Double peaked P1 visual evoked potentials in healthy ageing

Clinical Neurophysiology xxx (2013) xxx–xxx

Contents lists available at ScienceDirect

Clinical Neurophysiology journal homepage: www.elsevier.com/locate/clinph

Double peaked P1 visual evoked potentials in healthy ageing G. Stothart ⇑, A. Tales, C. Hedge, N. Kazanina School of Experimental Psychology, University of Bristol, 12a Priory Road, Bristol BS8 1TU, UK

a r t i c l e

i n f o

Article history: Accepted 15 November 2013 Available online xxxx Keywords: Visual evoked potentials Ageing Double peak P1 Gamma Compensation

h i g h l i g h t s  The P1 visual evoked potential is significantly affected by healthy ageing.  Older adults show reduced P1 amplitudes and in many cases a change in morphology to a double

peak.  Double peaked P1 evoked potentials may be more prevalent than previously estimated in healthy

older adults, with implications for both clinical and experimental research.

a b s t r a c t Objectives: To robustly examine the prevalence of the double peaked P1 visual evoked potential in healthy younger and older adult populations. Methods: The evoked potentials and spectral power changes to simple visual stimuli of 26 healthy younger (M = 20.0 y) and 26 healthy older adults (M = 76.0 y) were examined. Results: Group and individual analyses showed a clear effect of age on P1 morphology and amplitude. Older adults showed significantly lower P1 amplitude and 44% of older adults showed a double peaked P1 compared to 12% of younger adults. Double peaked P1 responses were associated with an increase in spectral power in the gamma range. Conclusions: The double peaked P1 may be more prevalent in older adults than previously demonstrated and may represent a de-synchronisation of the cortical sources of the visual P1 in healthy ageing. Increased power in post stimulus gamma in the double peak group may be indicative of compensatory neural processing. Significance: Clinically the prevalence of the double peaked P1 may have been underestimated, and its reflectance of demyelinating disease overestimated. Experimentally the results suggest that any investigation of visual processing in older adults must control for early changes in P1 morphology. Ó 2013 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved.

1. Introduction Vision is a primary sense for human beings, and one that is susceptible to many age-related changes. Whilst the tools for quantifying and measuring such changes in ocular and optic tract function prior to the cortex are well established and readily available, tools for assessing age-related changes at the cortical level of visual processing are scarcer. Visual evoked potentials (VEPs) provide the opportunity to measure these cortical processes. Cognitive theories of ageing have been based largely on evidence from behavioural responses to a wide variety of cognitive tasks. Neurophysiological tools, such as VEPs, can further inform these theories. They can help to examine the effects of age on early processing and provide a more complete picture of visual ⇑ Corresponding author. Tel.: +44 177 928 8450; fax: +44 117 9288588. E-mail address: [email protected] (G. Stothart).

information processing, from stimulus onset to behavioural response. With a greater understanding of the changes that occur in healthy ageing, the same tools have the potential to be used to examine pathological ageing. One of the major components of interest in the typical VEP complex is the visual P1 response, commonly elicited either using chequerboard pattern-reversal or flash stimuli. Peaking between 90 and 130 ms, with neural generators in the ventral–lateral extrastriate cortex (Brodmann area 19) (Clark et al., 1994; Clark and Hillyard, 1996; Di Russo et al., 2002; Proverbio et al., 2007), the P1 is considered to represent the processing of stimulus characteristics and visuo–spatial selection. How the P1 changes in healthy ageing is unclear as the P1 is often neglected in analyses with the focus on later components such as the N1 and P3. Studies that have examined the P1 have reported conflicting findings. Some report increases in the amplitude and latency, and/or changes in the morphology of P1 with healthy ageing

1388-2457/$36.00 Ó 2013 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved. http://dx.doi.org/10.1016/j.clinph.2013.11.029

Please cite this article in press as: Stothart G et al. Double peaked P1 visual evoked potentials in healthy ageing. Clin Neurophysiol (2013), http://dx.doi.org/ 10.1016/j.clinph.2013.11.029

