Brain oscillatory responses in patients with bipolar disorder manic episode before and after valproate treatment

Abstract

Background

GABA/Glutamatergic dysfunction and neural circuits which regulate cognitive processing are involved in the underlying pathology of bipolar disorder. Event related oscillatory neuroelectrical activity reflects integrative brain functioning, different frequency bands representing different cognitive functions.

Methods

Event Related Potentials to visual odd-ball paradigm in ten manic/hypomanic medication free, DSM-IV bipolar patients were measured before and after six weeks of valproate monotherapy in comparison to ten sex and age matched healthy controls. Different frequency band responses were obtained by digital filtration of ERPs. Young mania rating scale (YMRS) was used to assess clinical response. Repeated measures ANOVA, Wilcoxon and Mann Whitney U tests were used for statistical analysis.

Results

Patients showed significantly higher baseline occipital beta (18–30 Hz) (p: 0.014) response than healthy controls. They were devoid of the occipito-frontal alpha (8–13 Hz) dominance presented by the control group. Occipital beta response reduced significantly (p: 0.009) and became similar to controls after treatment. Post-treatment alpha responses were significantly lower than baseline in anterior temporal (p: 0.038) and occipital (p: 0.027) locations. Healthy controls displayed a significantly increased frontal alpha response at the second assessment but the patients did not. Mean YMRS score reduced significantly compared to baseline at the end of six weeks (p: 0.004).

Conclusions

Alpha response is the universal operator in the brain. Increased occipital beta response in mania may be compensatory to the dysfunctional alpha operation. Its reduction after valproate may be through modulation of glutamatergic and GABAergic mechanisms and indicate medication’s corrective effect on the underlying pathogenesis.

Introduction

Bipolar disorder is a chronic illness with a relapsing and remitting course. Mania is the core feature of the illness which gives rise to the definite diagnosis (DSM-IV, 1994). Manic state is characterized by increased energy and motor activity, decreased need for sleep, distractibility with a strong involvement of pleasure seeking and impulsive behavior. Manic patients also display signs of dysfunction in attentional measures, complex processing and memory as well as emotional processing. Having an acute episode of mania or depression is suggested to give way to damage to learning and memory systems (Bearden et al., 2001). Pathology has been shown to involve decreases in cortical thickness in multiple cortical areas which are associated with sensory function as well as emotional and cognitive processing such as left cingulat cortex, left middle frontal cortex, left middle occipital cortex, right fusiform cortex and bilateral postcentral cortices (Lyoo et al., 2006).

Among a wide range of neurotransmitters that are involved in bipolar disorder, Gamma amino butyric acid (GABA) plays an important role. It spreads in neural networks that are involved in cognitive and emotional processing and modulates noradrenergic, dopaminergic and serotonergic local neural circuitry (Brambilla et al., 2003). GABAergic interneurons which are the core component of cortico-limbic circuitry were found to be defective in cerebral cortex of bipolar patients (Benes and Berretta, 2001). Findings pointing to cortical inhibitory deficits were thought to be a neurophysiologic evidence for an association between bipolar disorder and disrupted cortical GABA related inhibitory neurotransmission (Levinson et al., 2007). Several studies revealed low plasma (Berettini et al., 1983, Kaiya et al., 1982) or cortical (Bhagwagar et al., 2007) GABA activity or altered genetic expression of GABA (Guidotti et al., 2000, Heckers et al., 2002) in bipolar disorder. Low GABA activity was thought to be a genetically determined trait creating a vulnerability to development of either mania or depression with contribution of environmental factors and suggested to return to baseline levels with remission (Petty, 1995). GABAergic activity is reciprocally regulated by dopamine, hyperactivity of which also plays a role in mania (Yatham et al., 2002). Alterations in modulation of the dopamine system may trigger the appearance of a defective GABA system (Benes and Berretta, 2001).

Valproate is an effective antimanic agent (Bowden, 2003). Evidence supports a GABA potentiating mechanism of action of valproate (O’Donnell et al., 2003). Valproate was shown to augment the ability of atypical antipsychotic medications to increase dopamine (DA) and acetylcholine (ACh) efflux in the rat hippocampus and medial prefrontal cortex (Huang et al., 2006). It was also shown to lead to a significant reduction in presynaptic dopamine function in manic patients. This was thought to be related to improvement in manic symptoms (Yatham et al., 2002). It regulates cell survival pathways such as cAMP-responsive element binding protein (CREB), Brain Derived Neurotrophic Factor (BDNF), bcl-2 and mitogen-activated protein kinases (MAP) which may underlie its neuro-protective and neurtrophic effects (Xiaohua et al., 2002, Löscher, 2002).

Oscillations constitute the most obvious observable type of electrical activity in the brain in response to well-defined sensory or cognitive events. Event related oscillations are either “evoked” (phase locked) or “induced” (temporally related) to the event and can be easily recorded from the scalp. After off-line eradication of any interference –vascular or muscular in origin– remaining EEG sweeps can be digitally filtered and oscillations are classified according to the natural frequencies of the brain such as delta (0.5–3.5Hz), theta (3.5–7 Hz), alpha (8–13 Hz), beta (18–30 Hz) and gamma (30–70 Hz) (Basar et al., 1999). In a more simple way, the oscillatory activity occurring in different frequencies can be defined as the building blocks of the P300 response. They provide new evidence to be real signals of the CNS. Oscillations are selectively distributed, controlling the integrative brain functions at all sensory and cognitive levels (Başar et al., 2001, Mountcastle, 1992). This suggests a paradigm change in neuroscience. It is now possible to achieve measurements on scalp electrodes of human subjects under various states of behavior and learning. Early experimental studies on large scale brain activity showed superposition of multiple oscillations in delta, theta, alpha, beta, gamma in various parts of the brain. This raised the necessity to use brain's multiple oscillatory activities for the analysis of all brain functions in both animals and humans (Başar, 1980, Başar et al., 1975). Generation of P300 to visual target stimuli involves frontal brain structures such as orbito-frontal cortex, anterior cingulate cortex as well as deeper brain structures such as hippocampus/parahippocampal areas, the insula, the temporal lobe and thalamus (Herrmann and Knight, 2001), the anatomic structures that constitute the mostly affected neural circuitry in bipolar disorder as shown by imaging studies (Soares and Mann, 1997, Strakowski et al., 2005). As to the neurotransmitter involvement, GABAergic interneurons and pyramidal cells were found to build and maintain complex interconnections which lead to large scale network oscillations, such as theta, gamma (40–100 Hz), and ultrafast (200 Hz) frequency bands (Benes and Berretta, 2001).

There has been limited number of earlier electrophysiology studies in symptomatic bipolar patients. A prolonged P300 latency and reduced P300 amplitude was found to be equivocal and most probably related to psychosis (Salisbury et al., 1999). It was once suggested to have an association with an underlying frontal lobe pathology as previously shown by different neurocognitive and imaging studies (Salisbury et al., 1999). A more recent study showed abnormal high frequency synchronization in response to auditory stimulus (O’Donnell et al., 2004) in manic bipolar patients.

The aim of this study was to assess oscillatory brain activity in manic phase of bipolar disorder before and after treatment with valproate as monotherapy in comparison to healthy controls. We hypothesized that the patients would show hyper-responsive low frequency oscillatory activity before treatment and valproate monotherapy would lead to reduction in the oscillatory responses. To our knowledge, this is the first study assessing oscillatory brain activity in a prospective and controlled design.