Defining the Neurocircuitry of Borderline Personality Disorder: Testing a Maturational

1

Neurocircuitry of borderline personality disorder

Defining the neurocircuitry of borderline personality disorder: Testing a maturational model by means of single trial ERP analysis

Russell Meares1, 2,Dmitriy Melkonian2, Allan Schore34, Evian Gordon1,2 and Leanne M. Williams1,22,3

1Department of Psychological Medicine, Western Clinical School, University of Sydney; 2Brain Dynamics Centre, Westmead Millennium Institute and Western Clinical School, University of Sydney, Westmead Hospital, 3School of Psychology, The University of Sydney, and34University of California at Los Angeles David Geffen School of Medicine

Address correspondence and reprint requests to: Russell Meares, Department of Psychological Medicine, University of Sydney at Westmead Hospital, Wentworthville, NSW 2145, Australia;

E-mail: .

Short title: Neurocircuitry of borderline personality disorder

Defining the neurocircuitry of borderline personality disorder: Testing a maturational model by means of single trial ERP analysis

Abstract

Borderline personality disorder (BPD) is a severe disturbance of personality functioning characterized by psychic discontinuities, deficits in emotion and impulse regulation, and persistent difficulties in interpersonal relationships. These patients commonly experience abusive and neglectful developmental backgrounds and severely disrupted attachments. We presenttest a developmental maturational model of the psychoneurobiological pathogenesis of BPD and then assess it usingby means of event related potentials (ERPs) using a standard auditory “oddball” paradigm. highly heritable brain function endophenotypes?? (noting that EEG is around 80% - and ERPs moderate)with event related potentials (ERPs) using a standard auditory “oddball” paradigm. REPEATED WORD – WE We examined tThe complex of late component ERPs in 17 BPD patients and, 17 matched healthy controls were examined and consideredwithin a developmental context formed by data on and

108 healthy control subjects

spanning fromspanning 14 to 70 years of age. The elicitation rate (ER), a novel probabilistic measure of single trial variability gave indications of late positive P3a and P3b endogenous potentials as major electrophysiologicalmamarkers of disorder. We found two distinct types of neural circuitry dysfunctions reflected by different topography and an opposite character of P3a and P3b abnormalities:P3a - increase of the voltage and ER, P3b - decrease of the voltage and ER. P3a abnormalities dominate in the frontal regions of the right hemisphere, while less pronounced spatial asymmetry of P3b indicates a prevalence of abnormalities in the left hemisphere. The P3a dysfunction is consistent with the disturbances of prefrontal function as predicted by a developmental model of BPD psychopathogenesis involving disruption of the “medial reward system”, specificallyparticularly in the right brain??? WHY SPECIFICALLY RIGHT – NO EVIDENCE OF THIS??.

Key words: BPD, borderline, psychopathogenesis, lateralization, ERP, P3a

INTRODUCTION

Borderline personality disorder (BPD) is a severe disturbance of personality functioning characterized by affective disturbances and impulse control associated with deficits in emotion regulation, as well as a pervasive pattern of instability in self-image and persistent difficulties in interpersonal relationships (American Psychiatric Association, 1994). A variety of imaging technologies ishave been used to now identifying dysfunctions in brain circuitry underlying BPD (Brendel , Stern, & Silbersweig, et al., 2005). An integratiintegrative approach to neuroscience supported by on of findings provided by differentcomplementary techniques has the potential to provide an preliminary explanatory frameworks for the understanding of the core clinical features of this and others conditions (WE SHOULD DEFINITELY REFERENCE EVIAN”S WORK HERE – Gordon, 2003 Neuropsychopharmacology, Gordon et al. 2005 Int. J Psychophysiology – BOTH ON OUR WEBSITE). An integrative approach can also incorporate neurodevelopmental impacts such as an that are increasingly seen as the outcome of the abusive and neglectful developmental background suffered by the typical BPD patient. Since borderline patients c characteristically experience such a disadvantageous environment during early life, current models regard BPD as a largely developmental psychopathology (Fonagy, 1991;Gunderson, 1996; Meares, 1993;Meares, Stevenson, & Gordon, Meares et al., 1999; Schore, 1996;).

I’D COLLAPSE THE BLUE HIGHLIGHT INTO ONE PARAGRAPH - FOCUSING ON THE INTEGRATIVE ASPECT (BRAIN AND MIND FOR INSTANCE) Psychodynamic developmental theories of BPD focus on environmental factors and interpersonal experiences which lead to the development of BPD (Clarkin, Yeomans, & Kernberg, Clarkin et al., 1999; Fonagy, Gergely, Jurist, & Target,Fonagy et al., 2002; Kernberg, 1975; Masterson & Rinsley, 1975; Schore, 1994).

