EEG theta activity and pain insensitivity in self-injurious borderline patients

Abstract

The principal aim of this study was to investigate possible neurophysiological underpinnings of self-injurious behavior in women with borderline personality disorder (BPD). Pain report and EEG power spectrum density during a laboratory pain procedure, a 4-min 10°C cold pressor test (CPT), were compared among four groups; female inpatients with BPD who do (BPD-P group, n=22) and do not (BPD-NP group, n=19) report pain during self-injury, female inpatients with major depression (n=15), and normal women (n=20). The BPD-NP group reported less pain intensity during the CPT compared to the other groups. Total absolute theta power was significantly higher in the BPD-NP group compared to the Depressed (P=0.0074) and Normal (P=0.0001) groups, with a trend toward being significantly higher compared to the BPD-P group (P=0.0936). Dissociative Experience Scale scores were significantly higher in the BPD-NP group compared to the Depressed and Normal groups (maximum P=0.0004), and significantly higher in the BPD-P group compared to the Normal group (P=0.0016). Beck Depression Inventory and Sheehan Patient Rated Anxiety Scale scores were significantly lower in the Normal group compared to all patient groups. Theta activity was significantly correlated with pain rating (Pearson partial r=−0.43, P=0.0001) and Dissociative Experiences Scale score (Pearson partial r=0.32, P=0.01).

Introduction

Self-injurious behavior (SIB) in individuals with borderline personality disorder (BPD) consists of behavior that causes direct tissue damage, tends to be repetitive and non-lethal in severity and intent, and may temporarily improve dysphoric mood (Kemperman et al., 1997a). A few investigators have begun exploring possible neurophysiological mechanisms underlying SIB in this population, most often focusing on the endogenous opioid (Coid et al., 1983, Russ et al., 1994) and serotonin systems (Gardner et al., 1990, Simeon et al., 1992). Available data are presently too sparse to draw any conclusions regarding possible neuromodulatory mechanisms.

We have been investigating pain perception in self-injurious borderline patients in an attempt to explore possible biological underpinnings of these maladaptive behaviors. The rationale for this approach is twofold. First, pain appreciation during self-injury is reported to be absent in about one-half of self-injurious borderline patients (Leibenluft et al., 1987). Second, pain perception and mood regulation may be related by common neuroanatomical and neurochemical substrates. We have conjectured that pain perception, particularly its more clearly identifiable neurophysiological components, may serve as a ‘window’ to brain processes that might mediate aspects of SIB.

We studied self-injurious borderlines who do (BPD-pain, BPD-P) and do not (BPD-no pain, BPD-NP) typically experience pain during self-injury and found that the BPD-NP group reported significantly less pain during a 10°C cold pressor test (CPT) compared with both the BPD-P group and normal control subjects (Russ et al., 1992, Russ et al., 1994). Moreover, unlike the other groups, the BPD-NP group experienced a significant improvement in depressed, anxious, and angry mood, as well as confusion, following the CPT. These findings, however, do not permit us to conclude that there are neurophysiological (neurosensory) differences in pain perception among the groups. Pain is a complex multi-dimensional phenomenon, consisting of neurosensory, affective, cognitive, socio-cultural, and motivational factors (Melzack and Casey, 1968, Chapman, 1977, Clark and Yang, 1983). Group differences in pain report, therefore, may reflect differences in any or all of these factors (Kemperman et al., 1997b).

The current study was undertaken in an effort to investigate possible EEG correlates of cold pressor pain in self-injurious borderline patients. Chen et al. (1989) found an increase in delta and beta activity recorded from frontal, parietal, temporal and occipital areas in normal subjects during a 3-min 1°C CPT compared to baseline using power spectrum density analysis. Moreover, they were able to distinguish pain-sensitive and pain-tolerant groups of subjects in that the former exhibited significantly higher delta power (but not beta power) than the latter under the cold pressor condition (Chen et al., 1989). Increases in delta power (Ferracuti et al., 1994) and beta power (Backonja et al., 1991) during the CPT have been reported by other investigators. These more recent studies reported changes in alpha power during the 1°C CPT, with an increase in the low alpha band (8–10 Hz, Backonja et al., 1991) and a decrease in the high alpha band (10–12 Hz, Ferracuti et al., 1994) described.

Our principal aim in this study was to determine whether changes in cortical brain electrical activity during the CPT differentiate self-injurious women with BPD who do or do not typically experience pain during self-injury. We also studied women with major depression but without BPD because of the high comorbidity between major depression and BPD (Gunderson and Phillips, 1991) and the reports of decreased pain sensitivity in depressed patients (Hall and Stride, 1954, Merskey, 1965, Davis et al., 1979, Dworkin et al., 1995). Finally, we included a control group of normal women. A demonstration of EEG power spectrum density differences related to pain report during the CPT between the BPD-NP group and other groups would be consistent with the view that diminished pain perception in the ‘analgesic’ group may in part have an identifiable neurophysiological basis. Although many methodological differences between our study and that of Chen et al. (1989) precluded the goal of replication, our study design was nonetheless stimulated and informed by the work of these investigators. Based on their findings, we anticipated that while all groups would demonstrate an increase in delta and beta power during the CPT, the increase in delta power would be least for the BPD-NP group.