Understanding age-related differences in the speed of information processing of complex object categories measured with electroencephalography (EEG)

Ageing is associated with differences in visual function which can be observed, for example, as a decline in performance on face and object processing tasks. One of the most prominent accounts of age-related decrement in perceptual and cognitive tasks alike is that of a reduction in information processing speed (Salthouse, Psychological Review 1996, 103:403). Differences in myelin integrity in some parts of the cortex, as well as in neuronal responsivity are physiologically plausible as the origins of the age- related slowing-down of information processing. However, little research to date has directly investigated age-related slowing-down of visual information processing in humans. Previously, Rousselet et al. (Frontiers in Psychology 2010, 1:19) reported a 1ms/year delay in face visual processing speed in a sample of 62 subjects aged ~20-80, using event-related potentials (ERPs). This result was replicated in another 59 subjects, and was independent of stimulus luminance and senile miosis (Bieniek et al. Frontiers in Psychology 2013, 4:268). To go beyond differences in average brain activity and interpret previous findings, in the first study (Chapter 2) we investigated what information is coded by early face ERPs in younger and older observers. In a detection task, young and older observers each categorized 2,200 pictures of faces and noise textures revealed through Gaussian apertures (“Bubbles”). Using reverse correlation and Mutual Information (MI), we found that the presence of the left eye elicited fastest detection in both age groups. Older observers relied more on the eyes to be accurate, suggesting a strategy difference between groups. In both age groups, the presence of the eye contralateral to the recording electrode modulated single-trial ERPs at lateral-occipital electrodes, but this association was weaker in older observers and delayed by about 40 ms. We also observed a differentiated coding of the eyes across groups: in younger observers, both the N170 latency and amplitude coded the contralateral eye, whereas it was only the N170 amplitude in older adults. The latency modulation in younger adults was also higher in the right than in the left hemisphere, but very similar across hemispheres in older adults. Our results suggest that face detection in ageing is associated with delayed and weaker processing of the same face features, and point to potential coding differences. On the notion that incomplete or occluded stimuli (such as Bubbled images) might differentially affect older adults’ ability to perform a perceptual task, in the second study (Chapter 3) we sought to understand whether the age-related differences in eye sensitivity were preserved in a face context. Two groups of observers, young and older, performed a face detection task in which the visibility of the eye region was modulated in a parametric manner by adding phase noise. This way, we could investigate the modulation of ERPs by increasing information available in the eye region, when the face context was preserved (or absent – in control conditions). In line with behavioural results reported in Chapter 2, modulating the visibility of the left eye had a greater effect on reaction times across older participants, and this modulation increased with decreasing face context information in older adults. Contralateral eye sensitivity was weaker than that reported in Chapter 2 and did not differ between young and older observers, suggesting that coding of the eye by the N170 acts differently when the eye is revealed through Bubble masks and when it is presented in the face context. In Chapter 4, we investigated potential origins of the large N170 responses to textures observed in a sample of older participants before (Rousselet et al. BMC Neuroscience 2009, 10:114), and quantified age-related delays in visual processing speed of stimuli other than faces: houses and letters. Two groups of participants performed three simple detection tasks: face detection, house detection, and letter detection. Perceiving textures in the context of a face detection task, but not house detection or letter detection, influenced ERP responses to textures in older participants only to a small extent and after 200 ms post- stimulus, suggesting that the large N170 responses to textures are unlikely due to a top- down influence of the task at hand. Furthermore, visual processing speed of faces, houses and letters was delayed to a smaller extent than that predicted by the original study and depended on the nature of categorical comparisons made. Overall, our results fill the big gap in the literature concerned with age-related slowing of information processing: using Bubbles, we have presented direct evidence that processing of the same facial information is slower (and weaker) in ageing. However, quantifying visual processing speed using categorical designs yielded mixed evidence for the theory of slower information processing in ageing, pointing to the need for carefully designed visual stimuli in ageing research, and for careful selection of control stimuli for comparisons