Age differences in brain systems supporting transient and sustained processes involved in prospective memory and working memory

YNIMG-12711; No. of pages: 11; 4C: 3, 6, 7 NeuroImage xxx (2015) xxx–xxx

Contents lists available at ScienceDirect

NeuroImage journal homepage: www.elsevier.com/locate/ynimg

3Q2

Nathalie Peira a,b, Maryam Ziaei c, Jonas Persson b,d,⁎ a

9

a r t i c l e

10 11 12 13

Article history: Received 20 March 2015 Accepted 25 October 2015 Available online xxxx

14 15 16 17 18

Keywords: fMRI Aging Working memory Prospective memory

Department of Psychology, Uppsala University, Box 1225, 751 42 Uppsala, Sweden Department of Psychology, Stockholm University, 106 91 Stockholm, Sweden School of Psychology, The University of QLD, St Lucia, QLD 4072, Australia d Aging Research Center (ARC) at Karolinska Institute and Stockholm University, Gävlegatan 16, 113 30 Stockholm, Sweden b

a b s t r a c t

P

i n f o

C

T

E

D

In prospective memory (PM), an intention to act in response to an external event is formed, retained, and at a later stage, when the event occurs, the relevant action is performed. PM typically shows a decline in late adulthood, which might affect functions of daily living. The neural correlates of this decline are not well understood. Here, 15 young (6 female; age range = 23–30 years) and 16 older adults (5 female; age range = 64–74 years) were scanned with fMRI to examine age-related differences in brain activation associated with event-based PM using a task that facilitated the separation of transient and sustained components of PM. We show that older adults had reduced performance in conditions with high demands on prospective and working memory, while no age-difference was observed in low-demanding tasks. Across age groups, PM task performance activated separate sets of brain regions for transient and sustained responses. Age-differences in transient activation were found in fronto-striatal and MTL regions, with young adults showing more activation than older adults. Increased activation in young, compared to older adults, was also found for sustained PM activation in the IFG. These results provide new evidence that PM relies on dissociable transient and sustained cognitive processes, and that age-related deficits in PM can be explained by an inability to recruit PM-related brain networks in old age. © 2015 Published by Elsevier Inc.

44 45 46 47 48 49 50 51 52 53 54 55

R

Implementing intentions is essential for successfully executing actions at some point in the future. The cognitive processes involved in remembering to perform a specific action are usually referred to as prospective memory (PM). In laboratory studies of PM, participants are typically asked to monitor the environment for the presence of a specific target cue or the arrival of a particular time, and interrupt their performance of an ongoing task to complete the intended action (Einstein GO and MA McDaniel, 1996). The components of PM can be characterized along several hypothetical dimensions including their reliance on long-term memory or executive functions, their involvement of controlled versus automatic processes, or whether they are transient (related to the cue event) or sustained over a delay prior to cue presentation (e.g. monitoring). An ongoing theoretical debate concerns to what extent cognitive operations involved in PM rely primarily on sustained attentional control processes, or automatic and spontaneous transient processes triggered by an environmental cue that is linked to a specific intention. These

N C O

42 43

Introduction

U

40 41

R

37 35 34

39

19 20 21 22 23 24 25 26 27 28 29 30 31 32

E

33 36 Q3 38

R O

c

O

4 5 6 7 8

F

2

Age differences in brain systems supporting transient and sustained processes involved in prospective memory and working memory

1Q1

⁎ Corresponding author at: Aging Research Center (ARC) at Karolinska Institute and Stockholm University, Gävlegatan 16, 113 30 Stockholm, Sweden. E-mail address: [email protected] (J. Persson).

transient and sustained processes may work in parallel during the execution of a PM task, and investigating them separately has proven difficult solely based on behavioral measures. Although only a handful of studies have investigated the neural underpinnings of component processes in PM using fMRI, results from these studies have revealed separate sets of regions involved in ongoing sustained processes related to intention maintenance and transient processes related to cue detection (Reynolds JR et al., 2009; Kalpouzos G et al., 2010; McDaniel MA et al., 2013). For example, Reynolds et al. (2009) used a mixed eventrelated/blocked design to investigate transient and sustained processes associated with PM and working memory. They were able to demonstrate that sustained increase in PFC activity was related to PM task demands and not to the need to implement working memory processes. In addition, prospective memory was associated with distinct transient activation in the temporal cortex during the presentation of PM targets. This indicates that distinct regions may subserve sustained (controlled) and transient (automatic) components of PM. Subjective PM complaints are common in old age (Zeintl M et al., 2006), and failures in performing an intended action might have severe implications in daily life (e.g. remembering to take medicine). Similar to findings on retrospective memory (e.g. episodic memory), older adults show a general impairment in PM, although age differences are more pronounced in some tasks contexts then others (West R and R Bowry,

http://dx.doi.org/10.1016/j.neuroimage.2015.10.075 1053-8119/© 2015 Published by Elsevier Inc.

Please cite this article as: Peira, N., et al., Age differences in brain systems supporting transient and sustained processes involved in prospective memory and working memory, NeuroImage (2015), http://dx.doi.org/10.1016/j.neuroimage.2015.10.075

56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78

133 134

Fifteen young (6 female; age range = 23–30 years) and 16 older individuals (5 female; age range = 64–74 years) participated in the study. One older participants' behavioral and neural data were removed from analysis due to technical problems during fMRI scanning. All participants were right-handed, Swedish speakers, and had no history of neurological or psychiatric problems. All participants were screened for claustrophobia, neurological and psychiatric medications, MRI contraindication, and all had normal or corrected to normal vision using scanner compatible glasses or contact lenses. All participants took part in two

104 105 106 107 108 109 Q5 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128

135 136 137 138 139 140 141

C

102 103

E

100 101

R

98 99

R

96 97

O

94 95

C

92 93

N

90 91

U

88 89

Procedure

156

F

Participants

86 87

O

132

85

The study consisted of two separate sessions that occurred on two separate days: first a behavioral test session on day 1, and a second fMRI scanning session on day 2. The time in between each test session was approximately one week. During the behavioral test session, participants completed the color-word Stroop test, a test if visual attention (Bundesen C, 1990; Vangkilde S et al., 2011), and the operation span working memory task (Unsworth N et al., 2005). Older individuals also completed the mini-mental state examination (MMSE). All participants also received instructions and performed practice runs of the scanner tasks in preparation for the scanning session. During the scanning session on day 2, participants again performed practice runs of the scanner task immediately prior to the scanning session.

R O

Methods

83 84

142 143

P

131 Q4

81 82

separate test sessions; one for behavioral assessment, and one for the fMRI scanning session. Informed consent was obtained from all participants. The investigation was approved by the Ethics Committee in Stockholm. Participants were paid 800 SEK for their participation. Behavioral and neuropsychological results, and group characteristics are reported in Table 1. As noted, older participants scored high on the MMSE, indicating that they were a high performing group of participants and free of dementia. There were no differences between the groups with regards to education or proportion of males/females. The cognitive data further indicates that the two age groups were representative of their respective cohorts, with an advantage for the young in operation span and Stroop task response latencies. Error rates in the Stroop task were low and did not differ significantly between age groups.

