Full DXM FAQ audit

FAQ update

DXM FAQ audit · 5 October 2026.

Updating the DXM FAQ

Scientific audit of William E. White’s version 4.0 and the contemporary model of DXM action

Edition dated 5 October 2026

The DXM FAQ retains value as a historical collection of observations about dextromethorphan’s effects. Its explanations involving PCP₂, sigma receptors and the sequential engagement of targets at different “plateaus” need revision. A contemporary model must account for DXM itself, its active metabolite dextrorphan, further metabolism and formulation. Clinical data confirm the importance of CYP2D6 and drug interactions, but do not allow individual subjective experiences to be assigned to specific receptors.

The subject of this audit is The Dextromethorphan FAQ, Text Version 4.0, with an introduction dated 15 March 1997. Later FAQ editions fall outside the scope of this assessment. White’s separate 1998 article on neurotoxicity and his 2004 withdrawal of its conclusions are considered separate publications. Unless stated otherwise, numbers following § refer to sections of the original FAQ; references 1–33 lead to this document’s bibliography.

This is a targeted source-critical and pharmacological audit of selected FAQ claims, rather than a systematic review of all DXM literature or a comprehensive check of every chapter. The scope is shown in the table in §18 of this document: each entry includes an original text fragment, an editorial decision and sources. Priority is given to primary studies and regulatory documents. Where only an abstract is available, conclusions are limited to its contents. Regulatory information concerns the United States and is current as of the edition date.

1. What White already knew

In §5.15 White explicitly acknowledges that both DXM and DXO are psychoactive. The claim that “the old FAQ considered only DXO active” is therefore incorrect for version 4.0. §9.5 already describes CYP2D6, alternative metabolism and variability in effects; other sections discuss polistirex, combination products, tolerance and withdrawal. These topics should be clarified while preserving correct historical attribution. 1

White also warns about fatal combinations with monoamine oxidase inhibitors. However, §6.2.9 labels serotonin syndrome “Theoretical” while citing a clinical case. This internal inconsistency needs correction, rather than a supposed complete absence of a warning. 1

The introduction mentions approximately 500 current and former users with whom the author communicated. These are not 500 standardized observations or a validated sample for estimating complication rates. White himself prefaces the pharmacological discussion with caveats about speculation. An updated FAQ must not turn these into confident scientific conclusions. 1

2. DXM and DXO in the contemporary metabolic model

Five principal metabolic transitions

The diagram shows reactions from the table in audit section 2. The reaction list is not exhaustive.

  1. O-demethylation predominantly by CYP2D6
  2. N-demethylation involving CYP3A
  3. N-demethylation involving CYP3A
  4. O-demethylation involving CYP2D6
  5. Conjugation by UGT enzymes, primarily the UGT2B family

Audit §2: pathways, conjugation and interpretation limits →

DXM is dextromethorphan; DXO is dextrorphan. Both molecules are pharmacologically active, but their profiles differ. Their ratio changes over time and depends on metabolism; it is not a fixed property of the dose taken. 2, 3, 4, 6, 7, 8

Starting substanceConversionPrincipal pathway
DXMDXOO-demethylation predominantly by CYP2D6
DXM3-methoxymorphinanN-demethylation involving CYP3A
DXO3-hydroxymorphinanN-demethylation involving CYP3A
3-methoxymorphinan3-hydroxymorphinanO-demethylation involving CYP2D6
DXODXO-O-glucuronideConjugation by UGT enzymes, primarily the UGT2B family

This outlines the principal pathways, rather than every reaction. Both demethylation pathways converge on 3-hydroxymorphinan. In one sentence in §9.5 of the original FAQ, the route through 3-methoxymorphinan incorrectly starts from DXO; its meaning and the molecular structures indicate that it starts from DXM. 1, 2, 3

Glucuronidation is an important pathway of further DXO metabolism. In the experiments by Lutz and Isoherranen, the investigated UGT2B4, UGT2B7, UGT2B15 and UGT2B17 formed DXO-O-glucuronide. A contemporary PBPK model treats DXM, DXO and its glucuronide as separate substances. 2, 3

When interpreting pharmacokinetic data, distinguish unconjugated DXO, total DXO after hydrolysis of conjugates, and the fraction of unconjugated DXO that is not bound to proteins. The English word free can designate different quantities in different methods. Total plasma DXO must not be substituted for the concentration of active unconjugated substance in the brain. 2, 3

The glucuronide is more polar; its presence changes the interpretation of the plasma DXM/DXO ratio. This does not imply that it is entirely inactive, cannot enter the CNS at all, or that DXM necessarily predominates in the brain. The percentage conjugated depends on conditions and measurement method. UGT variability warrants investigation, but the cited studies have not established its role as a predictor of subjective DXM effects.

