Naltrexone

Limits on what this shows

• Why no grade: Both conditions for ungraded remain met: no Tier B or Tier C source directly measures a mitochondrial structure/function/biogenesis endpoint for naltrexone specifically, and no circulating consumer claim on this topic was found on re-check of drugs.com, RxList, and Suzy Cohen's site. The PubMed channel, which does remain reachable, was genuinely re-searched and again returned only off-target (opioid-agonist) results, so the core conclusion is a literature gap rather than a pure tooling failure; however, Europe PMC and eutils are now confirmed blocked by network policy rather than rate limiting, which a later revision with different network access should specifically re-try.

Channels tried

This revision re-ran the category from scratch with a dedicated tool budget. PubMed search-listing pages, query "naltrexone AND mitochondria," re-executed successfully this session (this endpoint remained reachable). Europe PMC's REST API (www.ebi.ac.uk) and NCBI eutils were attempted again and this time the proxy's own diagnostics confirmed the block is an organization-level policy denial at the network gateway (HTTP 403 to the CONNECT request itself, not a site-side rate limit), for both hosts. OpenAlex's works-search API and Semantic Scholar's paper-search API were attempted with several phrasings and deliberate pacing (waits of 30-90 seconds between calls) and returned HTTP 429/503 on every attempt. DuckDuckGo's and Bing's HTML search endpoints were tried as WebSearch-budget workarounds and were blocked (server error / robots.txt disallow) respectively. WebSearch itself remained unavailable all session (200/200 exhausted, confirmed at the start of this revision).

Outcome

The repeated PubMed search returned the same result set as the prior review: opioid AGONIST or endogenous-opioid-peptide cardioprotection literature (delta- and kappa-opioid receptor activation protecting mitochondrial respiratory chain function against ischemia-reperfusion injury), mechanistically the opposite condition (receptor activation, not naltrexone's blockade), not carried forward as naltrexone evidence per the cohort test. No new naltrexone-specific mitochondrial structure/function/biogenesis study surfaced through any of the additional channels attempted this session; those channels were blocked outright rather than returning negative results, so their non-yield is a tooling limitation, not a searched-and-empty finding.

Limits on what this shows

• Null/negative findings: Lin 2005 found naltrexone did not suppress nitric oxide overproduction and did not attenuate iron-induced lipid peroxidation despite reducing superoxide and TNF-alpha, indicating the protective effect is not a uniform, non-specific antioxidant action but pathway-selective.
• At-risk subgroups: None established in humans; all direct evidence is rodent or invertebrate.

Regulatory status

Not addressed in FDA labeling for either formulation.

Population evidence

Lin SL, Lee YM, Chang HY, Cheng YW, Yen MH. J Biomed Sci, 2005. In vivo rat LPS-induced sepsis model, naltrexone 10 mg/kg IV pretreatment before LPS 10 mg/kg IV: naltrexone "significantly decreased the levels of plasma TNF-alpha and inhibited overproduction of superoxide anions in aortic rings," but "did not suppress the overproduction of NO" (nitric oxide) and did not attenuate non-enzymatic iron-induced lipid peroxidation in rat brain homogenates (see Iron, Ungraded, for that sub-finding). Dehe L, Mousa SA, Shaqura M, Shakibaei M, Schafer M, Treskatsch S. 2022. In vivo rat volume-overload model; title (PubMed-indexed, full text access-limited this review): "Naltrexone-Induced Cardiac Function Improvement is Associated With an Attenuated Inflammatory Response and Lipid Peroxidation in Volume Overloaded Rats," indicating reduced inflammation and lipid peroxidation alongside improved cardiac function; quantitative results not independently verified for this dossier. Li W, McIntyre R, Schomakers B, Kamble R, Luesink A, van Weeghel M, Houtkooper R, Gao A, Janssens G. iScience, 2024. In vivo C. elegans; title (bibliographic record confirmed via Crossref, full text access-limited this review): "Low-dose Naltrexone (LDN) extends healthspan and lifespan through activation of the transcription factor SKN-1/NRF2 in C. elegans," indicating naltrexone activates the invertebrate ortholog of the mammalian NRF2 antioxidant-response pathway; quantitative lifespan/healthspan extension figures not independently verified for this dossier. Two additional 2025-2026 animal studies in a chronic obstructive pulmonary disease model (Soni et al., PMID 40383174 and PMID 42635925) report very-low-dose naltrexone modulating redox homeostasis via Nrf2/Keap1/HO-1 and related pathways in rodents; both were identified only at the title/indexed level (individual PubMed abstract pages returned bot-verification challenges on repeated attempts) and are listed in Sources for completeness, not used to support this grade's direction or magnitude beyond noting a consistent thematic signal.

