NIH toolbox cognition battery (CB): Measuring episodic memory
P. Bauer,S. Dikmen,R. Heaton,D. Mungas,J. Slotkin,J. Beaumont
DOI: https://doi.org/10.1111/mono.12033
2013-08-01
Abstract:One of the most significant domains of cognition is episodic memory, which allows for rapid acquisition and long-term storage of new information. For purposes of the NIH Toolbox, we devised a new test of episodic memory. The nonverbal NIH Toolbox Picture Sequence Memory Test (TPSMT) requires participants to reproduce the order of an arbitrarily ordered sequence of pictures presented on a computer. To adjust for ability, sequence length varies from 6 to 15 pictures. Multiple trials are administered to increase reliability. Pediatric data from the validation study revealed the TPSMT to be sensitive to age-related changes. The task also has high test– retest reliability and promising construct validity. Steps to further increase the sensitivity of the instrument to individual and age-related variability are described. In this chapter, we introduce the NIH Toolbox Picture Sequence Memory Test (TPSMT), a measure developed as a test of episodic memory for ages 3– 85 years. Episodic memory permits rapid learning and retention of new information. It is the basis for formation of memories of the mundane—such as where the car was parked last—to the special events that constitute one’s life story or personal past. Subdomain Definition Evidence from nonhuman animals, patient populations, and typically developing children and adults makes clear that memory is not a unitary construct. Rather, it is comprised of different systems of information encoding, storage, and retrieval. One major distinction is the dichotomy between maintenance of information over the short term, and over the long term (long-term memory). Short-term or working memory is discussed in Tulsky and colleagues (Chapter 5, this volume); long-term memory is the subject of the current chapter. Within long-term memory there is a distinction between procedural (or implicit, or nondeclarative) memory and declarative (or explicit) memory (e.g., Squire, Knowlton, & Musen, 1993; Tulving, 2000). Procedural, implicit, or nondeclarative memories guide behavior but seemingly require no effort to retrieve and are not accessible to consciousness. Work with animal models and human patients makes clear that the procedural memory system remains relatively intact in aging and is virtually unaffected by neurological conditions and diseases that target the medial temporal lobe structures involved in the declarative memory system (Churchill, Stanis, Press, Kushelev, & Greenough, 2003; Reber, Martinez, & Weintraub, 2003). In contrast to procedural memory, declarative or explicit memory is effortful and involves conscious recollection of information that is potentially verbally accessible. The declarative memory system is further divided into semantic and episodic memory (e.g., Squire, 2004). Corresponding author: Patricia J. Bauer, Department of Psychology, Emory University, 36 Eagle Row, Atlanta, GA 30322, patricia.bauer@emory.edu. NIH Public Access Author Manuscript Monogr Soc Res Child Dev. Author manuscript; available in PMC 2014 August 01. Published in final edited form as: Monogr Soc Res Child Dev. 2013 August ; 78(4): 34–48. doi:10.1111/mono.12033. N IH PA Athor M anscript N IH PA Athor M anscript N IH PA Athor M anscript Semantic memory is specialized for storage of timeless, placeless facts, concepts, and the vocabulary to describe them. Knowledge stored in semantic memory typically is long lasting, which is the basis for reference to semantic memory as “crystallized intelligence.” Semantic memory is resistant to decline with age and to neurological insult. The construct of semantic memory is measured by the Toolbox Picture Vocabulary Test and the Toolbox Oral Reading Recognition Test (see Gershon et al., Chapter 4, this volume). In contrast, episodic memory is specialized for storage of unique events or experiences encoded in a time-specific manner. Episodic memory is fragile and time-limited (though some memories, such as those of special, personally relevant events, may endure over long periods of time), and it is sensitive to decay and interference, as well as to both normal aging and many brain diseases. Importance During Childhood Episodic memory allows for rapid, even one-trial, learning of new information and for retention of information for later retrieval. As such, it provides the building blocks for cognitive growth during development and throughout the lifespan. Its importance to mental life is nowhere more in evidence than in the historic case of the patient HM, who as a result of surgery to the medial–temporal lobe, lost the capacity for forming new episodic memories (e.g., Corkin, 2002). As well, the relatively protracted course of development of episodic memory is a major source of one of the most robust phenomena in the memory literature, namely, infantile or childhood amnesia: the relative paucity among adults of memories of unique events for the first 3–4 