Full Terms & Conditions of access and use can be found athttp://www.tandfonline.com/action/journalInformation?journalCode=pmem20

Memory

ISSN: 0965-8211 (Print) 1464-0686 (Online) Journal homepage: http://www.tandfonline.com/loi/pmem20

Acute pain disrupts prospective memory cuedetection processes

Margarida Pitães, Chris Blais, Paul Karoly, Morris A. Okun & Gene A. Brewer

To cite this article: Margarida Pitães, Chris Blais, Paul Karoly, Morris A. Okun & Gene A.Brewer (2018): Acute pain disrupts prospective memory cue detection processes, Memory, DOI:10.1080/09658211.2018.1491602

To link to this article: https://doi.org/10.1080/09658211.2018.1491602

Published online: 01 Jul 2018.

Submit your article to this journal

Article views: 6

View Crossmark data

Acute pain disrupts prospective memory cue detection processesMargarida Pitães, Chris Blais, Paul Karoly, Morris A. Okun and Gene A. Brewer

Department of Psychology, Arizona State University, Tempe, USA

ABSTRACTProspective memory refers to the planning, retention, retrieval, and execution of intentions forfuture behaviours and it is integral to the enterprise of daily living. Although prospectivememory relies upon retrospective memory and executive processes often disrupted by pain,limited research has explored the influence of acute or chronic pain on the ability tocomplete prospective memory tasks. In the present study we investigated the influence ofacute pain on prospective memory tasks that varied in their demands on executive processes(i.e., non-focal versus focal prospective memory cues). Complex-span working memory taskswere also administered to examine whether individual differences in working memorycapacity moderated any negative impact of pain on prospective memory. Acute painsignificantly impaired prospective memory performance in conditions that encouraged non-focal strategic processing of prospective memory cues, but not in conditions that encouragedmore spontaneous focal processing. Individual differences in working memory capacity didnot moderate the effect of acute pain on non-focal prospective memory. These findingsprovide new insights into prospective memory dysfunction created by painful experiences.

ARTICLE HISTORYReceived 24 April 2017Accepted 7 June 2018

KEYWORDSProspective memory; pain;working-memory; memoryfor future intentions

Cognitive performance is known to be hindered by acuteand chronic pain with memory and attention systemsamong those most commonly affected (Katz, Heard, Mills,& Leavitt, 2004; Muñoz & Esteve, 2005). While moststudies have focused on retrospective memory impair-ments (i.e., memory for past events and information; Berry-man et al., 2013), limited research has examined theinfluence of pain on prospective memory (i.e., memoryfor future intentions; Gatzounis, Schrooten, Crombez, &Vlaeyen, 2018; Ling, Campbell, Heffernan, & Greenough,2007; Miller, Basso, Candilis, Combs, & Woods, 2014). Pro-spective memory (PM) problems are the most frequentmemory failures in everyday life (e.g., forgetting to paybills on time, take prescribed medications, etc.; Kliegel &Martin, 2003; Unsworth, Brewer, & Spillers, 2012). Becauseof its importance and prevalence, remembering toperform delayed intentions at the appropriate futuremoment is essential for independent and successfulliving (Dismukes, 2008; Park & Kidder, 1996). While rela-tively little work has examined the impact of pain on PM,prior research has shown that many related cognitivephenomena are negatively impacted by pain, includingworking memory (Berryman et al., 2013), autobiographicalmemory (Erskine, Morley, & Pearce, 1990), memory forspecific painful experiences (Noel, Chambers, McGrath,Klein, & Stewart, 2012; Redelmeier & Kahneman, 1996),and task switching (Attridge, Keogh, & Eccleston, 2016).

The negative effect that pain has on various cognitiveprocesses holds important clinical implications.

Theoretically, the diverse effects of chronic and acutepain on cognition demand a biopsychosocial model thatincorporates differing perspectives on how pain leads topsychopathology (Karoly & Crombez, 2018). For example,the fear-avoidance model of chronic pain suggests thatpeople experiencing acute painful experiences may perse-verate on these painful experiences, thus leading to cata-strophizing, negative affect, and depression (Vlaeyen &Linton, 2012). Perseverative effects of acute pain negativelyimpacts several cognitive processes including workingmemory, retrospective memory, and potentially PM. Thisimpact may lead to chronic negative evaluations of thesource of pain, the experience of pain, and how pain willinfluence behaviour in future situations. This fear-avoid-ance loop traps some people into a vicious cycle of disabil-ity and suffering. Here, we hypothesise that individualswith higher working memory may be less susceptible tothis type of interference from pain as compared to thosewith low working memory. This prediction emerges fromresearch suggesting that pain hinders attention controland interrupts task-related cognitive functioning (Eccles-ton & Crombez, 1999; Moore, Eccleston, & Keogh, 2017).

Relatedly, neuropsychological evaluations for painpatients typically assess working memory as well as long-term retention of information to determine a patient’smental capacity because these cognitive processes areimpacted by acute and chronic pain (Mazza, Frot, & Rey,In Press). These cognitive factors are routinely implicatedin successful prospective memory where people must

© 2018 Informa UK Limited, trading as Taylor & Francis Group

CONTACT Gene A. Brewer gene.brewer@asu.edu Department of Psychology, Arizona State University, 950 S McAllister Ave, Tempe, AZ 85287-1104, USAThe data that support the findings of this study are openly available in our laboratory website at asumaclab.com.

MEMORYhttps://doi.org/10.1080/09658211.2018.1491602

encode, maintain, and retrieve intentions for future action(i.e., long-term memory functions). Moreover, recentresearch has shown that working memory functions arealso necessary for successful PM under certain conditions(Brewer, Knight, Marsh, & Unsworth, 2010; Meeks, Pitães,& Brewer, 2015; Smith & Bayen, 2005). Given that painnegatively impacts retrospective and working memory,we reasoned that pain would also negatively impact PM,especially under demanding circumstances. Furthermore,we predicted that individuals with higher workingmemory capacity would be more resilient to negativeeffects of pain on PM.

Successful PM relies on a range of frontally-mediatedexecutive processes, such as planning, goal maintenance,and interrupting/switching from irrelevant ongoing activi-ties to execute intentions (Kliegel, Altgassen, Hering, &Rose, 2011). All of these executive functions are relatedto individual differences in working memory capacity andare known to be disrupted by pain (Attridge et al., 2016;Berryman et al., 2014; Moore, Keogh, & Eccleston, 2012;Moriarty, McGuire, & Finn, 2011; Okun, Karoly, Mun, &Kim, 2016). Given the prevalence of painful experiences(Breivik, Collett, Ventafridda, Cohen, & Gallacher, 2006)and the adverse effects of PM failures on daily living, it isimportant to understand the conditions under whichpain affects PM performance. Moreover, it is critical tofind possible moderators of pain’s impact on behaviour(e.g., working memory capacity).

