Introduction

Movement disorders (MD) such as tardive dyskinesia (TD) and extrapyramidal-symptoms as parkinsonism and akathisia are common adverse effects of long-term treatment with antipsychotics (AP). Especially typical AP are known to cause severe MD [54, 77, 80, 86], whereas atypical AP are increasingly being used in juvenile patients [35, 39, 66, 84], due to numerous advantages concerning the incidence of MD, negative schizophrenic symptoms as well as heightened quality of life [7, 24, 64]. We had initially reported on the prevalence of MD in adolescent patients with onset of schizophrenia in childhood or adolescence. Patients treated with atypical AP displayed significantly fewer symptoms of parkinsonism and - as a trend - of akathisia compared to patients with typical AP [41].

While relations between MD occurrences as symptoms of schizophrenia itself and psychopathological symptoms have been discussed–e.g. in the case of catatonic schizophrenia [15, 71] or regarding motor compulsions in prodromal and residual schizophrenic phases [83], relations between AP-induced MD and schizophrenic psychopathological symptoms have rarely been analyzed in detail. In general, there are numerous studies which assessed MD as well as psychopathological symptoms but did not focus on relations between both. These studies are predominantly restricted to adults [e.g. 16, 28, 33] including residual and/or chronic schizophrenic patients [e.g. 12, 30]. In contrast, some studies focused more strongly on relations between MD and psychopathology, but were as well limited to adults [1, 11–13, 17, 18, 20, 21, 27, 29, 31, 34, 38, 48–52, 55, 56, 58, 62, 67, 70, 72, 75, 87, 88, 94–98]. The mean age (estimated from the raw data presented in the studies) is of about 40 years and mostly long-term treated patients were included in these studies. Interestingly, in their study of 45 adult patients with schizophrenia spectrum diagnoses (with an neuroleptic-untreated subgroup of 23 patients at admission), Peralta and Cuesta [75] found that negative, parkinsonian and catatonic symptoms are highly related features, while depressive symptoms seem to be a relatively independent dimension of psychopathology in schizophrenia. The authors hypothesized, that there may be a specific behavioural syndrome under AP treatment which may, at times, be clinically indistinguishable from negative, parkinsonian and catatonic symptoms.

Studies on adolescents as well have mostly described psychopathology and MD symptoms only separately without examining relations between both in detail [e.g. 44, 61, 77, 79, 80, 101]. However, there is, to our knowledge, only one study in 34 adolescents reporting on the relationship between AP-induced MD and schizophrenic psychopathological symptoms [57]. The aim of the current study was to investigate relations of MD with psychopathological findings in a large sample (n = 93) of adolescents who receive predominantly atypical AP for the treatment of schizophrenia spectrum disorders. Characteristics of MD and the use of AP in this sample have been described elsewhere [41].

Methods

Subjects

In brief, 93 patients with schizophrenia or schizoaffective disorder [per DSM-IV] with onset in childhood or adolescence recruited from a rehabilitation center participated in this cross-sectional/retrospective study (see Table 1). Patients were treated on average for 3.6 ± 1.9 years with antipsychotics throughout the course of illness. Three patients (3.2%) were never treated with atypical antipsychotics. At the time of examination 76 patients (81.7%) were being treated with atypical antipsychotics (clozapine 66.7%, olanzapine 10.8%, risperidone 2.2%, sulpiride 2.2%), 10 patients (10.8%) with typical antipsychotics (perazine 3.2%, haloperidol 2.2%, zuclopenthixol 2.2%, flupentixol 1.1%, fluphenazine 1.1%, pimozide 1.1%) and 7 patients (7.5%) with a combination of typical and atypical antipsychotics (clozapine 3.2%, risperidone 3.2%, olanzapine 1.1%; haloperidol 3.2%, levomepromazine 2.2%, perazine 1.1%, thioridazine 1.1%). The mean treatment duration with this medication at time of examination (current antipsychotic treatment) was 2.1 ± 1.4 years meaning that patients were at stable drug regiment. 4 patients (4.3%) were additionally under treatment with SSRIs, 3 patients (3.2%) with tricyclic antidepressants, 4 patients (4.3%) with mood stabilizers. Mean intelligence assessed retrospectively from clinical records was 92.2 ± 13.7 (no statistical gender difference; t-test). The study was approved by the Ethics Committee of the University of Marburg. Further characteristics of this sample are described elsewhere [41].