3. CYP2D6 and variability in human effects

In the 1996 study by Capon and colleagues, after 30 mg DXM, the area under the DXM plasma concentration curve, AUC, was approximately 150 times higher in poor metabolizers than in extensive metabolizers. Adding quinidine increased AUC in extensive metabolizers approximately 43-fold. These are results of a specific study, rather than universal multipliers for every dose, formulation or inhibitor. 4

The contemporary CYP2D6 categories are poor, intermediate, normal and ultrarapid metabolizer, abbreviated PM, IM, NM and UM. Phenotype is inferred from allele function and copy number using an activity score. The term extensive metabolizer, EM, is retained only when describing older studies: their classification criteria may differ from contemporary ones. Drug inhibition can change actual enzyme activity without changing genotype; this is called phenoconversion. 3, 5

Genotype does not determine all differences between people. The 2022 PBPK study by Rüdesheim and colleagues shows substantial variability even within groups with the same activity score. Formulation, timing, concomitant substances and subsequent metabolism also matter when interpreting effects. 3

The statement in §9.5 that PM experience fewer effects needs correction. The 1998 pilot study by Zawertailo and colleagues included four EM and two PM. At 3 mg/kg, PM showed greater impairment on some psychomotor measures, sedation and dysphoria, whereas EM showed greater positive reinforcement potential. The small sample precludes a rule for every individual, but demonstrates that lower DXO formation does not mean weak or safe effects. 6

A separate 2010 study by Zawertailo included eight volunteers. In a single-blind, within-subject study, quinidine increased DXM concentrations and reduced DXO; ratings of euphoria and drug liking decreased, while unpleasant effects increased. This supports the dependence of the subjective profile on metabolism, but does not reduce every experience to a single molecule. 7

4. Target affinity and the limits of numerical comparisons

DXM / DXO · Ki · µM

Xiao et al. table, audit §4. A common logarithmic axis; further left is a lower Ki in that assay.

The table compares Ki values from different biological systems. Ki describes binding; functional potency, clinical effect and brain target occupancy are distinct characteristics.

NMDA

● DXM 1.59 · ■ DXO 0.22 · n in the article = 4

σ1

● DXM 1.67 · ■ DXO 2.85 · n in the article = 2

α3β4 nAChR

● DXM 14.70 · ■ DXO 9.98 · n in the article = 2

SERT

● DXM 0.0084 · ■ DXO 0.14 · n in the article = 2

NET

● DXM 20.51 · ■ DXO 5.67 · n in the article = 2

n is the number of experiments reported in this article.

Audit reference №8 · Assay systems and conflict of interest · Plasma concentration and receptor occupancy

Below is the complete DXM and DXO Ki table from Xiao and colleagues’ study of SAL0114. Units are μM; a lower Ki means higher affinity in that assay. 8

TargetDXMDXOn in the article
NMDA1.590.224
σ11.672.852
α3β4 nAChR14.709.982
SERT0.00840.142
NET20.515.672

Under these conditions, DXO has approximately 7.2 times greater affinity for the NMDA site, DXM has 16.7 times greater affinity for SERT, and DXO has 3.6 times greater affinity for NET. The last comparison concerns binding, rather than demonstrated functional potency or a clinical contribution of NET. 8

The assays used different biological systems; for example, NMDA was investigated in rat brain material and σ1 in human Jurkat cells. Radioligand binding is not equivalent to channel blockade or transport inhibition. These numbers must not become a universal scale of target engagement; for α3β4 it is especially important to distinguish binding from functional assays. 8

This is a preclinical study of deuterated DXM with bupropion. All authors disclosed employment at Salubris. The conflict of interest requires disclosure and independent confirmation of the conclusions, but does not in itself invalidate the table. 8

5. Why plasma concentration does not give receptor occupancy

Free plasma DXM concentration and Ki cannot reliably determine brain target occupancy. Such a calculation would require concentrations at the target, plasma-to-brain distribution, DXO dynamics and a binding model applicable to the specific receptor. Ki from a competitive radioligand assay cannot automatically be treated as the dissociation constant in a living brain. For noncompetitive NMDA channel blockade, channel activity and state matter. 8, 29

Binding data allow discussion of the relative target profile, but do not establish actual occupancy or contribution to an effect. They do not demonstrate the absence of an NMDA component in the therapeutic range or its appearance only above a particular dose.

NMDA antagonism remains a supported explanation for a substantial part of DXM/DXO’s dissociative action. However, the precise contributions of the two molecules and other targets depend on exposure and have not been determined for each subjective effect. Regulatory descriptions of DXM as an NMDA antagonist do not resolve this quantitatively. 8, 13, 14, 29

6. Historical DM₂ and PCP₂

The FAQ compares the high-affinity DXM binding site DM₂ with PCP₂ and tentatively interprets PCP₂ as the dopamine reuptake complex. Explanations of stimulation, musical effects and the first “plateau” are derived from this. This is a historical hypothesis that must not be retained as established contemporary pharmacology. 1

Meoni, Tortella and Bowery did use the designation DM₂ in 1997. They demonstrated sodium-dependent high-affinity DXM binding in rat brain, a distribution similar to paroxetine’s, and reciprocal ligand displacement. The study supports an association between the investigated DM₂ component and the serotonin uptake system. 9

However, it does not establish that every site called PCP₂ in different experiments is unambiguously identical to contemporary SERT. Historical designations were defined by assay conditions, rather than identification of a single protein. It is accurate to state that some older high-affinity observations fit serotonin transport better than the earlier dopamine interpretation. 9

