Background
In 1985 Crow suggested a dichotomic concept for the diagnosis of schizophrenia by dividing schizophrenic disorders into type I- and type II-schizophrenia according to the prevalence of either positive or negative symptoms [
1]. Since schizophrenic patients often show symptoms of both dimensions and the psychopathology tends to change in the course of disease [
2], Crow’s approach did not suffice as a diagnosis system.
Nevertheless the concept of negative symptoms in schizophrenia has attracted much attention, since negative symptoms have been identified as important indicators for treatment response and prognosis. Negative symptoms, especially anhedonia and affective flattening, can be used as predictors of long-term outcome [
3],[
4]. Furthermore negative symptomatology in schizophrenia was found to intensify the severity of illness and strongly influence global functioning [
5]. Some authors found a specific influence of negative symptoms on vocational (work performance) and social functioning as well as functional skills [
6]-[
8]. Furthermore negative symptoms have been associated with social problem solving [
9]. Besides, they may also predict response to certain medication. Clozapine responders were identified by a lesser degree of negative symptomatology [
10].
The first rating scale for the assessment of negative symptomatology, the “Scale for the Assessment of Negative Symptoms” (SANS), was developed by Andreasen in the early 1980s [
11],[
12]. The SANS became the standard instrument for rating the extent of negative symptoms in schizophrenia. It defines five symptom complexes: affective flattening, impoverishment of language and thinking (alogia), reduced drive (avolition/apathy), the inability to feel happy (anhedonia/asociality) and disturbance of attention.
Negative symptoms might be differentiated from positive symptoms by a distinct pathological process. Traditionally negative symptoms in schizophrenia have been linked to frontal lobe anomalies, mainly due to the frequently described relationship between negative symptoms and hypofrontality while in a resting state or during the processing of cognitive tasks [
13]-[
18].
Due to these findings many studies regarding the correlation between negative symptoms and cognitive dysfunctions in schizophrenia have hypothesized and verified an inverse relationship between the extent of negative symptoms and so-called frontal lobe deficits. Inverse correlations were most often found with paradigms assessing logical thinking and categorization, e.g. the Wisconsin Card Sorting Test as well as indicators of verbal fluency [
19]-[
23].
Nevertheless inverse correlations with a variety of other cognitive dysfunctions have also been found. Pronounced negative symptoms have been linked to stronger impairments in recognition memory, psychomotor speed, attention or visuospatial abilities [
22],[
24]-[
27].
Several morphometric studies support the link between negative symptoms in schizophrenia and frontal lobe impairment. Reduced frontal lobe volumes have been found in correlation with anhedonia and the apathy level of schizophrenic patients, which are core symptoms of schizophrenic negative symptomatology [
28],[
29]. The extent of negative symptoms was also significantly correlated with smaller orbitofrontal, left insular, dorsolateral prefrontal regions, bilaterally in the medial frontal, anterior cingulate, inferior frontal and superior temporal regions gray matter volume [
30]-[
32] as well as reductions in prefrontal gray and white matter [
21],[
33],[
34]. Asami and coworkers found a positive correlation between frontolimbic and left superior temporal gyrus gray matter volume reduction and negative symptoms [
35]. Reduced gray matter volume in the middle prefrontal gyrus (Brodmann area 9/46 of frontal association cortex) as a correlate of negative symptoms have also been shown by other authors [
36]-[
38]. Following Talati and Hirsch the medial prefrontal gyrus, appendant to the dorsolateral prefrontal cortex, seems to be associated with executive mechanisms and decision- making abilities as a higher processing unit and might thereby modulate negative symptoms [
39]. A study using diffusion tensor imaging, a method to assess the integrity of neural fiber tracks, suggests an association between impaired white matter integrity in the inferior frontal region of schizophrenic patients and the severity of negative symptoms [
40]. However, in analogy to the neuropsychological findings, these results do not seem to be unambiguous, since other studies reported correlations between negative symptomatology and anomalies in other brain areas, especially the temporal regions, with the superior temporal gyrus [
34],[
41]-[
43] and the hippocampus being most often reported [
44].
So far distinct conclusions concerning the relationship between cognitive dysfunctions and negative symptoms in schizophrenia especially with respect to the cerebral correlate cannot be drawn. Based on present findings anomalies connected to the frontal and temporal lobes are primarily discussed.
