Digitized analysis of abnormal hand–motor performance in schizophrenic patients

Schizophrenia Research 45 (2000) 133–143 www.elsevier.com/locate/schres

Digitized analysis of abnormal hand–motor performance in schizophrenic patients Peter Tigges *, Roland Mergl, Thomas Frodl, Eva M. Meisenzahl, Ju¨rgen Gallinat, Andreas Schro¨ter, Michael Riedel, Norbert Mu¨ller, Hans-Ju¨rgen Mo¨ller, Ulrich Hegerl Department of Psychiatry, Ludwig-Maximilians-University Munich, Clinical Neurophysiology Section, Nußbaumstrasse 7, D-80336 Munich, Germany Received 3 September 1999; accepted 3 September 1999

Abstract Many studies have shown a high prevalence of discrete neuromotor disturbances in schizophrenic patients. It was hypothesized that these disturbances are lateralized and reflect a neurodevelopmental disorder underlying schizophrenia. A new method for assessing subtle motor dysfunction and hemispheric asymmetries is the registration of hand movements with a digitizing tablet. Using this method, we studied hand–motor dysfunction and its lateralization in schizophrenics, as compared with healthy controls. All subjects (27 schizophrenic patients, 13 of them without neuroleptic medication, the others under neuroleptics; 31 healthy controls) drew super-imposed concentric circles. We computed kinematic parameters reflecting velocity and automatization to quantify neurological soft signs (NSS). The patients had significant impairments of regularity of repetitive hand movements, as compared with the healthy controls (F≥5.35; p≤0.0241). Comparing differences of left- and right-hand performance between patients and controls, we found longer stroke duration (F=(15,98); p=0.000***) and decreased automatization (F=18,14; p= 0.000***), especially on the left side in schizophrenic patients. Measuring hand movements with a digitizing tablet is a sensitive method for assessing subtle motor dysfunction in schizophrenic patients, not reflected in the scores of clinical scales. Our findings show NSS in schizophrenic patients, independently of neuroleptics. Further, the hypothesis of lateralization of cerebral structures generating NSS towards the right hemisphere in schizophrenia is supported. © 2000 Elsevier Science B.V. All rights reserved. Keywords: Digitizing tablet; Hand movement; Laterality; Neuroleptics; Neurological soft signs; Schizophrenia

1. Introduction Neurological soft signs (NSS) are discrete motor and sensory disorders that cannot be linked to special cerebral lesions or dysfunction. * Corresponding author. Tel.: +49-089-5160-5544; fax: +49-089-5160-5542. E-mail address: [email protected] (P. Tigges)

Kraepelin was one of the first psychiatrists who observed NSS in many schizophrenic patients and described them ( Kraepelin, 1919). Schro¨der et al. (1992) could demonstrate that NSS are more often found in schizophrenic patients than in other psychiatric patients or healthy controls. They are independent of neuroleptic medication in this population (Gupta et al., 1995). We also know that there exists a close relationship between NSS,

0920-9964/00/$ - see front matter © 2000 Elsevier Science B.V. All rights reserved. PII: S0 9 2 0 -9 9 6 4 ( 9 9 ) 0 0 18 5 - 1

134

P. Tigges et al. / Schizophrenia Research 45 (2000) 133–143

especially disorders of motor coordination, and negative symptoms in schizophrenia (Heinrichs and Buchanan, 1988; Manschreck et al., 1990; Schro¨der, 1998). Several investigators have studied the asymmetry of NSS in schizophrenic patients, too. Some of these studies support the hypothesis of left-hemispheric deficiencies in schizophrenic patients (e.g., Torrey, 1980). This hypothesis is supported by many morphological and functional studies. Crow (1990) gives an overview of these findings and postulates that abnormally reduced cerebral asymmetries reflect a genetically determined disturbance of cerebral development in schizophrenic patients. However, some authors have reported motor abnormalities related to dysfunction in the right cerebral hemisphere in schizophrenia leading to significantly increased hemispheric asymmetries (e.g., Walker et al., 1994; Niethammer et al., 1998). In most of these studies, clinical rating scales and neurological examinations were used in order to register NSS. Direct kinematic measurements of hand–motor dysfunction in schizophrenic patients offer a new approach. They are an objective tool for investigating functional disturbances reflecting the pathophysiology of psychiatric disorders. In a methodological study, we could show that kinematic handwriting parameters have middle to high temporal stability in healthy subjects and that hand–motor performance is influenced by some variables, especially age (Mergl et al., 1999). Therefore, this variable has to be controlled in studies using digitized analysis of hand movements in psychiatric patients. So far, only a few scientists have used kinematic measurement methods for investigating hand– motor asymmetries in schizophrenic patients. Lohr and Caligiuri (1997) instructed 49 older schizophrenic patients, 6 manic patients, 10 patients with bipolar affective disorders and 30 healthy controls ‘‘to press and hold a stable level of force with their index finger placed on a strain gauge, while following a stable target representing that force on a computer monitor’’ (Lohr and Caligiuri, 1997, p. 196). The coefficient of force variation served as measure for hand-force instability. The hand–motor asymmetry score resulted from subtracting the right-hand force error score from the left one. A

