Running head: BIOMARKERS OF SUCCESSFUL AGING

Memory impairments are among the most troubling aspects of cognitive aging. Can older adults change how they encode events to compensate for impairments? We reveal a novel compensatory effect by examining how neural activity changes across protracted study periods. We examined oscillatory power in electroencephalographic recordings obtained while younger (18-30 years) and older (60+ years) adults studied lists of words. Power changed in a highly consistent way from word-to-word across the study period. Above 14 Hz, there were virtually no age differences in these neural gradients. But gradients below 14 Hz reliably discriminated between age groups. Older adults with the best memory performance showed the largest departures from the younger adult pattern of neural activity. These results suggest that age differences in the dynamics of neural activity across an encoding period reflect changes in cognitive processing that compensate for age-related decline. BIOMARKERS OF SUCCESSFUL AGING 3 Age-Related Changes in the Dynamics of Memory Encoding Processes Provide a Biomarker of Successful Aging Our memories define us as individuals, they record our personal histories on an autobiographical timeline. Memory is also central to our intellectual lives, as almost every cognitive task requires finding key information in memory. Sadly, our memory tends to get worse as we grow older, as revealed by both self-reports (Newson & Kemps, 2006; Zacks, Hasher, & Li, 2000) and laboratory studies (Craik & Jennings, 1992; Light, 1991; Salthouse, 1991; Stark, Yassa, & Stark, 2010). Identifying age differences in cognitive processing that contribute to, or compensate for, age-related memory impairments is a critical step in developing effective treatments. One potential source of age differences that has received little attention in the literature arises when we must encode a series of events or items that unfold over time. For example, the people you meet during a job interview, the grocery list your spouse dictates over the phone, or which of your medications you have already taken today. Researchers have studied this aspect of memory using the free recall task, in which participants study a list of sequentially presented items (e.g., words) and then recall the items in any order. The nature of the encoding processes participants engage changes from item-to-item as the list is studied. Most cognitive aging theories are silent about the contribution of these dynamics to memory impairments (Benjamin, 2010; Hasher & Zacks, 1988; Naveh-Benjamin, 2000; Salthouse, 1996; Stark et al., 2010). Perhaps because age differences in such encoding dynamics are difficult to detect by examining recall behavior, as such behavior is determined not only by encoding but also by retrieval processes. We argue, however, that there are two general categories of item-to-item changes in cognitive processing that are likely to show age differences—age differences that may be detectable by examining neural activity. First, some processes likely become less efficient as the list progresses. As one example, early list items may be encoded more strongly than later items due to fatigue of BIOMARKERS OF SUCCESSFUL AGING 4 the neural circuitry underlying encoding (Tulving & Rosenbaum, 2006) and these dynamics may show age differences. Indeed, using a model of memory encoding and search that simulates this “primacy gradient” (Lohnas, Polyn, & Kahana, 2015), we found that older adults’ encoding efficiency starts out higher than younger adults’ but decays rapidly, dropping below the young adult level within a few items (Healey & Kahana, 2016). The second class of processes likely to show age differences are those that ramp up as the list goes on. As an example, younger adults engage in rehearsal (thinking back to early list items) and elaborative encoding (forming connections between current and earlier items). Engagement of these processes should increase across a list, simply because the number of earlier items to think back to increases across the list. There is reason to suspect age differences in the rate at which these processes increase across the list. For example, asking participants to rehearse aloud reveals that whereas rehearsals increase across early serial positions for younger adults (Ward & Maylor, 2005) older adults rehearse fewer items across a list (Ward & Maylor, 2005). A natural prediction is that any such age-related changes in the dynamics of encoding processes would exacerbate memory impairment. Some cognitive changes, however, are compensatory (Buckner, 2004; Cabeza, Anderson, Locantore, & McIntosh, 2002; Daselaar et al., 2015; Gutchess et al., 2005; Lighthall, Huettel, & Cabeza, 2014; Zimmerman, Hasher, & Goldstein, 2011). Therefore, we raise the possibility that age differences in encoding dynamics are associated with reduced memory impairments, contributing to successful aging. For example, rehearsal and elaborative encoding require retrieving earlier list items (Laming, 2008). Given that older adults have impaired retrieval processes (Healey & Kahana, 2016), attempts to think back to earlier items will often fail and older adults may be well-served by forgoing rehearsal and elaborative encoding in favor of focusing on encoding the current item. We will test the hypotheses that there are age differences in the dynamics of cognitive processing across a list and that these processing differences may either contribute to, or BIOMARKERS OF SUCCESSFUL AGING 5 compensate for, age-related memory impairment. We have offered two examples of processes that might show age differences in across-item dynamics (encoding efficiency and rehearsal/elaborative encoding). Many other processes might show age differences as well. Any process that changes across encoding should leave its signature in neural activity. Therefore, our approach is to examine neural recordings taken while participants study lists and test for biomarkers of age that take the form of different gradients of neural activity across an encoding period.