I thought it might be of interest to some members of ISPS to summarize recent functional neuroimaging research in the schizophrenias. These findings are reported in the scholarly volume, Schizophrenia: From Neuroimaging to Neuroscience by Stephen Lawrie, Eve Johnstone, and Daniel Weinberger published in 2004 for Oxford University Press.
Much of the material for this summary of the research is taken from a chapter in this volume entitled “Functional magnetic resonance imaging (fMRI)” by Garry Honey, Phillip McGuire and Edward Bullmore.
The most common form of fMRI data acquisition is blood oxygenation level dependent (BOLD) imaging. There is a coupling of neuronal activity to vascular response. There are many advantages to fMRI: non-invasive methodology, optimal spatial and temporal resolution, and sensitivity to signal change.
Executive function
Executive functioning refers to a host of neurocognitive abilities which are associated with the prefrontal cortex, such as planning, problem solving, working memory, etc. ‘Hypofrontality,’ a state of reduced metabolic activity in the prefrontal cortex, has been observed in neuroleptic naive patients. Further research in verbal fluency has demonstrated that hypofrontality depends upon the severity of psychotic symptoms and demands of the task at the time of scanning. Attenuated frontal activation has been observed in a variety of tasks, including mental state attribution and motor response inhibition. There is also research demonstrating greater or equivalent prefrontal activation in persons with schizophrenia compared with controls.
Episodic and working memory
Working memory refers to a limited-capacity system that facilitates the simultaneous storage and processing of information, which can be utilized to guide subsequent behavior and action. It is considered to be fundamental to a range of cognitive processes, including reasoning, language comprehension and problem solving. Working memory deficits have been consistently observed in persons with schizophrenia, as well as first-degree asymptomatic relatives, which has resulted in many researchers considering it an endophenotype (i.e., phenotypic markers or traits that may represent more direct expressions of underlying genes and the broader disordered phenotype). Working memory deficits may have prognostic implications for psychosocial rehabilitation programs. Several studies have shown hypofrontality during working memory tasks in a dose-response relationship to cognitive complexity. Several studies have shown normal or even exaggerated prefrontal activation until a certain task complexity is reached. One research group (Callicot et al 2000) demonstrated a relationship between prefrontal activation and levels of N-acetyl-aspartate (NAA). The latter has been shown to be reduced in persons with schizophrenia, but it is also observed to be reduced in persons with bipolar disorder and PTSD. NAA is generally viewed as a measure of neuronal integrity. Therefore, hypofrontality is not static but dynamic and is in a dose-response relationship to task complexity and subject capacity.
Kazushige Mizoguchi (2004), in the pharmacology department of Central Research Laboratories in Ibaraki, Japan, demonstrated that acute stress activates the mesoprefrontal dopaminergic system in rats, while chronic stress reduces dopaminergic neurotransmission (hypofrontality) in the PFC (prefrontal cortex). The latter resulted in working memory impairment through a D1 receptor involvement. Chronic stress induces a depressive behavioral state (negative symptoms in schizophrenia?), caused by a reduction in dopaminergic and serotonergic neurotransmission in the PFC. This research is detailed in Kazushige Mizoguchi (2004) “Stress and prefrontal cortical dysfunction in the rat,” in Prefrontal Cortex: From Synaptic Plasticity to Cognition edited by Satoru Otani for Kluwer Academic Publishers.
A team of investigators at the Institute of Psychiatry in London (Wykes et al 2002) demonstrated that patients with schizophrenia who had received a psychological treatment, cognitive remediation therapy (CRT), had significantly increased brain activation in regions associated with working memory, i.e., frontocortical areas.
Episodic memory refers to the encoding, storage, and retrieval of information in long-term memory with reference to personal history as opposed to semantic memory which refers to factual and conceptual knowledge. fMRI studies have demonstrated fronto-hippocampal dysfunction during the encoding of verbal and visual episodic information.
Sensorimotor function
The research in this area has been inconsistent. Some studies have shown attenuated activation of the supplementary motor area (SMA) and sensorimotor cortex during the performance of simple motor tasks and other studies have failed to observe this effect. There is no post-mortem data implicating the primary motor cortex as a pathomorphological site for neurological soft signs (NSS) in persons with schizophrenia.
Symptomatology
A consistent pattern of brain dysfunction which could serve as a biological marker in schizophrenia has failed to emerge. As a result, researchers have focused on individual psychotic symptoms, e.g., auditory hallucinations and formal thought disorder. Some research demonstrated that when patients are experiencing auditory hallucinations, there is increased activation of inferior frontal and temporal cortex. Research has indicated that in persons prone to experiencing auditory hallucinations, when asked to imagine external speech, there is reduced activation in temporal, parahippocampal and cerebellar cortex, which is consistent with previous PET data and cognitive models of auditory hallucinations (e.g., Chris Frith and colleagues) implicating defective monitoring of inner speech and action. Some research has shown a reduced activation in the temporal cortex to external speech during periods of auditory hallucination (Woodruff et al 1997). In addition, fMRI and PET data suggest that patients who are prone to auditory hallucinations evidence reduced temporal activation when required to monitor inner speech.
Hubl and Dierks (2004), in brief, discovered the following temporal neural activation pattern in patients who are actively hallucinating: hippocampus, Broca’s area (speech activation), amygdala, PAC (primary auditory cortex-Heschl’s gyrus), and lastly, the primary motor cortex. These investigators concluded:
“We interpreted this sequence of pathological activation in the following way: memories come up [I would say emotional memories, possibly traumatic in origin-mediated by the hippocampus], nearly simultaneously the motor speech area [Broca’s area] is activated. This seems to be related to an erroneous functional connection between subjective memories and language system [I would understand this as purposive intentional activity designed to communicate anxiety and emotional pain to an other or Other, a primary, homeostatic, affect-regulating symbiotic maternal presence and/or possibly what some researchers in PTSD have called “speechless terror”]. Broca’s area is a motor speech area and thus is relevant for the generation of inner speech, one mechanism considered to be important for the generation of auditory hallucinations. Somewhat later, the activation of the amygdala contributes to the emotional coloration of the auditory hallucinations [fear, rage, etc], and the activation of the PAC makes this mixture of memory, inner speech and emotional content heard [it is the activation of the PAC which results in the inner voice being perceived as an external ‘real’ voice].
There is evidence in fMRI research which suggests an inverse correlation between severity of thought disorder and left superior temporal gyral activity and reduced inferior frontal activation.
Honey et al (2003) observed that negative symptoms were correlated with increased lateral and medial frontal response during a working memory task and positive symptoms were associated with reduced fronto-temporal response.
Honey et al (2004) proposed that some of these observations are congruent with a capacity model of cortical function-which suggests that different cognitive tasks may compete for common, capacity-limited neural resources. They noted: “...psychotic symptoms such as auditory hallucinations may represent an endogenous demand on fronto-temporal cortical processing resources which may be competitive with exogenous demands, such as processing external speech” (p. 265).
Honey et al (2004) concluded:
Functional imaging research in schizophrenia has demonstrated widespread deficits affecting a range of cognitive function throughout the brain. However, a pattern of brain dysfunction that would serve as a biological trait marker or predict treatment response has not emerged to date” (p.277). These researchers noted the importance of functional connectivity in schizophrenia and have suggested that the hypofrontality and fronto-temporal disconnectivity observed in research studies may be state-not trait-related. They suggested that it is not so much the level of activation of a given neural region that is critical, rather, the inter-regional functional integration that exists within a neurocognitive network.
Brian Koehler PhD
New York University
brian_koehler@psychoanalysis.net