2

G. Stothart et al. / Clinical Neurophysiology xxx (2013) xxx–xxx

(Sokol et al., 1981; Amenedo and Diaz, 1998; Yordanova et al., 2004; Falkenstein et al., 2006; Ceponiene et al., 2008; Plomp et al., 2012), while others report no change (Pfütze et al., 2002; Czigler and Balázs, 2005; Gazzaley et al., 2008). Consequently, with a wide variety of stimuli and experimental designs contributing to the variability of findings, a consensus on the neurophysiological impact of healthy ageing on the early stages of visual processing is yet to emerge. VEPs are also commonly used clinically for the characterisation and diagnoses of a variety of pathologies. Changes in the morphology and latency of the P1 are reported to be symptomatic of a wide variety of disorders including glaucoma (Parisi et al., 2006), retinitis pigmentosa (Janáky et al., 2007), optic neuropathy (Tsaloumas et al., 1994), multiple sclerosis and other demyelinating disease (Brigell et al., 1994; McDonald et al., 2001; Rousseff et al., 2005). The specific change in P1 morphology from a single peak (typical of healthy populations) to a double peak (also referred to as ‘‘bifid’’ or ‘‘w’’ shaped) is thought to be symptomatic of demyelinating disease. The reported rates of double peak prevalence within groups showing evidence of demyelinating disease ranged from 5% to 22% (Jones and Blume, 1985; Marra, 1990). In healthy young individuals the double peaked P1 is rarely observed, i.e. in 1–6% of cases (Blumhardt et al., 1980; Marra, 1990) and is attributed to individual variation in brain anatomy, in particular, the orientation of key neuronal populations in the extrastriate cortex. It can be temporarily induced in healthy populations through the administration of acetylcholine antagonists (Bajalan et al., 1986) and/or the occlusion of the central visual field (Brecelj et al., 1990) and abated with benzodiazepine antagonists (Aguglia et al., 1993). It has been reported that as the double peaked P1 is so rarely observed in healthy populations the presence of a double peaked P1 ‘‘should suggest the possibility of demyelinating disease and prompt further investigation and follow-up’’ (Rousseff et al., 2005). In a recent study examining the effects of age on visual mismatch negativity (see Stothart et al., 2013) we serendipitously observed that a significantly higher proportion of healthy older adults displayed a double peaked P1 to visual stimuli than would be expected according to the current literature. By considerably extending the sample of the original study, adding time frequency and behavioural analyses, we were able to comprehensively examine the double peaked P1. Therefore the aims of the current study were to present our preliminary findings on the prevalence and neurophysiological characteristics of the double peaked P1 in younger and healthy older adult populations, with a view to prompting future robust examination of this effect.

2. Method 2.1. Participants Twenty-six healthy younger adults (aged 18–31, mean age 20.3 (±2.6), 13 males) and 26 healthy older adults (aged 62–88, mean age 76.0 (±7.0), 14 males) gave consent to participate in the study. Data from seventeen participants from both the younger and older adult group has previously been published in a paper examining visual mismatch negativity (see Stothart et al., 2013). Younger adults were recruited from the University of Bristol student population and declared themselves to be in normal health, right hand dominant and with normal or corrected to normal vision. Older adults were recruited by the Avon and Wiltshire and South Gloucestershire Primary Care Trust memory service clinics at the Bristol Research into Alzheimer’s and Care of the Elderly Centre, Frenchay Hospital, and the Research Institute for the Care of Elderly People, Royal United Hospital, Bath. They were comprehensively assessed by clinicians at the memory clinics upon enrolment

onto a volunteer database within 12 months prior to taking part in the current study, in which they participated as part of a wider study into dementia as healthy controls. They displayed normal cognitive function in relation to their age (mean Mini-Mental State Examination score 28.2 (±1.3) (Folstein et al., 1975)) and educational attainment and none met criteria for dementia or any other neuropsychological disorder. No older adults had history or signs of stroke or transient ischemic attack, significant head injury, depression, or other psychiatric disorder, or major neurological disease and none were receiving medication (prescribed or nonprescribed) deemed likely to affect cognitive function. All had normal or corrected-to-normal vision, as assessed at the memory clinics, and were right hand dominant. Any participants who reported a history of major ocular health problems, such as glaucoma or macular degeneration, were excluded. 2.2. Stimuli Stimuli were presented using Presentation software version 12.2 (Neurobehavioral Systems, Inc). The fixation point was a blue square (visual angle = 1.5°  1.5°) at the centre of a monitor situated 0.5 m directly in front of them (see Fig. 1a) that remained on screen throughout the experiment. The standard stimuli were single vertical white bars (visual angle = 4.46°  1.5°, see Fig 1b) appearing above and below the fixation point. Deviant stimuli were double white bars equal to the standards in total area (visual angle = 4.46°  0.75°, see Fig 1c), and the target stimulus was a red square appearing at the central fixation point (visual angle = 1.5°  1.5°, see Fig 1d). As explained below and unless explicitly noted otherwise, only standard stimuli were used for the analysis. 2.3. Procedure Participants completed a visual oddball paradigm designed to elicit a visual mismatch negativity. Unless explicitly stated otherwise, for the purpose of this report only the neurophysiological responses to standard stimuli will be considered (the experimental procedure of the original study is described in full in Stothart et al., 2013). Deviant stimuli were rare, 40 presentations per participant, and averaged responses had a low signal to noise ratio. Target stimuli were also rare, 40 presentations per participant, and were found to elicit a very small P1 (c 0.5 uV) across all participants, additionally evoked responses may have been confounded by the target’s attentional salience. Standard stimuli however were the most numerous, the least physically complex and therefore the most appropriate for further analysis. Participants were instructed to fixate and attend exclusively to the fixation point, the standard stimuli were presented simultaneously above and below the central fixation point. Periodically, the target and deviant stimuli were pseudo-randomly presented. Participants were instructed to focus solely on the target stimuli and respond to it as quickly as possible by pressing a hand-held button whilst ignoring standard and deviant stimuli. The stimuli were presented with a randomised inter-stimulus interval (ISI) of 612–642 ms for 200 ms. The stimuli were shown in one block lasting 11 min containing 640 standards, 40 deviants and 40 targets, a ratio of 16:1:1. 2.4. EEG recording EEG signals were continuously recorded from 64 Ag/AgCl electrodes fitted on an elasticised cap in a standard electrode layout using a common FCz reference, this reference was maintained for all analyses unless explicitly stated otherwise. Signals were sampled at a rate of 1000 Hz using a BrainAmp DC amplifier (Brain

Please cite this article in press as: Stothart G et al. Double peaked P1 visual evoked potentials in healthy ageing. Clin Neurophysiol (2013), http://dx.doi.org/ 10.1016/j.clinph.2013.11.029

G. Stothart et al. / Clinical Neurophysiology xxx (2013) xxx–xxx

3

Fig. 1. Stimuli used to elicit visual evoked potentials (a) inter-stimulus screen; (b) standard; (c) deviant; (d) target. Unless otherwise noted, the analyses in the current study are based only on standard stimuli.