The claim that “PCP₂/DAT explains the first plateau” should be removed. DXM’s high affinity for SERT supplies a more contemporary pharmacological context, but likewise does not prove that SERT alone produces euphoria, stimulation or changes in music perception. 8, 9

7. Sigma receptors and the endopsychosin hypothesis

σ1 is a molecularly identified protein associated with the endoplasmic reticulum and regulation of cellular processes. It is usually described as a ligand-regulated chaperone and modulator; it is not a conventional neurotransmitter receptor with a single linear output. The crystal structure of human σ1 was published by Schmidt and colleagues in 2016. 10

For DXM, there are σ1 binding data and preclinical data on the functional involvement of this target. In Nguyen and colleagues’ 2014 study, the antagonist BD1063 significantly attenuated DXM’s antidepressant-like effect in the mouse forced swim test. For BD1047, the corresponding comparison showed only a statistically nonsignificant trend. It would be incorrect to state that both antagonists demonstrated the mechanism equally convincingly. 12

A behavioral test result in mice does not establish the mechanism of depression treatment in humans, much less explain specific subjective features of intoxication. Attributing “insight”, mystical experience, depth of dissociation or an individual “plateau” to σ1 remains hypothetical. 12, 29

σ2 was identified as TMEM97 in 2017. It is a different protein, so σ1 and σ2 must not be merged into a single undifferentiated sigma system. The older designation σ3 should not be used as an established third molecular target of DXM without separate confirmation. 11

Endopsychosin in FAQ §10.7 is a historical candidate for an endogenous system, proposed within the hypotheses of that period. None of this audit’s sources confirms that this proposed mediator explains DXM’s action. The section may be retained as a history of ideas with an explicit indication of uncertainty. 1

8. The status of the four plateaus

“Plateaus” can be retained as language for describing common subjective states. In the original FAQ, they summarize user observations, rather than clinically validated stages with reliable dose boundaries. 1

The sequence “first PCP₂, then sigma, then NMDA” has not been confirmed by direct measurements of human receptor occupancy. The balance of active substances, absorption rate and individual conditions can change effects without crossing thresholds that are identical for everyone. The diagram in §9.3 must not be reproduced as a measured relationship. 1, 3, 6, 7, 8

Nonlinear changes in subjective experience may reflect pharmacokinetics and neural network activity. This is a possible explanation, rather than evidence of exactly four biological phases. The number of “plateaus” must not be derived from the number of known targets.

9. Controlled comparisons of subjective effects

Carbonaro and colleagues compared DXM, psilocybin and placebo in 20 healthy participants experienced with hallucinogens. They used a within-subject, double-blind design, DXM 400 mg/70 kg and three psilocybin doses. These are the research conditions of a specific experiment, rather than a description of a typical outcome in every user. 13

DXM was more often accompanied by impaired balance, dizziness, vomiting and a sense of separation from the body. Psilocybin had stronger effects on some visual, mystical and music-related ratings. With DXM, 12 of 20 participants, 60%, selected the dissociative class as most similar; the mean similarity rating for that class was 65%. These are different measures. Vomiting occurred in 11 of 20, 55%; this percentage must not be generalized to all doses and formulations. 13

In Shram and colleagues’ 2023 study of esmethadone’s abuse potential, DXM was an additional comparator in the Ketamine Study. The crossover study, with 51 completers, compared oral DXM 300 mg, intravenous ketamine 0.5 mg/kg and other conditions. Maximum drug-liking ratings for DXM were lower than for ketamine. 14

This experiment involved selected recreational users and primarily investigated abuse potential. Routes of administration and pharmacokinetics differed; the doses were not demonstrated to be equivalent for dissociation. The study confirms that a direct comparison exists, but does not establish complete equivalence or fundamental incomparability of DXM and ketamine experiences. 14

10. Serotonin toxicity and drug interactions

Serotonin syndrome should be described as a documented clinical complication, particularly in combinations. DXM inhibits serotonin reuptake; it should not be described without qualification as a conventional serotonin-releasing agent. The case report and literature analysis by Monte, Chuang and Bodmer also demonstrate the importance of chlorphenamine, also known as chlorpheniramine, in combination products. 15

Interactions can combine pharmacodynamic enhancement of serotonin transmission with a pharmacokinetic increase in DXM exposure. These mechanisms should be considered separately.

Combination or classWhat an updated FAQ should reflectSources
MAOISubstantial risk of a severe interaction, already known to White. A fatal case of DXM and moclobemide overdose was described in 2025.1, 16, 29
Linezolid and intravenous methylene blueHave MAO-inhibiting activity and are explicitly listed among contraindicated Auvelity combinations.29
SSRI and other serotonergic medicinesPharmacodynamic amplification of toxicity is possible; the extent of the interaction depends on the substance and conditions.15, 29
Fluoxetine and paroxetineSerotonergic action combines with CYP2D6 inhibition; they must not be interpreted solely as a way to reduce DXO formation.15, 27, 29
DXM with chlorphenamineChlorphenamine itself affects serotonin uptake and has its own toxicity. A combination product’s risk must not be attributed solely to DXM.15
DiphenhydramineCan inhibit CYP2D6; it also causes sedation and antimuscarinic effects. The old “probably safe” in §6.1.3 needs removal.1, 17, 27

