《认知心理学》课程教学资源（书籍文献）反射性注意的返回抑制起始时间可显著预测认知老化 Onset Time of Inhibition of Return Is a Promising Index for Assessing Cognitive Functions in Older Adults

caesJ6eronm1BPeiOi5asaC3S03Ve5R65g8 ◆GERONTOLOGICAL doi:10.1093/geronb/gby7 Advance Access publication June6018 OXFORD SOCIETY OF AMERICA Research Article Onset Time of Inhibition of Return Is a Promising Index for Assessing Cognitive Functions in Older Adults Tingni Li,BSc,Lei Wang,MEng,Wanyi Huang, PhD,Yanfen Zhen,BSc, Chupeng Zhong,BSc,Zhe Qu,PhD,and Yulong Ding,PhD Brain and Cognition Laboratory,Department of Psychology,Sun Yat-Sen University,Guangzhou,China. rsity,132 Waihua E-mail:quzhe@mail.sysu.edu.cn Received:August 13,17Editorial Decision Date:May.018 Decision Edito Nicole Anderson,PhD,CPsych Abstract Objectives:Developing efficient tools for assessing general cognitive functions in older adults is essential.Previous studies found that inhibition of retu (IOR)occurred later in the older adults than in the younger.However,little is known abour Methods:In two studies,the IOR-OT of healthy younger and older adults was measured by a modified Posner peripheral kEcoencnicoehecmndhAhnokoscaeeEwm 47535:33575MsG sho nt correlation tween IOR-OT and cognitive functions as assessed by was ace owlcdge K score latively st IOR-OT and cognitive adults.These inding provide new evidence supporting deficit theory of aging and lay the foundation of using IOR-OT as an objective measure of cognitive functions in the aging population. Keywords:ACE-R,Cognitive aing,,Visual attention With population aging,the cognitive wellbeing of older Della Sala.Henry.2004).could lead to an abnorma adults has become an important issue for society.Cognitive cognitive decline-a major challenge for healthy aging.The screening of abnormal cognitive decline requires objective on 04 Ma with age in late adulthood.Many cogni age,such as process 20222 Verhaeghen &Cerela 2002)and memory (Krame Robins,Helzer,1983;Tombaugh Mclntyre,1992) et al.,2006).In the process of cognitive aging,some dis and Montreal Cognitive Assessment (MoCA)(Nasreddine eases.such as Alzheimer's disease (AD)(Amieva.Phillips et al,2005)are the most widely used cognitive assessment

© The Author(s) 2018. Published by Oxford University Press on behalf of The Gerontological Society of America. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com. 753 Research Article Onset Time of Inhibition of Return Is a Promising Index for Assessing Cognitive Functions in Older Adults Tingni  Li, BSc, Lei  Wang, MEng, Wanyi  Huang, PhD, Yanfen  Zhen, BSc, Chupeng Zhong, BSc, Zhe Qu, PhD, and Yulong Ding, PhD Brain and Cognition Laboratory, Department of Psychology, Sun Yat-Sen University, Guangzhou, China. Address correspondence to: Yulong Ding, PhD, Brain and Cognition Laboratory, Department of Psychology, Sun Yat-Sen University, 132 Waihuan Donglu, Higher Education Mega Center, Guangzhou 510006, China. E-mail: edsdyl@mail.sysu.edu.cn Or Zhe Qu, PhD, Brain and Cognition Laboratory, Department of Psychology, Sun Yat-Sen University, 132 Waihuan Donglu, Higher Education Mega Center, Guangzhou 510006, China. E-mail: quzhe@mail.sysu.edu.cn Received: August 13, 2017; Editorial Decision Date: May 22, 2018 Decision Editor: Nicole Anderson, PhD, CPsych Abstract Objectives: Developing efficient tools for assessing general cognitive functions in older adults is essential. Previous studies found that inhibition of return (IOR) occurred later in the older adults than in the younger. However, little is known about the relationship between the onset time of IOR (IOR-OT) and cognitive functions in the aging population. The present study examined this issue and investigated the potential of using IOR-OT as an index of cognitive functioning in older adults. Methods: In two studies, the IOR-OT of healthy younger and older adults was measured by a modified Posner peripheral cueing task, and cognitive functions of the older adults were evaluated with the Addenbrooke’s Cognitive Examination Revised (ACE-R). Results: Both studies showed a significant correlation (r = ~.5) between IOR-OT and cognitive functions as assessed by ACE-R in older individuals: later IOR-OT was accompanied by a lower ACE-R score. Discussion: To our knowledge, the present studies are the first to discover a relatively strong correlation between IOR-OT and cognitive functions in older adults. These findings provide new evidence supporting the inhibition deficit theory of aging and lay the foundation of using IOR-OT as an objective measure of cognitive functions in the aging population. Keywords: ACE-R, Cognitive aging, Inhibition ability, Visual attention With population aging, the cognitive wellbeing