Amphetamines Modulate Prefrontal Gamma Oscillations during Attention Processing

Abstract

Amphetamine-based medications robustly suppress symptoms of attention-deficit/hyperactivity disorder (ADHD), but their exact mechanisms remain poorly understood. Recent hemodynamic imaging studies have suggested that amphetamines may modulate the prefrontal and anterior cingulate brain regions, although few studies have been published and the results have not been entirely consistent. Meanwhile, several electrophysiology studies have shown that abnormal fast oscillations (in the gamma range) may be closely tied to inattention and other cardinal symptoms of ADHD. In this study, we utilized magnetoencephalography to examine how amphetamines modulate high-frequency brain activity in adults with ADHD. Participants performed an auditory attention task, which required sustained attention in one block and passive listening in a separate block. Participants completed the task twice (4 total blocks); in the on- and off-medication states. All data were analyzed using beamforming techniques to resolve cortical regions exhibiting event-related synchronizations and desynchronizations (ERS/ERD). Our primary findings indicated that oral amphetamine administration significantly decreased gamma-band ERD activity in the medial prefrontal area, and decreased ERS in bilateral superior parietal areas, left inferior parietal, and the left inferior frontal gyrus. These results suggest that psychostimulants strongly modulate gamma activity in frontal and parietal cortical areas, which are known to be central to the brain’s core attentional networks.

Introduction

Attention-deficit/hyperactivity disorder (ADHD) is the most common neurobehavioral disorder in children and adolescents [1]. Although the specific neural basis of the disorder remains an area of contention, several medications (e.g., methylphenidate) have proven robustly effective at suppressing the symptomatology. While their clinical efficacy is clear, the mechanisms by which these medications alleviate ADHD symptoms have been only vaguely characterized. Rubia and colleagues found that the anterior cingulate, caudate, cerebellum, and ventromedial frontal areas are modulated by methylphenidate treatment in ADHD children who are completing a continuous performance task [2]. In another functional magnetic resonance imaging (fMRI) study, Bush et al. [3] demonstrated that stimulant medications increase activation in the anterior cingulate, dorsolateral prefrontal area, and parietal regions in adults with ADHD during attention-demanding interference task.

Recent electrophysiological studies have indicated that patients with ADHD have aberrant gamma-frequency neural activity compared with their non-ADHD peers [4,5]. Gamma-band (30–120 Hz) activity occurs across the neocortex and extensive neurophysiological data has suggested that gamma oscillations are crucial to coordinating information processing (e.g., see [6]), and may have a role in integrating the disparate aspects of stimulus and/or intrinsic processes that are performed by separate regions of a distributed neural system [6,7]. In patients with ADHD, gamma-band deficits have been observed during periods of eyes-closed rest, during various cognitive tasks, and have been connected to the severity of inattention and hyperactivity symptoms [4,5,8].

In the current study, we investigated whether stimulant medications modulate high-frequency gamma activity during an attention-demanding auditory oddball task by recording magnetoencephalography (MEG) data before and after psychostimulant administration in adults with ADHD. Given the recognized role of the prefrontal cortex in attentional processing, our primary hypothesis was that amphetamines would enhance gamma-band activity in a network of brain regions that included the prefrontal cortical areas.

Methods and Materials

Subject Selection

We studied 11 adults (4 females) with attention-deficit/hyperactivity disorder (ADHD), inattentive type. Patients ranged from 34–58 years-old with 43.4 years being the average. All participants had shown widespread symptom suppression (i.e., the targeted clinical response) to a mixture of dextroamphetamine salts, extended release formula, and had maintained the same stable dosage of the study medication for at least 6 months prior to study enrollment. All patients were originally diagnosed during childhood or early adolescence, and were fully reassessed prior to enrollment by a board certified psychiatrist. This involved a semi-structured comprehensive psychiatric assessment with DSM-IV diagnostic criteria, the Adult ADHD Symptom Rating Scale (ASRS, v1.1; [9]), and collateral history. Exclusionary criteria included any medical illness affecting CNS function, neurological disorder, history of head trauma, and current substance abuse. Informed consent was obtained along the guidelines of the University of Nebraska Medical Center Institutional Review Board.

Experimental Paradigm

All participants were scheduled for MEG early in the morning (e.g., 07:30–08:00) and about 24 hours since their last stimulant dosage (i.e., morning of the previous day). Upon arrival each participant completed one block of binaural stimulation, which consisted of ~250 1kHz tones (duration = 50 ms; inter-stimulus interval = 1.2 s). Participants were instructed to monitor for the rare tones (7% of all stimuli), which were slightly quieter, and to their raise their right index finger each time such a tone was detected. These quiet tones were 10 dB less than the standard tones (93%), which were presented at 90 dB. Throughout the experiment, participants rested their hand on a laser-based response pad that detected finger movements. All stimuli were presented using TIP-300 transducers (Nicolet Biomedical, Madison, WI, USA) and foam ear inserts with 30 dB attenuation to exterior noise. Participants were then orally administered their standard daily dosage of amphetamine medication and moved to the patient waiting area. Approximately 75 minutes later, participants returned to the MEG room and completed a second (identical) block of the oddball paradigm.

