Meditation is known to achieve a particular class of altered states of consciousness. These states are experienced by the subject and are objectively measurable. Brain activity of meditation has been examined by EEG (electroencephalogram) for nearly 50 years, and various results are accumulated. Brain electric measurements reflect diverse subtle parameters of the working mechanisms of neural networks that constitute the different functional states of the brain.

Many meditation practices are designed to increase awareness towards occurring experiences by sustaining attention to the present moment in a detached manner. The meditation state itself is characterized by a detached monitoring of ongoing experience. Meditators claim that continuous meditation practice leads to a changed state of mind in everyday life. It is believed that long-term meditation training correlates to increased awareness and greater detachment. A detached attitude with reduced initial emotional evaluation of events is described. Indeed, meditation in general is practiced by many as a self-regulatory approach to emotion management. That is, non-engaging observation with no intent to analyze or judge or form expectations.

Several popular meditation exercises involve sustained attention. The exercises can be roughly divided into two categories depending on the focus of attention which can be on one single object or on no object. In 'focused attention' meditation, the focus of attention is kept on a single object (like the perception of one's own breathing, a mantra or a visualization); when a distracting thought occurs and is detected, the instruction is to move the focus of attention back to the primary object. In 'open monitoring' meditation, there is no object of attention. The instruction is to keep an open awareness and non-reactively monitor all internal and external experiences. A common characteristic of focused attention and open monitoring meditation is the above described detachment. It is conceivable that prolonged meditation training carries detachment over into the everyday lives of meditators and becomes an integral part of it after some time.

The question arises whether physiological training effects of long-term meditation practice can be found in meditators during no-task resting. We hypothesized that the neuroplasticity effects of meditation, specifically increased awareness and detachment, would be detectable in a no-task, eyes-closed resting state. By probing for differences in the resting between meditators and controls, we expected to find physiological evidence for the described detachment in the form of inhibited appraisal systems as well as increased awareness of themselves and their surround in the brain of meditators.

Considering the reviewed reports of various affected brain regions by meditation, we used LORETA functional tomography (low resolution electromagnetic tomography) for the analysis of EEG data in order to obtain unambiguous brain localizations. This method does not simply ascribe brain electric signals from a given head site to directly underlying brain sources, but computes intracerebral distributions of source strength from the head surface-recorded multichannel EEG data. Moreover, in order to provide experimental settings that are less invasive and more natural to daily situations, portable EEG device was used. It had minimum restriction of subjects' postures and movements.

EEG was recorded during task-free resting in 8 Qigong meditators and 9 normal healthy controls. 19 EEG electrodes were applied at the positions of the international 10/20 system using a electrode cap. All subjects were self-declared right-handed. The meditators had an average meditation experience of 11.5 years, whereas the controls had no experience in meditation. The age, gender and education level of the two groups had no statistical significant difference. And all subjects reported to have had no previous or current psychiatric diagnosis, head trauma or drug usage, and did not use any central active medication. After complete information about the study design, the subjects gave their written consent. The study was approved by the ethics committee of the University of Tokyo (#1364).

Off-line, scalp EEG recordings were computed for intracortical source localizations by applying LORETA software. The manually selected epochs were firstly digitally filtered via Fourier Transformation into the seven independent frequency bands which is determined by factorial analyses: Delta (1.5-6 Hz), Theta (6.5-8 Hz), Alpha-1 (8.5-10 Hz), Alpha-2 (10.5-12 Hz), Beta-1 (12.5-18 Hz), Beta-2 (18.5-21 Hz), Beta-3 (21.5-30 Hz) and Gamma (35-44 Hz). And 3-dimensional electric neuronal distributions were linearly calculated by solving the inverse problem by finding maximal smoothness of all solutions. It means that LORETA assumes equivalent generator strength among neighboring voxels. And it represents electrical activities as power of current density. These generators have no restriction for number of dipoles, locations or orientations within the brain. It applies a three-shell spherical head model (scalp, skull, and brain compartments). The brain compartment is restricted to the cortical gray matter, co-registered to the Talairach probability brain atlas and digitized Montreal Neurological Institute MRI template. The solution space is sampled at 7mm resolution, resulting in a total of 2394 voxels.

LORETA functional images were computed for each subject, separately in each of the 8 frequency bands. The LORETA functional images were scaled to unit average total power, i.e. the average of the power values over all voxels was scaled to unity (for each given subject and frequency band separately). Such a scaling procedure, commonly used in fMRI and PET imaging, has the effect of decreasing non-physiological sources of variance. The differences in brain electric activity between groups were assessed by exceedence proportion tests performed on LORETA images of t-statistics. These voxels were attributed to the corresponding Brodmann Areas (BAs), based on their MNI coordinates.

Comparing resting between meditators and controls, only the delta frequency band revealed significant differences in the exceedence proportion tests (p<0.05, corrected for multiple testing). Delta activity is known to correspond with functional inhibition. In meditators, parts of prefrontal cortex (BA 9, 10, 11, 44, 45, 46 & 47, preferably right lateralized) and anterior cingulate cortex (BA 32) showed inhibition (stronger delta activity), whereas motor, somatosensory (BA 4, 6 & 7) and visual association cortices (BA 18 & 19), left superior temporal gyrus (BA 22), left precuneus (BA 31), left temporo-parietal junction (BA 22, 39 & 40) and bilateral fusiform gyrus (BA 37), extending on the right to the parahippocampal gyrus (BA 30) showed stronger activation (less delta activity) compared to controls.