Differences in the working memory and attentiondemands inherent in detecting PM cues can be conceptu-alised in the multiprocess view of PM (McDaniel & Einstein,2000). In PM, information in the environment can cue theretrieval of an intended action (e.g., encountering a pillbottle cues retrieval of your intention to take medication).To mimic the demands of daily life, typical laboratory para-digms instruct participants to remember to perform aspecific action upon the occurrence of a cue that occurswhile they are busily engaged in an unrelated ongoingactivity (e.g., press the “/” key whenever you encounterthe word “doctor” while performing a lexical decisiontask; cf., Einstein & McDaniel, 2005). The ongoing activitydoes not change when the cue occurs, so for intentionretrieval to happen individuals must interpret the eventas a cue for action. According to the multi-process theoryof PM, the detection of cues and retrieval of deferred inten-tions varies with the extent to which capacity-demanding(versus more spontaneous) cognitive processes are necess-ary for performance. For example, spontaneous PM cuedetection is more likely to occur in cases where theongoing activity fully encourages processing of relevantfeatures of the PM cue (i.e., focal PM). Alternatively, if theongoing activity draws attention away from the PM cue(i.e., non-focal PM), then prospective remembering isassumed to require more strategic control processes tomonitor for the cue signalling the appropriateness for per-forming the intended action (Brewer et al., 2010; McDaniel& Einstein, 2000). For example, during a lexical-decision

task, which requires semantic processing of letter stringsto determine whether they constitute words, rememberingto press the “/” key when the word “tortoise” is presentedinvolves focal processing of the PM cue. However, thesame task is classified as non-focal when the cue is theappearance of the syllable “tor” because the informationextracted in the service of the ongoing task primarilyfosters focal attention to stimuli at the word level, butnot at a syllable level (Einstein et al., 2005). Therefore,while focal cues are assumed to spontaneously triggerintention retrieval, non-focal cues are assumed instead torequire a more controlled approach to prospectiveremembering.

The notion that PM demands (such as the degree ofcue-focality) influence the working memory demands forPM cue detection has promoted a more refined examin-ations of PM abilities in many cognitively impaired clinicalpopulations (Foster, McDaniel, Repovs, & Hershey, 2009;Kliegel et al., 2011; Kliegel, Jäger, & Phillips, 2008; Marshet al., 2009; Ordemann, Opper, & Davalos, 2014; Rendell,McDaniel, Forbes, & Einstein, 2007; Zuber, Kliegel, & Ihle,2016). Focal PM cues are typically found to enhance PMperformance among older and clinical populations whosePM performance is nearly perfect under such conditions.However, more PM failures occur in non-focal PM con-ditions. Similarly, healthy younger adults under divided-attention conditions exhibit less PM disruption for focalthan for non-focal PM tasks (Windy McNerney & West,2007), suggesting that successful non-focal PM perform-ance depends on some optimal level of executive function-ing (for similar findings on individual differences inworking-memory capacity see Brewer et al., 2010; Rose,Rendell, McDaniel, Aberle, & Kliegel, 2010). As previouslynoted, only two studies have explored the effects ofchronic pain on PM performance and one study has exam-ined the effects of acute pain.

Relatively few studies on pain and PM have been pub-lished. Ling et al. (2007) reported that patients with lowback pain evidenced demonstrable impairments in PMusing a questionnaire measure. Miller et al. (2014) used aperformance measure of PM (the Memory for IntentionsScreening Test; MIST) in patients with multiple sclerosisand found that pain accounted for approximately 10% ofthe variance in PM performance. In contrast, a recentstudy by Gatzounis et al. (2018) induced electro-cutaneouspain as well as threatening instructions and observed theireffects on a PM task that involved subjects performing apreviously-encoded intention whenever a specific cueword (i.e., a focal cue) was detected while completing aPM-irrelevant word categorisation task. Pain stimulationoccurred on 14.6% of the trials and PM cues were pre-sented on 2.65% of the trials. The investigators alsoassessed working memory (via reading span) and catastro-phizing tendencies as possible moderators. They reportedthat neither threat nor pain stimulation influenced PM, nordid working memory or catastrophizing play a role in PMintention performance. A primary aim of the current

2 M. PITÃES ET AL.

study is to address some of these inconsistencies by com-paring the focal PM task used in Gatzoiunis to a nonfocalPM task that is more demanding and commonly used inthe PM literature with healthy individuals.

Given the specific nature of the cognitive impairmentscaused by pain, and based on the above framework, thepresent study investigated the influence of experimen-tally-induced acute pain on PM tasks varying in thedemands of executive control required for prospectiveremembering (i.e., cue-focality). We hypothesised thatacute pain would impair non-focal, but not focal, PM per-formance. Further, if the expected pain-related PMdecline for non-focal tasks is indeed a consequence ofshared executive resources, the effect of acute pain onnon-focal PM performance should decrease as working-memory (WM) capacity increases (Rose et al., 2010). Thisaccount predicts that individual differences in workingmemory capacity should moderate the effects of pain onnon-focal PM. Critically, however, this account also predictsthat individual differences in WM capacity should not beassociated with focal PM performance. A final goal of thecurrent work was to examine how pain influences individ-uals’ metacognitive judgements about PM performance.This goal was accomplished by asking participants tomake postdictions regarding the impact of pain on theirPM performance at the end of the experiment. Specifically,we were interested in the degree to which participantswere metacognitively aware of any negative impacts ofpain on their own PM performance.

Methods

This study was approved by the Human Research Insti-tutional Review Board at Arizona State University.

Participants and design

Over the course of two semesters, a total of 239 youngadults (104 females) from Arizona State University partici-pated in this study (Mage = 22 years, SD = 2.25). We col-lected the data with no a priori power calculation nor anystopping rule based on significance of the findings. Infact, the study ended with a sample size deemed appropri-ate based on our experience conducting many individualdifferences studies of this nature at the end of thesecond semester of data collection. After providingconsent, each participant was tested individually in ses-sions that lasted approximately 75 min and all participantscompleted a battery of pain-related questionnaires (includ-ing the Mindful Attention Awareness Scale (MAAS); Positiveand Negative Affect Schedule (PANAS); Profile of ChronicPain and the Perception of Pain Scale); a set of three auto-mated complex-span working-memory tasks (Operation-span; Symmetry-span and Reading-span; see Oswald,McAbee, Redick, & Hambrick, 2015; Unsworth, Heitz,Schrock, & Engle, 2005 for full task details) and a PM task.All participants reported no existing chronic pain

experience. The information collected from the pain-related questionnaires was not relevant to the currentstudy and will not be further discussed.

