Increased event-related theta activity as a psychophysiological marker of comorbidity in children with tics and attention-deficit/hyperactivity disorders

Abstract

Objective:

The question as to whether coexisting tic disorder (TD) and attention-deficit/hyperactivity disorder (ADHD) in children represent a combination of two independent pathologies, a separate nosologic entity manifested by both tics and hyperactivity or a phenotype subgroup of one of the two major clinical forms has received increasing attention. The aim of the present study was to classify the TD+ADHD comorbidity in the neurocognitive domain and to elucidate the neurophysiological background of TD+ADHD coexistence by analyzing event-related electroencephalographic (EEG) oscillations in the theta (3–7.5 Hz) frequency band.

Methods:

Event-related potentials were recorded at 10 electrodes in 53 children (9–13 years old) from four groups (healthy controls, TD-only, ADHD-only, and combined TD+ADHD patients), while they performed an auditory selective attention task requiring a button press to a predefined target. Event-related theta oscillations were analyzed by means of time–frequency decomposition (wavelet analysis) in two latency ranges—early (0–200 ms) and late (200–450 ms). The effects of psychopathology factors (TD and ADHD) and task variables (attended channel and stimulus task relevance) on early (ETR) and late (LTR) theta responses were evaluated statistically. Theta response measures were further correlated with psychopathology scores and spontaneous theta EEG activity.

Results:

(1) The ETR was enhanced only in comorbid children and did not differ between the control, TD-only, and ADHD-only groups. (2) The LTR was larger in children with ADHD (ADHD-only and comorbid), but this effect was mediated by the spontaneous theta EEG activity. (3) The ETR was larger to attended stimuli at frontal–central electrodes contralateral to the side of attention, to the target stimulus type at frontal locations, and at the hemisphere contralateral to the side of the response. The functional reactivity and scalp distribution of ETRs were modulated by psychopathological factors.

Conclusions:

In the neurocognitive domain, the TD+ADHD comorbidity can be identified as a unique nosologic entity. Both the spontaneous theta activity and late event-related theta oscillations appear as neurophysiological markers of the ADHD condition. In children, the early event-related theta oscillations may be associated with representations of relevant target features in working memory.

Significance:

(1) A new model is proposed according to which TD+ADHD comorbidity can be classified at different levels (from neurobiological to cognitive). (2) The functional significance of stimulus-synchronized theta oscillations in children is described for the first time.

Introduction

Because of their prevalence and great social impact, hypermotor symptoms in children have been increasingly focused on in recent child psychiatric research (Buitelaar and Rothenberger, 2004). Hypermotor behavior is identified in attention-deficit/hyperactivity disorder (ADHD) and tic disorder (TD). The core clinical symptoms of ADHD are inattention, impulsivity, and general motor hyperactivity (e.g., Swanson et al., 1998). In contrast, though being a complex neuropsychiatric disturbance, TD is essentially characterized by multiple motor and/or phonic tics with fluctuating phenomenology (Rothenberger, 1991, Leckman et al., 1997, Leckman, 2002). Extensive literature has documented that in TD, comorbidity with ADHD ranges from 35 to 90% in different studies (average 52%, review: Spencer et al., 1998, Rothenberger and Banaschewski, 2005), while in ADHD patients, tic disorders are seen in 11–33% (MTA Cooperative Group, 1999a, MTA Cooperative Group, 1999b, Kadesjo and Gillberg, 2001), both of which are much above the by-chance rate of combination. The overlap and relationships between these two psychiatric disturbances have been recognized as critical with respect to treatment choice and clinical consequences. Accordingly, the question as to whether coexisting TD and ADHD in children represent a combination of two independent pathologies (additive model), a separate nosologic entity manifested by both tics and ADHD (interactive model), or a phenotype subgroup of one of the two major clinical forms (phenotype model) has received increasing attention (Yordanova et al., 1996, Yordanova et al., 1997, Rothenberger et al., 2000, Moll et al., 2001).