of older adults has become an important issue for society. Cognitive aging refers to the phenomenon that individuals’ cognitive functions decline with age in late adulthood. Many cogni￾tive functions decline with age, such as processing speed (e.g., Cerrella, 1985; Salthouse, 1996), attention (e.g., Verhaeghen & Cerella, 2002), and memory (e.g., Kramer et al., 2006). In the process of cognitive aging, some dis￾eases, such as Alzheimer’s disease (AD) (Amieva, Phillips, Della Sala, & Henry, 2004), could lead to an abnormal cognitive decline—a major challenge for healthy aging. The screening of abnormal cognitive decline requires objective and efficient tools for assessing cognitive functions in older adults. The Mini-Mental State Examination (MMSE) (Folstein, Robins, & Helzer, 1983; Tombaugh & McIntyre, 1992) and Montreal Cognitive Assessment (MoCA) (Nasreddine et al., 2005) are the most widely used cognitive assessment Journals of Gerontology: Psychological Sciences PSYCHOLOGICAL SCIENCES Journals of Gerontology: Psychological Sciences cite as: J Gerontol B Psychol Sci Soc Sci, 2020, Vol. 75, No. 4, 753–761 doi:10.1093/geronb/gby070 Advance Access publication June 6, 2018 Downloaded from https://academic.oup.com/psychsocgerontology/article/75/4/753/5033575 by Southern Medical University user on 04 May 2022

754 niicapracticc.Theseowevrae on (in ator able for evaluatine cognition comprehensively To address long SOA conditions (e.g>500 ms),the response pattern these limitations,researchers have developed a more sen- reverses(i.e.,responses to targets occurring at the cued loca menon is called IOR.Although 2000)ACE-R extends the content of MMSE me five major cognitive functions:attention and orientation ing,Hopfinger Mangun,2001),the prevailing view holds function,epis odic and semantic memory, verbal flue hat IOR,a ry after functio pre Dawson.Mirchell.Arnold-Hodges.2006).the ACE-R is tion may begin when atrention mov away from the cuce recommended as a suitable and efficient tool for evaluar location,or when the cue is presented(but the early,atten ons in older adults (Larner Mitchell. obscures the inhibition;Klein ding the with norma cognitive functions o on the However there are two maio sho for thes The IOR effect a neuropsychological tests in practical application.First,per facilitation,therefore any mechanism that decreases inhibi formance on neuropsychological tests may be influenced by the eve Prev tud 014 eof IOR cipants with limir 50 and 3 000 ms Castel.Chas ity or low education may have difficulty completing the Pratt (2003)compared the time course of IOR in younge assessment. and older adults.When rese rs plotted cueing effects as also and/o they found mpare in. SOA rhat th to develop more objective and efficient asses ssment tools onset time (IOR-OT)delayed A novel approach to this issue is to identify key cognitive This delay in IOR-OT has been repli ed in a numbe processes mpacted by cog ive aging and then to ex (e.g,Bao,Zhe Fu,2 abili in th Amon the many theories addressing th auses of co 2016:Wasch The inhibitory deficit theory may account for the delay of IOR-OT in older adults.Inhibi ory processes are consic to act the vice of goals to (a)preve rrelevan dthe mos g ac ally would not effectively inhibit irelevant information tion,and (c)restrain prepotent re nses (Hasher.Lustig nctions (e.g..slowdo might no Hash mer ding compre rom a and he by a sulting in a 014)R L. a weaker inhibitory ability woud It remain unclear howeyer whether the delay of nitive functions IOR-OT in the aging population is associated with their genera function (IOR)is w of IOp an teddcross ral cueing task develo and Cohen (1984).In d that IOR-OT and functionin sence of are related.However,in many conditions,if we merel Before the 0 mpar the ma (SOA) both foo and intelli short (e.g.,200 ms),the reaction time (RT)at the cued and adults,which leads to an apparent"correlation"that

tools in clinical practice. These tests, however, are insensi￾tive in differentiating levels of cognitive functioning (Luis, Keegan, & Mullan, 2009; Mitchell, 2009), and are not suit￾able for evaluating cognition comprehensively. To address these limitations, researchers have developed a more sen￾sitive and comprehensive cognitive test battery, known as the Addenbrooke’s Cognitive Examination Revised (ACE￾R) (Mathuranath, Nestor, Berrios, Rakowicz, & Hodges, 2000). ACE-R extends the content of MMSE, measuring five major cognitive functions: attention and orientation function, episodic and semantic memory, verbal fluency, language function, and visuo-spatial ability. With good val￾idity and cross-culture usability (Fang et al., 2014; Mioshi, Dawson, Mitchell, Arnold, & Hodges, 