High-density EEG recordings during a passive oddball paradigm characterized basic auditory processing, sensory memory, and involuntary attention in a sample of adults with cystinosis. Although statistically significant differences were observed in the P2, MMN, and P3a amplitudes between the groups, these differences were quite subtle, and by-and-large adults with cystinosis presented highly similar neural responses compared to those of age-matched controls.
In the N1 time window, no detectable differences were found between groups, suggesting intact sensory transmission through the auditory system in individuals with cystinosis. Further indicative of intact basic auditory processing in cystinosis was the modulation of the N1 as a function of SOA (i.e. rate of presentation). This modulation, consistently described in the literature for the neurotypical population (38) as reflecting typical habituation (39, 40) and/or refractoriness (41, 42), was present in both control and cystinosis groups. These results were fully consistent with those we previously observed in children and adolescents with cystinosis (17), suggesting that basic auditory sensory-perceptual processing is maintained across development in this population and does not suffer from decline with age, in contrast to visual memory and learning (13). Thirteen out of the fifteen individuals with cystinosis tested received a kidney transplant and thus these findings are not due to the inclusion of a non-representative, atypically healthy adult sample of cystinosis.
In the P2 time window, individuals with cystinosis presented increased amplitudes when compared to their peers, which was not observed in children and adolescents in our earlier study (17). An enhanced P2 has been previously suggested in cystinosis in a proceedings paper (16), though in a sample with a significantly wider range of ages, and during a task focused on spatial selective attention. The authors argued that the larger amplitudes observed in cystinosis could be explained by anatomical factors such as a thinner skull, caused by renal osteodystrophy present in this population (16). The amplitudes of the electrical potentials recorded on the scalp surface are affected by the conductive properties of the volume between the cortex and scalp surface: A greater electrical resistance of the volume (i.e., a thicker skull) results in reduced EEG amplitudes; a decreased electrical resistance of volume (i.e., a thinner skull) results in increased EEG amplitudes (43–45). Skull thinness could thus be a potential explanation for the increased P2 in cystinosis. However, one would then expect a similar increase to be observed across all AEP sensory components, which was not the case in the current study. Such an increase should likewise be observed in children and adolescents, but no differences were found in P2 in our previous study testing children and adolescents.
Though often conceptualized in the context of the N1/P2 complex, P2 can be dissociated from the N1 experimentally, developmentally, and topographically (for a review, see (46)). The findings herein, suggesting no differences between the groups in the N1, but an increased P2 in cystinosis, are in accordance with a dissociation between these two components. The functional significance of the P2 is, however, poorly understood. In the context of oddball paradigms, P2 has been argued to reflect attention modulation (47), to index stimulus classification (whereby a stimulus is considered to be target or non-target) (48), and/or to be related to consolidation processes associated with auditory memory formation and learned relevance (49, 50). Increased P2 amplitudes, however, have been mainly associated with attentional processes. While some argue that such enhancement may reflect a deficit in the capacity to withdraw attentional resources from stimuli (48) or the capture of more attention (51), others suggest that differences in P2 amplitude may simply be an indication of the involvement of different processes underlying attentional disengagement (46). When compared to their age-matched peers, individuals with cystinosis might have engaged attention differently while passively listening to tones. The consequences of these potential differences warrant further investigation. An active oddball paradigm, for instance, would allow one to measure deviant detection rate and test whether those with increased P2 amplitudes are more or less accurate in detecting the deviant tones. Behaviorally, in standardized tasks and during behavioral observations, children and adolescents with cystinosis appear to present some level of impairment in the domain of attention (52–54), but no study has yet focused on this process during adulthood. Thorough behavioral and electrophysiological investigations of voluntary and involuntary attention in this population are thus warranted.