At the level of clinical evaluation, previous epidemiologic and genetic studies have supplied evidence for each of the different models for TD+ADHD (e.g., Comings, 1995, Pauls et al., 1993, Spencer et al., 1998), which may at least partially stem from limitations imposed by symptom overlap. Therefore, to pursue the origins of comorbidity, the pathophysiological mechanisms underlying TD+ADHD coexistence need to be explored. At the neurophysiological level, motor processes in TD+ADHD have been examined by means of transcranial magnetic stimulation (TMS) and event-related brain potentials (ERPs). TMS application has supported the additive model by demonstrating that TD and ADHD may contribute independently to the excitability of the motor system in comorbid children (Moll et al., 2001). However, ERP studies have not yielded conclusive results (Dumais-Huber and Rothenberger, 1992, van der Meere et al., 1996, Yordanova et al., 1996, Yordanova et al., 1997). Analysis of slow negative potentials has shown that the preparatory motor cortical activation is similar between comorbid and TD-only group in support of the phenotype model (Yordanova et al., 1996). Further examination of the post-imperative negative variation has demonstrated that TD+ADHD may manifest itself as either a combined (additive model) or a unique (interactive model) psychopathology depending on motor response controllability (Yordanova et al., 1997). Thus, in TD+ADHD coexistence, two independent pathogenic sources may contribute to the inhibition/disinhibition mechanisms within the motor system (Moll et al., 2001), whereas the neurophysiological mechanisms subserving higher motor control may manifest deviations that do not result from a simple combination of TD- and ADHD-related deficits but may interact upon specific cognitive demands.

In this regard, analysis of external stimulus processing may provide important additional information about the neurocognitive background of TD+ADHD comorbidity. Previous ERP studies have demonstrated deviations in active stimulus processing during attention conditions in both TD-only (van de Wetering et al., 1985, van Woerkom et al., 1994, Oades et al., 1996) and ADHD-only children (Satterfield et al., 1990, Satterfield et al., 1994, Jonkman et al., 1997, Karayanidis et al., 2000, van der Stelt et al., 2001). In TD, ERP alterations have been attributed to difficulties in focusing, sustaining and allocating attention (Oades et al., 1996, Johannes et al., 2001), and to increased processing of non-relevant stimuli (van Woerkom et al., 1994). In ADHD, ERP variations are thought to reflect sustained and focused attention deficiency as well as problems of inhibition and working memory (Barkley, 1997, Tannock, 1998, Swanson et al., 1998). Despite the ERP evidence for specific changes of cognitive stimulus processing in pure TD and ADHD (Oades et al., 1996), in only one previous study, has stimulus processing been analyzed in combined TD+ADHD (Rothenberger et al., 2000). The results from an auditory attention task have shown that frontal and temporal mechanisms of stimulus selection may be similarly impaired in comorbid and ADHD-only children, but comparisons with pure TD have not been made. Thus, the mechanisms of cognitive stimulus evaluation in TD+ADHD need further examination.

In child psychiatric research, cognitive stimulus processing is conventionally explored by subtraction waveforms (e.g., Satterfield et al., 1990) or by ERP measures in the time domain (reviews: Tannock, 1998, Barry et al., 2003b, etc.). However, subtraction waveforms do not distinguish the source of differences and are prone to misinterpretations (van Boxtel, 2004). Likewise, a time domain ERP component may have a complex heterogeneous structure comprising several subcomponents that are completely or partially overlapped (Basar, 1980, Falkenstein et al., 1995). Such subcomponents may be functionally distinctive and may have specific frequency characteristics (delta, theta, alpha, gamma), but they remain undetected in the time domain ERPs (Basar, 1998, Kolev et al., 1997, Demiralp and Ademoglu, 2001, Yordanova et al., 2000, Yordanova et al., 2004). These EEG responses to external stimuli, called event-related oscillations or time–frequency ERP components, can be extracted by ERP decomposition in the time–frequency domain and can provide new refined information about neuroelectric dynamics of information processing (e.g., Basar, 1998, Kolev and Yordanova, 1997, Heinrich et al., 1999, Demiralp and Ademoglu, 2001, Makeig et al., 2004, Yordanova et al., 2004).

