has practical as well as theoretical interest for the nature of language and memory in adult development. Tasks include a narrative and a word-cued autobiographical memory procedure, and a bilingual language assessment. Participant populations recruited to help separate effects include adult Hispanics who migrated to either Anglo or Hispanic communities in the US, Poles who were granted asylum in Denmark, monolingual non-migrating matched controls, and older monolingual adults who migrated within the US to non-retirement communities, age segregated retirement communities, and long-term care facilities.

DESCRIPTION (provided by applicant): We formulate and test a theory of posttraumatic stress disorder (PTSD) based only on mechanisms that have solid support in scientific studies of cognition, personality, and emotion. The theory combines findings from experimental and individual differences research to make predictions that often contradict existing theories of PTSD, but that agree with the existing data. It is the first theory to combine measures of the severity of traumatic events, personality, and memory processes. Moreover, it predicts quantitatively clinical outcomes by assigning numerical weights to the various measures. Two main studies are proposed, which use structural equation models. In the first, 200 adults will be tested, who vary in symptom severity. Of these, at least 50 will meet standard clinical diagnostic inclusion criteria for current PTSD and 50 will meet standard exclusion criteria for lifetime PTSD so that, in addition to the correlational analyses, direct comparisons can be made between the prototypical groups used in the PTSD literature. To provide predictor variables, the participants will complete standardized tests of mood and personality. They will recall and record a narrative version of their most traumatic or stressful events, which will be scored for coherence and event severity. They will also recall their three most positive and three most important events and imagine that three traumatic scenarios occur to them in the future. The imagined future events will probe how the participants view and react to traumatic events in general. For each of their memories and the imagined events, participants will provide measures of emotional intensity and valence, vividness of sensory information, narrative coherence, and degree of reliving of the event. They will also provide ratings of involuntary, often intrusive, memories made as they come to mind for two weeks, recording their responses on personal data assistants. This online technique was developed by our group and its use is still unique to it in PTSD populations. It provides the most accurate measure of intrusive memories, a key symptom of PTSD. The rich data set obtained allows us to disentangle competing existing hypotheses and better understand the mechanisms that lead to and maintain PTSD. The second study uses the same theory and quantitative techniques, but takes advantage of an ongoing longitudinal study in its 20th year. It includes fewer, but similar, tests to the first study and a sample of several thousand adults approximately 60 years old. It includes tests of current PTSD symptom severity, mood and personality tests, as well as health data, to do a truly prospective study of current PTSD symptoms as predicted by measures obtained longitudinally. Additional tests and experiments will be conducted to test specific supporting hypotheses. In particular, we will examine the coherence of narrative of memories of traumatic, very positive, and important events in participants who have or do not have a clinical diagnosis of PTSD. PUBLIC HEALTH RELEVANCE: Based on measures of the severity of a traumatic event, the personality of the person experiencing the event, and the properties of the memories that the person has of the event and of other non-traumatic events, we attempt to predict which people will develop posttraumatic stress disorder (PTSD) and how severe their symptoms will be. We do this by trying to understand autobiographical memory for emotional, stressful events. Our quantitative theory, if successful, would help identify risk factors for PTSD and should suggest which properties of traumatic memories therapy should attempt to change.

[unreadable] DESCRIPTION (provided by applicant): Autobiographical memory - people's memory for events from their lives - is a basis of judgments they make about their future behavior, what they report in their medical histories, and what they use in many forms of psychotherapy and everyday life. Its malfunction is a presenting symptom in Alzheimer's disease, posttraumatic stress disorder, closed head injury, and nearly all forms of general memory loss. To date, studies of the neural basis of autobiographical, or episodic, memory have generally relied on materials learned in the laboratory with minimal emotional involvement, a highly controlled and simple sensory environment, over one very short time interval, and in one location. Thus, many of the factors known behaviorally to have profound effects on memory (i.e., emotional intensity, sensory input, and spatial and temporal context) are not varied enough at learning to have effects that can be observed at recall, providing a limited view of the brain systems involved in memory for real-world events. In the proposed studies, these factors are varied in realistic ways at encoding and their effects on the neural basis of recall are probed using functional magnetic resonance imaging. Three different kinds of tasks will be used to investigate the neural correlates of autobiographical memory: (1) recall of personal life events given various word cues, (2) recognition of digital photographs which participants take in the field or study in the laboratory, and (3) cued recognition of staged events. Variations in visual imagery processes, emotional content, and spatiotemporal contextual information will be measured at encoding and mapped onto their associated processing regions in the brain. New insight will be gained into the network components activated at different stages of memory retrieval, including the search and production of the memory and the monitoring of its accuracy and appropriateness. The proposed research will advance knowledge about the brain systems involved in retrieving information about autobiographical events by bridging between memory for real-world events and memory for laboratory stimuli. [unreadable] [unreadable]

DESCRIPTION (provided by applicant): We formulate and test a theory of posttraumatic stress disorder (PTSD) based only on mechanisms that have solid support in scientific studies of cognition, personality, and emotion. The theory combines findings from experimental and individual differences research to make predictions that often contradict existing theories of PTSD, but that agree with the existing data. It is the first theory to combine measures of the severity of traumatic events, personality, and memory processes. Moreover, it predicts quantitatively clinical outcomes by assigning numerical weights to the various measures. Two main studies are proposed, which use structural equation models. In the first, 200 adults will be tested, who vary in symptom severity. Of these, at least 50 will meet standard clinical diagnostic inclusion criteria for current PTSD and 50 will meet standard exclusion criteria for lifetime PTSD so that, in addition to the correlational analyses, direct comparisons can be made between the prototypical groups used in the PTSD literature. To provide predictor variables, the participants will complete standardized tests of mood and personality. They will recall and record a narrative version of their most traumatic or stressful events, which will be scored for coherence and event severity. They will also recall their three most positive and three most important events and imagine that three traumatic scenarios occur to them in the future. The imagined future events will probe how the participants view and react to traumatic events in general. For each of their memories and the imagined events, participants will provide measures of emotional intensity and valence, vividness of sensory information, narrative coherence, and degree of reliving of the event. They will also provide ratings of involuntary, often intrusive, memories made as they come to mind for two weeks, recording their responses on personal data assistants. This online technique was developed by our group and its use is still unique to it in PTSD populations. It provides the most accurate measure of intrusive memories, a key symptom of PTSD. The rich data set obtained allows us to disentangle competing existing hypotheses and better understand the mechanisms that lead to and maintain PTSD. The second study uses the same theory and quantitative techniques, but takes advantage of an ongoing longitudinal study in its 20th year. It includes fewer, but similar, tests to the first study and a sample of several thousand adults approximately 60 years old. It includes tests of current PTSD symptom severity, mood and personality tests, as well as health data, to do a truly prospective study of current PTSD symptoms as predicted by measures obtained longitudinally. Additional tests and experiments will be conducted to test specific supporting hypotheses. In particular, we will examine the coherence of narrative of memories of traumatic, very positive, and important events in participants who have or do not have a clinical diagnosis of PTSD.