negative score indicates less stability for the right hand, whereas a positive score reflects less stability for the left hand. There were significant differences between the asymmetry scores of the schizophrenic patients and those of the control groups: the asymmetry score of the schizophrenic patients was negative, the score of manic patients was positive, and that of bipolar patients in between. Healthy controls showed an asymmetry score of approximately 0. Lohr and Caligiuri (1997) interpret these results as indicating left-hemispheric dysfunctions in schizophrenia and right-hemispheric dysfunctions in patients with mania. In contrast, Gallucci et al. (1997) found little evidence of abnormal hemispheric specialization in schizophrenia concerning hand–motor function. The authors registered hand movements of schizophrenic patients and healthy controls who drew guirlands and arcades with both their dominant and non-dominant hand using a digitizing graphic tablet. In order to examine the NSS, they computed several kinematic parameters (e.g., stroke length and duration). Only one measure indicated reduced functional asymmetries of hand movements in schizophrenic patients (regularity of stroke duration). These findings do not confirm the theory of altered motor asymmetry in schizophrenia, but they do not falsify Crow’s theory of structural temporal lobe asymmetries in schizophrenia (Crow, 1990): functional asymmetries do not always reflect structural asymmetries and vice versa. In the present study, we used the new method of digitizing handwriting analysis for comparing schizophrenic patients with healthy controls concerning hand–motor dysfunction, assuming greater impairments for the patients. The second hypothesis, derived from the assumption of left-hemispheric dysfunction in schizophrenia, dealt with the differences between hand–motor performance of the dominant and the non-dominant hand. These differences were expected to be significantly reduced in schizophrenic patients compared with healthy controls. In an additional explorative analysis, we examined the relationship between medication, psychopathology and cognitive status on one side, and hand–motor performance in schizophrenia on the other.

P. Tigges et al. / Schizophrenia Research 45 (2000) 133–143

2. Methods and materials 2.1. Subjects We examined 27 partially remitted schizophrenic in-patients diagnosed according to ICD-10 criteria ( F20.x) (Dilling et al., 1993) (21 male and six female) and having mean ages of 31.04 (range 20–54; SD 8.98) years. Their mean illness duration was 7.04 years (range 0.2–27; SD 7.12). All patients had given written consent to their participation, according to the Declaration of Helsinki ( World Medical Association, 1997). Patients with neurological or internistic disease, mental retardation or pre-existing drug abuse were excluded from our study. Of the patients examined, 22 were paranoid schizophrenics (ICD-10: F20.0), two were hebephrenics (F20.1), there were two patients with the diagnosis schizophrenia simplex ( F20.6) and two with an undifferentiated schizophrenia (F20.3). We assessed the patients’ psychopathology using the Positive and Negative Symptom Scale (PANSS ) ( Kay et al., 1988), their positive symptom score being 19.48 (range 10–33; SD 6.42), their negative symptom score 21.15 (range 7–39; SD 8.87) and their global score 41.56 (range 31–59; SD 8.25). In order to determine the degree of cognitive disturbances in the schizophrenic examinees, we averaged their individual scores concerning the item ‘conceptual disorganization’ (P2) (mean: 3.00; SD: 1.11; range: 1–5). Psychiatrists screened the 13 neuroleptic-free patients as to extrapyramidal motor dysfunction ( EPMS) using the Simpson–Angus Scale (SAS) (Simpson et al., 1970) (mean: 1.62; SD: 2.29; range: 0–7) and the other patients using the Extrapyramidal Symptom Rating Scale ( ESRS) (Chouinard et al., 1980) (parkinsonism: mean: 3.93; SD: 1.64; range: 1–7; dyskinesia: mean: 0.07; SD: 0.27; range: 0–1; dystonia: mean: 0.21; SD: 0.43; range: 0–1). The interpretation of the results concerning the scales indicates that the patients treated with neuroleptics had only very low to low EPMS. Patients not under the influence of neuroleptics had no EPMS. All patients were righthanded [determined by the Edinburgh Handedness Inventory (Oldfield, 1971); mean: 89.13; SD: 17.93; range: 30–100].

