آیا سن حواس پرتی شنوایی را افزایش می دهد؟ ارتباط الکتروفیزیولوژیک عملکرد بالا و پایین در سالمندان

Abstract Aging usually affects the ability to focus attention on a given task and to ignore distractors. However, aging is also associated with increased between-subject variability, and it is unclear in which features of processing older high-performing and low-performing human beings may differ in goal-directed behavior. To study involuntary shifts in attention to task-irrelevant deviant stimuli and subsequent reorientation, we used an auditory distraction task and analyzed event-related potential measures (mismatch negativity), P3a and reorienting negativity) of 35 younger, 32 older high-performing, and 32 older low-performing participants. Although both high and low performing elderly individuals showed a delayed reorienting to the primary stimulus feature, relative to young participants, poor performance of the elderly participants in processing of deviant stimuli was associated with strong involuntary attention capture by task-irrelevant features. In contrast, high performance of the elderly group was associated with intensified attentional shifting toward the target features. Thus, it appears that performance deficits in aging are due to higher distractibility in combination with deficits in the orienting–reorienting mechanisms.

1. Introduction Healthy aging is associated with declines in various cognitive functions such as working memory capacity, processing speed, and inhibitory control (Van der Linden et al., 1999). In particular, elderly individuals often encounter difficulties in complex task settings where various concurring stimuli are present, and where top-down attentional control is needed to focus attention on a relevant event and to ignore irrelevant events to achieve efficient behavioral control (Duncan, 2006). As older adults are typically more readily distracted by task-irrelevant stimuli than are younger adults, goal-directed behavior of elderly persons could suffer from deficits in selectively promoting relevant stimuli and inhibiting irrelevant stimuli (reviewed by Kramer and Madden, 2008). The interplay of goal-directed and orientation-related cognitive processes in complex scenarios has been described within a three-stage model of distraction (Escera et al., 2000, Polich and Criado, 2006 and Schröger and Wolff, 1998): At the first stage of regularity extraction and deviance detection task-relevant information is filtered out of a stream of ongoing stimulation, and task-irrelevant information is automatically detected in terms of violations of regularities in sensory memory buffers. At the second stage of involuntary attention-switching the deviant information may lead to involuntary attention shifts that are, in a final stage of reorientation, to be compensated for by mechanisms restoring the optimal attention-set relevant for a given task. This distraction–orientation–refocusing cycle has experimentally been operationalized within an auditory distraction paradigm in which a sequence of repeating tones is intermixed with occasional irregular tones violating the repetition. Subjects have to respond to the standard tones, while ignoring deviant tone features ( Schröger et al., 2000). The basic processes taking place during the three stages can be distinguished by analysis of the event-related potentials (ERPs): Deviant stimuli typically evoke the fronto-central mismatch negativity (MMN) ( Näätänen et al., 1978) that is thought to be a physiological correlate of pre-attentive deviance detection ( Näätänen, 1990 and Sussman et al., 2003). The MMN is usually followed by the fronto-central P3a ( Friedman et al., 2001), a correlate of an involuntary attention-switching mechanism ( Escera et al., 2000, Friedman et al., 2001, Knight and Scabini, 1998 and Schröger, 1996). Finally, a late fronto-central negativity, so-called reorienting negativity (RON) ( Schröger and Wolff, 1998 and Schröger et al., 2000), is assumed to reflect re-allocation of attention to the relevant task after distraction by the deviant features (empirical evidence in Berti, 2008a and Hoelig and Berti, 2010). In contrast to the deviants, standard stimuli usually produce a fronto-central N1-P2 complex that is followed by the parietal P3b (reviewed by Polich and Criado, 2006). Although the N1 reflects an automatic processing of sensory input, the P2 has been related to attention allocation ( Potts, 2004), and the P3b to the allocation of working memory and processing resources ( Polich, 2007). Previous studies have shown that the interplay of deviance detection, involuntary attention