evipedia.ai
← Back

Low-Dose Naltrexone for Health & Longevity

Evidence Review created on 10/24/2026 using AI4L / Opus 5.5

Also known as: Low Dose Naltrexone, LDN, Naltrexone, Naltrexone Hydrochloride, ReVia

Motivation

Low-dose naltrexone (LDN) is naltrexone, a prescription drug approved for alcohol and opioid dependence, taken at roughly one-tenth or less of the approved dose. At these small doses it briefly blocks the body’s opioid receptors and is thought to calm immune cells in the brain and spinal cord. It is inexpensive, taken by mouth once a day, and prescribed outside its approved uses by integrative and longevity-focused physicians.

Interest grew from clinical reports in the 1980s and later from small trials in chronic pain and in autoimmune conditions. Longevity clinics have since added it to their programs, and laboratory work in worms has examined its effect on lifespan, which drew it into discussions about healthy aging. Because the drug is old, inexpensive and no longer patented, most of the research has come from academic groups and practicing clinicians rather than from drug companies.

This review examines the human evidence for and against low-dose naltrexone as a health and longevity intervention: how it acts in the body, which benefits and risks controlled studies support, how it is dosed, combined and monitored, and where the evidence is still missing.

Benefits - Risks - Protocol - Conclusion

Five resources give a high-level overview of low-dose naltrexone from integrative, pain-research, neuroscience, longevity and cancer-research perspectives.

Note: No relevant low-dose naltrexone content was found from Rhonda Patrick (FoundMyFitness) by web search or on-site search; the FoundMyFitness site search did not complete on any retrieval tier.

Grokipedia

  • Low-dose naltrexone

    AI-generated encyclopedia entry summarizing LDN dosing, the proposed opioid growth factor (a natural peptide that slows cell division) and immune-receptor mechanisms, off-label uses and safety; its claims need checking against primary studies.

Examine

No Examine article on low-dose naltrexone exists. Examine.com does not typically cover prescription medications, and naltrexone is a prescription drug.

ConsumerLab

No ConsumerLab article on low-dose naltrexone exists. ConsumerLab does not typically cover prescription medications, and naltrexone is a prescription drug.

Systematic Reviews

Five systematic reviews and meta-analyses (statistical pooling of results from several studies) cover fibromyalgia pain, with opposite conclusions, Crohn’s disease, long COVID (persistent symptoms after COVID-19 infection), and serious adverse events across all oral naltrexone doses.

Mechanism of Action

Naltrexone is a non-selective competitive antagonist (it occupies a receptor without activating it) of opioid receptors, with highest affinity for the mu receptor (main target of morphine and endorphins) (Toljan & Vrooman, 2018). Three explanations compete:

  • Rebound opioid signaling: a 4.5 mg dose blocks receptors for only hours; Penn State researchers argue this intermittent blockade raises endorphins and the opioid growth factor (OGF)–OGF receptor axis, which restrains cell division, whereas continuous blockade does the opposite (McLaughlin & Zagon, 2015).
  • Microglia (brain and spinal immune cells) calming: in rodent and cell studies, naltrexone isomers blocked toll-like receptor 4 (TLR4, an immune danger sensor) on microglia and reversed nerve-injury pain (Hutchinson et al., 2008); Younger’s group proposes this lowers cytokines (immune signaling proteins) in people (Younger et al., 2014).
  • Contrary findings: in mice, LDN did not change β-endorphin release (Metz et al., 2021), and in a depression trial it did not change high-sensitivity C-reactive protein (hsCRP, a blood inflammation marker) (Moloney et al., 2026). Neither mechanism is established in humans.

Pharmacology: first-pass metabolism (breakdown in gut wall and liver) leaves 5–40% bioavailability (fraction reaching the bloodstream); levels peak within an hour; half-life is about 4 hours for naltrexone and 13 hours for its active metabolite 6-β-naltrexol; wide distribution, 21% protein binding; metabolism runs through dihydrodiol dehydrogenase (DD4, also called AKR1C4, the liver enzyme forming 6-β-naltrexol), not the cytochrome P450 (CYP, the main drug-processing enzymes) system, with kidney excretion (US Food and Drug Administration (FDA) label; Porter et al., 2000).

Historical Context & Evolution

Naltrexone was approved by the FDA in 1984 to block the effects of opioids, with alcohol dependence added in 1994, at 50 mg daily (Drugs@FDA). In the same decade, Penn State researchers Ian Zagon and Patricia McLaughlin found in animals that brief and continuous opioid blockade had opposite effects on cell growth (McLaughlin & Zagon, 2015), and physician Bernard Bihari began giving 1.5–3 mg as an add-on treatment for acquired immune deficiency syndrome (AIDS), introducing LDN into clinical practice (Toljan & Vrooman, 2018).