2006), the ACE-R is recommended as a suitable and efficient tool for evaluat￾ing cognitive functions in older adults (Larner & Mitchell, 2014), including those with normal cognitive functions or unnoticeable cognitive declines. However, there are two major shortcomings for these neuropsychological tests in practical application. First, per￾formance on neuropsychological tests may be influenced by the language capacity or education level of the par￾ticipants (Gamaldo et al., 2018; Velayudhan et al., 2014). Specifically, participants with limitations in language abil￾ity or low education may have difficulty completing the assessment. Second, using these neuropsychological tests also requires professional assessment skills and/or lan￾guage skills of interviewers, which could consume social resources in personnel training. Therefore, it is necessary to develop more objective and efficient assessment tools. A novel approach to this issue is to identify key cognitive processes impacted by cognitive aging and then to examine whether these processes could account for individual differ￾ences in general cognitive ability in the aging population. Among the many theories addressing the causes of cog￾nitive aging (see Anderson & Craik, 2017 for a review), the inhibitory deficit theory (Hasher, Stoltzfus, Zack, & Rypma, 1991; Hasher & Zack, 1988) is arguably the most influential. This theory proposes that older adults gener￾ally would not effectively inhibit irrelevant information or stop prepotent responses, leading to a decline of cogni￾tive and behavioral functions (e.g., slowdown of response, reduced memory, and poor reading comprehension, Weeks & Hasher, 2014). Based on this theory, an older adult with a weaker inhibitory ability would have poorer general cog￾nitive functions. In the study of inhibitory processes, inhibition of return (IOR) is widely regarded as a reliable indicator of inhibitory attentional control. IOR is measured by the Posner periph￾eral cueing task developed by Posner and Cohen (1984). In this task, participants are asked to detect the presence of the target and respond as quickly as possible. Before the display of the target, an uninformative peripheral cue that does not predict the location of the target is presented. When the cue-target stimulus onset asynchrony (SOA) is short (e.g., < 200 ms), the reaction time (RT) at the cued location (valid trial) is faster than that at the uncued loca￾tion (invalid trial). This is called a facilitatory effect, indi￾cating a reflexive shift of attention to the cued location. In long SOA conditions (e.g., > 500 ms), the response pattern reverses (i.e., responses to targets occurring at the cued loca￾tion are slowed). This phenomenon is called IOR. Although the specific processes of information processing affected by IOR remain controversial (e.g., response processes, Pastötter, Hanslmayr, & Bäuml, 2008; or sensory process￾ing, Hopfinger & Mangun, 2001), the prevailing view holds that IOR, as an inhibitory aftereffect, reflects inhibition of the previously attended location (Dukewich & Klein, 2015; Gazzaniga, Ivry, & Mangun, 2014; Klein, 2000). The inhibi￾tion may begin when attention moves away from the cued location, or when the cue is presented (but the early, atten￾tionally-mediated facilitation obscures the inhibition; Klein, 2000). The IOR effect may depend on the relative strength of two opponent processes (i.e., facilitation vs inhibition). The IOR effect appears when the inhibition dominates the facilitation, therefore any mechanism that decreases inhibi￾tory control will delay the appearance of IOR. Previous studies found an age-related difference of IOR effects. For instance, with 11 cue-target SOAs that ranged between 50 and 3,000  ms, Castel, Chasteen, Scialfa, and Pratt (2003) compared the time course of IOR in younger and older adults. When researchers plotted cueing effects as a function of cue-target SOAs they found that, compared with younger adults, IOR effects appeared in longer cue￾target SOA conditions in older adults, meaning that the onset time of IOR (IOR-OT) in older adults was delayed. This delay in IOR-OT has been replicated in a number of subsequent studies (e.g., Bao, Zhou, & Fu, 2004; Langley, Friesen, Saville, & Ciernia, 2011; Muiños, Palmero, & Ballesteros, 2016; Wascher, Falkenstein, & Wildwall, 2011). The inhibitory deficit theory may account for the delay of IOR-OT in older adults. Inhibitory processes are consid￾ered to act in the service of goals to (a) prevent irrelevant information from gaining access to the focus of attention, (b) delete irrelevant information from the focus of atten￾tion, and (c) restrain prepotent