Fig. 2. Grand average responses to standard stimuli at the occipital region of interest (mean value of electrodes O1, Oz, O2, PO9, PO10, PO7 & PO8), for younger and older adults. The topographic map illustrates the location of the region of interest electrodes.

Products GmbH). Impedances were kept below 5 kO and all signals were online low-pass filtered at 250 Hz during recording. Recordings were analysed offline using Brain Electrical Source Analysis software version 5.3 (BESA GmbH). Artifacts including blinks and eye movements were corrected using BESA automatic artifact correction (Berg and Scherg, 1994) and any remaining epochs containing artifact signals > ± 100uV were rejected. The mean number of epochs comprising individual averages was 594 ± 27 (range 553–635) for younger adults and 599 ± 28 (range 557– 639) for older adults. 2.5. Electrode selection and VEP analysis Seven electrodes, O1,Oz,O2,PO9,PO10,PO7 & PO8, were selected for statistical analysis on the basis of existing reports on early visual evoked potentials, their values low pass filtered at 40 Hz, and averaged to form an occipital region of interest, see Fig. 2. Epochs from 100 to 500 ms were defined around stimulus onset, baseline corrected using the pre-stimulus interval ( 100 to 0 ms), and averaged. Examination of the grand average evoked responses re-referenced to a common average reference confirmed that the electrode selection was appropriate and that neural activity was highly consistent across the seven electrodes (see Supplementary Fig. S1 for 64 channel plots). Based on inspection of the group

grand average waveforms P1 latency was measured as the time of the largest positive peak occurring during the period of 50– 200 ms post stimulus onset and P1 amplitude as the mean amplitude during an epoch defined by one standard deviation around the mean peak latency. E.g. if for a given group the mean peak latency was 100 ms with a standard deviation of 10 ms, the epoch used to calculate the mean amplitude of the P1 peak for that group would be 90–110 ms. The effect of age on mean P1 amplitude was examined in a 1-way (young/old) ANOVA. For individuals with a double peaked P1 the latencies of the early and late peaks were measured as the largest positive peak in the epochs 70–120 ms and 120–170 ms post-stimulus onset respectively and the amplitudes as the mean amplitude during an epoch defined by one standard deviation around the mean peak latency. For individuals with an ambiguous P1 peak morphology latency and amplitude was calculated using both the single and double peak criteria. 2.6. P1 peak morphology classification Individual P1 evoked potential plots were anonymised and then presented to two researchers experienced in ERP techniques and naive to the purpose of the classification. They classified each peak as ‘‘single peak’’ ‘‘ambiguous’’ or ‘‘double peak’’. An inter-rater reliability analysis using the Kappa statistic showed a high level of

Please cite this article in press as: Stothart G et al. Double peaked P1 visual evoked potentials in healthy ageing. Clin Neurophysiol (2013), http://dx.doi.org/ 10.1016/j.clinph.2013.11.029

4

G. Stothart et al. / Clinical Neurophysiology xxx (2013) xxx–xxx

consistency between the classifications, Kappa = 0.76 (p < 0.001). A combined scale was calculated by reclassifying any discrepancies between the two ratings as either ‘‘ambiguous’’ in the case of diametrically opposite classifications or as either ‘‘single peak’’ or ‘‘double peak’’ when only one rater’s classification was ‘‘ambiguous’’. A chi-square test was performed to test for significant differences in classifications between younger and older adults. 2.7. Behavioural analysis In order to establish if age and/or P1 peak morphology affected task performance, median reaction time to targets and the percentage of targets detected were calculated. These were then examined in a 2  2 group (young/old)  peak (single/double) ANOVA. Participants with a peak morphology classified as ambiguous (n = 5) were excluded from this analysis. 2.8. Time frequency analysis In order to further characterise the double and single peaked P1 responses in the older adult participants, original epochs (i.e. on a trial-by-trial basis and not subjected to any offline filtering) from all 64 electrodes were re-referenced to a common average reference and transformed into the time–frequency domain using a complex demodulation approach, implemented by BESA Source Coherence module version 5.3 (Hoechstetter et al., 2004) and analysed using BESA Statistics v1.0 (BESA GmbH). Complex demodulation was applied using a sampling step of 2 Hz for frequencies between 4 and 50 Hz and a finite impulse response filter with a sampling step of 25 ms at latencies between 200 ms and 500 ms relative to the stimulus onset. A longer baseline was used than in the VEP analysis in order to ensure the validity of time–frequency data at low frequencies. Changes in spectral power were calculated relative to pre-stimulus baseline ( 200 to 0 ms) in two different ways. ‘Overall spectral power’ was calculated by averaging spectral content of all individual demodulated epochs and represented both stimulus phase locked and non-phase locked activity. ‘Non-phase locked spectral power’ was calculated by subtracting the participant’s average response in the time–frequency domain from the spectral content of each demodulated epoch, and then averaging the remaining activity (which was specifically non-phase locked to the stimulus). Differences in spectral power between single and double peakers from the older group were analysed using a non-parametric cluster based permutation approach using Fieldtrip software (Oostenveld et al., 2011). This approach, described by Maris and Oostenveld (2007), controls for multiple comparison testing when computing statistics across multiple frequency and time points. Firstly an independent samples t-test between the single and double peaked P1 older adult groups was calculated for each sample point. Significant values (alpha 6 0.01) were clustered based on their adjacency in time and frequency, and the t-values for all points in this cluster were summed. The critical p-value for each cluster was calculated using the Monte Carlo estimate. For each cluster this involved randomly dividing the data into two subsets and calculating a new summed t-value. This was repeated 10,000 times and the proportion of random partitions that resulted in a larger test statistic than the one observed in the real data identified. If the summed t-value of the observed data cluster was higher than 95% of the random partitions (i.e. less than an alpha-level of 0.05, two-tailed), then the cluster was considered to represent a significant difference between the two groups. The technique allows for the evolution of spectral activity across time to be observed without the need for reductive averaging across arbitrary time windows or grouping of frequencies into bands. It should be noted that the initial alpha value for cluster formation was lowered from alpha <0.05 to alpha <0.01 in order