Controlled studies followed from Jill Smith’s Penn State group in Crohn’s disease (Smith et al., 2011) and Jarred Younger’s Stanford group in fibromyalgia (Younger & Mackey, 2009). These small positive trials, the drug’s low cost and patient networks spread its use; after a 2013 Norwegian television documentary, 0.3% of Norway’s population collected a prescription within two years (Raknes & Småbrekke, 2017). Long COVID clinics and longevity clinics adopted it after 2020, and a 2024 worm study added a lifespan rationale (Li et al., 2024).

Opinion then shifted with new evidence: larger independent parallel-group (separate treatment and placebo groups) trials in fibromyalgia and Crohn’s disease did not reproduce the early results (Due Bruun et al., 2024; van de Pol et al., 2026). Supporters note that pooled analyses still show a modest fibromyalgia benefit (Vatvani et al., 2024) and that trials have tested single conditions for 12 weeks. The question remains open rather than settled.

Expected Benefits

High 🟩 🟩 🟩

No benefit reaches High: the positive placebo-controlled pain trials come mainly from one research group, and the largest independent parallel-group trials found no difference from placebo.

Medium 🟩 🟩

Chronic Pain Relief ⚠️ Conflicted

LDN may ease centrally amplified pain by calming spinal immune cells. Pooled fibromyalgia trials show a modest reduction (Vatvani et al., 2024; responder estimate corrected in 2026), echoing small early trials from Younger’s Stanford group (Younger et al., 2013). Parallel-group trials from Bruun’s Danish and Rodríguez-Freire’s Spanish groups found no difference (Due Bruun et al., 2024; Rodríguez-Freire et al., 2026); a chronic pain meta-analysis found no overall effect but a fibromyalgia benefit (Hegde et al., 2025). Net reading: a small fibromyalgia benefit is possible but unconfirmed by the largest trials.

Magnitude: Pooled pain reduction 0.86 points more than placebo on a 0–10 scale (95% CI [confidence interval, the range likely to contain the true effect] −1.20 to −0.51) (Vatvani et al., 2024); in the two largest trials, 0.34 points more than placebo (95% CI −0.95 to 0.27, not significant) (Due Bruun et al., 2024) and an adjusted 0.49 points less than placebo (not significant) (Rodríguez-Freire et al., 2026).

Low 🟩

Crohn’s Disease Activity ⚠️ Conflicted

In Smith’s 40-adult Penn State trial, 4.5 mg improved clinical and bowel-healing response (Smith et al., 2011); Cochrane rated it low certainty (Parker et al., 2018). An independent 41-patient Dutch trial stopped early for lack of benefit (van de Pol et al., 2026). Net reading: benefit is unconfirmed.

Magnitude: A 70-point drop in disease-activity score in 83% versus 38% on placebo (risk ratio [RR, how many times more likely] 2.22, 95% CI 1.14–4.32) (Parker et al., 2018); in the Dutch trial, clinical remission in 22.2% versus 70.0% on placebo (van de Pol et al., 2026).

Fatigue in Long COVID and Inflammatory Disease ⚠️ Conflicted

Four uncontrolled before-after studies report less long COVID fatigue, with no randomized results (Byambasuren et al., 2026). In Crohn’s disease, LDN reduced fatigue versus placebo (van de Pol et al., 2026), but a multiple sclerosis trial found no change in energy (Sharafaddinzadeh et al., 2010). Net reading: plausible but unproven.

Magnitude: Pooled before-after fatigue reduction Hedges’ g −0.74 (a standardized effect size; 0.8 counts as large) (95% CI −1.11 to −0.37), no control group (Byambasuren et al., 2026); in Crohn’s disease, median FACIT-Fatigue (validated fatigue questionnaire) change +2.5 versus −3 on placebo (p=0.012, the probability of such a difference arising by chance) (van de Pol et al., 2026).

Mental Well-Being and Quality of Life ⚠️ Conflicted

In multiple sclerosis, Cree’s San Francisco trial improved mental quality-of-life scores (Cree et al., 2010), but an Iranian trial of 96 patients found no difference (Sharafaddinzadeh et al., 2010). As add-on depression treatment, LDN matched placebo (Moloney et al., 2026). Net reading: no consistent effect.

Magnitude: 3.3 points more than placebo on the SF-36 Mental Component Summary (a validated quality-of-life scale) (Cree et al., 2010); Montgomery–Åsberg depression scores fell 10.5 versus 9.8 points on placebo, with no between-group difference (Moloney et al., 2026).

Lower Use of Rheumatic Medications

In Norway’s national prescription database, among 360 people with rheumatoid or seropositive (antibody-positive) arthritis starting LDN, those who kept taking it collected fewer analgesic, anti-inflammatory and disease-modifying drugs (drugs that slow joint damage) the following year (Raknes & Småbrekke, 2019). Prescriptions are only proxies for clinical effect.

