Substance use patterns in 9-10 year olds: Baseline findings from the adolescent brain cognitive development (ABCD) study.

BACKGROUND: The Adolescent Brain Cognitive Development ™ Study (ABCD Study®) is an open-science, multi-site, prospective, longitudinal study following over 11,800 9- and 10-year-old youth into early adulthood. The ABCD Study aims to prospectively examine the impact of substance use (SU) on neurocognitive and health outcomes. Although SU initiation typically occurs during teen years, relatively little is known about patterns of SU in children younger than 12. METHODS: This study aims to report the detailed ABCD Study® SU patterns at baseline (n = 11,875) in order to inform the greater scientific community about cohort's early SU. Along with a detailed description of SU, we ran mixed effects regression models to examine the association between early caffeine and alcohol sipping with demographic factors, externalizing symptoms and parental history of alcohol and substance use disorders (AUD/SUD). PRIMARY RESULTS: At baseline, the majority of youth had used caffeine (67.6 %) and 22.5 % reported sipping alcohol (22.5 %). There was little to no reported use of other drug categories (0.2 % full alcohol drink, 0.7 % used nicotine, <0.1 % used any other drug of abuse). Analyses revealed that total caffeine use and early alcohol sipping were associated with demographic variables (p's<.05), externalizing symptoms (caffeine p = 0002; sipping p = .0003), and parental history of AUD (sipping p = .03). CONCLUSIONS: ABCD Study participants aged 9-10 years old reported caffeine use and alcohol sipping experimentation, but very rare other SU. Variables linked with early childhood alcohol sipping and caffeine use should be examined as contributing factors in future longitudinal analyses examining escalating trajectories of SU in the ABCD Study cohort.

Atypical visual and somatosensory adaptation in schizophrenia-spectrum disorders.

Neurophysiological investigations in patients with schizophrenia consistently show early sensory processing deficits in the visual system. Importantly, comparable sensory deficits have also been established in healthy first-degree biological relatives of patients with schizophrenia and in first-episode drug-naive patients. The clear implication is that these measures are endophenotypic, related to the underlying genetic liability for schizophrenia. However, there is significant overlap between patient response distributions and those of healthy individuals without affected first-degree relatives. Here we sought to develop more sensitive measures of sensory dysfunction in this population, with an eye to establishing endophenotypic markers with better predictive capabilities. We used a sensory adaptation paradigm in which electrophysiological responses to basic visual and somatosensory stimuli presented at different rates (ranging from 250 to 2550 ms interstimulus intervals, in blocked presentations) were compared. Our main hypothesis was that adaptation would be substantially diminished in schizophrenia, and that this would be especially prevalent in the visual system. High-density event-related potential recordings showed amplitude reductions in sensory adaptation in patients with schizophrenia (N=15 Experiment 1, N=12 Experiment 2) compared with age-matched healthy controls (N=15 Experiment 1, N=12 Experiment 2), and this was seen for both sensory modalities. At the individual participant level, reduced adaptation was more robust for visual compared with somatosensory stimulation. These results point to significant impairments in short-term sensory plasticity across sensory modalities in schizophrenia. These simple-to-execute measures may prove valuable as candidate endophenotypes and will bear follow-up in future work.

Reward contingencies and the recalibration of task monitoring and reward systems: a high-density electrical mapping study.

Task execution almost always occurs in the context of reward-seeking or punishment-avoiding behavior. As such, ongoing task-monitoring systems are influenced by reward anticipation systems. In turn, when a task has been executed either successfully or unsuccessfully, future iterations of that task will be re-titrated on the basis of the task outcome. Here, we examined the neural underpinnings of the task-monitoring and reward-evaluation systems to better understand how they govern reward-seeking behavior. Twenty-three healthy adult participants performed a task where they accrued points that equated to real world value (gift cards) by responding as rapidly as possible within an allotted timeframe, while success rate was titrated online by changing the duration of the timeframe dependent on participant performance. Informative cues initiated each trial, indicating the probability of potential reward or loss (four levels from very low to very high). We manipulated feedback by first informing participants of task success/failure, after which a second feedback signal indicated actual magnitude of reward/loss. High-density electroencephalography (EEG) recordings allowed for examination of event-related potentials (ERPs) to the informative cues and in turn, to both feedback signals. Distinct ERP components associated with reward cues, task-preparatory and task-monitoring processes, and reward feedback processes were identified. Unsurprisingly, participants displayed increased ERP amplitudes associated with task-preparatory processes following cues that predicted higher chances of reward. They also rapidly updated reward and loss prediction information dependent on task performance after the first feedback signal. Finally, upon reward receipt, initial reward probability was no longer taken into account. Rather, ERP measures suggested that only the magnitude of actual reward or loss was now processed. Reward and task-monitoring processes are clearly dissociable, but interact across very fast timescales to update reward predictions as information about task success or failure is accrued. Careful delineation of these processes will be useful in future investigations in clinical groups where such processes are suspected of having gone awry.

