|Home | About | Journals | Submit | Contact Us | Français|
In this double-blind, placebo-controlled study, we examined the effects of subanesthetic doses of ketamine (an NMDA glutamate receptor antagonist) and thiopental (a GABA-A receptor agonist) on the event-related potential (ERP) correlates of deviant stimulus processing in 24 healthy adults. Participants completed three separate pharmacologic challenge sessions (ketamine, thiopental, saline) in a counterbalanced order. EEG data were recorded both before and during each challenge while participants performed a visual “oddball” task consisting of infrequent “target” and “novel” stimuli intermixed with frequent “standard” stimuli. We examined drug effects on the amplitude and latency of the P3(00) component of the ERP elicited by target (P3b) and novel stimuli (P3a), as well as the N2(00) and P2(00) components elicited by both targets and novels. Relative to placebo, both drugs reduced the amplitude of parietal P3b, and also tended to attenuate Target N2. Their effects on novelty processing were more differentiated. While both drugs reduced parietal P3a, this effect was greater for thiopental. Relative to placebo, ketamine also increased frontal P3a amplitude, shortened P3a latency, reduced N2, and enhanced P2 to novel stimuli.
Glutamate and gamma-aminobutyric acid (GABA) are the primary excitatory and inhibitory neurotransmitters in the brain and are likely involved in most aspects of cognition. The P3(00) component of the event-related potential (ERP) is generated by distributed cortical sources involved in processing contextually deviant stimuli, an integral component in a variety of cognitive operations. In this double-blind, placebo-controlled, pharmacologic challenge study, we examined the role of N-methyl-D-aspartate (NMDA) glutamate and GABA receptor function in modulating the P3 component in healthy adults. To our knowledge, this is the first study to administer a GABA-A agonist (the barbiturate, thiopental), a noncompetitive NMDA-antagonist (ketamine), and placebo in a within-subjects design to study the effects of these systems on the P3 complex.
The P3 has been extensively studied in a variety of contexts (for recent reviews, see Polich & Criado, 2006, Polich, 2007). It is frequently elicited using variations of the “oddball” paradigm in which participants detect infrequent “target” stimuli, and in some cases infrequent non-target stimuli, interspersed among frequent “standard” stimuli. The P3 represents a family of related but dissociable components (e.g., Courchesne et al., 1975; Squires et al., 1975) including the “Target” P3b (elicited by infrequent, task-relevant target stimuli) and the “Novelty” P3a (elicited by infrequent, task-irrelevant non-target stimuli). The P3b is commonly associated with voluntary attention and the updating of working memory (e.g., Donchin & Coles, 1988). It is generated by multiple cortical sources, including the temporo-parietal junction, frontal cortex, and parietal association areas (see Halgren et al., 1998; Polich & Criado, 2006; Soltani & Knight, 2000). The P3a likely represents an automatic orienting response to novel or otherwise salient stimuli (e.g., Courchesne et al., 1975; Knight, 1984) mediated by an interaction between prefrontal, temporal, and parietal cortical sources (e.g., Halgren et al., 1998; Knight, 1984; Knight, 1996; Soltani & Knight, 2000).
It is likely that the neuropharmacologies of the P3a and P3b are at least partially dissociable (see Polich & Criado, 2006). Although catecholaminergic transmitters are known to influence the P3 (e.g., Nieuwenhuis et al., 2005; Polich, 2007; Polich & Criado, 2006; Turetsky & Fein, 2002), both the GABAergic and glutamatergic systems have also been implicated (Frodl-Bauch et al., 1999). The latter are of particular interest because the underlying neuro-oscillatory activity that comprise ERPs like the P3 may depend on the interplay between excitatory NMDA and inhibitory GABAergic receptors (Ford et al., 2007).
Comparatively few studies have used selective pharmacologic challenges to examine the influences of NMDA and GABA activity on the P3. However, at least two groups have reported that ketamine attenuates auditory (Oranje et al., 2000) and visual P3b amplitude (Ahn et al., 2003). Several groups have also found that benzodiazapines attenuate auditory P3b amplitude (Hayakawa et al., 1999; Reinsel et al., 1991; Rockstroh et al., 1991), and one previous report (Fowler & Mitchell, 1997) found a barbiturate to affect the latency (but not amplitude) of the visual P3b.