A disadvantage of many neuropsychological studies examining the association between negative symptoms and neuropsychological dysfunctions is the restriction to a single cognitive faculty or a group of functions related to only one brain region, especially the frontal lobes. Furthermore only few studies have combined a neuropsychological assessment with a morphometric approach to explore the relationship with negative symptoms.
To address this question this study aims to investigate the cognitive and cerebral correlates of negative symptomatology in schizophrenia.
Schizophrenic patients and healthy controls were examined using the SANS, volumetric MRI-measures of different frontal and temporal brain structures and a comprehensive neuropsychological test battery covering frontal lobe (executive) as well as mediotemporal (memory) functions to assess the relationship between negative symptoms, neuropsychological deficits and structural brain anomalies in patients with schizophrenia.
Discussion and conclusions
In this study, schizophrenic patients displayed pronounced impairments in verbal and visuospatial memory as well as deficits concerning response inhibition. Dividing the schizophrenic group in patients with low and high SANS scores demonstrated especially strong memory, but also executive impairments in the HSS patients, whereas the LSS patients were relatively unimpaired. Correlations between SANS sub scores and global score, respectively, were mainly found with indicators of verbal memory. Anomalies in frontal or temporal brain regions or associations with neuropsychological variables were not found. These findings underline the association between negative symptomatology in schizophrenia and pronounced neuropsychological deficits and emphasise the relevance of verbal memory impairments.
The impact of negative symptoms on neuropsychological impairment in schizophrenia
The neuropsychological findings in our schizophrenic group are in general agreement with findings of other studies showing specific deficits concerning learning and memory in schizophrenic patients [
65].
Although the prevalence of executive dysfunctions in schizophrenia is estimated to be around 90% [
66] we could not confirm impairments in frontal lobe functioning with the exception of difficulties in response inhibition as assessed by the Stroop task. Results of a study by Hutton [
67] indicates, that the type of executive impairment might depend on the phase of illness. At the beginning of the disease process patients display deficits of planning and strategical thinking, whereas in chronic patients this shifts towards impairments concerning categorization and flexibility as assessed by the WCST. Since on average our schizophrenic patients have had only 2 schizophrenic episodes, this might explain the negative WCST results.
Studies have already shown the impact of negative symptomatology on certain neuropsychological functions in schizophrenic patients. Our results confirm the association between negative symptomatology in schizophrenia and pronounced neuropsychological deficits already found by others [
22],[
26],[
27],[
29]. Patients with a high SANS global score showed significantly more pronounced deficits especially concerning mnestic functioning than patients with a low SANS global score.
These results underline the close connection between negative symptomatology and verbal memory on the background of macroscopically unchanged structures in frontal and temporal cerebral areas. This is in contrast to other studies, which primarily reported a correlation between negative symptoms and executive functioning, sometimes associated with the corresponding structural changes in frontal lobe regions [
19],[
20],[
22].
Following the general line of thought a correlation between negative symptoms and executive – or so called frontal lobe - impairments would have been expected, since negative symptoms are assumed to be a consequence of frontal lobe impairment [
68]. On the other hand Hill and coworkers observed an amplifying effect of negative symptomatology in combination with disorganized behaviour on learning and memory deficits in schizophrenic patients, thereby confirming the results of this study [
69].
The schizophrenic group in this study showed a relatively high level of functioning in the neuropsychological categories studied here. The patients showed almost no executive impairments and no deficits concerning concentration and working memory. Patients with high and low negative symptoms could mainly be separated by their memory performance, even HSS patients could adhere their executive functioning. Memory, especially concerning the verbal modality, should be regarded as left temporal lobe functions in the majority of people and might be more vulnerable in regard to the cerebral changes associated with a schizophrenic disorder than other neuropsychological functions.
Interestingly the schizophrenic patients in this study did neither not show any change in the examined cerebral structures nor any relationship between neuropsychological measures and the examined anatomical faculties. Current findings report contradictory results. Some studies have found smaller volumes of brain structures in schizophrenic patients, especially of the hippocampus, the amygdala and the anterior cingulate cortex [
36],[
70]-[
72]. Whereas other findings could neither find any volumetric differences in the hippocampus-amygdala-complex nor in the different frontal lobe regions [
73],[
74]. Possible explanations for these inconsistencies might be differences in the kind of volume assessment, e.g. morphometric measurements using SPM or manual tracing of a structure according to validated protocols. However, such contradictions could also relate to the fact that the different study samples vary in terms of underlying pathophysiology [
75]. Especially in schizophrenia it is extremely difficult to assess a sufficiently homogenous population to be able to generalize findings to the results of other studies. The coexistence of normal volumes of the examined frontal and temporal faculties on the one hand and memory impairments as well as deficits in response inhibition on the other hand allow different interpretations. A functional recovery may have taken place before the structural regeneration. It could also be the case, that the macroscopically normal volume of a certain brain area might mask more subtle histological alterations or functional anomalies not related to structural change.
