Thought disorder: A developmental disability in pediatric epilepsy

Abstract

This study compared thought disorder (i.e., impaired use of language to formulate and organize thoughts) in 93 children with complex partial seizures (CPSs) and 56 children with primary generalized epilepsy with absence (PGE) and its relationship to age, seizure, cognitive, and linguistic variables. By the use of psychopathology, social competence, academic achievement, and school problem measures, the functional implications of thought disorder in these two groups were compared. When demographic variables were controlled for, there were no significant differences in thought disorder scores between the CPS and PGE groups. However, the profile of age, gender, seizure, and cognitive variables related to thought disorder differed in the CPS and PGE groups. Within each group, different aspects of thought disorder were associated with different seizure variables. Thought disorder was related to psychopathology, school problems, decreased academic achievement, and poor peer interaction in the CPS group, but with school problems in the PGE group. These findings suggest that CPS and PGE affect the normal maturation of children’s discourse skills, albeit through different mechanisms. The relationship of thought disorder to behavioral, academic, and social problems implies that these discourse deficits are one component of the developmental disabilities or comorbidities associated with pediatric CPS and PGE.

Introduction

Thought disorder is a clinical psychiatric term that denotes difficulty using language to formulate and organize one’s thoughts. It involves impairment in the form or manner in which a speaker presents thoughts to a listener, and has been characterized by clinical signs, such as illogical thinking, loose associations, incoherence, poverty of content of speech, tangentiality, and circumstantiality [1]. It represents abnormal organization, control, and processing of thinking [2], improper use of semantic and relational aspects of speech [3], and poor use of linguistic context to process and produce speech [4].

Although thought disorder was considered a hallmark of schizophrenia, it is observed in other psychiatric disorders, including mania [5] and depression [6], as well as in children with schizophrenia [7], attention-deficit hyperactivity disorder (ADHD) [8], and multiplex developmental disorder [9]. In schizophrenia, thought disorder is associated with both cognitive and higher-level linguistic or discourse linguistic deficits including impaired semantic processing [3], [10], semantic memory (see review in [11]), working memory [12], attention, and sequencing [13].

Given the association between a schizophrenia-like psychosis and temporal lobe epilepsy (see review in [14]), particularly left temporal lobe epilepsy, we began our initial studies on thought disorder in children with complex partial seizures (CPS) to identify possible organic substrates of thought disorder in childhood schizophrenia. We included subjects with primary generalized epilepsy with absence (PGE) as a control group in these studies, predicting that they, unlike the children with CPS, would not have thought disorder compared with age- and gender-matched normal children. In these studies we used the four types of thought disorder measures described in children with schizophrenia (Table 1). The first type, formal thought disorder, represents impaired organization of thoughts and includes illogical thinking (e.g., failure to present the listener with appropriate reasoning in causal and noncausal utterances or contradiction) and loose associations (e.g., unpredictable change in the topic of conversation without preparing the listener for the topic change) [15], [16]. The second type, cohesive deficits, involves poor use of linguistic ties, such as conjunctions and referential cohesion, that link ideas across sentences [17].

The third and fourth types, deficits in self-initiated repair and revision, respectively, involve impaired use of strategies to correct online errors during conversation that occur at the level of the organization of ideas or of the linguistic ties across sentences, word choice, or syntax [18], [19]. As a result of poor organization of ideas, impaired use of cohesive devices, and poor online correction of communication breakdown, the listener has difficulty following the reasoning, topic maintenance, ideational contiguity between sentences, and who and what the child with thought disorder is referring to. Normal children acquire the ability to organize their thinking and provide the listener with the necessary links to follow who and what they are talking about from the toddler period through adolescence [20], [21], [22], [23]. During this developmental process, thought disorder features are observed in normal children at young ages. Thus, young children do a poor job organizing their thoughts and presenting them to the listener in a coherent and cohesive manner with appropriate reasoning (e.g., illogical thinking). They also uncommonly use pragmatic devices to notify the listener of upcoming topic changes (e.g., loose associations), and cohesive devices to connect ideas across contiguous sentences (e.g., cohesion). They infrequently monitor and correct online errors as they speak (e.g., repair), and also have a limited repertoire of cohesive and repair devices.

Although normal children through age 10 use illogical thinking, they do this at significantly lower levels than younger children [7]. From age 7, loose associations are not heard in the conversation of normal children [7]. Similarly, with development, normal children also use cohesive and self-initiated repair devices more frequently during conversation and increase their repertoire of these devices (see review in [18]).

Our earlier studies demonstrated that younger schizophrenic children have significantly more thought disorder than older schizophrenic children [16], [17], [19]. These maturational differences suggested that onset of schizophrenia coinciding with development of children’s discourse skills might disrupt the normal developmental trajectory (i.e., thought disorder). We therefore predicted that onset of seizures or poor seizure control during active development of children’s discourse skills could impair this process.

Using the discourse measures listed in Table 1, we found that, contrary to our original prediction, both children with CPS and PGE of average intelligence failed to provide the listener with appropriate reasoning (e.g., illogical thinking) [24], [25] and underused linguistic devices to connect ideas expressed across contiguous sentences (e.g., conjunctions, referential cohesion) [26]. They repaired errors in the organization of their thoughts (e.g., fillers, false starts) infrequently, but made more frequent referential and syntactic revisions than normal children [27]. The children who had CPS and a schizophrenia-like psychosis differed from schizophrenic children by not having loose associations [25].

As demonstrated in children with schizophrenia [7] and ADHD [8], the younger subjects with CPS and PGE had significantly more thought disorder than the older subjects with these disorders. In more recent studies, we found that poor seizure control and earlier age of onset were related to illogical thinking in the PGE group [28]. In contrast to our predictions, in CPS, thought disorder was unrelated to lateralization of EEG findings. However, frontotemporal localization of epileptic activity was associated with underuse of cohesion and fillers, and temporal epileptic activity was associated with increased use of referential and syntactic revision [27], [28]. Unlike schizophrenic children [7], but similar to those with ADHD [8], IQ was associated with illogical thinking in CPS and PGE [28].

These findings suggested that both CPS and PGE might impact the development of children’s discourse skills. Similar mean thought disorder scores but differences in the seizure and IQ correlates of thought disorder implied differences in how illness variables affect the maturation of discourse skills in CPS and PGE.

To further our understanding of how age, seizure, cognitive, and linguistic variables play a role in the development of children’s discourse skills, it was important to reduce the dimensionality of the thought disorder measures to four scores and to compute a summary score. In addition, we were also interested in assessing the functional relevance of thought disorder in CPS and PGE. The study described in this article, therefore, had the following aims: (1) to compare a summary thought disorder score, described under Section 2.3, and four individual scores (formal thought disorder, cohesion, repair, revision) in a large sample of children with CPS and PGE; (2) to assess the relationship of these scores to age, seizure, cognitive, and linguistic variables separately in the CPS and PGE groups; and (3) to use measures of psychopathology, social competence, academic achievement, and school problems to examine functional implications of thought disorder in children with CPS and PGE.

Based on our earlier findings, we predicted no significant differences in the thought disorder scores of the CPS and PGE groups after controlling for demographic variables. We explored the hypothesis that age, seizure, cognitive, and linguistic variables would be differentially related to thought disorder measures in the CPS and PGE groups. Positing similar functional implications of having thought disorder in both epilepsy groups, we predicted that the children with more severe thought disorder would have more psychopathology, poor social competence, lower academic achievement scores, and increased school problems compared with those with less thought disorder.