In contrast to these prior challenge studies, we examined the effects of ketamine and thiopental on both the P3b and P3a components elicited during a visual oddball paradigm. Due to factors such as practice effects and test-retest habituation, we expected to observe attenuation of P3a/b amplitude from pre- to post-infusion on the placebo day. However, we predicted that both thiopental and ketamine would lead to greater attenuations of P3a/b amplitude. We also predicted that both ketamine and thiopental would alter P3a/b latency. In addition, we explored whether the effects of these drugs on P3a/b were related to their behavioral performance effects. Finally, we examined the effects of the drugs on the N2(00) and P2(00) components elicited by target and novel stimuli.
This protocol was approved by the Human Subjects Subcommittee of the VA Connecticut Healthcare System (West Haven, CT) and the Human Investigations Committee of the Yale University School of Medicine (New Haven, CT). Participants’ safety was reviewed on an ongoing basis by the Data Safety Monitoring Board of the NIAAA Center for the Translational Neuroscience of Alcoholism. All participants gave written informed consent prior to the onset of the study.
Twenty-four healthy adults were recruited by public advertisement. One was excluded from analysis due to excessive EEG artifacts, leaving a sample of 23 (8 female; Age m ± sd = 24.55 ± 2.59 years). Five of the participants had a family history of alcoholism, a distinction that is the focus of a larger study for which data collection is ongoing, but is not addressed in the current report.
All participants were medically and neurologically normal (based on history, physical exam, EKG, and screening laboratories), and underwent psychiatric screening using the Structured Clinical Interview for DSM-IV (First et al., 2002). Participants had previously used alcohol but had no history of alcohol abuse or dependence, other Axis I psychopathologies, or a family history of psychosis. Participants remained alcohol-free for three days prior to the first test day and for the duration of the testing period, and completed urine-toxicology and alcohol breathalyzer tests on each test day. Participants fasted the night before and for the duration of each session to minimize the risk of nausea.
Three separate, double-blind, drug challenge sessions were conducted with at least a three-day interval between sessions. Each session consisted of a 60-minute intravenous infusion of saline (placebo), ketamine (0.23 mg/kg loading and infusion rate 0.58 mg/kg/hr), or thiopental (1.5mg/kg loading and infusion rate of 40 mcg/kg/minute), and was supervised by an anesthesiologist (AP). Doses were selected to have mild-to-moderate psychogenic and sedating effects based on previous studies (Krystal et al., 1994; Krystal et al., 2005a; Krystal et al., 2005b) and were well tolerated.
ERP data were collected as part of a larger study comparing the psychiatric and behavioral effects of ketamine and thiopental. To assess subjective arousal and intoxication, ratings from Visual Analog Scales (VAS) were obtained at multiple time points. Analyses from this larger sample revealed that the two drugs did not lead to significantly different levels of sedation, although the ketamine had greater euphoric effects than thiopental (Dickerson et al., unpublished observations).
Participants were seated in a dimly lit, sound-attenuated, testing booth and completed two runs of a 3-stimulus visual oddball task approximately 90 minutes before and 45 minutes after the initiation of infusion. The task consisted of standard (small blue circle, white background, 80%), target (large blue circle, white background, 10%), and novel (non-repeating fractal images, 10%) stimuli presented in pseudo-random order on a computer display. Stimuli were presented for 500 ms with an inter-stimulus-interval of 1500 ms. Each run consisted of three blocks of 150 stimuli. Participants were instructed to respond only to the target stimulus by pressing a button using their right hand.
EEG data were acquired using sintered electrodes from three midline scalp sites (Fz, Cz, Pz) using a linked-earlobe reference and a fronto-central ground. The electro-oculogram was recorded horizontally (HEOG) from electrodes placed at the outer canthi of the eyes and vertically (VEOG) from electrodes placed above and below the right orbit. The data were digitized at 1000 Hz with a gain of 500, and were bandpass filtered between 0.1 and 100 Hz during acquisition. Electrode impedances did not exceed 10 kΩ. Data were low-pass filtered at 30 Hz offline prior to epoching.
ERP epochs were obtained from −100 ms to 1000 ms following stimulus presentation. An automated ocular correction routine (Gratton et al., 1983) was applied to remove blink and eye-movement artifacts. Epochs were baseline corrected using a −100 to 0 ms interval. Any corrected epochs containing EEG amplitudes exceeding +/− 75 µV were excluded from analysis. Error trials were also excluded. ERP averages were created for each Stimulus Type (standards, targets, novels) X Drug Type (placebo, ketamine, thiopental) X Time (pre- and post-infusion) combination.