144 145 146 147 148 149 150 151 152 153 154 155

157 158 159 160 161 162 163 164 165 166 167 168 169

fMRI task

T

129 130

2005; McDaniel MA and GO Einstein, 2007; Kliegel M et al., 2008). Age differences may stem from a decreased efficiency of controlled preparatory attentional processes that underlie the detection of PM cues (West R and R Bowry, 2005; Smith RE and UJ Bayen, 2006), a decline in mechanisms related to the retrospective component of PM (Einstein GO et al., 1992; Zimmermann TD and B Meier, 2006; Zöllig J et al., 2007; Gonneaud J et al., 2011) or impairments in underlying working memory and executive control processes associated with PM (Mäntylä T, 2003; Bisiacchi PS et al., 2008; Rose NS et al., 2010; Gonneaud J et al., 2011; Schnitzspahn KM et al., 2013). Recent neuroimaging and neuropsychological evidence implicate the prefrontal cortex (PFC) as critical for executing delayed intentions in prospective memory tasks. In particular, the medial and lateral part of the anterior PFC has consistently been activated across time-based and event-based PM tasks (see Burgess PW et al., 2011 for a review). Additionally, the dorsolateral and inferior PFC, the inferior parietal cortex, precuneus, and the anterior cingulate cortex (ACC) have commonly shown enhanced activation during PM tasks (Okuda et al., in press; Simons JS et al., 2006; Haynes JD et al., 2007; Reynolds JR et al., 2009; Poppenk J et al., 2010; Bisiacchi PS et al., 2011; Burgess PW et al., 2011; Hashimoto T et al., 2011; Gonneaud J et al., 2014). In line with observations from fMRI and PET, several ERP studies have observed a frontal positivity effect associated with PM cue trials (see West R, 2011 for a review). Activation in the medial temporal lobe (MTL) has also often been observed in individuals during PM tasks performance (Okuda J et al., 1998; Martin T et al., 2007; Poppenk J et al., 2010). Thus, successful PM may rely on frontal control processes that modulate other brain regions responsible for maintaining the intention and/or for processing incoming information to identify event occurrences. Given that pronounced age-related alterations in prefrontal systems is a hallmark of normal aging (Rajah MN and M D'Esposito, 2005; Raz et al., in press; Persson J and L Nyberg, 2006), they have been assumed to underlie impairments in PM functioning with increasing age (McFarland CP and EL Glisky, 2011; West R, 2011). While this idea seems reasonable, very few studies have attempted to specifically investigate the brain correlates of age-related changes in PM. Moreover, previous studies that have examined the link between PM-related brain activation and aging have used electrophysiological methods, and, while providing excellent temporal resolution, lack the ability to test predictions about brain regional specificity. Indeed, it is noteworthy that no study to date has examined transient and sustained components of PM in relation to aging using fMRI. The current study was therefore intended to investigate age-related differences in brain activation patterns associated with transient and sustained components of PM in relation to aging using fMRI. The task was designed to facilitate the separation of transient (event-related) and sustained (blocked) components of PM. The major objectives of this study were to further investigate the previous observation of agerelated deficit in PM, to examine the neural underpinnings of ongoing and cue-related components across participants in young and older adults, and investigate age-differences in brain networks underlying these components.

D

79 80

N. Peira et al. / NeuroImage xxx (2015) xxx–xxx

E

2

A slightly modified version of a previously used PM task (Reynolds JR et al., 2009) was used as the scanner task. In the scanner, participants performed four different tasks, with each task consisting of a sequence of colored words (see Fig. 1). Oddball and PM target colors varied randomly within and between individuals, and each participant was never presented with the same target color twice within each condition (Fig. 2). In the oddball task, participants were informed of a target color, and asked to respond with their right index finger each time the target color was presented, and with their right middle finger if the word was presented in any other color. The oddball targets occurred with low frequency (~10%) and served as a control task for the low-frequency PM cues used in the PM task.

170 171

Table 1 Demographics and cognitive performance for young and older adults.

t1:1 t1:2

Demographics N Age, years (range, SD) Gender (f/m) Education, years (range, SD) Cognitive scores Operation span MMSE (range, SD) Stroop task (RT) Neutral Congruent Incongruent Stroop task (error rate)*

Young adults

Older adults

P

15 22.4 (20–26, 1.8) 6/9 2.7 (1–3, 0.5)

15 68.1 (64–74, 3.6) 5/10 2.5 (1–3, 0.8)

46.9 (20.9)

21.0 (12.6) 27.9 (26–30, 1.3)

.001

816.7 851.2 1029.3 1.81

1311.8 1410.1 1654.9 2.16

.001 .001 .001 n.s.

n.s. n.s.

172 173 174 175 176 Q6 177 178 179 180 181 182

t1:3 t1:4 t1:5 t1:6 t1:7 t1:8 t1:9 t1:10 t1:11 t1:12 t1:13 t1:14 t1:15 t1:16 t1:17

Note: Values are means (range, SD) except for Gender that represent number of partici- t1:18 pants. MMSE = mini-mental state examination. Education equals the number of years t1:19 after high school. P = p-value for the comparison of young and older adults. RT = t1:20 Reaction time. n.s. = non-significant. *Error rates were only available for all conditions t1:21 combined.

Please cite this article as: Peira, N., et al., Age differences in brain systems supporting transient and sustained processes involved in prospective memory and working memory, NeuroImage (2015), http://dx.doi.org/10.1016/j.neuroimage.2015.10.075

3

F

N. Peira et al. / NeuroImage xxx (2015) xxx–xxx

196 197 198 199 200 201

P

D

E

T

194 195

C

192 193

E

190 191

R

188 189

There were 3 scanning runs in total, each with a duration of 9 min and 45 s. Each of the three runs was composed of 4 task blocks alternating with 3 fixation blocks. The three runs were composed of one repetition each of the four task conditions (OB, 1B, 3B, PM-1B). Each fixation block had a duration of 20 s. Each task block consisted of 36 trials (4 PM or OB targets, 10 NB targets, and 22 non-targets). Each task block started with a 5000 ms instruction screen. Each trial was presented for 2000 ms with an variable inter-trial interval (ITI) that was jittered in steps of 1500 ms, with a minimum ITI of 1500 and a maximum ITI of 9000 ms (allowing for an independent estimation of the BOLD response on a trial-by-trial basis). Similar to Reynolds et al. (2009), stimuli consisted of common one- or two-syllable words, which were pseudorandomized and presented in a center position on the screen. The stimuli could appear in one of five colors (green, blue, magenta, red or yellow). The words assigned to each task condition (OB, 1B, 3B, PM-1B) were counterbalanced across participants. Each block started with an instruction screen that referred to the relevant task: (1) press when the target color is presented and ignore n-back (oddball), (2) press when the target word was presented 1 word back and ignore color

R

186 187

In the n-back WM tasks (Braver TS et al., 1997; Cohen JD et al., 1997) a series of words appeared and participants were instructed to press the response button with their index finger as quickly and accurately as possible when a word matched the one presented three words earlier (3-back) or 1 word earlier (1-back). If the current word was not identical to the word preceding or the word presented three words earlier, participants were instructed to respond by pressing with their middle finger. Memory load (3-back vs. 1-back) varied across blocks. In the nback conditions, and similar to previous fMRI studies using this task (Braver et al., 1997; Cohen et al., 1997), target trials occurred on approximately 30% of the trials. In the PM task, participants simultaneously performed the oddball detection and the n-back WM tasks. Participants performed the 1back task unless a target color was presented, in which case they were instructed to make a response with their right index finger. The fact that participants needed to be sensitive to infrequent target colors while at the same time as performing the ongoing n-back task results in high demands on PM processes that are not present in the other conditions.

N C O

184 185

U

183

R O

O

Fig. 1. Overview of the cognitive tasks. The oddball task required the participants to respond to an oddball target which was defined as a word occurring in a color specified at the beginning of the current task block. In the N-back task, a target was defined as a word that was repeated from the previous trial (1-back) or a repetition of a word presented three trials previously (3–31 back). The prospective memory (PM) task, required participants to 1 respond to color targets, while simultaneously performing the 1-back task if the current stimulus is not a color target.

Fig. 2. Behavioral results. Mean performance on the oddball, working memory and prospective memory tasks task (top — number of accurate responses; bottom — response time in milliseconds). Bars represent 1 S.E.M. *P b 0.05, **P b 0.01. From left to right, each bar represents the response to the oddball (OB), 1-back (1B), 3-back (3B), and prospective memory 1-back (PM-1B), respectively.

Please cite this article as: Peira, N., et al., Age differences in brain systems supporting transient and sustained processes involved in prospective memory and working memory, NeuroImage (2015), http://dx.doi.org/10.1016/j.neuroimage.2015.10.075

202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220

Image acquisition Magnetic resonance imaging was performed on a 3-Tesla General Electric scanner MR750 equipped with a 32 channel head coil. Functional data was acquired with a gradient echo-planar imaging sequence (37 transaxial slices, odd–even interleaved, 2 mm in plane resolution, thickness: 3.4 mm, gap: 0.5 mm, repetition time [TR]: 2000 ms, echo time [TE]: 30 ms, flip angle: 80°, field of view: 25 × 25 cm). In order to allow for progressive saturation of the fMRI-signal, 10 dummy scans were collected and discarded prior to experimental image acquisition. High-resolution T1-weighted structural images were also collected with a 3D fast spoiled gradient echo sequence (180 transaxial slices, with a 1 mm thickness in plane resolution, TR = 8.2 ms, TE = 3.2 ms, flip angle 12°, field of view 25 × 25 cm). Each scanner task was presented to the participants on a computer screen, seen through a mirror mounted on the head coil, and responses were collected on a scannercompatible response box. Participants were given headphones and earplugs to dampen scanner noise, and cushions inside the head coil to minimize head movements.

246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282

Image preprocessing All fMRI data were analyzed within statistical parametric mapping software (SPM8, Welcome Department of Imaging Neuroscience, University College London, UK) implemented in Matlab 2010b (Mathworks Inc., MA). Pre-processing included slice-timing correction, motion correction, co-registration of functional images to participants' anatomical scans, spatial normalization using the EPI-MNI template provided in SPM8 (volumes re-sampled to a resolution of 2 × 2 × 2 mm), and smoothing (8-mm full-width half maximum [FWHM] Gaussian kernel).

