Does repetition engender the same retrieval

NEUROREPORT

AGING

Does repetition engender the same retrieval processes in young and older adults? Doreen Nesslera, David Friedmana, Ray Johnson Jrb and Michael Bersicka a

Cognitive Electrophysiology Laboratory, NYS Psychiatric Institute and bDepartment of Psychology, Queens College/CUNY, Flushing, New York

Correspondence to Doreen Nessler, PhD, Cognitive Electrophysiology Laboratory, NYSPI,1051 Riverside Drive, Unit 6, New York, NY10032, USA Tel: + 1212 543 5729; fax: + 1212 543 6002; e-mail: [email protected] Received 23 July 2007; accepted17 August 2007

Aging di¡erentially a¡ects retrieval processes underlying recognition memory: familiarity is maintained, whereas recollection declines. We determined whether word repetition across two study-test phases enhanced older adults’ use of recollection. DuringTest 1, frontal episodic memory e¡ects, suggestive of familiarity-based processes, were age invariant, whereas only the young showed a parietal episodic memory e¡ect, suggestive of recollection. Repetition did not modulate the frontal episodic

memory e¡ect in either group, but increased the parietal episodic memory e¡ect in the young. Importantly, older adults showed a parietal episodic memory e¡ect at Test 2, suggesting that repetition did enable recollection. Only older adults, however, showed a left frontal negativity, implying that they may have used additional processes to recover episodic memories. NeuroReport c 2007 Wolters Kluwer Health | Lippincott 18:1837^1840 Williams & Wilkins.

Keywords: aging, compensation, episodic memory, event-related potentials, familiarity, frontal episodic memory e¡ect, parietal episodic memory e¡ect, recollection, repetition

Introduction Aging is characterized by a decline in episodic memory performance [1], but does not seem to impact equally the two processes thought to support recognition memory, familiarity and recollection [2,3]. Familiarity-based recognition, the acontextual knowing that an item has been encountered previously, appears to remain intact as people age. In contrast, recollection-based recognition, characterized by the remembering of detailed, item-specific, contextual features, is thought to deteriorate with increasing age [4,5]. Recent data from our laboratory (Nessler, Johnson, Bersick and Friedman, unpublished) suggest that agerelated changes in memory retrieval occur independently from age-related encoding deficits [6]. The independent assessment of retrieval activity was possible due to a fortuitous finding in an earlier study [7]; the older adults’ engagement in deeper semantic encoding equated their encoding-related event-related potential (ERP) activity and subsequent recognition performance with those of the young. Despite these equivalencies, we found qualitative differences between the two age groups in the corresponding ERPs at retrieval. Items that were correctly recognized (hits) elicited greater positivity at frontal locations (300–500 ms) than correctly rejected new items for both groups (frontal episodic memory effect [8]). However, for young adults only, hits elicited a larger late positive component than correct rejections at left parietal sites (500–800 ms) a difference labeled the parietal episodic memory effect [8]. The specific pattern of retrieval-related ERPs found by Nessler et al. (unpublished) was interpreted as further

c Wolters Kluwer Health | Lippincott Williams & Wilkins 0959- 4965

support for theoretical accounts suggesting age-invariant familiarity but age-related deficits in recollection [2,3]. The frontal episodic memory effect does not vary with depth of processing [9] and is elicited by lures semantically related to studied items erroneously judged to be old [10], providing evidence for its association with familiarity-related processes [8,11,12]. In contrast, the magnitude of the parietal episodic memory effect increases with depth of processing [9] and is larger for words remembered with contextual information compared with words recognized only with the feeling that they have been studied before (knowing) [13]. These and other findings denote the parietal episodic memory effect as an indicator of recollection-related processing [8,11,12]. Thus, it can be inferred that the ageinvariant frontal episodic memory effect and the presence of the parietal episodic memory effect for young but not older adults (Nessler et al., unpublished) reflect, respectively, intact familiarity-based processes and impaired recollection for older adults. The question remains whether or not these results reflect a general age-related qualitative difference in the processes used to recover information from memory or whether older adults are able, under certain conditions, to use recollective as well as familiarity-based processes. One manipulation known to increase recognition performance is repetition [14]. It is conceivable that the beneficial effects of repetition for older adults might enable them to engage processes similar to those of the young. Alternatively, older adults might achieve increased recognition performance as a result of using processes that compensate for deficiencies in recollection [15,16]. In line with this compensation hypothesis [17], age-related,

