[Importance and limits of statistical methods in psychiatry]. / Intérêt et limites des outils statistiques en psychiatrie.

Clinical reasoning and statistical treatment of psychiatric data interact to reduce the risk of random diagnosis and prognostic probabilities. Errors of statistical procedures are found in an important amount of papers in the most reliable psychiatric journals. Guide-lines for reporting statistics are given advising the authors about mistaking statistic application. Unfortunately, on a more elementary level, representation of data are often criticizable for they are incomplete, or make difficulties of decoding. Harmonization of data representation is then desirable. Some remarks are expressed about paucity of negative studies in literature.

[Dysthymic disorder and depressive personality]. / Trouble dysthymique et personnalité dépressive.

Depressive personality is a clinic category of Kurt Schneider psychopathic personalities. Despite definition of DSM III Axe II personality disorders criteria is Kurt Schneider' definition of psychopathic personalities, no depressive personality disorders is mentionned in Axe II. Axe I dysthymic disorder is a chronic state of depression with depressive traits. The hypothesis that some depressive personality disorders are included in this category cannot be rejected. More generally, DSM III classification regulation discriminate states, but, less operationally, traits which include affective variables with questionnable inter-rating reliability.

[Delayed-action neuroleptics and maintenance treatments]. / Le cas des neuroleptiques à action prolongée et des traitements d'entretien.

Neuroleptic therapeutic trial is a short time experience, usually lasting three months, or even less. A reference treatment (sometimes a placebo) or the studied neuroleptic are given respectively to two randomized samples from the same population. One variable is fixed, chemotherapy, and the other therapeutic variables--psychologic and socio-familial--are supposed to be constant during that short interval of time, though they are linked to the evaluation criteria. In a long term perspective, this robust methodology is questionnable. Psychosocial variables cannot be abstracted from the experience, and the treatment itself, volens nolens, is a composite variable in which biologic and psychologic variables interact. The singularity of the neuroleptic efficacy cannot be measured. Short term neuroleptic treatment answers a first question: how to cure a schizophrenia? Long term treatment, and incidentally long acting neuroleptics, respond to a second question: how to cure and prevent schizophrenia in one particular schizophrenic patient? Some empirical guidelines are given.

Striatal D2 dopaminergic receptors assessed with positron emission tomography and [76Br]bromospiperone in untreated schizophrenic patients.

Striatal D2 dopaminergic receptors of 12 drug-free schizophrenic patients and 12 normal subjects were investigated with positron emission tomography and [76Br]bromospiperone. Patients were classified according to DSM-III criteria, and their clinical symptoms were rated according to Andreasen's negative and positive symptom scales. The ratio of striatal to cerebellar radioactivity was taken as an index of striatal D2 dopamine receptor density. There was no significant difference between the control subjects and the overall schizophrenic group and no significant relationship between this index and the symptom ratings. However, state-dependent variables could partly account for the striatal D2 receptor density variability.

[Caffeine and EEG mapping: effects of visuo-spatial task in healthy volunteers]. / Caféine et cartographie EEG: effets d'une tache visuospatiale chez des volontaires sains. Stratégie d'analyse des données électropharmacologiques.

A group of 10 subjects, normal volunteers (6 men, 4 women), mean age 27.2 years, right-handers, has been selected for a double-blind study of caffeine effects (400 mg per os) versus placebo. One week delay separated the two EEG sessions quantified by spectral analysis. During each session (16 EEG channels, common reference), each subject was recorded four times: 2.5 min, in eyes-closed condition, under diffuse attention, followed by 2.5 min, during a visuo-spatial task ("eyes open" condition: presentation of a picture). These two EEG recordings were obtained before the oral administration and repeated 1 hr after the administration of caffeine or placebo. Individual and group results presented globally: 1) an increase in mean frequencies of alpha activity in "eyes closed" condition, and of delta activity in "eyes open" condition; 2) a decrease of alpha activity amplitudes (both RMS and % values) in both conditions; 3) a decrease of total RMS amplitudes in "eyes closed" condition; 4) an increase of relative % amplitudes for beta 2% fast activity in the "eyes open condition". These results characterize the psycho-stimulant effects of caffeine, especially over the left temporal area which was confirming an initial statistical hypothesis of specific cerebral local activation. The non-parametric permutation tests of Fisher, were not always reaching statistical significance for the same EEG channel quantified by one of the 17 analyzed spectral parameters, except on the left temporal area. During the visuo-spatial task, the decrease in alpha amplitudes (RMS and %) was also statistically significant, but over larger areas: over left and right temporal, central and parietal regions. In both conditions, eyes closed and eyes open, the alpha RMS amplitude was also decreased over the right anterior frontal area.

[Cerebral imaging and depressive disorder]. / Imagerie cérébrale et dépression.