Magnitude: 73.3 fewer defined daily doses (standard dose units) per patient per year (95% CI −120.2 to −26.4), a 13% relative reduction among persistent users, no control group (Raknes & Småbrekke, 2019).

Inflammatory Skin Disorders

A systematic review, mainly of LDN case reports and series, found it effective in Hailey-Hailey disease (inherited blistering skin disorder) and lichen planopilaris (inflammatory scarring hair loss) (Ekelem et al., 2019); a 2025 review covered related scarring hair loss (Bai et al., 2025). No controlled trials exist.

Magnitude: Not quantified in available studies. Only case reports and small uncontrolled series exist, without pooled outcome figures.

Cancer Outcomes

A systematic review found case reports of prolonged survival with 3–5 mg daily, plus animal and cell studies in which intermittent blockade slowed tumor growth (Liubchenko et al., 2021). A 110-patient glioma (brain tumor) trial found no quality-of-life benefit (Peters et al., 2022).

Magnitude: Not quantified in available studies. Survival evidence consists only of case reports without comparison groups.

Speculative 🟨

Lifespan and Healthspan Extension

In Caenorhabditis elegans (a roundworm), a low but not a high naltrexone dose extended lifespan and healthspan via SKN-1 (worm counterpart of the antioxidant regulator NRF2) (Li et al., 2024). Animal data only.

Systemic Inflammation Reduction ⚠️ Conflicted

In an eight-woman fibromyalgia study without a placebo group, LDN lowered many plasma cytokines (Parkitny & Younger, 2017); a randomized depression trial found no hsCRP change (Moloney et al., 2026). Net reading: unestablished.

Benefit-Modifying Factors

  • Genetic variants: The OPRM1 (mu-opioid receptor gene) variant rs1799971 modified full-dose naltrexone response in some alcohol studies but not in genotype-stratified trials (Counts et al., 2025); no LDN-specific pharmacogenetic data exist.
  • Baseline inflammation: In the first fibromyalgia pilot, a higher erythrocyte sedimentation rate (ESR, a blood marker of inflammation) predicted larger responses (Younger & Mackey, 2009); in depression, baseline hsCRP did not modify response (Moloney et al., 2026).
  • Sex: Most fibromyalgia trials enrolled only women, and 74% of Norwegian users were female (Raknes & Småbrekke, 2017), so effects in men are largely untested; Crohn’s disease and multiple sclerosis trials enrolled both sexes without reporting sex-specific results.
  • Type of pain: Benefit signals concentrate in centrally amplified and inflammatory conditions; a crossover (each participant gets both treatments) trial in mainly nociceptive (tissue-damage) arthritis pain found no effect (Beaudette-Zlatanova et al., 2023).
  • Expectations: In the 12-month INNOVA trial, exploratory analyses suggested some improvement among participants who believed they had received LDN, although effects were inconsistent (Rodríguez-Freire et al., 2026).
  • Pre-existing conditions: Signals come from fibromyalgia, inflammatory bowel disease, multiple sclerosis and long COVID; in healthy adults without a target condition, no human benefit has been measured.
  • Age: The FINAL trial enrolled women aged 18–64 (Due Bruun et al., 2024), and no trial reports age-stratified efficacy, leaving benefit in adults over 65 untested.

Potential Risks & Side Effects

High 🟥 🟥 🟥

Vivid Dreams and Sleep Disturbance ⚠️ Conflicted

Vivid dreams, and less often insomnia, are the most consistent LDN effect, usually mild and fading within weeks; the mechanism is unknown. A meta-analysis of five fibromyalgia trials from groups in Denmark (Bruun), the USA (Younger) and Egypt found a clear excess (Nazir et al., 2025), as did a second pooled analysis (Vatvani et al., 2024). The Cochrane Crohn’s review, with 46 participants, found no significant difference (Parker et al., 2018). Net reading: the larger fibromyalgia data show a real but mild, reversible excess.

Magnitude: Risk ratio for vivid dreams 2.41 versus placebo (95% CI 1.77–3.28) (Nazir et al., 2025); the source abstract does not report the rate in each group.

Medium 🟥 🟥

Blocked Opioid Analgesia and Precipitated Withdrawal

Naltrexone competitively blocks opioid receptors, so opioid pain medicines work poorly; large opioid doses taken to overcome the blockade can cause fatal overdose, and in anyone physically dependent on opioids it can trigger withdrawal severe enough for hospital or intensive care; after stopping, formerly dependent people are more sensitive to opioids and at overdose risk (FDA label). This is established at 50 mg; at LDN doses blockade is shorter, a dose gap that caps the grade. Among Norwegian LDN users, opioid consumption fell, consistent with avoidance (Raknes & Småbrekke, 2017).