shifts, and top-down attentional control mechanisms change with increasing age (Cooper et al., 2006, Horváth et al., 2009, Mager et al., 2005 and Ruzzoli et al., 2012), indicating an increased susceptibility to distracting stimuli. However, different cognitive subprocesses involved in the distraction–orientation–refocusing cycle may play a role: On the one hand, an age-related reduction of MMN relative to younger adults suggests specific deficits in encoding or retention of sensory information (Alain and Woods, 1999, Bertoli et al., 2002, Cooper et al., 2006, Czigler et al., 1992, Karayanidis et al., 1995, Pekkonen et al., 1993 and Ruzzoli et al., 2012). On the other hand, age-related changes in P3a (Czigler et al., 2006, Gaeta et al., 2001, Horváth et al., 2009 and Mager et al., 2005) and RON (Horváth et al., 2009 and Mager et al., 2005) suggest attentional orienting and reorienting to contribute to deficits observed in elderly persons. Beside this deficit view, it has been hypothesized that older adults may compensate for increased distractibility by stronger focusing of attention on the task-relevant stimuli (Horváth et al., 2009). In line with the decline-compensation hypothesis, elderly adults invest more effort than young adults to achieve the same level of performance (reviewed by Dennis and Cabeza, 2008). Neurophysiological studies thus demonstrated that especially high-performing older adults recruit extra cognitive resources, relative to their low-performing counterparts (Cabeza et al., 2002). In speech comprehension, for example, high-performing older adults efficiently counteracted age-related neural declines through increased allocation of attentional resources that were associated with additional activation of frontal brain areas (Getzmann, 2012 and Getzmann and Falkenstein, 2011). According to the effortfulness hypothesis (Wingfield et al., 2005), this extra effort requires and inflates cognitive resources. In sum, age-related declines in performance in the auditory distraction paradigm could be based on (1) specific deficits in the distraction–orientation–refocusing cycle, and (2) insufficient cognitive resources to completely compensate for these deficits. The present study investigated auditory distraction in younger and older adults, focusing on the question in which features of processing older high-performing and low-performing adults may differ in goal-directed behavior. Specifically, given the strong inter-individual variability in cognitive performance in elderly (Hultsch et al., 2002), we asked whether high performance of older participants was associated with preservation of processing capacities (enabling the distraction-orientation-refocusing cycle to be performed as efficient as the young) or compensation of actual deficits. To answer this question, a total of 129 older and 35 younger adults performed an auditory distraction task well suited to examine the different processes underlying the 3-stage model introduced above (Getzmann et al., 2013, Horváth et al., 2009 and Mager et al., 2005). Participants had to discriminate the duration of short and long tones that were either of high-probability standard frequency or of low-probability deviant frequencies (Schröger and Wolff, 1998). Thus, the participants had to attend to the task-relevant tone feature (i.e., its duration), while ignoring the distracting task-irrelevant tone feature (i.e., its pitch). To clarify the sources of performance differences within the older group, a high-performing (Old-High) and low-performing (Old-Low) subgroup was taken from the entire group of older participants, and ERPs on standard and deviant stimuli were analyzed for both older groups and the younger group. Significant differences in ERPs between the Old-High and Old-Low group, in combination with minor differences between the Old-High and Young group, would suggest preservation of processing capacities as a crucial factor. In this case, specific deficits of the Old-Low group in cognitive subprocesses involved in the distraction-orientation-refocusing cycle should become manifest by selective analyses of N1, MMN, P3a, and RON: differences in N1 and MMN would suggest deficits in sensory encoding and deviance detection, respectively, while differences in P3a and RON would suggest deficits in attentional orienting and re-orienting. Significant ERP differences between the Old-High and Young group, possibly in combination with increased frontal activation in the Old-High group relative to the Old-Low group, would suggest compensatory mechanisms to be at work, according to the decline-compensation hypothesis.