Results

338

Behavioral results

339

Accuracy and reaction time measures were analyzed separately. Reaction times for non-target trials were unfortunately not collected, and are therefore not included in the behavioral analysis. Median RTs were used in order to avoid undue influence from deviant reaction times. The first set of analyses included accuracy data.

340

T

244 245

C

242 243

fMRI data analysis Event-related and sustained effects were modeled in the framework of the general linear model (GLM) as implemented in SPM8. For individual subject analyses, and event-related transient responses, a fixed effects event-related design was implemented using multiple linear regression time series analyses to determine the location and extent of brain activations associated with different trial types. In order to estimate event-related activation for different trial types, separate regressors for oddball targets, and 1-back and 3-back targets, were included in the GLM, along with regressors coding for events in the PM task; PM cues and PM 1B targets. For the estimation of sustained activation, regressors were included that coded for each task block (oddball, 1-back, 3-back, PM and fixation). The frequent non-target trials may be less reliably separable from the constant of the GLM once they have been convolved with the HRF as they are close together in time and the HRF is sustained. Consequently, non-target trials and inter-trial fixations were not explicitly modeled in the design matrix. Hemodynamic responses for estimating transient activation were modeled using the SPM canonical hemodynamic response function (HRF). Sustained (blocked) responses were modeled as a box-car function, and convolved with the canonical HRF. An AR(1) model was used to estimate and correct for non-sphericity of the error covariance. Single-subject statistical contrasts were set up using the general linear model, and group data were analyzed in a random-effects model. Only onsets of trials with correct responses were used for further analyses. In addition, three translational (x, y, z) and three rotational (pitch, roll, yaw) regressors obtained from the realignment step were included as covariates of no interest in the individual fixed effect analysis to account for inscanner movement. Single-subject statistical contrasts were set up using the GLM, and group data were analyzed in a random-effects

E

240 241

R

238 239

R

236 237

O

234 235

C

232 233

N

230 231

U

228 229

F

226 227

O

fMRI scanning

283 284

R O

225

model that differentiated between conditions. Statistical parametric maps were generated using t statistics to identify regions activated according to the model. All results are reported in MNI space. The Marsbar toolbox (http://marsbar.sourceforge.net/) was used to create regions-of-interest (ROIs), and to extract each ROI's mean BOLD parameter estimate value for each condition within each subject. ROIs were derived from functional peaks obtained in group level contrasts of interest, and were defined as a functional cluster that survived a threshold of P b 0.001 uncorrected for multiple comparisons. ROIs contained a minimum of 30 contiguous voxels. The Marsbar toolbox was also used to extract each ROI's mean parameter estimate value (% signal change) across all voxels contained in each cluster for each of the 5 conditions. These values were then used for plotting the results in SPSS, as well as for performing brain–behavior correlations. Importantly, no inferential statistics were performed on the BOLD signal estimates to investigate between condition differences or group by condition interactions. A set of contrasts were used to ensure that brain activity associated with PM was distinct from processes engaged during the ongoing task (n-back) and color target detection (oddball). Moreover, we were interested in the separate contributions of sustained processes across trials and transient processes that are trial specific on brain activation during PM performance. As a first step, activation associated with transient and sustained components of PM was investigated across age groups and for young and older adults separately. Sustained PM activation was identified by contrasting blocks of PM trials with blocks of oddball and 1-back trials (PM-1B–(OB + 1B); uncorrected threshold: P b 0.001). Moreover, PM blocks had to show increased activation compared to fixation baseline blocks. Using an exclusive masking procedure with the contrast of 3-back vs. 1-back, we were able to ensure that activation was specifically linked to PM, and not to oddball activation or working memory load (3B–1B; uncorrected threshold: P b 0.001). Using a 3-back vs. 1-back inclusive mask, we were able to address the question of similarities (overlap) in activation between PM and WM load (3B–1B; uncorrected threshold: P b 0.001). For transient responses, we contrasted PM cue trials with oddball, PM 1-back, 1-back trials (PMcue–(PM-1B + OB + 1B)). Similar to the analyses of sustained responses, we used a masking procedure to investigate unique contribution of PM processing that was not related to WM load, and brain responses that were common to PM and WM (see above). In a second series of analyses, brain correlates of age-difference in PM was investigated. Age comparisons consisted of directly comparing transient and sustained brain responses for young N older adults and the reverse using the contrast files as described above. For example, increased sustained PM activation in young, compared to older adults were investigated using the following contrast: young (PM-1B–(OB + 1B)) N older (PM-1B–(OB + 1B)) using an uncorrected threshold of P b 0.001, and masked with WM related activation across all participants (3B–1B; uncorrected threshold: P b 0.001). In order to investigate the relationship between behavioral performance and brain activation, we performed correlation analyses (Pearson's correlation coefficient) on the BOLD signal estimates extracted from selected ROIs and reaction time and accuracy measures of PM for younger adults, older adults, and all participants together.

P

223 224

targets (1-back), (3) press when the target word was presented 3 words back and ignore color targets (3-back), and (4) press when the target color appears or when the target word was presented one word back (PM-1B).

D

221 222

N. Peira et al. / NeuroImage xxx (2015) xxx–xxx

E

4

Please cite this article as: Peira, N., et al., Age differences in brain systems supporting transient and sustained processes involved in prospective memory and working memory, NeuroImage (2015), http://dx.doi.org/10.1016/j.neuroimage.2015.10.075

285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337

341 342 343 344

N. Peira et al. / NeuroImage xxx (2015) xxx–xxx

366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410

Transient and sustained PM effects across age groups As a first step, we investigated brain activation patterns associated with prospective and working memory across age groups. As noted earlier, we were particularly interested in effects that were specifically related to PM load, but not WM load, but also in activation shared across the two tasks (i.e. commonly modulated by WM and PM). Transient and sustained effects were investigated separately. Inclusive/exclusive masking was applied using an uncorrected threshold of P b 0.001 for the mask. All masked as well as unmasked contrasts were evaluated using 2-tailed T-contrasts under an uncorrected alpha level of 0.001, and a minimum cluster size of 10 contiguous voxels. Transient effects associated with PM were investigated by contrasting PM cue trials with OB, 1B and PM-1B trials. For activation specific to PM, this contrast was exclusively masked with results from the 3B N 1B contrast and thus ensuring that activation was specifically related to PM load and not WM load. This contrast revealed a set of regions including right middle temporal cortex, left caudate nucleus, left posterior cingulate, left thalamus, left medial temporal lobe, and left inferior occipital cortex (Fig. 3). In order to investigate regions commonly activated for PM and WM, the PM contrast was inclusively masked with the contrast for WM load. No significant regions were commonly activated by PM and WM. Sustained effects were investigated using the contrast of PM N OB and 1B. Similarly to the analysis of transient effects, this contrast was exclusively masked with the results from the contrast of 3B N 1B to make sure that the effect were not related to WM load. Results from this contrast revealed PM-related activation in left putamen, left inferior frontal gyrus (IFG), right anterior cingulate cortex (ACC), and right hippocampus (Fig. 4). Regions that were commonly activated by PM and WM included left IFG, left parietal cortex, right IFG, and the cerebellum (Fig. 4). To ensure the commonality for these brain networks associated with PM across both age groups, an inclusive mask technique was also used. Hence, the activation pattern for PM related transient and sustained activation, respectively, for the young group was used to inclusively mask the activation pattern associated with PM in older adults. We found that the activation patterns overlapped between the groups, and were therefore able to conclude that the activation patterns was present in both age groups. While not being the main scope of the current paper, we also examined brain activation related to working memory load by comparing the 3-back task to the 1-back task (3B N 1B). This analysis included both young and older adults. Transient activation associated with working memory load, was mainly associated with frontal, occipital and parietal activation. Sustained activation association with working memory load engaged a network that included a fronto-parieto-striatal network (Supplemental Table 3). This is in line with much evidence showing increased activation in these regions in response to working memory demands (Courtney SM et al., 1998; D'Esposito M et al., 1998; Jonides J et al., 1998; Pessoa L et al., 2002; Owen AM et al., 2005; D'Ardenne K et al., 2012; Takahashi E et al., 2013).