Cerebral imaging of depressive syndromes has been studied since the early eighties in search of organic or functional anomalies in the central nervous system. There are fewer studies than with schizophrenia. Computed tomography, scintiscan, positron emission tomography (PET) are being used in various depressive states. For methodological reasons, results with PET are only preliminary. EEG mapping, a new technique, has, above all, measured the abnormalities of organic brain syndromes, especially dementia, which has to be ruled out in the diagnosis of depression. CT scan has not substantiated any cerebral defects in depressed patients in the comparative case-control studies, except for some clinical types (aged population or forms associated with delusional or hallucinatory syndromes) where images of cortico-subcortical atrophy have been observed. Scintiscan and PET scan have shown a decreased cerebral blood flow, with a Xenon 133 test, and PET scan a decreased consumption of glucose on condition that depressed subjects are cognitively resting, in comparison to matched controls. In EEG mapping an index of spacetiation of the basal quantitative EEG activity typifies cases of evolution towards dementia, which has to be ruled out in the diagnosis of involutional depression.

[Value of quantitative EEG and EEG mapping in medicine]. / Intérêt de l'EEG quantitatif et de la cartographie EEG en médecine.

Quantitative EEG consists in digitising the EEG from one or several leads and then mathematically processing the signal usually by Fourier spectral analysis. A reduction of the digital information gives characteristic parameters (frequency, amplitude or power, asymmetry, etc.), which can themselves be averaged and submitted to different statistical analysis. Quantitative EEG is particularly suitable for clinical electropharmacology for the assessment of the profiles of psychotropic drugs, for pharmacokinetic correlations of dose-effect relationships and studies of drug bioavailability. The signals obtained from schizophrenic and depressed patients have well defined quantitative EEG characteristics: dissymmetry of the amplitude and hypovariability of the EEG signals. EEg mapping is performed by computer processing of 16 simultaneous EEG lead recordings. This new form of medical imaging is fast, relatively economical and non-invasive, and it is used in psychophysiology to study cognitive tasks and the states of vigilance and sleep. It is also used to study the EEG topography of cerebrovascular accidents, brain tumours, cranial trauma, cerebral degeneration and dementia, and the EEGs of psychiatric patients on psychotropic drugs. EEG mapping represents not only a static and statistical approach by drawing the averaged maps of groups of patients but also a dynamic approach by the recording of sequential individual maps presented as an animated film. The results of this method can be correlated with other medical imaging techniques used to investigate the central nervous system.

Comparative study of four evaluation criteria in a clinical trial with schizophrenic patients.

In order to compare the level of agreement between four evaluation criteria, data from an 18-month clinical trial with 181 schizophrenic patients were reanalyzed. In that study, assessments of the subjects were carried out with two psychiatric rating scales, Brief Psychiatric Rating Scale and Nurses' Observation Scale for Inpatient Evaluation (BPRS and NOSIE), and two clinical evaluations (psychiatrist and nurse). Results show that there was total agreement on the four outcome measures for only half of the subjects. After further analyses, we observed that the best agreement was between the two clinical assessments. The NOSIE had poor agreement with the other criteria and appeared less sensitive to moderate levels of improvement in the patients' clinical status. These results are discussed with respect to reliability and validity issues.

Electroencephalographic Cartography. I. By means of mini- or microcomputers. Reliability and interest of this electrical non-invasive brain imagery.

We have recorded control subjects, neurological and psychiatric patients (n = 217 sequences recorded over both hemispheres). Minicomputers (HP Fourier analyzer, HP 1000) for spectral analysis provided 10 spectral parameters over 5 spectral frequency bands (delta, theta, alpha, beta 1, beta 2, raw EEG). For each recorded sequence, 90 EEG maps could be computed over both hemispheres. Topo-EEGs were stored in an EEG image data bank. An EEG mapping microcomputer system linked with a digital polygraph (Alvar Electronic, REEGA 2000) has been used simultaneously. White noises have been fed into both computers for testing spatial resolution. Ten topo-EEGs have been recorded in control subjects and patients. The microcomputer system has provided very reliable topographical results when compared to similar maps generated by the minicomputer. A common average reference has been used. First clinical applications have been studied (brain strokes, brain tumors). The method appears very reliable in comparison with CT scans.

Bouffée délirante: an examination of its current status.

Bouffée délirante is an historic and unique French diagnostic term for a short-lived psychosis. The key diagnostic features are acute, floridly psychotic symptoms with complete remission. Its use in a Paris hospital has been examined, and it appears that the term is declining in popularity. A case-controlled study indicated that the diagnosis is likely to be given to migrants on first admission. A re-diagnosis of case summaries indicated no particular correspondence of bouffée délirante to any one ICD category. However, the usefulness of having a special diagnostic term for psychosis with a good outcome is discussed.

