Dronabinol

Limits on what this shows

• Null findings: None found, but none were sought in people: no human biopsy, blood-cell respiration or lactate study of dronabinol users was located. The labels report no mitochondrial toxicity. A StatPearls chapter on cannabinoids discusses no mitochondrial effects [6].
• At-risk subgroups: None established. Lung data come from a cell line and the neuronal data from mice.

Direction

Functional inhibition of mitochondria in the laboratory. Not a depletion or an excess of any nutrient.

Regulatory status

Not mentioned in the Marinol or Syndros labels [1][2]. The label mechanism section describes complex CNS effects and central sympathomimetic activity only [1].

Population evidence

None in people. Tier B studies using delta-9-THC (cohort test: delta-9-THC is the molecule dronabinol consists of; the studies did not use Marinol or Syndros as far as the retrieved text shows). Sarafian 2003 [8]: A549 human lung epithelial cells; concentration-dependent ATP loss with IC50 7.5 ug/mL at 24 hours; diminished mitochondrial membrane potential (JC-1) at concentrations as low as 0.5 ug/mL within 1 hour; effect resembled the uncoupler FCCP and was suppressed by cyclosporin A (permeability transition pore); read from a reconstructed abstract. Benard 2012 [9]: per the journal abstract as returned by a summarizing fetch tool, exogenous cannabinoids (THC, WIN55,212-2, HU-210 reported) and endocannabinoids acting on mitochondrial CB1 decreased cAMP, PKA activity, complex I activity and respiration in mouse neuronal mitochondria. Hebert-Chatelain 2016 [10]: mouse neurons and mitochondria; THC and WIN at 800 nM to 1 uM reduced respiration through an intra-mitochondrial Galpha-i, soluble adenylyl cyclase, PKA, complex I pathway, and genetic removal of mitochondrial CB1 prevented the memory impairment (summarizing fetch tool).

Mechanism

Named and receptor-dependent in neurons (mitochondrial CB1 signaling to complex I) [9][10]; receptor-independent uncoupling and permeability-transition involvement suggested in lung cells [8].

Counter-evidence held on record

Counter-evidence and confounding The concentrations (0.5 ug/mL and above; 800 nM to 1 uM) are not compared with plasma levels in treated patients because the retrieved label sections give no plasma concentration figures; relevance to therapeutic dosing is unknown. Sarafian also tested marijuana smoke extract, which is not dronabinol and was not used here. The neuronal findings are described by their authors in the context of memory and synaptic function, not toxicity. Framing caution Laboratory mechanism only. This entry is not a claim that dronabinol damages human mitochondria or that any antioxidant or cofactor supplement is needed. Do not recommend supplements from this entry.

Limits on what this shows

• Null findings: No measurement of malondialdehyde, glutathione, or antioxidant enzymes in dronabinol-treated people or animals was located; a Crossref query for THC with lipid peroxidation and glutathione returned no relevant records. The uncoupling reported in Sarafian [8] could in theory raise reactive oxygen species, but ROS were not measured there as far as the reconstructed abstract shows.
• At-risk subgroups: None established.

Direction

Reverse (protective antioxidant chemistry in vitro). Not a depletion.

Regulatory status

Not mentioned in the labels [1][2].

Population evidence

Tier B only. Hampson 1998 [11]: rat cortical neuron cultures exposed to glutamate; cannabidiol and delta-9-THC gave equivalent protection across NMDA, AMPA and kainate receptor pathways, independent of cannabinoid receptors, and both were antioxidants on electrochemical analysis; cannabidiol was more effective than ascorbate or alpha-tocopherol in preventing oxidative damage. Read from a reconstructed abstract; the PMC full text returned a bot-verification page.

Mechanism

Present in vitro: electrochemical analysis showed antioxidant behavior [11]; the specific chemistry was not seen in the abstract.

Counter-evidence held on record

Counter-evidence and confounding The antioxidant result is a test-tube and neuron-culture finding with no in vivo marker; the sign could differ in tissue where mitochondrial inhibition [8][9][10] occurs. The two yellow entries point in opposite directions and neither is clinical. Framing caution Not a finding that dronabinol is an antioxidant supplement or that it protects patients. Do not recommend antioxidant supplements, or advise against them, from this entry.