3. Results 3.1. Behavioral data The 3 groups of participants differed significantly in the rate of correct responses (main effect of Group: F2,96 = 9.92; p < 0.001; partial η2 = 0.17): Pairwise post-hoc tests showed that Old-Low participants produced fewer correct responses (71.6%) than Young participants (87.2%; p < 0.001) and Old-High participants (81.0%; p < 0.05; all values Bonferroni corrected), whereas the latter 2 groups did not differ from one another (p > 0.05). There was a higher rate of correct responses with standard than with deviant tones (main effect of Stimulus: F1,96 = 101.05; p < 0.001; partial η2 = 0.51). A significant interaction of Group and Stimulus (F2,96 = 87.98; p < 0.001; partial η2 = 0.65) indicated that the 3 groups differed significantly when responding to deviant tones, but not standard tones ( Fig. 2A). Accordingly, additional 1-way ANOVAs indicated a main effect of Group for deviant tones (F2,96 = 27.09; p < 0.001; partial η2 = 0.36), but not standard tones (F2,96 = 2.68; p > 0.05; partial η2 = 0.05). Pairwise post-hoc tests indicated that the Old-Low group (62.6%) produced fewer correct responses to deviant tones than the Young group (86.2%; p < 0.001) and the Old-High group (81.6%; p < 0.001), whereas there was no difference between the Young and Old-High group (p > 0.05). It should be noted that the percentage of correct responses of the Old-Low group differed significantly from change level (t[31] = 6.93; p < 0.001), indicating that, despite their relatively low performance, these participants were yet able to perform the task with the deviant tones. There were no significant differences in responses to standard tones between the Young (88.1%), Old-High (80.5%), and Old-Low groups (80.7%; all p > 0.05). Rates of correct responses (A) and response times (B) for the younger and the ... Fig. 2. Rates of correct responses (A) and response times (B) for the younger and the older high-performing and low-performing groups, shown separately for the frequent standard tones and the rare deviant tones. Error bars indicate standard errors across participants (younger: N = 35; older: N = 32). Figure options There was no main effect of Group on RTs (F2,96 = 0.49; p > 0.05; partial η2 = 0.02), and no interaction of Group and Stimulus (F2,96 = 1.20; p > 0.05; partial η2 = 0.02), but a significant main effect of Stimulus (F1,96 = 94.17; p < 0.001; partial η2 = 0.50), indicating higher RTs with deviant than with standard tones ( Fig. 2B). In sum, the accuracy in duration discrimination of the deviant tones was impaired in the Old-Low group, relative to the Old-High and Young groups (as expected according to the classification chosen). However, the 3 groups of participants did not differ in duration discrimination of the standard tones. No group differences in RTs were found. 3.2. ERPs to standard tones Grand average ERP-waveforms for standard tones at Fz, FCz, Cz, and Pz are shown in Fig. 3A for each group. Standard tones produced a typical fronto-central N1-P2 complex peaking at 94 ms and 169 ms, and a pronounced parietal P3b peaking at about 600 ms. Grand average event-related potentials (ERPs) of the frequent standard tones ... Fig. 3. Grand average event-related potentials (ERPs) of the frequent standard tones (A), the rare deviant tones (B), and difference waves (deviant minus standard tones; C) at Fz, FCz, Cz, and Pz, shown for the younger and the older high-performing and low-performing groups. Vertical lines reflect the onset of the tone stimulus. ERP components (N1, P2, P3a, P3b, MMN, and RON) are marked at the waveform of maximal amplitude. Figure options 3.2.1. N1 The N1 peaked at 94 ms after tone onset and did not differ in amplitude or latency between the groups (amplitude: F2,96 = 1.05; p > 0.05; partial η2 = 0.02; latency: F2,96 = 1.06; p > 0.05; partial η2 = 0.02); nor was there a significant between-group contrast (all p > 0.05). 