Neurophysiological Electroencephalography Signatures of Pediatric Sedation: A Comprehensive Comparative Review
Article information
Abstract
The utility of pediatric electroencephalography (EEG) is frequently compromised by motion artifacts and limited patient cooperation, necessitating pharmacological sedation to ensure signal quality. However, sedative agents produce distinct neurophysiological signatures that can obscure diagnostic landmarks or interfere with modern automated analysis pipelines. This review evaluates a five-agent framework comprising dexmedetomidine, chloral hydrate, midazolam, ketamine, and melatonin, with emphasis on their molecular mechanisms, developmental effects, and spectral influences on the EEG background. We highlight that biomimetic agents such as melatonin and dexmedetomidine preserve natural sleep architecture and interictal epileptiform discharges, whereas gamma-aminobutyric acid (GABA)-ergic (GABAergic) modulators such as midazolam and N-methyl-D-aspartate antagonists such as ketamine induce disruptive beta crowding and gamma bursts, respectively. The review also addresses age-dependent responses, including the neonatal GABAergic polarity shift, which may lead to paradoxical excitation. By integrating clinical safety profiles with signal-integrity requirements, we propose a tiered clinical decision-making framework that prioritizes diagnostic yield. Ultimately, a mechanism-based approach to sedative selection may help neurophysiologists optimize both patient safety and the precision of visual and digital EEG interpretation in the evolving landscape of pediatric neurology.
Introduction
In pediatric neurology, electroencephalography (EEG) remains a cornerstone for diagnosing epilepsy and monitoring encephalopathic states. However, its diagnostic utility is fundamentally constrained by raw signal quality. Pediatric patients present unique challenges, particularly physical movement and limited cooperation. To obtain recordings of sufficient duration and clarity, clinicians often rely on pharmacological sedation.
The primary objective of sedation during pediatric EEG is to reduce motion-related artifact while minimizing environmental interference. However, obtaining a diagnostic-quality recording requires clinicians to navigate a critical trade-off between patient quiescence and signal integrity. Although the sedative must reliably induce stillness to maintain electrode contact and reduce muscle artifact, it must do so without suppressing the underlying neurophysiological signal. Specifically, the selected agent should preserve true cortical oscillations so that both interictal epileptiform discharges (IEDs) and essential elements of sleep architecture, such as spindles and K-complexes, remain identifiable for accurate interpretation.
Traditionally, a ‘clean’ EEG was defined by a neurophysiologist’s visual ability to distinguish epileptiform spikes from background rhythms. However, this definition has evolved with the emergence of computer-assisted analysis and automated spindle detection.
For the modern clinician, a clean signal now implies preservation of the excitation-to-inhibition balance and of spectral characteristics that allow machine-learning models to function without high false-positive rates [1]. Many sedative agents introduce high-frequency pharmacological noise that may appear ‘calm’ on visual inspection but creates substantial spectral clutter for automated pipelines. This review examines five primary agents currently used in pediatric neurophysiology—dexmedetomidine, chloral hydrate, midazolam, ketamine, and melatonin—with a focus on their mechanisms of sedation, dosing strategies for EEG acquisition, and effects on EEG signal characteristics.
Pharmacological Profiles and Neurobiological Mechanisms
The primary neuroanatomical targets and site-specific mechanisms of action of the sedative agents discussed in this section are summarized in Table 1 [2-13].
Summary of neuroanatomical targets and site-specific mechanisms of action for five key sedative agents
1. Dexmedetomidine: α2-adrenergic agonism and the ventrolateral preoptic nucleus pathway
Dexmedetomidine is a highly selective α2-adrenoceptor agonist with a receptor affinity substantially greater than that of clonidine. Its sedative effects are mediated primarily through the locus coeruleus (LC) in the pons. At the cellular level, dexmedetomidine binds to presynaptic α2 receptors in the LC, thereby inhibiting norepinephrine release. This reduction in noradrenergic output disinhibits and subsequently activates the ventrolateral preoptic nucleus (VLPO), a critical sleep-promoting center. Once activated, the VLPO modulates the global arousal state by sending inhibitory gamma-aminobutyric acid (GABA)-ergic (GABAergic) and galaninergic projections to major wake-promoting systems, including histaminergic, orexinergic, and cholinergic pathways [14].
