Naltrexone Drug Interactions (Including Low-Dose LDN)

Low-dose naltrexone (LDN) is naltrexone hydrochloride, the same molecule sold at 50 mg under the brand names ReVia and Depade for alcohol and opioid use disorder, compounded instead at 1.5 to 4.5 mg and taken once nightly. It is not a different drug and it is not FDA-approved at this dose range; every LDN prescription is an off-label, compounded use of an approved active ingredient. This distinction matters for interactions, because the FDA label, black box warning, and pharmacokinetic data that exist all come from studies of the 50 mg product, and the low-dose regimen has not been separately reviewed by the FDA.

There is one absolute, non-negotiable interaction: naltrexone at any dose, including 1.5 mg, can precipitate acute opioid withdrawal in a person who has recently used an opioid agonist. Beyond that single hard rule, the rest of the LDN interaction picture is thin. No large, dedicated interaction trials exist for the 1.5 to 4.5 mg range; most of what is written about LDN and immunosuppressants, thyroid drugs, or antidepressants is extrapolated from small pilot studies, case series, and pharmacologic reasoning rather than direct interaction data. This article separates what the naltrexone label and basic pharmacology actually establish from what is plausible but unproven, and flags where a precise number should not be trusted without checking the primary paper.

The one contraindication that cannot be missed

No patient currently taking, or recently exposed to, an opioid agonist should start LDN. Naltrexone binds mu-opioid receptors with high affinity; in someone who is opioid-dependent, that binding displaces the opioid and triggers withdrawal, which can begin within minutes and includes severe cramping, vomiting, diarrhea, tachycardia, and agitation. The FDA label for naltrexone 50 mg tablets instructs that patients be opioid-free, confirmed where possible by urine drug screen, for a minimum of 7 to 10 days before starting naltrexone, and this guidance is the basis for LDN washout practice as well (FDA naltrexone label).

Longer-acting opioids need a longer washout than the label's general figure suggests. Methadone and other long-acting or extended-release opioids are commonly held for 10 to 14 days in clinical practice because residual receptor occupancy persists longer than with short-acting agents; this extended window is standard prescribing caution rather than an FDA-specified number, and prescribers should confirm current institutional protocol.

Partial agonists and "hidden" opioids carry the same risk and are the ones most often missed:

• Buprenorphine products (Suboxone, Sublocade) bind mu-receptors tightly and will be displaced by naltrexone.
• Tramadol is frequently perceived as a mild analgesic, but its active metabolite has meaningful mu-agonist activity, so it should be treated as a full opioid for washout purposes.
• Kratom (mitragynine) and tianeptine act on opioid receptors and are not reliably detected on standard urine screens, which makes patient disclosure essential rather than optional.
• Opioid-containing cough syrups (codeine, hydrocodone) and supratherapeutic loperamide use also carry theoretical risk and should be disclosed before starting LDN.

Does LDN interact with immunosuppressants?

LDN is proposed to work partly through endorphin upregulation and toll-like receptor 4 (TLR4) antagonism on glial and immune cells, a mechanism distinct from how corticosteroids, biologics, or antimetabolite immunosuppressants work. Because the mechanisms differ, most clinicians treat LDN and immunosuppressive therapy as pharmacodynamically separate rather than competing.

• Corticosteroids. No pharmacokinetic interaction is established. Some case series and small trials in inflammatory bowel disease describe LDN being used alongside low-dose corticosteroids without a described adverse signal, but this is observational and clinician-reported rather than a controlled interaction study, and any specific percentage of concurrent use should be verified against the original paper before being repeated as fact.
• Methotrexate, azathioprine, mycophenolate. No published interaction is documented at the pharmacokinetic level. Because methotrexate and azathioprine both carry independent hepatotoxicity risk, and naltrexone's own hepatic safety margin at low doses is not fully characterized in this population, periodic liver function testing is reasonable when LDN is added to these regimens.
• Biologic agents (anti-TNF drugs, rituximab). These act on specific cytokines or cell-surface markers upstream or downstream of LDN's proposed TLR4 mechanism. Small open-label studies in Crohn's disease report no obvious pharmacokinetic alteration of either drug when combined, but sample sizes are small and this should be treated as reassuring rather than conclusive.
• Calcineurin inhibitors (tacrolimus, cyclosporine). Both are CYP3A4 substrates, and naltrexone is a minor CYP3A4 substrate, so competitive inhibition at LDN doses is theoretically possible but not documented. Because both calcineurin inhibitors and naltrexone can affect the liver, periodic liver function monitoring is a reasonable precaution when they are combined.

