Low-Dose Naltrexone Drug

## What it is Naltrexone is a long-acting oral opioid receptor antagonist developed in the 1960s and FDA-approved in 1984 (Trexan, later ReVia) for opioid-use disorder at 50 mg daily, and in 2010 (Vivitrol) as a monthly extended-release injectable for opioid and alcohol-use disorders. Low-dose naltrexone (LDN) refers to off-label use at 1.5 to 4.5 mg daily, approximately one-tenth the addiction-treatment dose, for indications spanning autoimmune disease, fibromyalgia, chronic pain, and various inflammatory conditions. The LDN concept originated with Bernard Bihari, a New York physician who in the mid-1980s observed that very low doses of naltrexone appeared to improve symptoms in HIV/AIDS patients before the antiretroviral era. He extended the protocol to autoimmune and chronic pain populations through the 1990s and 2000s. His work was largely uncontrolled and built a patient community that pushed academic interest into low-dose naltrexone in the 2010s. The regulatory shape of LDN is unusual. Naltrexone itself is FDA-approved for opioid-use disorder and alcohol-use disorder at the 50 mg dose, but no manufacturer has sought approval for low-dose indications. As a result, all LDN dispensed in the US comes from compounding pharmacies that prepare 1.5, 3, and 4.5 mg capsules from bulk naltrexone hydrochloride. This is legal compounding for an off-label indication and does not require an approved low-dose product. Insurance coverage is rare; cash prices range from 20 to 80 USD per month. The evidence base is heavy on enthusiast literature and light on rigorous trials. The Younger 2014 review is the most-cited academic synthesis. Across fibromyalgia, Crohn's disease, multiple sclerosis-related fatigue, complex regional pain syndrome, and Hashimoto's thyroiditis, the trial portfolio consists of small open-label studies, single-blind crossovers, and a handful of placebo-controlled trials with sample sizes typically under 50. The honest grade across nearly all LDN indications is C: signal exists, mechanism is plausible, evidence base is preliminary. ## Mechanism of action The proposed LDN mechanism is fundamentally different from the addiction-treatment use. At 50 mg daily, naltrexone produces sustained mu-opioid receptor blockade, preventing opioid agonists from producing reinforcing effects. At 1.5 to 4.5 mg, plasma levels rise briefly (peak 1 to 2 hours after dose), produce transient receptor blockade lasting roughly 4 to 6 hours, and then drop below pharmacologically active levels for the remainder of the day. Two mechanistic frames coexist in the LDN literature. The first, often called the 'rebound endorphin' hypothesis, proposes that brief opioid receptor blockade triggers a compensatory upregulation of endogenous opioids (beta-endorphin, met-enkephalin) and opioid receptor density. After the naltrexone clears, the elevated endogenous opioid tone produces analgesic and immune-modulating effects. This hypothesis is invoked to explain the dosing pattern (single bedtime dose, rather than continuous coverage) and why higher doses are reportedly less effective. The second mechanism is non-opioid. Naltrexone (and its primary metabolite 6-beta-naltrexol) is a TLR4 antagonist. Toll-like receptor 4 is expressed on microglia and immune cells and contributes to inflammatory signaling in chronic pain and autoimmune contexts. TLR4 antagonism could plausibly explain the analgesic effect in fibromyalgia and the anti-inflammatory effects across autoimmune conditions, independent of any opioid pathway. The truth probably involves both mechanisms in different proportions across different indications. The mechanistic case is strong enough to motivate continued investigation; the clinical evidence has not caught up. ## Evidence base by outcome ### Fibromyalgia The largest LDN evidence base. Younger 2009 and Younger 2013 (Stanford) ran small placebo-controlled crossover trials in fibromyalgia patients and reported pain reductions of roughly 30% versus placebo at 4.5 mg LDN. Sample sizes were 10 and 31 respectively. Bruun-Plesner 2020 (Denmark) ran a larger placebo-controlled trial (n=99) and reported smaller and less consistent effects, with subgroup signals but failure to meet primary endpoint cleanly. A 2025 systematic review concluded that signal exists but evidence base is inadequate for a routine clinical recommendation. ### Crohn's disease Smith 2007 (n=17 open-label) and Smith 2011 (n=40 placebo-controlled) reported endoscopic and clinical remission improvements in active Crohn's. The trials are small and methodologically limited but consistent in direction. The Cochrane review (Segal 2014) found insufficient evidence for routine use but noted the consistent direction of effect across small trials. ### Multiple sclerosis fatigue and quality of life Cree 2010 (n=80 placebo-controlled crossover) reported small improvements in self-reported quality of life in MS patients on LDN, with no effect on objective disability measures. Other small trials have been mixed. LDN is a popular self-prescribed adjunct in MS communities; the trial