Magnitude: Withdrawal begins within five minutes and lasts up to 48 hours after 50 mg in opioid-dependent people (FDA label); the literature reports no outcome figure at LDN doses. Persistent Norwegian LDN users used 46% less opioid than the prior year, no control group (Raknes & Småbrekke, 2017).

Nausea, Headache and Other Mild Adverse Events ⚠️ Conflicted

Nausea, headache, dizziness and mild diarrhea are the usual non-sleep complaints, typically transient. A meta-analysis of chronic pain trials found more adverse events overall on LDN than placebo (Hegde et al., 2025), and pooled fibromyalgia data showed more nausea but not more headache (Vatvani et al., 2024). The two largest fibromyalgia trials, from Bruun’s and Rodríguez-Freire’s groups, found similar rates on LDN and placebo (Due Bruun et al., 2024; Rodríguez-Freire et al., 2026). Net reading: any excess is small and mild.

Magnitude: Incidence rate ratio (IRR, the ratio of event rates) 1.4 versus placebo (95% CI 1.12–1.75) (Hegde et al., 2025); any adverse event in 84% versus 86% on placebo (Due Bruun et al., 2024) and 68.8% versus 72% on placebo (Rodríguez-Freire et al., 2026).

Low 🟥

Liver Injury

Naltrexone raised transaminases (liver enzymes) in a dose-related way, significantly only at 300 mg daily and more often from age 40 (Pfohl et al., 1986); the label keeps a liver-toxicity warning (FDA label). LDN doses are about 1/65 of 300 mg, an indirect dose gap.

Magnitude: Elevations differed from placebo only at 300 mg daily, with none at the 50 mg dose (Pfohl et al., 1986); the literature reports no outcome figure at LDN doses.

Depression and Suicidal Thoughts ⚠️ Conflicted

The label lists postmarketing (after-approval) reports of depression and suicidality with 50 mg naltrexone in opioid dependence, without demonstrated causation, a dose gap from LDN (FDA label). In an LDN depression trial, scores did not differ from placebo (Moloney et al., 2026). Net reading: no excess shown.

Magnitude: New-onset depression in 0–15% on naltrexone versus 0–17% on placebo in alcohol-dependence studies (FDA label); depression scores fell 10.5 versus 9.8 points on placebo (Moloney et al., 2026).

Opioid Hypersensitivity During Use

A case report described a man taking LDN for multiple sclerosis who became unresponsive after a single 5 mg oxycodone dose, needing naloxone (an opioid-reversal drug) and one night of intensive care (Leonard et al., 2017). Proposed mechanism: intermittent blockade upregulates opioid receptors. Evidence is one uncontrolled report.

Magnitude: Not quantified in available studies. Only a single case report exists, without a comparison group (Leonard et al., 2017).

Speculative 🟨

Harm in Pregnancy

Naltrexone increased early fetal loss in rats and rabbits at 5–18 times human doses, and no adequate human studies exist (FDA label). The basis is animal data only.

Thyroid Hormone Over-Replacement

Practitioners report that LDN may lower thyroid hormone needs in autoimmune thyroiditis (Hashimoto’s disease), risking over-replacement; national prescription data showed no dispensing change (Raknes & Småbrekke, 2020). The basis is isolated reports.

Cell Growth with Continuous Blockade

In cell and animal studies, continuous opioid-receptor blockade enhanced cell proliferation while intermittent blockade suppressed it (McLaughlin & Zagon, 2015); slow-release LDN could shift toward continuous blockade. The basis is mechanistic only.

Risk-Modifying Factors

  • Genetic variants: Variable activity of dihydrodiol dehydrogenase (DD4/AKR1C4) alters 6-β-naltrexol formation (Porter et al., 2000); adolescents carrying reduced-activity AKR1C4 variants had 3.2-fold higher naltrexone exposure (Stancil et al., 2022); consequences at low doses are unknown.
  • Baseline liver tests: When transaminases rose on naltrexone, pre-existing alcoholic liver disease, hepatitis B or C, or other liver-toxic drugs were often contributing (FDA label), so abnormal baseline ALT or AST (alanine and aspartate aminotransferase, liver enzymes) marks higher liver risk.
  • Liver and kidney function: Naltrexone exposure rises about 5-fold in compensated and 10-fold in decompensated (failing) cirrhosis (liver scarring), and the label advises caution in kidney impairment because drug and metabolite are excreted in urine (FDA label).
  • Sex: Pregnancy is the main sex-specific concern, based on animal fetal-loss data (FDA label); LDN trials report no sex differences in adverse events.
  • Opioid use: Any current opioid use, including tramadol, methadone or buprenorphine, turns a low-risk drug into a risk of precipitated withdrawal and blocked analgesia (FDA label).
  • Autoimmune thyroid disease: Practitioners report LDN may reduce thyroid hormone needs (theoretical), but Norwegian prescription data showed no change in thyroid hormone dispensing (Raknes & Småbrekke, 2020).
  • Age: At high doses, transaminase elevations were more frequent in people aged 40 and over (Pfohl et al., 1986); older adults also more often need surgery with opioid pain control.