Consequently, this pharmacological pathway effectively mimics the endogenous transition to non-rapid eye movement sleep. Because dexmedetomidine acts upstream of primary cortical loops rather than producing widespread direct GABAergic depression, it preserves much of the brain’s natural oscillatory architecture [15]. This preservation is particularly advantageous for pediatric EEG because it allows a quiescent state while maintaining the integrity of diagnostic neurophysiological markers.
2. Chloral hydrate: the classical GABAergic standard
Despite its long history in clinical practice, chloral hydrate remains a commonly used sedative-hypnotic agent that functions primarily as a prodrug undergoing rapid hepatic metabolism. After administration, it is converted by alcohol dehydrogenase to its active metabolite, trichloroethanol (TCE). TCE acts as a positive allosteric modulator of the GABAA receptor, with a mechanism similar to that of barbiturates. By binding to the receptor complex, it prolongs chloride channel opening, thereby hyperpolarizing the postsynaptic neuron and enhancing inhibitory neurotransmission [16].
From a neurophysiological perspective, this mechanism induces a state of pharmacological sleep that is spectrally distinct from that produced by benzodiazepines. Importantly, chloral hydrate allows patients to cycle through the N1, N2, and N3 stages of sleep, thereby preserving the natural sleep architecture needed to capture sleep-activated epileptiform discharges. However, clinicians should note that TCE often produces a heavy delta background. This increase in low-frequency power may create a masking effect, potentially obscuring low-voltage fast spikes or subtle paroxysmal activity that would be more apparent against the cleaner background of melatonin-induced sleep [17].
3. Midazolam: benzodiazepine modulation and receptor saturation
As a rapid-acting, water-soluble benzodiazepine, midazolam is frequently used for procedural sedation and as a rescue intervention in cases of extreme agitation. Its primary mechanism of action is positive allosteric modulation of the GABAA receptor through binding at the interface of the α and γ subunits. In the presence of endogenous GABA, midazolam increases the frequency of chloride channel opening. This enhancement of inhibitory neurotransmission leads to widespread neural inhibition across the cerebral cortex and hippocampus [2].
This potent and direct action on cortical GABAergic circuits produces a distinct neurophysiological profile characterized by beta crowding, or a marked increase in high-frequency EEG activity [18]. Although these properties make midazolam an effective antiseizure agent, they also present a significant diagnostic challenge. Specifically, suppression of cortical excitability may mask the IEDs required for accurate epilepsy diagnosis, thereby limiting its utility when the primary objective is seizure-focus localization [19].
4. Ketamine: NMDA antagonism and cortical disinhibition
Classified as a dissociative anesthetic, ketamine induces a unique clinical state in which the patient may appear awake, often with open eyes, yet remains entirely nonresponsive to external sensory input. This phenomenon is driven primarily by its action as a noncompetitive antagonist of the N-methyl-D-aspartate (NMDA) receptor. Although ketamine inhibits excitatory glutamatergic transmission more broadly, its specific sedative profile is thought to result largely from its high affinity for NMDA receptors located on inhibitory GABAergic interneurons [20].
The resulting neurophysiological state is one of paradoxical disinhibition: by suppressing the interneurons responsible for regulating cortical excitability, ketamine permits uncontrolled firing of excitatory pyramidal neurons. This disruption manifests on EEG as high-frequency cortical oscillations, specifically gamma bursts, together with a profound breakdown of normal thalamocortical coordination. Unlike traditional sedatives that globally depress neural activity, ketamine reorganizes cortical dynamics into a dissociated, highly active, yet nonfunctional state [21].