Limits on what this shows

• Null findings: [25] above. Drugs.com consumer pages list no iron or anemia effect [20][21]. A consumer page on cannabis and iron deficiency that was opened states that cannabis has not been shown to cause iron deficiency and cites no studies (Tier D, not evidence) [27]. A PubMed listing for cannabis with ferritin returned three records, none about dronabinol or THC pharmacology; one 1994 pregnancy study of illicit-drug serum levels and nutritional status (folate, B12, ferritin, ascorbate) was located, but its abstract as retrieved reports lower folate and does not state a ferritin result, so it is not used [28][29].
• At-risk subgroups: None established. Anemia in HIV, cancer and chemotherapy is common and indication-linked (confounding by indication is presumed).

Direction

Neutral to reverse (no reduction in serum iron). Not a depletion and not an excess-retention signal. No acid-suppression, chelation or mucosal-injury mechanism applies to dronabinol.

Regulatory status

The Marinol and Syndros labels contain no iron, anemia or hematologic statement [1][2].

Population evidence

CLASS EXTRAPOLATION, Tier C. NHANES 2009 to 2018 [25]: 16,038 US adults aged 18 to 59, self-reported smoked marijuana. Baseline characteristics: median serum iron 84 ug/dL (IQR 63 to 110) in current users versus 80 ug/dL (IQR 60 to 106) in never users, with no p-value reported for that comparison in the paper; current users had lower red-cell folate; hemoglobin was 0.111 g/dL higher in current users (95% CI 0.021 to 0.201), and sensitivity analyses adjusting for serum iron and red-cell folate left the hemoglobin result robust. Cohort test: pharmaceutical THC (dronabinol) was not specifically included, so THIS drug was not in the cohort and no THC-identical data exist for iron. Read through fetch summaries of the journal page, not verified full text.

Mechanism

Absent for dronabinol in the retrieved material. A mouse and cell literature on cannabinoid receptors and iron homeostasis appeared in search results (CB2 and hepcidin; cannabidiol and iron-induced mitochondrial injury) but concerned other cannabinoids and iron overload, was not opened, and is not used.

Counter-evidence held on record

Counter-evidence and confounding The serum iron figures come from a descriptive baseline table in a survey of smoked cannabis users, are unadjusted, and carry no reported test of significance; smoking itself raises hemoglobin. Users of recreational marijuana differ from AIDS and cancer patients in age, comorbidity and diet. No ferritin, transferrin saturation or hepcidin measurement in a dronabinol user was located. Framing caution Not a finding that dronabinol protects iron status or is safe for it. Context flag only; no iron testing or iron supplement is suggested by this entry.

Limits on what this shows

• Why no grade: Dual condition met: no Tier B or C source was opened to support even a yellow grade (the cannabis NHANES survey [25] reports no copper measure), and no circulating copper-depletion claim for dronabinol or THC was found to rebut. This is a disclosed gap, not a finding of safety. Re-check when search access returns, using ceruloplasmin and serum copper terms with dronabinol, Marinol and delta-9-THC.

What was searched

Original pass: Marinol and Syndros label sections (no copper or mineral mention) [1][2]; StatPearls dronabinol and cannabinoids chapters (no mention) [5][6]; Crossref bibliographic query "tetrahydrocannabinol copper serum ceruloplasmin trace element" (8 results, none about THC, cannabis or dronabinol); Drugs.com pages [20][21]. Repair pass: PubMed listing for (cannabis OR cannabinoid OR tetrahydrocannabinol OR dronabinol) AND (copper OR ceruloplasmin) returned 10 records, none measuring copper or ceruloplasmin in users (they concerned copper and cannabis plant production, heavy metals in cannabis vapes, THC electrochemistry and copper toxicity in fish) [26]; a web search for delta-9-THC with copper, zinc, serum levels and trace elements returned only general THC reviews, a plant-contaminant news item and pharmacology pages, none opened as relevant; a PubMed listing for tetrahydrocannabinol with zinc and serum returned 429.

Channel outcomes

Labels and two reference chapters opened: clean zero. PubMed copper listing opened: clean zero for human biomarker studies. Heavy-metal contamination of cannabis vape products is a product-quality topic, not an effect of THC on copper status, and does not involve dronabinol. Remaining limits: the zinc and mitochondrial-copper PubMed listings were rate-limited, and web search budget ran out, so some tooling limitation remains. No copper-chelation, transporter or ceruloplasmin mechanism for dronabinol appeared in the labels.