Based on inspection of the grand-average waveforms, the P3b component was defined as the largest positive-going peak in a latency window of 300 to 500 ms following target stimulus onset. The P3a component was defined as the largest peak in a latency window of 250 to 450 ms following novel stimulus onset. The P2 component was defined as the largest positive-going peak between 160 to 240 ms post-stimulus, and the N2 component was defined as the negative-going peak immediately preceding the P3a/b component. An automated computerized routine was used to measure the peak amplitudes (relative to the pre-stimulus baseline) of the P3a/b and P2 components and the peak latencies of the P3a/b components. A semi-automatic, computerized routine was used to measure the peak amplitude of the N2 component, subject to verification by one of the authors (TDW).
Behavioral performance measures included the number of omission errors to target stimuli (misses), the number of commission errors to novel stimuli (false-alarms), and the mean response time (RT) to targets. One participant’s behavioral data for the placebo day were unavailable due to computer error.
Analysis of the larger sample from the behavioral study (Dickerson et al., unpublished observations) revealed ketamine had greater euphoric effects than thiopental. Accordingly, we examined VAS scores measuring subjective intoxication (“High” and “Buzzed”) as potential covariates for ERP analyses. To capture the overall euphoria levels during the infusion period, we averaged the 15 and 45-minute post-infusion scores of each item. One participant’s 45-minute post-infusion VAS ratings for the thiopental session were unavailable. The ketamine-induced euphoria scores did not significantly correlate with changes in P3a or P3b amplitude from pre- to post-infusion. Accordingly, these VAS variables were not included as covariates in the ERP analyses.
ERP and behavioral data were analyzed using repeated-measures Analyses of Variance (ANOVA). Behavioral data were analyzed using a Drug Type X Time design, while ERP data were analyzed using a Drug Type X Electrode X Time design. An alpha of .05 was used for all analyses. Greenhouse-Geisser corrections for non-sphericity were used for all comparisons with more than two levels. Effect sizes are reported as partial-eta squares.
Our primary interest was in assessing ERP and behavioral changes from pre- to post-infusion runs within a drug challenge session. The magnitude of these changes was expected to differ between drugs and across electrodes. Accordingly, simple interaction contrasts (Keppel, 1991) were employed to isolate the simple effects of significant higher order interactions in all ERP and behavioral analyses. For each significant Drug Type X Electrode X Time interaction, we tested the interaction of each pair-wise combination of Drug Type and Time separately for each electrode.
Pearson product-moment correlation coefficients (alpha = .05, two-tailed) were used to examine the relationship between changes in P3a/b amplitude following drug administration (post-infusion minus pre-infusion) and changes in target RT following drug administration (post-infusion minus pre-infusion).
No significant effects of Drug Type or Drug Type X Time were found for either omission errors to targets or false alarm errors to novels. However, a significant Drug Type X Time interaction was found for target RT (F[2, 42] = 6.37, p < .01, ηp2 = .23). Interaction contrasts revealed that relative to placebo, both thiopental (F[1, 21] = 8.73, p < .01, ηp2 = .29) and ketamine (F[1, 21] = 5.63, p = .03, ηp2 = .21) significantly slowed target RT. This effect tended to be larger for thiopental than ketamine (p = .07, ηp2 = .14).
Figures 1 and and22 display the pre- and post-infusion grand average ERPs elicited by target and novel stimuli on each test day. Table 1 displays pre- and post-infusion amplitudes for all measured components elicited by target and novel stimuli, peak latencies for the P3a/b components and a summary of the primary significant effects. Table 2 lists other significant effects found in the overall ANOVAs.
The primary finding was a significant Drug Type X Electrode X Time interaction effect (F[4, 88] = 2.65, p = .05, ηp2 = .11; Figure 3). Baseline amplitude differences between conditions were not significant at any electrode site. The results of pairwise interaction contrasts are as follows:
There was a significant Drug Type X Electrode X Time interaction (F[2, 44] = 3.48, p = .04, ηp2 = .14). Relative to placebo, ketamine significantly attenuated P3b amplitude at electrode Pz (F[1, 22] = 6.73, p = .02, ηp2 = .23).
There was a significant Drug Type X Electrode X Time interaction (F[2, 44] = 4.88, p = .02, ηp2 = .18). Relative to placebo, thiopental significantly attenuated P3b amplitude at electrode Pz (F[1, 22] = 13.98, p = .001, ηp2 = .39) and at the trend level at Cz (p = .07)
No significant differences were found between the effects of ketamine and thiopental on P3b amplitude.
