| Abstract: | PURPOSE: To determine the effect of schizophrenia on patient handwriting and to catalogue handwriting deformities typical of schizophrenic graphism. The additional purpose is to explore and ascertain whether the self-inspired handwriting samples of schizophrenic patients contain language phenomenons typical of schizophasia.
METHOD: Two study groups were recruited for the handwriting analysis, each of them comprising 30 subjects. The patient group consisted of subjects diagnosed with schizophrenia, according to the behavioural criteria outlined in ICD-10. All they have been treated by the specialist psychiatrist for a relatively long time and have been receiving neuroleptic treatment for at least 5 years. The collected handwriting samples were provided by the individuals diagnosed with schizophrenia: women (age range 3050 years) and men (age range 30-55 years). The precise diagnoses in this study groups included paranoid schizophrenia (17 subjects), simple schizophrenia (11 subjects) and undifferentiated schizophrenia (2 subjects). Therefore, the collected study material encompassed 30 handwriting samples of 15 men and 15 women. The mean age of subjects in the study group was approx. 44.6 years, with the mean age in the women subgroup of approx. 45 years and the mean age in the men subgroup of approx. 44 years. In order to conduct the comparative analysis of graphism traits in patients with schizophrenia, the handwriting samples were also collected from the control group, consisting of 30 healthy individuals, who never used any mental health care provisions before and did not present with any detectable mental abnormality. The controls were matched to the subjects in terms of sex, age and level of education. Eventually, the material provided by 30 individuals (15 women and 15 men) was analysed, with the mean age in the women subgroup of approx. 43 years and the mean age in the men subgroup of approx. 55 years. Two types of written samples were collected: Same content samples - all subject were requested to write down two stanzas and a chorus of Polish national anthem. The study samples enabled the author to analyse the effect of schizophrenia on the handwriting thus allowing the attempt to determine what the handwriting traits of schizophrenia sufferers are. Non-subject-specific, free piece of handwriting, requested only from the individuals diagnosed with schizophrenia. The analysis of this handwriting sample was meant to
detect the mental state alterations, such as impaired judgement and conclusion, with particular emphasis on delusional attitudes. Based on the handwriting variable system and guided by the operational definitions of the variables provided by T. Widła, the analysis of 29 specific variables and 178 corresponding sub-variables was performed. These pertained to four aspects of graphism, namely its structure (form), alignment, as well as the form and degree of connection. Both study groups were included in this test – the experimental group composed of subjects diagnosed with schizophrenia and the control groups composed of healthy individuals. In the author's opinion, it is only the comparative research, including the control group consisting of healthy individuals without the obvious signs and symptoms of schizophrenia, to enable the actual comparison of differences between the graphism samples provided by both groups. Thus, only comparative research makes it possible to determine what is really specific of the patients as compared to the healthy individuals. Only this research plan enables drawing conclusions regarding cause-andeffect relationships between the handwriting traits and psychosis while minimising the potential for errors. The ex post facto approach was used for the empirical analysis of the manuscripts. A supplementary table was developed for each subgroup and each measured variable. The presence or absence of an individual variable was recorded in the table using the binary system. Next, the results obtained for each variable group were entered in the corresponding four-fold table. The strength of association between the handwriting variables and the dependent variables (these graphic aspects which give the manuscript the form typical of schizophrenic sufferer) was measured using the Kendall correlation coefficient Q. This association was further tested for significance using Pearson's chi2 test. The following were considered critical values: Q coefficient of 0.5 and alpha value over 0.05 (df = 1). These critical values did not limit the number of preliminarily confirmed hypotheses concerning the association between the analysed variables and the diagnosis of schizophrenia in an individual to provide the manuscript for the analysis. On the other hand, it also limited the potential for errors and wrong choices. After the handwriting traits which are associated with the writer's psychosis significantly enough to be considered suggestive of schizophrenia were identified and catalogued, the issue of accuracy and reliability of the catalogue in everyday use was addressed. For this purpose, the test typically employed for the assessment of sensitivity and reliability of a given diagnostic method was used. Hence, it was assumed that the catalogue of schizophrenia-specific handwriting traits is a diagnostic test, allowing for diagnosing schizophrenia in an individual to provide a handwriting sample. The following were calculated: sensitivity and specificity of the diagnostic test, the positive and negative predictive value, prevalence rate, positive and negative likelihood ratio and accuracy. Additionally, the neurological and psychiatric aspects of non-subject-specific handwriting samples provided by schizophrenia sufferers were assessed. This approach enabled evaluation of these aspects of handwriting which may reflect neurological
abnormalities (i.e. tremor causing letters and words to be deformed) and mental abnormalities, especially personality psychopathology, signs of schizophasia referred to as linguistic pathology typical of schizophrenia, including but not limited to the lack of logic, derailment, echolalia and all other language phenomena linked to language disintegration.