Mechanism

Not fully characterized at the systemic human level. Proposed mechanisms across the animal/invertebrate literature include reduced TNF-alpha-driven and inflammation-driven reactive oxygen species production (Lin 2005, Dehe 2022) and direct activation of the SKN-1/NRF2 antioxidant transcription factor pathway (Li 2024, and thematically the Soni et al. COPD papers). These are mechanistically related but not identical pathways, and no single study establishes both in the same system.

Counter-evidence held on record

All direct evidence is preclinical (rat, and one invertebrate model). This should not be read as established human antioxidant benefit, and does not support any change in antioxidant supplementation, for which no dosing evidence exists in naltrexone-treated patients. See Overlap map: the Lin 2005 source is shared with the Iron ungraded entry's disclosure.

Limits on what this shows

• Why no grade: Both conditions for ungraded remain met: no Tier B or Tier C evidence exists to support even a yellow grade, and no circulating claim specific to naltrexone and copper was found on re-check of the three consumer sources. The prior review's completed PubMed search is treated as the main evidence this gap is genuine rather than purely tooling-limited, since it returned substantive, on-topic-adjacent results rather than an empty or blocked page; this session's five additional blocked channels (disclosed above) mean a broader secondary search could still not be completed, and a later revision with different network access should specifically re-attempt Europe PMC, Crossref, OpenAlex, and Semantic Scholar for this category before treating the gap as fully closed.

Channels tried

Prior review: PubMed search-listing pages, query "naltrexone AND (zinc OR copper)" plus a copper-only follow-up. This revision attempted to extend that with OpenAlex's works-search API and Semantic Scholar's paper-search API (multiple phrasings: "naltrexone copper," "naltrexone ceruloplasmin"), Europe PMC's REST API, Crossref's API, and a repeat PubMed search-listing attempt, all with deliberate pacing (waits of 30-90 seconds between calls, spread over roughly 15 minutes). Every one of these five channels was blocked this session: Europe PMC, Crossref, and NCBI eutils returned an explicit HTTP 403 policy denial at the network proxy's own CONNECT layer (confirmed via the proxy status diagnostics, not merely an application-level error), OpenAlex and Semantic Scholar returned HTTP 429/503, and the repeat PubMed search-listing attempt was itself rejected by the proxy on the same pacing pattern used successfully for the Mitochondrial burden category earlier in this session. DuckDuckGo and Bing HTML search were also tried as WebSearch-budget workarounds and both failed (server error / robots.txt disallow). Consumer sources re-checked directly: drugs.com, RxList, Suzy Cohen's site.

Outcome

No new naltrexone-copper evidence, human or animal, was located this session. This is disclosed as a tooling-limited non-result for the new channels attempted, distinct from the prior review's PubMed search, which did complete and returned clean, substantive but off-target results (reviews mentioning naltrexone and zinc/copper only in unrelated contexts within the same document, not a studied interaction). The mechanism screen finding (no chelation, gastric-pH, or renal-wasting pathway by which naltrexone would plausibly alter copper status) is unchanged and still stands as the main basis for the absence of a yellow-tier mechanistic grade.

Limits on what this shows

• Why no grade: The dual condition for ungraded is met on balance and is unchanged by this session's re-verification: the one directly relevant finding (Lin 2005) is a null result on a narrow, non-classical endpoint (iron-induced oxidative damage in brain tissue, not iron status), which does not itself establish mechanistic plausibility for a yellow grade, and no circulating consumer claim linking naltrexone to iron was found to rebut as white on re-check. This null finding remains disclosed here in full rather than omitted, consistent with Step 5's guidance, and is cross-referenced with the Oxidative stress entry in the Overlap map since both draw on the same source. This session's blocked channels (Europe PMC, Crossref, OpenAlex, Semantic Scholar) mean a broader secondary search for classical iron-status endpoints specifically could still not be completed, and a later revision with different network access should re-attempt them before treating this gap as fully closed.

Channels tried

Prior review: PubMed search-listing pages, query "naltrexone AND iron metabolism," reviewed in full (this is the source of the Lin 2005 null finding below). This revision re-attempted the category via OpenAlex works-search, Semantic Scholar paper-search, Europe PMC REST API, Crossref API, and a repeat PubMed search-listing query ("naltrexone AND (ferritin OR transferrin OR hemoglobin)"), with the same pacing protocol used for Copper and Mitochondrial burden above. All five channels were blocked this session in the same pattern already disclosed for Copper: Europe PMC, Crossref, and eutils gave a confirmed network-policy 403 at the proxy's CONNECT layer; OpenAlex and Semantic Scholar gave HTTP 429/503; the repeat PubMed attempt was proxy-rejected on the same pacing that had briefly worked earlier in this session. Consumer sources re-checked directly (drugs.com, RxList, Suzy Cohen's site) for any circulating iron-related claim.