to reduce the likelihood of large clusters spanning the entire dataset, a potential problem in cluster based permutation testing highlighted recently by Mensen and Khatami (2013). 3. Results Younger and older adults both displayed a clear P1 in response to the standard stimuli, see Fig. 2. As discussed below, the P1 component had two peaks rather than the single peak observed in the younger adults and its amplitude was lower in older adults. The early differences in P1 amplitude affected the subsequent evoked responses until 350–400 ms. That these later differences were indeed consequences of the initial difference in P1 amplitude is illustrated by the significant positive correlation between P1 and N1 amplitudes (Pearson’s r (52) = 0.321, p = 0.02), i.e. the smaller the P1 amplitude the greater (more negative) the subsequent N1 amplitude. Full-head 64-electrode VEP plots referenced to the original FCz reference and re-referenced to an average reference are provided in Supplementary Fig. S1. The average referenced plot clearly demonstrates that the majority of neural activity was located in the occipital region. A frontal negative counterpart of occipital P1 can be observed somewhat weakly in both plots, however this is likely to be an artefact of the (re-)referencing process. 3.1. P1 peak morphology classifications and topographies Twenty-three younger adults (88%) and 10 older adults (38%) had single peaks, 11 older adults (42%) and three younger adults (12%) had double peaks, five older adults (20%) had ambiguous peaks, see Figs. 3 and 4. A chi-square test indicated that the proportion of single vs double peakers was significantly different between younger and older adults, X2(2) = 14.69, p < 0.001, uC = 0.53. As shown in Fig. 5 and 11 older adults who had a double peaked P1 response to standard stimuli, also showed a double peaked P1 to deviant and target stimuli. 3.2. P1 peak characteristics P1 amplitude was significantly lower in older adults than younger adults across morphology type (see Table 1), as confirmed by a significant main effect of age on P1 amplitude (F 1, 51 = 14.0, p < 0.001, g2 = 0.21). Older adults with a single peaked P1 showed significantly smaller amplitudes than younger adults with a single peaked P1, (F 1,32 = 15.5, p < 0.001, g2 = 0.33). The number younger adults with a double peaked P1 was too low (n = 3) to meaningfully compare the mean amplitude of P1a and P1b between younger and older adults. Older adults with a double peaked P1 had a greater group mean age than older adults with a single peaked P1 (see Table 1). Fig. 3 illustrates the topographic voltage distribution of the P1 peaks across group and morphology types. Peak amplitude for both single and double peaked P1 VEPs is consistently concentrated within the occipital region of interest used for analyses and no differences were observed between younger and older adults. 3.3. Behavioural analysis There was no effect of P1 morphology type (Single peak: mean = 395 ms (±50), Double peak: mean = 396 ms (±38), F 1,46 = 0.21, p = 0.651) or group (Young: mean = 392 ms (±54), Old: mean = 398 ms (±34), F 1,46 = 0.88, p = 0.352) on median reaction time to target stimuli and no morphology type x group interaction. There was no effect of P1 morphology type (Single peak: mean = 97.6% (±=3.3), Double peak: mean = 98.6% (±1.89), F 1,46 = 0.83,

Please cite this article in press as: Stothart G et al. Double peaked P1 visual evoked potentials in healthy ageing. Clin Neurophysiol (2013), http://dx.doi.org/ 10.1016/j.clinph.2013.11.029

G. Stothart et al. / Clinical Neurophysiology xxx (2013) xxx–xxx

5

Fig. 3. Individual VEPs to standard stimuli at the occipital region of interest (mean value of electrodes O1, Oz, O2, PO9, PO10, PO7 & PO8), split by P1 peak morphology classification. For each morphology group topographic maps illustrate the mean voltage distribution for the P1, P1a or P1b peak, re-referenced to a common average reference.