We implemented a 2 (PM condition: focal versus nonfo-cal PM) × 2 (Pain condition: pain versus no-pain) mixed-fac-torial design to assess whether cue-focality interacted withpain to produce selective PM deficits. The PM task involvedtwo between-subjects conditions ( focal and nonfocal). Ofthe 239 participants, 101 participants (62 females) wererandomly assigned to the focal condition and 138 (42females) to the non-focal condition. Acute pain wasinduced by having participants place their non-dominantindex finger into a pressure algometer while performingthe PM task using their dominant hand (Figure 1). Thetwo levels of Pain (Pain and No-pain) were manipulatedwithin-subjects and counterbalanced such that all partici-pants completed the PM task both in pain and not inpain. Importantly, all of the results reported in thecurrent study are unaltered when controlling for the coun-terbalancing factor of putting the participants in painduring the first versus the second block. The lack of a coun-terbalancing effect indicates that the acute painful experi-ence did not have any detectable and lasting effects on PMwhen participants transitioned to no pain blocks.

Materials

Pain manipulation and assessment

Pain intensity was rated on a numerical rating scale where1 stands for “No pain/physical discomfort” and 10 for “Themost intense pain/physical discomfort ever”. Each participantrated their level of pain at three points during the study: (1)upon arrival on the day of testing (Baseline rating); (2) afterperforming all WM tasks (Pain rating); and (3) before per-forming the PM task in the No-pain condition (No-painrating). In order to ensure sufficiently intense nociceptiveinput, the weight for the Pain condition (Pain rating) wasdetermined by having the experimenter add weight tothe algometer until the participant reported between 6–7on the pain scale (note that participants were blind tothis pain threshold determination). For both Pain and No-

Figure 1. Acute pain was induced using a pressure algometer device. Thecup filled with weight was placed on the end opposite the hinge.

MEMORY 3

pain conditions, the subject’s index finger remained in thealgometer device during the entire PM task. However, inthe No-pain condition the experimenter removed allweight from the algometer.

Complex-span working memory tasks

Operation spanParticipants solved a series of math operations while tryingto remember a set of unrelated letters. Immediately afterthe letter was presented the next operation was presented.When recall was required, letters from the current set couldbe recalled in the correct order by clicking on the appropri-ate letters. One trial of each set-size (3–7) was presented,with the order of set-size varying randomly. Consistentwith prior research, the dependent variable for all of thecomplex-span measures was the proportion of correctitems recalled in the correct serial position.

Symmetry spanParticipants were required to recall sequences of redsquares within a matrix while performing a symmetry-judg-ment task. In the symmetry-judgment task, participantswere shown an 8 × 8 matrix with some black squares. Par-ticipants decided whether the design was symmetricalabout its vertical axis. The pattern was symmetrical halfof the time. After each symmetry judgment, participantswere presented with a 4 × 4 matrix with one of the cellsfilled in red. At recall, participants were required to recallthe sequence of red-square locations in the preceding dis-plays in the order they appeared by clicking on the cells ofan empty matrix. Set-size ranged from 2–5, with oneadministration for each set-size.

Reading spanParticipants read a set of sentences while trying to remem-ber a set of unrelated letters. For each sentence, partici-pants determined whether the sentence made sense ornot (half of the sentences made sense). After each sentencedecision, participants were presented with a letter for recallat the end of the set. At recall, letters from the current set-size ranged from 3–7, with one administration for each set-size.

Prospective memory task

The parameters of this task and the word and nonwordstimuli follow a well-validated PM paradigm known tomanipulate executive demands within a PM task whileholding all other task demands equal (Brewer et al., 2010;see Figure 2). Participants first received instructions forthe lexical-decision task. They were told that their taskwas to decide, as quickly and accurately as possible,whether a string of letters formed valid English words ornot by pressing one of two keys (“J” and “F” keys, respect-ively). Word and nonwords were presented one at the timeand stayed on the screen until participants responded.

Participants were also told to press the spacebar (withtheir thumb) when a “waiting” message appearedbetween each trial, which triggered presentation of thenext lexical-decision trial. After a 20-trial practice block, par-ticipants encoded the PM intention. For both focal andnonfocal PM conditions, the participants were told thatwe were also interested in their ability to remember toperform a specific future action. Therefore, in addition toperforming the lexical-decision task, participants wereinstructed to make a special key press (“/” key) duringthe “waiting” message whenever they encountered a PMcue. In the focal condition the cue was the specific word(“DOCTOR”) which appeared 8 times in the context of theongoing task, and for the nonfocal condition the cue wasthe syllable (“TOR”) which also occurred 8 times in thecontext of the ongoing task but in different words. Thetwo PM conditions were exact replications of oneanother, and only the nature of the PM cues differedbetween them.

For both conditions 8 PM cues occurred once every 25trials beginning on trial 25 and ending on trial 200. EachPM block consisted of a total of 208 lexical-decision trials,and participants performed two blocks (one block in Painand one in No-pain). PM performance was defined as theproportion of cues successfully responded to with a “/”response during the “waiting” message immediately fol-lowing the cue. After the experimenter was satisfied thatparticipants fully understood all the instructions, a 10-mindistractor task was administered (i.e., a visual puzzle-task).In PM paradigms, instituting delays (filler distractor tasks)between intention encoding and execution is cruciallyimportant to remove the intention from primary memoryand prevent the prospective task from becoming a vigi-lance task (Einstein & McDaniel, 1990). Therefore, followingthis delay, the experimenter asked the participant to placetheir finger into the algometer and initiated the first blockof the PM task without making any reference to the PMtask. Depending on the counterbalancing order, partici-pants performed the first PM block either in Pain or No-pain. After completing the first PM block, and before par-ticipants initiated the second block, all participants ratedtheir pain level with all weight removed from the alg-ometer (No-pain rating). This last pain rating is importantto insure that participants that performed the first PMblock in Pain initiated the second PM block reporting notbeing in pain. The experimenter then either added theweights (or kept all weight removed) from the algometer,and participants initiated the second PM block in thePain or No-pain condition.

