Our data of COMT genotype and allele frequency in our control group is in agreement with the data of Parliament [5] for Russian population (49% - COMT*H, 51% - COMT*L) and for some other European populations (Finns, Irish).

Literature data of the COMT gene investigations in schizophrenia and behavioral disorders are controversial. R.D. Strous et al. [12] reported that schizophrenic patients (N=37) who were homozygous for the low activity allele (COMT*L) were judged to be at higher risk for aggressive and dangerous behavior that those who were homozygous for the high activity allele (COMT*H).

H.H. Lachman et al. [3] reported association between the COMT*L/*L genotype and aggressive behavior in schizophrenic and schizoaffective patients (N=55). Control group consisted of patients without aggressive behavior.

M. Kotler et al. [14] investigated the character of COMT genotype distribution in 3 groups: homicidal schizophrenic patents (N=30), nonviolent schizophrenic patients (N=62) and control subjects (N=415). Significant excess homozygosity of COMT*L/*L was observed only in homicidal schizophrenic group compared with healthy control group. K.A. Nolan et al. [15] revealed that the COMT*L allele was more frequent in males with schizophrenia who had attempted suicide by violent means.

These findings support a hypothesis of common neurobiological substrate for self- and out-directed aggression. In the present study we found the Val158/108Met polymorphism of the COMT gene was associated with both self - and out - directed aggressive behavior in subjects without diagnosis of schizophrenia. According to family, twin and adoption studies patients with schizophrenia tend to have an increased risk of aggressive behavior [18]. We suggest that catecholaminergic alterations may contribute to these behaviors independently from such psychiatric disorder as schizophrenia.

All above-mentioned studies have reported a significant association between the COMT low activity genotype and allele and aggressive behavior in schizophrenic patients. However, G. Jones et al. [19] found association between the COMT*H/*H genotype and higher level of aggression in schizophrenic patients (N=180). So, our data is in agreement with this study. We suggest that the COMT*H/*H genotype is a marker of increased risk for schizophrenia (OR=2.56, 95%CI 1.60-4.12), both out-directed (OR=1.84, 95%CI 1.08-3.16) and in-directed aggressive behavior (OR=2.28, 95%CI 1.39-3.77).

But in some studies association between the COMT genotypes and schizophrenia were not seen [20, 21]. Some reasons for controversial results can be possible. Mental diseases — this combination of various symptoms, which, unfortunately, not always allow dividing the patients precisely using certain phenotype criteria. Problem of exact and authentic diagnostics of mental distresses seems to be especially important. Differential diagnostics in psychiatry is based on careful analysis of all symptoms and traits of disease, complaints of the patient, of behavior features, etc. Also the structured questionnaires and quantitative diagnostic criteria are used frequently in psychiatric practice. Some diagnostic systems are too circumscribed and specialized; on the contrary, others include superfluous quantity of assessments. Therefore validity of any diagnosis remains the main problem in psychiatry.

Secondly, the different explorers dealt with groups of the patients with a various degree of aggressive behavior and probably with other forms of schizophrenia. In the present study persons with extremely high level of aggression (murder, numerous physical offences, attempt of suicide by violent means) and patients with paranoid schizophrenia were included.

Thirdly, in comparison with samples of the patients in the previous researches of mental distresses, the sample of the persons in our work was much more: 236 patients with paranoid schizophrenia, 182 suicide attempters, 182 violent offenders and 203 healthy subjects.

At last, the controversial results can be explained by features of genetical structure of populations, which the samples of the surveyed persons were enrolled from.