Fig. 4. P1 peak morphology type as determined by expert classification.

p = 0.367) or group (Young: mean = 97.7% (±3.5), Old: mean = 98.2% (±2.3), F 1,46 = 0.13, p = 0.910) on the percentage of targets detected and no morphology type x group interaction.

spectral power. These differences between single- vs doublepeaked were confirmed by the results of cluster based permutation analysis that revealed a significant increase between 300 and 400 ms in overall spectral power in the typical gamma range, i.e. 25–50 Hz, in older adults with a double peaked P1 compared to older adults with a single peaked P1; Monte-Carlo p = 0.022, see Fig. 6. This difference was greatest at central, right hemisphere electrode sites; see Supplementary Fig. S2 for 64 electrode plots. The analysis of non-phase locked spectral power change also revealed a significant increase of power in the typical gamma range, i.e. 25–50 Hz, in older adults with a double peaked P1 compared to older adults with a single peaked P1; Monte-Carlo p = 0.017, see Fig. 6. The difference was again greatest at central, right hemisphere electrode sites; see Supplementary Fig. S2 for 64 electrode plots. The maintenance of the difference in gamma power change in the non-phase locked analysis suggested that it largely resulted from non-phase locked oscillatory change. There were no significant differences between single and double-peakers in gamma power in other time intervals (including the interval around P1) or in the lower frequencies, i.e. theta, alpha and beta ranges, in either the overall or non-phase locked analyses.

3.4. Time frequency analysis 4. Discussion In order to further characterise the double and single peaked P1 responses in older adults responses to standard stimuli were transformed into the time–frequency domain. As shown in Fig. 6, there was an increase in overall and non-phase locked spectral power with a maximum around 100–200 ms post-stimulus for both single- and double-peakers, corresponding to the P100 component in the temporal domain. In single-peakers (but not double-peakers), this early effect was followed by a prominent decrease in (overall or non-phase locked)

Healthy older adults displayed lower P1 amplitudes than younger adults and a significant proportion were found to display a double peaked P1 VEP. Younger adults typically displayed a single peaked P1 VEP. Classification of individual participants’ P1 morphology demonstrated that a significantly higher proportion of older adults displayed a double peaked P1 compared to younger adults. The proportion of older adults who displayed a double peaked P1 was also considerably higher than has previously been

Please cite this article in press as: Stothart G et al. Double peaked P1 visual evoked potentials in healthy ageing. Clin Neurophysiol (2013), http://dx.doi.org/ 10.1016/j.clinph.2013.11.029

6

G. Stothart et al. / Clinical Neurophysiology xxx (2013) xxx–xxx

Fig. 5. Grand average waveforms to standard, deviant and target stimuli for older adults who displayed a double peaked P1 to standard stimuli (n = 11) at the occipital region of interest (mean value of electrodes O1, Oz, O2, PO9, PO10, PO7 & PO8).

Table 1 Peak latencies (ms) and mean (SD) amplitudes (lV) of P1, P1a and P1b at the occipital region of interest (mean value of electrodes O1, Oz, O2, PO9, PO10, PO7 & PO8) in younger and older adults, split by P1 peak morphology classification. Morphology

Group

Age

n

Peak characteristics P1

Single peak Double peak Ambiguous

Young Old Young Old Young Old

20.3 71.7 20.3 76.8

(2.7) (6.0) (2.1) (5.7)

83.0 (6.3)

23 10 3 11 0 5

P1a

P1b

Peak latency

Mean amplitude

Peak latency

Mean amplitude

Peak latency

Mean amplitude

123.1 (9.2) 123.5 (22.8) – – – 134.9 (13.5)

4.4 (2.2 1.4 (1.2) – – – 1.7 (1.2)

– – 113.1 (0.9) 105.4 (10.3) – 106.0 (13.6)

– – 0.8 (1.0) 2.2 (1.3) – 1.2 (0.7)

– – 136.9 (0.3) 140.1 (10.2) – 141.7 (7.7)

– – 1.6 (1.3) 2.5 (1.4) – 2.1 (1.6)

Fig. 6. Time–frequency plots of percentage change in overall and non-phase locked spectral power at electrode C6 compared to baseline ( 200–0 ms). A – Older adults with a double peaked P1. B – Older adults with a single peaked P1. C – Difference plot (A minus B). Significant differences, p < 0.05 after Monte Carlo permutation correction for multiple comparisons, are masked in green. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

demonstrated in healthy populations. Older adults who exhibited a double-peaked P1 to standard stimuli, also showed a doublepeaked response to deviant and target stimuli, demonstrating that it generalised across conditions. Older adults who displayed a double peaked P1 showed significantly greater overall and nonphase locked spectral power in the gamma range following P1, i.e. 300–400 ms post-stimulus.

We propose that the double peaked P1 VEP may reflect a slowing of cortico-cortical transfer in older adults and essentially an ‘‘unpacking’’ of the different processes that occur with sufficient temporal proximity in younger adults to coalesce into a single peak. In Di Russo et al. (2002, 2005) exhaustive dipole modelling of the P1 in younger adults there emerged up to six distinct sources of the underlying neural activity, with distinct separate dipoles

Please cite this article in press as: Stothart G et al. Double peaked P1 visual evoked potentials in healthy ageing. Clin Neurophysiol (2013), http://dx.doi.org/ 10.1016/j.clinph.2013.11.029