135

Thirty-one healthy volunteers served as controls and were comparable to the patients for age, gender and handedness. They were recruited from the scientists and their associates of the Department of Psychiatry of the University (LMU ) of Munich. They were not paid for participating in our study. Table 1 presents a review of the most relevant demographic characteristics of the sample. 2.2. Drugs Thirteen patients were not under the influence of neuroleptics. Twelve of these patients had shown insufficient benefits from previous neuroleptic treatment (except clozapine) or suffered from serious side effects of conventional neuroleptics, and one patient was neuroleptic-naive. The duration of the washout phase for the patients previously treated with neuroleptics was between 2 and, maximally, 9 days. The other 14 patients were treated with neuroleptics. Nine patients were medicated with ‘typical’ neuroleptics [haloperidol (2), perazine (2), perphenazine (2), trifluoperazine (2), fluphenazine (1) and flupentixol (1)]. At the time of the study, the mean neuroleptic dose of these patients was 506.94 (SD: 378.27; range: 150–1250) chlorpromazine equivalents according the computational procedure by Hollister (1970). Five patients were treated with atypical neuroleptics: four patients received clozapine (mean daily dose: 343.75±241.85 mg; range: 125–600 mg) and one patient ziprasidone — a drug being not officially permitted for therapeutic purposes in Germany as yet (daily dose: 8 mg). Three patients were also on biperidin because of the development of extrapyramidal motor disorders induced by typical neuroleptics. Six patients without neuroleptic treatment had additional medication (three oxazepam, two lorazepam and one paracetamol ). Of those nine patients treated with ‘typical’ neuroleptics, five patients were also on other psycho-active drugs (one chloralhydrate, one diazepam, one lithium carbonate, one lorazepam and one zopiclon). 2.3. Apparatus and materials We used a digitizing graphic tablet ( WACOM IV ), with a maximal sampling rate of 200 Hz, and

136

P. Tigges et al. / Schizophrenia Research 45 (2000) 133–143

Table 1 Description of the sample Variable

Schizophrenic patients (n=27)

Healthy controls (n=31)

pa

Gender (female/male) Age (years) [Mean (SD)] Educational level Secondary Modern School Secondary School High School University Writing hand (right/left) Handedness (LQd) [Mean (SD)]

6/21 31,0 (9) – 13 2 5 6 27/0 89.1 (17.9)

10/21 33,3 (11,3) – 5 7 3 16 31/0 94.6 (8.7)

0.39b 0.64c 0.01*c – – – – 1b 0.42

a Significance level; p<0.05. b Group comparisons using Pearson’s chi-square tests. c Group comparisons using Mann–Whitney U-tests. d LQ=Laterality Quotient in the Edinburgh Handedness Inventory (Oldfield, 1971).

a spatial resolution of 0.05 mm, in order to record the x–y-coordinates of the subjects’ hand movements. The coordinates were concurrently transmitted to the PC interface. Data registration and the following signal processing were done by means of a commercially available program (CS Version 4.3; Marquardt and Mai, 1992). The subjects had to draw superimposed concentric circles with an approximate diameter of 12 mm as fast as possible, with the dominant and then the non-dominant hand, for 30 s. Thus we were able to register their degree of fine motor dysdiadochokinesia (compare Jahn et al., 1995). 2.4. Data analysis To determine the quantitative parameters of handwriting-motion, the curve, represented in an x, y coordinate system, is divided into time-segments or so-called strokes. A stroke is one halfcycle in the ongoing movement, defined by two sequential extreme points (maxima/minima) of the position curves. The first and second derivations (velocity and acceleration) of each stroke were calculated and smoothed, using non-parametric kernel estimation (Marquardt and Mai, 1994). The endpoints of each stroke correspond to changes of direction and, therefore, to 0-points of velocity. Similarly, the borders of a segment for velocity curves are 0-points of acceleration, so that

these points are identical with the local extreme scores of velocity. For subjects and trials the following motion parameters were calculated: $ mean peak acceleration (mm/s2); $ Mean peak velocity (mm/s); $ mean stroke duration (ms); $ mean stroke length (mm); $ number of changes of direction of acceleration (NCA) per stroke; $ number of changes of direction of velocity (NCV ) per stroke; $ relative standard deviation of peak acceleration (%); $ relative standard deviation of peak velocity (%); $ relative standard deviation of stroke duration (%); $ relative standard deviation of stroke length (%); and $ skewness coefficient (%). The mean peak acceleration and velocity are defined as the arithmetical means of the acceleration and velocity peaks of every stroke. They reflect the basic dynamics of hand movements. Jahn et al. (1995) demonstrated a relative retardation of repetitive hand movement patterns in schizophrenia with this parameter. Since all subjects were instructed to draw the circles as fast as possible, the mean stroke duration is the average amount of time needed to draw one