responses (Hasher, Lustig, & Zack, 2007). Thus, due to their deficits in inhibitory abil￾ity, older adults might not efficiently prohibit and/or recover from attentional capture by a distracting cue, resulting in a longer dwell time and hence a later onset of IOR. It remains unclear, however, whether the delay of IOR-OT in the aging population is associated with their decline in general cognitive functions. Previous cross￾sectional studies have identified an age-related decline in both onset of IOR and general cognitive functioning. These results suggested that IOR-OT and cognitive functioning are related. However, in many conditions, if we merely compare the data of different age groups and neglect the confounding effect of age, we may find an unreliable “cor￾relation” between factors. For example, there are differ￾ences in both foot size and intelligence between children and adults, which leads to an apparent “correlation” that 754 Journals of Gerontology: PSYCHOLOGICAL SCIENCES, 2020, Vol. 75, No. 4 Downloaded from https://academic.oup.com/psychsocgerontology/article/75/4/753/5033575 by Southern Medical University user on 04 May 2022

Journals of Gerontology:PSYCHOLOGICAL SCIENCES,2020,Vol.75,No.4 those with bigger feet have higher intelligence.But when were functionally normal using the screening criteria of Fang controlling for et al.(2014).To ensure reliable asse ment of ACE-R,olde correlation were redu the use of IOR-OT in ing cognitive functionsinolder individuals.To the best of Sun Yat-Sen Universiry. our knowl dge,however,no research has been conducted directh Apparatus and procedure OR-OT cognitive functions in older individuals.Specifically,we The part hhrghrmforanedhe a sound attenuate given a posne eripheral cueing task and an ACE-R assess ate ment within half a day.The order of the ACE-R and Posner general cognitive functioning in older adults:the later one's .200 IOR-OT i its Chinese translation version Hodges (2013),translated on,thi by Xiong,Liu Yang and edited by Zhou). The Posner peripheral cueing task was similar to that of stand what is"normal"in cognitive aging and can help tal.(2003).Parti define what is“abn in future studies (e.g..Petersen. with a 1.024 768 resolution (32 cm19 cm Res were made on an external numeric kevpad.Participants were asked to stare at a fixation cross at the center of the creen.I e of cach t is sh wn In l Study 1 nd on each trial a white oss fixation(0.5° en lor-OT and general cognitive functions in older adults We adopted a Posner peripheral cue play and these remained on the screen until a response was with six SOAs tim oxes w of the initial dis observer showed an IOR effect of at least 10 ms,the pr play,a bigger and concentric white box (the cue,1.4 sent study innovatively evaluates IOR-OT for each individ. 1.4)app red at the outer o one of the two horizonta ual using a quadratic polynomial fitting approach boxes for 50 ms.In 80 of the trials,a (th the cent one of th T Method c0 uld be50.100.250,500,750,0r1,000ms;these S0A randomly selected with withir Th older adults (56-85 =70.39,1 rget w as the cue alid(valid m the from the community in Zhuhai,China.They had an averag 12.79(SD ties The remaining 20%of trials served as catch trials in 2.8 range:6- 16)years of edu tion.For repl cating the Cn the to respond as acc (18-23 vears.mean age=20.59.8males)were also recruited. wa presente o0 ms Yat-Sen Un 139 (including 7 catch trials)with five short breaks.Twelve d younger participants were practice trials were given before the formal study. fluent in Chinese.All participants reported normal or cor Data prepro cessing For the ACE-R evaluation,the total score of each partic known genera In the Posne r peripheral cucing task and had an ACE-R score not which means they RTs bevond SD above or below a given participant's mear

those with bigger feet have higher intelligence. But when controlling for age, the correlation between foot size and intelligence would disappear. The correlation between IOR-OT and general cognitive functions among a group of older adults would support the use of IOR-OT in assess￾ing cognitive functions in older individuals. To the best of our knowledge, however, no research has been conducted to explore their relationship. The main purpose of this research was to directly explore the relationship between IOR-OT and the general cognitive functions in older individuals. Specifically, we used the Posner peripheral cueing task to measure individu￾al’s IOR-OT and the ACE-R to evaluate cognitive functions of older adults. Based on the inhibitory deficit theory, we expect to find a negative association between IOR-OT and general cognitive functioning in older