Sleep-EEG in borderline patients without concomitant major depression: a comparison with major depressives and normal control subjects

Abstract

The link between borderline personality disorder (BPD) and the affective disorders remains controversial. The aim of this study was to examine the relationships between BPD and major depression (MD) from the perspective of sleep parameters and to contribute to the characterisation of the sleep-EEG in BPD. We compared 20 off-medication BPD in-patients without co-existing MD with 20 sex- and age-matched MD patients without BPD and 20 sex- and age-matched control subjects. BPD patients had a greater prevalence of drug or alcohol abuse and suicide attempts than MD patients. MD patients had higher scores on the Hamilton Depression Rating Scale (HDRS). Both BPD and MD patients had less total sleep time, more prolonged sleep onset latency, and a greater percentage of wakefulness than control subjects. BPD patients and control subjects had more stage 2 sleep than MD patients. BPD patients had a longer duration of rapid eye movement (REM) sleep, and less stage 3, stage 4 and slow wave sleep than MD patients and control subjects. REM latency did not differentiate the three groups. BPD and MD patients shared sleep-continuity characteristics, but sleep architecture differentiated the two groups. BPD patients with a past history of MD had more wakefulness and less slow wave sleep than BPD patients without a history of MD; other sleep parameters, age, sex and HDRS scores were not statistically different in the two BPD subgroups. Although BPD and MD may coexist, the present study offers more arguments favouring the concept that they are not biologically linked and that BPD patients with depressive symptoms often experience an affective syndrome different from that in MD patients without BPD, in terms of quality and duration of symptoms and of the biological substrate.

Introduction

Borderline personality disorder (BPD) manifests a wide variety of symptoms. BPD has been associated with several pathological conditions such as epilepsy, affective disorders, schizophrenia, schizoaffective disorder, and atypical psychoses (Andrulonis et al., 1982, Post et al., 1982, Schulz et al., 1989). Although the hypothesis of a relationship with epilepsy is difficult to support, there is evidence of an excess of electroencephalographic (EEG) slow activity, which can indicate brain dysfunction, in patients with BPD (De la Fuente et al., 1998). Other methodologies, such as positron emission tomography, have also shown the existence of brain abnormalities in BPD (De la Fuente et al., 1997).

The link between BPD and the affective disorders is controversial (Korzekwa et al., 1993, Coid, 1993, Rogers et al., 1995). Arguments for a link between BPD and affective disorders (Schulz et al., 1989) are contrasted with those that suggest the two disorders to be independent (Soloff et al., 1989, Benson et al., 1990). Gunderson and Phillips (1991) conclude that the two disorders may coexist, but are not etiologically related. Some authors have suggested that depression associated with BPD is distinguished from depression without BPD by the quality and duration of symptoms (Westen et al., 1992, Coid, 1993, Rogers et al., 1995). Moreover, other authors have proposed that the depressed symptoms in BPD might have a biological substrate that is distinct from that in depressive illness without concomitant BPD characteristics (Krishnan et al., 1984, Korzekwa et al., 1993, Kavoussi et al., 1993, De la Fuente and Mendlewicz, 1996). Pharmacological studies have shown differences in medication response between BPD and MD patients. Conventional antidepressants have been reported to lead to poor response and even worsening in BPD (Cole et al., 1984, Soloff et al., 1989).

The dexamethasone suppression test (DST) and the thyrotropin-releasing hormone stimulation test (TRH-ST), which have proved valuable in the study of affective illness, have also been studied in BPD (Korzekwa et al., 1991, De la Fuente and Mendlewicz, 1996). In some of these studies, positive findings with the DST and TRH-ST seem to be more related to an Axis I diagnosis of MD than to BPD itself (Soloff et al., 1991). The only study evaluating the DST and the TRH-ST in BPD patients without a co-existing diagnosis of MD (De la Fuente and Mendlewicz, 1996) provided no evidence for a biological link between BPD and the MD.

Another approach to the putative affective nature of BPD is the study of sleep-electroencephalography (EEG), which has been useful in the biological characterisation of depressive and other psychiatric syndromes (Benca et al., 1992). To our knowledge, there have been at least eight studies on the sleep-EEG in BPD. Most have disregarded or not taken into account the existence of a concomitant Axis I MD in the BPD patients. Bell et al., (1983) described the ‘characteristic short REM latency reported in MD’ in a group of eight Research Diagnosis Criteria (RDC)-primary major depressive patients with borderline symptoms and 11 primary depressive patients without borderline symptoms. McNamara et al. (1984) found sleep continuity disturbances, greater rapid eye movement (REM) activity and shortened REM latency in 10 depressed borderline women and MD patients compared with controls. Reynolds et al. (1985) compared 20 retrospective and prospective BPD (MD was not an exclusion criterion and most patients had RDC diagnoses of depression) to 10 MD patients and to controls; they found similar REM latency values in both patient groups. King et al. (1987) compared 18 BPD (eight with current or past history of MD) patients with 16 MD patients and controls; BPD patients could not be distinguished from controls, whereas MD patients had shortened REM latencies compared with the other two groups. Lahmeyer et al. (1988) found similar EEG-sleep abnormalities in 17 BPD patients (eight also had RDC diagnoses of MD) and 20 MD patients; stage 1 sleep was lower and sleep efficiency was higher in the BPD group. In these five studies, the sleep-EEG perturbations could be more directly linked to a concomitant MD than to BPD itself (Lahmeyer et al., 1989).

Only three studies have clearly considered the existence of MD in their BPD patients. Akiskal et al. (1985) compared 24 non-MD BPD patients (not diagnosed with the Diagnostic Interview of Borderlines) to 30 non-BPD MD patients, 16 patients with other personality disorders, and 14 controls (the non-MD BPD patients had significantly higher scores on the Beck Depression Inventory than the major depressives); they found similar REM latencies in the BPD and MD groups. In the study of Benson et al. (1990), 18 BPD patients with and without a present or past history of affective disorder were compared to each other and to controls. The three groups could not be distinguished in terms of REM latency; the BPD group had less total and delta sleep and more time awake after sleep onset than controls. Battaglia et al. (1993) studied 10 ‘never-depressed’ BPD patients and controls; they found reduced REM latency in the BPD group.

Taken together, the phenomenology and duration of the depressive episodes in BPD, along with their poor response to conventional antidepressants and the findings of some TRH-ST, DST and EEG sleep studies, provide suggestive evidence against the existence of a biological link between BPD and MD. Furthermore, the depressive symptoms observed in BPD patients might not have the same biological substrates as those found in MD patients.

The aim of our without Major Depression study was to further examine the relationships between BPD present and MD patients from the perspective of sleep-EEG parameters and to contribute to the characterisation of the sleep-EEG in BPD patients without Major Depression. Specifically, we wanted to test the hypothesis that, as already proposed by us in studies with the DST and the TRH-ST, sleep-EEG patterns in BPD without Major Depression patients and MD patients would be different, therefore not giving arguments favouring the existence of a biological link between BPD and the MD nor between the substrates of the depressive symptoms in BPD and non-borderline MD.