Outcome

No new naltrexone-iron evidence was located this session; the five extension channels were blocked outright rather than returning new null or positive results. The prior review's finding stands unchanged and is re-verified as the operative evidence: Lin et al. 2005 (the same rat sepsis study cited under Oxidative stress) directly measured that naltrexone did not attenuate non-enzymatic iron-induced lipid peroxidation in rat brain homogenates, a genuine null finding specific to naltrexone and iron-driven oxidative damage, but not a measure of classical iron absorption, transport, or storage status. No study was found, this session or the prior one, examining naltrexone and classical iron absorption, ferritin, transferrin saturation, or hemoglobin-iron status.

Limits on what this shows

• Null/negative findings: Both directly relevant human studies are themselves null/negative findings for naltrexone specifically (no significant microbiota difference in Gicquelais 2020's antagonist subgroup n=6; "limited effects" in Liu 2026's 12-week AUD cohort), and are reported here as the primary evidence rather than as a caveat.
• At-risk subgroups: None established for naltrexone. The Gicquelais 2020 antagonist subgroup (n=6) and combined-therapy subgroup (n=4) are both small; a true effect could be missed at this sample size (underpowered null, disclosed as a limitation rather than definitive reassurance).

Regulatory status

Not addressed in FDA labeling for either formulation.

Population evidence

Gicquelais RE, Bohnert ASB, Thomas L, Foxman B. 2020. Observational cross-sectional study, n=46 outpatient addiction-treatment patients providing stool samples, 16S rRNA V4 sequencing. Participants using opioid agonists only (heroin/prescription opioids, n=5) showed significantly lower microbiota alpha diversity, reduced Bacteroides enterotype, and decreased Roseburia (butyrate-producing) and Bilophila (bile-acid-metabolizing) abundance versus controls (n=31, no opioid exposure in the prior month). No microbiota differences were found between the antagonist-only group (naltrexone, n=6), the combined agonist-antagonist group (buprenorphine-naloxone, n=4), and controls. Liu XL, Meng SC, Hung YJ, Hsu SH, Huang MC, Wu LS. 2026. Human alcohol use disorder cohort followed for 12 weeks during naltrexone treatment; title (PubMed-indexed, full abstract access-limited this review by repeated bot-verification challenges): "Gut Microbiome Dynamics and Alcohol Use Outcomes During Naltrexone Treatment: A 12 Week Follow-up Study," with the search-result snippet indicating "limited effects on overall gut microbiota structure"; quantitative detail not independently verified for this dossier. For contrast (class-adjacent, opposite receptor direction, not naltrexone): Wang F, Meng J, Zhang L, Johnson T, Chen C, Roy S. 2018, PMID 29483538, in vivo murine morphine-dependence model, found a "significant shift in the gut microbiome" with morphine, an opioid AGONIST; cited only to illustrate that the agonist/antagonist directions differ, not as naltrexone evidence.

Mechanism

Opioid-receptor AGONISM, not antagonism, is the pathway implicated in opioid-associated gut dysbiosis in both the human (Gicquelais 2020) and murine (Wang 2018) literature, via effects on gut motility, mucosal barrier integrity, and bile-acid metabolism at mu-opioid receptors. Naltrexone blocks rather than activates these receptors, which is mechanistically consistent with, though not proof of, an absence of the agonist-associated dysbiosis signal.

Counter-evidence held on record

This finding should not be read as definitive proof of no effect; both human studies are small or access-limited, and the antagonist subgroups in Gicquelais 2020 are underpowered. It should also not be confused with alcohol's own, separately well-documented effects on gut barrier integrity and microbiota in alcohol use disorder generally, which arise from alcohol exposure itself and the underlying disease, not from naltrexone (confounding by indication, not assessed as a graded entry in this dossier since it is not a naltrexone-specific mechanism).

Limits on what this shows

• Null/negative findings: The apnea-hypopnea index (a standard periodic-breathing severity measure) did not differ significantly between naltrexone and placebo (P=0.134) despite the significant SpO2 and AMS findings, indicating the effect is not fully explained by a change in discrete apnea/hypopnea events alone.
• At-risk subgroups: Healthy adults acutely ascending to high altitude (~3,800 m or higher) while taking standard-dose oral naltrexone. Not established for Vivitrol's more gradual, sustained-release plasma profile, for chronic altitude residence, or for lower altitudes.