adults: the later one’s IOR-OT is, the worse are his/her general cognitive func￾tions. In addition, this research focuses on the assessment of cognitive function in healthy older adults. The data from cognitively normal older adults allow us to better under￾stand what is “normal” in cognitive aging and can help define what is “abnormal” in future studies (e.g., Petersen, 2004; Sperling et al., 2011). Study 1 The main purpose of Study 1 is to directly explore the rela￾tionship between IOR-OT and general cognitive functions in older adults. We adopted a Posner peripheral cueing task with six SOAs to characterize the time course of IOR. Unlike previous studies (e.g., Castel et  al., 2003), which defined individual IOR-OT as the first SOA where each observer showed an IOR effect of at least 10 ms, the pre￾sent study innovatively evaluates IOR-OT for each individ￾ual using a quadratic polynomial fitting approach. Method Participant Thirty-three older adults (56–85 years, mean age = 70.39, 11 males) participated in Study 1. Older adults were recruited from the community in Zhuhai, China. They had an average 12.79 (SD = 2.88; range: 6–16) years of education. For repli￾cating the age difference in the time course of IOR reported in previous studies (i.e., Castel et al., 2003), 22 younger adults (18–23 years, mean age = 20.59, 8 males) were also recruited. The younger adults were undergraduate students from Sun Yat-Sen University and had an average 14.32 (SD = 1.39) years of education. All older and younger participants were recruited as paid volunteers. Informed consent was obtained from all participants before the study. All participants were fluent in Chinese. All participants reported normal or cor￾rected-to-normal vision. None of the participants suffered any known general psychiatric or neurological disease. All older participants self-reported no obvious behavioral impairment and had an ACE-R score not less than 85, which means they were functionally normal using the screening criteria of Fang et al. (2014). To ensure reliable assessment of ACE-R, older participants were required to have at least 6 years of educa￾tion (i.e., primary school education). The study was approved by the Ethics Committee of the Department of Psychology, Sun Yat-Sen University. Apparatus and procedure The participants performed the tasks in a sound attenuated room with bright illumination. All older participants were given a Posner peripheral cueing task and an ACE-R assess￾ment within half a day. The order of the ACE-R and Posner peripheral cueing task administration was counterbalanced across participants. The measurement of ACE-R was in strict accordance with the standards (Hodges, 2007; and its Chinese translation version Hodges (2013), translated by Xiong, Liu & Yang and edited by Zhou). The Posner peripheral cueing task was similar to that of Castel et al. (2003). Participants were seated 60 cm in front of a 14-inch color monitor of an IBM Thinkpad laptop, with a 1,024 × 768 resolution (32 cm × 19 cm). Responses were made on an external numeric keypad. Participants were asked to stare at a fixation cross at the center of the screen. The sequence of each trial is shown in Figure 1. Full contrast stimuli (white or green) were presented on a black background. On each trial, a white cross fixation (0.5° × 0.5°) and two white boxes (1° × 1°) served as an initial dis￾play and these remained on the screen until a response was made. The boxes were centered 5° from the fixation cross and located horizontally to the left and right of the fixation. About 1,000–1,200 ms after the presence of the initial dis￾play, a bigger and concentric white box (the cue, 1.4° × 1.4°) appeared at the outer of one of the two horizontal boxes for 50 ms. In 80% of the trials, a green circle (the target, 0.7°) was presented at the center of one of the two boxes for 2,000 ms or until response. The cue-target SOA could be 50, 100, 250, 500, 750, or 1,000 ms; these SOAs were randomly selected with equal probabilities within blocks of trials. The location of the target was either same as the cued location (valid trial) or different from the cued location (invalid trial), randomized with equal probabili￾ties. The remaining 20% of trials served as catch trials in which no target was presented. The participants were asked to respond as accurately and quickly as possible if the tar￾get was presented. A 500 ms blank display was presented after a response. The entire session consisted of 360 trials (including 72 catch trials) with five short breaks. Twelve practice trials were given before the formal study. Data preprocessing For the ACE-R evaluation, the total score of each partici￾pant was calculated. In the Posner peripheral cueing task, the RTs of noncatch trials were computed and analyzed. RTs beyond 3 SD above or below a given participant’s mean Journals of Gerontology: PSYCHOLOGICAL SCIENCES, 2020, Vol. 75, No. 4 755 Downloaded from https://academic.oup.com/psychsocgerontology/article/75/4/753/5033575 by Southern Medical University user on 04 May 2022