At the end of the experiment, all participants completeda PM post-experimental questionnaire that asked them torecall the intended action and PM cue (Scullin, Bugg,McDaniel, & Einstein, 2011). The use of this type of ques-tionnaire is suggested in the PM literature to confirm par-ticipant’s understanding of the PM task and todisentangle retrospective memory failures (i.e., failuresremembering the intention content, or what action to

4 M. PITÃES ET AL.

perform and when to perform it) and actual PM failures (i.e.,failures remembering the intention itself, or that one needsto do something at a certain point in the future; Ellis & Kva-vilashvili’s, 2000; Scullin et al., 2011). Participants were alsoasked to make metacognitive judgments about whetherthey believed pain to affect their PM ability by rating theperceived difficulty of remembering to perform the PMintention while they were in pain versus no-pain. Ratingswere made in an 8-point Likert scale ranging from 0 “Notdifficult at all” to 8 “Extremely difficult”. These judgementsmay provide insight into compensatory strategies devel-oped by participants to mitigate perceived negativeimpacts of acute pain.

Results

For all statistical analyses, a conventional alpha level of .05was used and counterbalancing had no effect on any of thedependent variables so that factor was excluded (ps > .05).

Descriptive statistics

Average performance for all WM tasks can be found inTable 1. WM task performance was similar to previouslyreported research from our laboratory and from other lab-oratories (Oswald et al., 2015; Redick et al., 2012). Also, esti-mates of skew and kurtosis were at reasonable levels (Kline,2015).

Table 2 shows descriptive statistics for the three subjec-tive pain ratings. As the reader can note from the baselineratings, all participants reported that they were not in painupon arrival on the day of testing. Pain and No-pain ratingswere significantly different, confirming that our painmanipulation was a valid means to induce genuinephysical pain, F(1, 237) = 1478.580, p < .001, h2

p = .862.Important for the present purposes, neither Pain orNo-pain ratings differ between the focal and nonfocal PMconditions (ps > .05).

Lexical decision task accuracy and responselatencies

Before conducting the analysis on PM performance, weexamined mean lexical decision latency scores (averagedacross trials that did not contain PM cues) as a functionof pain (between subject variable) and PM condition (cuefocality manipulated within subjects). In line with previousresearch, results of this analysis showed that responselatencies were slower in the non-focal (M = 959, SE = 14)than focal (M = 829 ms, SE = 16) possibly reflectingincreased monitoring for PM cues in the non-focal con-dition, F(1, 237) = 39.954, p < .001, h2

p = .144. There wasalso a significant main effect of pain condition, reflectingfaster lexical decision latencies when individuals wereplaced in pain (M = 874, SE = 11) compared to when theywere not in pain (M = 914, SE = 12), F(1, 237) = 19.710, p< .001, h2

p = .077. Finally, there was a significant PM con-dition × Pain interaction, F(1, 237) = 17.053, p < .001, h2

p

= .067. Planned comparisons confirmed that pain

Figure 2. Schematic illustration for nonfocal PM procedure (procedure for focal PM task was a direct replication). (1) Before initiating the first PM block, allparticipants were asked to place their finger in the algometer. Depending on counterbalancing order, the experimenter then added weight (or kept noweight) in the algometer and participants performed the first PM block in Pain or No-pain condition, respectively; (2) After completing the first PMblock, all participants rated their level of pain with all weights removed from the algometer (No-pain rating); (3) Participants who performed the firstPM block in pain (i.e., with weights in the algometer) performed the second PM block in the No-pain condition (and vice-versa).

Table 1. Descriptive statistics for subjective pain ratings.

Focal PM (N = 101) Nonfocal PM (N = 138)Condition M (SD) M (SD)

Baseline 1.52 (0.5) 1.56 (0.5)No-Pain 1.55 (0.9) 1.58 (0.7)Pain 6.66 (2.4) 6.9 (1.8)

Notes: Pain intensity was rated on a numerical rating scale where 1 standsfor “No pain/physical discomfort” and 10 for “Themost intense pain/physicaldiscomfort ever”.

Table 2. Descriptive statistics for all working memory tasks.

Measure Mean SD Skew Kurtosis

Ospan 16.90 5.00 −1.322 1.527Rspan 15.84 3.89 −.567 .037Symspan 10.01 2.75 −.538 −.215

MEMORY 5

reduced latencies in the non-focal PM condition (M = 921,SE = 14) versus (M = 998, SE = 15), t(137) = 5.760, p < .001,but not in the focal PM condition (M = 827, SE = 17versus (M = 830, SE = 18), t(100) = .269, p = .789.

Similarly, we report an analysis of lexical decision taskaccuracy (averaged across trials that did not contain PMcues) as a function of pain (between subject variable)and PM condition (manipulated within subjects). Therewas no difference in lexical decision accuracy betweenthe nonfocal (M = .92, SE = .01) and focal conditions (M= .91, SE = .01), F(1, 237) = 3.604, p = .059, h2

p = .015. Therewas a significant main effect of pain on lexical decisionaccuracy, reflecting worse ongoing task performancewhen individuals were placed in pain (M = .91, SE = .01)compared to when they were not in pain (M = .92, SE= .01), F(1, 237) = 17.469, p < .001, h2

p = .069. There wasnot a significant PM condition × Pain interaction, F(1,237) = 3.604, p = .059, h2

p = .015.

Prospective memory performance

The proportion of PM cues successfully responded to wassubmitted to a general linear model (GLM) with PM con-dition (focal, non-focal) as the between-subjects factorand pain condition (pain, no-pain) as the within-subjectsfactor. Replicating previous work (e.g., Brewer et al., 2010;Einstein et al., 2005) more PM cues were detected in thefocal (M = .95, SE = .02) than in the non-focal condition(M = .57, SE = .02), F(1, 237) = 157.378, p < .001, h2

p = .399.There was also a significant main effect of pain condition,reflecting worse PM performance when individuals wereplaced in pain (M = .72, SE = .02) compared to when theywere not in pain (M = .80, SE = .02), F(1, 237) = 25.807, p< .001, h2

p = .098. However, this main effect was qualifiedby a significant PM condition × Pain interaction, F(1, 237)= 14.173, p < .001, h2

p = .060. Follow up analysis showedthat the main effect of pain was significant in the non-focal condition with participants performing higher in theno pain (M = .63, SE = .02) than in the pain condition (M= .50, SE = .02), t(137) = 5.526, p < .001. In contrast, in thefocal condition, participants’ performance in the no paincondition (M = .96, SE = .02) did not differ from that in thepain condition (M = .95, SE = .03), t(100) = 1.596, p = .114.As hypothesised, pain selectively affected highly demand-ing PM tasks (Figure 3). It should be noted that perform-ance in the focal condition was at high levels, whichcould limit the opportunity to observe an effect of painin this condition. The observed pattern of results ishowever consistent with previous research using similarfocal cues to the ones used here (Brewer et al., 2010; Ein-stein et al., 2005). Also, we conducted pairwise compari-sons after removing participants from focal conditionswho detected all of the cues and the pattern of resultsdid not change, i.e., we still did not find an effect of painon focal PM performance, t(38) = 1.617, p = .114.