Effects of ZNF804A on auditory P300 response in schizophrenia.

The common variant rs1344706 within the zinc-finger protein gene ZNF804A has been strongly implicated in schizophrenia (SZ) susceptibility by a series of recent genetic association studies. Although associated with a pattern of altered neural connectivity, evidence that increased risk is mediated by an effect on cognitive deficits associated with the disorder has been equivocal. This study investigated whether the same ZNF804A risk allele was associated with variation in the P300 auditory-evoked response, a cognitively relevant putative endophenotype for SZ. We compared P300 responses in carriers and noncarriers of the ZNF804A risk allele genotype groups in Irish patients and controls (n=97). P300 response was observed to vary according to genotype in this sample, such that risk allele carriers showed relatively higher P300 response compared with noncarriers. This finding accords with behavioural data reported by our group and others. It is also consistent with the idea that ZNF804A may have an impact on cortical efficiency, reflected in the higher levels of activations required to achieve comparable behavioural accuracy on the task used.

Cortical cross-frequency coupling predicts perceptual outcomes.

Functional networks are comprised of neuronal ensembles bound through synchronization across multiple intrinsic oscillatory frequencies. Various coupled interactions between brain oscillators have been described (e.g., phase-amplitude coupling), but with little evidence that these interactions actually influence perceptual sensitivity. Here, electroencephalographic (EEG) recordings were made during a sustained-attention task to demonstrate that cross-frequency coupling has significant consequences for perceptual outcomes (i.e., whether participants detect a near-threshold visual target). The data reveal that phase-detection relationships at higher frequencies are dependent on the phase of lower frequencies, such that higher frequencies alternate between periods when their phase is either strongly or weakly predictive of visual-target detection. Moreover, the specific higher frequencies and scalp topographies linked to visual-target detection also alternate as a function of lower-frequency phase. Cross-frequency coupling between lower (i.e., delta and theta) and higher frequencies (e.g., low- and high-beta) thus results in dramatic fluctuations of visual-target detection.

Generation of the VESPA response to rapid contrast fluctuations is dominated by striate cortex: evidence from retinotopic mapping.

The VESPA (visual-evoked spread spectrum analysis) method derives an impulse response function of the visual system from scalp electroencephalographic (EEG) data using the controlled modulation of some feature of a visual stimulus. Recent research using VESPA responses to modulations of stimulus contrast has provided new insights into both early visual attention mechanisms and the specificity of visual-processing deficits in schizophrenia. To allow a fuller interpretation of these and future findings, it is necessary to further characterize the VESPA in terms of its underlying cortical generators. To that end, we here examine spatio-temporal variations in the components of the VESPA as a function of stimulus location. We found that the first two VESPA components (C1/P1) each have a posterior dorsal midline focus and reverse in polarity across the horizontal meridian, consistent with retinotopic projections to calcarine cortex (V1) for the stimulus locations tested. Furthermore, the focal scalp topography of the VESPA was strikingly constant across the entire C1-P1 timeframe (50-120 ms) for each stimulus location, with negligible global scalp activity visible at the zero-crossing dividing the two. This indicates a common focal source underpinning both components, which was further supported by a significant correlation between C1 and P1 amplitudes across subjects (r=0.54; p<0.05). These results, along with factors implicit in the method of derivation of the contrast-VESPA, lead us to conclude that these responses are dominated by activity from striate cortex. We discuss the implications of this finding for previous and future research using the VESPA.

Visual object processing as a function of stimulus energy, retinal eccentricity and Gestalt configuration: a high-density electrical mapping study.