56 of:PSYCHOLGICAL SCIENCES,0,Vol.75,No.4 Typical participants + Cu-tame sr Gpn Group moans te square) sed 6 cu target SOA tch trials)to discourage anticipatory response SOA in Study 1 and 2.Each poir 1e0 2.09 n of stimulus onset asynchrony.The curved ine repres unger adults rest ectively res se accuracies includin hit rates of noncatch trials and false-alarm rates of catchtr r adults:bl dults).Th als,were also analyzed. Cueing effects (RTs in in alid trial the time e of the cue typica minus RIs h pa rive the IOR-OT,which is defined as the SOA at which the means for each age group in Studies1and. cucing effect is zero(typical cases are shown in Figure 2A). IOR-OT reflects the me poin ,996±0.002，Mean±SEFA:.007±0.002)and the younger e)to ar (hit99300o4h:08±002，mhn (sce Fig idity (valid,invalid)and SOA (six cu SOAs)as within ftcipansineach OA co subject factors and Age (old,young)as a between-subject polynomi led IOR-OT for S0A5.256 perspective,the time course function of IOR effects looks 1,53=42.134.p<001.n2 Age illustrated that the you ger pa ipants (344.8+6.1 ms us studies (e.g.,C responded faster than theo er adults (463.8=14.3 ms).The IOR effects an (E(1 53) 298g alidin xS0AF5,2651=40.749,p 2=435 as well as a significant 3-way interaction of Cue validity (0. group,respe a tempo I dynamic o .579 (0.053)and .529 (0.046)for the older group and the 2B).The Cue younger group,respectively. interaction indicated that the tempora dynamic pattern of cueing effects was different between olde adults and yo see Figure 2B) Result unger adults that In the ACE-R 2 particip nd ld 500 ms (ps<.05),but not at SOA 750 and 1.000 ms (s>43) Posner peripheral cueing task,the hit rates were high and Consistent with the significant three-way inter the false-alarm(FA)rates were low for both the older (hit: action described above,the IOR-OT estimated by

were removed from further analysis. On average, 2.0% and 2.2% of noncatch trials were removed for the older and the younger adults, respectively. Response accuracies, including hit rates of noncatch trials and false-alarm rates of catch tri￾als, were also analyzed. Cueing effects (RTs in invalid trials minus RTs in valid trials) in the six SOAs for each partici￾pant were fitted with quadratic polynomial function to de￾rive the IOR-OT, which is defined as the SOA at which the cueing effect is zero (typical cases are shown in Figure 2A). IOR-OT reflects the time point that the cueing effect begins to change from a facilitatory effect (positive value) to an IOR effect (negative value). We also computed the onset time of group average IOR effect for each age group (see Figure 2B). For each group, we averaged the cueing effects across all par￾ticipants in each SOA condition and then used the quadratic polynomial curve fitting to compute the group IOR-OT. Quadratic polynomial curve-fit was chosen to compute IOR-OT for the following reasons. First, from the graphic perspective, the time course function of IOR effects looks like a quadratic polynomial curve both in our results (see Figure 2) and in previous studies (e.g., Castel et al., 2003). Second, from the data-driven perspective, the time course of IOR effects in the present study fits well with the quadratic polynomial curve. The mean fitting degree R2 s using quad￾ratic polynomial curve fitting are .706 (0.041, SE) and .731 (0.032) for the older group and the younger group, respect￾ively. These R2 s are much better than the fitting degree R2 s of linear fitting (ps < .001, Cohen’s ds > 0.47), which are .579 (0.053) and .529 (0.046) for the older group and the younger group, respectively. Result In the ACE-R assessment, the older participants obtained a mean score of 92.67 (SE = 0.753; range: 85–100). In the Posner peripheral cueing task, the hit rates were high and the false-alarm (FA) rates were low for both the older (hit: .996 ± 0.002, Mean ± SE; FA: .007 ± 0.002) and the younger participants (hit: .993 ± 0.004; FA: .008 ± 0.002), with no significant differences between groups (ps > .5). A three-way mixed-design analysis of variance (ANOVA) with Cue val￾idity (valid, invalid) and SOA (six cue-target SOAs) as within￾subject factors and Age (old, young) as a between-subject factor was conducted on RTs. This revealed significant main effects of Cue validity (F(1, 53) = 