P. Tigges et al. / Schizophrenia Research 45 (2000) 133–143

half-cycle reflecting possible influences of bradykinesia. The average stroke length represents the way of the pen around the circle, measured per half-cycle, and shows micro- vs macrographia relative to the expected size of the circles. Highly automated hand movements are characterized by bell shaped and smooth acceleration and velocity profiles, being single peaked for every stroke. The parameters NCA and NCV reflect the average number of local extremes of the acceleration and velocity curve per stroke. For single peaked and bell shaped acceleration and velocity profiles, we find the theoretical ideal values for NCA and NCV of 1. Higher values reflect a disturbance of the execution of open loop movements and correspond to multi-peaked acceleration and velocity profiles found in the execution of closed loop movements (e.g., copying a sentence on sight). Eichhorn et al. (1996) used NCA and NCV for handwriting analysis in parkinsonism, showing the detection of dopamimetic effects of drugs on handwriting kinematics in parkinsonian patients. The relative standard deviations of acceleration and velocity reflect the degree of intra-individual variability regarding the acceleration and velocity profile of writing movements. Each parameter is a percentage obtained by the division of the value of the standard deviation of the parameter by its mean value. The relative standard deviation of stroke duration and stroke length are further parameters to assess the regularity of the movements. If the movement replications are totally identical, these scores are 0. These parameters may be understood as variation coefficients of their respective parameters. The skewness coefficient is the ratio of the acceleration phase and the total movement time of each stroke and represents ‘‘the relative proportion of time spent in acceleration during the writing stroke’’ (Gallucci et al., 1997, p. 831). The kinematic parameters permit quantitative analyses of the quality of handwriting movements (for a detailed discussion see Mergl et al., 1999). To test the hypotheses and for further explorative studies, the kinematic parameters were statistically analyzed using SPSS (version 7.5). The dependent variables were submitted to a

137

2×2 analysis of variance (ANOVA) for repeated measurements, in order to compare between groups (patients, controls) with respect to hand (dominant, non-dominant). The main effects of the groups reflect the first hypothesis, that schizophrenics show poorer motor performance than healthy controls. Group and hand interactions indicate abnormal motor asymmetries, confirming the second hypothesis, that the differences between the dominant and the non-dominant hand are lower in the group of schizophrenic patients. Though there were no significant differences between the groups concerning gender, we examined possible interactions between gender and the kinematic parameters via an additional 2×2×2 analysis of variance. In the explorative part, the influence of neuroleptic medication [no neuroleptic medication (n= 13), atypical neuroleptics (n=5) and typical neuroleptics (n=9)] on hand–motor performance was studied by means of Jonckheere statistic, a nonparametric test for group comparisons for non-normally distributed dependent variables (Jonckheere, 1954). In order to test the strength of the former findings, we computed Mann–Whitney U-tests for the comparison of unmedicated and treated patients. We also computed Spearman–Brown correlations between the PANSS scores and the selected hand–motor parameters in order to clarify the relationship between schizophrenic psychopathology and hand–motor disturbances. For the group of patients treated with neuroleptics, we determined partial correlations between clinical variables and kinematic parameters with EPS effects controlled for. Since, in this part of the study, we were interested in explorative analysis rather than confirming or rejecting hypotheses, we did not correct the significance level (5%) by use of Bonferroni’s method (Abt, 1981). We suggest this explorative approach to be adequate for detecting relationships between clinical features and hand–motor performance disorders in schizophrenia.

3. Results Mean kinematic scores from the hand–motor examination are presented in Table 2.

138

P. Tigges et al. / Schizophrenia Research 45 (2000) 133–143

Table 2 Mean kinematic parameter scores for the right and left hands by diagnostic group [values are means (SD)] Parameter

Mean peak acceleration (mm/s2) Mean peak velocity (mm/s)b Mean stroke duration (ms) Mean stroke length (mm) No. of changes of direction of acceleration (NCA)b No. of changes of direction of velocity (NCV )b Relative SD of peak acceleration (%)b Relative standard deviation of peak velocity (%) Relative SD of stroke duration (%)b Relative SD of stroke length (%)b Skewness coefficient (%)

Schizophrenic (n=27)

Healthy controls (n=31)

Group effecta

Group by hand interactiona

Right

Left

Right

Left

2504.2 (1111.7) 103.2 (38.7) 170.5 (55.9) 10.0 (3.2) 2.73 (1.66)

1476.4 (654.5) 76.9 (31.3) 327.2 (83.8) 11.6 (4.6) 6.59 (2.19)