Limits on what this shows

• Null findings: Giorgini [15]: no within-arm microbiota change in humans. Gorelick [13]: THC effects weak compared with CBD. No study of dronabinol and C. difficile or fecal microbiota in patients was located.
• At-risk subgroups: None established. People with HIV and cancer patients have disease-related and treatment-related dysbiosis (antiretrovirals, chemotherapy, antibiotics) that is a stronger confounder than the drug.

Direction

Composition change, direction varies by study; no depletion or excess of a nutrient.

Regulatory status

Not mentioned in the labels [1][2]. Abdominal pain (3 to 10%) and nausea or vomiting are listed; diarrhea is 1% or less [1].

Population evidence

Tier B (THC alone, mice). Cluny 2015 [12]: male mice, daily THC 2 mg/kg for 3 weeks then 4 mg/kg for 1 week; THC reduced body weight gain, fat mass and energy intake in diet-induced obese mice and prevented diet-induced changes in select gut microbiota (taxa not captured in the reconstructed abstract). Gorelick 2022 [13]: mice fed a high fat-cholesterol diet for 6 weeks with THC in the diet; THC produced weak responses, only slightly reducing inflammatory-related gene expression and dysbiosis; CBD had the larger effects. Tier B (THC with CBD, mice): Al-Ghezi 2019 [14] used THC plus CBD at 10 mg/kg each by injection in autoimmune encephalomyelitis; Akkermansia muciniphila and Proteobacteria fell and short-chain fatty acids rose; THC alone at 10 mg/kg was ineffective on disease and its microbiome effect was not reported separately (combination cohort, class extrapolation).

Class cohort test

CLASS EXTRAPOLATION for the human data. Giorgini 2025 [15]: randomized pilot in people with HIV on antiretroviral therapy, 10 randomized (5 to CBD plus THC capsules, 5 to CBD alone), 8 completing 12 weeks; 16S rRNA sequencing of stool. Dronabinol was not named; the product was an oral THC/CBD capsule and delta-9-THC was in only one arm and confounded by CBD. Within-arm microbiota were not altered by cannabinoid exposure; between arms the THC/CBD group had lower Coprobacillus and Lachnospiraceae UCG001 and higher Collinsella, against pre-existing baseline differences. Source read through a reconstructed abstract. Capped at yellow.

Mechanism

Not defined for THC alone in the retrieved material. A fecal transplant experiment in the THC plus CBD mouse model suggested microbiome-mediated protection [14]. In the mouse studies the same cannabinoid also changed food intake and body weight, so a diet-mediated effect cannot be separated from a direct one.

Counter-evidence held on record

Counter-evidence and confounding Doses in mice are not scaled to patient doses here. The only human study is a 10-person pilot with arms that differed at baseline. Mouse body-weight change from THC is the opposite of the weight gain sought in patients (see Dietary intake entry), so extrapolation to human patients is weak. Framing caution Not a finding that dronabinol disrupts the human microbiome. No probiotic or prebiotic recommendation follows from this entry.

Limits on what this shows

• Null findings: No human ATP, phosphocreatine or muscle energetics study of dronabinol was located.
• At-risk subgroups: None established.

Direction

Reduction of cellular energy output in vitro. Not a nutrient deficiency.

Regulatory status

Not mentioned in the labels [1][2].

Population evidence

Tier B only. ATP was directly measured by Sarafian 2003 [8] (A549 cells, IC50 7.5 ug/mL at 24 hours). Respiration, a surrogate of ATP synthesis, fell in mouse neuronal mitochondria [9][10]. The fetch summary of [10] says the pathway decreased cellular respiration and ATP production; ATP values were not separately seen.

Mechanism

Complex I inhibition through mitochondrial CB1 [9][10] or uncoupling [8]. Both are shared with the Mitochondrial burden entry (see Overlap map).