[Topographic aspects of drug-induced fast rhythms in quantitative topo-electroencephalography]. / Aspects topographiques des rythmes rapides médicamenteux en topo-électroencéphalographie quantitative.

The topographical aspects of fast rhythms induced by different psychotropic drugs have been studied in patients by means of a previously described method. This method is based on the spectral analysis of 28 EEG channels over both cerebral hemispheres. The spectral parameters, computed over 2 min 40 sec, are put into a data base. The clinician can then decide which of these parameters will be extracted from the data base and then displayed by means of maps on a graphic terminal. Those maps are useful for immediate comparison not only of the fast rhythms' characteristic values but also of those from other spectral bands. As quantitative EEG analysis is more and more frequently used in psychotropic drug studies, a good knowledge of the topography of the rhythms induced by these drugs will help in the selection of the most useful derivations.

[Computerized quantitative topo-electroencephalography]. / Topo-électroencéphalographie quantitative par ordinateur.

A method is described which has been developed to topographically display the values of spectral EEG analysis from 11 electrodes on each hemisphere and 5 electrodes on the midline. The EEG from 16 channels at a time is multiplexed and sampled at 3.2 kHz. From power spectra averaged every 10 sec for 5 min a few spectral parameters are computed among which are the spectral intensities within the delta, theta, alpha and beta bands. The whole values are then displayed on maps representing lateral views of cerebral hemispheres by means of spectral iso-intensity surfaces with extrapolation of the intermediate points. Different spectral intensities are coded by means of different gray levels. These maps clearly illustrate the topographical distribution of EEG rhythms and their peculiarities related to the side of the head recorded.

[Value of quantitative electroencephalography in psychiatry]. / Intérêt de l'électroencéphalographie quantitative en psychiatrie.

Computerized quantitative EEG processes the signal in four sequential steps: 1) EEG sampling with rejection of artefacts; 2) Fourier analysis (power spectrum according to frequency); 3) data reduction with computation of spectral parameters in five-minute recordings on at least four EEG channels (mean frequencies, mean amplitudes in microvolts and %, with standard deviations for delta, theta, alpha and beta frequency bands); 4) data base and multivariate statistical analyses. Five main results are obtained by computerized quantitative EEG analysis in psychiatry and clinical pharmacology: 1) longitudinal studies of patients before and after treatment; 2) transversal studies of homogeneous EEG groups (discrimination between tracings of residual-type or paranoid-type schizophrenic patients) comparatively with control groups (high and low alpha subjects); 3) study of psychotropic drugs: profiles, prediction of action, dose-effect curves, pharmacokinetic studies, biodisponibility, correlations with plasmatic levels and behavioural or clinical psychiatric rating scales; 4) typologic classification and discrimination by multivariate analysis, studies of functional activities: functional interhemispheric lateralisation of amplitudes and its variations during the course of psychotic processes; 5) new techniques: computerized topo-electroencephalography (CT-EEG, BEAM technique) based on 16 EEG recordings, comparison with other imaging techniques.

Prevalence of psychosis in epilepsy as a function of the laterality of the epileptogenic lesion.

Retrospective analysis of the psychiatric diagnoses in a group of patients surgically relieved of medically intractable epilepsy tested the hypothesis that patients with left-sided temporal lobe epileptogenic lesions are at greater risk for the development of a so-called schizophrenic-like psychosis than are those with right-sided temporal lobe epileptogenic lesions. The data confirmed the hypothesis and also demonstrated an increased prevalence of sinistrals in the psychotic group. Thus, epilepsy involving the dominant hemisphere at the inception of the seizure disorder is the significant risk factors. The data also indicated that a psychosis is unlikely to develop in patients with other (nontemporal) forms of focal epilepsy. On the basis of these data and data from other studies, the prevalence of psychosis in patients with poorly controlled temporal lobe epilepsy was estimated to be approximately 10% to 15%.

Intra- and interhemispheric EEG differences quantified by spectral analysis. Comparative study of two groups of schizophrenics and a control group.

In this work we have studied comparatively the relationships between alpha intensity, alpha frequency, root-mean-square amplitude and their standard deviations, versus the location of the four recorded EEG channels: left and right rolando-parietal, left and right parieto-occipital. Twenty-four spectral parameters were computed for each 5-min EEG recording, in the eyes closed situation. The hebephrenic group (36 EEGs from six patients) presented higher values of alpha intensity and RMS amplitudes with low coefficient of variation when compared with the other groups. The left/right rolando-parietal amplitude ratio was 0.73 versus 1.52 for the parieto-occipital ratio. An amplitude hypovariability in the central regions, probably associated with hyperarousal, especially in the left hemisphere, can then be observed in hebephrenics, together with high alpha intensity on the left parieto-occipital area. The two groups of schizophrenics differ from the control group (27 EEGs from 12 subjects) in terms of spectral parameters, the hebephrenic group presenting a possible interhemispheric dysfunction.