Sign in to Peak Brain Path
Sign in to access your courses, books, and progress tracker. New here? Signing in creates your account automatically.
Want to explore courses first?
Browse courses and booksSign in to access your courses, books, and progress tracker. New here? Signing in creates your account automatically.
Want to explore courses first?
Browse courses and booksModule 14
Chapter 14 · 2 h · 8 quiz items · pass at 80%
This module covers IQCB Domain VI (Psychopharmacology), 10% of the exam. Medication status is a mandatory part of every QEEG interpretation: a benzodiazepine or a stimulant rewrites the map, and a practitioner who does not recognize the fingerprint will misread a drug effect as a trait finding. The quiz confirms the learner can identify a drug-class signature and state the documentation and washout requirements.
A recording arrives for interpretation. The data are clean, the artifacting is careful, the database is age-matched, and the map almost writes its own report: elevated frontal theta, a textbook inattentive picture. Then the intake form surfaces the line nobody flagged at scheduling. The client took an extended-release stimulant four hours before the cap went on. The numbers have not changed, but the clinical object has. You are no longer looking at this brain's resting physiology. You are looking at the same brain near a stimulant's peak, and the frontal theta you were about to call a phenotype may be the theta the medication failed to fully suppress, or the residue of a poor night's sleep the stimulant is papering over. The map is real. The interpretation you were about to commit to is not safe.
This is the single most common way a competent-looking QEEG goes wrong, and it is why Domain VI sits on the IQCB blueprint at all. Medication and substance exposure change the EEG you record and the norm you compare it against. A Diplomate who interprets a map without a documented medication history, and who signs a report that does not name the recording state, is reading a moment as though it were a trait. This chapter gives you the drug-class-by-QEEG thread you need at the map: which classes move the signal which way, why that matters against a normative database, when washout is even a question, and how the IQCB expects medication status documented on a report that has to stand up to review. The full pharmaco-EEG framework, the receptor-to-spectrum mechanisms, the worked medicated reports, and the per-compound detail live in the companion volume Your Brain on Drugs. This chapter is the certification subset: enough to interpret correctly and document defensibly, with the deeper reference one shelf over.
Normative databases are built almost entirely from unmedicated subjects. NeuroGuide, qEEG-Pro, the NYU lineage, and the open multinational sets all screen out or statistically down-weight medication effects in their reference samples. When you compare a medicated client to one of those norms, the z-score you compute is the distance between two states, the client's medicated brain and an unmedicated reference, not a measure of the client's own stable physiology. A benzodiazepine-driven beta excess scored against a benzodiazepine-free database reads as "hyperarousal" when it is a pharmacological GABA-A signature (Mandema et al., 1992). A stimulant-suppressed frontal theta scored against an unmedicated norm can read as unremarkable when the untreated brain might deviate sharply (Clarke et al., 2002). The database cannot tell you which. Only the medication history can.
The interpretive failure has a name worth carrying: treating a medicated, caffeinated, sleep-variable recording as a clean phenotype. The medication does not invalidate the map. It changes which question the map answers. A recording near a stimulant's peak answers "how does this brain look on medication," which is the right question if the client functions and trains on medication. It does not answer "what is this brain's untreated baseline," and scoring it against an unmedicated database as though it did is the error. Before the cap goes on, you have to know what state you are recording, because that determines which question the z-scores can address and what your report is allowed to say.
The rule is simple and non-negotiable for every map you interpret. A full medication and substance history is taken before the recording and documented on the record: prescription psychotropics, over-the-counter sleep and allergy agents, supplements, caffeine, nicotine, cannabis, alcohol, and any recent change to any of them. For each entry, capture compound, dose, formulation, last-dose timing, and whether that timing is typical. The most consequential medication fact is often the one nobody thought to ask, and on a QEEG report it is the difference between a defensible interpretation and a confident mistake.
Methylphenidate and the amphetamines raise catecholamine signaling and, in the EEG, push toward less slow activity and more fast activity. The primary pediatric anchor found that single-dose methylphenidate reduced frontal and central theta and increased posterior beta in children with ADHD (Clarke et al., 2002). The broader review literature supports the same direction with real heterogeneity across samples and ages (Loo & Barkley, 2005). Adult ADHD samples do not always show the pediatric beta-deficit pattern, so do not assume that pediatric normalization targets transfer to an adult map.