421

O

F

420

R O

364 365

fMRI results

P

362 363

411 412

D

360 361

significant main effect of condition (F (1, 29) = 61.22, P b .001, ηp2 = .679), but the main effect of age (F (1, 29) = 1.28, P = .264, ηp2 = .043), and the age by condition interaction did not reach significance (F (1, 29) = 1.18, P = .286, ηp2 = .039). Finally, the 2 (group; young/older) by 2 (condition; PM effect/WM load effect) repeated-measures ANOVA showed a main effect of condition (F (1, 29) = 31.23, P b .001, ηp2 = .519), but the main effect of age (F (1, 29) = 0.817, P = .373, ηp2 = .027), and the age by condition interaction was not significant (F (1, 29) = .917, P = .346, ηp2 = .031).

E

358 359

T

356 357

C

354 355

E

352 353

R

351

R

349 350

N C O

347 348

First, a 2 (group; young/older adults) by 2 (condition; PM cue/oddball) repeated-measures ANOVA was performed to investigate the PM interference effect of PM cue targets compared to oddball targets. In this analysis, we found a main effect of condition (F (1, 29) = 26.95, P b .001, ηp2 = .482), a main effect of age group (F (1, 29) = 6.71, P = .015, ηp2 = .188), and a significant group by condition interaction (F (1, 29) = 4.98, P = .034, ηp2 = .147) showing that older adults performed disproportionally worse on PM cue targets compared to oddball targets. The PM interference effect on 1-back targets was examined in a 2 (group; young/older) by 2 (condition; PM 1B/1B) repeated-measures ANOVA. In this analysis we found a significant main effect of condition (F (1, 29) = 19.9, P b .001, ηp2 = .408), and age (F (1, 29) = 6.52, P = .016, ηp2 = .184), and a significant age by condition interaction (F (1, 29) = 5.89, P = .022, ηp2 = .169). For non-targets, the main effect of condition (F (1, 29) = 3.68, P = .065, ηp2 = .113), the main effect of age (F (1, 29) = 3.52, P = .071, ηp2 = .029), and the age by condition interaction (F (1, 29) = .877, P = .357, ηp2 = .029) were all nonsignificant. This shows that, at least for target trials and similar to the analysis on oddball trials, participants performed worse on the PM 1B trials compared to the 1-back, and that older adults performed disproportionally worse compared to young adults on PM 1B trials. The effect of working memory load on performance was investigated in a 2 (age; young/older) by 2 (condition; 1B/3B). The analyses on target trials showed that performance on the 3B task was significantly lower compared to the 1B task (F (1, 29) = 197.3, P b .001, ηp2 = .872). Moreover, older adults performed worse compared to young adults (F (1, 29) = 14.74, P b .001, ηp2 = .337), and this effect was larger for 3B compared to 1B as showed by the significant age by condition interaction (F (1, 29) = 15.59, P b .001, ηp2 = .350). For non-target trials, performance in the 3B condition was lower compared to the 1B condition (F (1, 29) = 6.62, P = .015, ηp2 = .186). The main effect of age approached significance (F (1, 29) = 3.95, P = .056, ηp2 = .120), but the age by condition interaction was not significant (F (1, 29) = .076, P = .784, ηp2 = .003). Finally, we examined whether performance was more affected by increased working memory load or demand for PM. This was performed by calculating estimates for the PM effect (PM cue–oddball) and working memory load (3B–1B). These estimates were included in a 2 (group; young/older) by 2 (condition; PM effect/WM load effect) showing that the WM load effect was larger than the PM effect (F (1, 29) = 36.68, P b .001, ηp2 = .558). Moreover, older adults performed worse compared to young adults (F (1, 29) = 15.24, P b .001, ηp2 = .344), and this difference was similar for the PM and WM effects, as indicated by the non-significant age by condition interaction (F (1, 29) = 0.421, P b .522, ηp2 = .014). For non-target trials, a main effect of condition (F (1, 29) = 29.59, P b .001, ηp2 = .505) indicated that the working memory load effect (i.e. reduced accuracy for 3B compared to 1B) was more pronounced compared to the PM effect. Neither the main effect of age (F (1, 29) = .096, P b .759, ηp2 = .003), nor the age by condition interaction was significant (F (1, 29) = 1.93, P = .175, ηp2 = .063). A similar analysis was performed on the median response time data for correct target trials. The 2 (group; young/older adults) by 2 (condition; PM cue/oddball) repeated-measures ANOVA revealed a significant main effect of condition (F (1, 29) = 61.28, P b .001, ηp2 = .679), a significant main effect of age (F (1, 29) = 9.12, P = .005, ηp2 = .239). The age by condition interaction was not significant (F (1, 29) = .982, P = .333, ηp2 = .033). The PM effect of 1-back trials from the 2 (group; young/older) by 2 (condition; PM 1B/1B) repeated-measures ANOVA showed a significant main effect of condition (F (1, 29) = 12.23, P = .002, ηp2 = .298), while neither the main effect of age (F (1, 29) = .224, P = .639, ηp2 = .008), nor the group by condition interaction was significant (F (1, 29) = .038, P = .846, ηp2 = .001). The analysis on working memory load using a 2 (age; young/older) by 2 (condition; 1B/3B) repeated-measures ANOVA revealed a

U

345 346

5

413 414 415 416 417 418 419

422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470

Age differences in transient and sustained PM effects 471 Similar to previous analyses, transient and sustained effects were 472 investigated separately. Age comparisons consisted of whole-brain 473

Please cite this article as: Peira, N., et al., Age differences in brain systems supporting transient and sustained processes involved in prospective memory and working memory, NeuroImage (2015), http://dx.doi.org/10.1016/j.neuroimage.2015.10.075

N. Peira et al. / NeuroImage xxx (2015) xxx–xxx

R O

O

F

6

U

N

C

O

R

R

E

C

T

E

D

P

Fig. 3. Transient BOLD signal changes in response to prospective memory demands across young and older adults. From left to right, each bar represents the response to the oddball (OB), 1-back (1B), 3-back (3B), prospective memory 1-back (PM 1B), and prospective memory oddball (PM Cue) tasks, respectively.

Fig. 4. Sustained BOLD signal changes in response to prospective memory, or to both prospective memory and working memory, across young and older adults. From left to right, each bar represents the response to the oddball (OB), 1-back (1B), 3-back (3B), and the prospective memory (PM) tasks, respectively.

Please cite this article as: Peira, N., et al., Age differences in brain systems supporting transient and sustained processes involved in prospective memory and working memory, NeuroImage (2015), http://dx.doi.org/10.1016/j.neuroimage.2015.10.075

N. Peira et al. / NeuroImage xxx (2015) xxx–xxx

494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515

F

O

Discussion

Brain–behavior correlations To further elucidate the role brain activation may play in successfully executing an intended action, BOLD signal estimates in regions showing transient PM effects were correlated with RT and accuracy for detecting

R O

492 493

This study investigated age-differences in brain mechanisms underlying PM performance. Behavioral results show that older adults performed with lower accuracy in conditions with high demands on prospective and working memory. No age differences were observed in low demanding tasks, such as the oddball and 1-back tasks. Older adults also performed differentially slower on PM cue trials than oddball trials compared to younger adults. Regional differences in brain activity level across age groups associated with transient PM processes were found in a network including striatal, MTL, thalamic, temporal, medial and lateral occipital regions. In contrast, sustained activity associated with PM was found largely in lateral frontal regions and the ACC, along with MTL, parietal, striatal and cerebellar regions, some of which overlapped with activation related to WM processing. Agedifferences in transient activation were predominantly found in a fronto-striatal network, and the MTL, in which young adults showed more activation compared to older adults. More pronounced activation in young adults compared to older adults was also found for sustained PM activation in the right IFG. Finally, direct brain–behavior correlations showed a negative relationship between BOLD signal in inferior frontal cortex and RTs for detecting PM cues, supporting the notion that brain activation in this region is related to the behavioral outcome of this task.

P

490 491

Complementary analyses in each age group separately Given that we could not find support for anterior PFC activation for neither transient nor sustained effects across young and older participants, which has been a common finding in several previous neuroimaging studies in young adults, we also performed similar analyses as described above but included only young adults. For transient effects (PM cue N OB, 1B, PM-1B, masked with 3B N 1B) activation was found in bilateral parietal cortex, left middle temporal gyrus, right precentral gyrus, left insula and left precuneus/parahippocampal gyrus (Supplemental Table 1). Significant sustained effects (PM N OB & 1B, masked with 3B N 1B) were observed in bilateral IFG, left anterior PFC, left temporal pole, and left insula (Supplemental Table 1). This finding replicates previous observations in young adults that consistently have found activation in anterior lateral and medial frontal regions during PM tasks (Burgess PW et al., 2008). Moreover, in order to examine brain activation patterns that were specific for the older group, we performed a similar analysis for older adults only. For transient effects (PM cue N OB, 1B, PM-1B, masked with 3B N 1B) activation was found in sub-cortical regions (caudate nucleus, thalamus, and parahippocampal gyrus), along with activation in medial frontal, occipital, inferior temporal, and cerebellar activation (Supplemental Table 2). Significant sustained effects (PM N OB & 1B, masked with 3B N 1B) were observed in middle and inferior frontal regions, putamen and the hippocampus (Supplemental Table 2).

