INTERNATIONAL JOURNAL OF PSYCHOPHYSIOLOGY ELSEVIER International Journal of Psychophysiology 21 (1996) 189-196 Short Communication Hemispheric differences for P300 amplitude from an auditory oddball task d Joel E. Alexander a, Lance 0. Bauer b, Samuel Kuperman ‘, Sandra Morzorati , Sean J. O’Connor d, John Rohrbaugh e, Bemice Porjesz f, Henri Begleiter f, John Polich g** a Deparhnent of Psychology, Western Oregon State College, Monmouth, OR 97361, USA b Department of Psychiatry, Uniuersity of Connecticut Health Center., Farmington. CT 06030, USA ’ Department of Psychiatry, University of Iowa, Iowa City, IA 52242, USA ’ Department of Psychiatry, Washington Uniuersiv, St. Louis. MO 63108, USA d Department of Psychiatry, Institute of Psychiatric Research, Indianapolis, IN 46202, USA f Department of Psychiatry, SUNY Health Sciences Center, Brooklyn. NY 11203, USA ’ Department of Neuropharmacology. TPC-IO, The Scripps Research Institute. 10664 North Torrey Pines Road, La Jolla, CA 92037, USA Received 8 August 1995; revised 19 November 1995; accepted 20 November 1995 Abstract The P3(00) event-related potential (ERP) was elicited in 80 normal, right-handed male subjects using a simple auditory stimulus discrimination task, with electroencephalographic (EEG) activity recorded at 19 electrodes. P300 amplitude was larger over the right compared to left hemisphere electrode sites primarily at anterior-medial locations (F3/4, C3/4) for both target and standard stimuli. The NIOO, P200, and N200 components also demonstrated several similar, albeit less robust, hemispheric asymmetries. No hemispheric effects for P300 latency were observed, with few consistent latency findings for any of the other components obtained. The results suggest that the discrimination process underlying P300 generation may originate with right frontal activation. Keywords: P300; Event-related potential (ERP); Hemispheric differences; Auditory stimulus When an auditory ‘oddball’ paradigm is used to the midline scalp distribution provides substantial elicit the P3(00) event-related brain potential (ERP), information about the attentional and mnestic pro- subjects are required to discriminate between two cesses thought to contribute to P300 generation stimuli that vary on some dimension by responding (Donchin et al., 1986; Donchin and Coles, 1988; to a designated target stimulus. The P300 often is Johnson, 1993; Picton, 1992). However, relatively measured at the central (Fz, Cz, Pz) electrode sites little is known about its possible hemispheric differ- with lateral electrodes typically not assessed, since ences asymmetries, perhaps because of the typically small ERP amplitude hemispheric asymmetries ob- served (e.g., Brown et al., 1985; Molfese et al., * Corresponding author. Tel: 619-554-8176; Fax: 619-554-6393; 1975; Neville, 1980), which may originate from E-mail: [email protected] inadequate numbers or placement of electrodes and 0167.8760/96/$15.00 0 1996 Elsevier Science B.V. All rights reserved SSDI 0 167-8760(95)00047-X 190 J.E. Alexander et al./Internotionul Journal ofPsychophysiology 21 (1996) 189-196 the often small, inhomogenous samples employed in from an auditory oddball task is hemispherically ERP studies that have reported laterality effects asymmetric. If laterality differences for the P300 and (Donchin et al., 1977). Although cognitive hemi- other components can be demonstrated, such asym- spheric differences are observed readily using behav- metries may help disentangle the possible influences ioral techniques for auditory (Kimura, 1993; Ivry and of neuroanatomical factors on ERP morphology (Ford Lebby, 19931, visual (Sergent, 1991; Hellige, 1993; et al., 1994; Myslobodsky et al., 1989; Pfefferbaum Polich, 1993a), and tactile (O’Boyle et al., 1987; and Rosenbloom, 1989). Furthermore, these data also Reitan et al., 1992) stimuli, hemispheric asymmetries may be helpful in the evaluation of ERP scalp for late endogenous components such as the P300 distribution changes that occur with normal aging and N4(00) have been found primarily with verbal and clinical disorders (Holinger et al., 1992; Pfeffer- tasks in the visual modality (e.g., Friedman et al., baum et al., 1989; Ford and Pfefferbaum, 1991; 1975; Kutas and Hillyard, 1982; Kutas and Van Polich, 1993b; Polich and Luckritz, 19951. Petten, 1988). Thus, lateralization of the P300 and A total of 80 young adult right-handed males other cognitive ERPs has been observed for some (mean = 22.6, SD = 1.8 years) served as subjects for tasks, but not in all modalities. pecuniary remuneration, with several laboratories us- Several recent studies that utilized multiple scalp ing the same equipment and methods contributing to electrodes have suggested that hemispheric ampli- the data set (cf. Alexander et al., 1994). Males were tude differences for the P300 ERP from an auditory used exclusively to maximize the likelihood of ob- oddball paradigm may exist, such that P300 ampli- taining cognitive based hemispheric differences tude from the right hemisphere is larger than that (Halpem, 1992). Based