Results from the PM post-experimental questionnairerevealed that both focal and non-focal PM conditions

elicited nearly perfect post-test recall, and only two partici-pants from the focal condition failed to recall the PMintended action. Thus, it seems likely that the PM failuresdescribed above were due to problems in detecting thecues rather than retrieving the intended action when thecue occurred (i.e., PM problems), rather than to retrospec-tive memory failures for the task demands.

Correlations and moderation analysis

Correlations between all WM capacity measures (Ospan,Rspan and Symspan) ranged between .22 and .45, all ps< .05. Therefore, z-scores were created for each task andthese measures were averaged together to create asingle WM capacity score. Consistent with previousresearch (Brewer et al., 2010), WM capacity positively corre-lated, albeit weakly, with nonfocal PM performance, r = .18,p < .05, but not with focal PM, r = .03, p > .05. The WMcapacity score was therefore included as a covariate inthe GLM analysis using only data from the non-focal PMcondition and an interaction between WM capacity andpain on PM performance was assessed to test for moder-ation. WM capacity was predictive of non-focal PM, F(1,135) = 4.49, p < .05, h2

p = .03, replicating previous work(Brewer et al., 2010; Rose et al., 2010). Contrary to our pre-dictions, however, individual differences in WM capacitydid not moderate the decline in nonfocal PM performanceunder acute pain conditions, F(1, 135) = .001, p = .998, indi-cating that high and low WM participants were similarlyaffected by pain. Thus, pain seems to impact non-focalPM performance through some alternative mechanism.

Perceived difficulty

To evaluate metacognitive aspects of PM performance inthis experiment, the same 2 (PM condition: focal vs. non-focal) × 2 (Pain condition: pain vs. no pain) mixed-factorialANCOVA was conducted on subjective ratings of PM taskdifficulty with WMC as a covariate. Participants in thenon-focal condition perceived the PM task to be signifi-cantly more difficult than participants in the focal

Figure 3. Proportion of correct prospective memory (PM) responses as afunction of prospective memory condition and pain condition. Error barsrepresent standard errors.

6 M. PITÃES ET AL.

condition, F(1, 236) = 50.281, p < .001, h2p = .176. There was

also a significant main effect of pain condition, F(1, 236) =198.652, p < .001, h2

p = .457, that was qualified by a signifi-cant interaction with PM condition, F(1, 236) = 43.901, p< .001, h2

p = .157. This interaction reflects that, participantsin the non-focal, but not focal, condition felt that painmade the PM task more difficult.

Interestingly, when the composite WM capacity scorewas entered as a covariate on only the non-focal PM con-dition, the ANCOVA revealed a significant interaction ofWM and pain condition, F(1, 135) = 40.978, p < .001, h2

p

= .233. To examine the nature of this interaction, we did amedian split of the data based on WM performance whichshowed that this interaction is driven by the fact that highWM capacity participants did not believe that the painmanipulation had an impact on the proportion of non-focal cues participants responded to successfully (M =5.86), whereas lowWMcapacity participantsweremore sen-sitive in their retrospective judgments regarding the nega-tive impact of the acute pain exposure (M = 2.67), t (66)= .803, p = .409 and t(69) = 11.312, p < .00, respectively.The fact that high-WMC participants showed similardecreases in non-focal PM performance in the pain con-dition relative to low-WMC participants may thereforereflect that these participants did not perceive any greaterdifficulty in PM remembering while they were in pain.

Thus, high-WM capacity individuals may not haveengaged in compensatory strategies to ensure successfulintention fulfilment that resulted in decreased PM perform-ance under acute pain conditions leading to the absence ofthe intended moderating effect of WM capacity on pain’seffects on PM (see Einstein & McDaniel, 2007 for a similarargument). While this latter analysis is exploratory innature, it underscores the importance of evaluating indi-vidual differences in PM through the lens of metacognitivemonitoring and control processes in future studies.

General discussion

Given the ubiquity of pain and PM demands in everydaylife, and considering that both pain and unsuccessful PMperformance disrupt daily living, it seems remarkable thatonly three studies have investigated the effects of painon PM performance. Drawing from the multi-process-theory of PM, our study aimed to provide a more refinedexamination of the nature and extent of PM impairmentdue to acute pain. To this end, we used a well-validatedparadigm known to manipulate PM demands, and com-pared PM performance while participants were underexperimentally-induced pain compared with no pain.

As hypothesised, we found that pain impaired event-based PM in conditions that place high demands on theexecutive processes required for successfully detectingthe PM cue (i.e., non-focal PM). In contrast, we replicatedGatzounis et al. (2018) in showing that pain does notimpair focal PM cue detection. These results are consistentwith experimental research showing no impoverished PM

performance, even under highly demanding divided atten-tion-conditions, when individuals rely on focal targets tocue them to perform some future action (Marsh,Hancock, & Hicks, 2002; Windy McNerney & West, 2007).These results also dovetail with prior research showingthat cue-focality reduces age and clinical-related PMdeficits (Foster et al., 2009; Kliegel et al., 2008; Rendellet al., 2007). Theoretically, our findings not only confirm aspecific prediction of the multi-process-view for PM retrie-val, but they also support the prediction of a mediationalmechanism proposed in a recent process-model of theclinical neuropsychology of PM (Kliegel et al., 2011).Namely, that PM deficits in a specific populations onlyemerge if there is a mismatch between PM demands(e.g., high level of controlled demands as found in nonfocaltasks) and the available cognitive resources (e.g., impairedexecutive control).

Studies of both experimentally-induced acute andchronic pain have demonstrated a detrimental effect onmore complex-tasks such as divided-attention, task switch-ing and memory updating tasks, but not on less cognitivelydemanding tasks (Moore et al., 2012; Moriarty et al., 2011).Given the current debate about whether pain-related cog-nitive interference is dependent on task-related factors andwhat sort of tasks are most susceptible to disruptive paineffects (Buhle & Wager, 2010; Veldhuijzen, Kenemans, deBruin, Olivier, & Volkerts, 2006), our results extend previousfindings by providing the first evidence that acute painfulexperiences can have deleterious effects on PM, but onlyunder certain conditions (i.e., highly demanding PMtasks). These effects were obtained with a group of partici-pants under experimentally-induced acute pain. Based onthese results and those reported by Ling and colleagues(Ling et al., 2007) and Miller and colleagues (Miller et al.,2014), there seems to be an open line of inquiry regardingchronic pain populations and their PM abilities under focalversus nonfocal conditions.