The Auvelity label specifies 14-day separation when switching between the medicine and MAOI and separately lists linezolid and intravenous methylene blue. This is a rule from a specific label; it must not be turned into a universal protocol for all substances with any degree of MAO inhibition and all emergency situations. 29

The human diphenhydramine study confirms an interaction with another CYP2D6 substrate, metoprolol. It supports the inhibition mechanism, but does not specify an exact multiplier for changes in DXM concentrations. 17

In 2008, Forget and colleagues described life-threatening intoxication from DXM with amitriptyline in a PM, and a rechallenge investigation. This is an important clinical signal, but a single case does not determine risk frequency or allow the whole interaction to be reduced to further inhibition of already low-activity CYP2D6. 18

11. CYP3A and grapefruit juice

CYP3A participates in alternative DXM metabolism and further DXO conversion. Discussion of CYP2D6 alone is therefore incomplete. 2, 3

Di Marco and colleagues investigated grapefruit and Seville orange juice in 11 volunteers. The authors interpreted the pharmacokinetic changes in terms of intestinal CYP3A and possible involvement of P-glycoprotein or related efflux systems. Strauch and colleagues’ 2009 study additionally investigated dependence of the interaction on juice quantity using urinary metabolite profiles. 32, 33

These data do not support viewing grapefruit as a selective, predictable CYP2D6 inhibitor. They cannot establish a universal “potentiation” multiplier, calculate the degree of dissociation, or demonstrate changes in DXM transport across the blood–brain barrier. Investigating intestinal pharmacokinetics is not equivalent to measuring brain exposure.

12. Formulation and accompanying ingredients

Immediate-release DXM and extended-release polistirex must be distinguished. In polistirex, DXM is bound to an ion-exchange resin. The Delsym label states the equivalent of 30 mg DXM hydrobromide in 5 ml and a 12-hour formulation. Equivalent active-substance quantity does not mean identical Cmax, Tmax or time courses of effects. The FAQ’s suggestion that DXM hydrobromide may be absorbed in 30 minutes must not be interpreted as a universal time to complete absorption. 1, 19

The phrase “equivalent to DXM hydrobromide” expresses dosage and does not itself prove an identical bromide load. Bromism associated with the hydrobromide salt must not automatically be generalized to polistirex. Extended release nevertheless does not eliminate DXM’s own toxicity or the possibility of prolonged intoxication. 19, 27

Auvelity is also called an extended-release tablet, but contains immediate-release DXM and extended-release bupropion. This combination must not be equated with DXM polistirex. 29

Combination cold remedies require separate description. Paracetamol adds a risk of liver injury; sedating antihistamines, including doxylamine, diphenhydramine and promethazine, have their own central and antimuscarinic effects; pseudoephedrine and phenylephrine add sympathomimetic effects. Chlorphenamine also has a serotonergic component. Guaifenesin’s presence also changes product tolerability. These risks must be tied to the specific composition, rather than the label “DXM syrup”. 15, 27

13. Bromism and the anion gap

Bromism from chronic bromide exposure is real. Ng and colleagues’ 1992 article describes a patient after years of using a combination syrup containing DXM hydrobromide and paracetamol. The abstract explicitly states pseudohyperchloremia and a negative anion gap. 20

Bromide can interfere with chloride measurement in some analytical methods: the result is falsely elevated and the calculated anion gap decreases, sometimes becoming negative. This is analytical interference, rather than necessarily a true chloride elevation. The FAQ’s wording about an increased anion gap needs correction. 20

A low anion gap is not equivalent to measuring bromide concentration and is not specific to bromism alone. The advice in §6.8.3 that these tests amount to practically the same thing is therefore also incorrect. Risk depends on bromide load and exposure conditions, rather than only on the presence of the psychoactive molecule DXM. 1, 20

14. Neurotoxicity and structural findings

Olney lesions, or classic NMDA antagonist neurotoxicity, NAN, refer to specific pathological changes described in animal models. This term must not be used as a synonym for any memory deterioration, MRI changes or toxic effects of dissociatives.

In Carliss and colleagues’ 2007 study, no corresponding vacuolation or neurodegeneration was found in the retrosplenial and posterior cingulate cortex in rats under the tested oral DXM regimens. However, other signs of toxicity were noted, including impaired balance, reduced activity and mortality. A negative result for a particular injury type does not mean the substance is harmless. 21

Positive preclinical findings also exist. In 2016, after high-dose intraperitoneal DXM, 80 mg/kg, Tran and colleagues found myelinoid bodies and mitochondrial dysfunction in rat hippocampus. This is a different route of administration and a different pathological outcome; it does not demonstrate classic Olney lesions in humans. Such a dose cannot be directly extrapolated to humans on an mg/kg basis. 22

White withdrew the strong conclusions of his separate 1998 article in a response dated 28 December 2004. This was the author’s withdrawal of conclusions from an internet publication, rather than a journal article retraction. He acknowledged that the collected reports were insufficient and biased for concluding that NAN occurs in humans. This episode must be distinguished from the 1997 FAQ text. 23

In humans, Qiu and colleagues compared MRI scans of 38 adolescents and young adults dependent on DXM-containing syrups with 18 controls. They found increased cortical thickness in several regions and increased volumes of some subcortical structures. The study was published online in 2016 and in a journal issue in 2017. 24

This is a cross-sectional group comparison, rather than proof of neuronal death from pure DXM. It does not establish the direction of causality, rule out baseline differences, or determine the role of every syrup ingredient. The appropriate conclusion is that structural associations have been described, but these studies do not demonstrate specific human Olney lesions or their frequency. Neither the absence of all structural findings nor established universal brain damage can be asserted.