Thus, during resting, those brain areas involved in reflection about current experiences (BA 10), attending to subjective emotions (BA 32), emotion regulation (suppression of sadness: BA 11; reappraisal of sad film excerpts: BA 9, 10, 11, 32 & 47), anticipation (BA 11), expectation (BA 11), semantic searching and decision making (BA 32, 9 & 47) and analogical reasoning (BA 11) were all inhibited in meditators as compared to controls. This inhibition of appraisal systems in the brain is in line with a detachment from experience, defined by an attenuation of analysis, judgment and expectation, implying a lesser degree of processing of ongoing experience. Interestingly, in normal subjects, increased delta and theta EEG in the right but not left prefrontal cortex (BA 8, 9, 10 & 46) during pre-experimental resting correlated with greater propensity for later decreased cognitive control measured as risk-taking. This supports our conclusion that increased delta in the same areas indicates increased detachment in the meditators.

The areas with stronger activity (reduced delta activity) in meditators mostly comprised the unimodal sensory association cortices and the multimodal temporo-parietal junction and the secondary motor cortex, all part of a network known to be involved in the detection of changes in the sensory environment. In particular, the temporo-parietal junction has been related to the integration of multimodal sensory information about the own body and the body in space. The precuneus is an important part of the default mode network and is associated with monitoring the external world. Thus, the meditators keep those areas activated during resting that are involved in the detection and integration of internal and external sensory information: this suggests that they stay more aware of their internal and external sensory environment than controls.

In search for evidence of an effect of continued meditation practice, the hypotheses were supported by the results of LORETA functional tomography, suggesting a training effect of long-term meditation practice leading to a trait change in non-meditating states. The appraisal systems of the meditators were inhibited and the brain areas involved in the detection and integration of internal and external sensory information showed increased activation.

Two further points should be noted: Firstly, resting is not the same for meditators and non-meditators, because meditators often report an inability for non-meditative resting. After years of practice it apparently becomes quite impossible for them not to meditate. It has been proposed earlier that a lack of differences between meditation and resting in meditators could be due to this inability for 'normal' resting. The differences of resting state characteristics between meditators and non-meditators should be clarified by ratings of their subjective experiences in future studies. In which way this quasi-meditation resting state in meditators differs from actively engaged meditation states is another issue to be addressed in future studies. As discussed above, the observed EEG effects of long-term meditation when interpreted as possibly resulting from the trait change could well account for the meditators' inability for 'normal' resting. The observations give additional support to neuroplasticity effects in meditation in a growing body of findings as reviewed in the introduction. A second point that warrants caution in the interpretation of the observed differences is the possibility of trait differences between subject groups. Only a longitudinal study starting before the uptake of meditation practice could shed light on this issue.

In sum, the findings of differing brain-electric activation patterns during no-task eyes-closed resting between Qigong practitioners and normal controls reflect the subjective reports of experienced meditators of an acquired increased awareness and greater detachment from ongoing experience during non-meditation times in everyday life, suggesting a training effect of continued meditation practice.

本研究は、既に臨床応用が行われ始めている瞑想法に着目し、この長期訓練効果を脳機能の変化としてとらえる事を試みたもので、下記の結果を得ている。

1. 瞑想実践群と瞑想をしない健常群の安静閉眼時の脳波から、それぞれの群の3次元脳内電位発生源を調べると皮質抑制を示すとされるデルタ波(1.5-6.0 Hz:低周波帯域)のみで2群間に有意差が認められた。

2. 瞑想をしない健常群に比較し、瞑想実践者では、脳前頭部(前頭前野の一部、前帯状回)でデルタ活動上昇、そして、頭頂/後頭部(運動野、2次体性感覚野、視覚野)でデルタ活動減少が認められた。デルタ活動は、皮質の抑制化を示すものと考えられている為、瞑想実践群では、瞑想をしない健常群に比較し、脳前頭部が皮質活性化、頭頂/後頭部では抑制化が、推測された。これらの脳部位の局在機能に鑑み、瞑想群では、高度な認知活動中ではなく、安静閉眼中では、現在体験中の行為に対する、熟考、主観的感情の評価、分析、論理的理解などの機能が、健常群に比べて減弱化している可能性が推測されうる。そして逆に、健常群に比べて、体外、体内環境からの、感覚情報の収集、統合の機能は増強している可能性が推測される。この結果は、長期瞑想者の主観である、現在進行中の出来事への無執着化、知覚の鋭敏化の傾向と一致し、日常生活時にも 持続されている事が推測されうる。そしてこれは、一過性の状態(state)変化から持続的変化(trait)の可能性を示唆するものと考えられる。

3. 瞑想の、attention training(意識を注意や集中でコントロールするトレーニング)としての側面に注目し、その訓練効果を脳の変化(trait)として、とらえられる可能性が示唆された。

以上、本論文は、特に心療内科領域における認知行動療法の長期的効果の解明に重要な貢献をなすと考えられ、学位の授与に値するものと考えられる。