4.658, p = .035, ηp 2  = .081), SOA (F(5, 256)  =  6.835, p < .001, ηp 2   =  .114), and Age (F(1, 53) = 42.134, p < .001, ηp 2  = .443). The main effect of Age illustrated that the younger participants (344.8 ± 6.1 ms) responded faster than the older adults (463.8 ± 14.3 ms). The analysis also showed significant 2-way interactions of Cue validity × Age (F(1, 53) = 29.891, p < .001, ηp 2  = .361), and Cue validity × SOA (F(5, 265) = 40.749, p < .001, ηp 2  = .435), as well as a significant 3-way interaction of Cue validity × SOA × Age (F(5, 265) = 4.713, p < .001, ηp 2  = .082). The Cue validity × SOA interaction reflected a temporal dynamic of cueing effects across SOAs, from facilitatory effects at shorter SOAs to IOR effects at longer SOAs (see Figure 2B). The Cue validity × SOA × Age interaction indicated that the temporal dynamic pattern of cueing effects was different between older adults and younger adults (see Figure 2B). Further analyses showed that the cueing effects were significantly different between younger and older adults at SOA 50, 100, 250, and 500 ms (ps < .05), but not at SOA 750 and 1,000 ms (ps > .43). Consistent with the significant three-way inter￾action described above, the IOR-OT estimated by Figure 2. Time courses of cueing effects in Study 1 and 2. Each point (red circle for the old and blue triangle for the young) represents a cue￾ing effect (reaction time [RT] for invalid trials minus RT for valid trials) as a function of stimulus onset asynchrony. The curved line represents the quadratic polynomial curve indicating the time course of cueing effects (red = older adults; blue = younger adults). The abscissa where the fitted line crosses the horizontal axis represents the IOR-OT. Panel A and C show the time course of the cueing effects of a typical partici￾pant in each age group in Studies 1 and 2, respectively. Panel B and D show the time course of the cueing effects derived from the group means for each age group in Studies 1 and 2, respectively. Figure 1. The sequence of trial events in the Posner peripheral cueing task in Study 1 and 2. In this task, participants pressed a key in response to a target (a green circle) that was preceded by a non-predictive cue (a bigger white square). Study 1 used 6 cue-target SOAs, whereas Study 2 used 58 cue-target SOAs. The target was not present on 20% of tri￾als (catch trials) to discourage anticipatory responses. SOA = Stimulus onset asynchrony. 756 Journals of Gerontology: PSYCHOLOGICAL SCIENCES, 2020, Vol. 75, No. 4 Downloaded from https://academic.oup.com/psychsocgerontology/article/75/4/753/5033575 by Southern Medical University user on 04 May 2022

Journals of Gerontology:PSYCHOLOGICAL SCIENCES,2020,Vol.75,No.4 757 quadratic polynomial fitting was much later in the older factors,such as general processing speed and age.Is such 359 ms) ms;(53) eters use ,or reprod nnemof1oROofiheoa adopted a modifed os task with 58 that of the younger group (sSupplementary Figure 1B) OIs in th younger lation IOR-OT and gen eral cognitive functions of the older adults,Pearson's cor Method relation analyses were carried out As shown in Figure 3A corr the ParicipantoderadhlspanidipaiedinsSndy2,olde adults were recruited from the community in Zhuhai 311 524.=0021.Th s that.for older adults,worse performance on the ACE-R assessment smen use he ha led to 85 Th ut the ning 27 p s5993 :Salth 19g 5 males)were included in further analyses,with an aver tial correlation analyses were used to further verify the cor age 11.37 (SD=3.28;range:6-16)years of education rty-hve young adults (16-2 an age 19. 11 gage mean R and had an average 13.84 (SD =0.67)vears of cducation 6 IOR OT ACE R remained significant (s)and these partial correla- All older and younger participants were recruited as paid ed was obtai ed fron all par tions did not significantly differ from the simple correlation ween IOR-OT nd ACE-R 73).These re Chin had at least 6years of education.Non of the participants suffered any k nown general psychiatri The study was by the Study2 In Study 1,a significant negative correlation between ACE-R performance in old r individuals was ion was invulnerable to confounding Apparatus and procedure The procedure in Study 2 was similar to Study 1,except that Study 1:6 SOAs Study 2 58 SOAS in Study 2,the Posner peripheral cueing task contained 58 Table 1.Partial Correlations Between IOR-OT and ACE-R Performance in Older Adults After Controlling for Processing Speed (RT or RT/ACC of the Posner peripheral cueing task). ti and Age,Respectively Contro factor RT RTIACC Age Study 1 ,三-517* =-516 ,=-524* p=.002 p=.002 p=.002 M酒 Study 2 4 516 516 2 Combined p<.001 p<.001 pc.001 OR-OT and ACE-R in th