years of life (see Bauer, 2007, for a review). Thus, this cognitive ability is critical for achieving a concept of self that is continuous over time, for independence, education, and success in personal and professional activities of daily life throughout the lifespan. The centrality of episodic memory explains why it is the most frequently measured form of memory and why it is included in the CB. The course of development of episodic memory is protracted, with pronounced changes throughout the first two decades of life (see Bauer, 2007; Bauer, Larkina, & Deocampo, 2011, for reviews). In infancy, the ability is measured using nonverbal tasks, such as elicited and deferred imitation (props are used to produce a specific action or sequence of actions that the infant is permitted to imitate either immediately, after a delay, or both; e.g., Bauer & Shore, 1987; Bauer & Mandler, 1989). Use of imitation-based tasks in the first 3 years of life has revealed age-related increases in the length of time over which memory is preserved, in the robustness of memory, and in the reliability with which it is observed (e.g., Bauer, Wenner, Dropik, & Wewerka, 2000). Developmental changes are especially apparent in the ability to remember the temporal order of events (e.g., Bauer et al., 2000), making this aspect of episodic memory a target for the CB. Episodic memory continues to develop throughout childhood and into adolescence. There are age-related increases in the amount of information that children remember. For example, relative to younger children, older children remember longer lists of items (see, e.g., Bjorklund, Dukes, & Brown, 2009, for a review), thus making list length a prime target for exploitation in tests of episodic memory designed for wide age ranges (see below). With development, children’s memory becomes more deliberate and strategic, with resulting increases in the organization that children impose on to-be-remembered material (e.g., Bjorklund et al., 2009). In addition, children become more aware of their own and others’ memory processes (i.e., increases in metamemory), enabling them to recruit informationprocessing resources in the service of increased memory demands. As a result, we may expect to see age-related increases in the amount of information children are able to remember, and in their ability to bring organization to it. Bauer et al. Page 2 Monogr Soc Res Child Dev. Author manuscript; available in PMC 2014 August 01. N IH PA Athor M anscript N IH PA Athor M anscript N IH PA Athor M anscript The overall importance of episodic memory to typical development is brought into stark relief in pediatric populations in whom the function is impaired. The most striking example is a population who as infants and very young children, sustained damage to the medial– temporal lobe structures that support declarative memory (see below). Without exception, individuals in this population of so-called developmental amnesics experienced difficulty learning in school and deficits on episodic memory tasks (Gadian et al., 2000). Importantly, the damage sustained by this population is confined to the neural structures involved in declarative memory in general and episodic memory in particular, described next. Relations of Domain With Brain Function Encoding, storage, and retrieval of episodic memories depend on a multi-component network involving the temporal lobe (including hippocampus and surrounding cortices) and other cortical area (including prefrontal cortex and limbic/temporal association areas) (e.g., Eichenbaum & Cohen, 2001; Zola & Squire, 2000). Specifically, the process of encoding of new memories begins as the elements that constitute an event register across primary sensory areas (auditory, somatosensory, visual). Inputs from the primary cortices are projected to unimodal association areas, where they are integrated into whole percepts of what objects sound, feel, and look like. Unimodal association areas in turn project to polymodal prefrontal, posterior, and limbic association cortices where inputs from the different sense modalities are integrated and maintained over brief delays (seconds; e.g., Petrides, 1995). For maintenance beyond the short term, the inputs must be stabilized or consolidated, a task attributed to medial temporal structures, in concert with cortical areas (McGaugh, 2000). Eventually, new traces become stabilized, permitting long-term storage in the neocortex. The prefrontal cortex is implicated in memory retrieval (e.g., Cabeza et al., 2004; Maguire, 2001). Demands on the temporal–cortical network are especially high when tasks require free recall versus recognition, and memory for temporal order information versus for items alone (e.g., Shimamura, Janowsky, & Squire, 1990). As discussed earlier, these conditions are those under which the most pronounced age-related differences are observed, thus informing design of the Toolbox measure for episodic memory. Each of the brain regions involved in episodic memory, as well as the connections between them, undergoes substanti