Does combining LDN with other drugs raise liver injury risk?

The FDA's hepatotoxicity black box warning for naltrexone comes from studies of the 300 mg/day dose used in early obesity research, a dose roughly 65 to 200 times higher than a typical LDN dose (FDA naltrexone label). The label states that naltrexone "has the capacity to cause hepatocellular injury when given in excessive doses," and that framing, tied explicitly to high-dose exposure, is the FDA's own language rather than an LDN-specific finding.

Direct hepatotoxicity data for the 1.5 to 4.5 mg range is limited to small observational cohorts and patient-reported surveys, and any specific incidence figure (for example, a particular percentage of patients with elevated liver enzymes) should be checked against the original source before being treated as an established rate; the underlying studies were not designed as controlled safety trials and their numbers vary between publications.

What is reasonable regardless of the exact incidence: LDN should be used cautiously, with baseline and periodic liver function testing, in anyone also taking another hepatotoxic drug. Relevant co-medications include:

• Regular acetaminophen use above roughly 2 g/day
• Statins metabolized through CYP3A4 (atorvastatin, simvastatin)
• Methotrexate and thiopurines (azathioprine, 6-mercaptopurine)
• Valproic acid, particularly in patients with underlying mitochondrial disease
• Calcineurin inhibitors, as above

Naltrexone itself is contraindicated in acute hepatitis and liver failure per the FDA label, independent of dose.

Does LDN change how much thyroid medication you need?

There is no direct pharmacokinetic interaction between naltrexone and levothyroxine (T4), liothyronine (T3), or desiccated thyroid extract. The interaction that matters here, if it occurs, is indirect and immunological. Some small studies and case series in Hashimoto's thyroiditis describe reduced thyroid peroxidase antibody levels after several months of LDN use, with a proposed mechanism of reduced autoimmune thyroid inflammation. If thyroid inflammation genuinely decreases, the gland's own hormone output could recover, which would make a previously stable levothyroxine dose too high over time.

This mechanism is plausible but not established as a reliable, dose-predictable effect, and the antibody studies behind it are small. The practical implication does not depend on resolving that uncertainty: anyone with autoimmune thyroid disease who starts LDN should have TSH and free T4 checked roughly every 6 to 8 weeks in the months after starting, so a dose adjustment is not missed. Desiccated thyroid products complicate this because their fixed T4:T3 ratio makes fine dose titration harder than with synthetic T4 alone.

Antidepressants, benzodiazepines, and neuropathic pain drugs

Naltrexone is metabolized mainly through CYP2D6 reduction to its active metabolite, 6-beta-naltrexol, with a minor CYP3A4 contribution. SSRIs and SNRIs such as fluoxetine, paroxetine, and duloxetine are CYP2D6 inhibitors, so in principle they could raise naltrexone plasma levels. At the 50 mg standard dose this interaction would be worth watching; at LDN doses, the absolute quantity of drug involved is small enough that the theoretical shift is unlikely to be clinically meaningful, and small pilot trials of LDN in fibromyalgia allowed concurrent SSRI use without a reported interaction signal, though these trials were not designed to detect a pharmacokinetic interaction specifically.

Bupropion is combined with naltrexone at full dose in the FDA-approved product Contrave (naltrexone 32 mg / bupropion 360 mg daily) for weight management. LDN's naltrexone component is far lower than the Contrave dose, and no dose adjustment of either drug is required when LDN and bupropion are prescribed separately.