Key Interactions & Contraindications

  • Opioid analgesics (morphine, oxycodone, hydrocodone, fentanyl, tramadol, codeine) — absolute contraindication: pain relief is blocked and precipitated withdrawal is possible; the label contraindicates naltrexone with opioid analgesics (FDA label). Pausing LDN before planned opioid use avoids the conflict.
  • Opioid maintenance (methadone, buprenorphine) — absolute contraindication: severe precipitated withdrawal, possibly requiring hospitalization; vulnerability persists up to two weeks after stopping these drugs (FDA label).
  • Opioid-containing over-the-counter products (loperamide, codeine cough syrups where sold without prescription): reduced or absent effect of the product (FDA label). Non-opioid alternatives avoid the conflict.
  • Benzodiazepines (sedative anti-anxiety drugs such as diazepam): in a human study, 50 mg naltrexone delayed peak diazepam levels from 75 to 135 minutes and increased sedation, fatigue and anxiety ratings (Swift et al., 1998). Separating doses limits overlap.
  • Thioridazine (an antipsychotic): lethargy and somnolence (abnormal sleepiness) were reported with naltrexone (FDA label). Watching for excess drowsiness after starting is the usual step.
  • Liver-toxic drugs (disulfiram; others such as high-dose acetaminophen or ketoconazole theoretical): additive liver-injury risk; the label advises against combining two potentially liver-toxic drugs (FDA label). Liver tests at baseline and after starting address this.
  • Acamprosate (an alcohol-dependence drug): a human pharmacokinetic (drug-level) study found no change in naltrexone levels and no adverse safety or cognitive interaction, while acamprosate exposure rose about 25–33% (Mason et al., 2002). No dose change is needed.
  • Thyroid hormone (levothyroxine, liothyronine): a proposed fall in thyroid hormone needs could cause over-replacement (theoretical); Norwegian data found no dispensing change (Raknes & Småbrekke, 2020). Thyroid-stimulating hormone (TSH) is rechecked 4–8 weeks after starting.
  • Immunosuppressants (tacrolimus, cyclosporine, mycophenolate): immune-modulating effects could oppose transplant or autoimmune immunosuppression (theoretical). Coordination with the prescribing specialist addresses this.
  • Kratom (mitragynine) — supplement, avoid: kratom alkaloids act on mu-opioid receptors, so LDN may blunt kratom effects or precipitate withdrawal in regular users (theoretical).
  • Additive supplements (palmitoylethanolamide, curcumin, omega-3 fatty acids): may add to LDN’s proposed anti-inflammatory and pain-relieving effects (theoretical); no combination studies exist.
  • Other interventions (rapamycin, metformin, NAD+ [nicotinamide adenine dinucleotide, a cellular energy cofactor]): combined with LDN in longevity programs and in a registered pilot trial (NCT07475546); rapamycin’s immune suppression could blunt or mask LDN’s proposed immune effects (theoretical), and no interaction data exist.
  • Cannabis (THC, tetrahydrocannabinol, its main psychoactive compound): in heavy cannabis users, 50 mg naltrexone increased the positive subjective effects of oral THC without changing THC blood levels (Haney et al., 2003); effects at LDN doses are untested.
  • Alcohol: naltrexone produces no disulfiram-like reaction (flushing and nausea on drinking) with alcohol (FDA label); at 50 mg it reduces drinking, while LDN effects on intake are untested.

Populations who should avoid Low-Dose Naltrexone:

  • People taking opioid analgesics or currently dependent on opioids, including methadone or buprenorphine maintenance, and those opioid-free for less than 7–10 days after short-acting opioids or under two weeks after methadone or buprenorphine (FDA label)
  • Anyone with a positive urine opioid screen or a failed naloxone challenge test (a test dose of a short-acting opioid blocker that reveals hidden dependence) (FDA label)
  • People with symptoms or signs of acute hepatitis (FDA label)
  • People with decompensated liver cirrhosis, in whom exposure rises about 10-fold (FDA label)
  • Pregnant people, given the absence of adequate human studies (FDA label)
  • People with known hypersensitivity to naltrexone (FDA label)
  • People under 18 outside a supervised study, as safety is not established (FDA label)

Risk Mitigation Strategies

Doses and timings below follow common practice unless cited.