For interpretation against a database, the hazard is masking. A stimulant can suppress the very frontal theta a database would flag, so a near-peak recording can hide the deviation you would otherwise report. If the client is on medication during ordinary life, the treated map is the clinically relevant one, and you say so in the report. If the referral question is the untreated phenotype, an off-medication recording requires prescriber coordination and tight control of sleep and caffeine, because a skipped-dose, under-slept, energy-drink morning is not a clean baseline (one of the most common reasons a "clean" map is anything but). The non-stimulant attention agents differ and should not be collapsed in: atomoxetine is a norepinephrine reuptake inhibitor with delayed onset and CYP2D6-dependent exposure, read as steady-state rather than same-day peak (Clarke et al., 2009). Theta/beta ratio in adolescents and young adults warrants extra caution, because the ratio shifts with maturation and a high value can reflect development, sleep timing, or cannabis as much as attention physiology.
Benzodiazepines are positive allosteric modulators at GABA-A receptors, and they produce the most counterintuitive signature in this chapter. They augment beta while sedating the client (Mandema et al., 1992). The map looks fast while the person is slowed, foggy, or calm. The recognized qualitative fingerprint is spindling beta in roughly the 14 to 18 Hz range, often diffuse across the scalp rather than focal. On a z-score map this drives beta and high-beta deviations that look exactly like an endogenous hyperarousal pattern, and a benzodiazepine-associated beta and an anxiety-driven beta can be visually identical on a topographic plot. They do not mean the same thing. Treating them as interchangeable is the most common error in medicated-QEEG beta interpretation, and the report has to disambiguate them by exposure history, not by the map alone.
The documentation here is exacting. "Recorded two hours after PRN alprazolam" tells the next reader far more than "takes alprazolam as needed," because PRN use means the client may be unexposed on most days and acutely exposed during the recording. Chronic use is its own state: tolerance can leave the client feeling unaffected while the EEG still carries GABAergic shaping, and long half-life agents such as diazepam stay relevant long after the last pill. Discontinuation is both an interpretive and a safety matter. Benzodiazepine withdrawal can produce beta rebound, hyperexcitability, and seizure risk, so a recently tapered or missed-dose recording is labeled a withdrawal-state map, never a "pristine off-benzodiazepine baseline."
Antidepressants are among the most common exposures you will interpret, and they resist a single EEG signature. Effects vary by compound, dose, responder status, sex, and recording method, and the directions reported across studies are inconsistent (Saletu, Grunberger & Linzmayer, 1986). The practitioner-relevant caution is frontal alpha asymmetry, the antidepressant marker most likely to be over-read on a map: its predictive value is sex-dependent and method-sensitive, and a recent meta-analysis found a near-zero pooled effect for resting frontal alpha asymmetry in depression (Luo, Tang & Fan, 2025), so it should never be presented as a universal SSRI response biomarker. Alpha and individual alpha frequency also shift with reproductive-hormone state, so a low-alpha finding in a woman on an SSRI carries menstrual-cycle, contraceptive, and perimenopausal context that has to be weighed before the medication takes the credit.
SNRIs add noradrenergic activation on top of the serotonergic profile and may push arousal and beta a little harder, but the same conditional posture applies. For the report, name the compound, dose, duration, and any reported activation (insomnia, agitation, restlessness) or sedation, and adjust confidence in arousal-based findings rather than assigning them to a stable trait. Discontinuation produces its own state that is not a never-medicated baseline. A client who stopped venlafaxine two weeks ago is in a discontinuation map, and the report says so.