Table 1 Characteristics of the patient groups (with/without MD symptoms) and of the total study sample (n = 93; 52 male, 41 female)
Full size table

Assessment

Psychopathological symptoms were assessed using the Schedule for Assessment of Negative/Positive Symptoms (SANS-K/SAPS-K; 3–6, 85) and the Brief Psychiatric Rating Scale (BPRS; 73) including the 5 subscales “anxiety/depression”, “anergia”, “thought disturbance”, “activation”, “hostility/suspiciousness”. MD symptoms were assessed using the Tardive Dyskinesia Rating Scale (TDRS; 90), the Abnormal Involuntary Movement Scale (AIMS; 47), the Extrapyramidal Symptom Scale (EPS), a German version of the Simpson Angus Scale [89] and the Barnes Akathisia Scale (BAS; 8, 9). All used instruments are sufficiently reliable and valid [2, 5, 6, 8, 9, 22, 23, 45, 46, 53, 68, 74, 90]. Further details on assessment features and patient groups are given elsewhere [41].

Statistics

To investigate relations of MD with psychopathological findings categorial (group comparison), dimensional (Pearson correlations) and multivariate (superfactor-analysis) statistical approaches were used. Group differences in psychopathology between patients with motor symptoms (n = 37; M-group) and without motor symptoms (n = 56) were studied using two-tailed t-tests or chi-square tests. For the total sample (n = 93) correlations between sumscores of the scales and subscales were evaluated with the Pearson correlation coefficients. Results were calculated as explorative measured values. Furthermore, a superfactor-analysis (with varimax-rotation; n = 93) was performed for the scales’ sumscores. All p values were two-tailed; the significance level was set at .05. All statistical analyses were performed using the Statistical Package of the Social Sciences (SPSS 10.0 for Windows) software.

Results

At the time of examination the group of all patient with motor-symptoms with either TD-, parkinsonism- or akathisia-symptomatology criteria (n = 37; see 41) showed significantly higher scores in the SANS- und SAPS-scales (Student’s t-test; p = 0.026 and p = 0.033 respectively), the BPRS-scale (p = 0.001) and the anergia-subscale of the BPRS (p = 0.005) as well as heightened sumscores regarding the BPRS-subscales anxiety/depression and thought disturbance (p = 0.051 and p = 0.066, respectively), if compared to patients without motor-symptoms (see Table 2).

Table 2 Sumscores of the psychopathology scales, which showed significant differences or trends towards differences between the patient groups with or without MD symptoms, respectively
Full size table

Patients of the group with MD symptoms (n = 37) displayed a significantly younger age at onset of schizophrenia (p = 0.019) and at onset of antipsychotic treatment (p = 0.012) compared to patients without MD symptoms (n = 56) (see Table 1). However, they did not significantly differ with respect to duration or type (typical/atypical) of antipsychotic treatment.

Table 3 shows trends as well as significant correlations between the MD-scales (TDRS, AIMS, EPS, BAS) and the psychopathology scales (SANS, SAPS, BPRS, subscales of BPRS) of the total study sample.

Table 3 Trends and significant correlations between MD and psychopathology scales (n = 93)
Full size table

A superfactor-analysis of the sumscores of all MD- and psychopathology scales of the total study sample revealed four factors, explaining 70.2% of variance within the scales: 1. “negative symptoms/parkinsonism/global psychopathology”, 2. “productive symptoms/thought disturbances”, 3. “tardive dyskinesia”, 4. “akathisia/activation” (Table 4).

Table 4 Factors of superfactor-analysis according to sumscores of MD- and psychopathology scales (n = 93)
Full size table

Discussion

Although MD symptoms are a common phenomenon for patients with schizophrenia under long-term AP treatment, associations between MD symptoms and psychopathological symptoms have been little examined solitarily in adolescent schizophrenic patients treated predominantly with atypical AP. Compared to patients without motor-symptoms all patients with MD symptoms (n = 37) showed higher scores within the psychopathology scales (p = 0.001–0.033), indicative for their aggravated illness; they showed mainly more anergia-symptoms (p = 0.005) as well as tendencies towards more anxiety/depression- (p = 0.051) and thought disturbance symptoms (p = 0.066). This could be explained by patients with worse global psychopathology being treated more often with AP and therefore showing more MD symptoms. Though patient groups did not significantly differ with respect to duration or type (typical/atypical) of antipsychotic treatment, patients with MD symptoms were significantly younger at onset of schizophrenia (p = 0.019) and antipsychotic treatment (p = 0.012), what might suggest, that these patients are predisposed to an increased disease severity with earlier onset of schizophrenia, possibly due to a higher genetic loading.

A superfactor-analysis performed on the scores of the MD- and psychopathology-scales revealed four factors (“negative symptoms/parkinsonism/global psychopathology”, “productive symptoms/thought disturbances”, “tardive dyskinesia”, “akathisia/activation”), which may indicate the existence of different syndrome-like entities as already suggested by Peralta & Cuesta [75].