  • Opioid-free interval before starting: stopping all opioids, including tramadol, 7–10 days before the first dose, and two weeks after methadone or buprenorphine, prevents precipitated withdrawal (FDA label).
  • Pausing around surgery: pausing LDN about 72 hours before procedures likely to need opioid pain control, and restarting after opioids end, avoids blocked pain relief.
  • Medical identification: a card or note stating naltrexone use alerts emergency staff that opioid analgesia may be blocked (FDA label).
  • Slow titration: starting at 1.5 mg nightly and adding 1.5 mg weekly, within a four-week titration as in the FINAL protocol (Bruun et al., 2021), reduces vivid dreams and insomnia during adjustment.
  • Dose reduction or morning dosing: a review reports that vivid dreams and headaches eased when the dose was lowered to 3 mg (Toljan & Vrooman, 2018); morning dosing is the alternative for persistent sleep disruption.
  • Baseline liver tests: ALT (alanine aminotransferase) and AST (aspartate aminotransferase) at baseline detect pre-existing liver injury; use stops with symptoms or signs of acute hepatitis (FDA label).
  • Thyroid retesting: for people on thyroid hormone, a TSH check 4–8 weeks after starting catches any over-replacement (palpitations, insomnia) early.
  • Immediate-release preparations: avoiding slow-release fillers keeps blockade brief, away from the continuous pattern linked to enhanced cell growth in animal work (McLaughlin & Zagon, 2015).
  • Pregnancy exclusion: a pregnancy test before starting and stopping on conception limit fetal exposure, given animal fetal-loss data (FDA label).
  • Time-limited trial: judging response after 8–12 weeks against a baseline symptom score, an N-of-1 (single-person trial) approach suggested by reviewers (Gouda et al., 2026), prevents open-ended side-effect exposure without benefit.

Therapeutic Protocol

Timing, titration steps and other parameters without a citation reflect common practice.

  • Standard dose: 4.5 mg once nightly, the dose in the main fibromyalgia, Crohn’s disease and multiple sclerosis trials (Younger et al., 2013; Smith et al., 2011; Cree et al., 2010).
  • Higher-dose variant: 6 mg once daily in the Danish FINAL trial (Due Bruun et al., 2024); dose-finding estimated 3.88 mg effective in half and 5.40 mg in 95% of women with fibromyalgia (Bruun-Plesner et al., 2020).
  • Lower-dose origin: Bernard Bihari, who popularized LDN, used 1.5–3 mg; very-low (1 µg–1 mg) and ultra-low (<1 µg) doses are distinct regimens used mainly alongside opioids (Toljan & Vrooman, 2018).
  • Titration: a 1.5 mg starting dose with 1.5 mg weekly increases; the FINAL protocol used a four-week titration (Bruun et al., 2021); integrative clinicians use 0.5–1 mg steps for sensitive patients.
  • Neuropathy regimen: 2 mg daily, raised to 4 mg if pain fell less than 20%, in the diabetic neuropathy trial (Srinivasan et al., 2021).
  • Long COVID range: 1–6 mg daily across clinic-based studies (Byambasuren et al., 2026).
  • Competing approaches: integrative practitioners such as Chris Kresser regard LDN as a low-harm long-term option for autoimmune conditions, with most patients settling at 2.5–3 mg (Kresser); critical reviewers place it after approved therapies, for treatment-resistant cases (Gouda et al., 2026).
  • Time of day: bedtime dosing follows the “nightly” trial regimens (Cree et al., 2010); morning dosing is the usual switch when sleep is disturbed.
  • Half-life: about 4 hours for naltrexone and 13 hours for 6-β-naltrexol (FDA label), so receptor blockade lasts only part of each day.
  • Single versus split dose: LDN is taken as one daily dose; splitting would lengthen blockade, contrary to the intermittent-blockade rationale (McLaughlin & Zagon, 2015).
  • Genetics: no genotype guides LDN dosing; OPRM1 rs1799971 testing is not supported even for full-dose naltrexone (Counts et al., 2025).
  • Sex: dose-finding and most fibromyalgia data come from women (Bruun-Plesner et al., 2020); no sex-specific dosing exists.
  • Age: no age-specific dosing exists; with reduced kidney function in older adults, lower starting doses are common practice, given urinary excretion (FDA label).
  • Baseline biomarkers: higher baseline ESR predicted response in fibromyalgia (Younger & Mackey, 2009); no biomarker sets the dose.
  • Pre-existing conditions: cirrhosis raises exposure 5–10-fold (FDA label), favoring lower doses; current opioid use excludes LDN.