Typical (first-generation) antipsychotics are dopamine D2 antagonists that shift the EEG toward slowing: increased delta and theta, alpha slowing and profile change, sometimes reduced fast activity (Itil et al., 1987). The direction is dependable; precise magnitudes are not, and you should not attach a specific hertz value to the alpha shift, because the topographic literature does not support a clean number (Saletu et al., 2002). This slowing can mimic underarousal, frontal theta excess, a post-concussive picture, or an encephalopathic pattern on a z-score map, which matters because these clients frequently also carry anticholinergic adjuncts, benzodiazepines, nicotine, and disrupted sleep, all pushing the same direction. The slowing is dose-dependent: a recent dose increase or an acute intramuscular dose produces more than stable maintenance, and a recently hospitalized or acutely medicated client is a transition state, not a baseline.
Atypicals (second-generation) are more varied and should not be read as one class. Risperidone increased absolute delta and theta across leads after a single dose in healthy volunteers (Lee et al., 1999). Clozapine carries strong EEG effects and genuine seizure-threshold concerns, and its exposure is sensitive to smoking, because clozapine is a CYP1A2 substrate and quitting smoking can raise plasma levels and increase slowing on an unchanged dose, which puts smoking status in the record at every recording. Olanzapine slows less than clozapine in comparative work but is not EEG-neutral. Aripiprazole does not fit the sedating-slowing model and can produce activation or akathisia, so a beta-heavy or movement-contaminated map under aripiprazole should not be read automatically as anxiety. Low-dose quetiapine prescribed for sleep is a different clinical object than antipsychotic-dose treatment, and the report names the indication when it bears on interpretation.
This class is heterogeneous by mechanism and demands compound specificity. A global "mood stabilizer" label is too vague to interpret on a map. Lithium produces qualitative slowing and reduced alpha organization, not the beta increase sometimes loosely attributed to it, and a slowed map in a lithium-treated client is read alongside serum level, hydration, thyroid state, and toxicity context, because unusual slowing can signal a medical problem rather than a trait (Reischies & Neu, 2000). Lamotrigine runs opposite to the common misconception: it reduces delta and theta power while preserving alpha mean frequency, making it relatively EEG-sparing rather than alpha-elevating, so do not write that lamotrigine raises alpha (Clemens et al., 2007). Topiramate is associated with slowing and the cognitive-slowing complaints clients report on it, which on a map can mimic an underarousal phenotype (Cai, Chen & Wu, 2003). Valproate and carbamazepine can be sedating and slowing for some clients and carry their own medical and drug-interaction context. Gabapentin and pregabalin reach the EEG mainly through drowsiness and the conditions they treat (pain, sleep, anxiety) rather than a clean spectral fingerprint, and the broader antiepileptic literature supports class-level orientation without blurring compound differences (Perucca & Perucca, 2016).
Because several of these agents reduce slow activity or stabilize the record, a normal-looking map under them does not certify a normal untreated baseline. It may show treated stabilization, and a lamotrigine-normalized map is not the same object as an unmedicated one. Missed doses and tapers are labeled transition states, and for clients with epilepsy, anticonvulsant discontinuation is a seizure-safety matter that belongs to the prescriber.
Opioids act primarily at the mu receptor and tend toward slowing and altered arousal. The clearest evidence is in methadone maintenance, where resting EEG shows qualitative theta and beta differences versus controls (Liang et al., 2015). Evidence for many other opioid contexts (acute postoperative use, buprenorphine maintenance, fentanyl patches, tramadol) is sparse, so generalize cautiously and lean on practice-level synthesis where direct data are absent (Kerson, 2023). The interpretive trap is composite: in chronic pain the slowing you see on a map may be the opioid, the pain itself, poor sleep, a gabapentinoid, or an obstructive-sleep-apnea load, and the picture is rarely one clean cause. Tramadol deserves separate caution, because it is opioid-like and serotonergic-noradrenergic and carries seizure-threshold concerns, so a tramadol map is not a simple opioid map. Withdrawal flips the direction toward hyperarousal, insomnia, and beta or artifact elevation, and is a transformation state rather than an off-medication baseline. Stable maintenance on methadone or buprenorphine is best read as the treated state, named as such, not removed from the interpretation.