15. Tolerance, dependence and withdrawal

These topics appear in the FAQ and need clarification with clinical data. Dependence and withdrawal symptoms have been described in the literature; Mutschler and colleagues presented a relevant case report in 2010. Olives and colleagues’ 2019 study described a long history of repeated intoxications with signs of tolerance. 1, 25, 26

Such reports show that complications are possible, but do not allow their prevalence or typical time to development to be calculated. Tolerance to a subjective effect does not mean equivalent tolerance to all toxic effects. The FAQ’s suggestions about a specific “excitotoxic rebound” mechanism, a universal “50-trip limit” and self-directed withdrawal regimens must not be presented as established patterns. 1, 25, 26

16. Naloxone

The antidote wording in §6.10 should be narrowed: naloxone is not a specific, reliable means of reversing DXM intoxication. The toxicology consensus by Chyka and colleagues allows its use by medical professionals for marked sedation or coma, particularly with respiratory depression; the recommendation is Grade C. This does not mean demonstrated reversal of all NMDA-dependent and serotonergic effects. 1, 27

17. Nuedexta and Auvelity

Nuedexta, approved by the FDA in 2010 for pseudobulbar affect, combines DXM with quinidine. Quinidine inhibits CYP2D6 and increases DXM exposure. It is also an antiarrhythmic medicine with its own pharmacology: the label warns about QT prolongation and important contraindications. It must not be considered a neutral excipient. 28

Auvelity combines DXM and bupropion. Bupropion inhibits CYP2D6 and has its own antidepressant pharmacology. The combination’s effect therefore cannot be attributed entirely to DXM or solely to a change in the DXM/DXO ratio. The FDA initially approved the medicine for major depressive disorder in 2022 and, on 30 April 2026, for agitation associated with dementia due to Alzheimer’s disease in adults. This is the first approval of a non-antipsychotic medicine for this agitation, rather than approval to treat Alzheimer’s disease itself or its progression. 29, 30

Agitation study data are heterogeneous. Below are results reported by the developer in an SEC document; the study type and endpoint must be retained alongside the result. 31

StudyDesign or measureResult and its limits
ADVANCE-1Placebo-controlled study of agitation symptomsPrimary endpoint met; one of the two studies named by the FDA as the basis for efficacy.
ADVANCE-2408 participants; CMAI change over 5 weeksPrimary endpoint not met: a 13.8-point reduction versus 12.6 on placebo was not statistically significant.
ACCORD-2Randomized withdrawal after sustained response; 167 randomized participantsLower relapse risk: HR 0.276, p = 0.001. The selected responder population differs from the population starting treatment.

Approval of the combination confirms the clinical applicability of a particular medicine for particular indications. It does not demonstrate DXO’s inactivity, the efficacy of DXM alone at the same exposure, or a benefit of recreational use for depression. The FAQ’s claim that dissociatives are inherently poor antidepressants should be replaced with a description of specific clinical data. 1, 29, 30, 31

The FDA describes DXM as a noncompetitive NMDA antagonist and σ1 agonist while noting that the mechanism of depression and agitation treatment is unclear. Listing the targets does not prove which is necessary for the clinical effect. Nuedexta and Auvelity are different medicines with their own indications, pharmacology and evidence bases. 28, 29

18. Summary audit of FAQ claims

The status in the table denotes an editorial decision. “Keep” means the main proposition withstood checking; “Update” means add limitations or contemporary context; “Correct” means replace an incorrect claim; “Hypothesis” means retain only as an explicitly unproven explanation. These statuses do not denote evidence levels. Evidence design and limitations are described in the corresponding sections above.

Quotations are verbatim in the original language, with line breaks merged. Short fragments identify the claim being checked and do not replace the entire section context. The author’s caveats are preserved. All quotations refer to version 4.0 1; the other references in the last column support the contemporary wording.