Discontinuation & Cycling

  • Duration: trials ran 8 weeks to 12 months (Rodríguez-Freire et al., 2026); in practice LDN continues long-term when it helps, without controlled safety data beyond one year.
  • Withdrawal effects: none expected; the label states naltrexone causes no tolerance or dependence (FDA label).
  • Tapering: not required; stopping at once is usual, though symptoms LDN was controlling may return over days to weeks.
  • Cycling: no evidence supports cycling for efficacy; the intermittent-blockade theory (McLaughlin & Zagon, 2015) leads some clinicians to skip occasional doses (theoretical).
  • Temporary stops: pauses before planned opioid analgesia are standard, with restart once opioids are cleared.
  • Opioid sensitivity after stopping: formerly opioid-dependent people become more sensitive to opioids after stopping naltrexone, raising overdose risk (FDA label).
  • Stopping for non-response: without meaningful change from baseline after 8–12 weeks, discontinuation ends exposure without benefit.

Sourcing and Quality

  • Compounded product: low doses need compounding from naltrexone (Cote et al., 2018); compounded drugs are not FDA approved, so their safety, effectiveness and quality are not reviewed before marketing (FDA).
  • Formulation: capsules (0.5–6 mg) or liquids; immediate-release fillers such as microcrystalline cellulose (Avicel) keep blockade brief, whereas slow-release fillers prolong it.
  • Stability: LDN capsules in Avicel stayed within 90–110% of labeled potency for 360 days at room temperature away from light (Cote et al., 2018).
  • What to look for: a licensed pharmacy with compounding accreditation, such as PCAB (Pharmacy Compounding Accreditation Board), batch potency testing, and stated fillers; established LDN compounders include Belmar Pharma Solutions and Skip’s Pharmacy in the US.
  • Liquid from tablets: some prescribers have patients dissolve a generic 50 mg tablet in 50 mL water (1 mg/mL) and measure with an oral syringe; accuracy depends on mixing.
  • Telehealth suppliers: online clinics such as AgelessRx prescribe and ship LDN and earn revenue from it, so their efficacy claims carry a commercial interest.

Practical Considerations

  • Time to effect: trials judged outcomes after 8–12 weeks (Cree et al., 2010; Smith et al., 2011); clinicians typically allow two to three months before deciding.
  • Common pitfalls: combining with opioids, using slow-release fillers, quitting during early vivid dreams, judging response without a baseline score, and assuming a proven longevity effect.
  • Regulatory status: naltrexone is FDA-approved at 50 mg for alcohol dependence and opioid blockade (FDA label); LDN use is off-label and relies on compounding (FDA).
  • Cost and access: LDN is inexpensive (Gouda et al., 2026); access depends on finding a prescriber willing to use it off-label.
  • Funding and payer incentives: off-patent naltrexone attracts no manufacturer funding for large trials; payers would gain if an inexpensive drug displaced costly biologics (engineered protein drugs), so payer incentives favor testing LDN, while makers of costly alternatives have none.
  • Commercial interests: much public enthusiasm comes from compounding pharmacies and telehealth clinics that sell LDN, alongside academic groups without such ties.

Interaction with Foundational Habits

  • Sleep: direct, potentially disrupting: pooled fibromyalgia trials showed more vivid dreams than placebo (Nazir et al., 2025), and in fibromyalgia LDN did not improve sleep (Younger et al., 2013). Morning dosing is the usual adjustment.
  • Nutrition: direct, exposure-raising: in adolescents, taking naltrexone with food significantly increased blood exposure versus fasting (Stancil et al., 2022), so consistent timing relative to meals keeps levels steadier; no specific diet is required, and alcohol causes no disulfiram-like reaction (FDA label).
  • Exercise: possibly blunting (theoretical): exercise-induced analgesia and mood lift partly depend on endorphins, which blockade could dampen in the hours after dosing; bedtime dosing keeps daytime training outside that window. No human LDN exercise study exists.
  • Stress management: indirect (theoretical): endogenous opioids help regulate stress responses, and brief nightly blockade could transiently heighten them; no human LDN study has measured cortisol or stress outcomes.

Monitoring Protocol & Defining Success

Before starting, baseline testing confirms safety and sets a personal reference point: liver enzymes, kidney function, a urine opioid screen where any opioid use is possible, thyroid tests for people taking thyroid hormone, a pregnancy test where relevant, and a validated symptom score for the target condition. Success is defined against this baseline, for example a 30% or greater fall in pain, a responder threshold used in fibromyalgia trials (Bruun et al., 2021).

Ongoing monitoring follows this cadence: thyroid tests and symptom scores 4–8 weeks after starting or after a dose change, a response review at 12 weeks, then liver and kidney tests every 6–12 months while use continues; symptoms of hepatitis prompt immediate testing (FDA label).

