Sequential processing in young and older adults in the equiprobable auditory Go/NoGo task

Abstract

Objective: We recently proposed a sequential processing schema for the equiprobable auditory Go/NoGo task, based on a principal components analysis (PCA) of event-related potentials (ERPs) from a university student sample. Here we sought to replicate the schema, and use it to explore processing in well-functioning older adults. Methods: We compared behavioural responding and ERPs of 20 independent-living older adults (M age = 68.2 years) to data from a sex-and handedness-matched group of university students (M age = 20.4 years). ERPs had substantial latency differences between the groups, and hence were subjected to separate group temporal PCAs. Results: Component latencies were systematically increased in the older group by some 26%, with no significant increase in RT or error rates. Despite some differences in their identified components, each group displayed differential component responsivity to Go versus NoGo; this was reduced in the older participants. Conclusion: The results support our processing schema, and provide insight into the processing stages in well-functioning older adults. Significance: Understanding the perceptual and cognitive processing stages in normal ageing is a pre-requisite for research on mild cognitive impairment and dementia. This study may also provide a simple paradigm and schema suitable for further exploration of functionality in ageing. Highlights• A PCA derived sequential processing Go/NoGo schema was confirmed in young adults, with similar processing in healthy older adults.• Older adults had comparable Go RT and Go/NoGo error rates, increased latencies and smaller, less differentiated Go/NoGo P3s, and N2c was absent.• Go/NoGo processing by Older adults may reflect less specialist processing regions, requiring more effort. ABSTRACT Objective:We recently proposed a sequential processing schema for the equiprobable auditory Go/NoGo task, based on a principal components analysis (PCA) of event-related potentials (ERPs) from a university student sample. Here we sought to replicate the schema, and use it to explore processing in well-functioning older adults. Methods:We compared behavioural responding and ERPs of 20 independent-living older adults (M age = 68.2 years) to data from a sex-and handedness-matched group of university students (M age = 20.4 years). ERPs had substantial latency differences between the groups, and hence were subjected to separate group temporal PCAs.Results: Component latencies were systematically increased in the older group by some 26 %, with no significant increase in RT or error rates. Despite some differences in their identified components, each group displayed differential component responsivity to Go versus NoGo; this was reduced in the older participants. Conclusion:The results support our processing schema, and provide insight into the processing stages in well-functioning older adults.Significance: Understanding the perceptual and cognitive processing stages in normal ageing is a pre-requisite for research on mild cognitive impairment a...

ERP Go/NoGo condition effects are better detected with separate PCAs

“…In temporal order, based on their latency, scalp topography, and similarity to components in previous fixed SOA studies De Blasio, 2013, 2015;Barry et al, 2014aBarry et al, , 2014bBarry et al, , 2014cBarry et al, , 2016, these components were identified as N1-1, PN, a P2/N2b complex, N2c, P3, SW, and LP. Fig.…”

“…We expected that a Combined PCA, following our standard procedure and inputting ERPs from both Go and NoGo conditions, would replicate the previous young adult schema studies (Barry and De Blasio, 2013;Barry et al, 2014aBarry et al, , 2014bBarry et al, , 2014cBarry et al, , 2016. In contrast, we expected Separate PCAs, conducted independently on each of the Go and NoGo ERP sets, to yield components similar to those identified in the processing schema, but with latency shifts that reflected differences such as those apparent between P3a and P3b in the raw ERPs.…”

“…Further, we found that even when PCA separation does occur, substantial misallocation of variance between the derived subcomponents may remain. This suggests that even when P3a and P3b were separated by the combined PCAs in our child and older-adult studies Barry et al, 2014aBarry et al, , 2016, the obtained components may not have been optimal, but rather, were probably contaminated by misallocated variance. The use of separate PCAs has been shown here to adequately recover the different prototypes, and this provides a simple solution to avoid these problems with minimal corruption from component latency jitter.…”

“…With minor conceptual updating (see Introduction of Barry and De Blasio, 2015), this has been found to apply substantially across the lifespan, from children Barry et al, 2014a), through young adults (Barry and De Blasio, 2013;Barry et al, 2014aBarry et al, , 2014bBarry et al, , 2014cBorchard et al, 2015), to healthy older adults (Barry et al, 2016), covering an age range from 8 to 75 years. In this regard, the simple paradigm and processing schema may provide generalpurpose tools to probe the development of cognitive functioning in different populations and clinical disorders, and also to explore pharmacological effects (e.g., caffeine: Barry et al, 2014b).…”

“…This is likely to be close to the truth for early components (such as N1-1, the dominant N1 subcomponent), which may not show obvious condition separation in the raw ERPs. But it is unlikely to be a good approximation for the P3a and P3b subcomponents of the P3/P300/Late Positive Complex (LPC) in the Go/NoGo task, which usually appear separated in timing in the raw ERPs (with P3a typically occurring before P3b), but are extracted together in one PCA component in young adult samples (Barry and De Blasio, 2013;Barry et al, 2014aBarry et al, , 2014bBarry et al, , 2014cBarry et al, , 2016. Interestingly, we have found in the same paradigm that the P3a and P3b were extracted as different PCA components (with P3a before P3b) in children (Barry et al, 2014a;Barry and De Blasio, 2015) and older adults (Barry et al, 2016).…”