D

488 489

E

487

a PM cue. A false discovery rate (FDR) correction was applied for multiple statistical tests. In a first step, these correlation analyses was performed across all participants. Activation in one region, left IFG were negatively correlated with RT, indicating that increased activation in this region was associated with faster response times (Fig. 6). Multiple regression was thereafter used to model the interaction between RT and age on brain activation. This analysis showed that age did not moderate the relationship between PM RT and left IFG activation (P = .75), and therefore age was not used as a covariate in the analysis. Moreover, when correlations were performed independently for each age group, we found that another region within the left IFG were negatively correlated with RT, but this relationship was only seen in young adults. Older adults showed a positive, although non-significant relationship between BOLD signal and RT in this particular region. The difference between these correlations was statistically significant (Z = 2.42, P b .05), further suggesting that young and older adults recruit this region differently in response to high demands on PM cue detection.

T

485 486

C

483 484

E

481 482

R

480

R

478 479

N C O

476 477

analysis in which we directly compared transient and sustained brain responses for young N older adults and using the contrasts as described above (i.e. an age by condition interaction in SPM). Transient effects associated with PM were investigated by contrasting PM cue trials with OB, 1B and PM-1B trials, and this contrast was masked with results from the 3B N 1B contrast and thus ensuring that activation was specifically related to PM load and not WM load. Age differences in this contrast showed that young adults had more pronounced engagement of bilateral IFG, ACC, MTL and the caudate nucleus relative to older adults (Fig. 5A). No regions were significantly more activated for older adults compared to young adults. Age differences for sustained PM processing (age differences in PM N OB and 1B, masked with the WM load contrast) was only found in one region, the right IFG, for which young adults had stronger activation compared to older adults (Fig. 5B).

U

474 475

7

Fig. 5. Age-differences in transient (A) and sustained (B) brain activation associated with prospective memory. (A) From left to right, each bar represents the response to the oddball (OB), 1-back (1B), 3-back (3B), prospective memory 1-back (PM 1B), and prospective memory oddball (PM Cue) tasks, respectively. (B) From left to right, 1 each bar represents the response to the oddball (OB), 1-back (1B), 3-back (3B), and the prospective memory (PM) tasks, respectively.

Please cite this article as: Peira, N., et al., Age differences in brain systems supporting transient and sustained processes involved in prospective memory and working memory, NeuroImage (2015), http://dx.doi.org/10.1016/j.neuroimage.2015.10.075

516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533

534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556

N. Peira et al. / NeuroImage xxx (2015) xxx–xxx

O

F

8

563 564 565 566 567 568 569 570 571

C

561 562

E

559 560

Task-related transient responses across age groups

573

The first set of analyses revealed that, across age groups, transient PM activation was found in MTL, thalamic, striatal, temporal and occipito-parietal regions. Somewhat surprisingly, no activation was found in frontal regions. However, a similar network of regions have often been associated with transient effects during episodic memory retrieval (Rugg MD and KL Vilberg, 2013), and the MTL has previously been implicated in PM processes (Okuda J et al., 1998; Martin T et al., 2007; Kalpouzos G et al., 2010; Gordon BA et al., 2011). One possibility is that these regions are part of a network supporting associative retrieval acting in an automatic, non-strategic fashion in which attended information (PM cues) automatically produce interactions between the cue and a memory trace (i.e. performing an action). Alternatively, activation in these regions reflect operations involved in detecting a matching cue that is consistent with an individual's specific goal (i.e. memory representation for a color), a process that may rely also on top-down signals. This second view is supported by recent evidence of “match enhancements” in the hippocampus (Duncan K et al., 2009), and that this effect is determined by the association between a test probe and an explicit goal state, in which a response to matching test probes reflects a memory signal denoting the match of external inputs with internal goals. In line with this notion, fMRI data has shown that retrieval responses are not automatically triggered by memory cues, but rather are effected by goal-directed attention (e.g. Dudukovic NM and AD Wagner, 2007). Additionally, many of the regions showing transient PM responses have been identified as belonging to a network of supporting different

582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598

R

O

C

580 581

N

578 579

U

576 577

R

572

574 575

Task-related sustained responses across age groups

612

The current findings of activation in a fronto-striatal network is largely in line with previous fMRI investigations of PM which consistently have shown brain activation related to sustained attentional control (Simons JS et al., 2006; Reynolds JR et al., 2009; Burgess PW et al., 2011; McDaniel MA et al., 2013). The finding of sustained activation in the MTL is also in line with previous observations. For example, neuropsychological studies that included patients with lesions restricted to the hippocampus have reported impairments in PM tasks (Adda CC et al., 2008), and structural MRI findings have demonstrated an association between MTL volume and PM performance (Gordon BA et al., 2011). We did not find any evidence of activation in the anterior PFC for the PM condition across young and older individuals, a finding that has been reported in a number of previous studies that included only young adults. However, in two complimentary analyses that were performed on each age group separately, sustained activation was observed in the PFC, including the anterior PFC (Supplemental Tables 1 and 2). The anterior PFC regions that were activated separately for young and older adults differed slightly and this finding might have contributed to the lack of anterior PFC activation across participants. Regardless, this finding corroborate previous observations of anterior PFC activation in PM tasks (Burgess PW et al., 2011), and further suggests that activation in this particular region is primarily associated with sustained processes of PM (see Reynolds JR et al., 2009 for a similar observation). Sustained fronto-striatal network activation across PM trials suggests that these regions might assist in maintaining an intention in parallel to performing another separate task, in order to execute the action when an appropriate cue appears. The ACC and putamen may be involved in top-down context processing important for PM in order to monitor and sustain attention throughout task periods. While these particular regions are typically linked to WM (Owen AM et al., 2005;

613

P

The behavioral results showing reduced accuracy and slower responses for older adults for high demand prospective and working memory conditions corroborate previous observations that older adults display impaired PM performance (West R, 2005; McDaniel MA and GO Einstein, 2007; Kliegel M et al., 2008; Gonneaud J et al., 2011). Importantly, no age differences were observed when individuals only responded to color cues in the oddball task, indicating that an age deficit was only present when the demands on PM were high. A similar pattern of results were found for the working memory task, where age differences were present only for the more demanding 3-back task, and not for the 1-back task. This is in agreement with numerous observations of older adults showing greater decline in performance with increased WM load compared to young adults (e.g. Jaeggi SM et al., 2009; Nagel IE et al., 2009; Kessels RP et al., 2011; Nagel IE et al., 2011).

599

D

558

forms of declarative memory retrieval, namely episodic memory, semantic memory, and autobiographical memory (Burianova H et al., 2010). Such results indicate that these regions constitute a functional network supporting common processes of memory retrieval. These commonalities further support evidence that cognitive processes are shared between retrospective and prospective memory, and that certain aspects of retrospective memory are required for prospective memory. Such common underlying processes may include selective attention, error monitoring, and response verification. Additionally, parietal, occipital and temporal lobe activation may be associated with carrying representational (color) information that supports decisionmaking processes when the PM cue is presented (see also Gonneaud J et al., 2014).

E

Behavioral results

T

557

R O

Fig. 6. (A) Negative correlation between transient activation and RT in the PM cue condition in the left IFG (−44 10 18) across young and older adults (with age as a covariate), and (B) negative correlation between transient activation and RT in the left IFG (−44 28 −2) for young, but not older, adults.