This report concerns a testing of two interrelated hypothesis concerning the genesis of BPD which have major neurobiological components. The first draws A central issue of a developmental neurobiological model of BPD drawing upon Hughlings Jackson’s theories of dissolution is the character of neural dysfunctions caused by these psychosocial factors (Meares et al., 1999). This maturational model of BPD pathogenesis is based upon

Jackson’s proposaled that the evolutionary development of the prefrontal cortex allows for the emergence of a higher order, or reflective consciousness, which he called “self” (Jackson, 1931, 1932). He conceived “self” as the product of an increased capacity for coordination of brain systems brought about by a late evolutionary development of pre-frontal activity. This co-ordination results in increased control, i.e. monitoring, evaluation, and modulation over systems which had evolved earlier, e.g. the limbic system (Meares, 1999).

A crucial concept that follows from this phylogeny of the human brain and puts the maturation of cognitive functions into an environmental framework is the He also postulated that the higher order functions are “incomplete”, implying that an environmental provision is necessary for their maturation. Important clinical inferences from this theory are that maturation failure will be associated, first, with less coordinated or aggregated, activity between various neural generators, resulting in disaggregation or discontinuities in psychic life, and secondly, with diminished cortical inhibitory control over earlier evolved subcortical brain systems (Meares, 1999). Utilizing this Jacksonian perspective, the model proposes that those suffering this condition will show disturbance of prefrontal function, relative disconnection between neural generators and a failure of frontally mediated inhibitory control (Meares et al., 1999).

The second hypothesis THE FOLLOWING SECTION IS GETTING TOO LONG – AND I THINK TOO FOCUSED ON OTHER PEOPLE’S THEORIES.

I WOULD STRONGLY SUGGEST BASING THIS ON OUR OWN THEORETICAL MODEL (SEE LATEST IN JIN 2006 ATTACHED) AND A FOCUS ON SIGNIFICANCE AS A GENERIC CONCEPT THAT INCLUDES THE ODDBALL – AND THE CONTINUUM THRU TO EMOTIONAL VALENCE PER SE

THIS SHOULD REPLACE THIS BELOW – THAT LINKS INTO THE FOCUS ON P3a, P3B BELOW

Our hypothesis supposes that this failure of pre-frontally mediated inhibitory control will affect, in particular, the operations of the emotional processing system which Damasio has seen as underpinning the consciousness of self. The neurocircuitry of the emotional processing system involves the ventro-medial and orbitofrontal cortices, anterior cingulate, and the amygdala. It is sometimes called the “medial reward system”. Its functioning can be contrasted to a lateral system involving the dorsolateral prefrontal cortex and the hippocampus (Williams et al., 2001).

A third element of the hypothesis concerning the psychopathogenesis of BPD draws upon the neurophysiological evidence of a dominantmajor role of the right hemisphere in determining the establishment of selfhood (Keenan, Nelson, O’Connor, Pascual-Leone, 2001;

Keenan, Rubio, Racioppi, Johnson, & Barnacz, 2005; Feinberg, Keenan, 2005). The right hemisphere is also particularly implicated in inhibitory control for inhibitory control (Garavan, Ross, Stein, Garavan et al., 1999). and aA growing literaturewhich indicates that right hemispheric functions are essential to the processing and regulatory control of emotions (see Schore, 1994, 2003a,b) and of risk-taking behavior characteristic of the borderline syndrome (Knoch, Gianotti, Pascual-Leone, Treyer, Regard, Hohmann, & Brugger, 2006).

The right hemisphere is in a growth spurt in the first two years of life, a period of right brain dominance, and so right hemisphere resources are the first to develop (Chiron, Jambaque, Nabbout, Lounes, Syrota, & Dulac,Chiron et al., 1997; Decety & Chaminade, 2003; Schore, 1994). This growth is not totally encoded in the genome, but is indelibly shaped by experiences with the environment, especially the social environment. The neurobiological maturation of the emotion processing right hemisphere in the early critical period of the first two years of life is thus “experience-dependent”. It is specifically the affect communicating and regulating transactions within the infant-mother attachment bond that impact the experience-dependent maturation of prefrontal cortical-limbic circuits of the early developing right cortical hemisphere(Schore, 2005). Lyons and colleagues demonstrate that conditions that have an impact on early maternal variability in infancy produce “significant differences in right but not left adult prefrontal volumes, with experience-dependent asymmetric variation most clearly expressed in ventral medial cortex measured in vivo by magnetic resonance imaging” (Lyons, Afarian, Schatzberg, Sawyer-Glover, & Moseley,Lyons et al.,2002, p. 51).