Cannabis resists a single signature, and the most important correction is that chronic use does not produce one consistent alpha direction. The heterogeneous picture across cohorts includes reduced delta and theta with elevated beta and gamma in some chronic-use samples (Prashad, Dedrick & Filbey, 2018), and regionally dissociated alpha findings, with frontal and posterior patterns diverging, in long-term users during abstinence (Struve et al., 2008). Acute intoxication is a state recording, not a phenotype map. Product, route, potency, and last use all matter, and a single intake checkbox for "marijuana" does not separate a high-THC concentrate, a balanced edible, and wellness-dose CBD (which has little reliable resting-EEG signature and is not the same exposure as THC). Inhaled forms clear faster than edibles, which undergo hepatic conversion to a longer-acting metabolite, so the abstinence interval for a clean baseline is longer for edibles. Adolescent cannabis use intersects with developmental maturation and should not be read through adult frameworks. The older claim that acute cannabis reliably raises resting alpha does not hold up in the audited literature, so the heterogeneity itself is the interpretive caution; the full per-compound treatment is in Your Brain on Drugs.
Alcohol is common, underreported, and strongly state-dependent, and separating acute, withdrawal, chronic-use, and abstinent states is the whole interpretive game. Acute low-to-moderate alcohol tends to increase alpha (Lukas et al., 1986). The chronic and familial-risk literature points the other way at the trait level, supporting elevated beta as a persistent signature in alcohol-use disorder and family-risk populations (Rangaswamy et al., 2002; Ehlers et al., 1991). The certification-critical point is the pristine-baseline problem: a recording during abstinence after heavy use is not a never-exposed baseline, because elevated beta and other signatures can persist for months to years (Bauer, 2000). Elevated beta on the map of a client with an alcohol history reads differently from beta in a caffeine-anxious client, and alcohol history belongs in the differential before you label that beta as anxiety, trauma-related hyperarousal, or artifact. Withdrawal is a safety matter, carrying seizure and kindling risk, and a medical-care priority rather than a phenotype. Ask about drinking directly and without judgment, because shame produces inaccurate histories and an inaccurate alcohol history weakens every z-score downstream.
Whether to record on or off a substance is a clinical decision tied to the question the map has to answer, and it is constrained by safety. The literature does not support precise universal washout intervals for most agents, so the standards below are cautious heuristics, not validated rules, and the report frames them as such.
For acute exogenous substances the client can reasonably be asked to hold, caffeine, nicotine, cannabis, and alcohol, standard pre-recording cutoffs apply, and the cleaner the hold the cleaner the baseline, provided the hold does not itself create a withdrawal state that is less representative than ordinary use. For prescribed psychotropics taken daily, the usual approach is to record before that day's morning dose, capturing a defined trough without a medication skip, then have the client dose immediately after. A true off-medication washout is a heavier decision. For stimulants, an extended hold is sometimes used to evaluate the unmedicated phenotype, and because that is a timing hold rather than a discontinuation, it still requires prescriber coordination and control of sleep and caffeine to mean anything. For benzodiazepines, antipsychotics, anticonvulsants, and antidepressants, you do not direct an off-medication recording on your own initiative: discontinuation of these classes carries medical and psychiatric risk and belongs to the prescriber. The boundary is firm. A QEEG practitioner does not start, stop, or change medication and does not ask a client to do so unsupervised, no matter how much cleaner the map would be.
The medication note is a load-bearing part of the QEEG record, not a formality, because the next person to read the map, a referring clinician, a re-assessing Diplomate, an opposing expert, or the future you, cannot reconstruct the recording state from the z-scores alone. For every map, the record captures each medication and substance with compound, dose, and formulation; last-dose timing and whether it is typical; recent regimen changes (new starts, dose adjustments, tapers, missed doses); caffeine, nicotine, cannabis, and alcohol in the relevant window; and sleep the night before.
The report language then names the state explicitly and keeps every claim conditional. "Recorded near the expected peak of an extended-release stimulant; frontal theta was not elevated under this treated condition, so the unmedicated theta phenotype cannot be determined from this recording alone" is defensible. "Stimulants normalized the QEEG" is not. "Beta excess is present under recent benzodiazepine exposure; interpretation as endogenous hyperarousal is reduced, because benzodiazepines augment beta while sedating" is defensible. "The beta excess indicates anxiety" is not. The phrasing that protects you and serves the client names the exposure, names what it does to confidence, and stops short of causal claims the recording cannot support.