Structural Magnetic Resonance Imaging (MRI)

High-resolution T1-weighted neuroanatomic images were acquired with a Philips Achieva 3T X-series scanner and an eight channel head coil using a 3D fast field echo sequence with the following parameters: field of view: 24 cm, slice thickness: 1 mm (no gap), in-plane resolution: 1.0 × 1.0 mm, sense factor: 1.5. The structural volumes were aligned parallel to the anterior and posterior commissures and used for MEG coregistration.

MEG Data Acquisition & Coregistration

With an acquisition bandwidth of 0.1 – 330 Hz, neuromagnetic responses were sampled continuously at 1 kHz using an Neuromag system with 306 magnetic sensors (Elekta, Helsinki, Finland). Using MaxFilter (v2.2; Elekta), MEG data from each condition (pre- and post-drug) and subject were individually-corrected for head movements during the recording, and subjected to noise reduction using the signal space separation method with a temporal extension [10].

Prior to measurement, four coils were attached to the subject’s head and the locations of these coils, together with the three fiducial points and scalp surface, were determined with a 3-D digitizer (Polhemus Navigator Sciences, Colchester, VT, USA). Once the subject was positioned inside the MEG room, an electric current was fed to each of the coils. This induced a measurable magnetic field and allowed coils to be localized in reference to the sensor array throughout the acquisition session. Since coil locations were also known in head coordinates, all MEG measurements could be transformed into a common coordinate system. With this coordinate system (including the scalp surface points), each participant’s MEG data was co-registered with their native space structural T1-weighted MRI data prior to source analyses.

MEG Pre-Processing

Artifact rejection was based on a fixed threshold method (MEG level exceeding +/− 1.2 pT), supplemented with visual inspection. Epochs were of 1.0 s duration, including a 0.2 s pre-stimulus baseline, with 0 ms defined as the stimulus (tone) onset. Artifact-free epochs from each condition were transformed into the time-frequency domain using complex demodulation, and the resulting spectral density power estimations per sensor were averaged over trials. These data were normalized by dividing the power of each time-frequency bin by the respective frequency’s baseline power, calculated as the mean power during the 200 ms baseline period (−0.2 to 0 s). This normalization procedure allowed task-related power fluctuations to be readily visualized in sensor space as event-related synchronizations (ERS; power increases) or desynchronizations (ERD; power decreases), and once identified the neural regions generating these responses could be scrutinized by examining time-frequency windows with a beamformer. In these data, the most robust gamma-band response across the sample was in the 68–88 Hz window and peaked in the 25–200 ms latency period. These responses concentrated mostly in the anterior aspect of the MEG sensor array.

MEG Source Imaging

Cortical networks were imaged through an extension of the linearly constrained minimum variance vector beamformer [11], which employs spatial filters in the frequency domain to calculate source power for the entire brain volume. The single images are derived from the cross spectral densities of all combinations of MEG sensors averaged over the time-frequency range of interest, and the solution of the forward problem for each location on a grid specified by input voxel space. In principle, the beamformer operator generates a spatial filter for each voxel (i.e., grid point), which passes signals without attenuation from the given neural region while minimizing interference from activity in all other brain areas.

In summary, source power was computed for the selected frequency band (68–88 Hz) and latency period (25–200 ms) over the subject’s entire brain volume using BESA 5.3 (MEGIS Software GmbH, Grafelfing, Germany). Subsequently, all structural MRI data were transformed into the Talairach coordinate system [12] using BrainVoyager QX (Brain Innovations, The Netherlands), and the same transform parameters were then applied to the subject’s functional volume. Activation patterns for the task main effects (attend, no-attend) and the effects of medication status were probed using a mass univariate approach based on the general linear model.

Results

Behavioral Data

Un-medicated participants successfully identified the oddball tones 97.06% (mean) of the time, whereas medicated persons identified the oddballs 97.77% on average. This difference was not significant, t(10) = 0.45 (p=0.66)

Main Effect of Attention Task

A cluster-level correction method (i.e., 40 contiguous voxels) was used to control for Type I errors. The statistical parametric maps were inspected for regions of significant ERS or ERD after thresholding (p < 0.001, cluster corrected). In the unmedicated condition, patients with ADHD showed significant ERD in bilateral medial prefrontal regions (Talairach Coordinates: −9,34,17) and left inferior frontal gyrus (−4,52,38), whereas significant ERS was detected in the right superior temporal sulcus and gyrus (52,−22,0; 60,−40,11). In the medicated condition, significant ERD activity was noted in right parietal (21,−7,46), right superior parietal lobule (22,7,62), and the left inferior parietal region (−10,−78,9)...

Main Effect of Medication Status

Administration of amphetamine medication significantly (p < 0.005) decreased ERD in the medial prefrontal regions (1,53,21), and decreased the ERS in bilateral superior parietal regions (−34,−72,45; 35,−75,38), the left inferior frontal gyrus (−64,10,27), left inferior parietal region (−8,−82,38) and the left parietal occipital sulcus (−18,−89,7; see Figure 1).