3301.6 (1914.4) 116.3 (39.6) 157.3 (70.8) 9.8 (1.6) 2.08 (1.59)

2023.5 (1130.3) 92.8 (35.4) 238.6 (83.0) 11.1 (2.4) 4.02 (2.14)

0.032* 0.071† 0.005** 0.628 0.001***

0.428 0.320 0.000*** 0.643 0.075†

1.68 (.98)

4.23 (1.63)

1.24 (.59)

2.16 (1.38)

0.000***

0.000***

17.9 (7.7)

27.6 (11.9)

15.9 (15.9)

22.3 (6.3)

0.024*

0.890

13.5 (4.7)

22.5 (10.4)

11.1 (5.4)

18.4 (6.1)

0.021*

0.484

9.6 (6.2)

15.7 (9.3)

6.4 (2.2)

12.7 (3.7)

0.007**

0.227

10.0 (4.6) 49.5 (4.1)

16.6 (11.2) 48.2 (5.6)

7.9 (3.6) 52.1 (4.7)

13.4 (4.4) 51.5 (2.4)

0.111 0.001***

0.178 0.667

a †p<0.10; *p<0.05; **p<0.01; ***p<0.001. b Two-way variance analysis for repeated measures was computed for log-transformed (base 10) variables, because normalization of the values was necessary.

Overall, the repetitive hand movements of schizophrenic patients were less consistent and automatized than those of the healthy controls; time spent in acceleration and mean stroke duration were significantly longer in patients. Anomalous functional asymmetries occur for stroke duration and automatization. Whereas patients’ stroke duration was obviously longer for the left than the right hands, this effect was less apparent for the healthy controls (see Fig. 1). Regarding the degree of automatization reflected by NCV, the right–left hand differences were significantly larger for the schizophrenic patients than for the healthy controls. No significant effects were found with respect to gender, which had no measurable effects on the reported interactions of groups with hand. The Jonckheere statistic revealed that medication did affect the relative intra-individual standard deviation of mean stroke length (J1=2.58; p= 0.01**, two-tailed ) and that of mean peak velocity (J1=2.17; p=0.03*, two-tailed ) with respect to the left hand (see Fig. 2). Stroke length irregularity was least in patients not treated with neuroleptic

medication, greater in patients treated with typical neuroleptics, and highest for patients treated with atypical neuroleptics. Movement velocity irregularity in patients without any neuroleptic treatment and patients under atypical neuroleptics was less than in patients under the influence of typical neuroleptics (see Table 3). These effects are not due to group differences in psychopathology (J1≥76.0; p≥0.09) or cognitive disturbances (J1≥82.0; p≥0.14). The Mann–Whitney U-test showed significant differences between treated and unmedicated patients regarding the regularity of stroke length (U=35.0; p≤0.006**) and velocity (U=44.0; p≤0.02*). Stroke length and velocity were less regular in the group of treated patients. Spearman–Brown correlations between schizophrenic psychopathology measured by use of PANSS ( Kay et al., 1988) and our selected hand– motor parameters indicate that there is no significant relationship between positive, negative and global schizophrenic psychopathology and patients’ hand–motor performance. Special analyses of the correlations between psychopathology

P. Tigges et al. / Schizophrenia Research 45 (2000) 133–143

139

Fig. 1. The bars represent bradykinesia [greater mean stroke duration (ms)] and reduced automatization of movement (higher NCV= No. of changes of direction of velocity per half cycle) for the left and right hand in schizophrenic patients (n=27) and healthy controls (n=31). Asterisks indicate significance of t-test for paired samples (mean stroke duration) and Wilcoxon’s signed-ranks test (dysautomatization of movement) for inner group comparisons between left and right hands.

of the medicated patients and their kinematic measures with EPS effects ( ESRS parkinsonism score) controlled for, revealed no significant relationships either. However, it is apparent that cognitive disturbances (measured by the PANSS subitem P2) are significantly correlated with impairments in simple hand–motor performance in the context of schizophrenia. There are significant negative Spearman– Brown correlations between ‘conceptual disorgani-

zation’ (PANSS: Item P2) and the intra-individual relative standard deviations of mean peak velocity (r=−0.439; p=0.022*), of mean peak acceleration (r=−0.420; p=0.029*) and of mean stroke length (r=−0.396; p=0.041*), relative to the left hand. After controlling the EPS effects in the group of medicated patients, we found only one significant (partial ) correlation between ‘conceptual disorganization’ and the skewness coefficient,

140

P. Tigges et al. / Schizophrenia Research 45 (2000) 133–143

Fig. 2. The bars represent reduced stroke length regularity (increased relative standard deviation of mean stroke length) and velocity regularity (increased relative standard deviation of mean peak velocity) for the left hand in schizophrenic patients treated with atypical neuroleptics (n=5) or typical neuroleptics (n=9), compared with patients without neuroleptic medication (n=13). J1 indicates Jonckheere statistic.

concerning the right hand (r=0.61; p=0.028*). This result reflects prolonged acceleration phases for conceptually disorganized patients.