To reveal the fundamental processes underlying the different stages of visual object perception, most studies have manipulated relatively complex images, such as photographs, line drawings of natural objects, or perceptual illusions. Here, rather than starting from complex images and working backward to infer simpler processes, we investigated how the visual system parses and integrates information contained in stimuli of the most basic variety. Simple scatterings of a few points of light were manipulated in terms of their numerosity, spatial extent, and organization, and high-density electrophysiological recordings were made from healthy adults engaged in an unrelated task. We reasoned that this approach permitted an uncontaminated view of the spatio-temporal dynamics of the related neural processes. We were guided in our predictions by the "frame-and-fill" model for object perception, whereby fast inputs to the dorsal stream of the visual "where" system first frame the spatial extent of visual objects, which are subsequently "filled-in" by the slower activation of the ventral stream of the visual "what" system. Our findings were consistent with this view, showing a rapidly-onsetting effect of spatial extent in dorsal stream sources, and later-onsetting effects due to dot number and symmetry, which were deemed to be more closely tied to the details of object identity, from ventral stream sources. This collection of observations provides an important baseline from which to understand the spatio-temporal properties of basic visual object perception, and from which to test dysfunction of this system in clinical populations.

Neural correlates of oddball detection in self-motion heading: a high-density event-related potential study of vestibular integration.

The perception of self-motion is a product of the integration of information from both visual and non-visual cues, to which the vestibular system is a central contributor. It is well documented that vestibular dysfunction leads to impaired movement and balance, dizziness and falls, and yet our knowledge of the neuronal processing of vestibular signals remains relatively sparse. In this study, high-density electroencephalographic recordings were deployed to investigate the neural processes associated with vestibular detection of changes in heading. To this end, a self-motion oddball paradigm was designed. Participants were translated linearly 7.8 cm on a motion platform using a one second motion profile, at a 45° angle leftward or rightward of straight ahead. These headings were presented with a stimulus probability of 80-20 %. Participants responded when they detected the infrequent direction change via button-press. Event-related potentials (ERPs) were calculated in response to the standard (80 %) and target (20 %) movement directions. Statistical parametric mapping showed that ERPs to standard and target movements differed significantly from 490 to 950 ms post-stimulus. Topographic analysis showed that this difference had a typical P3 topography. Individual participant bootstrap analysis revealed that 93.3 % of participants exhibited a clear P3 component. These results indicate that a perceived change in vestibular heading can readily elicit a P3 response, wholly similar to that evoked by oddball stimuli presented in other sensory modalities. This vestibular-evoked P3 response may provide a readily and robustly detectable objective measure for the evaluation of vestibular integrity in various disease models.

Changes in effective connectivity of human superior parietal lobule under multisensory and unisensory stimulation.

Previous event-related potential (ERP) studies have identified the superior parietal lobule (SPL) as actively multisensory. This study compares effective, or contextually active, connections to this region under unisensory and multisensory conditions. Effective connectivity, the influence of one brain region over another, during unisensory visual, unisensory auditory and multisensory audiovisual stimulation was investigated. ERPs were recorded from subdural electrodes placed over the parietal lobe of three patients while they conducted a rapid reaction-time task. A generative model of interacting neuronal ensembles for ERPs was inverted in a scheme allowing investigation of the connections from and to the SPL, a multisensory processing area. Important features of the ensemble model include inhibitory and excitatory feedback connections to pyramidal cells and extrinsic input to the stellate cell pool, with extrinsic forward and backward connections delineated by laminar connection differences between ensembles. The framework embeds the SPL in a plausible connection of distinct neuronal ensembles mirroring the integrated brain regions involved in the response task. Bayesian model comparison was used to test competing feed-forward and feed-backward models of how the electrophysiological data were generated. Comparisons were performed between multisensory and unisensory data. Findings from three patients show differences in summed unisensory and multisensory ERPs that can be accounted for by a mediation of both forward and backward connections to the SPL. In particular, a negative gain in all forward and backward connections to the SPL from other regions was observed during the period of multisensory integration, while a positive gain was observed for forward projections that arise from the SPL.

Objects are highlighted by spatial attention.

Selective attention may be focused upon a region of interest within the visual surroundings, thereby improving the perceptual quality of stimuli at that location. It has been debated whether this spatially selective mechanism plays a role in the attentive selection of whole objects in a visual scene. The relationship between spatial and object-selective attention was investigated here through recordings of event-related brain potentials (ERPs) supplemented with functional magnetic brain imaging (fMRI). Subjects viewed a display consisting of two bar-shaped objects and directed attention to sequences of stimuli (brief corner offsets) at one end of one of the bars. Unattended stimuli belonging to the same object as the attended stimuli elicited spatiotemporal patterns of neural activity in the visual cortex closely resembling those elicited by the attended stimuli themselves, albeit smaller in amplitude. This enhanced neural activity associated with object-selective attention was localized by use of ERP dipole modeling and fMRI to the lateral occipital extrastriate cortex. We conclude that object-selective attention shares a common neural mechanism with spatial attention that entails the facilitation of sensory processing of stimuli within the boundaries of an attended object.