In the MMN time window, despite showing robust MMN responses across SOAs, individuals with cystinosis showed a slightly decreased response to the 1800 ms SOA when compared to their age-matched peers. In children and adolescents, such reductions were likewise observed, not only for the 1800 ms, but also for the 900 ms SOA, which we interpreted as a sign of reduced short-term auditory sensory memory (55, 56). Hence, such difficulties might still be present in adulthood, though less markedly so. Auditory sensory memory, a preattentive memory system that allows an individual to retain traces of sensory information after the termination of the original stimulus (57), could impact subsequent processing in working memory (30), a conscious cognitive system responsible for the temporary holding, processing, and manipulation of information (58). And, indeed, the MMN has been associated with performance in memory tasks in both neurotypical and clinical populations (30–32). Here, however, and despite the differences found between the groups in the longer SOA MMN, individuals with cystinosis performed similarly to controls on a standardized behavioral working memory task. Thus, this warrants further investigation of the neurophysiology of sensory and working memory processes in this population, and consideration of their consequences for perceptual and cognitive function.
One of the express purposes of employing three rates of presentation in the current study was to parametrically tax the auditory sensory memory system. That is, when the tones are presented at a rapid rate (> 2 Hz; 450 ms SOA), this rapid presentation establishes a strong auditory sensory memory trace for the standard tones such that the longer duration deviants strongly “pop out”. As the rate slows (~ 0.5 Hz; 1800 ms SOA), the sensory memory representation of the ongoing standards tends to weaken, and the general thesis is that the longer deviants are less perceptible at these slower rates (i.e., they do not pop out as much). As can been seen from the MMN data herein, even at the slowest rate, a robust MMN was generated for both groups. In contrast, in data collected from individuals with Rett Syndrome using the same paradigm, an MMN was only detected at the fastest rate and was completely absent at the two slower rates (56). Clearly then, this manipulation was not sufficient to assay potentially more modest differences in this clinically less affected group. The reader will also note that the duration deviant used here (180 ms vs. a 100 ms standard) is very large, and consequently highly perceptible. A potentially more sensitive test could involve the use of progressively less noticeable duration deviants, which would allow for more finely calibrated testing of these more subtle phenotypes in cystinosis (e.g. (59)). It is also worth noting that there are both hierarchically early sensory generators of the duration MMN, as well as higher-order cortical contributions from regions in the frontal and parietal lobes (60). Given the current findings of intact early sensory processing, as represented by the wholly intact N1, it is possible that subtle deficits seen in the duration MMN at the slowest rate here may be specifically attributable to deficits in these higher-order cortical generators. It will fall to future work to determine if this is the case, but such a finding would point to the emergence of subtle higher-order deficits rather than early sensory processing.
Lastly, in the P3a time window, and similar to what was observed for the P2, individuals with cystinosis presented increased amplitudes for the 900 and 1800 ms SOA conditions. There are several interpretations of the role of the P3 component (see (61)). One common view is that it represents context updating when stimulus events require that the model of the environment be revised (62). The value, significance, and relevance of the stimulus dictate the extent to which the updating process is activated (63). Another view, and perhaps a better alternative in the context of passive paradigms, is that P3 is associated with event categorization: Each stimulus is evaluated and categorized as expected, unexpected, related to the task, etc., but with little controlled processing involved (64). A larger P3 in cystinosis could then be suggestive of an easier categorization process (64). One could argue that if attentional resources were allocated differently in cystinosis in earlier stages of information processing (which the increased P2 in this population could be indicative of) (65), the later categorization of the stimulus could have been facilitated in this group. An active oddball paradigm with behavioral outcomes will be useful in understanding the impact of the increased P3 on task performance.
In summary, this study provides the first neural evidence of auditory sensory memory and involuntary attention differences in adults with cystinosis and pinpoints areas that warrant more research. It is nevertheless important to stress that despite the differences found between the groups, the adults with cystinosis tested here presented overall similar neural responses to controls. Larger samples would have allowed for more detailed analyses focused on the associations between neural, cognitive, and behavioral outcomes. An active oddball paradigm would have been particularly useful in understanding the impact of the P2 and P3a amplitude differences in the successful detection of deviants.