There were no significant effects of drug type, time, or their interaction, on target P3b latency.
There was a significant Drug Type X Time interaction effect (F[2, 44] = 4.39, p = .02, ηp2 = .17). Relative to placebo, thiopental significantly reduced (F[1, 22)] = 12.29, p < .01, ηp2 = .36), and ketamine tended to reduce (p =.07) the N2. No significant differences were found between the effects of thiopental and ketamine on the N2.
There was a trend-level (p = .08) Drug Type X Electrode X Time interaction effect, with thiopental tending to increase P2 at Cz and Pz relative to both placebo and ketamine.
There were significant Drug Type X Time (F[2, 44] = 3.46, p = .05, ηp2 = .14) and Drug Type X Electrode X Time (F[4, 88] = 6.21, p = .001, ηp2 = .22; Figure 4) interactions. Pre-infusion P3a amplitude was significantly greater (at Fz and CZ) for the placebo session than for the two active drug sessions (p <. 05 in both cases). The results of pairwise interaction contrasts are as follows:
There was a significant Drug Type X Electrode X Time interaction (F[2, 44] = 11.10, p = .001, ηp2 = .34). Relative to placebo, ketamine significantly reduced P3a amplitude at Pz (F[1, 22] = 7.39, p = .01, ηp2 = .25) but significantly increased its amplitude at Fz (F[1, 22] = 4.78, p = .04, ηp2 = .18).
There was a significant Drug Type X Electrode X Time interaction (F[2, 44] = 7.51, p < .01, ηp2 = .26). Relative to placebo, thiopental significantly reduced P3a amplitude at Pz (F[1, 22] = 13.65, p = .001, ηp2 = .38).
There was a significant Drug Type X Time interaction (F[1, 22] = 4.16, p = .05, ηp2 = .16). Averaged across electrodes, thiopental reduced P3a amplitude significantly more than ketamine.
There was a significant Drug Type X Time interaction (F[2, 44] = 3.71, p = .04, ηp2 = .14; Figure 5). Ketamine produced a trend level reduction in novelty P3a latency relative to placebo (p = .06) and a significant latency reduction relative to thiopental (F[1, 22] = 8.48, p < .01, ηp2 = .28). No significant differences were found between the effects of thiopental and placebo on P3a latency.
There were significant Drug Type X Time (F[2, 44] = 7.69, p < .01, ηp2 = .26), and Drug Type X Electrode X Time (F[4, 88] = 4.91, p < .01, ηp2 = .18) interactions. Analysis of the Drug Type X Time interaction revealed that relative to placebo, both thiopental and ketamine produced significant attenuations of N2 amplitude, (F[1, 22] = 5.23, p = .03, ηp2 = .19 and F[1, 22] = 10.56, p < .01, ηp2 = .32, respectively). However, ketamine also produced a significantly greater attenuation of N2 amplitude than thiopental, (F[1, 22] = 4.72, p = .04, ηp2 = .18). Analysis of the three-way interaction revealed that relative to placebo, ketamine significantly reduced N2 amplitude at all three electrodes, whereas thiopental significantly reduced N2 at only Fz and Cz.
There was a significant Drug Type X Time interaction effect (F[2, 44] = 3.35, p = .05, ηp2 = .13). Simple interaction contrasts revealed that relative to placebo, ketamine significantly increased P2 amplitude (F[1, 22] = 11.51, p < .01, ηp2 = .34). However, at baseline, P2 amplitude was larger (p = .01) on the placebo test day than the ketamine day. No significant differences were found between the effects of ketamine and thiopental, or thiopental and placebo, on P2 amplitude.
There were significant correlations between changes in both P3a and P3b amplitude and changes in target RTs following the infusion of thiopental, but not ketamine. Specifically, delays in target RT following thiopental (but not ketamine), were significantly correlated with attenuations of target P3b (r = −.60, p < .001) and novelty P3a (r =−. 50, p < .001) amplitude at electrode Pz.
Consistent with previous studies of ketamine and GABA agonists (Ahn et al., 2003; Hayakawa et al., 1999; Oranje et al., 2000; Reinsel et al., 1991; Rockstroh et al., 1991), we found that both ketamine and thiopental attenuated parietal P3b amplitude. Overall, the data suggest a role of both the NMDA-glutamatergic and GABAergic receptor systems in the modulation of the Target P3b. Interestingly, whereas we found that thiopental affected P3b amplitude but not its latency, a previous report found a different barbiturate affected P3b latency but not its amplitude (Fowler & Mitchell, 1997). However, a combination of drug, dose, and task differences between the two reports may have contributed to the conflicting results.