FAQ sectionOriginal text fragmentStatusWording for an updated versionSources
§5.15“some of the effects of DXM are due to the DXM itself”; “some are due to its metabolite dextrorphan”KeepDXM and DXO are both active; their profiles differ. White already acknowledged this.6, 7, 8
§9.5“DXM is converted to DXO by a liver enzyme called cytochrome P450-2D6”KeepCYP2D6 substantially affects exposure and effects.3, 4, 6, 7
§9.5“will experience less effect from DXM (or more specifically, won’t experience the effects of DXO)”CorrectPM do not necessarily experience weak effects; greater DXM exposure and different tolerability are possible.4, 6
§9.5“have multiple copies of the P450 gene”UpdateUse contemporary PM/IM/NM/UM, activity score and phenoconversion; older EM do not always equal contemporary NM.3, 5
§9.5“the second goes from DXO to 3-methoxymorphinan (3MM) and then to 3HM”CorrectIn this sentence, the route to 3-methoxymorphinan should start from DXM.1, 2, 3
§9.5“DXM is metabolized via two pathways”UpdateSupplement the principal demethylation pathways with glucuronidation and the distinction between unconjugated, total and unbound DXO.2, 3
§9.2.2“The PCP2 binding site is probably the dopamine reuptake complex”HypothesisPCP₂/DAT and the explanation of the first plateau are not established mechanisms.8, 9
§9.2.1“The second receptor is almost certainly the PCP2 receptor”UpdateSome historical DM₂ observations fit serotonin transport; complete identity of all PCP₂ sites has not been demonstrated.9
§10.2“We don’t know if they’re on the outside or inside of cells.”Updateσ1 has been molecularly identified; σ2 corresponds to TMEM97.10, 11
§10.2.1.2“some of the bizarre effects of DXM on motion may be related to sigma activity”HypothesisAn association between σ1 and individual subjective experiences has not been established.12, 29
§10.2.3“Sigma3 receptors are a new discovery”HypothesisRetain σ3 only as a historical designation, without asserting an established third molecular target of DXM.1, 10, 11
§10.7“As I write this, nobody has managed to positively identify an endopsychosin”HypothesisDiscuss endopsychosin as a historical hypothesis, not confirmed by the cited sources as a mechanism of DXM’s action.1
§9.2.4“The NMDA receptor is the central site of dissociative action”KeepNMDA antagonism is important to dissociative action; the precise contributions of targets are unknown.8, 13, 14, 29
§9.3.1“some people have asserted that there are plateaus in between the first three”UpdatePlateaus are descriptive categories with variable boundaries.1, 3, 6, 7
§9.3.1“the first plateau may correspond to predominantly PCP2 activity”; “it is a qualitative drawing, not a quantitative one”HypothesisThe receptor ladder and drawn receptor occupancy are not measured facts; White already qualified the diagram as qualitative.1, 8
§10.6“The resemblance is still fairly limited.”UpdateA direct controlled comparison of DXM with ketamine exists within abuse-potential assessment; it does not cover the whole phenomenology or establish complete equivalence of effects.14
§6.2.9“Category: Theoretical (especially DXM + antidepressants)”CorrectSerotonin syndrome is documented; remove the Theoretical status.15, 16, 29
§6.2.9“combining DXM with a MAOI has been repeatedly fatal!”KeepThe MAOI warning was already correct; add contemporary drug examples and the limits of specific labels.1, 16, 29
§9.5.1“Many drugs inhibit P450-2D6, notably including fluoxetine (ProzacTM).”UpdateFluoxetine and paroxetine cause serotonergic and pharmacokinetic interactions.15, 27, 29
§6.1.3“Diphenhydramine (BenadrylTM) is a good OTC antihistamine that is probably safe to combine with DXM”CorrectDo not consider diphenhydramine an inherently safe addition.17, 27
§6.3.15“bromism can be detected by increased anion gap”CorrectBromism can involve pseudohyperchloremia and a low, including negative, anion gap.20
§6.8.3“check anion gap, which amounts to the same thing”CorrectAnion gap does not replace bromide measurement.20
§6.10“IV naloxone is considered the antidote for DXM overdose”CorrectNaloxone is not a specific antidote; possible medical use is limited by the clinical context.27
§6.7“Once tolerance has built, withdrawal has the potential to cause more serious problems.”UpdateDependence and withdrawal have been described; universal time courses and the proposed mechanisms have not been validated.25, 26
§7.6“He suggested a limit of about 50 trips”; “the exact number seems to vary from person to person”UpdateThe 50-trip limit is a user observation, not an established universal limit.1, 25, 26
§6.3.1“I have received a few anecdotal reports of DXM-induced degradation of mental performance that are consistent with this type of damage.”UpdateDistinguish animal injury, human MRI associations and unproven human Olney lesions; separately account for the author’s withdrawal of the 1998 article’s conclusions.21, 22, 23, 24
§9.5“within 30 minutes, all of it may have entered the bloodstream”; “polistirex compound is intended for continuous absorption”UpdateDistinguish immediate and extended release; do not treat 30 minutes as a universal time to complete absorption or automatically transfer bromide load between formulations.19, 27, 29
§6.1.3“one should not use DXM+antihistamine tablets above the second plateau”UpdateThe risks of combination products depend on the full composition, including antihistamines and paracetamol; plateau number does not define a combination’s safety boundary.15, 27
§6.8“dissociatives make poor antidepressants, since they have numerous side effects.”CorrectAssess antidepressant applicability using specific medicines and trials.29, 30, 31

19. Final working model

The following model is sufficient for an updated FAQ:

  • DXM and DXO concentrations change after administration; CYP2D6 affects DXO formation, while CYP3A and UGT participate in further pathways.

  • The plasma DXM/DXO ratio is meaningful only alongside sampling time, formulation and the method for measuring conjugates and protein binding.

  • DXM/DXO act on multiple targets. Affinity in a single assay does not determine the sequence of subjective effects or plateau dose boundaries.