5. Melatonin: chronobiological induction and signal purity
As an exogenous form of the pineal hormone that regulates circadian rhythm, melatonin acts as a physiological mediator rather than a traditional sedative. Its effects are mediated by two high affinity G-protein-coupled receptors, MT1 and MT2, located primarily within the suprachiasmatic nucleus (SCN) of the hypothalamus. Activation of MT1 receptors suppresses alerting signals generated by the SCN, whereas MT2 receptors mediate phase-shifting effects that synchronize the biological clock [22].
The resulting neurophysiological state is unique among agents used for pediatric sedation because melatonin does not force a change in consciousness through generalized neural inhibition. Instead, it promotes a natural sleep state by modulating endogenous sleep-wake transitions. Consequently, the resulting EEG is indistinguishable from physiological sleep, making melatonin the cleanest pharmacological option for preserving the native neurophysiological signal and maintaining diagnostic integrity [23].
Developmental Considerations in Signal Modulation
The pediatric brain is not a miniature version of the adult brain; rather, it is a rapidly evolving system characterized by fluctuating receptor densities, synaptic pruning, and shifting ion gradients. Consequently, EEG responses to the five sedative agents are highly nonlinear and age dependent [24]. Understanding these developmental milestones is critical for accurate interpretation of sedated EEG recordings.
1. The maturation of the thalamocortical loop
The rhythmic oscillations observed on scalp EEG, most notably sleep spindles and alpha rhythms, arise from complex interactions between the thalamus and cerebral cortex. During the neonatal period and first few months of life, however, these thalamocortical pathways have not yet achieved full myelination or synaptic synchronization [25].
This developmental immaturity of neonatal circuitry fundamentally alters the pharmacological induction of EEG signatures. In adults, dexmedetomidine reliably generates distinctive spindle activity at approximately 13 Hz by engaging endogenous sleep-wake circuits, but these rhythmic transients were historically considered absent in infants younger than 6 months [26]. Early observations suggested that dexmedetomidine administration in neonates produced only nonspecific slow-wave activity or generalized voltage attenuation. However, recent evidence has challenged the notion of complete absence: although no visually discernible drug effects were identified in newborns despite computational shifts in cortical synchrony [27], sleep spindles and N3 sleep stages have been documented in a cohort that included infants younger than 6 months [28]. These findings suggest that although the neural architecture required for robust spindle transients is still maturing, the capacity for dexmedetomidine-induced rhythmic oscillations may emerge earlier than previously recognized.
2. GABAergic polarity shifts: the KCC2/NKCC1 ratio
Perhaps the most important developmental factor in pediatric sedation is the phenomenon known as the GABA switch. Early in neural development, intracellular neuronal chloride concentrations remain high because of predominant expression of the sodium-potassium-chloride cotransporter 1 (NKCC1) transporter and relatively low expression of the potassium chloride cotransporter 2 (KCC2) exporter. Under these conditions, activation of GABAA receptors may cause chloride efflux rather than influx. This outward flow leads to postsynaptic depolarization and paradoxical excitation [29].
3. Spindle ontogeny and drug-induced morphological variants
Sleep spindles undergo a predictable developmental progression that must be considered during clinical evaluation. During the early postnatal period (0 to 8 weeks), spindles are largely absent from the EEG trace [30]. Between 2 and 6 months of age, these transients begin to emerge but are often asynchronous, frequently appearing over one hemisphere at a time rather than as a unified vertex phenomenon [31,32]. Only from approximately 2 years of age onward do spindles become consistently synchronous and assume the characteristic waxing-and-waning morphology of mature N2 sleep [33].
When dexmedetomidine is used, the resulting dexmedetomidine-induced spindles follow this same ontogenetic timeline, although they often have greater amplitude than endogenous spindles. Clinicians should recognize that the absence of spindles in a sedated 3-month-old infant is a normal developmental finding. This absence should not be misinterpreted as evidence of cortical dysfunction or pharmacological failure, but rather as a reflection of the patient’s current stage of thalamocortical maturation.