RESULTS: Handwriting traits of schizophrenia sufferers 1. Out of 178 variables listed in the catalogue developed during the research, 7 are linked to the writer's psychosis significantly enough to consider them typical of schizophrenic patients. They are itemized below, in the order of the correlation coefficient (Q) magnitude.
2. The seven aforementioned traits should be considered characteristic of handwriting who provided the handwriting samples in this study group. The control experiment also gave some findings, that is three traits, considered to be linked significantly to schizophrenia, which have not occurred in our control handwriting samples provided by the healthy individuals. Hence, the main traits to differentiate the manuscripts of subjects with schizophrenia from the ones of the healthy individuals were the angular ending connective form of the letter "n" as well as the space after the letters "ą" and "ę". The other graphism traits found in the samples provided by the schizophrenic subjects also occurred in the handwriting of the healthy individuals. Nevertheless, the frequency of occurrence was significantly lower or they were plainly rare.
The accuracy and reliability of the schizophrenic handwriting trait catalogue application 1. 90% of subjects diagnosed with schizophrenia, whose manuscripts were assessed for the presence of trait from the predefined graphism catalogue, were diagnosed correctly. This means that their manuscripts were classified correctly as the handwriting of a schizophrenic sufferer. 2. 44.6% of healthy individuals with no formal mental health problem were diagnosed correctly. This means that their manuscripts were classified correctly as the
No. Variable Q value
chi-squared value
Significance level (p)
1.
The angular ending connective form of the letter "n"
1 12 0,001
2. Space after the letter "ą" 1 5,45 0,025 3. Space after the letter "ę" 1 4,29 0,05
4.
The angular beginning connective form of the letter "n"
0,83 6,41 0,025
5. Wavy baseline 0,65 8,15 0,005 6. Slender handwritten letter form 0,6 5,93 0,025
7.
The angular beginning connective form of the letter "m"
0,56 4,8 0,05
handwriting of a healthy person, based on the absence of traits from the predefined graphism catalogue. 3. The positive predictive value is 62.79%. Hence, discovering a single trait from the predefined catalogue in a handwriting sample gives 62.79% of probability that an individual actually suffers from schizophrenia. 4. The negative predictive value is 82.35%. Hence, the absence of traits from the predefined catalogue in a handwriting sample gives 82.35% of probability that an individual does not actually suffer from schizophrenia. 5. The possibility that the individual classified as schizophrenic sufferer based on the handwriting analysis (i.e. the sample contained at least one trait typical of schizophrenia) actually suffers from the disease is 1.68-fold higher than the possibility that the healthy individual, with no mental health pathology, will have the positive result of the handwriting assessment. 6. Therefore, the possibility that the individual classified as healthy based on the handwriting analysis (i.e. the sample did not contain any trait typical of schizophrenia) actually suffers from the disease equals just 21.42% of the possibility that the healthy individual, with no mental health pathology, will have the negative result of the handwriting assessment. 7. The likelihood of the correct diagnosis (schizophrenia sufferer vs. healthy individual) made using the described handwriting analysis for the graphism traits typical of schizophrenia listed in the said catalogue is approximately 68.33%.
