LDN Side Effects: What Low-Dose Naltrexone Trials Found Compared With Placebo

Side effects were about as common on placebo in the largest fibromyalgia trials

Most placebo-controlled side-effect data for LDN come from fibromyalgia trials. Poor sleep and fatigue are also fibromyalgia symptoms, which makes side effects harder to pin on the drug.³

The INNOVA 21.1% and 8.5% are shares of reported events (16 of 76 and 7 of 82), not of the women in each group.³ In Crohn's disease, a Cochrane review of two small trials (46 people) found no rise in specific side effects on LDN, with low-quality evidence.⁹

Vivid dreams and nausea were more common than on placebo

Across two fibromyalgia trials with 127 people, the odds of vivid dreams were about three times higher on LDN (odds ratio 2.97; 95% CI, 1.30-6.78).⁵ In one research group's LDN trials, about 37% of participants reported more vivid dreams, and a minority had nightmares. Vivid dreams were also the most common complaint while those participants took placebo, so expectation may explain some cases.⁶

Trouble sleeping is less clear. In a 10-woman pilot, one person had nausea and insomnia for the first few nights, and two reported more vivid dreams.¹⁰ In the Bested trial, no one reported worse sleep on LDN and 3 did on placebo.⁷ LDN is commonly taken about an hour before bed; in one research group's practice, people with insomnia were moved to a morning dose.⁶ Sleepiness and the 50 mg tablet are covered on does naltrexone make you sleepy?

Nausea. Pooled across three fibromyalgia trials with 179 people, the odds of nausea were higher on LDN (odds ratio 2.75; 95% CI, 1.35-5.61).⁵ Two of the 58 people in the Bested trial dropped out with nausea and vomiting right after starting.⁷ Diarrhea did not differ significantly from placebo.⁵

Headache. Pooled headache odds did not differ (odds ratio 1.02; 95% CI, 0.54-1.93), and the Bested trial counted 9 people with headache in each phase.^5,7 One research group saw slightly more headaches on LDN than on placebo, too few to test, and notes headaches are common in fibromyalgia.⁶

Dizziness, tiredness and anxiety. Pooled dizziness was borderline and not significant (odds ratio 2.16; 95% CI, 0.99-4.73).⁵ In the Bested trial, 9 people reported fatigue on LDN and 4 on placebo.⁷ Some doctors have described anxiety and a fast heartbeat on LDN, but trials have not recorded them.⁶

How long LDN side effects last

None of the trial reports read for this page gives a typical duration in days or weeks. The closest thing is the 12-month INNOVA trial, which counted events at each check. LDN had more events than placebo at 1 month and fewer at 12 months, and the authors describe the frequency over time as similar in the two groups.³

Most events (84.1%) had resolved by the end of the trial. The INNOVA team started everyone on the full 4.5 mg and suggested this may have added to short-lived nervous-system effects.³ Early researchers saw vivid dreams start with the first dose and ease over time, and cut the dose to 3 mg for people with side effects.⁶ Dose and timing changes are for your prescriber to decide.

What 12 months of trial data leave out

INNOVA is the longest placebo-controlled LDN trial we found: 98 women took 4.5 mg or placebo for a year. About a third of each group left before 12 months (63.8% on LDN and 65.3% on placebo stayed).³

Serious adverse events in INNOVA. The results table lists serious adverse events in 4 of 48 women on LDN and 1 of 50 on placebo, none judged treatment-related; the paper's discussion says there were none, and it does not describe the 5 events.³

What a year of data cannot show:

• Rare problems. Only 48 women took LDN, too few to pick up uncommon effects.³
• Blood and liver tests. INNOVA did no routine laboratory tests and excluded people with abnormal liver or kidney function.³
• Several years of use. A 2014 review by LDN researchers said "we know very little about the long-term safety of the drug when used chronically in low dosages"; a 2026 review still found safety studies sparse.^2,6

Across all doses, a 2019 review of 26 placebo-controlled naltrexone trials (4,960 people) found no increase in serious adverse events (risk ratio 0.84; 95% CI, 0.66-1.06). Only 5 trials used a target dose of 4.5 mg or less, and the authors caution that results for doses under 26 mg rest on few studies.¹¹

Liver: the label's warnings come from 50 to 300 mg doses

LDN has no label, so quoted liver warnings come from the 50 mg tablet, whose label reports hepatitis cases with naltrexone in general. Its dose-related warning comes from studies at up to 300 mg a day, which found that naltrexone "causes hepatocellular injury in a substantial proportion of patients exposed at higher doses."¹² The dosing section adds: "There may be a higher risk of hepatocellular injury with single doses above 50 mg."¹²

Evidence at 4.5 mg is thin. The authors of the first LDN fibromyalgia pilot noted that liver-enzyme changes had not been found at 4.5 mg in earlier work, yet suggested liver function may need monitoring.¹⁰ A 2026 review found no evidence on whether the 50 mg cautions for liver disease extend to low doses.² In a 20-year US liver-injury registry, the only case blamed on naltrexone was a woman with known fatty liver taking 4.5 mg a day; the injury was rated mild and probably drug-caused, and she recovered within a week of stopping.¹³ One case cannot give a rate. Liver tests and the 50 mg warning are covered on naltrexone and your liver.

Stopping LDN: no withdrawal seen, and no taper has been tested

Researchers who ran early LDN trials wrote: "We have observed no withdrawal symptoms when LDN treatment is stopped." They saw no dependence or tolerance, and after stopping, symptoms made a "slow return" to where they were before.⁶ Case reports in Hailey-Hailey disease, a rare skin condition, often noted relapse after LDN was withdrawn.²

No trial has compared stopping at once with tapering, so the advice online is practice opinion. The LDN Research Trust, an advocacy charity, says "there is no need to taper down" when stopping. Its Q&A adds that people restarting after a break are often started below their original starting dose.^14,15

Talk to your prescriber before you stop, especially if LDN is part of treatment for a condition such as Crohn's disease or multiple sclerosis. If you have used opioids, ask how the 50 mg label's warning about overdose after stopping naltrexone applies to you; the warning comes from 50 mg use.¹²

Opioid painkillers may work less well on LDN

Even a low dose may weaken opioid pain medicines, and early LDN trials excluded people taking them.⁶ Opioids, kratom, cough and diarrhea medicines are graded one by one in what to avoid on low-dose naltrexone. Planning around an operation is in surgery and procedures.

Your experience: If you have taken LDN, you can share your story on our naltrexone reviews page. LDN stories will be labeled off-label.

How we checked

• Read in full on October 17, 2026: INNOVA, Bested, Vatvani (its 2026 correction does not change the side-effect figures), Younger 2009 and 2014, Gouda and Bolton. Read as abstracts: the Danish trial, Hegde and the Cochrane Crohn's review.^4,8,9,16
• INNOVA figures were checked against its event tables; serious adverse events come from the table, not the discussion. Its supplementary table S12 was not available.