3.2.2. P2 The P2 peaked at 170 ms after tone onset. There was no effect of Group on P2 at the vertex position (amplitude: F2,96 = 0.67; p > 0.05; partial η2 = 0.01; latency: F2,96 = 1.76; p > 0.05; partial η2 = 0.04). At the frontal position, however, the Old-High and Old-Low groups showed a pronounced positivity, whereas the Young group did not show a frontal positivity at all ( Fig. 3A). ANOVA indicated a significant between-group effect on P2 amplitude at Fz (F2,96 = 12.27; p < 0.001; partial η2 = 0.20), and significant contrasts occurred between the Young and Old-High group (t96 = 4.63; p < 0.001), and the Young and Old-Low group (t96 = 3.76; p < 0.001), but not between the Old-High and Old-Low group (t96 = 0.85; p > 0.05; Fig. 4A). Also, there was a between-group effect on P2 latency at Fz (F2,96 = 3.14; p < 0.05; partial η2 = 0.06), resulting from latencies being shorter in the Young group than the Old-Low group (t96 = 2.44; p < 0.05), whereas the 2 older groups (t96 = 0.76) and the Young and Old-High group (t96 = 1.66; both p > 0.05) did not differ from one another in P2 latency. Mean amplitudes (left) and latencies (right) of P2 at Fz (A) and P3b at Pz (B) ... Fig. 4. Mean amplitudes (left) and latencies (right) of P2 at Fz (A) and P3b at Pz (B) to the frequent standard tones, shown for the younger and the older high-performing and low-performing groups. Error bars indicate standard errors across participants (younger: N = 35; older: N = 32). Figure options To investigate a potential relationship between the frontal positivity of the older participants and their performance in duration discrimination, correlations of P2 amplitude at Fz and IE in the processing of standard tones were computed for the old participants (averaged across the Old-High and Old-Low groups) and for the young participants. There was a significant correlation for the old participants (r = −0.26; p < 0.05), but not for the young participants (r = 0.03; p > 0.05). Thus, superior performance in duration discrimination of standard tones came along with a strong P2 amplitude in the older participants, whereas performance of the younger participants was not related to P2 amplitude. 3.2.3. P3b The P3b peaked over parietal areas at 600 ms, that is, at 400 ms after the end of the short tone stimulus. The Young group had a pronounced P3b, whereas the older groups, especially the Old-Low group, showed a decrease in amplitude of the P3b (Fig. 4B). There was a significant between-group effect on P3b amplitude at Pz (F2,96 = 4.14; p < 0.05; partial η2 = 0.08), and the contrast analysis indicated a weaker P3b amplitude of Old-Low than Young participants (t96 = 2.71; p < 0.01). No significant differences occurred between the Young and Old-High group (t96 = 1.39) or between the older groups (t96 = 1.73; both p > 0.05). There was a significant between-group effect on P3b latency (F2,96 = 7.64; p < 0.001; partial η2 = 0.14): The Young group had a shorter P3b latency than the Old-High group (t96 = 4.23) and the Old-Low (t96 = 2.72; both p < 0.01), whereas there was no difference between the older groups (t96 = 0.77; p > 0.05). In sum, the analysis of the ERPs to standard tones revealed that the older participants had a pronounced early frontal positivity (P2) that was associated with a higher performance in duration discrimination, and that was nearly absent in the younger group. On the other hand, the activation over parietal areas (P3b) was delayed (and slightly reduced) in the older group relative to the younger group. No significant differences in P3b occurred between the Old-High and Old-Low group with standard tones. 3.3. Deviance-related ERPs Grand average ERP-waveforms for deviant tones at Fz, FCz, Cz, and Pz are shown in Fig. 3B, and the difference waveforms (deviant minus standard tones) are shown in Fig. 3C. 3.3.1. ERPs to deviant tones: N1 The N1 on deviant tones peaked at 97 ms. The N1 peak amplitude and latency did not differ between the groups (amplitude: F2,96 = 1.29; p > 0.05; partial η2 = 0.03; latency: F2,96 = 0.39; p > 0.05; partial η2 = 0.01), and there was no significant between-group contrast (all p > 0.05). 