The IQCB holds the Diplomate's report to a documentation standard, and medication is one of the elements a reviewer checks. A report on a medicated client is expected to identify the normative database and version used, state the recording conditions, and explicitly disclose the medication and substance state under which the data were collected, so that a second qualified reader can judge whether the z-scores mean what the report says they mean. A medication-confounded map is not disqualified from interpretation, but it is interpreted within stated limits: the report describes the quantitative findings, names the agents that plausibly contribute to them, and constrains the conclusions accordingly rather than reporting deviations as though the client were unmedicated.
Two requirements follow directly. First, the report does not present a z-score deviation that a documented medication is known to produce as evidence of an inherent pattern. It flags the confound and lowers confidence. Second, when the medication picture is too entangled to interpret cleanly, the report says so plainly and, where appropriate, recommends prescriber coordination or a repeat recording under a defined state, rather than forcing a phenotype label the data cannot carry. A report that omits medication status, or that reads a drug effect as a trait, is the kind of report the mentoring review process exists to catch before it reaches a client or a court.
Polypharmacy is the rule rather than the exception in real practice, and it concentrates in older adults, where psychiatric agents stack on top of medical ones and clearance slows (Mojtabai & Olfson, 2010). An older client may arrive on an antidepressant, a sedative-hypnotic, an antipsychotic for sleep or agitation, a gabapentinoid for pain, an anticholinergic for bladder or allergy, and cardiac or metabolic medications, several of which independently push the EEG toward slowing. The problem is additive-slowing: combined sedative and anticholinergic load can produce diffuse slowing and reduced alpha organization that, scored against a young-adult-weighted database, reads as marked deviation when much of it is pharmacological and age-related rather than a discrete phenotype.
Two cautions belong in any older-adult interpretation. First, attribution defeats you when three or more psychoactive agents are present: you can describe the overall pattern and flag what persists despite medications that should suppress it, but you cannot assign individual spectral features to individual drugs, and the report says that rather than pretending to a clean decomposition. Second, the comparison itself needs an age-appropriate norm, because alpha slowing and frontal theta increase modestly with normal aging, and a database that does not match the client's age inflates the apparent deviation before any drug is considered. The defensible report on a polymedicated older adult names the stack, uses an age-matched database, interprets conservatively, and recommends prescriber review when the medication burden or an unexpected slowing pattern warrants it.
Strip this chapter to what you do when a map lands for interpretation. You confirm a full medication and substance history was taken and documented before the recording, every time, including caffeine, nicotine, cannabis, and alcohol, with compound, dose, and last-dose timing. You establish which state was recorded and therefore which question the z-scores can answer, and you keep your report conditional and state-specific. You never read a documented drug effect as a trait, you label withdrawal and transition states as such, and you never direct an unsupervised washout of a class that is dangerous to stop. You meet the IQCB documentation standard by naming the database, the recording conditions, and the medication state on the report, and by constraining conclusions on any medication-confounded map. And you forward-reference Your Brain on Drugs when a regimen raises a question this chapter does not answer, because the full pharmaco-EEG reference lives there.
For the IQCB exam, hold the directions and the traps. Stimulants reduce theta and raise beta, and a near-peak recording masks the attention phenotype. Benzodiazepines augment beta while sedating, so benzodiazepine beta is not anxiety beta, and the spindling 14 to 18 Hz pattern is the tell. Typical antipsychotics and many opioids slow the record; topiramate slows it too. Lithium produces slowing, not beta. Lamotrigine reduces slow activity rather than raising alpha. SSRIs are variable, and frontal alpha asymmetry is not a universal antidepressant biomarker. Cannabis is heterogeneous, not uniformly alpha-elevating. Alcohol raises alpha acutely while chronic alcohol use augments beta, and abstinence is not a never-exposed baseline. Normative databases assume an unmedicated subject, which is why medication status is mandatory for every map and why the recording state must be documented on every report a Diplomate signs.