Vol 18 No 17 19 November 2007

18 3 7

NEUROREPORT prefrontal, hemodynamic overactivations for older adults have been documented for a wide range of processes, including episodic memory [16]. Recent ERP research has identified a sustained left-frontal negativity (600–1200 ms) in older but not young adults that has been observed for retrieval conditions that require the remembering of contextual (i.e. source) information from the study phase, suggesting that it may reflect compensatory activity [18,19]. The present report is the third in a series ([7]; Nessler et al., unpublished) analyzing age-related effects on encoding and retrieval assessed in one experimental session. Here we examined whether the qualitative difference in retrievalrelated ERP activity observed earlier (Nessler et al., unpublished) persisted in older adults even after enhanced recognition performance engendered by repetition. If the repetition-induced increase in older adults’ performance is due to the engagement of recollection, then, in addition to a frontal episodic memory effect, a parietal episodic memory effect should be present in the older adults’ ERPs. Alternatively, older adults might recruit compensatory processes at left frontal locations to increase their performance.

Methods Participants Sixteen healthy right-handed young (M¼22.8 years, range: 18–29) and older (M¼71.6 years, range: 62–86) adults participated [7] after signing informed consent. The study was approved by the NYS Psychiatric Institute’s Institutional Review Board. Experimental design Participants faced a 1700 computer monitor about 100 cm from the screen and held a response box on their laps. All stimuli (48 pt Arial) were presented within a centrally located grey presentation box on a black background. Participants were informed that there were two initial study-test blocks (Study 1, Test 1) and two repeated study-test blocks (Study 2, Test 2). In each of the two blocks in Study 1, participants made, in random order, 30 low-selection and 30 high-selection decisions on nouns (total of 120 studied nouns). In the low-selection task participants decided whether a S2 noun did or did not match a preceding S1 picture of an object (50% match, 50% no-match). In the high-selection task participants indicated whether a preceding S1 adjective described a characteristic of a S2 noun. S1 was shown for 500 ms and replaced for 500 ms by a blank box. Presentation of the to-be-remembered S2 noun (300 ms) ensued and was replaced by a blank box for a randomly jittered 2400–2900 ms interval before the next trial. Recognition tests were given 5 min after the end of each study block. For each of the two blocks in Test 1, the 60 nouns from each study block were combined with 60 new words. Studied and new words appeared randomly for 300 ms followed by a jittered 1900–2400 ms intertrial interval. Speeded and accurate old/new responses were required. All items were repeated in Study 2 and tested again in Test 2, with a different set of new words. Event-related potential recordings Electroencephalogram (EEG) activity was recorded with 62 sintered Ag/AgCl electrodes mounted in an elastic cap (Neuromedical Supply, Compumedics, El Paso, Texas, USA) in accord with the extended 10–20 system (ground: right

18 3 8


NESSLER ETAL.