Neurological findings suggest that the orbitofrontal areas of the right hemisphere are more critical to emotional functions than those of the left (Tranel, Bechara, & DenburgTranel et al., 2002)and are centrally involved in “the emotional modulation of experience” (Mesulam, 1998). Current formulations of the pathogenesis of BPD suggest that a developmental impairment of the experience-dependent maturation of this system results in disorders of

attachment leading to later failure of maturation of higher order consciousness which first appears at about 4 years of age (Flavell, Green, & FlavellFlavell et al., 1993; Meares Orlay, 1988), and a related disturbance in the achievement of intimacy (Meares, 1993).

WE CAN LINK THE ABOVE FRAMEWORK INTO P3a, P3B (I EXPLICITLY COVER THEM) – AND THAT ALSO LINKS TO OUR GAMMA PAPER WHICH ARGUES FOR FRONTAL VS POSTERIOR NETWORKS IN BPD

(RELEVANT TO P3A VS P3B, RESPECTIVELY) – A preliminary electrophysiological test of this developmental maturational hypothesis has been conducted in earlier single trial study of P3a and P3b auditory event related potentials (ERPs) which revealed distinctive disturbances of the P3a in BPD, namely, significantly enhanced amplitude and failure to habituate (Meares, Melkonian, Gordon, & WilliamsMeares et al., 2005). Abnormal changes of late component ERPs, particularly P3a and P3b, are regarded as reliable objective markers of cognitive disorders in clinical research (Polich, 2004).The evidence that abnormalities of ERPs may reflect long-lastingdysfunctions inemotion processing and selective attention caused byearly experience of abuse and neglect has been obtained for different age groups using emotional stimuli(Cicchetti, & Curtis, 2005; Parker, NelsonParker et al., 2005) and selective attention paradigms (Pollak, & Tolley-Schell, 2003).

An important benefit of ERP single trial analysis as compared with older averaging techniques isthat measures of trial-to-trial variability provide additional clues for the understanding of cognitive implications of ERP abnormalities. A methodological approach established in the previous study of auditory ERPs has been extended is several aspects. First, the whole complex of late component ERPs, N1-P2-N2-P3a-P3b, has been analysed. Secondly, the component elicitation rate(ER) has been estimated. This novel parameter of single trial variability of ERP components shows the probability with which a task relevant stimulus elicits a reaction, a basic concept supporting the very notion and a basic mechanism of an endogenous potential (Donchin, Ritter, & McCallumDonchin et al., 1978). Finally, the topographical aspects of ERP spatial distribution for frontal, medial and parietal scalp locations have been analysed.

Thethree major objectives guided this work. A first objective of this study addressed to BPD patients and age and sex matched control group was to outline a group of ERP components the abnormalities of which are most sensitive markers of BPD. A second objective addressed to the same comparison groups was to investigate temporal and spatial aspects of ERP abnormalities, particularly the inter-hemispheric asymmetry in BPD patients.A third objective of this study was to find ERP abnormalities which might be linked to developmental factors. For this purpose the ERP data from BPD patients were compared with the normative age dependencies that were derived from the normative group of 106 healthy subjects from 14 to 70 years of age.

RESEARCH DESIGN AND METHODS

Subjects

Seventeen patients with BPD (4 males and 13 females; range = 20-44 years; mean age = 31.6) took part in the study. The BPD patients came from an ongoing program for the treatment and evaluation of BPD patients. The diagnosis was made by two independent raters (psychiatrist and psychologist) according to DSM-III-R criteria in a diagnostic interview that included the Diagnostic Interview for Borderline Patients. Patients were free of medication for at least 30 days at the time of the study.

Exclusion criteria for healthy subjects for control and normative groups were left-handedness, recent history of substance abuse, epilepsy or other neurological disorders, and mental retardation or head injury [assessed using Section M from the Composite International Diagnostic Interview (Robins, Wing, Wittchen, Helzer, Babor, et al.,Robins et al., 1989) and the Westmead Hospital Clinical Information Base questionnaire]. Subjects were asked to refrain from smoking or drinking caffeine for three hours prior to the recording session. Written consent was obtained from all subjects prior to testing in accordance with National Health and Medical Research Council guidelines.