Although individual differences in working-memorycapacity were positively correlated with nonfocal (but notfocal) PM performance, working memory capacity did notmoderate the pain-related decline in nonfocal PM perform-ance. Also, the correlation between working memory andPM was smaller than previously reported findings whichis important to note. This finding suggests that the inter-ruptive function of acute pain on nonfocal PM appears tobe modulated by factors other than goal maintenance(Crombez, Eccleston, Baeyens, & Eelen, 1998; Karoly &Crombez, 2018).

As a potential explanation for the lack of moderation ofpain’s effects on PM via WM, it is worthy of note that highand low WM participants differed in their metacognitiveratings of the perceived difficulty of the nonfocal PM taskunder pain. This may have altered their approach to theexperiment and ultimately hindered our ability to findthe intended moderating effect. Another possibilityinvolves our assessment of WMC. Specifically, we assessedit using a reduced subset of the number of trials in the

MEMORY 7

complex-span tasks, and in so doing, we may have reducedthe variance in WMC scores, thus decreasing our power todetect a moderation effect (Oswald et al., 2015). Consider-ing the high prevalence of pain in the general population(Breivik et al., 2006; Karoly & Ruehlman, 2007), ourfindings of pain-related PM impairment underscore theneed for further investigation into the potential mechan-isms that may underlie these effects. This work will beimportant for developing training and support interven-tions that can help mitigate pain’s impact on day-to-daycognitive functioning where goal attainment is hindered.

The present experiment examined the role of acute painon prospective memory—a topic that has received littleempirical attention. By studying acute pain induced exper-imentally in healthy college-age volunteers, we ruled outextraneous factors (such as the influence of age-relatedmedical conditions, drug effects, and personality differ-ences) while carefully tracking short-term cue-detectionprocesses. Strong acute pain that is not modified by a cog-nitive or pharmacologic intervention yielded PM disruptionon in nonfocal conditions but not in focal conditions.

Furthermore, numerous studies of the placebo effect inpain have demonstrated the intricate connection betweensensory processing (nociception) and cognition (Moriartyet al., 2011; Pancyr & Genest, 1993), with the meaning,interpretation, or appraisal of pain now widely understoodas possessing as much or greater predictive power as theraw, interoceptive experience of pain or distress itself(Crombez, Vervaet, Baeyens, Lysens, & Eelen, 1996). Asplacebo treatments pivot on the creation of a positiveexpectancy of pain relief (Price, Finniss, & Benedetti,2008) and as such expectancies are basically predictionsor simulations of future state changes, the link betweenfelt pain and dimensions of prospective processingdeserves expanded empirical attention. Because PM is aform of episodic future thinking (Brewer & Marsh, 2010;Szpunar, Nathan Spreng, & Schacter, 2014) and becauseindividuals in pain tend, over time, to develop deficits per-taining to the cognitive control of attention, planning, andmemory for delayed task execution, the systematic study ofpain and PM has the potential to contribute to both basicand applied cognitive and clinical psychology (cf., Karoly,2018; Seligman, Railton, Baumeister, & Sripada, 2013).

There are several limitations of the current study thathighlight important directions for future research in thedomain of pain and PM. The sample consisted of universitystudents (18–35 years of age) which leaves open the ques-tion about whether a different pattern of results might befound with other populations (e.g., older adults, chronicpain patients, and various other clinical group thatexhibit PM deficits). The influence of acute and chronicpain on future-oriented behaviours like PM is an importantavenue for future research in these groups and we hopethe methods in the current study turn out to be usefulfor researchers. Another limitation of the current study isthat the experimental paradigm unfolded over a shorttime period, used non-personally relevant cues, and had

no consequences for PM success or failure. These factorsall have important influences on PM and may interactwith pain in theoretically meaningful ways. One moreimportant shortcoming of the current study is that noongoing task words were repeated in a manner similar tohow the focal cue repeated, and thus the lack of aneffect of pain on the focal condition is somewhat ambigu-ous (i.e., cues in the focality condition also differed in termsof item-repetition perhaps explaining the lack of an effectof pain on PM in this condition). This issue should beaddressed in future research. Finally, another interestingarea for future research is in time-based PM. Futurestudies should examine the impact of pain on otherforms of PM (e.g., time- and activity-based PM). Forexample, the impact of pain on time-based PM can beexamined from the perspective of temporal gate modelsof time perception. Specifically, it is our prediction thatpain will adjust attentional gating processes leading toshifts in time perception (Block & Zakay, 1997). This pain-induced shift in attention should greatly impact time per-ception and therefore time-based PM.

We believe that the incorporation of experimental dataalong with theories of PM is a crucial step toward under-standing how and why pain adversely affects daily func-tioning. The current research provides new insights intothe existing literature on pain-related cognitive impair-ment. Specifically, it points to a specific cognitive functionthat is disrupted by acute pain: the ability to remember toexecute delayed intentions. The effects of PM failures onquality of life are not minor. Emerging literature suggeststhat PM deficits predict functional dependence in everydaycapacity above and beyond impairment in other cognitiveabilities (Park & Kidder, 1996; Woods et al., 2009). Forexample, failures to notice prompts to take prescribedmedications are likely to result in poor treatment adher-ence and clinical outcomes. Therefore, if we can betterunderstand which specific task characteristics influencePM performance under varying pain conditions, we canimprove our ability to predict which individuals are mostat risk for PM disruption. Importantly, the results of thisstudy may suggest particularly effective strategies forimproving PM performance under acute pain conditions,such as using an approach of prominent, external cuesthat reduce executive demands and automaticallyprompt PM cue detection (McDaniel & Einstein, 2007).Accordingly, future studies might extend the current lab-oratory finding to a more naturalistic PM task.

Disclosure statement

No potential conflict of interest was reported by the authors.

References

Attridge, N., Keogh, E., & Eccleston, C. (2016). The effect of pain on taskswitching: Pain reduces accuracy and increases reaction timesacross multiple switching paradigms. Pain, 157(10), 2179–2193.