  • Drug interactions can simultaneously change exposure and amplify toxicity through the accompanying substance’s own pharmacology.

  • User accounts help describe experiences. Causal mechanisms and complication rates require other research methods.

This model preserves the observational part of the FAQ and allows explanations to be updated as data emerge, without presenting a receptor hypothesis as an established cause of an experience.

20. Brief glossary

TermMeaning in this document
DXMDextromethorphan, the parent substance.
DXODextrorphan, an active DXM metabolite.
CYP2D6 and CYP3AOxidative metabolism enzymes; CYP3A designates a family.
UGTUDP-glucuronosyltransferases, conjugation enzymes.
AUCArea under the concentration–time curve, a measure of exposure.
Cmax and TmaxMaximum concentration and the time to reach it.
KiInhibition constant in a specified assay; here, a measure of competitive binding.
SERT and NETSerotonin and norepinephrine transporters.
NMDAA subtype of glutamate receptor associated with an ion channel.
α3β4 nAChRNicotinic acetylcholine receptor with the specified subunits.
PM, IM, NM, UMCYP2D6 activity categories from low to ultrarapid; not a scale of experience intensity.
PBPKPhysiologically based pharmacokinetic model.
MAOI and SSRIMonoamine oxidase inhibitors and selective serotonin reuptake inhibitors.
Anion gapCalculated anion gap, dependent on electrolyte measurements.
NANSpecific NMDA antagonist neurotoxicity in particular experimental models.
CMAI and HRCohen–Mansfield Agitation Inventory; hazard ratio in time-to-event analysis.

21. Sources

Journal articles list author, year, title and DOI or PMID. DOI links lead to the publication; PubMed pages and available full texts allow the design to be clarified. Document versions are given for FDA and drug labeling sources. Web sources were checked as of 5 October 2026. The versions used for mutable labels [19] and [29] are recorded; the current pages’ contents may subsequently change. An ellipsis in a title indicates bibliographic abbreviation of a long title.

[1] White WE. The Dextromethorphan FAQ. Text Version 4.0. 1997. Text version dxmfaq40.txt with an introduction dated 15 March 1997 was checked; quotations from it appear in §18 of this document. FAQ archive: Erowid.

[2] Lutz JD, Isoherranen N. 2012. Prediction of Relative In Vivo Metabolite Exposure from In Vitro Data Using Two Model Drugs: Dextromethorphan and Omeprazole. Drug Metabolism and Disposition. DOI: 10.1124/dmd.111.042200. PMID: 22010218.

[3] Rüdesheim S et al. 2022. Physiologically-based pharmacokinetic modeling of dextromethorphan to investigate interindividual variability within CYP2D6 activity score groups. CPT: Pharmacometrics & Systems Pharmacology. DOI: 10.1002/psp4.12776.

[4] Capon DA et al. 1996. The influence of CYP2D6 polymorphism and quinidine on the disposition and antitussive effect of dextromethorphan in humans. Clinical Pharmacology & Therapeutics. DOI: 10.1016/S0009-9236(96)90056-9. PMID: 8841152.

[5] Caudle KE et al. 2020. Standardizing CYP2D6 Genotype to Phenotype Translation: Consensus Recommendations from the Clinical Pharmacogenetics Implementation Consortium and Dutch Pharmacogenetics Working Group. Clinical and Translational Science. DOI: 10.1111/cts.12692.

[6] Zawertailo LA et al. 1998. Psychotropic effects of dextromethorphan are altered by the CYP2D6 polymorphism: a pilot study. Journal of Clinical Psychopharmacology. DOI: 10.1097/00004714-199808000-00014. PMID: 9690700.

[7] Zawertailo LA et al. 2010. Effect of metabolic blockade on the psychoactive effects of dextromethorphan. Human Psychopharmacology: Clinical and Experimental. DOI: 10.1002/hup.1086. PMID: 20041473. Published online 29 December 2009; journal year: 2010.

[8] Xiao Y et al. 2024. SAL0114: a novel deuterated dextromethorphan-bupropion combination with improved antidepressant efficacy and safety profile. Frontiers in Pharmacology. DOI: 10.3389/fphar.2024.1464564. Table 1; Conflict of interest section.

[9] Meoni P, Tortella FC, Bowery NG. 1997. An autoradiographic study of dextromethorphan high-affinity binding sites in rat brain: sodium-dependency and colocalization with paroxetine. British Journal of Pharmacology. DOI: 10.1038/sj.bjp.0701043. PMID: 9105700.

[10] Schmidt HR et al. 2016. Crystal structure of the human σ1 receptor. Nature. DOI: 10.1038/nature17391. PMID: 27042935.

[11] Alon A et al. 2017. Identification of the gene that codes for the σ2 receptor. Proceedings of the National Academy of Sciences. DOI: 10.1073/pnas.1705154114. PMID: 28559337.

[12] Nguyen L et al. 2014. Involvement of Sigma-1 Receptors in the Antidepressant-like Effects of Dextromethorphan. PLOS ONE. DOI: 10.1371/journal.pone.0089985. Results for BD1063 and BD1047: Results and Discussion sections of the full text.