Figure 1.

MEG Activation Differences Following Amphetamine Medication Administration. Gamma-band (68–88 Hz) activity was compared in medicated and un-medicated adults with ADHD. In (A), 2D brain slices show the neural regions where significant (p<0.005) reductions in event-related desynchronization (ERD; blue) and event-related synchronization (ERS; red) amplitude were observed following medication administration. In (B), the same comparison images showing differences between un-medicated and medicated patients are shown as 3D renditions to illustrate the frontal brain areas relative to each other. Medicated ADHD patients exhibited significantly reduced gamma ERD amplitude in the medial prefrontal region (blue), and reduced gamma ERS amplitude in the superior parietal (red), left inferior parietal, and the left inferior frontal gyrus (red). All images are shown at (p < 0.005) using a cluster-based correction method.

Discussion

We evaluated cortical activation differences in adults with ADHD before and after amphetamine medication in a demanding auditory attention task using MEG-based beamforming. Prior to psychostimulant medication, we observed neuronal desynchronization in the gamma-band frequency (68–88 Hz) in the medial prefrontal cortex which was reduced after psychostimulant administration. Such enhanced medial prefrontal cortex activity supports our original hypothesis and reflects improved suppression of default-mode activity. Also, we observed reduced gamma ERS activity in the left inferior parietal area, superior parietal, and the left inferior frontal gyrus after medication administration. Overall, these results suggest psychostimulants do modulate gamma activity in cortical regions that are critical for attention processes in adults with ADHD. Below, we relate these data to other pertinent pharmacological imaging studies and discuss the implications of these findings for the role of gamma activity in attention, as well as general understandings of the systems-level mechanism(s) that may underlie stimulant medication efficacy in adult ADHD.

A key finding in this study was gamma-band desynchronization in the medial prefrontal cortex of un-medicated patients, which was strongly attenuated after the administration of psychostimulants. The medial prefrontal cortex is a key cortical region of the default-mode network (DMN), a group of brain regions that are more active in the alert, resting-state compared to an active task. At rest, the brain’s DMN has been linked with self-reflection and mind-wandering and appears to have an inverse correlation with task positive regions. Failure to effectively deactivate the DMN has been linked with lapses in attention, increased distractibility, and ADHD [9,23,24]. More specifically, abnormalities in the medial prefrontal node of the DMN have been linked with ADHD, and studies have shown that psychostimulants at least partially normalize activity in this brain area (e.g., [13]). Moreover, a recent MEG study demonstrated resting-state neurophysiological aberrations in the medial prefrontal cortex, but not other nodes, of the DMN in un-medicated ADHD patients compared with controls, suggesting a critical importance or locus for the MPFC in ADHD pathophysiology [14]. Overall, these findings suggest that not only does gamma-band activity appear to be critical for those with ADHD, but that it can be modulated with the use of psychostimulants. The current findings are also consistent with prior studies highlighting the importance of gamma-band activity [4,5,8] and further suggest that modulating the medial prefrontal cortex may be vital to symptom suppression in ADHD treatment.

The inferior frontal gyrus is widely believed to contribute to response inhibition processes or the ability to attend to a certain stimuli and inhibit responses to other (generally, pre-potent) stimuli, a skill that is essential for top-down attentional processing. A recent fMRI study demonstrated that children with ADHD do not activate typical cortical regions associated with top-down attention (i.e., right inferior frontal gyrus) and instead use multiple compensatory cortical regions [15]. This study showed no differences in behavioral or reaction time measures in children with ADHD compared to controls, thus supporting the notion that the activation disparities reflected a compensatory mechanism [15].Furthermore, in a study evaluating the effects of stimulants on cortical thickness, more rapid cortical thinning was detected in the left inferior frontal gyrus of adolescents with ADHD who were stimulant-naïve compared to those who had regularly taken stimulant medications for their ADHD [16]. The current data appear to be largely consistent with these previous findings, as the medication strongly modulated gamma band activity in the left inferior frontal gyrus of patients who were known responders to stimulant-based pharmacotherapy. The medication effects detected in the parietal regions (i.e. bilateral superior and left inferior parietal regions) are also suggestive to disruption in attention networks (i.e. fronto-parietal) in ADHD.

Conclusion

Our study identified medial prefrontal cortex ERD and left inferior frontal gyrus and parietal ERS in the gamma-frequency range in adults with ADHD in a demanding attention task. These ERD and ERS responses were attenuated with the administration of psychostimulant medications in all regions. Overall, these findings strongly suggest that stimulants modulate gamma-frequency activity in frontal and parietal regions, which are widely known as critical brain areas for core attentional networks. Such modulation is likely central to the efficacious effects of stimulant-based medications in ADHD. To close, it is important to recognize the limitations of this study including the relatively small sample size (11 subjects) and the lack of placebo-control and healthy control groups. Future work will need to examine larger groups of patients with different ADHD subtypes, include a well-matched group of non-ADHD controls, as well as extent these observations to younger persons with ADHD. In sum, our findings suggest that the medial prefrontal cortex plays a vital role in high-level attentional processes in the gamma-band frequency in adults with ADHD.