8 M. PITÃES ET AL.

Berryman, C., Stanton, T. R., Bowering, K. J., Tabor, A., McFarlane, A., &Moseley, G. L. (2014). Do people with chronic pain have impairedexecutive function? A meta-analytical review. Clinical PsychologyReview, 34(7), 563–579.

Berryman, C., Stanton, T. R., Jane Bowering, K., Tabor, A., McFarlane, A.,& Lorimer Moseley, G. (2013). Evidence for working memory deficitsin chronic pain: A systematic review and meta-analysis. Pain, 154(8),1181–1196.

Block, R. A., & Zakay, D. (1997). Prospective and retrospective durationjudgments: A meta-analytic review. Psychonomic Bulletin & Review, 4(2), 184–197.

Breivik, H., Collett, B., Ventafridda, V., Cohen, R., & Gallacher, D. (2006).Survey of chronic pain in Europe: Prevalence, impact on daily life,and treatment. European Journal of Pain, 10(4), 287–333.

Brewer, G. A., Knight, J. B., Marsh, R. L., & Unsworth, N. (2010). Individualdifferences in event-based prospective memory: Evidence for mul-tiple processes supporting cue detection. Memory & Cognition, 38(3), 304–311.

Brewer, G. A., & Marsh, R. L. (2010). On the role of episodic future simu-lation in encoding of prospective memories. Cognitive Neuroscience,1(2), 81–88.

Buhle, J., & Wager, T. D. (2010). Performance-dependent inhibition ofpain by an executive working memory task. Pain, 149(1), 19–26.

Crombez, G., Eccleston, C., Baeyens, F., & Eelen, P. (1998). Attentionaldisruption is enhanced by the threat of pain. Behaviour Researchand Therapy, 36(2), 195–204.

Crombez, G., Vervaet, L., Baeyens, F., Lysens, R., & Eelen, P. (1996). Dopain expectancies cause pain in chronic low back patients? A clini-cal investigation. Behaviour Research and Therapy, 34(11–12), 919–925.

Dismukes, R. K. (2008). Prospective memory in aviation and everydaysettings. In M. Kliegel, M. A. McDaniel, & G. O. Einstein (Eds.),Prospective memory: Cognitive, neuroscience, developmental, andapplied perspectives (pp. 411–431). New York: Lawrence ErlbaumAssociates.

Eccleston, C., & Crombez, G. (1999). Pain demands attention: A cogni-tive–affective model of the interruptive function of pain.Psychological Bulletin, 125(3), 356–366.

Einstein, G. O., & McDaniel, M. A. (1990). Normal aging and prospectivememory. Journal of Experimental Psychology: Learning, Memory, andCognition, 16(4), 717–726.

Einstein, G. O., & McDaniel, M. A. (2005). Prospective memory multipleretrieval processes. Current Directions in Psychological Science, 14(6),286–290.

Einstein, G. O., & McDaniel, M. A. (2007). Prospective memory andmetamemory: The skilled use of basic attentional and memory pro-cesses. Psychology of Learning and Motivation, 48, 145–173.

Einstein, G. O., McDaniel, M. A., Thomas, R., Mayfield, S., Shank, H.,Morrisette, N., & Breneiser, J. (2005). Multiple processes in prospec-tive memory retrieval: Factors determining monitoring versus spon-taneous retrieval. Journal of Experimental Psychology. General, 134(3), 327–342.

Ellis, J., & Kvavilashvili, L. (2000). Prospective memory in 2000: Past,present, and future directions. Applied Cognitive Psychology, 14(7),S1–S9.

Erskine, A., Morley, S., & Pearce, S. (1990). Memory for pain: A review.Pain, 41(3), 255–265.

Foster, E. R., McDaniel, M. A., Repovs, G., & Hershey, T. (2009). Prospectivememory in Parkinson disease across laboratory and self-reportedeveryday performance. Neuropsychology, 23(3), 347–358.

Gatzounis, R., Schrooten, M. G., Crombez, G., & Vlaeyen, J. W. (2018).Forgetting to remember? Prospective memory within the contextof pain. European Journal of Pain, 22(3), 614–625.

Karoly, P. (2018). Chronic pain and psychopathology: Exploring themotivational context and its clinical implications. In P. Karoly, & G.Crombez (Eds.), Motivational perspectives on chronic pain: Theory,research, and practice (pp. 355–403). New York: Oxford UniversityPress.

Karoly, P., & Crombez, G. (Eds.). (2018). Motivational perspectives onchronic pain: Theory, research and practice. New York: OxfordUniversity Press.

Karoly, P., & Ruehlman, L. S. (2007). Psychosocial aspects of pain-related life task interference: An exploratory analysis in a generalpopulation sample. Pain Medicine, 8(7), 563–572.

Katz, R. S., Heard, A. R., Mills, M., & Leavitt, F. (2004). The prevalence andclinical impact of reported cognitive difficulties (fibrofog) inpatients with rheumatic disease with and without fibromyalgia.Journal of Clinical Rheumatology: Practical Reports on Rheumatic &Musculoskeletal Diseases, 10(2), 53–58.

Kliegel, M., Altgassen, M., Hering, A., & Rose, N. S. (2011). A process-model based approach to prospective memory impairment inParkinson’s disease. Neuropsychologia, 49(8), 2166–2177.

Kliegel, M., Jäger, T., & Phillips, L. H. (2008). Adult age differences inevent-based prospective memory: A meta-analysis on the role offocal versus nonfocal cues. Psychology and Aging, 23(1), 203–208.

Kliegel, M., & Martin, M. (2003). Prospective memory research: Why is itrelevant? International Journal of Psychology: Journal Internationalde Psychologie, 38(4), 193–194.

Kline, R. B. (2015). Principles and practice of structural equation model-ing. New York, NY: Guilford Publications.

Ling, J., Campbell, C., Heffernan, T. M., & Greenough, C. G. (2007). Short-term prospective memory deficits in chronic back pain patients.Psychosomatic Medicine, 69(2), 144–148.

Marsh, R. L., Brewer, G. A., Jameson, J. P., Cook, G. I., Amir, N., & Hicks, J.L. (2009). Threat-related processing supports prospective memoryretrieval for people with obsessive tendencies. Memory (Hove,England), 17(6), 679–686.

Marsh, R. L., Hancock, T. W., & Hicks, J. L. (2002). The demands of anongoing activity influence the success of event-based prospectivememory. Psychonomic Bulletin & Review, 9(3), 604–610.

Mazza, S., Frot, M., & Rey, A. E. (In Press). A comprehensive literaturereview of chronic pain and memory. Progress in Neuro-Psychopharmacology & Biological Psychiatry.