Neurological and psychiatric abnormalities reflected in handwriting 1. Neurological abnormalities are caused not only by the organic (structural) damage to the nervous system but also as a result of functional impairment of the complex, highest level neural structures responsible for speech and language. The former were reflected in the study material as the letter and word deformity, the absence of evenness or homogeneous slant and the presence of amplification, thus symptomatising tremor and ataxia. The latter, though, were mostly symptomatised by agraphia (omitting letters or words in the text or inserting wrong letters or words into it), perseveration (on the level of characters, syllables and words) and paraphrasia.
2. The research partly confirmed the observations made by other scholars studying the handwriting of schizophrenia sufferers. The quality of handwriting of these patients deteriorates with the main changes including:
a severe decrease in handwriting legibility, its aesthetic aspects and the linear alignment, b spelling individual words the way they are pronounced, c uneven handwriting, including numerous corrections, d omitting the graphic elements of letters, including diacritical marks in particular, e punctuation errors (mostly including the absence of punctuation marks),
f numerous spelling mistakes. 3. The analysis of non-subject specific handwriting samples provided by the subjects was intended to indicate the psychiatric disorders manifesting as the handwriting abnormalities. However, the conclusions seem to be relatively pessimistic from this point of view. Except for manuscript # 10, whose content appears to be the incongruent and nonsense scribbled piece in a form of a word salad, full of bizarre associations and pseudo-religious pathos, all other handwriting samples (except for manuscripts #22 and #27) do not demonstrate significant mental process abnormalities. The manuscripts #22 and #27 include the weird and an incomprehensible juxtaposition of fact as well as the atypical logics and conclusions. The other manuscripts were neither unnatural nor presented with stilted speech, weird pathos, symbolism and associations or unusual style and exaltation. The pressure to write (graforrhea) was not observed, neither was weird juxtaposition of facts and conclusions, addressing pseudo-philosophical topics, perseverations or racing thoughts. 12 subjects categorically refused to provide a non-subject-specific handwriting sample. In a single case, the patient with obvious imperative auditory hallucinations and the delusional reality interpretation, which he had verbalised for some time, refused to provide a 'non-subject-specific' handwriting sample, explaining that the 'voices' had forbidden him to do so. However, others despite their illness provided manuscripts which appeared normal in the assessment and did not reveal any psychopathology.
CONCLUSIONS: 1. The attempt to differentiate the graphism of schizophrenia sufferers and healthy individuals ended with a partial success; it also confirmed to some extent the previous findings published by other researchers. The empirical and statistical analyses led to identification of seven handwriting traits, which are closely and significantly related to schizophrenic psychosis. The claim that the said variables must be present in each and every handwriting sample provided by each and every schizophrenia sufferer is obviously wrong. The set of variables is not too extensive to make a good test for presence/ absence of schizophrenia in an individual. However, it is beyond all doubt that the direction of the research is the right one, without the need to generalise that "the graphic signs of schizophrenia" occur in the handwriting of every schizophrenia sufferer.