4. Discussion Our results clearly confirm the first hypothesis that the significantly higher prevalence of NSS in schizophrenic patients can be detected with digitizing graphic tablets analyzing fine motor performance. Particularly, we could demonstrate that the patients’ simple, repetitive hand–motor performance is significantly impaired as to the degree of automatization, mean peak acceleration, mean stroke duration and regularity of velocity, accelera-

tion and stroke duration. The loss of automatization with respect to repetitive hand movements reflects fine motor dysdiadochokinesia (compare Jahn et al., 1995). The schizophrenic patients’ increased hand–motor variability shown during the drawing of concentric circles can be interpreted as a symptom of the neuro-integrative dysfunction considered to be the biological substrate of schizophrenia by Meehl (1990). We attribute these changes to the disease, since the impairments occurred both in medicated and unmedicated patients. We found anomalous functional asymmetries in schizophrenic patients, contrary to Crow’s hypothesis of reduced cerebral asymmetries in schizophrenia. They occur for mean stroke duration and degree of dysautomatization. Regarding these variables, the left–right hand differences were significantly greater in schizophrenic patients than in healthy controls. A significant group by hand interaction was found for NSS, being stronger for the left hand in schizophrenic patients. These neuromotor findings indicate right-hemispheric dysfunction in schizophrenia. However, many morphological and functional studies confirm Crow’s hypothesis of left-hemispheric structural lesions in schizophrenia (compare Crow, 1990). These lesions are related to brain structures not primarily involved in motor processes (e.g., the superior temporal gyrus; compare Barta et al., 1990). Moreover, structural asymmetries are not always reflected by functional asymmetries. For these reasons, our motor asymmetry findings do not contradict Crow’s theory. They are in line with one of the findings of Walker et al. (1994), who found a predominance of motor abnormalities on the left body side in their pre-schizophrenic subjects, considered to be associated with right-hemispheric brain damage. Niethammer et al. (1998) found these asymmetries not only in schizophrenic patients, but also in the non-psychotic monozygotic twins of schizophrenics. They conclude that the lateralization of NSS is genetically determined in schizophrenia. Our results show this lateralization only for some neuromotor variables. The correlation of these kinematic parameters with brain imaging data is necessary in order to eluci-

141

P. Tigges et al. / Schizophrenia Research 45 (2000) 133–143

Table 3 Mean kinematic parameter scores of schizophrenic patients for the right and left hands by medication group [values are means (SD)] Parameter

Mean peak acceleration (mm/s2) Mean peak velocity (mm/s)a Mean stroke duration (ms) Mean stroke length (mm) No. of changes of direction of acceleration (NCA)a No. of changes of direction of velocity (NCV )a Relative SD of peak acceleration (%)a Relative SD of peak velocity (%) Relative SD of stroke duration (%)a Relative SD of stroke length (%)a Skewness coefficient (%)

Unmedicated (n=13)

Typical neuroleptics (n=9)

Atypical neuroleptics (n=5)

Right

Left

Right

Left

Right

Left

2439.2 (1087.2)

1454.1 (585.3)

2242.7 (922.5)

1227.7 (534.0)

3144.0 (1451.1)

1981.9 (851.9)

101.7 (40.5) 183.7 (71.3) 10.2 (3.9) 3.07 (2.13)

78.1 (35.7) 346.9 (80.2) 12.8 (5.7) 6.94 (2.31)

95.4 (26.5) 165.9 (40.2) 9.7 (2.3) 2.48 (1.29)

68.2 (23.4) 344.7 (85.1) 9.7 (2.6) 7.01 (2.23)

121.2 (53.3) 144.2 (18.3) 10.3 (2.9) 2.31 (0.67)

89.0 (33.0) 244.3 (36.9) 11.6 (3.8) 4.89 (0.88)

1.91 (1.23)

4.61 (1.72)

1.53 (0.73)

4.29 (1.72)

1.33 (0.55)

3.15 (0.71)

17.37 (9.95)

25.34 (13.80)

17.08 (3.19)

31.49 (11.39)

20.92 (7.64)

26.25 (5.39)

13.23 (5.62)

18.22 (8.38)

13.11 (1.56)

26.93 (10.58)

15.04 (6.23)

25.82 (12.17)

9.15 (4.42

14.74 (8.61)

7.97 (3.18)

16.88 (12.51)