Relations between movement disorders

Tardive dyskinesia. Dyskinetic symptoms correlate with global psychopathology (p = 0.040), what goes in line with studies in adult patients [49, 67]. Further, in our study dyskinetic symptoms showed a trend towards correlation with negative symptoms (p = 0.068) and symptoms of anergia (p = 0.065), what has also been shown in most of the adult study samples [17, 18, 38, 49, 58, 72, 87, 95–98, overview: 94], whereas other authors could not reveal any linkages [11, 52]. Studies on relations between positive symptoms and TD are comparably rarer [57]. In their study of a small sample (n = 34) of adolescent patients, Kumra et al. [57] found more pronounced positive symptoms in patients with dyskinetic symptoms. Based on the small sample size, these results might be less representative than those found in the current study sample (n = 93). Yuen et al. [100] suggest that the presence of TD may be associated with positive symptoms, while the severity of TD may be related to negative symptoms. According to Waddington et al. [98], the association between negative symptoms and late onset dyskinesia increases with age. A study conducted in untreated adult patients with schizophrenia found no correlations between dyskinesia and psychopathology [62]. Evidence from multiple lines of study indicates that mood disorders, particularly depression, are a risk factor for developing comparably severe TD under AP treatment [26].

Parkinsonism. In the current study, parkinsonism symptoms show correlations with negative symptoms (p = 0.036) and anergia (p = 0.023), a finding which accords with other studies [13, 21, 31, 34, 27, 51, 75]. We found no associations with depressive or aggressive symptoms. Results of former studies about associations of parkinsonism with depressive symptoms are ambiguous [1, 34, 55, 56, 75]. Cheung et al. [29] found more severe parkinsonism side effects in highly aggressive patients when compared to patients with low aggressive behaviour. On the other side it is well known that primary idiopathic Parkinson’s disease goes hand in hand with emotional instability and slowed psychic functions, similar to motor functions.

Akathisia. Akathisia symptoms are correlated with the global psychopathology (p = 0.080) and symptoms of hostility (p = 0.013). As well, two studies reported significant correlations between aggressive behaviour and akathiform symptoms [29, 50]. Some studies revealed connections of akathisia with negative [17, 87] or with depressive symptoms [12, 20], whereas a study of Halstead et al. [48] failed to describe these connections. We could also not identify such relationships, possibly due to the low prevalence of fully pronounced akathisia in our study group (one patient with akathisia and additional 10 patients with subthreshold akathisia symptoms). Nair et al. [70] showed akathisia to predict an increased psychopathology, especially regarding symptoms such as “activation” and “thought disturbance”, what is partly reflected by factor 4 of the superfactor-analysis of our study.

Pathogenic considerations about relationships of MD and psychopathology

Neurochemical hypotheses for TD, parkinsonism and akathisia include changes in dopaminergic, cholinergic, GABAergic and noradrenergic systems, as well as a genetic vulnerability [59, 66, 81, 102]. The anatomical correlative of the linkage between psychopathological and MD phenomena can be seen in the basal ganglia. Especially the striatum is assumed to regulate both motility and mental functions [25]. A dysfunction of the dopaminergic and glutamate system, for example, is likely to result in positive schizophrenic symptoms [25, 53, 76], whereas an imbalance of serotonin and noradrenalin appears to be linked to negative symptoms [14, 60]. Psychomotor activity is determined by interactions of glutamate, dopamine, noradrenaline, serotonin, and acetylcholine [25]. Pharmacological blockage of the D2-receptor may have AP effects within the limbic system [42], although it may be the reason for MD, due to an imbalance of dopamine and acetylcholine [35, 59].

Relationships between psychopathology and MD might be explained by different reasons. First, it is generally assumed that psychopathology results from the pathophysiology of schizophrenia whereas motor symptoms represent side effects of AP treatment. According to this assumption, bias effects based on the cooccurence of more residual symptoms with long-term AP treatment could play a role in relationships between psychopathology and MD. One could hypothesize that patients, which are treated–due to their highly pronounced psychopathology–with high doses of AP, show an enhancement of MD symptoms compared to patients with a reduced global psychopathology. In addition, not only motor but also psychopathological symptoms may be influenced by AP by affecting the basal ganglia [35]. One of the main results of the current study indicating that an increased MD symptomatology is associated with pronounced global psychpathological signs may contribute to this hypothesis.