forehead; electrode impedance o5 kO). All activity was initially nosetip referenced and rereferenced offline to averaged mastoids. Vertical electrooculogram (EOG) was recorded from electrodes placed above and below the left eye and horizontal EOG from electrodes placed at the outer canthus of each eye. EEG and EOG (DC; 100 Hz highfrequency cutoff; 500 Hz digitization rate) were recorded continuously with Synamp amplifiers (Compumedics). Data analysis Test trials with reaction times (RTs) between 200 and 2000 ms were analyzed. Eye movements were corrected [20]. Trials with visible artifact were rejected. EEG epochs extended from 200 ms prestimulus to 1600 ms poststimulus, with the prestimulus interval serving as baseline. Averages were computed across all items for hits in Test 1 (young trial numbers, range: 55–102; older: 53–111) and Test 2 (young: 79–113; older: 69–111) and for correct rejections in Test 1 (young: 60–117; older: 58–109) and Test 2 (young: 61–116; older: 50–114). Based on previous research and the grand mean ERPs, the frontal episodic memory effect for young adults was measured by comparing the averaged voltages for hits and correct rejections between 300 and 500 ms at Fz. A later interval (400–600 ms) and adjacent electrode site (F4) was used to capture frontal episodic memory effects in the older adults [21]. In line with its left-parietal maximum, the parietal episodic memory effect was quantified as the difference in late positive component amplitude between hits and correct rejections at P3. Peak latencies of this activity typically differ between young and older adults [13], hits and correct rejections, and decrease as RT decreases across test repetitions [14]. Therefore, we determined mean late positive component peak latencies between 500 and 800 ms at P3 for all conditions in both age groups. Rounded to the nearest 10 ms, these latencies defined the center of the time interval used for the quantification of the averaged voltages (young hits Test 1: 490–790 ms, Test 2: 450–750 ms; correct rejections Tests 1 and 2: 500–800 ms; older hits Tests 1 and 2: 520–820 ms; correct rejections Test 1: 560–860 ms, Test 2: 550–850 ms). To determine the presence of the left-frontal negativity, averaged voltages at AF7 for two different time intervals (500–900, 900–1600 ms) were used. Repetition effects on memory sensitivity, Pr (hits – false alarms), and response bias, Br {(False Alarms/100)/[1(Pr/ 100)]} [22] were assessed in between-group ANOVAs (analyses of variance) with the repeated-measures factor of Repetition (Test 1, Test 2). Analyses for RT and ERP data also used the repeated-measures factor of Stimulus (hits, correct rejections).

Results Behavioral data Pr was larger for young (Test 1: 73.3; Test 2: 90.3) than older adults (Test 1: 58.7; Test 2: 82.1) [F(1,30)¼11.0, Po0.01] and repetition increased Pr in both age groups [F(1,30)¼128.5, Po0.001]. The Group difference in Pr of 14.6 at Test 1 decreased at Test 2 to 8.1, indicating that older adults benefited marginally more from repetition than the young [Group by Repetition F(1,30)¼3.2; P¼0.08]. No reliable age differences were found for Br (young Test 1: 0.29, Test 2: 0.58; older Test 1: 0.31, Test 2: 0.52) or RT (young hits Test 1: 884 ms, Test 2: 768 ms; correct rejections Test 1:

NEUROREPORT

MEMORY RETRIEVAL IN YOUNG AND OLDER ADULTS

TEST 1

(a)

TEST 2

Frontal episodic memory effect

300 −500 ms Fz

F4

Fz

F4

+ (b) − 0

Parietal episodic memory effect

2 µV

500

1000

1500 ms 500 − 800 ms

500 − 800 ms

500 −800 ms P3

400 − 600 ms

300 − 500 ms

400 − 600 ms

P3

P3

P3

∆ =0.2 µV Left frontal negativity

Correct rejections Hits (c)

900− 1600 ms

900− 1600 ms

Af7

Af7

Young

Af7

Older

Af7

Young

500 − 900 ms

Older

Fig. 1 Grand-averaged correct rejection and hit event-related potentials (ERPs) inTests1 (left) and 2 (right) at (a) Fz (young), F4 (older adults), (b) P3 and (c) Af7. Shading and hatching in the waveforms indicate signi¢cant di¡erences between hits and correct rejections. The maps show reliable topographic di¡erences between hits and correct rejections. Shading¼negativity; unshaded areas¼positivity.

931 ms, Test 2: 864 ms; older hits Test 1: 933 ms, Test 2: 831 ms; correct rejections Test 1: 950 ms, Test 2: 943 ms) [F’so2.0, P’s40.05]. Frontal episodic memory effect The ANOVA on the averaged voltages at Fz (young) and F4 (older) indicated that hits elicited more positivity than correct rejections [F(1,30)¼51.9, Po0.001] in both groups [Group by Stimulus: F(1,30)¼3.3, P40.05; Fig. 1a]. Repetition did not influence the magnitude of the frontal effect (all interactions with Repetition Fso1). Parietal episodic memory effect The between-age group ANOVA on late positive component amplitudes at P3 confirmed that hits were larger than correct rejections [F(1,30)¼47.9, Po0.001]. The Group by Stimulus interaction [F(1,30)¼15.2, P¼0.001] indicated that the parietal effect was larger for young than older adults (Fig. 1b). Follow-up analyses carried out separately for each group showed main effects of Stimulus [young F(1,15)¼54.1, Po0.001; older F(1,15)¼5.0, Po0.05] and Repetition by Stimulus interactions [young F(1,15)¼5.1, Po0.05; older F(1,15)¼5.9, Po0.05]. However, while these effects indic-

