Neuroimaging Insights Into Thought Disorder In Children

SAN DIEGO—Multimodal neuroimaging has given researchers insight into the neural circuits and brain structures involved in children’s discourse skills, specifically, thought disorder, according to studies presented at the 53rd Annual Meeting of the American Academy of Child and Adolescent Psychiatry.

WHAT IS THOUGHT DISORDER?

“Thought disorder is actually an impaired development of higher-level linguistic functions,” according to Rochelle Caplan, MD, Professor of Psychiatry and Director of the Brain Research Institute at the University of California, Los Angeles (UCLA). “It has been found in quite a wide variety of pediatric neurobehavioral disorders, including schizophrenia spectrum disorders, schizophrenia, prodromal syndrome, schiz­o­typal personality disorder, attention-deficit/hyperactivity disorder, high-functioning autism, multiplex developmental disorder, and epilepsy.” Dr. Caplan continued, “In point of fact, thought disorder refers to poor use of language to formulate and organize one’s thought. It is a higher-level linguistic process that involves basic linguistic skills, as well as the ability to formulate and organize one’s thinking and put it in a way that the listener is able to follow what one is talking about.”

Overall, Dr. Caplan said, there are three measures that are assessed, each with different functions. The first measure is formal thought disorder. This is an indicator of how well the speaker is organizing his or her thoughts at the level of the paragraph. The second measure is the connection across sentences. The third measure is how well the speaker monitors what he or she is saying, regarding the ability to speak a clear message or self-correct when the message comes across as unclear. This includes using fillers such as “like” or “well,” words commonly used to let the listener know it is still the speaker’s turn to continue speaking.

DETECTING DISCOURSE COHERENCE IN EARLY-ONSET SCHIZOPHRENIA

Children and adolescents with schizo­­phrenia recruit abnormal frontal-temporal networks when trying to make sense of discourse, according to Mirella Dapretto, PhD, Associate Professor of Psychiatry at UCLA.

“Based on studies from people with brain lesions as well as neuroimaging data, it is clear that both hemispheres of the brain are very important in understanding language and discourse,” Dr. Dapretto explained. “The left hemisphere is thought to be very important to perform a fine brain level analysis of speech. The left hemisphere, and particularly the frontal region, is important in deductive and syllogistic reasoning and in achieving coherence in speaking and listening to people talking. The right hemisphere, on the other hand, is important in comprehending the nonliteral aspect of speech, appreciating humor, jokes, and metaphors, and plays a critical role in the comprehension of the gist of a story or a narrative.”

In schizophrenia, Dr. Dapretto noted, a reduced left frontal activity is connected with a number of language tasks, particularly those that involve the generation of speech, and several studies have shown this reduced leftward symmetry.

To examine the integrity of neu­ral networks underlying discourse coherence in early-onset schizophrenia, investigators used an fMRI discourse coherence paradigm that had been previously validated in both normal adults and children. In this task, individuals listened to two people having short conversations and determined if the verbal exchange made sense. Unbeknownst to the individuals listening, features critical to this determination varied, so that the listeners’ reasoning or topic-maintenance abilities could be gauged.

The reasoning condition required detection of a possible break in the logic of a conversation. The patient heard a dialogue with one speaker asking “Why are you wearing a rain coat?” The second speaker replied “So I won’t get wet.” This is a plausible, logical answer. If, however, the individual heard “Why are you wearing a rain coat?” followed by “So I won’t get sad,” this raised a red flag, as there is no causal relationship between wearing a raincoat and avoiding sadness. In the topic-maintenance condition, the answer did not make sense because there was an abrupt change in topic. For example, if the individual heard things like “Do you believe in angels?” and a reply of “I have my own special angel,” that is a plausible answer. If, on the other hand, the second speaker responded with “I have my own special sandwich,” once again, the red flag goes up.

In this study, there were 24 children who met DSM-IV criteria for schizophrenia and 24 children who were age-matched and had no history of psychiatric, neurologic, or learning disorders. All of the children in the schizophrenia group, except one, were on typical neuroleptics, SSRIs, or stimulants. In order to assess the impact that medication had on brain activity, researchers computed chlorpromazine equivalents, then used the scores and regression analysis to see where brain activity was modulated by the amount of medications the patients were taking.

Overall, the participants listened to 24 question-and-answer pairs that were presented in two runs with the two activation conditions being counterbalanced. All the while, their behavioral responses were being documented through brain imaging.

