7 May 2007. A fortunate conflux of studies reveals functional and structural deficits in the early processing of information about sounds in schizophrenia. Their findings may shed light both on why people with the illness experience auditory hallucinations and have difficulty reading other people’s emotions and meaning in conversation. The studies also counter the view that these symptoms stem primarily from problems with higher-level functions such as attention and memory.
In one of the studies, Robert A. Sweet of the University of Pittsburgh, Pennsylvania, and associates find that the brains of people with schizophrenia have relatively sparse axon terminals in areas responsible for processing basic auditory information. Daniel C. Javitt, at the Nathan S. Kline Institute for Psychiatric Research in Orangeburg, New York, and colleagues report that white matter anomalies in early auditory processing regions in schizophrenia may underlie the lack of ability to discern emotion, and perhaps other kinds of meaning, in speech. Two other studies, directed by Judith M. Ford of Yale University, New Haven, Connecticut, show how the brain’s mechanism for differentiating self- versus other-generated speech malfunctions in schizophrenia.
Unsound Structures, Pitch Troubles, Tied to Social Deficits
People with schizophrenia do poorly at identifying speakers’ emotions from their tone of voice. In the March American Journal of Psychiatry, Javitt, first author David I. Leitman, and colleagues, present two studies looking at impairments in the ability to detect prosody, or language-related rhythm and tones, in schizophrenia.
The first focused on affective prosody and brain structure in 19 subjects with schizophrenia or schizoaffective disorder and 19 healthy controls. Leitman and colleagues used fractional anisotropy derived from diffusion tensor magnetic resonance imaging to gauge the diffusion of water along axonal fibers and myelin, a reflection of their structural integrity.
Subjects with schizophrenia did worse than controls on tasks that involved spotting wrong notes in common melodies and decoding emotions from the speaker’s tone of voice. Javitt says people with schizophrenia “can’t detect emotions from people’s voices just because they can’t detect the music in the voices.”
On both tasks, poor performance was associated with lower anisotropy in early auditory sensory pathways. Not surprisingly, it also correlated with lower anisotropy in pathways to the amygdala, an area thought to play a large role in emotion, and for the emotion identification task, in the orbitofrontal cortex and corpus callosum. The findings “suggest that functional and structural deficits within early sensory regions contribute to the overall pattern of cognitive dysfunction in schizophrenia,” Leitman and colleagues write.
The second study tested whether deficits in musical pitch perception would coincide with impairments in detecting nonaffective prosody. To assess the latter, Leitman and colleagues asked 24 subjects with schizophrenia or schizoaffective disorder and 17 healthy controls to discriminate between sentences that differed in emphasis between subject and object, or between statements and questions based on tone of voice.
The performance of subjects with schizophrenia lagged behind that of controls on all prosody tasks. Relationships between pitch perception and different kinds of prosodic detection imply that tone-processing deficits may hinder recognition of emotional and nonemotional meaning in schizophrenia. Javitt says it was once thought that individuals with the disease are in “a kind of confusional state,” but “it’s only recently that people are coming around to saying maybe the information just isn’t getting in properly.”
Structures Lacking in Auditory Cortex
In the second paper, published in the April 1 issue of Biological Psychiatry, Sweet and colleagues at the University of Pittsburgh also provide evidence of structural abnormalities in parts of the auditory cortex involved in low-level processing (see also a commentary on this article, Shi, 2007). Prior research by his group found that people with schizophrenia have pyramidal cell bodies of abnormally low volume in deep layer 3 of Brodmann's areas 41 and 42, but not layer 5 of area 42 (Sweet et al., 2003; 2004). The layer 3 cells provide feedforward connections that help rush sensory input from one cortical region to another; layer 5 cells provide feedback circuits that allow contextual processing.
Since excitatory input from feedforward projections supports cell size and growth; “this pattern of abnormalities suggests that structural alterations in auditory cortex of subjects with schizophrenia are present in feedforward, but not feedback, circuits,” Sweet and colleagues write. They tested this hypothesis using brain tissue from autopsies of 15 people with schizophrenia or schizoaffective disorder and 15 comparison subjects matched on age, gender, and tissue storage duration.
Compared to controls, subjects with schizophrenia had 13.6 percent lower density of a marker of axon terminals in deep layer 3 of area 41, which gets input from the thalamus and from within itself. In contrast, the marker indicated similar terminal densities between patients and controls in layer 1 of area 41, an endpoint for feedback projections that start and stop in the cortex, but also in deep layer 3 of area 42, a higher level of the feedforward pathway.
In order to address the possible confound of antipsychotic drug treatment, the researchers replicated the experiments in nonhuman primates. They found no evidence that haloperidol exposure could affect the density of axon terminals within auditory cortical areas.
Area 41, the primary auditory cortex, helps distinguish pure tones, localize sounds, and define beginnings and ends of sounds; area 42 responds to more complex groupings of sounds. Sweet says the results point to “reductions in the structures that support the rapid spread of auditory sensory information in the earliest stages of processing in the cerebral cortex.”