Please cite this article as: Peira, N., et al., Age differences in brain systems supporting transient and sustained processes involved in prospective memory and working memory, NeuroImage (2015), http://dx.doi.org/10.1016/j.neuroimage.2015.10.075

600 601 602 603 604 605 606 607 608 609 610 611

614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642

N. Peira et al. / NeuroImage xxx (2015) xxx–xxx

667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706

C

665 666

E

663 664

R

661 662

R

659 660

N C O

657 658

U

Interestingly, one region, the right IFG showed stronger sustained activation in young compared to older adults. This finding corroborates previous observations in young adults (Simons JS et al., 2006), and further suggests that sustained processes subserved by this region are affected in older adults. As discussed above, the right PFC has often has been associated with sustained attentional functions (Corbetta M and GL Shulman, 2002), and with prospectively maintaining task-sets over longer timescales (Coull JT et al., 1996; Coull JT et al., 1998; Lawrence NS et al., 2003; Kim J et al., 2006). Thus, it is most likely that older adults are impaired in maintaining attention towards the task-cue across trials in order to detect these PM cues. This is also in line with previous demonstrations that older adults appear to rely more on reactive compared to proactive strategies for executive control; a shift that might be mediated by declining dopamine availability and/or to a reactive strategy, likely mediated by declining dopamine availability and compromised PFC functioning (Paxton JL et al., 2008; Braver TS et al., 2009; Jimura K and TS Braver, 2010; Andrews-Hanna JR et al., 2011) It should be noted that many PM tasks, including the one used in the present study, are designed as dual task paradigms where participants are instructed to carry out a primary task in parallel to a PM task. Since dual task paradigms are known to require recruitment of cognitive resources, PM interference effects and associated brain activation may also reflect increased dual task demands in the PM task. Indeed, cognitive processes implicated in dual task paradigms may be also critically involved in PM tasks in which participants, once they have detected the PM cue, is required to shift attention from the external stimuli of the ongoing task to the internal representation of the prospective intention, to disengage from the motor plan associated with the ongoing task to that of the PM task, and subsequently again focus attention to the ongoing task. The processes associated with dual-task performance and those implicated in PM execution cannot be entirely dissociated with the current task design. Importantly for the interpretation of the present results, ample evidence has demonstrated that age differences are more pronounced during dual task processing compared to performing a single task (e.g. Salthouse, 1991). A limitation of the present study is the small sample size, and that older adults most likely consisted of a group of high performing older adults compared to the general population. Clearly, future studies with larger sample sizes are needed to confirm our data. With regards to sample characteristics, it is most likely that the group of older adults constitutes a group of relatively high performing individuals and may therefore not be representative for the whole population of older adults. It is also plausible that older adults are a more select group compared to young adults, which may consist of individuals more typical for their age group. While this is a general issue in aging research, it should also be noted that older adults, while potential consisting of a sample of high-performing individuals, still showed reduced performance on most cognitive measures. Another limitation is that the participants were instructed to respond to PM cue targets and targets 1-back targets with the same response button. This may confound the results by not being able to dissociate between correctly identifying a PM cue/1-back targets and a non-target false alarm. However, given the very low frequency of 1-back/PM false alarms, we believe that this may not have significantly affected the results. Finally, one limitation is that nontarget trials were not included in the statistical model for estimating brain activity associated with the different conditions. A much discussed distinction in work on PM is between focal vs. non-focal PM tasks. A focal PM task involves detailed processing of the PM cue as part of the ongoing task, while a non-focal PM task, involves

712

F

655 656

711

O

Age-differences in transient responses related to correct responses to PM cues were found in a fronto-parietal network, and in the MTL. In all these regions, young adults showed stronger activation compared to their older counterparts. Indeed, results from several previous behavioral studies have indicated that older adults perform differentially worse on tasks that load heavily on frontal function, compared to young adults (Cabeza R and NA Dennis, 2013), including PM tasks (McDaniel MA et al., 1999; McFarland CP and EL Glisky, 2009, 2011; McDaniel MA et al., 2013). Moreover, behavioral evidence suggest a link between inhibitory control functions and prospective memory performance (Kliegel M and T Jäger, 2010), suggesting that age-related impairments in PM may, at least partly, be related to deficits in executive control functions subserved by the PFC. Alternatively, findings from studies using event-related potentials (ERPs) have led to proposals that age-related differences in prospective memory result from a decline in the functional integrity of the brain system that sustain the detection of PM cues (West R, 2005). These views may not be exclusive, however, and age related impairments may be related to one of these functions, or both. Critically, activation in one left IFG region was differentially related to PM response time, with young adults showing a robust negative correlation, while the correlation for older adults was non-significant. This relationship provides a link between performance and left IFG activation, and suggests a reduced capacity for older adults to engage task-relevant brain regions in response to PM task demands. Although the current findings provide the first evidence of agerelated reduced frontal-striatal and MTL functioning in PM tasks, observations from episodic memory tasks have commonly implicated these regions as critically involved in memory impairments in older adults (Persson J and L Nyberg, 2006; Grady CL, 2008; Nyberg L et al., 2010). For example, frontal regions have shown to be commonly co-activated across task domains (Duncan J and AM Owen, 2000; Fletcher PC and R Henson, 2001; Corbetta M and GL Shulman, 2002; Nyberg L et al., 2003; Marklund et al., 2007a, 2007b; Burianova H et al., 2010), suggesting that they are involved in processes that are shared among different memory tasks. Moreover, frontal and striatal regions are functionally connected (Burianova et al., 2010). Unfortunately, the present study cannot resolve the question of how certain PM sub-components are impaired by increasing age, and how these specifically related to changes in this fronto-parietal–MTL network. Clearly, more studies are needed to address this issue. The finding of reduced PM activation in the MTL for older compared to young individuals corroborates numerous findings of age-related impairments in MTL functioning (e.g. Raz N et al., 2004; Sperling RA, 2007; Persson J et al., 2012). While age-related alteration of MTL activation during PM has never previously been reported, this finding clearly suggest that impaired PM performance in older adults may be linked to an inability to recruit the MTL when the task demands for detecting PM cues is high. This inability could either be related to the retrospective component for remembering task-related cue-response mappings, or other strategic processes involved in retrieving information about whether a specific cue matches a current goal state (i.e. PM cue color). Moreover, the basal ganglia, including the caudate nucleus, has been shown to have a role in the execution of intentions, and the present results suggest that older adults, who also perform worse on the PM task,

Age-differences–sustained responses

R O

653 654

649

707 708

P

Age differences–transient responses

647 648

show reduced activation in these regions. Together, these results show that regions with a critical role in PM are differentially involved in young and older adults, and that deficiencies in older adults' ability to recruit these regions may underlie age differences in PM performance.

D

652

645 646

T

650 651

Marklund et al., 2007a, 2007b), it might be that these regions are associated with sustained processes during PM, but transient responses in the WM task. Importantly, activation in the putamen, MTL, and ACC was specific to PM and did not reflect processes engaged during working memory. However, in inferior frontal cortex, left parietal cortex and the cerebellum, sustained activation was related to both PM and WM load. Parietal activation has been observed for both PM and WM tasks, and has previously been reported as showing sustained activity in a similar paradigm (Reynolds JR et al., 2009).

E

643 Q7 644

9

Please cite this article as: Peira, N., et al., Age differences in brain systems supporting transient and sustained processes involved in prospective memory and working memory, NeuroImage (2015), http://dx.doi.org/10.1016/j.neuroimage.2015.10.075

709 710

713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 Q8 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770

793

Appendix A. Supplementary data

794 795

Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.neuroimage.2015.10.075.

796

References

797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839

Adda, C.C., Castro, L.H., Além-Mar e Silva LC, de Manreza ML, Kashiara R., 2008. Prospective memory and mesial temporal epilepsy associated with hippocampal sclerosis. Neuropsychologia 46, 1954–1964. Andrews-Hanna, J.R., Mackiewicz Seghete, K.L., Claus, E.D., Burgess, G.C., Ruzic, L., Banich, M.T., 2011. Cognitive control in adolescence: neural underpinnings and relation to self-report behaviors. PLoS One 6, e21598. Bisiacchi, P.S., Tarantino, V., Ciccola, A., 2008. Aging and prospective memory: the role of working memory and monitoring processes. Aging Clin. Exp. Res. 20, 569–577. Bisiacchi, P.S., Cona, G., Schiff, S., Basso, D., 2011. Modulation of a fronto-parietal network in event-based prospective memory: an rTMS study. Neuropsychologia 49, 2225–2232. Braver, T.S., Cohen, J.D., Nystrom, L.E., Jonides, J., Smith, E.E., Noll, D.C., 1997. A parametric study of prefrontal cortex involvement in human working memory. NeuroImage 5, 49–62. Braver, T.S., Paxton, J.L., Locke, H.S., Barch, D.M., 2009. Flexible neural mechanisms of cognitive control within human prefrontal cortex. Proc. Natl. Acad. Sci. U. S. A. 6, 7351–7356. Bundesen, C., 1990. A theory of visual attention. Psychol. Rev. 97, 523–547. Burgess, P.W., Dumontheil, I., Gilbert, S., Okuda, J., Schölvinck, M., Simons, J., 2008. On the role of rostral prefrontal cortex (area 10) in prospective memory. In: Kliegel, M., McDaniel, M.A., Einstein, G.O. (Eds.), Prospective Memory: Cognitive, Neuroscience, Developmental, and Applied Perspectives. Mahwah, Erlbaum. Burgess, P.W., Gonen-Yaacovi, G., Volle, E., 2011. Functional neuroimaging studies of prospective memory: what have we learnt so far? Neuropsychologia 49, 2246–2257. Burianova, H., McIntosh, A.R., Grady, C.L., 2010. A common functional brain network for autobiographical, episodic, and semantic memory retrieval. NeuroImage 49, 865–874. Cabeza, R., Dennis, N.A., 2013. Frontal lobes and aging: deterioration and compensation. In: Stuss, D.T., Knight, R.T. (Eds.), Principles of Frontal Lobe Function. Oxford University Press, New York. Cohen, J.D., Perlstein, W.M., Braver, T.S., Nystrom, L.E., Noll, D.C., Jonides, J., Smith, E.E., 1997. Temporal dynamics of brain activity during a working memory task. Nature 386, 604–608. Corbetta, M., Shulman, G.L., 2002. Control of goal-directed and stimulus-driven attention in the brain. Nat. Rev. Neurosci. 3, 201–215. Coull, J.T., Frith, C.D., Frackowiak, R.S., Grasby, P.M., 1996. A fronto-parietal network for rapid visual information processing: a PET study of sustained attention and working memory. Neuropsychologia 34, 1085–1095. Coull, J.T., Frackowiak, R.S., Frith, C.D., 1998. Monitoring for target objects: activation of right frontal and parietal cortices with increasing time on task. Neuropsychologia 36, 1325–1334. Courtney, S.M., Petit, L., Maisog, J.M., Ungerleider, L.G., Haxby, J.V., 1998. An area specialized for spatial working memory in human frontal cortex. Science 279, 1347–1351.