Both drugs also attenuated the Target N2, although this effect only reached the trend level for ketamine. The N2 is frequently interpreted as an index of response inhibition (e.g., Falkenstein, 2006; Falkenstein et al., 1999) and/or conflict monitoring (e.g., Donkers & van Boxtel, 2004; Nieuwenhuis et al., 2003; Van Veen & Carter, 2002a; Van Veen & Carter, 2002b) mediated by the anterior cingulate cortex (ACC) and other prefrontal regions (e.g., Bekker et al., 2005; Jonkman et al., 2007; Mathalon et al., 2003; Van Veen & Carter, 2002a; Van Veen & Carter, 2002b). Accordingly, the effects of ketamine and thiopental on N2 amplitude might be result of the disruption of ACC activity required for the detection and processing of infrequent target stimuli.
Overall, the data indicate that ketamine and thiopental modulate both the early attentional processes (indexed by the N2) and later, top-down cognitive processes (indexed by the P3b) involved in target processing. Additionally, both drugs also led to significant delays in RT to target stimuli, suggesting that their basic behavioral consequences were also similar.
However, while the neural and behavioral effects of ketamine and thiopental on target processing were generally comparable, they were not identical. Overall, thiopental tended to produce greater ERP changes than ketamine, although these differences failed to reach significance. Thiopental also exhibited a trend to slow target RT more than ketamine. Finally, P3b amplitude attenuations following thiopental, but not ketamine, were significantly correlated with a slowing of RT to targets. Thus, although the active drugs had similar effects on the ERP and behavioral correlates of target processing; they were dissociable in terms of the coupling of the P3b changes with behavioral slowing.
The effects of ketamine and thiopental on novelty processing differed more than their effects on target processing. Although both drugs attenuated parietal P3a amplitude relative to placebo, thiopental led to greater overall reductions than ketamine. This suggests that GABA agonism leads to greater disruptions of the fronto-parietal networks thought to be involved in generating the P3a (e.g., Knight, 1984; Soltani & Knight, 2000). In addition, P3a amplitude attenuations following thiopental, but not ketamine, were significantly correlated with a slowing of RT to targets.
Ketamine was associated with additional alterations in P3a topography and latency. Specifically, ketamine produced a small but significant increase in frontal P3a amplitude relative to placebo, as well as a significant reduction in overall P3a latency relative to thiopental and a trend level reduction relative to placebo. Reduced P3 latencies are often interpreted as an index of enhanced processing speed or stimulus classification speed in healthy adults (see, Picton, 1992). It is therefore reasonable to speculate that P3a latency reductions following ketamine indicate that NMDA-antagonism alters the speed with which novel stimuli are detected and processed.
Ketamine also attenuated novelty N2 amplitude to a greater extent than both thiopental and placebo and increased novelty P2 amplitude relative to placebo. Increased P2 amplitudes have been found to reflect the effortful allocation of attention in a variety of tasks (Crowley and Colrain, 2004; Falkenstein et al., 2003). Thus, while the P3a latency reductions suggest that ketamine increases the processing speed/classification of novel stimuli; NMDA antagonism also appears to alter the efficiency of the early attentional processes indexed by the N2 and P2.
Thiopental led to attenuations of both P3a and P3b amplitude, the magnitude of which tended to be greater than for ketamine. Thiopental also tended to slow RTs more than ketamine. Finally, the degree of both P3a and P3b attenuations following thiopental (but not ketamine) were strongly correlated with delays in RTs to targets. Taken together, these data suggest that the GABAergic (but not the NMDA-glutamatergic) mechanisms responsible for P3a and P3b attenuations are related to a to a general cognitive-behavioral slowing.
At modestly sedating doses, thiopental appears to produce broad inhibitory effects that are relatively uniform across the electrophysiological and behavioral correlates of target and novelty processing. In contrast, ketamine appears to produce more uneven effects. Although ketamine affected both the N2 and P3b elicited by targets, its most striking effects were on the ERP correlates of novelty processing. Some ERP changes following ketamine appeared to reflect a decrease in the efficiency of certain aspects novelty processing (e.g., the N2, P2, and parietal P3a amplitude effects) while other aspects were preserved or enhanced (e.g., P3a latency reductions and a relative increase in frontal P3a amplitude). Overall, these findings are consistent with previous reports in both the human and animal literatures describing the importance of the NMDA system in novelty processing (e.g., Gironi Carnevale et al., 1990; Grunwald & Kurthen, 2006; Harich et al., 2007).