ated larger parietal episodic memory effects in Test 2 [F(1,15)¼86.5, Po0.001] than Test 1 [F(1,15)¼17.7, Po0.001] for the young, for older adults they revealed that the parietal effect only obtained at Test 2 [F(1,15)¼10.5, Po0.01] but not Test 1 [Fo1] (Fig. 1b). Group differences in the magnitude of the parietal episodic memory effect at Test 2 were assessed on the hit minus correct rejection difference ERPs at P3. The young produced larger effects than older adults [F(1,30)¼16.4; Po0.001], a result consistent with their better recognition performance.

Left-frontal negativity The ANOVA carried out on averaged voltages at AF7 between 500 and 900 ms yielded a Group by Stimulus interaction [F(1,30)¼5.7, Po0.05]. Separate analyses showed that hits and correct rejections did not differ for the young [Fo1]. For older adults the Stimulus by Repetition interaction [F(1,15)¼5.1, Po0.05] indicated greater negativity for hits than correct rejections at Test 2 [F(1,15)¼17.1, P¼0.001] but not Test 1 [F(1,15)¼2.2, P40.05] (Fig. 1c). For the 900–1600 ms interval, the Group by Stimulus interaction [F(1,30)¼5.6, Po0.05] indicated that hits and correct rejections did not differ for the young [Fso2.5,


18 3 9

NEUROREPORT Ps40.05]. For older adults hits were more negative than correct rejections [F(1,15)¼20.4, Po.001] for both Tests 1 and 2 [Stimulus by Repetition F(1,15)¼1.3, P40.05; Fig. 1c].

Discussion Current theories of age-related memory decline posit that the two processes hypothesized to support episodic recognition are differentially affected by age, with recollection, but not familiarity, impaired as people age [2,3]. In line with this hypothesis, Nessler et al. (unpublished) found that while young and older adults showed a frontal episodic memory effect, a presumed correlate of familiarity, only young adults evinced a parietal episodic memory effect, a putative indicator of recollection [12]. This qualitative age-related difference in brain activity was also observed in the initial test phase (Test 1) of the current study. Most important, however, the present results demonstrate that under certain conditions, such as following the repetition of study items, older adults also show a parietal episodic memory effect in the test phase (Test 2). Although this effect in older adults remained smaller than that of the young, this result suggests that older adults are able to engage in recollection-related retrieval. Nonetheless, age-related differences in the pattern of retrieval activity remained, because only older adults seemed to utilize additional processes as reflected by the presence of reliable left-frontal negativity. It has been suggested that older adults achieve adequate recognition performance as a result of using additional processes, not recruited by the young, to compensate for deficiencies in memory retrieval [15,16]. One possible marker for such processes is the left frontal negativity [18]. Although the processes reflected by this negativity are not known with certainty [18,19,23], one possibility is that this activity reflects attentional control mechanisms that focus on task-relevant attributes of the stored memory traces. Older adults may have been able to bring these processes on line to inform their old/new decisions only after the items’ memory strengths had been augmented by repetition. This could account for the reliable 500–900 ms negative effect observed only at Test 2. As the timing of this effect was coincident with the parietal episodic memory effect at Test 2 (Fig. 1b and c), the processes it reflects could have supported recollection-based retrieval in older adults. Although this is a plausible, albeit speculative, scenario, additional work will be required to uncover the precise nature of the processes reflected by this activity and by the subsequent left-frontal negativity (900–1600 ms) that primarily followed the RT response.

Conclusion Repetition seemed to create the circumstances necessary for older adults to begin using recollective processes in their recognition decisions. However, qualitative age-related differences remained, because, in addition to the limited use of recollection, older adults recruited putatively compensatory left-frontal mechanisms.

18 4 0


NESSLER ETAL.

Acknowledgements This study was supported by Grant AG05213 from NIA and by the NYS Department of Mental Hygiene.