evidence is C-tier. ### Hashimoto's thyroiditis and other autoimmune conditions Mostly anecdotal and observational. A handful of case series report reduced TPO antibody titers and improved subjective well-being. No well-designed RCT has tested LDN in Hashimoto's. The evidence is D-tier on hard endpoints, C-tier on subjective improvement. ### Complex regional pain syndrome (CRPS) Chopra 2013 case series (n=2) reported substantial pain improvement with LDN. Other small case series have followed. Evidence is D-tier. ### Chronic Lyme and post-treatment Lyme syndrome No controlled trials. Anecdotal use only. D-tier. ### Long COVID and post-viral fatigue A small number of pilot trials and case series have reported LDN improvements in long COVID symptoms. The O'Kelly 2022 pilot (n=37) reported subjective improvements; placebo-controlled data are pending. C-tier on emerging signal. ## Dosage and administration The canonical LDN protocol starts at 1.5 mg taken at bedtime, titrating to 3 mg after 1 to 2 weeks and 4.5 mg after another 1 to 2 weeks if tolerated. Some protocols use 4.5 mg from the start. Bedtime dosing is the most common pattern, on the rationale that the brief receptor blockade aligns with peak nighttime endogenous opioid release. A morning-dosing variant exists for users who experience vivid dreams at bedtime. No cycling is part of standard protocol. Most LDN users continue dosing indefinitely as long as the indication persists and tolerability holds. Some practitioners try 1 to 2 week washouts to assess whether benefit is medication-driven or coincidental. The dose ceiling is empirical. Doses above 4.5 mg appear to lose the LDN-pattern benefit, possibly because sustained receptor blockade outlasts the rebound window. Doses below 1.5 mg appear to be sub-therapeutic. This U-shaped dose-response is part of the mechanistic case for the brief-blockade-rebound hypothesis but has not been mapped rigorously in trials. ## Side effects and safety The overall safety profile is favorable. Vivid dreams or sleep disruption affect roughly 20 to 40% of users in the first 1 to 2 weeks; these typically resolve and can be managed with morning dosing. Mild GI upset, headache, and nausea occur in a smaller fraction. The major contraindication is concurrent opioid use. LDN will precipitate opioid withdrawal in opioid-dependent patients and will block the analgesic effect of opioids in surgical or acute-pain contexts. Patients on chronic opioids for pain must discontinue opioids 7 to 10 days before starting LDN, or vice versa. This is the single most important practical consideration. Liver function should be checked at baseline and periodically. The 50 mg dose has rare hepatotoxicity reports; the LDN dose has not been associated with liver injury but the precaution carries over. Drug interactions are limited at the LDN dose. The opioid antagonism is the dominant interaction. Combination with thyroid hormone replacement does not require dose adjustment in the available data, but autoimmune thyroid patients should monitor TSH after starting LDN in case the immunomodulation alters thyroid replacement requirements. Pregnancy and lactation data are insufficient. Routine use is not recommended. ## Stack interactions and timing LDN does not have well-characterized stacks. It is sometimes paired with other immunomodulators in autoimmune protocols, but the combinatorial evidence is essentially absent. Pairing with vitamin D, omega-3 fatty acids, and standard autoimmune nutritional supplements is common and uncomplicated. The critical interaction is with opioids: avoid concurrent use. This includes prescription opioid analgesics, codeine cough syrups, and tramadol. Surgical planning requires LDN discontinuation 24 to 72 hours before procedures requiring opioid analgesia. Dosing timing is bedtime by convention; switch to morning if dreams disrupt sleep. ## Practical notes LDN must be obtained from a compounding pharmacy with a prescription. Compounded capsules are typically supplied as 1.5, 3, or 4.5 mg in 30 or 90 day quantities. Liquid LDN preparations exist for dose flexibility, particularly for pediatric or sensitive users; storage is refrigerated. Quality control varies across compounding pharmacies. Pharmacies certified by PCAB (Pharmacy Compounding Accreditation Board) provide more reliable dosing accuracy than uncertified compounders. Expect an evaluation period of 6 to 12 weeks before judging response. Some users report immediate effects; the trial data show effect sizes accumulating over the first 8 to 12 weeks of consistent dosing. The honest framing for the LDN use case: the mechanistic plausibility is real, the safety record is reassuring, and the evidence base across all indications is preliminary. It is a reasonable trial in patients with fibromyalgia, Crohn's, or autoimmune conditions where standard therapy has been inadequate, but it should not be presented as established treatment. Users and clinicians considering LDN should treat it as an off-label experimental option with a favorable safety profile, not as a settled clinical recommendation.