G. Stothart et al. / Clinical Neurophysiology xxx (2013) xxx–xxx

responsible for an ‘‘early 80–110 ms’’ and ‘‘late 110–140 ms’’ P1. Despite the separate dipole sources the P1 VEP appears as a single peak in the participants’ grand averages. A temporal de-synchronisation of these multiple sources would result in a double peaked evoked potential, an explanation also proposed by Aguglia et al. (1993) in a study examining the role of benzodiazepine receptors in the geniculo–striate visual pathways. The timing of double peaks observed in our study accords well with the early and late ranges in Di Russo et al. study (P1a: 113 & 105 ms, P1b: 137 & 140 ms for younger and older double-peakers, respectively, see Table 1) and suggests that P1a and P1b are each produced by a distinct subset of dipoles which normally yield a single peaked P1. Furthermore, that P1a and P1b latencies fall at the opposite sides of the early and late ranges respectively is a likely reason why double peaks were not masked in our VEP data, unlike in Di Russo et al. study. Another possible explanation is the recruitment of a separate neural source in older adults that elicits a negative peak that ‘‘interrupts’’ the single P1 peak typically observed in younger adults. However, the negative trough in the double peak P1 occurs at approximately the same time as the positive peak in a typical single peak P1 occurs, i.e. circa 120 ms. If the negative trough represented a separate neural source to those generating the positive P1, a more likely consequence would be a reduction in the amplitude of the P1 or a cancelling out of the peak altogether. The increase in gamma oscillatory activity associated with the double peaked P1 may reflect a compensatory response to impaired initial sensory processing. Gamma oscillatory activity has been shown to be associated with the binding of perceptual features and object construction (Herrmann et al., 2010; TallonBaudry and Bertrand, 1999). If the assumption is taken that the double peaked P1 reflects a less efficient cortical response to a stimulus, then the increase in gamma may reflect the increased effort required to bind the perceptual features of the stimulus into a coherent visual object. That gamma activity may compensate for non-typical initial sensory processing is also suggested by the lack of differences in behavioural reaction time to target stimuli between the older participants who exhibited a single vs doublepeaked P1. Future studies should explore the relationship between gamma activity and behavioural performance at an individual subject level. If the assumption is taken that gamma activity that follows a double peaked P1 is critical for compensatory processing, then behavioural responses should be directly influenced by the magnitude of the gamma activity. One possibility that must be considered is that the increase in gamma activity observed is due to the spike potential associated with micro-saccades rather than neuronal activity. Yuval-Greenberg et al. (2008) demonstrated that bursts of post-stimulus gamma were time locked to microsaccades rather than to stimulus onset. Without eye-tracking data we are unable to rule this out, however it should be noted that the bursts observed in Yuval–Greenberg’s study were of considerably shorter duration than those observed in the current study. Our findings suggest that previous reports may have underestimated the prevalence of double peaked P1 VEPs in healthy older populations (e.g. Blumhardt et al., 1980; Jones and Blume, 1985; Marra, 1990; Rousseff et al., 2005). The grand average P1 responses for younger and older adults in the current study are highly comparable to those presented in a recent study by Vaden et al. (2012) in which simple visual stimuli (horizontal and vertical bars of varying orientation) were used to examine changes in alpha power associated with selective attention in older adults. The grand average VEP plots presented also show a clear double peaked P1 in the older adult group, but the morphology change is not discussed. Störmer et al. (2013) recently presented a report on the effects of ageing on visual attention during object tracking in which a clear double peaked P1 was present in the older adults’ data. There was also a double peaked P1 to some degree in the

7

younger adults’ data, however it was not consistent across the electrodes presented and the double peak morphology was less well defined as in the older adults’ data. Additionally topographic differences were reported with the later peak having a more centro–parietal distribution than the occipito–parietal distribution of the early peak. Visual inspection of the topographic distributions of the P1a and P1b peaks in the current study showed no clear differences between the two peaks. Much of the initial research into aging and visual evoked potentials was conducted using slower sampling rates, sparser electrode arrays (e.g. Blumhardt et al., 1980) and used high-pass filters that have the potential to distort or mask small changes in peak morphology. The unexpectedly high prevalence of the double peaked P1 in older adults in the present study compared to previous studies may be due to changes in EEG recording and analysis techniques since many of these studies were conducted. It may be beneficial to re-examine some of the early work in ageing and VEPs with these new methodological advantages. The stimuli used in this study were closer to a simple flash stimulus, rather than the more commonly used chequerboard stimuli used to elicit pattern reversal VEPs. Consequently any paradigm designed to elicit VEPs that have a rapid visual onset/offset, i.e. closer to a flash than a pattern change, may elicit a double peaked P1 in older adults. It is unlikely that the observed double-peak pattern in older adults is caused solely by the specific visual stimulus presented in the study. The 12% prevalence of double peaked P1 observed in the younger group is somewhat higher than in the previous reports, the highest previously reported rate was 6% by Blumhardt et al. (1980), and may be higher with flash VEP paradigms. However if the stimulus alone caused a double peaked P1 response, we would expect a much higher occurrence rate in the younger group, yet, the prevalence of double peak in the younger group was 12% compared to 42% in the older group. The increase in prevalence of the double peaked P1 with age, in the absence of any significant pathology, has implications for both clinical and experimental work. Current clinical suggestions, such as that of Rousseff et al. (2005), that a double peaked P1 is symptomatic of demyelinating disease and warrants further investigation may be too general, and specific to pattern reversal VEPs. Additionally it may be that there is far greater individual variability in P1 in healthy ageing than previously considered. From an experimental perspective the P1 is present in most VEP research and any age related change in its morphology should be taken into account or controlled for. With further studies there is the potential to tease apart which processes that typically coalesce into a single peaked P1 in younger adults are delayed in older adults, potentially furthering our understanding of visual processing across all ages. In particular, origins of double peaked P1 may be elucidated via dipole modelling or high resolution fMRI region of interest analysis of the P1 in older participants who exhibit a double peaked response. Whilst many studies may be concerned with later components such as the N1 and P3 they must take account of the preceding components, as changes in the earlier components may well impact on the amplitude and latency of subsequent components. This is clearly the case in the current data. The amplitude difference between younger and older adults’ evoked responses begins at the P1 peak but continues until 350–400 ms. If either the N1 or P3 peaks were analysed in isolation, i.e. without taking the preceding differences into account, they would show a clear effect of group. This would be highly misleading as the significant positive correlation between P1 and N1 amplitude demonstrates. These findings must however be treated with caution as the study was not originally designed to fully explore the effects of age on P1, instead the stimuli and experimental design were optimised to elicit a visual mismatch negativity (see Stothart et al., 2013). Nevertheless the clear difference between younger and