Benzodiazepines have no known pharmacodynamic or pharmacokinetic interaction with naltrexone; naltrexone does not act on GABA receptors. Gabapentin and pregabalin are renally cleared with no meaningful hepatic metabolism, so they do not compete with naltrexone's CYP pathways and are commonly co-prescribed for fibromyalgia and neuropathic pain without a documented interaction.

GLP-1 medications, diabetes drugs, and alcohol

Semaglutide and tirzepatide are metabolized by proteolytic cleavage and fatty acid processing, entirely independent of the CYP enzymes involved in naltrexone metabolism, so no pharmacokinetic interaction is expected. Metformin is renally cleared. Sulfonylureas are CYP2C9 substrates, a different pathway than naltrexone's. The one soft consideration: if LDN meaningfully improves insulin sensitivity, an effect that is proposed but not established, patients on insulin or sulfonylureas could theoretically see lower glucose readings, which supports routine glucose monitoring rather than a specific dose change.

Alcohol has no pharmacokinetic interaction with naltrexone; the two are cleared through different hepatic pathways. The practical concern is additive hepatic burden in someone with existing liver disease or heavy alcohol use, which is a reason for caution rather than an absolute contraindication at LDN doses. Standard-dose naltrexone (50 mg) is FDA-approved to reduce alcohol craving; LDN's brief and partial receptor blockade is not expected to reproduce that effect, and LDN should not be substituted for standard-dose naltrexone when alcohol use disorder is the treatment target.

Practical prescribing notes

LDN is compounded by 503A pharmacies because no FDA-approved low-dose product exists. Fillers, capsule design, and release characteristics vary between compounding pharmacies, and bioequivalence between pharmacies is not guaranteed, so a patient who switches pharmacies should be watched for a change in effect or side effects rather than assuming the new capsule behaves identically to the old one.

A common titration pattern is 1.5 mg nightly for two weeks, then 3 mg for two weeks, then a target dose of 4.5 mg, though exact titration should follow the prescribing clinician's plan rather than a fixed public schedule. Sleep disturbance and vivid dreams are commonly reported in the first one to two weeks and typically resolve; how common this is varies across the surveys that report it, and a specific percentage should not be treated as a fixed clinical fact.

Before starting LDN in a patient on other medications, a reasonable baseline includes liver function testing, confirmation of opioid-free status (ideally with a urine drug screen, understanding its limits for kratom and tianeptine), and a follow-up lab check at roughly 4 and 12 weeks if other hepatotoxic or immunosuppressive drugs are involved.

What is established, what is plausible, and what is not established

Established: Naltrexone at any dose, including LDN doses, can precipitate opioid withdrawal in someone with recent opioid exposure; this is stated in the FDA label and is the basis for mandatory opioid washout before starting LDN. Naltrexone's hepatotoxicity black box warning is based on 300 mg/day dosing, a dose far above the LDN range. LDN has no direct pharmacokinetic interaction with levothyroxine, GLP-1 agonists, gabapentin, pregabalin, benzodiazepines, or metformin.

Plausible but unproven: That LDN reduces thyroid antibody levels in Hashimoto's thyroiditis enough to require levothyroxine dose reduction in some patients. That LDN improves insulin sensitivity enough to matter for glucose monitoring. That SSRIs meaningfully raise naltrexone exposure at LDN doses. That co-prescribing LDN with corticosteroids or biologics is free of any additive immune effect over long follow-up.

Not established: A precise incidence rate for LDN-associated liver enzyme elevation, a precise percentage of LDN users who experience sleep disturbance, or a validated pharmacokinetic interaction study between LDN and any specific immunosuppressant. Where the source literature for these numbers could not be independently verified for this article, the specific figures have been removed or described qualitatively rather than stated as fact; anyone relying on a precise percentage from other sources should check the original study before citing it.

LDN interaction decision framework

Use this to decide what level of caution a given co-medication requires, and what the next concrete step is.

If a medication does not clearly fit one of these tiers, or if the evidence behind a specific claim in this article needs primary-source confirmation before a clinical decision is made, that is a reason to hold the LDN start and ask the prescriber to check the original literature rather than proceed on an assumed safety margin.