4. Synaptic pruning
As the brain matures and undergoes synaptic pruning, the cortical landscape of GABAA receptors undergoes substantial structural and functional reorganization. Although GABAA receptor binding, as measured by flumazenil positron emission tomography, is highest at approximately 2 years of age and declines exponentially thereafter, pharmacological sensitivity to benzodiazepines appears to increase with age [34]. This greater potency is thought to result from developmental shifts in receptor architecture, specifically an 8-fold increase in GABAergic synaptic frequency and increased expression of the α1 subunit [35]. Consequently, the pharmacological fingerprint of these agents becomes more prominent as inhibitory networks become more refined, even as overall receptor density declines. However, the clinical predictability of these transitions remains complex because evolving pharmacodynamics often interact with pharmacokinetic changes to produce variable responses across pediatric age groups [36].
Clinical Administration and Safety
The clinical utility of a sedative agent in pediatric neurophysiology is determined by a multidimensional trade-off: the agent must be potent enough to ensure successful recording on the first attempt while remaining sufficiently safe for use in potentially unmonitored outpatient settings. Practical administration parameters, including dosing ranges, onset times, and safety profiles for the five agents reviewed here, are summarized in Table 2 [37-53].
1. Procedural success and induction latency
Success in the EEG suite is often quantified by the induction-to-recording interval, which reflects the efficiency with which a patient transitions from wakefulness to a stable recording condition. Among commonly used agents, chloral hydrate and dexmedetomidine, particularly when dexmedetomidine is administered intranasally, have shown notable consistency in separate studies. The shift toward intranasal dexmedetomidine is increasingly regarded as a practical alternative for outpatient clinics, with some reports indicating that it produces a stable sedative state in approximately 90% of cases without the need for painful intravenous cannulation [37,38].
Chloral hydrate remains a commonly used option for pediatric EEG sedation, with reported procedural success rates ranging from 82.2% to 91% across specialized clinical studies [39,40]. Large laboratory-based reviews have also supported its reliability, with successful study completion documented in 91% of sedation attempts within those cohorts [40]. Although cross-study comparisons are limited by the absence of head-to-head trials, chloral hydrate appears to achieve sedation success rates similar to those of melatonin while maintaining high clinical utility for sleep induction [41]. Recent data continue to support its efficacy, showing that even conservative doses of 30 to 50 mg/kg yield sleep induction rates of approximately 82.2% [39].
In contrast, although melatonin has one of the most favorable safety profiles, it has been associated in some settings with clinical failure rates of up to 17% [41]. Consequently, melatonin-based protocols may more often require adjunctive approaches to complete the study when initial induction is unsuccessful [54].
2. Hemodynamic and respiratory safety profiles
Safety remains the principal barrier to routine pharmacological sedation in outpatient clinics. The risk profiles of these agents differ substantially, particularly with respect to respiratory drive and cardiovascular stability. Midazolam and chloral hydrate carry some risk of partial upper-airway obstruction and hypoventilation, especially in children with enlarged tonsils [55,56]. In contrast, dexmedetomidine has a distinctive hemodynamic profile, most notably a tendency to cause transient bradycardia [57]. Continuous heart-rate monitoring is therefore essential [58]. Melatonin has a generally favorable safety profile [59]. It is the only agent in the current clinical repertoire that can be administered safely with minimal specialized monitoring, which represents a major advantage in resource-limited or outpatient settings. Ketamine serves as a robust alternative in these settings, with a favorable safety profile relative to traditional sedatives. In emergency department settings, ketamine (14.6%) has shown complication rates comparable to those of midazolam (15.8%) and lower than those of propofol (22.2%) [60]. Pediatric data further support this stability, with studies documenting only transient respiratory effects and no serious complications such as laryngospasm or aspiration [61]. Although clinicians must manage unique trade-offs, including hypertension and emergence phenomena in 10% to 20% of adults, ketamine’s primary clinical advantage is preservation of respiratory drive [42].
