Learning and memory in the aging brain: The function of declarative and nondeclarative memory over the lifespan.

Abstract

Learning and memory systems Human memory is not a unitary faculty. Rather, it consists of distinct learning and memory systems, each contributing in unique ways to the acquisition, retention, and subsequent retrieval of information. The current chapter focuses on age-related changes to long-term learning and memory systems defined as the acquisition and retention of information over long intervals of time (i.e., beyond working memory capacity). Long-term memory is often subdivided into declarative memory (the acquisition and retention of knowledge) and nondeclarative memory (experience-induced changes in performance). In this chapter, we draw on both behavioral and neuroimaging studies of young adults and older adults to describe the functional-anatomic architecture of multiple memory systems including declarative and nondeclarative memory. First, we present the neural correlates of the declarative memory system, followed by a discussion of how age-related changes to declarative memory mechanisms affect episodic memory while leaving semantic memory relatively intact. Then, we present the neural correlates of the nondeclarative memory system, followed by a discussion of how age-related changes to mechanisms of priming, classical conditioning, procedural, and reinforcement learning affect associated learning and memory performance with age. Finally, we discuss interactions across and between memory systems with age, and potential future directions in research on learning and memory function in the aging brain. Neural correlates of the declarative system Declarative memory encompasses the acquisition, long-term retention, and retrieval of events, facts, and concepts (Squire, 2009). Such knowledge can be retrieved at will and used in a variety of contexts. Declarative memory is typically subdivided based on whether memories are concerned with personally relevant events (i.e., episodic memory) or impersonal information (i.e., semantic memory). Broadly speaking, declarative learning and memory processes depend predominantly on the prefrontal cortex (PFC), medial temporal lobes (MTL), and lateral temporal cortex (see Figure 1). Research suggests that the PFC plays a supervisory role over other brain regions, including the MTL (Norman & Shallice, 1986; Miller & Cohen, 2001), via consciously controlled bias mechanisms (Buckner, 2003). During encoding, the PFC implements the processes that organize input to the MTL and during retrieval, the PFC mediates search and post-retrieval monitoring processes (Cabeza, 2006; Moscovitch, 1992). In contrast, the MTL, particularly the hippocampus, serves to bind the conceptual, perceptual, and affective components of an episodic event into an integrated memory trace (Eichenbaum, Yonelinas, & Ranganath, 2007; Moscovitch, 1992). These binding processes operate both during the initial registration of novel events and when subjects retrieve recently acquired information (Gabrieli,

Normative data for the Spanish versions of the CVLT, WMS-Logical Memory, and RBMT from a sample of middle-aged and old participants.

“…As shown in the tables of adjustments (), performance scores tend to decrease as age increases, with a gradual decline in EM performance in a normal aging process. This decline has been also found in laboratory conditions where older and old–old adults do not perform as well as middle-age adults, as pointed out by Clark et al (2018), Lighthall et al (2019), Spaan (2015), and Williams et al (2020), among others. At the same time, the tables of adjustments show an improvement in EM scores as the level of education increases, in accordance with the cognitive reserve hypothesis for which educational level is one of the most important proxies for reserve (Constantinidou et al, 2014; Lojo-Seoane et al, 2014).…”

“…Cohort studies are recommended for obtaining normative data to exclude those participants who subsequently develop cognitive impairment (Fine et al, 2012; Grober et al, 2015). The influence of sociodemographic variables on EM scores should be analyzed, as lower age, female sex, and higher education have been found to be associated with better performance of EM tasks (Asperholm et al, 2019; Clark et al, 2018; Constantinidou et al, 2014; Golchert et al, 2019; Juncos-Rabadán et al, 2014; Lighthall et al, 2019; Lojo-Seoane et al, 2014; Lundervold et al, 2014).…”