One possible way that ketamine affects novelty processing is by altering the perceptual relationships of the different categories of stimuli in the oddball task. For example, it has been demonstrated that the relationship between target and nontarget (standard) stimuli strongly influences the P3a component (see Polich, 2007), and that the processing of novel stimuli engages more cortical resources when overall task difficulty increases (Hagen et al., 2006). Specifically, when it is difficult to discriminate between target and nontarget stimuli, P3a amplitude increases and latency decreases (Polich & Comerchero, 2003). If ketamine alters the perceived relationship between the target (large blue circle) and standard stimuli (small blue circle) in the oddball task, it could potentially account for the P3a latency reductions and increases in frontal P3a amplitude following infusion.
Attenuations in N2 and parietal P3a amplitude and the increase in P2 amplitude to novel stimuli following ketamine suggest that NMDA antagonism affected multiple cognitive operations related to novelty processing. Decreases in N2 amplitude have been interpreted in terms of impaired inhibition of irrelevant information (e.g, Bertoli & Probst, 2005; Wang et al., 2003), while P2-like increases have been implicated in processing current task demands to avoid invalid behavioral responses or in the suppression of irrelevant stimulus features to improve performance (e.g., Potts, 2004). If ketamine alters the perceived salience of the novel stimuli, this could necessitate changes in the cognitive resources required to avoid inappropriate responses.
Previous studies have shown that ketamine disrupts cognitive processing mediated by the prefrontal cortex and the hippocampus (e.g., Grunwald & Kurthen, 2006; Honey et al., 2004; Honey et al., 2005;), regions that have also been implicated in generating the P3a (e.g, Knight, 1996; Knight, 1984; Soltani & Knight, 2000). Accordingly, it is possible that ketamine’s effects on novelty processing result from an alteration of both prefrontal and temporal/parietal activity, affecting the efficiency with which the cortex processes engaging, but task irrelevant, stimuli.
It should be noted that this study was intended to be primarily exploratory, and had several limitations. First, we only studied the effects of a single dose of ketamine and thiopental. Accordingly, it is not possible to determine from the current data if the GABAergic and NMDA glutamatergic effects on the P3 complex vary in a dose dependent manner. Second, the oddball task used in this report is limited in terms of the behavioral dependent variables it generates. For example, participants did not respond to novel stimuli in any way and it was therefore impossible to directly determine if novelty processing speed was affected by ketamine. In the future, the effects of these drugs on the ERP correlates of cognitive processing should be studied with more complex and behaviorally rich paradigms.
The authors acknowledge the important contributions of Angelina Genovese, R.N.C., M.B.A., Elizabeth O’Donnell, R.N., and Brenda Breault, R.N., B.S.N., of the Neurobiological Studies Unit of the VA Connecticut Healthcare System, West Haven Campus, West Haven, CT. In addition, the authors acknowledge support for this work from the Department of Veterans Affairs (VA Merit Review Grant, Alcohol Research Center) and National Institute on Alcohol Abuse and Alcoholism (KO5 AA 14906-04, 2P50-AA012870-07, T-32 AA 015496-02).
Portions of these data were initially presented at the 2007 annual meeting of the Research Society of Alcoholism (Chicago, IL, and the 2007 annual meeting of the American College of Neuropsychopharmacology (Boca Raton, FL).
Statement of Interests
Dr. Krystal has served as a paid scientific consultant for years 2005, 206, 2007 to: Alkermes, Astra Zeneca, Aventis Pharmaceuticals, Bristol-Myers Squibb, Cypress Bioscience Inc, Eli Lilly and Co., Fidelity Bioscience, Forest Laboratories Inc, Glaxo-SmithKline, Janssen Research Foundation, Merz Pharmaceuticals, Organon Pharmaceuticals Inc, Pfizer Pharmaceuticals, Sumitomo Pharmaceuticals America Ltd., Takeda Industries, UCB Pharma and US Micron. He is also co-sponsor of two pending patents related to the use of glutamatergic agents to treat psychiatric disorders and antidepressant effects of oral ketamine. Dr Petrakis holds a grant funded through Forest Laboratories Inc. Dr Mathalon holds a grant funded through Astra Zeneca.