quadratic polynomial fitting was much later in the older participants (597.2  ±  45.9  ms) than in the younger adults (252.8 ± 33.5 ms; t(53) = 6.055, p < .001, Cohen’s d = 1.663; see Supplementary Figure 1A). The distribution pattern of IOR-OT of the older group was different from that of the younger group (see Supplementary Figure 1B). Specifically, IOR-OTs of the older group were mostly dis￾tributed after 300 ms, while the majority of IOR-OTs in the younger group were distributed prior to 300 ms. To explore the relationship between IOR-OT and gen￾eral cognitive functions of the older adults, Pearson’s cor￾relation analyses were carried out. As shown in Figure 3A, a significant negative correlation between the individual IOR-OT and ACE-R scores in the older group was found (Pearson’s r(31) = −.524, p = .002). This means that, for older adults, worse performance on the ACE-R assessment was accompanied by a later onset of IOR. In order to rule out the confounding effect of age, as well as the possible impacts of general processing speed (Cerella, 1985; Muiños et al., 2016; Salthouse, 1996), par￾tial correlation analyses were used to further verify the cor￾relation between IOR-OT and ACE-R score in older adults. As shown in Table 1, taking age, mean RT, or inverse effi￾ciency (RT/accuracy) of the Posner cueing task as controlled factors, correlations between IOR-OT and ACE-R score remained significant (ps < .01), and these partial correla￾tions did not significantly differ from the simple correlation between IOR-OT and ACE-R (ps > .73). These results indi￾cated that the correlation could not be simply explained by decline of general processing speed or an increase of age. Study 2 In Study 1, a significant negative correlation between IOR-OT and ACE-R performance in older individuals was found, and this correlation was invulnerable to confounding factors, such as general processing speed and age. Is such a correlation limited to the specific experimental param￾eters used in Study 1, or reproducible with different experi￾mental parameters? To address this question, in Study 2, we adopted a modified Posner peripheral cueing task with 58 SOAs (similar to the task used in Song, Meng, Chen, Zhou, & Luo, 2014), and explored the relationship between IOR-OT and ACE-R performance in an older group again. Method Participant Twenty-eight older adults participated in Study 2.  Older adults were recruited from the community in Zhuhai, China. One of the older participants was excluded after the ACE-R assessment, because he had failed to meet the crite￾ria of normal cognitive function, with a score below 85. The remaining 27 participants (59–83 years, mean age = 65.74, 5 males) were included in further analyses, with an aver￾age 11.37 (SD  =  3.28; range: 6–16) years of education. Forty-five young adults (16–23 years, mean age = 19.89, 11 males) also participated in Study 2. The younger adults were undergraduate students from Sun Yat-Sen University and had an average 13.84 (SD = 0.67) years of education. All older and younger participants were recruited as paid volunteers. Informed consent was obtained from all par￾ticipants before the study. All participants were fluent in Chinese. All participants reported normal or corrected-to￾normal vision and had at least 6 years of education. None of the participants suffered any known general psychiatric or neurological disease. The study was approved by the Ethics Committee of the Department of Psychology, Sun Yat-Sen University. Apparatus and procedure The procedure in Study 2 was similar to Study 1, except that in Study 2, the Posner peripheral cueing task contained 58 Figure 3. The correlation between IOR-OT and ACE-R score in the older group for Study 1 (A), Study 2 (B), and Studies 1 and 2 combined (C). Table 1. Partial Correlations Between IOR-OT and ACE-R Performance in Older Adults After Controlling for Processing Speed (RT or RT/ACC of the Posner peripheral cueing task), and Age, Respectively Control factor RT RT/ACC Age Study 1 r = −.517** r = −.516** r = −.524** p = .002 p = .002 p = .002 Study 2 r = −.504** r = −.507** r = −.485* p = .009 p = .008 p = .012 Studies 1 and 2 Combined r = .516*** r = −.515*** r = .516*** p < .001 p < .001 p < .001 Notes: RT = Reaction time; ACC = Accuracy. *Correlation is significant at the .05 level (two-tailed). **Correlation is sig￾nificant at the .01 level (two-tailed). ***Correlation is significant at the .001 level (two-tailed). Journals of Gerontology: PSYCHOLOGICAL SCIENCES, 2020, Vol. 75, No. 4 757 Downloaded from https://academic.oup.com/psychsocgerontology/article/75/4/753/5033575 by Southern Medical University user on 04 May 2022