Please cite this article in press as: Stothart G et al. Double peaked P1 visual evoked potentials in healthy ageing. Clin Neurophysiol (2013), http://dx.doi.org/ 10.1016/j.clinph.2013.11.029

8

G. Stothart et al. / Clinical Neurophysiology xxx (2013) xxx–xxx

older adults’ P1 responses warranted further investigation given the clear departure from a typical P1 response displayed by many of the older adults. The visual mismatch negativity paradigm used would only be expected to elicit differences between groups following the P1 response. Therefore components occurring prior to the visual mismatch negativity response remained unaffected by the experimental design. The optical health of the healthy older adults was not controlled for beyond ensuring that they had ‘‘normal or corrected to normal’’ vision, assessed during interview at memory clinics and excluding any participants with significant optical problems, e.g. macular degeneration, glaucoma. Additionally while all older adults were screened to ensure they were not taking any medication known to affect cognition, other medications were not controlled for. Given the previous evidence that has demonstrated that acetylcholine antagonists can temporarily induce a double peaked P1 VEP (Bajalan et al., 1986) it is possible that other prescription or over the counter medications may also have a similar affect. Despite these potential confounds, it is important to note that research involving healthy older adult populations will invariably include individuals taking medications and individuals with minor optical health concerns, e.g. cataracts, making these findings relevant to any VEP research involving older adults. Future studies should account for the optical health of participants, prescribed or over the counter medications, and include a wider range to stimuli in order to confirm that our findings are not stimulus-specific and fully assess the different stimulus characteristics, visual field locations and orientations that can influence the visual P1. To conclude, we present the novel finding that a double peaked P1 VEP appears to be a consequence of healthy ageing in a substantial proportion of individuals. In the absence of any known pathology almost half of the older adults in the study displayed a double peaked P1. The prevalence of this morphology has both clinical and experimental consequences and should be explored fully. Acknowledgements The authors would like to thank Dr. Judy Haworth at the Avon and Wiltshire and South Gloucestershire Primary Care Trust memory service clinics at the BRACE Centre, Frenchay Hospital, Abi Wright at the Research Institute, for Care of the Elderly and BRACE-Alzheimer’s research registered charity No. 297965 for their help with recruitment and the BRACE charity for support with equipment funding. We would also like to thank Jenna Todd Jones for her assistance in data analysis. This work was supported by the Biotechnology and Biosciences Research Council. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.clinph.2013.11. 029. References Aguglia U, Oliveri RL, Gambardella A, Quattrone A. Functional integrity of benzodiazepine receptors of the geniculo-striate visual pathways in Creutzfeldt-Jakob disease. J Neurol 1993;240:25–7. Amenedo E, Diaz F. Aging-related changes in processing of non-target and target stimuli during an auditory oddball task. Biol Psychol 1998;48:235–67. Bajalan AA, Wright CE, Vliet VD. Changes in the human visual evoked potential caused by the anticholinergic agent hyoscine hydrobromide: comparison with results in Alzheimer’s disease. J Neurol Neurosurg Psychiatry 1986;49:175–82. Berg P, Scherg M. A multiple source approach to the correction of eye artifacts. Electroenceph Clin Neurophysiol 1994;90:229–41. Blumhardt LD, Barrett G, Kiss A, Halliday AM. The pattern-evoked potential in lesions of the posterior visual pathways. Ann NY Acad Sci 1980;338:264–89. Brecelj J, Strucl M, Hawlina M. Central fiber contribution to W-shaped visual evoked potentials in patients with optic neuritis. Doc Ophthalmol 1990;75:155–63.