With this background, the main objective of the present study was to further classify the TD+ADHD comorbidity in the neurocognitive domain by (1) analysis of cognitive stimulus processing and (2) using time–frequency analysis of ERPs. As in the study of Rothenberger et al. (2000), here different models of TD+ADHD were tested in an auditory selective attention task, in which stimulus task relevance was varied on the base of internally guided spatial attention and physical stimulus features. Four groups of children (controls, TD-only, ADHD-only, and TD+ADHD) were studied with a balanced statistical design (Yordanova et al., 1996). The following specific goals were targeted.

(1) A main goal was to explore whether and how the processing of stimulus task relevance would depend on major psychopathology symptoms presented with TD, ADHD, and their combination. As a relevant analytic tool, stimulus-synchronized oscillations from the theta frequency range were used because of the following reasons: (a) time–frequency decomposition of ERPs from the current data set revealed that the ERPs contained prominent components from the theta (3–7.5 Hz) frequency band (see Results), which indicated that theta oscillations represented a most relevant EEG signal in this task. (b) Theta responses have been consistently associated with focused attention (Demiralp and Basar, 1992, Basar-Eroglu et al., 1992) and working memory processes (Yordanova and Kolev, 1996, Yordanova and Kolev, 1997, Yordanova and Kolev, 1998a, Yordanova and Kolev, 1998b, Kolev et al., 1997), thus appearing as an adequate correlate of cognitive stimulus evaluation. (c) Theta EEG frequency has been most frequently related with activations of frontal lobe networks (Basar et al., 2001, Sarnthein et al., 1998, von Stein and Sarnthein, 2000, Niedermeyer, 2001). Since frontal networks have been suggested to be differentially impaired in TD and ADHD, analysis of the theta frequency band would be especially appropriate to discriminate TD, ADHD, and TD+ADHD conditions.

(2) Another goal was to extend current views on the neurophysiological background of TD+ADHD by differentiating neurobiological from neurocomputational deficits. This was done by analyzing the spontaneous EEG theta activity in parallel with event-related theta oscillations. Spontaneous EEG rhythms reflect the neurobiological organization of frequency-specific networks in the brain, whereas event-related oscillations reflect the reorganization of these networks in relation to event-specific computational demands (Basar, 1998, Yordanova and Kolev, 1998a, Yordanova and Kolev, 1998b). It is noteworthy that children with ADHD and other psychiatric disorders have larger spontaneous theta EEG activity than normal children (Clarke et al., 2001; for review, see Barry et al., 2003a). This implies that irrespective of specific processing demands, there exists a basic (unspecific) alteration of the neuroelectric signaling in these children (Niedermeyer and Naidu, 1997, Niedermeyer and Naidu, 1998, Niedermeyer, 2001, Clarke et al., 2001). Thus, analysis of both the spontaneous and event-related theta activity would allow to assess whether the TD+ADHD comorbidity may be associated (a) with basic alterations of the neurobiological substrate, (b) with deficits emerging only upon specific processing demands, or (c) with both.

(3) A third goal of the present study was to shed further light on the functional significance of event-related theta responses. Despite the evidence for theta involvement in cognitive processes such as mental arithmetic, will, attention, episodic memory, memory gating, spatial learning, navigation, etc. (Basar et al., 2001, Klimesch, 1999, Kahana et al., 2001, Raghavachari et al., 2001, Caplan et al., 2003; etc.), the precise functional correlates of theta oscillations require further clarification. As stated earlier (Demiralp and Basar, 1992, Basar-Eroglu et al., 1992, Basar et al., 2001, Yordanova and Kolev, 1997, Yordanova and Kolev, 1998a, Kolev et al., 1997), analysis of event-related EEG is expected to increase most substantially the understanding of theta functional significance (Kahana et al., 2001). The present study therefore aimed at refining the functional correlates of stimulus-synchronized theta oscillations elicited in an auditory selective attention task in children.