Jumping the gun: is effective preparation contingent upon anticipatory activation in task-relevant neural circuitry?

Subjects switched between tasks that rely on separable "low-level" neural circuits, a motion and a color task. Using functional magnetic resonance imaging, we assessed anticipatory processes within these circuits during preparation to switch between tasks. Once the switch was made, we could then compare activation levels within the circuit associated with the newly relevant task to continuing activity in the circuit associated with the irrelevant task, allowing us to assess both the effectiveness of anticipatory switching mechanisms and the subsequent competition between alternative stimulus-response contingencies. Subjects prepared effectively for the color task, being equally fast and accurate on switch trials as on repeat trials, and this successful preparation was associated with robust preparatory activity within well-known color-processing regions. In contrast, subjects showed considerable behavioral costs when switching to the motion task, evincing a lack of effective preparation, borne out by the fact that motion circuits were silent during the preparatory period.

A topography of executive functions and their interactions revealed by functional magnetic resonance imaging.

We used fMRI to study the brain processes involved in the executive control of behavior. The Sustained Attention to Response Task (SART), which allows unpredictable and predictable NOGO events to be contrasted, was imaged using a mixed (block and event-related) fMRI design to examine tonic and phasic processes involved in response inhibition, error detection, conflict monitoring and sustained attention. A network of regions, including right ventral prefrontal cortex (PFC), left dorsolateral PFC (DLPFC) and right inferior parietal cortex, was activated for successful unpredictable inhibitions, while rostral anterior cingulate was implicated in error processing and the pre-SMA in conflict monitoring. Furthermore, the pattern of correlations between left dorsolateral PFC, implicated in task-set maintenance, and the pre-SMA were indicative of a tight coupling between prefrontally mediated control and conflict levels monitored more posteriorly. The results reveal that the executive control of behavior can be separated into distinct functions performed by discrete cortical regions.

The role of response requirements in task switching: dissolving the residue.

Task-switching paradigms, which are regularly used to assay 'executive control' processes in humans, almost invariably reveal a decrement in subjects' performance on the first trial following a switch of task. That is, subjects are slower to respond and more error prone on the switch trial, a difference in performance that has been termed the 'switch-cost'. This switch cost has then been taken to reflect the time taken by neural control processes. Previous studies have shown that while performance improves as more time is provided to prepare for the switch, switch costs persist, even over very long intervals. In the present study, however, we find that changing the response regimen (choice reaction time vs go-no-go) has profound effects on the switch cost. A task switching paradigm was used in which subjects randomly switched between two tasks, based on a cue that was presented at varying intervals prior to the presentation of the imperative stimulus. While switch costs were found in all conditions in the choice reaction time blocks, they were completely abolished in the go-no-go blocks when sufficient preparation time was provided (500 or 800 ms). This is important because the only difference between the choice reaction time and go-no-go conditions was the response requirement: these conditions did not differ in the stimuli used, in the tasks performed or in the preparation time provided. These data call into question models of executive control that interpret switch costs as reflecting the time taken by neural processes to switch the system from a readiness to perform one task to a readiness to perform another.

Task switching: a high-density electrical mapping study.

Flexibly switching between tasks is one of the paradigmatic functions of so-called "executive control" processes. Neuroimaging studies have implicated both prefrontal and parietal cortical regions in the processing necessary to effectively switch task. Beyond their general involvement in this critical function, however, little is known about the dynamics of processing across frontal and parietal regions. For instance, it remains to be determined to what extent these areas play a role in preparing to switch task before arrival of the stimulus to be acted upon and to what extent they play a role in any switching processes that occur after the stimulus is presented. Here, we used the excellent temporal resolution afforded by high-density mapping of brain potentials to explore the time course of the processes underlying (1) the performance of and (2) the preparation for a switch of task. We detail the contributions of both frontal and parietal processes to these two aspects of the task-switching process. Our data revealed a complex pattern of effects. Most striking was a period of sustained activity over bilateral parietal regions preceding the switch trial. Over frontal regions, activity actually decreased during this same period. Strongest sustained frontal activity was in fact seen for trials on which no switch was required. Further, we find that the first differential activity associated with switching task was over posterior parietal areas (220 ms), whereas over frontal scalp, the first differential activity is found more than 200 ms later. These and other effects are interpreted in terms of a "competition" model in which preparing to switch task is understood as the beginning of a competition between the potentially relevant tasks that is resolved during the switch trial. Our findings are difficult to account for with models that posit a strong role for frontal cortical regions in "reconfiguring" the system during switches of task.