References 1. Craik FI, Bialystok E. Cognition through the lifespan: mechanisms of change. Trends Cogn Sci 2006; 10:131–138. 2. Spencer WD, Raz N. Differential effects of aging on memory for content and context: a meta-analysis. Psychol Aging 1995; 10:527–539. 3. Yonelinas AP. The nature of recollection and familiarity: a review of 30 years of research. J Mem Lang 2002; 46:441–517. 4. Jacoby LL, Shimizu Y, Velanova K, Rhodes MG. Age differences in depth of retrieval: memory for foils. J Mem Lang 2005; 52:493–504. 5. Daselaar SM, Fleck MS, Dobbins IG, Madden DJ, Cabeza R. Effects of healthy aging on hippocampal and rhinal memory functions: an eventrelated fMRI study. Cereb Cortex 2006; 16:1771–1782. 6. Balota DA, Dolan PO, Duchek JM. Memory changes in healthy older adults. In: Tulving E, Craik FIM, editors. The Oxford handbook of memory. New York: Oxford University Press; 2000. pp. 395–409. 7. Nessler D, Johnson R Jr, Bersick M, Friedman D. On why the elderly have normal semantic retrieval but deficient episodic encoding: a study of left inferior frontal ERP activity. NeuroImage 2006; 30:299–312. 8. Friedman D, Johnson R Jr. Event-related potential (ERP) studies of memory encoding and retrieval: a selective review. Microsc Res Tech 2000; 51:6–28. 9. Rugg MD, Mark RE, Walla P, Schloerscheidt AM, Birch CS, Allan K. Dissociation of the neural correlates of implicit and explicit memory. Nature 1998; 392:595–598. 10. Nessler D, Mecklinger A, Penney TB. Event related brain potentials and illusory memories: the effects of differential encoding. Cogn Brain Res 2001; 10:283–301. 11. Mecklinger A. Interfacing mind and brain: a neurocognitive model of recognition memory. Psychophysiology 2000; 37:565–582. 12. Rugg MD, Curran T. Event-related potentials and recognition memory. Trends Cogn Sci 2007; 11:251–257. 13. Trott CT, Friedman D, Ritter W, Fabiani M, Snodgrass JG. Episodic priming and memory for temporal source: event-related potentials reveal age-related differences in prefrontal functioning. Psychol Aging 1999; 14:390–413. 14. Johnson R Jr, Kreiter K, Russo B, Zhu J. A spatio-temporal analysis of recognition-related event-related brain potentials. Int J Psychophysiol 1998; 29:83–104. 15. Rajah MN, D’Esposito M. Region-specific changes in prefrontal function with age: a review of PET and fMRI studies on working and episodic memory. Brain 2005; 128:1964–1983. 16. Reuter-Lorenz PA, Lustig C. Brain aging: reorganizing discoveries about the aging mind. Curr Opin Neurobiol 2005; 15:245–251. 17. Cabeza R. Hemispheric asymmetry reduction in older adults: the HAROLD model. Psychol Aging 2002; 17:85–100. 18. Swick D, Senkfor AJ, Van Petten C. Source memory retrieval is affected by aging and prefrontal lesions: behavioral and ERP evidence. Brain Res 2006; 1107:161–176. 19. Czernochowski D, Fabiani M, Friedman D. Use it or lose it? SES mitigates age-related decline in a recency/recognition task. Neurobiol Aging (in press). 20. Semlitsch HV, Anderer P, Schuster P, Presslich O. A solution for reliable and valid reduction of ocular artifacts, applied to the P300 ERP. Psychophysiology 1986; 23:695–703. 21. Wegesin DJ, Friedman D, Varughese N, Stern Y. Age-related changes in source memory retrieval: an ERP replication and extension. Cogn Brain Res 2002; 13:323–338. 22. Snodgrass JG, Corwin J. Pragmatics of measuring recognition memory: applications to dementia and amnesia. J Exp Psychol Gen 1988; 117:34–50. 23. Li J, Morcom AM, Rugg MD. The effects of age on the neural correlates of successful episodic retrieval: an ERP study. Cogn Affect Behav Neurosci 2004; 4:279–293.