[13] Carbonaro TM et al. 2018. Double-blind comparison of the two hallucinogens psilocybin and dextromethorphan: similarities and differences in subjective experiences. Psychopharmacology. DOI: 10.1007/s00213-017-4769-4. Full text: PDF.

[14] Shram MJ et al. 2023. The novel uncompetitive NMDA receptor antagonist esmethadone (REL-1017) has no meaningful abuse potential in recreational drug users. Translational Psychiatry. DOI: 10.1038/s41398-023-02473-8. The Ketamine Study section includes DXM as an additional comparator.

[15] Monte AA, Chuang R, Bodmer M. 2010. Dextromethorphan, chlorphenamine and serotonin toxicity: case report and systematic literature review. British Journal of Clinical Pharmacology. DOI: 10.1111/j.1365-2125.2010.03747.x. PMID: 21175434.

[16] Harada T et al. 2025. A fatal intoxication case due to the serotonin syndrome induced by dextromethorphan and moclobemide overdose. Legal Medicine. DOI: 10.1016/j.legalmed.2025.102604. PMID: 40107075.

[17] Hamelin BA et al. 2000. Significant interaction between the nonprescription antihistamine diphenhydramine and the CYP2D6 substrate metoprolol… Clinical Pharmacology & Therapeutics. DOI: 10.1067/mcp.2000.106464.

[18] Forget P et al. 2008. Life-Threatening Dextromethorphan Intoxication Associated with Interaction with Amitriptyline in a Poor CYP2D6 Metabolizer… Journal of Pain and Symptom Management. DOI: 10.1016/j.jpainsymman.2007.09.006. PMID: 18359183.

[19] DailyMed. DELSYM (DEXTROMETHORPHAN), extended-release suspension. Label dated 22 May 2026, version 5. DailyMed.

[20] Ng YY et al. 1992. Spurious Hyperchloremia and Decreased Anion Gap in a Patient with Dextromethorphan Bromide. American Journal of Nephrology. DOI: 10.1159/000168457. PMID: 1481876.

[21] Carliss RD et al. 2007. Oral administration of dextromethorphan does not produce neuronal vacuolation in the rat brain. NeuroToxicology. DOI: 10.1016/j.neuro.2007.03.009. PMID: 17573115.

[22] Tran H et al. 2016. High-dose dextromethorphan produces myelinoid bodies in the hippocampus of rats. Journal of Pharmacological Sciences. DOI: 10.1016/j.jphs.2016.10.001. PMID: 27802908.

[23] White WE. 2004. Response to “The Bad News Isn’t In”: Please Pass the Crow. Erowid. Version dated 28 December 2004. Erowid. The author’s withdrawal of conclusions from the separate 1998 publication.

[24] Qiu YW et al. 2017. Potential gray matter unpruned in adolescents and young adults dependent on dextromethorphan-containing cough syrups… Brain Imaging and Behavior. DOI: 10.1007/s11682-016-9628-0. PMID: 27738991. Online publication: 13 October 2016.

[25] Mutschler J et al. 2010. Dextromethorphan Withdrawal and Dependence Syndrome. Deutsches Ärzteblatt International. DOI: 10.3238/arztebl.2010.0537. PMID: 20737060.

[26] Olives TD et al. 2019. Ten Years of Robotripping: Evidence of Tolerance to Dextromethorphan Hydrobromide in a Long-Term User. Journal of Medical Toxicology. DOI: 10.1007/s13181-019-00706-1. PMID: 30903576.

[27] Chyka PA et al. 2007. Dextromethorphan poisoning: an evidence-based consensus guideline for out-of-hospital management. Clinical Toxicology. DOI: 10.1080/15563650701606443. PMID: 17849242. Naloxone recommendation: Grade C.

[28] FDA. NUEDEXTA prescribing information. Initial approval in 2010; Indications, Warnings and Precautions, and Clinical Pharmacology sections. FDA, PDF. The document is used for approval history and pharmacology, rather than as a complete current prescribing reference.

[29] AUVELITY prescribing information. Edition 08/2026; DailyMed SPL, version 18, effective 17 August 2026, published 19 August 2026. Sections 1, 2.7, 4, 5.8, 7.1, 11, 12.1 and 14. DailyMed. Historical edition 04/2026 corresponding to the expanded indications: FDA, PDF.

[30] FDA. FDA Approves First Non-Antipsychotic Drug to Treat Agitation Associated with Dementia. 30 April 2026. FDA.

[31] Axsome Therapeutics. Announcement of completion of the AXS-05 study program for agitation associated with Alzheimer’s disease. 30 December 2024. Developer document published as an exhibit to an SEC filing. SEC. ADVANCE-2, ACCORD-2 and information on preceding studies.

[32] Di Marco MP et al. 2002. The effect of grapefruit juice and seville orange juice on the pharmacokinetics of dextromethorphan… Life Sciences. DOI: 10.1016/S0024-3205(02)01799-X. PMID: 12095536.

[33] Strauch K et al. 2009. Dose-response relationship for the pharmacokinetic interaction of grapefruit juice with dextromethorphan investigated by human urinary metabolite profiles. Food and Chemical Toxicology. DOI: 10.1016/j.fct.2009.05.004. PMID: 19445995.

Reference

Open the source and related combinations →