Regulatory status

Not addressed in FDA labeling for either the oral tablet or Vivitrol; high-altitude periodic breathing is not a labeled indication, warning, or studied population for naltrexone.

Population evidence

Foster K, Anholm JD, Foster G, Thapamagar S, Subedi P. High Altitude Medicine & Biology, 2025 (published online 2024). Double-blind, placebo-controlled crossover, n=9 healthy volunteers (4 female, 5 male, mean age 27.9, BMI <27), baseline screening at ~370 m, two overnight trips to 3,810 m (Barcroft Laboratory, California) at least 2 weeks apart, 50 mg naltrexone or placebo at bedtime. Mean overnight SpO2: 83.1% placebo vs 81.2% naltrexone (P=0.040). Minimum overnight SpO2: 74.5% placebo vs 69.9% naltrexone (P=0.020). Groningen sleep-quality score: 7.3 vs 11.2 (P=0.033, higher is worse). Stanford Sleepiness Scale: 2.3 vs 3.6 (P=0.038). Lake Louise (acute mountain sickness) score: 2 vs 5 (P=0.025). Seven of nine participants reported AMS on naltrexone versus two of the same nine on placebo (mild symptoms only). Apnea-hypopnea index was numerically higher on naltrexone (63.1 vs 55.4) but did not reach significance (P=0.134).

Mechanism

Not fully established. The investigators hypothesized that blocking mu-opioid receptors would increase ventilatory drive and improve oxygenation at altitude; the result was the reverse, and the authors state the mechanism "remain[s] unclear," proposing further investigation into mu-receptor effects on cerebral physiology and ventilatory response as next steps.

Counter-evidence held on record

This is not a nutrient-deficiency finding and no supplement or dietary change addresses it. It is a direct pharmacologic effect on nocturnal oxygenation and mountain-sickness risk specific to acute high-altitude ascent, observed with the oral tablet at a standard maintenance dose. Patients on naltrexone (particularly oral) who are planning travel to high altitude should be informed of this finding and may wish to discuss timing or monitoring with their prescriber; this single small trial (n=9) has not been replicated and its relevance to Vivitrol's different pharmacokinetic profile is unknown.

Limits on what this shows

• Why no grade: Both conditions for ungraded remain met: no Tier B or Tier C source directly measuring ATP production for naltrexone exists, now including the one previously unverified candidate, and no circulating consumer claim specific to naltrexone and energy/ATP was found on re-check. Unlike the prior review, this is no longer a mixed tooling-and-evidence gap for the Li 2024 source specifically, that source has now been fully read and definitively does not bear on ATP; the remaining gap (no naltrexone-specific ATP study exists anywhere in the channels reachable this session) is a genuine literature gap, though OpenAlex's and Semantic Scholar's search endpoints remained blocked this session (see Sources access note) and were not exhaustively searched beyond the phrasings above.

Channels tried

This revision re-ran the category from scratch. PubMed search-listing pages, queries "naltrexone AND mitochondria" and "naltrexone AND ATP production." OpenAlex works-search API (multiple phrasings: "naltrexone ATP mitochondrial," "naltrexone copper," repeated with pacing waits up to 90 seconds between attempts) and Semantic Scholar's paper-search API were both attempted and both returned HTTP 429 or 503 on every try. Most importantly, the one candidate flagged in the prior review, Li et al. 2024 iScience (the C. elegans SKN-1/NRF2 paper), was specifically re-attempted via Unpaywall's API (by DOI), which this time succeeded and surfaced a PMC open-access full-text copy (PMC11126937) that the earlier review had not located. Consumer sources re-checked for any circulating "naltrexone and energy/ATP" claim: drugs.com, RxList, Suzy Cohen's site.

Outcome

The full text of Li et al. 2024, now read via the PMC copy, contains no mention anywhere of ATP, ATP production, ATP levels, oxidative phosphorylation, mitochondrial respiration, the electron transport chain, or mitochondrial membrane potential; its findings are confined to SKN-1/NRF2-mediated oxidative-stress and innate-immune gene expression and lifespan/healthspan endpoints, not a direct energy-production assay. This resolves the prior review's disclosed uncertainty about this specific source into a confirmed null: the one candidate that might have contained an ATP-specific naltrexone finding does not. No other naltrexone-specific ATP or oxidative-phosphorylation study was identified; the repeat PubMed mitochondria/ATP searches returned the same opioid-agonist and endogenous-opioid-peptide cardioprotection literature as before (receptor activation, not naltrexone's blockade), not carried forward as naltrexone evidence per the cohort test.