13.45 (12.02)

16.13 (4.21)

9.79 (5.28)

10.15 (3.60)

9.39 (2.19)

21.58 (11.31)

11.86 (6.31)

24.47 (16.07)

48.8 (3.7)

47.8 (5.8)

48.7 (3.4)

49.7 (5.4)

52.9 (5.6)

46.6 (6.0)

a Two-way variance analysis for repeated measures was computed for log-transformed (base 10) variables, because normalization of the values was necessary.

date the biological substrate of NSS lateralization in schizophrenia. Medication did influence stroke length regularity and variability of mean peak velocity in schizophrenic patients. It is remarkable that stroke length irregularity was higher in patients treated with atypical neuroleptics (predominantly clozapine) and not with typical neuroleptics. Regarding velocity, patients treated with typical neuroleptics showed the highest degree of velocity irregularity, followed by patients treated with atypical neuroleptics. These results suggest a different influence of typical and atypical neuroleptics on hand–motor function via different pathways of neurotransmission. We could not replicate the finding by Gallucci et al. (1997), that medication influences stroke duration regularity in schizophrenia. Perhaps differences in motor tasks (drawing of guirlands and arcades versus drawing of concentric circles) are due to this lack of replication. Furthermore, fast drawing of superimposed circles is supposed to be more automatized than producing guirlands and arcades, and therefore represents other motor demands.

Since the medication for the patients was not randomly assigned in this study, the impairments of the patients treated with atypical neuroleptics may be confounded by pretreatment with typical neuroleptics. The patients in our study were not schizophrenics with first manifestations of their psychosis. Therefore, we must consider them to have been treated with neuroleptics previously. Thus it is not possible to distinguish between premorbid indicators of vulnerability and secondary cerebral dysfunction due to effects of prior medication. Positive and negative schizophrenic symptomatology had little relation to hand–motor function in our patients. The relationship of conceptual disorganization to hand–motor function showed decreased regularity of dynamics and size for disorganized subjects. When EPS effects were controlled for, we found conceptual disorganization to be associated with a significantly prolonged acceleration phase of the right hand. These results can be considered to reflect attention deficits of disorganized patients. There are some issues limiting the value of our

142

P. Tigges et al. / Schizophrenia Research 45 (2000) 133–143

findings. The medicated patients were under the influence of different (classic and atypical ) neuroleptics, which can be expected to have quite different effects on the motor system. Due to the medication history of the patients, the separation of morbogenic and pharmacogenic hand movement disorders cannot be done exactly, even when statistical evaluations seem to suggest clear answers. Our investigation is a pilot study; it is not a longitudinal study which would allow us to decide whether the observed results reflect state or trait markers of the disease. However, there are hints that motor impairments in schizophrenia rather have the features of trait markers (compare Gu¨nther, 1992). Additionally, our sample does not contain any catatonic patients, who are known to manifest severe motor disorders. A further issue is the composition of our control sample, which may turn out as a hypernormal one because of different educational levels. On the other hand, the observed impairments, or side effects, are often below the subjective threshold of perception of the patients as well as the clinicians. Due to this, the assessment of these dysfunctions via rating scales in the clinical routine is difficult or even impossible. These difficulties are reflected by the low average SAS and ESRS scores, and their limited variance. This highlights the sensitivity of our method for the assessment of side effects of atypical neuroleptics on motor function.

Acknowledgements We gratefully acknowledge the assistance of cand. med. Willi Flatz and cand. med. Safet Sokullu in carrying out the hand motor investigations of the schizophrenic patients and Julian Rihl and Marlies Karsch in examining healthy controls’ hand–motor performance. The patients are gratefully acknowledged for their readiness to participate in our study.

References Abt, K., 1981. Problems of repeated significance testing. Control. Clin. Trials 1, 377–381.