3. Recovery kinetics: post-sedation monitoring
The hangover effect, or duration of postprocedural impairment, is an important consideration for outpatient workflow and patient safety after discharge. Chloral hydrate is known to produce prolonged sedation; because of the slow metabolic clearance of its active metabolite, TCE, some children may remain drowsy or ataxic for 4 to 12 hours after the procedure [62]. This extended recovery period necessitates prolonged clinical observation and complicates scheduling of subsequent diagnostic appointments. In contrast, dexmedetomidine is associated with faster recovery of consciousness, and ketamine may permit even more rapid recovery than midazolam [63]. Compared with midazolam alone, melatonin is also more favorable with respect to recovery kinetics [64]. Melatonin does not produce hangover effects, as assessed by mood and performance tests administered after treatment [65]. This rapid return to normal functioning is well suited to high-volume clinical environments because it minimizes the need for dedicated post-sedation recovery space and personnel.
Detailed Effects on EEG Signal Architecture
The hallmark of pharmacological sedation in neurophysiology is its specific modulation of cortical rhythms. These spectral signatures are not merely artifacts; rather, they reflect underlying alterations in thalamocortical and corticocortical communication. For clinicians, recognizing these patterns is essential to avoid mistaking drug-induced oscillations for pathological slow waves or paroxysmal discharges. A comparative overview of these drug-specific spectral signatures and their effects on power distribution across frequency bands is provided in Table 3 [13,66-76].
1. Spectral power shifts: delta, theta, and beta dominance
Sedative agents substantially alter the EEG power spectrum, with each pharmacological class driving rhythmic activity toward specific frequency bands. Dexmedetomidine primarily enhances low-frequency power in the delta and theta ranges [66]. This spectral shift closely resembles the energy distribution observed during natural slow-wave sleep [77]. In contrast, midazolam induces a marked rightward shift in the power spectrum toward the beta (12 to 30 Hz) range [67]. This beta crowding phenomenon departs from natural sleep architecture and may obscure lower-frequency components that are often important in diagnostic studies. Ketamine exhibits a distinctive bimodal power distribution that differentiates it from GABAergic and α2-adrenergic agents. Under ketamine sedation, extreme low-frequency (delta) and high-frequency (gamma) power coexist, whereas the intermediate alpha and beta bands are largely attenuated [78]. This spectral pattern reflects unconscious-state mechanisms and disruption of typical thalamocortical synchrony [79].
2. Morphological signatures: dex-spindles vs. gamma bursts
Visual inspection of the EEG reveals distinct transient waveforms that serve as reliable biomarkers of the administered sedative. Under dexmedetomidine, dexmedetomidine-associated spindles at approximately 13 Hz are a hallmark feature [26]. These transients are characterized by increased slow-delta oscillations across the scalp, increased theta oscillations in occipital regions, increased spindle oscillations in frontal regions, and decreased beta oscillations across the scalp [66]. In contrast, ketamine produces a readily identifiable gamma burst pattern [78]. This morphological signature consists of high-voltage delta activity punctuated by nested gamma oscillations. The pattern differs strikingly from any physiological sleep state and is often described as a dissociative rhythm. These high-frequency oscillations superimposed on slow-wave activity reflect paradoxical disinhibition of pyramidal neurons [80].
3. Preservation of sleep landmarks
Preservation of native sleep architecture is a critical diagnostic indicator of overall brain health and maturational integrity. Among the available sedatives, melatonin appears to provide superior signal preservation and diagnostic yield. It has been reported to identify electrographic abnormalities at a rate of 52.56%, more than double the 21.57% rate reported with chloral hydrate [41]. This pattern, observed across more than 348 pediatric patients, suggests that melatonin interferes less with spontaneous neurophysiological activity. Similarly, dexmedetomidine preserves EEG morphology with high fidelity. Although quantitative power shifts occur in slow-frequency bands, qualitative measures, including sleep-spindle density and spindle characteristics, remain comparable to those seen with chloral hydrate [68]. However, chloral hydrate consistently demonstrates the greatest risk of interpretive interference. Recent evidence indicates that melatonin is at least as effective as chloral hydrate for sleep induction while providing a more reliable baseline for pediatric neurodiagnostic interpretation [81].