Cognitive control processes during an anticipated switch of task.

For successful negotiation of our environment, humans must be readily able to switch from one task to another. This ability relies on 'executive control' processes and despite extensive efforts to detail the nature of these processes, there is little consensus as to how the brain achieves this critical function. Behavioural studies show that as subjects are given more time to prepare to switch task, performance improves; yet even with the longest preparation intervals, there remains an ineradicable performance cost on switch trials. As such, some elements of the switching process must wait until the stimulus to be acted upon has actually been presented. Here, using the methods of high-density mapping of brain potentials, we show that early visual processes are substantially different on switch trials than on later trials. Our data show that while there is clearly a degree of preparatory processing that occurs prior to a predictable switch of task, some elements of switching are only achieved after the switch stimulus has been presented. Our findings are discussed in the context of a new model of executive control processes that suggests that preparing to switch task may not be a separate (control) process per se, but rather, the beginning of a competition between the potentially relevant tasks, a competition that is ultimately resolved during the switch trial.

Early visual processing deficits in schizophrenia: impaired P1 generation revealed by high-density electrical mapping.

Integrity of early visual sensory processing in schizophrenia was assessed using the well characterized P1 and N1 components of the visual evoked potential (VEP) as our dependent measures. VEPs were recorded in response to successively less fragmented line drawings of common objects. P1 amplitudes were significantly reduced across all stimulus conditions for patients versus controls. Further, this decrement was relatively greater at parieto-occipital than occipito-temporal electrode sites. No differences in N1 amplitude were found. The finding of P1 deficits in patients, particularly over dorsal scalp, supports the view that schizophrenia is associated with impairment of early dorsal visual stream processing. On the other hand, the finding of normal N1 amplitudes in patients suggests that early stages of ventral stream processing may be relatively more intact. These results imply that the cognitive impairment seen in schizophrenia is not just due to deficits in higher order aspects of cognition but also encompasses significant deficits in early sensory processing.

Visual activation of frontal cortex: segregation from occipital activity.

Studies in primates have found visually responsive neurons that are distributed beyond cortical areas typically described as directly involved in vision. Among these areas are premotor cortex, supplementary motor area, dorsolateral prefrontal cortex and frontal eye fields. Given these findings, visual stimulation would be expected to result in activation of human frontal cortex. However, few human studies have described sensory activations in frontal regions in response to simple visual stimulation. Such studies have classically described event-related potential (ERP) components over occipital regions. The present study sought to further characterize the spatiotemporal dynamics of visually-evoked electrocortical responses elicited by simple visual stimuli using scalp current density measures derived from high-density ERP recordings, with particular emphasis on the distribution of stimulus-related activity over frontal cortex. Hemiretinal stimuli were viewed passively and during a simple ipsi- or contramanual (RT) task. The motor requirement was included to investigate the effects of response preparation on premovement frontal activations. The results indicate early frontocentral activation, particularly over the right hemisphere (peak magnitude 124-148 ms) that is independent of input visual field or motor response requirement, and that is clearly separate in timecourse from the posterior responses elicited by visual input. These findings are in accord with the multiplicity of visual inputs to frontal cortex and are discussed in terms of frontal lobe functions as may be required in these tasks.

Attention-dependent suppression of distracter visual input can be cross-modally cued as indexed by anticipatory parieto-occipital alpha-band oscillations.