Biomarker Optimal Functional Range Why Measure It? Context/Notes
ALT 7–56 U/L (standard reference range) Safety check: liver injury stops use Label hepatotoxicity warning (FDA label); no fasting needed; pair with AST; ranges are laboratory-specific
AST 10–40 U/L (standard reference range) Safety check: liver injury stops use Pair with ALT; strenuous exercise in the prior 48 hours can raise AST
eGFR ≥60 mL/min/1.73 m² (standard reference range) Safety check: kidney impairment changes use eGFR (estimated glomerular filtration rate, kidney filtering capacity); drug and metabolite are excreted in urine (FDA label)
Urine opioid screen No established target; track a negative result before the first dose Safety check: a positive result stops use Positive screen is a label contraindication (FDA label); some screening tests may cross-react
TSH 0.4–4.0 mIU/L (standard reference range) Safety check: over-replacement changes thyroid dosing Only for people on thyroid hormone; morning sample, same laboratory each time
Pregnancy test No established target; track a negative result before starting Safety check: pregnancy stops use hCG (human chorionic gonadotropin) urine or blood test; label reports animal fetal loss (FDA label)

Qualitative markers:

  • Target symptoms: pain, fatigue or disease activity versus baseline
  • Sleep: dream intensity, insomnia, sleep quality
  • Mood and energy: daily well-being
  • Side effects: nausea, headache, dizziness

Emerging Research

  • Brave-LC (long COVID fatigue): a National Institutes of Health (NIH) RECOVER-TLC trial of LDN for long COVID symptoms in people aged 6–25, protocol in development, no NCT ID registered as of October 2026 (RECOVER). A positive result would give the first large trial support; a null result would undercut clinic-based signals.
  • British Columbia post-COVID fatigue trial: double-blind (participants and staff unaware of allocation), phase 2 (mid-stage efficacy testing), 160 participants, primary outcome fatigue intensity (NCT05430152; Naik et al., 2024); completed February 2026, no results posted. Positive results would strengthen the fatigue item; null results would weaken it.
  • LIFT trial: phase 2, 160 participants with myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS) or long COVID, testing LDN and pyridostigmine against placebo on exercise capacity (NCT06366724); recruiting, completion 2027. Positive results would add objective functional evidence; null results would weaken the fatigue case.
  • ME/CFS dose-finding: University of Alabama at Birmingham study of 75 participants comparing 1.5, 3.0, 4.5 and 6.0 mg on a PROMIS (patient-reported outcome) fatigue score (NCT07285473); not yet recruiting. A dose-response would support a real effect; a flat response would suggest none.
  • Combination gerotherapeutics (drugs targeting aging itself) pilot: AgelessRx-sponsored randomized study of 30 adults combining LDN with rapamycin, metformin, NAD+ and other agents (NCT07475546); April 2026 completion passed, status “active, not recruiting”, no results posted. Either result cannot isolate LDN; the sponsor sells these drugs.
  • Complex regional pain syndrome (persistent limb pain after injury) trial: Stanford randomized placebo-controlled study, 120 participants, primary outcome pain severity (NCT02502162); recruiting, completion 2027. Positive results would extend controlled pain evidence beyond fibromyalgia; null results would narrow it further.
  • Diabetic neuropathy crossover trial: Dartmouth placebo-controlled phase 2 study, 35 participants, primary outcomes pain disability and pain rating (NCT04678895); recruiting, completion December 2026. Positive results would add placebo control to the active-comparator signal; null results would weaken it.
  • Evidence weakening the case: a 12-month fibromyalgia trial (Rodríguez-Freire et al., 2026) and a Dutch Crohn’s trial (van de Pol et al., 2026) found no benefit, and a 2026 review found early positive results rarely replicated (Gouda et al., 2026).
  • Longevity research gap: LDN extended worm lifespan through SKN-1 (Li et al., 2024), but a PubMed search in October 2026 found no published mammalian lifespan study; such a study would decide whether a longevity rationale exists.

Conclusion

Low-dose naltrexone is an inexpensive, decades-old addiction drug taken at a small fraction of its usual dose, in the hope of easing pain, fatigue and immune-driven inflammation. For health-focused adults weighing it within a longevity program, the evidence divides sharply by question.

For chronic pain, especially fibromyalgia, small early trials from one research group and studies combining several trials point to a modest benefit, but the largest independent trials found no difference from placebo. For Crohn’s disease, multiple sclerosis, mood and long COVID fatigue, results range from encouraging findings without comparison groups to clearly negative. Claims of slower aging rest on worm experiments and laboratory markers; no study has measured aging or lifespan in people or other mammals.

Safety is a genuine strength. Across many controlled trials serious harms were not increased, and the usual cost is vivid dreams, with occasional nausea or headache. The firm hazards involve opioid pain medicines and opioid treatment programs, where the drug blocks pain relief or can set off severe withdrawal; liver concerns come only from far higher doses.

The evidence base is small, short and built on single conditions. Because the drug is no longer patented, no manufacturer pays for large trials, while much of the enthusiasm comes from custom-mixing pharmacies and online clinics that sell it. The overall picture is a low-risk drug with uncertain, condition-specific benefits and no demonstrated effect on longevity itself.

Top - Benefits - Risks - Protocol