This leaves the question of the organic basis for the observed deficits and especially the correlation between negative symptomatology and verbal memory impairments unanswered. As macro-structural aberrations do not seem to be the organic basis for the correlation between negative symptoms and verbal memory impairment, biochemical or physiological mechanisms as the underlying common cerebral correlate have to be considered. Glutamate seems to be a good candidate to explain our correlation results. The glutamate hypothesis of schizophrenia was developed after it had been recognized that Phencyclidine (PCP) leads to an exacerbation of symptoms in schizophrenic patients [
76] and produces a schizophrenia-like state in normal individuals with positive and negative symptoms as well as cognitive disturbances [
77],[
78]. PCP serves as a noncompetitive antagonist at the glutamatergic NMDA receptor [
79]. Glutamate acts via different receptor types, but so far the main focus of the relationship between glutamate and schizophrenia has been on the NMDA receptor. It plays a major role in long-term potentiation, a synaptical process involved in memory formation [
80]. Besides its role concerning mnestic functions glutamate also affects schizophrenic negative symptomatology. Using D-serine, a full agonist at the glycine modulatory site of the NMDA receptor, Tsai et al. (1998) have found significant improvements in negative symptoms, psychosis, and executive function as measured by the cognitive subscale of the Positive and Negative Syndrome Scale (PANSS) and performance on the Wisconsin Card Sorting Test [
81]. Similarly D-cycloserine, a partial agonist at the glycine modulatory site of the NMDA receptor, has led to a diminution of negative symptomatology either alone or added to conventional antipsychotics [
82],[
83]. Considering these results glutamate might be the latent variable underlying the relationship between mnestic deficits and negative symptomatology observed in this study.
Methodological issues
Some methodological issues have to be considered. Dividing a group by a median split could be criticised due to the artificial allocation of patients to a certain group, who would otherwise be described on a continuum. In this study the highest SANS score of the LSS group was 16, whereas the lowest SANS score of the HSS group was 24. This yielded a clear differentiation of schizophrenic patients with low and high negative symptoms.
The chosen neuropsychological tests had to exist in a German version thereby limiting the range of possible assessment procedures. Nevertheless only those tests were taken that are profoundly validated and broadly accepted as neuropsychological diagnostic tools. The neuropsychological tests were administered by a trained and experienced neuropsychologist, guaranteeing an adherent administration. The volumetric method used is sound and has been described in a number of other publications [
59]-[
61]. However, according to our measurement protocol we did not separate gray from white matter and thus we cannot commend on whether there might have been a volume reduction of one of these compartments in the different groups.
The patient assessment was done by an experienced team of clinicians on a ward specialized on the treatment of chronic schizophrenia and the psychopathology was validated and quantified using internationally accepted psychometric tools.
Summary
In summary in this study we present data of a carefully diagnosed sample of patients with schizophrenia and control subjects. Following broad and parallel clinical, psychometric and MRI-based volumetric assessment of temporal and frontal lobe areas we did not identify group overall volumetric differences in frontal or temporal brain volumes. High scores of negative symptoms were categorically and dimensionally related in particular to verbal memory deficits rather than to dysexecutive dysfunction. Therefore the main finding of this study is, that from a functional point of view and in spite of absence of clear macroscopic volumetric brain differences negative symptoms are more related to left temporal brain properties (verbal memory deficits) than to frontal lobe dysfunction (dysexecutive symptoms). Further research should specifically adress this important research question and should possibly test the putative role of temporal glutamatergic dysfunction and its relationship to negative symptoms.
Ethic committee
AZ: 23–7532.22-11/1 L TvE, Ethik-Kommission der Albert-Ludwigs-Universität, Engelbergerstr. 21, 79106 Freiburg.
Competing interests
The authors declare that they have no competing interests. We certify that there is no conflict of interest with any financial organization or non- financial competing interests regarding the material discussed in the manuscript.
No reimbursements, fees, funding or salary from an organization that could lose financially was received by any author in the past five years or any stocks or shares are hold. The authors don’t hold any patents relating to the content of this manuscript.