Recent studies show that in addition to enhancing neural processing for attentionally relevant stimuli, selective attention also operates by suppressing the processing of distracter stimuli. When subjects are pre-cued to selectively deploy attention during voluntary (endogenous) attentional tasks, these mechanisms can be set up in advance of actual stimulus processing. That is, the brain can be placed in a biased attentional state. Two recent cueing studies have provided evidence for the deployment of such biased attentional states [J.J. Foxe, G.V. Simpson, S.P. Ahlfors, Neuroreport 9 (1998) 3929-3933; M.S. Worden, J.J. Foxe, N. Wang, G.V. Simpson, J. Neurosci. 20:RC63 (2000) 1-6]. Specifically, these studies implicated oscillatory activity in the alpha frequency-band (8-14 Hz) as an anticipatory mechanism for suppressing distracter visual stimulation. The current study extends these findings by showing that this alpha-suppressive effect is also invoked by cross-modal cues. Auditory symbolic cues were used in an intermodal attention task, to direct subjects' attention to a subsequent task in either the visual or auditory modality. Cueing attention to the auditory features of the imminent task stimuli resulted in significantly higher parieto-occipital alpha amplitude in the period preceding onset of this stimulus than when attention was cued to the visual features. Topographic mapping suggests that this effect is generated in regions of the inferior parietal cortex, areas that have been repeatedly implicated in the engagement and maintenance of visual attention. Taken together, the results of this series of studies suggest that these parietal regions are capable of integrating sensory cues from multiple sensory modalities in order to program the subsequent deployment of visual attention.

Visuo-spatial neural response interactions in early cortical processing during a simple reaction time task: a high-density electrical mapping study.

The timecourse and scalp topography of interactions between neural responses to stimuli in different visual quadrants, straddling either the vertical or horizontal meridian, were studied in 15 subjects. Visual evoked potentials (VEPs) were recorded from 64 electrodes during a simple reaction time (RT) task. VEPs to single stimuli displayed in different quadrants were summed ('sum') and compared to the VEP response from simultaneous stimulation of the same two quadrants ('pair'). These responses would be equivalent if the neural responses to the single stimuli were independent. Divergence between the 'pair' and 'sum' VEPs indicates a neural response interaction. In each visual field, interactions occurred within 72-86 ms post-stimulus over parieto-occipital brain regions. Independent of visual quadrant, RTs were faster for stimulus pairs than single stimuli. This replicates the redundant target effect (RTE) observed for bilateral stimulus pairs and generalizes the RTE to unilateral stimulus pairs. Using Miller's 'race' model inequality (Miller J. Divided attention: evidence for coactivation with redundant signals, Cognitive Psychology 1982;14:247-79), we found that probability summation could fully account for the RTE in each visual field. Although measurements from voltage waveforms replicated the observation of earlier peak P1 latencies for the 'pair' versus 'sum' comparison (Miniussi C, Girelli M, Marzi CA. Neural site of the redundant target effect: electrophysiological evidence. Journal of Cognitive Neuroscience 1998;10:216-30), this did not hold with measurements taken from second derivative (scalp current density) waveforms. Since interaction effects for bilateral stimulus pairs occurred within 86 ms and require interhemispheric transfer, transcallosal volleys must arrive within 86 ms, which is earlier than previously calculated. Interaction effects for bilateral conditions were delayed by approximately 10 ms versus unilateral conditions, consistent with current estimates of interhemispheric transmission time. Interaction effects place an upper limit on the time required for neuronal ensembles to combine inputs from different quadrants of visual space ( approximately 72 ms for unilateral and approximately 82 ms for bilateral conditions).

Determinants and mechanisms of attentional modulation of neural processing.

This review contrasts the most-studied variety of attention, visuospatial attention, with several types of nonspatial visual attention. We: 1) discuss the manner in which spatial and nonspatial varieties of attention are experimentally defined, and the ecological validity of the paradigms in which they are studied, 2) review and compare differing effects of spatial and nonspatial attention on neural processing, 3) discuss the manner in which attention operates within the framework of an anatomical visual hierarchy, as well as 4) how attention relates to the temporal dynamics of visual processing, 5) describe cellular circuits and physiological processes that appear to be involved in attention effects, 6) discuss the relationship of attentional physiology to the perceptual and cognitive effects of attention, and 7) consider the strengths and limitations of several current models of selective attention. Throughout, we attempt to integrate the findings of monkey and human studies whenever possible. We have three main conclusions. First, two models, the Neural Specificity Model of Harter and colleagues and the Feature Similarity Gain Model of Treue and colleagues best incorporate findings in relation to both spatial and nonspatial varieties of attention. Significantly, these models explicitly note that the specific neuronal components used in attentional modulation of processing are flexible and determined by task demands. Second, current evidence also provides strong bases for deriving testable hypotheses about the specific brain mechanisms utilized by attention. Cellular processes, brain circuits and neurotransmitter components can and should be incorporated into our models of attention. Finally, it is increasingly evident that we can and should analyze temporal patterns of attentional modulation, both within and across brain areas. These patterns provide critical information on the dynamics of attention.