2. Three handwriting traits from the developed catalogue (a straight ending connective of the letter "n"; the angular beginning connectives of the letters "n" and "m") make the schizophrenic handwriting take a form of the so-called angular, rigid handwriting. which is the trait typical of schizophrenia, according to many authors. When looking at the degree of connection, the spaces were found after the letters "ę" and "ą". Undoubtedly, the degree of connection is an instrument to record the handwriting proficiency. It is also a sensitive factor reflecting the effect of diseases, age, psychophysical fitness etc. on the handwriting of an individual. Hence, the schizophrenic psychosis is reflected by taking the pen/ pencil out of the sheet when
writing letters including diacritical marks. How to explain the aforementioned phenomena? The author postulates that one of the factors impacting the the angular connections of letters "m" and "n" is the long-term effect of anti-psychotic medications. As a result of neuroleptic intake, the extra-pyramidal symptoms (EPS) develop, which manifest as muscle rigidity and stiffness. As a result the patient fails to perform the fluent and accurate fine motor manipulation which is a part of handwriting. Instead, the schizophrenia sufferers tend to simplify the letter formation pattern and use angular connectives instead of garland or arcade ones. This can also explain spaces occurring after the letters "ą" and "ę". Due to fine motor disorder, the patient cannot write the diacritical mark in a continuous flow motion, without taking the pen (pencil) out of the sheet. Besides, as the absence of diacritical marks is also typical of schizophrenia, it is likely that the subjects had initially written the letters "ą" and "ę" without the cedillas and having spotted the mistake they corrected their handwriting. Therefore, it is also likely that the mental disease impairs proper associations between the graphic symbols and their phonetic counterparts and as a result the above symptoms emerge in the handwriting. 3. The majority of presented results is not compliant with the published findings of other researchers. The hypotheses suggesting the inclinations of individuals with schizophrenia to produce handwriting of variable slant, elongated ascendents or descendents, falling lines, ascending or descending baseline, wavy lines or increased spaces between words and rows were not confirmed. How can this be explained? The authors of these reports when did not have a control group consisting of healthy individuals when analysing the handwriting of schizophrenia sufferers. The graphism of subjects with schizophrenic psychosis was their only point of focus. As a matter of fact, it is only the comparative research protocol, including the control group consisting of healthy individuals without the obvious signs and symptoms of psychopathology, to enable the actual comparison of differences between the graphism samples provided by both groups. Thus, only comparative research makes it possible to determine what is really specific of the patients as compared to the healthy individuals. Assuming such research plan allows for statistical analysis and drawing conclusions concerning the cause-and-effect relationship between the handwriting traits and psychosis. It also minimises the risk of explaining the observed associations by means of alternative theories and hypotheses. 4. The assessment of the provided manuscripts against the neurological criteria showed the abnormalities also observed by W. Chłopicki and J. S. Olbrycht. Tremor, ataxia, word and letter deformity, lack of evenness (size of individual letters, slant uniformity), agraphia (omitting letter or words in the text, or inserting wrong letters or words into it), perseverations and paraphrasia were found in the analysed handwriting samples. 5. The analysis of patient's non-subject-specific handwriting pieces against the psychiatric criteria showed mental state alterations in 3 cases only. Others, despite their illness, provided manuscripts which appeared normal in the assessment and did not reveal any psychopathology. How can this be explained? This may, for
instance, be attributed to the relatively good adaptation skills of these patients, which are able to adjust their production to social norms. On the other hand, they might intentionally try to mislead others, especially the person collecting their written verbal production samples. However, the author argues that the key factor contributing to such outcomes was medical treatment used in all patients. Medical treatment of schizophrenia mainly affects its positive symptoms such as delusions, hallucinations, disorganized speech and behaviour. It affects the negative symptoms of schizophrenia, such as anhedonia or autism, to much lesser extent. All subjects had been on the anti-psychotic treatment for at least 5 years. Therefore, it can be assumed that the treatment they received proved to be effective in preventing symptoms of schizophrenia, especially resolving axial symptoms and disorganisation. Additionally, it promoted relatively good subjective and objective social function of the affected individuals. Even if the above statement is too farreaching, it should be noted that even the use of inadequately effective medications in chronically ill patients (and so were our subjects), limits the destructive effect of disease symptoms and improves the objective and subjective measures of real-life functioning To sum up, apart from the three described manuscripts, the study did not confirm language pathology or symptoms of thought disorder as key characteristics of written verbal production of schizophrenic patients. Perhaps different conclusions could be drawn from the speech analysis of patients with schizophrenia who are not treated. However, further research is necessary to investigate into the matter in a more conclusive manner. 6. From the methodology perspective, the research contributed to the insight into the language pathology in schizophrenia and in particular into the handwriting abnormalities found in the sufferers' manuscripts. The ultimate solution to the problem of graphism and linguistic pathology in schizophrenia is still ambiguous, if it can be solved at all. And so it is going to remain, unless the psychiatrists and neuroscientists determine conclusively what the actual mechanism of schizophrenia development is. |