on responses to a compre- from the left hemisphere (cf. Kamiski and Blair, hensive survey, all subjects reported an absence of 1989; Holinger et al., 1992; Naumann et al., 1992). psychiatric or neurologic problems and were screened However, these effects seem to occur primarily at the for alcohol and drug use. Handedness was evaluated anterior and central locations, which is somewhat by a self-report questionnaire, derived from standard surprising given that the P300 is usually largest over assessment methods (Bryden, 19771, that assessed parietal areas (Johnson, 1989; Polich, 1989) - at use of dominant hand, foot, and eye use on a variety least for young adult subjects (cf. Friedman et al., of tasks in addition to experimenter observation of 1993; Vesco et al., 1993). Moreover, although P300 writing, with all subjects reported being strongly hemispheric asymmetries have been reported they right-handed. are not completely consistent in their strength or EEG activity was recorded monopolarly using an location. The sources of these discrepancies are un- EC1 electrode-cap at 19 electrode sites (FP1/2, clear but are likely related to the nonhomogeneous F3/4, C3/4, P3/4, F7/8, T7/8, P7/8, 01/2, Fz, samples that are compromised by a lack of control Cz, Pz) referred to the nose, with a forehead ground over variables that can affect laterality differences and impedances at 5 k0 or less. Electra-ocular (e.g., left/right handedness, male/female subjects, (EOG) activity was assessed with two channels ref- etc.), the relatively small samples are typically em- erenced to the nose. One electrode was placed at the ployed (e.g., n = IO-20), and the use of linked ears outer canthus of the left eye to measure vertical eye or mastoids as a reference (this method does not movement and the second electrode was located on appear to affect asymmetry magnitudes appreciably, the forehead to monitor horizontal eye movement. but it has caused some controversy in this area; cf. The filter bandpass was 0.02-50 Hz (3 dB down, 6 Andino et al., 1990; Faux et al., 1990; Nunez, 1981; dB octave/slope) and the EEG was digitized at 3.9 Senulis and Davidson, 1989). When taken together, ms/point for 1500 ms, with a 187 ms prestimulus previous studies suggest a possible P300 hemispheric baseline. ERP data were averaged on-line, with the difference for nonlateralized stimulus presentations same computer also used to control the stimulus and simple discrimination tasks, but the nature of presentation and artifact rejection. Trials on which these effects is uncertain. the EEG or EOG exceed +73.3 PV were rejected The present study was designed to determine in a automatically. comprehensive fashion whether the P300 component ERPs were elicited with 400 auditory binaurally J.E. Alexander et al./lnternational Journal of Psychophysiology 21 (1996) 189-196 191 presented stimuli consisting of 600 Hz (standard) tifying amplitudes and latencies of the NlOO, P200, and 1600 Hz (target) tones presented at 60 dB SPL N200, and P300 components at each electrode site (10 ms r/f, 60 ms plateau). The interstimulus inter- by locating the most negative or positive component val was 1.5 s and the target tone occurred randomly within the latency windows of 75-125, 100-200, with a probability of 0.125. Subjects were instructed 125-275, and 250-450 ms, respectively. Amplitude to press a key pad with their forefinger whenever a was measured relative to the prestimulus baseline, target tone was detected and to refrain from respond- with peak latency defined as the time point of maxi- ing when the standard tone was presented. Response mum positive or negative amplitude within the la- hand was counterbalanced across subjects. Stimulus tency window. presentation was concluded when 25 target and 75 Results for the midline electrode sites (Fz, Cz, Pz) standard artifact-free stimuli were acquired. have been reported elsewhere as part of a large-scale All analyses of variance employed Greenhouse- study on inter-laboratory reliability and will not be Geisser corrections to adjust for violations of the considered further (Alexander et al., 1994). To as- sphericity assumption inherent in repeated measures sess within-hemisphere effects the left lateral (F7, designs. Only probability values from corrected df T7, W), left medial (F3, T3, P3), right medial (F4, are reported here. Task performance was virtually T4, P4), and right lateral (F8, T8, P8) electrode perfect with fewer than 2% of the target trials re- locations were analyzed (preliminary evaluation indi- jected and/or misperceived across subjects. Mean cated that the Fp1/2 and 01/2 electrode sites response time CRT) to the target stimulus taken over yielded no hemispheric effects). Additional analyses all subjects was 379 ms. Waveforms for both target of peak latency found no reliable P3 hemispheric and standard stimuli were analyzed visually by iden- results and only a few inconsistent effects for the MEDIAL LATERAL TARGETS STANDARDS TARGETS STANDARDS F3/4 .__I--- -jf---- -bL cu4 --*--*-- --If=-- ji P3/4 ---c--- W/6 --._.