The control group of age and sex matched healthy subjects used for comparisons with the patient group included 4 males and 13 females (range = 20-47 years ; mean age = 34.3).The subjects for normative groups were 54 normal females and 54 normal males between the ages of 14 and 70. There were 8 males and 8 females at each age from

14 to 22 (mean age=17.9), from 23 to 31 (mean age=26), from 32 to 40 (mean age=36.2), from 41 to 50 (mean age=45.4), from 51 to 60 (mean age=55.5) and from 61 to 70 (mean age=65.6). An almost homogenous distribution of age parameters across the groups provided opportunity to examine different age diapasons in order to find statistically significant age dependencies for different ranges.

Procedure

The ERP data were collected according to a standard auditory “oddball” paradigm in a method similar to that used in several previous studies (Lagopoulos, Gordon, Barhamali, Lim, Li, Clouston, & Morris, Lagopoulos et al., 1998;Melkonian, Gordon, & Bahramali, Melkonian et al., 2001; Williams, Gordon, Bahramali, Wright, & Meares, Williams et al., 2000). Each subject had their eyes open and was instructed to look at a coloured dot in the centre of a screen in order to minimize eye movements. Auditory tones(50-ms duration, 10-ms rise/fall time) were presented binaurally via stereo headphones at a constant interval of 1.3 s. Fifteen percent were target (task-relevant) tones presented at 1500 Hz. The remaining 85% were background (task-irrelevant) tones delivered at 1000 Hz. All tones were presented at 60 dB above each individual subject’s auditory threshold, pseudorandomly, with the constraint that two high tones (targets) were not presented in succession. Subjects were asked to ignore the background tones and to press two reaction time buttons “as fast and accurately as possible” using the index finger of each hand when they identified a target tone. The identification of each target was regarded as correct, if a button press response was obtained within 1 s of the target tone. Reaction time was recorded for each button press.

Recording and preprocessing

EEGs were recorded from midline and bilateral placements over frontal (F3, Fz, F4), central (C3, Cz, C4), and parietal (P3, Pz, P4) scalpelectrode sites according to the 10-20 international system using a DC acquisition system.Electrooculograms (EOGs) were recorded bipolarly, vertically from the middle of the supraorbital and infraorbital regions of the left eye and horizontally from two electrodes, placed 1 cm lateral to the outer canthus of each eye. Skin resistance at each site was <5 kΩ. The voltages were continuously digitized at 250 Hz and digitally stored. The preprocessing stage of the data analysis included the two procedures. First, EOG correction using a technique based on Gratton, Coles, and Donchinet al. (Gratton et al. 1983). This procedure retained unchanged the format of original EEG time series. Second, extraction of standard stimulus locked EEG segments. For each electrode site,the successive EEG segments time locked to correctly detected target tones were tested on the presence of excessive eye movement and electrical artifacts. An automated artifact rejection technique was employed that combines several criteria to identify abnormal activities (Harris, Melkonian, Williams, & Gordon,Harris et al., 2006). An accepted trial was then digitally filtered usingan equal weight boxcar filter (Cook & Miller, 1992) with the window length of 8 ms to attenuate high frequency activity above 50-60 Hz (the bandpass is variable because of the adaptive filtering). A 251 time points time series extracted from 0.2 s pre-stimulus to 0.8 s post-stimulus was digitally stored and the procedure continued until 40 eligible segments were acquired.

Single-trial analysis

Single trial EEG segments were analysed using technique of high resolution fragmentary decomposition described in detail elsewhere (Melkonian et al., 2001,; Melkonian, Blumenthal, & Meares, 2003). This method identifies the major peaks in single trial segment and applies to each peak combinedtime and frequency domain waveform analysis to estimate component parameters using a component template (model ERP) called a generic mass potential (GMP),

where

is function termed a bud,and , and  are the GMP parameters.The bud has a steeply rising left flank and a slowly decreasing right flank. The maximum of function is at x=1.2: bud(1.2)0.356.

After positioning of identified GMPs according to their peak latencies, the procedure of the component overlap correction is applied to provide coincidence of the peak amplitudes of the summary waveform with the peak amplitudes of empirical single trial (Melkonian et al., 2003). The measures we employ to characterize the ERP component are: A=0.356 - peak amplitude,O= - onset time, L=O+1.2 - peak latency. Reasonable accuracy of these measures applied to non-stationary electrophysiological signals has been shown elsewhere (Blumenthal & Melkonian, 2003).

Extended criteria for ERP classification from single trial records were applied which include the amplitude and shape parameter as additional descriptive parameters of the component under analysis (Meares et al., 2005). Besides the component identification, the amplitude and shape discrimination removeseffectively from the analysis insignificantly small and artifact contaminated peaks.