In the normal controls, activity in the frontal-temporal networks was strongly left-lateralized during the reasoning conditions, but bilateral in the topic-maintenance condition, according to Dr. Dapretto. Such task-specific lateralization profiles were not observed in the schizophrenic group. Instead, the schizophrenic patients showed bilateral prefrontal activity when appraising the logic of a conversation, and greater activity in left temporal regions, although they failed to engage the right hemisphere to the same extent during topic maintenance.

It should be noted, however, that the participants with schizophrenia who were not as impaired as some in the study group had similar results to the participants without schizophrenia, both in the reasoning task and in the topic maintenance condition. This finding suggests that the abnormality observed in the networks is related to the clinical symptomatology. The researchers ­also discovered a correlation between higher amounts of medication and a greater amount of activity in the left hemisphere frontal temporal network. Dr. Dapretto believes that this finding might suggest that the medications are allowing the patients to better compensate for their deficit by enabling them to recruit language cortices to a greater extent.

THOUGHT DISORDER IN THE PSYCHOSIS PRODROME

Initial results using functional neuroimaging in adolescents who are at ultra-high risk for psychosis suggest that thought disorder, particularly illogical utterances and poverty of content of speech, may predict who is most at risk for a psychotic illness, said Carrie E. Bearden, PhD, Assistant Professor of Psychiatry at the Semel Institute for Neuroscience and Human Behavior, UCLA.

“It appears that frontal-temporal language networks may be disrupted in adolescents that are at ultra-high risk for developing psychosis,” Dr. Bearden declared. “The ultra-high-risk adolescent appears to recruit additional brain regions when performing a language task, and at baseline, adolescents who later developed psychotic disorder showed a distinct activation pattern as compared to those who did not convert to psychosis.”
In this study, the thought disorder and discourse paradigms discussed by Dr. Dapretto were extended to examine thought and com­munication disorders in adolescents at ultra-high risk for psychosis, using the complementary perspectives of behavioral and neural changes in language-related systems that may occur during the prodromal state. The sample included 30 ultra-high-risk patients and 30 healthy controls, with the groups being similar in terms of gender, ethnicity, and age. The groups were similar in terms of IQ, but the Global Assessment Functioning (GAF) scale in ­the ultra-high-risk group was about 44, indicating that the group was fairly functionally impaired.

Patients were qualified as ultra-high risk according to several criteria, Dr. Bearden stated. Those included in the study had met criteria for either a brief intermittent psychotic disorder that had occurred within the past three months; attenuated positive symptoms, which included people with subthreshold psychotic symptoms that had worsened in the past year; or a family history of psychosis in a first-degree relative and a functional decline, defined as a 30% drop in the past year in terms of their GAF score. Transcribed speech samples were elicited from the 30 ultra-high-risk patients and the 30 healthy controls using the Story Game Interview, Dr. Bearden continued. This interview allows the examiner to elicit much more elaborate speech samples than those elicited in a standard structure clinical interview, particularly in participants who are difficult to interview.

All of the study participants were given the Story Game Interview at baseline, then at two follow-up time points, using three alternative versions of the story game. Speech was coded with the Kiddie Formal Thought Disorder Rating Score and an adaptation of the Halliday and Hasan evaluation of linguistic cohesion. In addition, fMRI was used in a subset of 25 ultra-high-risk individuals and 14 healthy controls who had undergone behavioral testing to investigate the neural networks involved in detecting discourse coherence.

At baseline, the ultra-high-risk participants who later had psychosis showed a trend toward increased illogical thinking and poverty of content of speech. There were 22 participants who did not convert to psychosis and eight who did, as indicated by the illogical patterns of thought disorder. To summarize the first part of the behavioral findings, despite the absence of fully psychotic symptoms, individuals with prodromal symptoms demonstrated signs of thought disturbance, suggesting the possibility of a preclinical diagnosis as well as an important predictor of outcome in the early development of psychotic disease.

—Lawrence M. Prescott, PhD

Suggested Reading
1. Rachelefsky GS, Shapiro GG, Bergman D, et al. Pediatric Asthma: Promoting Best Practice. 1999. Milwaukee, Wis: American Academy of Allergy, Asthma & Immunology Inc; 1999.
2. Mannino DM, Homa DM, Pertowski CA, et al. Surveillance for asthma-United States, 1960-1995. Mor Mortal Wkly Rep CDC Surveill Summ. 1998;47:1-27.

Return to table of contents