Flawed Response to "Own" Speech
Auditory hallucinations may result when the brain cannot tell its own thoughts from others’ voices. In the March Archives of General Psychiatry, Ford, first author Theda H. Heinks-Maldonado, of Albert-Ludwigs-Universitaet in Germany, Daniel H. Mathalon, also of Yale, and colleagues relate hallucinations to a failure of the brain to respond differently to self- versus other-generated actions.
Before someone talks, a neural signal travels from speech production areas to auditory cortex. It carries a prediction of the speech sounds based on a copy of the motor command and triggers a “corollary discharge.” If the corollary discharge matches what the speaker hears, as it should when the sound is self-produced, the sensory experience is reduced. The sensory experience thus carries a tag declaring it "self-produced," which presumably helps us avoid being distracted by our own talking.
Previous work by this group showed that the N100 component of the electroencephalogram, which Ford says “may represent one of the first representations of the auditory stimulus in the cortex,” typically responds with less amplitude to speaking than listening (Ford et al., 2005). In schizophrenia, this suppression may not occur, perhaps due to a faulty corollary discharge system. To investigate further, Heinks-Maldonado and colleagues compared N100 responses to speaking and listening in 20 men with schizophrenia, half of whom regularly heard voices, and 17 age-matched healthy men.
In the speaking task, the men said “ah” every few seconds and received instantaneous feedback over headphones. Feedback varied between their own voice and another speaker's (or "alien") voice, both either unchanged or with the pitch shifted down. The subjects’ own sounds were recorded and played back during the listening task. The subjects then indicated whose voice they heard.
Hallucinators made more voice misattributions in all conditions during talking, and unexpectedly, while listening. During talking, controls showed a dampened N100 response, especially over the left hemisphere, to their own unchanged voice compared to the other voice feedback conditions. Some non-hallucinating patients also showed normal suppression, but hallucinators’ N100 amplitude did not differentiate the four types of feedback, suggesting that their central nervous system failed to temper its response to self-generated sounds. Less suppression was linked to more errors during speaking (r = 0.32, P = .04, 1-tailed) and, in the left hemisphere, to worse hallucinations (r = 0.49, P = .02).
Ford explained, “In healthy, normal controls, the corollary discharge is a precise representation of what you expect, and there is suppression when what you hear is what you expected. In schizophrenics, that relationship is off; and it is no longer precise.”
Neurons Out of Sync
Finally, in the March American Journal of Psychiatry, Ford, Mathalon, and colleagues reasoned that if the motor and sensory systems together generate the corollary discharge, then the involved neurons should become entrained to oscillate in sync, or simultaneously fire at the same frequency, before speaking. They expected less synchronous firing in schizophrenia. To test this theory, they gathered EEG data from 24 subjects with schizophrenia or schizoaffective disorder and 25 healthy controls as they spoke and listened. They measured neural synchrony across trials.
Controls showed more synchrony before talking than during listening. This difference was found in subjects with schizophrenia as well, but to a lesser extent, corroborating the notion of an inexact corollary discharge in schizophrenia. Those with the worst hallucinations showed the least pre-speech synchrony. Preparatory synchrony and suppressed N100 response were related in control subjects but not in those with schizophrenia.
“We suggest that this premovement burst of synchronous neural activity is a reflection of the forward model preparing the CNS for the sensory consequences of its own actions,” Ford and colleagues write. They note that “a wide range” of schizophrenia symptoms “could result from a deficient message being sent to the sensory cortex that the action is ‘self-generated.’”
Together, these four papers attest to fundamental problems in the processing of information about sounds that may contribute to symptoms previously ascribed to association areas, frontal cortex, or limbic areas. As far as the researchers could tell, antipsychotic medication did not cause the deficits. However, most of the studies included few female subjects, leaving their generalizability unclear.
In conclusion, Mathalon says “schizophrenics may have difficulty with auditory discrimination in general,” from appreciating prosody to listening to voices. To him, “the results suggest that during speech production there is an additional, and perhaps, independent, deficiency.”—Victoria L. Wilcox.
Ford JM, Roach BJ, Faustman WO, Mathalon DH. Synch before you speak: Auditory hallucinations in schizophrenia. Am J Psychiatry. 2007;164:458-466. Abstract
Heinks-Maldonado TH, Mathalon DH, Houde JF, Gray M, Faustman WO, Ford JM. Relationship of imprecise corollary discharge in schizophrenia to auditory hallucinations. Arch Gen Psychiatry. 2007:64:286-296. Abstract
Leitman DI, Hoptman MJ, Foxe JJ, Saccente E, Wylie GR, Nierenberg J, Jalbrzikowski M, Lim KO, Javitt DC. The neural substrates of impaired prosodic detection in schizophrenia and its sensorial antecedents. Am J Psychiatry. 2007;164:474-482. Abstract
Sweet RA, Bergen SE, Sun Z, Marcsisin MJ, Sampson AR, Lewis DA. Anatomical evidence of impaired feedforward auditory processing in schizophrenia. Biol Psychiatry. 2007:61:854-864. Abstract