3.3.2. MMN The MMN peaked at 141 ms over fronto-central brain areas. Although there was no effect of Group on MMN latency (F2,96 = 1.83; p > 0.05; partial η2 = 0.04), the MMN amplitude successively decreased from the Young across Old-High to Old-Low group ( Fig. 5A). There was a significant effect of group (F2,96 = 3.44; p < 0.05; partial η2 = 0.07): The MMN was smaller in the Old-Low group than in the Young group (t96 = 2.61; p < 0.05; Fig. 5A), while the Old-High and Young groups (t96 = 1.06) and the 2 older groups (t96 = 1.52; both p > 0.05) did not differ from one another in MMN amplitude. Mean amplitudes (left) and latencies (right) of mismatch negativity (MMN) (A), ... Fig. 5. Mean amplitudes (left) and latencies (right) of mismatch negativity (MMN) (A), P3a (B), and reorienting negativity (RON) (C) at FCz to the rare deviant tones, shown for the younger and the older high-performing and low-performing groups. Error bars indicate standard errors across participants (younger: N = 35; older: N = 32). Figure options 3.3.3. P3a The frontal P3a peaked at 303 ms and was smaller in the Old-High group than in the Young and Old-Low groups (Fig. 5B). The ANOVA indicated a significant between-group effect on the P3a amplitude (F2,96 = 5.87; p < 0.005, partial η2 = 0.11), and significant contrasts occurred between the Old-High and Old-Low group (t96 = 3.51), and the Old-High and Young group (t96 = 3.08; both p < 0.005), but not Old-Low and Young group (t96 = 0.01; p > 0.05). There was also a significant between-group effect on P3a latency (F2,96 = 3.20; p < 0.05, partial η2 = 0.06): The Young group had a shorter P3a latency than the Old-Low group (t96 = 2.53; p < 0.05), whereas the Old-High and Young group (t96 = 1.25) and the 2 older groups (t96 = 1.25; both p > 0.05) did not differ from one another. 3.3.4. RON The RON peaked at 503 ms over fronto-central brain areas. There was a significant between-group effect on RON amplitude (F2,96 = 4.25; p < 0.05; partial η2 = 0.08), resulting from a more pronounced RON in the Young group, relative to the Old-High group (t96 = 2.67; p < 0.01) and the Old-Low group (t96 = 2.30; p < 0.05; Fig. 5C). The Old-High and Old-Low groups did not differ from one another (t96 = 0.36; p > 0.05). There was also a significant between-group effect on RON latency (F2,96 = 7.38; p < 0.005; partial η2 = 0.13): Relative to the Young group, the Old-High participants (t96 = 3.11) and the Old-Low participants (t96 = 3.46; both p < 0.005) showed a delayed RON, while the latter two groups did not differ from one another (t96 = 0.34; p > 0.05). In sum, the analysis of the ERPs to deviant tones revealed that the Old-Low group showed a decrease in MMN, but no reduction in P3a, relative to the Young group. In contrast, the Old-High group showed a less pronounced P3a than the Young and Old-Low groups. Both older groups showed a decreased and delayed RON relative to the Young group. 3.3.5. Correlations To investigate potential relationships between deviance-related ERPs and the sizes of the participants' individual distraction effects, correlations of MMN, P3a, and RON amplitudes at FCz and the percentage change in IE relative to the standard tones were computed for the older participants (averaged across the Old-High and Old-Low groups) and for the younger participants. For the older participants, there were significant correlations between MMN amplitude and change in IE (r = −0.25; p < 0.05), and between P3a amplitude and change in IE (r = 0.46; p < 0.001), but not between RON amplitude and change in IE (r = 0.03; p > 0.05). No such correlation was found for the young participants (all p > 0.05). Thus, a strong distraction effect (i.e., decline in performance in duration discrimination of deviant tones relative to the standard tones) of the older participants came along with a weak MMN and a strong P3a amplitude.