Second, a common basic pathophysiological principle of both entities could also be postulated, while motor symptoms might not only reflect an integral part of the schizophrenia disease process [31, 62, 83] but also of a specific (genetic) dispostion for the development of drug-induced MD [59, 102] and might thus be triggered by AP [49]. Taking into consideration, that triggering a (genetic) disposition for MD by antipsychotics may be related to the extent of a long-term antipsychotic treatment, this point does not necessarily exclude a common pathophysiology of both entities, which may be manifested in the significant connection of both, MD and global psychopathological symptomatology pronouncement in the current study. In former studies there are as well hints, that antipsychotic medication has not obligingly be present to induce MD symptoms. Reduced premorbid functioning and withdrawal correlate with more negative symptoms and increased extrapyramidal symptoms, even in untreated patients [51, 91]. Imbalances in transmitter pathways could result in motor dysfunctions as well as psychopathological symptoms [25, 32, 36, 40, 58, 65, 75, 81, 99]. As well, neurodegenerative processes may be involved in the pathogenesis of TD, e.g. decreased plasma brain-derived neurotrophic factor levels in schizophrenic patients with TD have been found [92]. However, though TD do obviously occur spontaneously in untreated patients with schizophrenia [27, 37, 38, 43, 51, 62, 93], in two studies no correlations with psychopathology were found in these patients [27, 62]. Honer et al. [51] found a tendency for correlation of the severity of MD and negative symptoms in untreated patients.

Third, one could speculate that psychopathological symptoms could in some respect represent secondary reactions to MD symptoms. It has been shown that MD may lead to especially depressive symptoms resulting in a reduced subjective quality of life [12, 34, 82]. According to a recent study under conventional AP treatment subjective well-being depends strongly on motor side effects, whereas in atypically treated and drug-naive schizophrenic patients, subjective well-being is mainly influenced by psychopathological status [78]. Patients of our study sample were mostly treated with atypical AP (92.5%) and MD symptoms in our study sample were generally less pronounced [41], the induction of e.g. depressive symptoms seems rather unlikely. Vice versa, it is known that psychopathology can also lead to MD phenomenons, e.g. anxiety resulting in motor activation.

Limitations

The limitations of the current study have been discussed in detail elsewhere [41]. In brief, due to the cross-sectional/retrospective design causal relationships cannot be inferred and statistical data represent explorative data; we therefore refrained from correcting for multiple testing and as well from covariate analysis due to insufficient retrospective data (e.g. considering dosage data). The listed p-values are therefore merely of an explorative nature. As well, correlations coefficients are comparably low, possibly due to the low prevalence and severity of MD symptoms in this study sample under treatment with predomaninantly (92.5%) atypical antipsychotics, so that sumscores of MD sumscores are mostly very low (see 41). However, the naturalistic character of the study represents the typical clinical features of relations between MD symptoms and psychopathology in this patient population. Second, the generalizability of this study may be limited by the fact that adolescents undergoing rehabilitative treatment were included, which makes it even more likely for MD of a long-term AP treatment and residual symptoms to occur together. Third, some motor and psychopathological symptoms may show partially overlays (e.g. akathisia and activation), whereas other symptoms are clearly differentiated from each other (e.g. tremor and social withdrawal). Thus, for the reason of clear differentiation psychomotor testing with specific reliable and valid instruments has been done (see 10). However, it can not be excluded that there might be syndrome-like entities as revealed by our explorative correlation tests and our superfactor-analysis and as has previously been described by Peralta and Cuesta [75]. Fourth, concerning depressive symptoms only a subscale of BPRS was used, while we suggest a specific depression instrument for future studies.

Conclusions

The current study conducted under naturalistic conditions in the, to our knowledge for the given topic, so far largest sample with adolescent patients undergoing a long-term treatment with predominantly atypical AP suggests that MD symptoms and psychopathology are in some respects connected. These relations could be due to [1] AP treatment, [2] pathophysiology of the disease itself or [3] bias effects based on the cooccurence of residual symptoms and long-term AP treatment. Clinical examinations in AP-naïve and AP treated patients suggest that both psychopathological and motor symptoms occur in schizophrenia, while the latter are additionally enforced by AP treatment. Thus, a (anatomical/pathophysiological) correlative of the linkage between psychopathological and MD symptoms playing a crucial role in both pathophysiology disease and AP treatment, seems to be likely. Due to the exploratory nature of the study causal relationships cannot be inferred and the study allows therefore for further differentiation (e.g. in prospective studies) between MD and psychopathological symptoms in patients with schizophrenia. Laboratory and more biologically driven tests that study endophenotypes of negative symptoms or neuromotor abnormalities are required to disentangle the overlapping phenotypic presentations of movement disorders and psychopathology, aiming at better understanding their genesis and building a foundation for a systematic therapy of the disease as well as the pharmacogenetic side-effects.