Discussion
For patients with suspected epilepsy, the primary objective of pharmacological intervention is preservation of IEDs. Achieving a diagnostic yield that reflects the patient’s baseline pathology requires careful selection of agents that minimize interference with cortical excitability. Some evidence suggests that melatonin may provide greater diagnostic sensitivity than traditional sedatives, with seizure-activity detection rates reported at 52.56%–53%, compared with 21.57%–46% for chloral hydrate [17,41]. Despite these higher detection rates, melatonin’s pharmacological profile warrants caution; its GABAergic enhancement suggests potential anticonvulsant properties that could theoretically suppress IED yield, although empirical findings on this issue remain contradictory [82]. In contrast, dexmedetomidine appears to preserve IED integrity without substantial reduction across diverse seizure foci, producing qualitative EEG patterns comparable to those achieved with chloral hydrate [68,83]. Although chloral hydrate’s utility is often limited by reported antiseizure activity that reduces discharge detection [41], other agents such as midazolam and ketamine currently lack sufficient evidence regarding their specific effects on IED preservation. Consequently, although melatonin and dexmedetomidine appear to be favorable options for sedated EEG, the potential anticonvulsant effect of melatonin and the limited data on other commonly used sedatives necessitate a cautious, evidence-based approach to agent selection.
To preserve signal integrity and ensure accurate interpretation, clinicians must consider the specific ways in which pharmacological agents alter the EEG spectral background. The choice of sedative determines the level of background noise and the rhythmic distortions through which the neurophysiologist must interpret the recording. Available studies suggest that dexmedetomidine and melatonin preserve electrographic profiles more effectively than traditional sedative-hypnotics. Specifically, dexmedetomidine is associated with increased slow-wave power in the delta and theta ranges but lower total spectral power than chloral hydrate, while qualitative clinical interpretations remain largely consistent between the two agents [68]. In addition, dexmedetomidine induces distinct spindle activity at approximately 13 Hz that remains physiologically distinguishable from the alpha-range oscillations typically associated with propofol administration [26]. Melatonin has also shown high clinical utility, with a significantly higher rate of abnormal EEG detection (52.56%) than chloral hydrate (21.57%), suggesting a greater capacity to capture genuine epileptiform activity without excessive interference [41]. By contrast, chloral hydrate is associated with more pronounced high-frequency distortion, specifically elevated alpha and beta power. Finally, although midazolam and ketamine are frequently used in clinical practice, the current literature still lacks rigorous spectral analyses of their specific effects on EEG background architecture.
Conclusion
In conclusion, a clinical framework for pediatric EEG sedation that incorporates pharmacological mechanisms may be useful because each agent produces distinct neurophysiological signatures that can influence diagnostic features and automated analytic outputs. In the absence of large-scale, head-to-head pediatric trials directly comparing these agents, no single sedative can be considered optimal for all patients. Instead, selection should be individualized, balancing biomimetic options such as melatonin and dexmedetomidine against the specific diagnostic requirements and the patient’s clinical profile. Using different agents according to the monitoring environment and the goals of the study may maximize diagnostic yield while maintaining patient safety.
Notes
Conflicts of interest
Jaeso Cho and Anna Cho are managing editors of the journal, but they were not involved in the peer reviewer selection, evaluation, or decision process of this article. No other potential conflicts of interest relevant to this article were reported.
Author contribution
Conceptualization: JC, AC, and HK. Data curation: IY, IY, and HJY. Formal analysis: IY (Ilha Yune). Funding acquisition: HK. Methodology: IY (Ilha Yune). Project administration: JC. Writing-original draft: IY (Ilha Yune). Writing-review & editing: IY (Ina Youn), HJY, AC, and HK.
Acknowledgments
This research was supported and funded by SNUH Lee Kun-hee Child Cancer & Rare Disease Project, Republic of Korea (grant number: 26B-025-0200).