McDaniel, M. A., & Einstein, G. O. (2000). Strategic and automatic pro-cesses in prospective memory retrieval: A multiprocess framework.Applied Cognitive Psychology, 14(7), S127–S144.

McDaniel, M. A., & Einstein, G. O. (2007). Prospective memory: An over-view and synthesis of an emerging field: An overview and synthesis ofan emerging field. Thousand Oaks, CA: Sage Publications.

Meeks, J. T., Pitães, M., & Brewer, G. A. (2015). The compensatory role ofimplementation intentions for young adults with low workingmemory capacity. Applied Cognitive Psychology, 29(5), 691–701.

Miller, A. K., Basso, M. R., Candilis, P. J., Combs, D. R., & Woods, S. P.(2014). Pain is associated with prospective memory dysfunction inmultiple sclerosis. Journal of Clinical and ExperimentalNeuropsychology, 36(8), 887–896.

Moore, D. J., Eccleston, C., & Keogh, E. (2017). Cognitive load selectivelyinfluences the interruptive effect of pain on attention. Pain, 158(10),2035–2041.

Moore, D. J., Keogh, E., & Eccleston, C. (2012). The interruptive effect ofpain on attention. Quarterly Journal of Experimental Psychology, 65(3), 565–586.

Moriarty, O., McGuire, B. E., & Finn, D. P. (2011). The effect of pain oncognitive function: A review of clinical and preclinical research.Progress in Neurobiology, 93(3), 385–404.

Muñoz, M., & Esteve, R. (2005). Reports of memory functioning bypatients with chronic pain. The Clinical Journal of Pain, 21(4), 287–291.

Noel, M., Chambers, C. T., McGrath, P. J., Klein, R. M., & Stewart, S. H.(2012). The influence of children’s pain memories on subsequentpain experience. Pain, 153(8), 1563–1572.

Okun, M., Karoly, P., Mun, C. J., & Kim, H. (2016). Pain-contingent inter-ruption and resumption of work goals: A within-day diary analysis.The Journal of Pain, 17(1), 65–75.

Ordemann, G. J., Opper, J., & Davalos, D. (2014). Prospective memory inschizophrenia: A review. Schizophrenia Research, 155(1–3), 77–89.

MEMORY 9

Oswald, F. L., McAbee, S. T., Redick, T. S., & Hambrick, D. Z. (2015). Thedevelopment of a short domain-general measure of workingmemory capacity. Behavior Research Methods, 47(4), 1343–1355.

Pancyr, G., & Genest, M. (1993). Cognition and pain experience. In K.Dobson, & P. C. Kendall (Eds.), Psychopathology and cognition (pp.121–158). San Diego, CA: Academic Press.

Park, D. C., & Kidder, D. (1996). Prospective memory and medicationadherence. In M. Brandimonte, G. Einstein, & M. McDaniel (Eds.),Prospective memory: Theory and applications (pp. 369–390).Hillsdale, NJ: Erlbaum.

Price, D. D., Finniss, D. G., & Benedetti, F. (2008). A comprehensivereview of the placebo effect: Recent advances and currentthought. Annual Review of Psychology, 59, 565–590.

Redelmeier, D. A., & Kahneman, D. (1996). Patients’memories of painfulmedical treatments: Real-time and retrospective evaluations of twominimally invasive procedures. Pain, 66(1), 3–8.

Redick, T. S., Broadway, J. M., Meier, M. E., Kuriakose, P. S., Unsworth, N.,Kane, M. J., & Engle, R. W. (2012). Measuring working memorycapacity with automated complex span tasks. European Journal ofPsychological Assessment, 28(3), 164–171.

Rendell, P. G., McDaniel, M. A., Forbes, R. D., & Einstein, G. O. (2007).Age-related effects in prospective memory are modulated byongoing task complexity and relation to target cue.Neuropsychology, Development, and Cognition. Section B, Aging,Neuropsychology and Cognition, 14(3), 236–256.

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 regu-larity, and cue focality. Psychology and Aging, 25(3), 595–605.

Scullin, M. K., Bugg, J. M., McDaniel, M. A., & Einstein, G. O. (2011).Prospective memory and aging: Preserved spontaneous retrieval,but impaired deactivation, in older adults. Memory & Cognition, 39(7), 1232–1240.

Seligman, M. E. P., Railton, P., Baumeister, R. F., & Sripada, C. (2013).Navigating into the future or driven by the past. Perspectives onPsychological Science, 8(2), 119–141.

Smith, R. E., & Bayen, U. J. (2005). The effects of working memoryresource availability on prospective memory: A formal modelingapproach. Experimental Psychology, 52(4), 243–256.

Szpunar, K. K., Nathan Spreng, R., & Schacter, D. L. (2014). A taxonomyof prospection: Introducing an organizational framework for future-oriented cognition: Fig. 1. Proceedings of the National Academy ofSciences, 111(52), 18414–18421.

Unsworth, N., Brewer, G. A., & Spillers, G. J. (2012). Variation in cognitivefailures: An individual differences investigation of everyday atten-tion and memory failures. Journal of Memory and Language, 67(1),1–16.

Unsworth, N., Heitz, R. P., Schrock, J. C., & Engle, R. W. (2005). An auto-mated version of the operation span task. Behavior ResearchMethods, 37(3), 498–505.

Veldhuijzen, D. S., Kenemans, J. L., de Bruin, C. M., Olivier, B., & Volkerts,E. R. (2006). Pain and attention: Attentional disruption or distrac-tion? The Journal of Pain, 7(1), 11–20.

Vlaeyen, J. W., & Linton, S. J. (2012). Fear-avoidance model of chronicmusculoskeletal pain: 12 years on. Pain, 153(6), 1144–1147.

Windy McNerney, M., & West, R. (2007). An imperfect relationshipbetween prospective memory and the prospective interferenceeffect. Memory & Cognition, 35(2), 275–282.

Woods, S. P., Dawson, M. S., Weber, E., Gibson, S., Grant, I., & Atkinson, J.H. & HIV NEUROBEHAVIORAL RESEARCH CENTER GROUP (2009).Timing is everything: Antiretroviral nonadherence is associatedwith impairment in time-based prospective memory. Journal ofthe International Neuropsychological Society, 15(1), 42–52.

Zuber, S., Kliegel, M., & Ihle, A. (2016). An individual difference perspec-tive on focal versus nonfocal prospective memory. Memory &Cognition, 44(8), 1192–1203.

10 M. PITÃES ET AL.