C

E

R

R

O

C

784 785

N

782 783

U

780 781

F

791 792

The project was supported by the Swedish Research Council grant 2007-1895 and 2006-1290 to J.P. We also would like to thank Dr Gregoria Kalpouzos for helpful suggestions and comments, and Rachel Elands and William Fredborg for assistance during the data collection.

778 779

O

789 790

777

R O

Acknowledgments

775 776

P

788

773 774

D'Ardenne, K., Eshel, N., Luka, J., Lenartowicz, A., Nystrom, L.E., Cohen, J.D., 2012. Role of prefrontal cortex and the midbrain dopamine system in working memory updating. Proc. Natl. Acad. Sci. U. S. A. 109, 19900–19909. D'Esposito, M., Ballard, D., Aguirre, G.K., Zarahn, E., 1998. Human prefrontal cortex is not specific for working memory: a functional MRI study. NeuroImage 8, 274–282. Dudukovic, N.M., Wagner, A.D., 2007. Goal-dependent modulation of declarative memory: neural correlates of temporal recency decisions and novelty detection. Neuropsychologia 45, 2608–2620. Duncan, J., Owen, A.M., 2000. Common regions of the human frontal lobe recruited by diverse cognitive demands. Trends Neurosci. 23, 475–483. Duncan, K., Curtis, C., Davachi, L., 2009. Distinct memory signatures in the hippocampus: intentional states distinguish match and mismatch enhancement signals. J. Neurosci. 29, 131–139. Einstein, G.O., McDaniel, M.A., 1996. Retrieval processes in prospective memory: theoretical approaches and some new empirical findings. In: Brandimonte, M., Einstein, G.O., McDaniel, M.A. (Eds.), Prospective Memory: Theory and Applications. Lawrence Erlbaum Associates, Inc., Hillsdale, NJ, pp. 115–142. Einstein, G.O., Holland, L.J., McDaniel, M.A., Guynn, M.J., 1992. Age-related deficits in prospective memory: the influence of task complexity. Psychol. Aging 7, 471–478. Fletcher, P.C., Henson, R., 2001. Frontal lobes and human memory: insights from functional neuroimaging. Brain 124, 849–881. Gonneaud, J., Kalpouzos, G., Bon, L., Viader, F., Eustache, F., Desgranges, B., 2011. Distinct and shared cognitive functions mediate event- and time-based prospective memory impairment in normal ageing. Memory 19, 360–377. Gonneaud, J., Rauchs, G., Groussard, M., Landeau, B., Mézenge, F., de La Sayette, V., Eustache, F., Desgranges, B., 2014. How do we process event-based and time-based intentions in the brain? An fMRI study of prospective memory in healthy individuals. Hum. Brain Mapp. 35, 3066–3082. Gordon, B.A., Shelton, J.T., Bugg, J.M., McDaniel, M.A., Head, D., 2011. Structural correlates of prospective memory. Neuropsychologia 49, 3795–3800. Grady, C.L., 2008. Cognitive neuroscience of aging. Ann. N. Y. Acad. Sci. 1124, 127–144. Hashimoto, T., Umeda, S., Kojima, S., 2011. Neural substrates of implicit cueing effect on prospective memory. NeuroImage 54, 645–652. Haynes, J.D., Sakai, K., Rees, G., Gilbert, S., Frith, C., Passingham, R.E., 2007. Reading hidden intentions in the human brain. Curr. Biol. 17, 323–328. Jaeggi, S.M., Schmid, C., Buschkuehl, M., Perrig, W.J., 2009. Neuropsychology. Neuropsychol. Dev. Cogn. B Aging Neuropsychol. Cogn. 16, 80–102. Jimura, K., Braver, T.S., 2010. Age-related shifts in brain activity dynamics during task switching. Cereb. Cortex 20, 1420–1431. Jonides, J., Smith, E.E., Marschuetz, C., Koeppe, R.A., Reuter-Lorenz, P., 1998. Inhibition in verbal working memory revealed by brain activation. Proc. Natl. Acad. Sci. U. S. A. 95, 8410–8413. Kalpouzos, G., Eriksson, J., Sjölie, D., Molin, J., L N., 2010. Neurocognitive systems related to real-world prospective memory. PLoS One 5, e13304. Kessels, R.P., Meulenbroek, O., Fernández, G., Olde Rikkert, M.G., 2011. Spatial working memory in aging and mild cognitive impairment: effects of task load and contextual cueing. Neuropsychol. Dev. Cogn. B Aging Neuropsychol. Cogn. 17, 556–574. Kim, J., Whyte, J., Wang, J., Rao, H., Tang, K.Z., Detre, J.A., 2006. Continuous ASL perfusion fMRI investigation of higher cognition: quantification of tonic CBF changes during sustained attention and working memory tasks. NeuroImage 31, 376–385. Kliegel, M., Jäger, T., 2010. Delayed-execute prospective memory performance: the effects of age and working memory. Dev. Neuropsychol. 30, 819–843. Kliegel, M., Jäger, T., Phillips, L.H., 2008. Adult age differences in event-based prospective memory: a meta-analysis on the role of focal versus nonfocal cues. Psychol. Aging 23, 203–208. Lawrence, N.S., Ross, T.J., Hoffmann, R., Garavan, H., Stein, E.A., 2003. Multiple neuronal networks mediate sustained attention. J. Cogn. Neurosci. 15, 1028–1038. Mäntylä, T., 2003. Assessing absentmindedness: prospective memory complaint and impairment in middle-aged adults. Mem. Cogn. 31, 15–25. Marklund, P., Fransson, P., Cabeza, R., Larsson, A., Ingvar, M., Nyberg, L., 2007a. Unity and diversity of tonic and phasic executive control components in episodic and working memory. NeuroImage 36, 1361–1373. Marklund, P., Fransson, P., Cabeza, R., Petersson, K.M., Ingvar, M., Nyberg, L., 2007b. Sustained and transient neural modulations in prefrontal cortex related to declarative long-term memory, working memory, and attention. Cortex 43, 22–37. Martin, T., McDaniel, M.A., Guynn, M.J., Houck, J.M., Woodruff, C.C., Bish, J.P., Moses, S.N., Kicić, D., Tesche, C.D., 2007. Brain regions and their dynamics in prospective memory retrieval: a MEG study. Int. J. Psychophysiol. 64, 247–258. McDaniel, M.A., Einstein, G.O., 2007. Prospective Memory: An Overview and Synthesis of An Emerging Field. Sage, Thousand Oaks (CA). McDaniel, M.A., Glisky, E.L., Rubin, S.R., Guynn, M.J., Routhieaux, B.C., 1999. Prospective memory: a neuropsychological study. Neuropsychology 13, 103–110. McDaniel, M.A., Lamontagne, P., Beck, S.M., Scullin, M.K., Braver, T.S., 2013. Dissociable neural routes to successful prospective memory. Psychol. Sci. 24, 1791–1800. McFarland, C.P., Glisky, E.L., 2009. Frontal lobe involvement in a task of time-based prospective memory. Neuopsychologia 47, 1660–1669. McFarland, C.P., Glisky, E.L., 2011. Implementation intentions and prospective memory among older adults: an investigation of the role of frontal lobe function. Neuropsychol. Dev. Cogn. B Aging Neuropsychol. Cogn. 18, 633–652. Nagel, I.E., Preuschhof, C., Li, S.C., Nyberg, L., Bäckman, L., Lindenberger, U., Heekeren, H.R., 2009. Performance level modulates adult age differences in brain activation during spatial working memory. Proc. Natl. Acad. Sci. U. S. A. 106, 22552–22557. Nagel, I.E., Preuschhof, C., Li, S.C., Nyberg, L., Bäckman, L., Lindenberger, U., Heekeren, H.R., 2011. Load modulation of BOLD response and connectivity predicts working memory performance in younger and older adults. 23 8 pp. 2030–2045.

