Low-Dose Naltrexone and Sleep Architecture: What the Evidence Actually Shows

Does LDN change REM or deep sleep?

Sleep architecture is the stage structure of a night: wake, N1, N2, N3 (slow-wave or deep sleep) and REM, scored from a polysomnography recording. The LDN trials described below did not record those stages.

The cited LDN trials measured outcomes such as pain, function, mood, sleep quality and adverse events, but not polysomnographic sleep stages [1][7][8]. Popular explanations describe a brief opioid-receptor blockade followed by a rebound in endogenous opioid signaling, or an effect on microglial toll-like receptor 4. A 2026 narrative review of 105 studies, 15 of them randomized trials, calls the proposed LDN mechanisms 'largely theoretical and speculative' [3]. A proposed mechanism is not a measured change in REM time or slow-wave sleep.

What the fibromyalgia trials measured

The clearest sleep datapoint in the LDN literature comes from a randomized, double-blind, placebo-controlled crossover trial published in Arthritis & Rheumatism in 2013. Thirty-one women with fibromyalgia took 4.5 mg naltrexone daily or placebo. Pain fell 28.8% from baseline on LDN versus 18.0% on placebo (P=0.016), and mood and life satisfaction improved, but the authors reported no improvement in fatigue or sleep [1]. Thirty-two percent met a composite response criterion on LDN versus 11% on placebo (P=0.05) [1].

Later trials were larger and did not reproduce the pain result. A 2024 Danish trial randomized 99 women to 6 mg daily or placebo for 12 weeks; pain intensity on an 11-point scale changed by -1.3 points with LDN and -0.9 with placebo, a between-group difference of -0.34 (95% CI -0.95 to 0.27, p=0.27) [7]. The 2026 INNOVA trial randomized 98 women to 4.5 mg daily or placebo for 12 months; at 3 months pain changed by -0.33 points on LDN and -0.64 on placebo, an adjusted difference of 0.49 (p=0.236) [8]. A 2023 crossover trial of 58 patients reported a median difference in revised Fibromyalgia Impact Questionnaire score of -1.65 (P=0.3) [9].

A 2024 meta-analysis of four randomized trials (222 patients) found pain lower with LDN by a mean of 0.86 points (95% CI -1.20 to -0.51) but no significant difference on the revised Fibromyalgia Impact Questionnaire, and more vivid dreams and nausea with LDN [10].

Where the insomnia and vivid-dream reports come from

A 2009 fibromyalgia pilot of ten women, a single-blind crossover, listed insomnia and vivid dreams among side effects it described as rare, minor and transient [11]. That is a tolerability note, not a sleep measurement, and a pilot of 10 people cannot establish how often either effect occurs [11]. The 2024 meta-analysis did find vivid dreams and nausea reported more often with LDN than with placebo [10].

Vivid-dream reports are patient-reported; they are not a measure of REM or deep-sleep time.

What the approved naltrexone labels record

Rate data for sleep complaints exist for the FDA-approved naltrexone products, which use far higher doses. In the 50 mg tablet label's open-label safety study of about 570 people treated for alcoholism, new-onset insomnia was reported in 3% and somnolence in 2% [5]. In the tablet's opioid-addiction trials, difficulty sleeping was reported by more than 10% at baseline and during the trials, and nightmares and bad dreams by less than 1% [5].

The Vivitrol label reports insomnia or sleep disorder in 14% of people on the 380 mg injection versus 12% on placebo in alcohol dependence (205 versus 214 patients), and insomnia in 6% versus 1% in opioid dependence (126 versus 124 patients) [6].

Those labels were written for 50 mg tablets and 380 mg injections and give no guidance on LDN doses [5][6][3]. A 3% insomnia rate at 50 mg is not a prediction for 4.5 mg.

The long COVID data that did measure sleep

A 2026 systematic review of LDN for long COVID searched to 5 May 2026 and identified no randomized trials. The four uncontrolled before-and-after studies it pooled (155 people, LDN 1 to 6 mg/day) reported improved sleep quality, with a Hedges g of -0.60 (95% CI -0.91 to -0.30, p=0.0001) [2].

Uncontrolled before-and-after data cannot show benefit over placebo. The outcome was reported sleep quality, not scored sleep stages [2].

Is morning or evening dosing better for sleep?

The multiple sclerosis pilot gave 4.5 mg nightly for 8 weeks to 80 enrolled patients, 60 of whom completed, and reported improved mental-health quality-of-life scores [4]. The fibromyalgia trials used once-daily dosing [1][7]. These trials tested one schedule against placebo rather than comparing a morning schedule with an evening one, so they do not show which time of day is better for sleep.

Opioids matter more here than sleep stages

Naltrexone is a pure opioid antagonist that blocks opioid effects by competitive binding at opioid receptors, and the blockade is potentially surmountable [5]. The labels contraindicate it in patients receiving opioid analgesics, in people currently dependent on opioids including methadone or buprenorphine maintenance, and in acute opioid withdrawal [5][6]. The contraindication is written without a dose exception.

Precipitated withdrawal in an opioid-dependent person can be severe enough to require hospital admission, with symptoms usually appearing within 5 minutes of the dose and lasting up to 48 hours [5]. The tablet label recommends a minimum of 7 to 10 opioid-free days for short-acting opioids and notes that people switching from buprenorphine or methadone may remain vulnerable for as long as 2 weeks [5]. Opioid-containing cough, cold and antidiarrheal products are antagonized too [5].

Compounding and product variability

No low-dose naltrexone product is FDA-approved; LDN is used off-label, and compounded drugs are not FDA-approved [3][12]. FDA says it does not verify the safety, effectiveness or quality of compounded drugs before they are marketed [12]. A 2026 review notes that inconsistencies in compounded formulations add uncertainty [3]. HealthRX.com offers low-dose naltrexone by prescription.