Brigell M, Kaufman DI, Bobak P, Beydoun A. The pattern visual evoked potential. Doc Ophthalmol 1994;86:65–79. Ceponiene R, Westerfield M, Torki M, Townsend J. Modality-specificity of sensory aging in vision and audition: evidence from event-related potentials. Brain Res 2008;1215:53–68. Clark VP, Hillyard SA. Spatial selective attention affects early extrastriate but not striate components of the visual evoked potential. J Cogn Neurosci 1996;8:387–402. Clark VP, Fan S, Hillyard SA. Identification of early visual evoked potential generators by retinotopic and topographic analyses. Hum Brain Mapp 1994;2: 170–87. Czigler I, Balázs L. Age-related effects of novel visual stimuli in a letter-matching task: an event-related potential study. Biol Psychol 2005;69:229–42. Di Russo F, Martínez A, Sereno MI, Pitzalis S, Hillyard SA. Cortical sources of the early components of the visual evoked potential. Hum Brain Mapp 2002;15: 95–111. Di Russo F, Pitzalis S, Spitoni G, Aprile T, Patria F, Spinelli D, et al. Identification of the neural sources of the pattern-reversal VEP. NeuroImage 2005;24:874–86. Falkenstein M, Yordanova J, Kolev V. Effects of aging on slowing of motor-response generation. Int J Psychophys 2006;59:22–9. Folstein M, Folstein S, McHugh P. ‘‘Mini-mental state’’. A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 1975;12:189–98. Gazzaley A, Clapp W, Kelley J, McEvoy K, Knight RT, D’Esposito M. Age-related topdown suppression deficit in the early stages of cortical visual memory processing. Proc Natl Acad Sci 2008;105:13122–6. Herrmann CS, Fründ I, Lenz D. Human gamma-band activity: A review on cognitive and behavioral correlates and network models. Neurosci Biobehav R 2010;34:981–92. Hoechstetter K, Bornfleth H, Weckesser D, Ille N, Berg P, Scherg M. BESA source coherence. a new method to study cortical oscillatory coupling. Brain Topogr 2004;16:233–8. Janáky M, Pálffy A, Horváth G, Tuboly G, Benedek G. Pattern-reversal electroretinograms and visual evoked potentials in retinitis pigmentosa. Doc Ophthalmol 2007;117:27–36. Jones DC, Blume WT. Aberrant wave forms to pattern reversal stimulation: clinical significance and electrographic ‘‘solutions’’. Electroenceph Clin Neurophysiol 1985;61:472–81. Maris E, Oostenveld R. Nonparametric statistical testing of EEG- and MEG-data. J Neurosci Meth 2007;164:177–90. Marra TR. The clinical significance of the bifid or ‘‘W’’ pattern reversal visual evoked potential. Clin Electroenceph 1990;21:162–7. McDonald WI, Compston A, Edan G, Goodkin D, Hartung HP, Lublin FD, et al. Recommended diagnostic criteria for multiple sclerosis: guidelines from the international panel on the diagnosis of multiple sclerosis. Ann Neurol 2001;50: 121–7. Mensen A, Khatami R. Advanced EEG analysis using threshold-free clusterenhancement and non-parametric statistics. NeuroImage 2013;67:111–8. Oostenveld R, Fries P, Maris E, Schoffelen JM. FieldTrip: open source software for advanced analysis of MEG, EEG, and invasive electrophysiological data. Comp Intell Neurosci 2011;2011:1–9. Parisi V, Miglior S, Manni G, Centofanti M, Bucci MG. Clinical ability of pattern electroretinograms and visual evoked potentials in detecting visual dysfunction in ocular hypertension and glaucoma. Ophthalmology 2006;113:216–28. Pfütze EM, Sommer W, Schweinberger SR. Age-related slowing in face and name recognition: evidence from event-related brain potentials. Psychol Aging 2002;17:140–60. Plomp G, Kunchulia M, Herzog MH. Age-related changes in visually evoked electrical brain activity. Hum Brain Mapp 2012;33(5):1124–36. Proverbio AM, Del Zotto M, Zani A. Inter-individual differences in the polarity of early visual responses and attention effects. Neurosci Lett 2007;419:131–6. Rousseff RT, Tzvetanov P, Rousseva MA. The bifid visual evoked potential: normal variant or a sign of demyelination? Clin Neurol Neurosur 2005;107:113–6. Sokol S, Moskowitz A, Towle V. Age-related changes in the latency of the visual evoked potential: Influences of check size. Electroenceph clin Neurophys 1981;51:559–62. Störmer VS, Li SC, Heekeren HR, Lindenberger U. Normal aging delays and compromises early multifocal visual attention during object tracking. J Cogn Neurosci 2013;25:188–202. Stothart G, Tales A, Kazanina N. Evoked potentials reveal age-related compensatory mechanisms in early visual processing. Neurobiol Aging 2013;34:1302–8. Tallon-Baudry C, Bertrand O. Oscillatory gamma activity in humans and its role in object representation. Trends Cogn Sci 1999;3:151–62. Tsaloumas MD, Good PA, Burdon MA, Misson GP. Flash and pattern visual evoked potentials in the diagnosis and monitoring of dysthyroid optic neuropathy. Eye 1994;8:638–45. Vaden RJ, Hutcheson NL, McCollum LA, Kentros J, Visscher KM. Older adults, unlike younger adults, do not modulate alpha power to suppress irrelevant information. NeuroImage 2012;63:1127–33. Yordanova J, Kolev V, Hohnsbein J, Falkenstein M. Sensorimotor slowing with ageing is mediated by a functional dysregulation of motor generation processes: evidence from high resolution event related potentials. Brain 2004; 127:351–62. Yuval-Greenberg S, Tomer O, Keren AS, Nelken I, Deouell LY. Transient Induced gamma-band response in EEG as a manifestation of miniature saccades. Neuron 2008;58:429–41.

Please cite this article in press as: Stothart G et al. Double peaked P1 visual evoked potentials in healthy ageing. Clin Neurophysiol (2013), http://dx.doi.org/ 10.1016/j.clinph.2013.11.029