Barta, P.E., Pearlson, G.D., Tune, L.E., Powers, R.E., Richards, S.S., 1990. Superior temporal gyrus volume in schizophrenia. Schizophr. Res. 3, 22 Chouinard, G., Ross-Chouinard, A., Annable, L., Jones, B.D., 1980. The Extrapyramidal Symptom Rating Scale. Can. J. Neurol. Sci. 7, 233 Crow, T.J., 1990. Temporal lobe asymmetries as the key to the etiology of schizophrenia. Schizophr. Bull. 16, 433–443. Dilling, H., Mombour, W., Schmidt, M.H. ( Eds.), Internationale Klassifikation psychischer Sto¨rungen ICD-10, chapter V, 2nd ed., 1993. Huber, Bern. Eichhorn, T., Gasser, T., Mai, N., Marquardt, C., Arnold, G., Schwarz, J., Oertel, W.H., 1996. Computational analysis of open loop handwriting movements in Parkinson’s disease: a rapid method to detect dopamimetic effects. Mov. Disord. 11, 289–297. Gallucci, R.M., Phillips, J.G., Bradshaw, J.L., Vaddadi, K.S., Pantelis, C., 1997. Kinematic analysis of handwriting movements in schizophrenic patients. Biol. Psychiatry 41, 830–833. Gu¨nther, W., 1992. Gesto¨rte Hirnfunktionen bei Schizophrenie. Kovac, Hamburg. Gupta, S., Andreasen, N.C., Arndt, S., Flaum, M., Schultz, S.K., Hubbard, W.C., Smith, M., 1995. Neurological soft signs in neuroleptic-naive and neuroleptic-treated schizophrenic patients and in normal comparison subjects. Am. J. Psychiatry 152, 191–196. Heinrichs, D.W., Buchanan, R.W., 1988. Significance and meaning of neurological signs in schizophrenia. Am. J. Psychiatry 145, 11–18. Hollister, L.E., 1970. Choice of antipsychotic drugs. Am. J. Psychiatry 127, 186–190. Jahn, T., Cohen, R., Mai, N., Ehrensperger, M., Marquardt, C., Nitsche, N., Schrader, S., 1995. Untersuchung der feinund grobmotorischen Dysdiadochokinese schizophrener Patienten: Methodenentwicklung und erste Ergebnisse einer computergestu¨tzten Mikroanalyse.. Z. Klin. Psychol. 24, 300–315. Jonckheere, A.R., 1954. A distribution-free k-sample test against ordered alternatives. Biometrika 41, 133–145. Kay, S.R., Fiszbein, A., Opler, L.A., 1988. Positive and Negative Symptom Scale (PANSS ) for schizophrenia. Schizophr. Bull. 13, 21–76. Kraepelin, E., 1919. Dementia Praecox and Paraphrenia. Livingstone, Edinburgh. Lohr, J.B., Caligiuri, M.P., 1997. Lateralized hemispheric dysfunction in the major psychotic disorders: historical perspectives and findings from a study of motor asymmetry in older patients. Schizophr. Res. 27, 191–198. Manschreck, T.C., Keuthen, N.J., Schneyer, M.L., Celada, M.T., Laughery, J., Collins, P., 1990. Abnormal involuntary movements and chronic schizophrenic disorders. Biol. Psychiatry 27, 150–158. Marquardt, C., Mai, N., 1992. CS-Computerunterstu¨tzte Analyse der Bewegungsabla¨ufe beim Schreiben. MedCom, Munich. Marquardt, C., Mai, N., 1994. A computational procedure for

P. Tigges et al. / Schizophrenia Research 45 (2000) 133–143 movement analysis in handwriting. J. Neurosci. Meth. 52, 39–45. Meehl, P.E., 1990. Towards an integrated theory of schizotaxia, schizotypy and schizophrenia. J. Personality Disord. 4, 1–99. Mergl, R., Tigges, P., Schro¨ter, A., Mo¨ller, H.-J., Hegerl, U., 1999. Digitized analysis of handwriting and drawing movements in healthy subjects. Methods, results and perspectives. J. Neurosci. Meth. 90, 157–169. Niethammer, R., Weisbrod, M., Sauer, H., 1998. Lateralization of neurological soft signs is genetically transmitted. Eur. Arch. Psychiatry Clin. Neurosci. 248 (2), S10. Oldfield, R.C., 1971. The assessment and analysis of handedness: the Edinburgh inventory. Neuropsychologia 9, 97–113. Schro¨der, J., 1998. Subsyndrome der chronischen Schizo-

143

phrenie: Untersuchungen mit bildgebenden Verfahren zur Heterogenita¨t schizophrener Psychosen. Springer, Berlin. Schro¨der, J., Niethammer, R., Geider, F.J., Reitz, C., Binkert, M., Jauss, M., Sauer, H., 1992. Neurological soft signs in schizophrenia. Schizophr. Res. 6, 25–30. Simpson, G.M., Angus, C.B., Angus, J.W.S., 1970. A rating scale for extrapyramidal side effects. Acta Psychiatr. Scand. 212, 11–19. Torrey, E.F., 1980. Neurological abnormalities in schizophrenic patients. Biol. Psychiatry 15, 381–388. Walker, E.F., Savoie, T., Davis, D., 1994. Neuromotor precursors of schizophrenia. Schizophr. Bull. 20, 441–451. World Medical Association ( Eds.), World Medical Association declaration of Helsinki. Recommendations guiding physicians in biomedical research involving human subjects. J. Am. Med. Assoc. 277 1997, 925–926.