T

786 787

monitoring of the environment or features not specific for the ongoing task. While the current study was not designed to tease apart the neural contribution associated with each of these task types, we recognize that an important avenue for future research could be to investigate agedifferences in sustained/transient brain activation related to focal vs. non-focal task conditions. To conclude, the current results provide evidence that PM is relying on multiple dissociable transient and sustained cognitive processes which are supported by a distinct set of brain regions. Our findings further demonstrate that older individuals perform worse on PM tasks compared to tasks with low cognitive demands. Older adults showed reduced PM related transient activation in a network of regions that included the PFC, striatum, and MTL regions, along with reduced sustained activation in the right IFG, compared to young adults. Thus, the observed age-related behavioral impairment in PM could be explained by an inability for older adults to recruit task-relevant brain networks.

D

771 772

N. Peira et al. / NeuroImage xxx (2015) xxx–xxx

E

10

Please cite this article as: Peira, N., et al., Age differences in brain systems supporting transient and sustained processes involved in prospective memory and working memory, NeuroImage (2015), http://dx.doi.org/10.1016/j.neuroimage.2015.10.075

840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925

N. Peira et al. / NeuroImage xxx (2015) xxx–xxx

E

D

P

R O

O

F

Rose, N.S., Rendell, P.G., McDaniel, M.A., Aberle, I., Kliegel, M., 2010. Age and individual differences in prospective memory during a “Virtual Week”: the roles of working memory, vigilance, task regularity, and cue focality. Psychol. Aging 25, 595–605. Rugg, M.D., Vilberg, K.L., 2013. Brain networks underlying episodic memory retrieval. Curr. Opin. Neurobiol. 23, 255–260. Schnitzspahn, K.M., Stahl, C., Zeintl, M., Kaller, C.P., Kliegel, M., 2013. The role of shifting, updating, and inhibition in prospective memory performance in young and older adults. Dev. Psychol. 49, 1544–1553. Simons, J.S., Schölvinck, M.L., Gilbert, S.J., Frith, C.D., Burgess, P.W., 2006. Differential components of prospective memory? Evidence from fMRI. Neuropsychologia 44, 1388–1397. Smith, R.E., Bayen, U.J., 2006. The source of adult age differences in event-based prospective memory: a multinomial modeling approach Journal of Experimental Psychology: Learning. Mem. Cogn. 32, 623–635. Sperling, R.A., 2007. Functional MRI studies of associative encoding in normal aging, mild cognitive impairment, and Alzheimer's disease. Ann. N. Y. Acad. Sci. 1097, 146–155. Takahashi, E., Ohki, K., Kim, D.S., 2013. Dissociation and convergence of the dorsal and ventral visual working memory streams in the human prefrontal cortex. NeuroImage 65, 488–498. Unsworth, N., Heitz, R.P., Schrock, J.C., Engle, R.W., 2005. An automated version of the operation span task. Behav. Res. Methods 37, 498–505. Vangkilde, S., Bundesen, C., T CJ., 2011. Prompt but inefficient: nicotine differentially modulates discrete components of attention. Psychopharmacology 218, 667–680. West, R., 2005. The neural basis of age-related decline in prospective memory. In: Cabeza, R., Nyberg, L., Park, D.C. (Eds.), Cognitive Neuroscience of Aging. Oxford University Press, New York, pp. 246–264. West, R., 2011. The temporal dynamics of prospective memory: a review of the ERP and prospective memory literature. Neuropsychologia 49, 2233–2245. West, R., Bowry, R., 2005. Effects of aging and working memory demands on prospective memory. Psychophysiology 42, 698–712. Zeintl, M., Kliegel, M., Rast, P., Zimprich, D., 2006. Prospective memory complaints can be predicted by prospective memory performance in older adults. Dement. Geriatr. Cogn. Disord. 22, 209–215. Zimmermann, T.D., Meier, B., 2006. The rise and decline of prospective memory performance across the lifespan. Q. J. Exp. Psychol. 59, 2040–2046. Zöllig, J., West, R., Martin, M., Altgassen, M., Lemke, U., Kliegel, M., 2007. Neural correlates of prospective memory across the lifespan. Neuropsychologia 45, 3299–3314.

N C O

R

R

E

C

T

Nyberg, L., Marklund, P., Persson, J., Cabeza, R., Forkstam, C., Petersson, K.M., Ingvar, M., 2003. Common prefrontal activations during working memory, episodic memory, and semantic memory. Neuropsychologia 41, 371–377. Nyberg, L., Salami, A., Andersson, M., Eriksson, J., Kalpouzos, G., Kauppi, K., Lind, J., Pudas, S., Persson, J., Nilsson, L.G., 2010. Longitudinal evidence for diminished frontal cortex function in aging. Proc. Natl. Acad. Sci. U. S. A. 107, 22682–22686. Okuda, J., Fujii, T., Yamadori, A., Kawashima, R., Tsukiura, T., Fukatsu, R., Suzuki, K., Ito, M., Fukuda, H., 1998. Participation of the prefrontal cortices in prospective memory: evidence from a PET study in humans. Neurosci. Lett. 253, 127–130. Okuda, J., Fujii, T., Ohtake, H., Tsukiura, T., Yamadori, A., Frith, C.D., Burgess, P.W., 2015. Differential involvement of regions of rostral prefrontal cortex (Brodmann area 10) in time- and event-based prospective memory. International Journal of Psychophysiology (in press). Owen, A.M., McMillan, K.M., Laird, A.R., Bullmore, E., 2005. N-back working memory paradigm: a meta-analysis of normative functional neuroimaging studies. Hum. Brain Mapp. 25, 46–59. Paxton, J.L., Barch, D.M., Racine, C.A., Braver, T.S., 2008. Cognitive control, goal maintenance, and prefrontal function in healthy aging. Cereb. Cortex 18, 1010–1028. Persson, J., Nyberg, L., 2006. Altered brain activity in healthy seniors: what does it mean? Prog. Brain Res. 157, 45–56. Persson, J., Pudas, S., Lind, J., Kauppi, K., Nilsson, L., Nyberg, L., 2012. Longitudinal structure–function correlates in elderly reveal MTL dysfunction with cognitive decline. Cereb. Cortex 22, 2297–2304. Pessoa, L., Gutierrez, E., Bandettini, P.A., Ungerleider, L.G., 2002. Neural correlates of visual working memory: fMRI amplitude predicts task performance. Neuron 35, 975–987. Poppenk, J., Moscovitch, M., McIntosh, A.R., Ozcelik, E., Craik, F.I., 2010. Encoding the future: successful processing of intentions engages predictive brain networks. NeuroImage 49, 905–913. Rajah, M.N., D'Esposito, M., 2005. Region-specific changes in prefrontal function with age: a review of PET and fMRI studies on working and episodic memory. Brain 128, 1964–1983. Raz, N., Rodrigue, K.M., Head, D., Kennedy, K.M., Acker, J.D., 2004. Differential aging of the medial temporal lobe: a study of a five-year change. Neurology 62, 433–438. Raz, N., Lindenberger, U., Rodrigue, K.M., Kennedy, K.M., Head, D., Williamson, A., Dahle, C., Gerstorf, D., Acker, J.D., 2015. Regional brain changes in aging healthy adults: general trends, individual differences and modifiers. Cereb. Cortex (in press). Reynolds, J.R., West, R., Braver, T., 2009. Distinct neural circuits support transient and sustained processes in prospective memory and working memory. Cereb. Cortex 19, 1208–1221.

U

926 927 928 929 930 931 932 933 934 935 Q10 936 937 938 Q11 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 Q12 960 961 Q13 962 963 964

11

Please cite this article as: Peira, N., et al., Age differences in brain systems supporting transient and sustained processes involved in prospective memory and working memory, NeuroImage (2015), http://dx.doi.org/10.1016/j.neuroimage.2015.10.075

965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999