Counter-evidence held on record

Counter-evidence and confounding ATP loss in Sarafian occurred at 7.5 ug/mL, with cell death only above 10 ug/mL, in a cancer-derived cell line. In the AIDS and cancer populations the drug is used in, fatigue and cachexia are produced by the underlying disease (confounding by indication). Framing caution Do not present this as "dronabinol lowers your energy" or as a reason to take CoQ10, B vitamins or other energy supplements. It is an unreplicated laboratory signal and shares its evidence with the Mitochondrial burden entry.

Limits on what this shows

• Null findings: [24]: no methemoglobin, no P50 change in males, a small change of doubtful relevance in females. Labels and reviews report no hematologic effect of dronabinol [1][2][5][19]. No hemoglobin, hematocrit or oxygen-saturation study of dronabinol users was located.
• At-risk subgroups: None established. People with HIV or cancer have disease- and chemotherapy-related anemia that is a far stronger determinant of hemoglobin than any laboratory signal here.

Direction

Mixed and unresolved. Oxygen affinity rose in [22] (lower P50 means hemoglobin holds oxygen more tightly) but was essentially unchanged in [24]. Not a nutrient deficiency and not an intended drug effect (Hard Rule 2).

Regulatory status

The Marinol and Syndros labels document no hematologic adverse reaction or laboratory effect [1][2].

Population evidence

None in dronabinol users. Tier B, pure delta-9-THC arms (cohort test: delta-9-THC is the molecule dronabinol consists of; the studies spiked THC into blood and did not use Marinol or Syndros; combination THC plus CBD extract arms in [23] are excluded from this entry). James 2020 [22]: blood from 24 healthy non-smoking adults aged 18 to 40; THC at 0, 2.5, 25, 50 and 100 ng/mL; P50 measured on a Hemox analyzer; P50 decreased with rising THC and the decrease was significant (p below.05) at all tested concentrations, suggesting increased oxygen affinity (P50 values themselves not captured; read from a summarizing fetch of the journal landing page, OpenAlex held no abstract). James 2022 [23]: blood from 24 healthy non-smoking adults aged 21 to 42, same concentrations; THC alone raised whole-blood viscosity (about 6.45 mPa s at 25 ng/mL to about 11.60 mPa s at 100 ng/mL), elasticity rose significantly only in combination extracts, and red-cell morphology was damaged at higher concentrations (OpenAlex reconstructed abstract plus fetch summary). Frisch 2024 [24]: blood from 11 healthy volunteers aged 22 to 33; delta-9-THC at 0.05, 0.1 and 0.5 ug/mL; no significant methemoglobin formation under any condition; of all P50 comparisons only one delta-9-THC change was significant (a rise of 1.4 mmHg, in females only, against a normal range of 24 to 28 mmHg), and the authors questioned its clinical relevance (fetch summary of the journal abstract).

Class cohort test

CLASS EXTRAPOLATION, hemoglobin in people. NHANES 2009 to 2018 [25]: cross-sectional, 16,038 US adults aged 18 to 59, self-reported marijuana use (smoked cannabis); dronabinol and pharmaceutical THC were not specifically included, so dronabinol was not in the cohort. Current use was associated with hemoglobin 0.111 g/dL higher than never use (95% CI 0.021 to 0.201); the authors interpret this as subclinical hypoxemia presumably from inhaled combustion products, noting that cigarette smoking showed a larger association and that carboxyhemoglobin was not available. Read through fetch summaries. This mechanism does not apply to oral dronabinol. Capped at yellow.

Mechanism

Not established. Direct interaction of THC with hemoglobin or the red-cell membrane is suggested in vitro [22][23]; no mechanism was named in the retrieved summaries. Methemoglobin formation, the most obvious oxygen-carrying mechanism, was not seen [24].

Counter-evidence held on record

Counter-evidence and confounding The studies add THC to blood in a dish; the in vitro concentrations are not compared with plasma levels in treated patients because no plasma concentration figures were retrieved from the label. Conflicting direction between [22] and [24] means no stable finding exists. Tachycardia, hypotension and syncope on the label are the drug's own hemodynamic effects and are not oxygen-delivery findings. The NHANES association is attributed by its authors to combustion products and is not evidence about an oral capsule or solution. Framing caution Not a finding that dronabinol reduces oxygen delivery or causes anemia. No iron, B12 or other supplement follows from this entry, and none should be inferred from the in vitro viscosity or oxygen-affinity data.