Ivermectin Side Effects: Why the Same Drug Shows 2.8% Itching in One Use and 27.5% in Another

Search for ivermectin side effects and you get one list, as though the drug behaves the same way in everyone who takes it. The FDA label does not report it that way. It reports two separate sets of figures, one for each approved use, and they are not remotely alike.

Itching appears in 2.8% of patients treated for strongyloidiasis. In patients treated for onchocerciasis it appears in 27.5%, roughly ten times as often, from the same drug at a similar dose. Fever does not appear on the strongyloidiasis list at all and reaches 22.6% in onchocerciasis.

That gap is the single most useful thing to understand about ivermectin's safety profile, and it is not a contradiction in the data. This guide explains what causes it, sets out both tables in full, and covers the one situation where ivermectin is dangerous.

The short answer

Ivermectin taken for an intestinal worm infection is well tolerated: the common side effects sit between 0.9% and 2.8%, most of them mild and short-lived. Ivermectin taken for onchocerciasis produces reactions in a fifth to a quarter of patients, but the label attributes most of those reactions to the body responding to large numbers of dying parasites rather than to the drug. The most serious risk is in people heavily infected with the Loa loa parasite, found in parts of West and Central Africa, who can very rarely develop a fatal encephalopathy. The current label also warns of neurotoxicity in people without Loa loa: changes in consciousness from drowsiness to coma, confusion, and deaths, reported at recommended doses as well as in overdose, with no frequency given.

How ivermectin works, and why people tolerate it

Ivermectin belongs to the avermectin class. The label explains that these drugs bind to glutamate-gated chloride channels in the nerve and muscle cells of invertebrates. Chloride floods in, the cells stop firing, and the parasite is paralysed and dies.

Two facts explain why that does not happen to the person taking it. Some mammals do not have glutamate-gated chloride channels at all, and avermectins have low affinity for the chloride channels that mammals do have. The label also notes that ivermectin does not readily cross the blood-brain barrier in humans. Those are the reasons the side effect lists below are short when the parasite load is small.

Why the same drug produces two different pictures

Ivermectin paralyses and kills parasites. When the parasite burden is small, as in a routine intestinal worm infection, very little dies at once and the body barely notices.

Onchocerciasis is different. A person with river blindness can carry enormous numbers of microfilariae in the skin and eyes. One dose kills a large share of them at the same time, and the immune system reacts to that die-off with itching, rash, fever, swollen lymph nodes and joint pain.

That cluster has a name: the Mazzotti reaction. The label describes these reactions as probably due to allergic and inflammatory responses to the death of microfilariae, and notes they were first seen with older microfilaricidal drugs such as diethylcarbamazine. It is the reason the onchocerciasis table looks alarming next to the strongyloidiasis one.

The practical implication is that the number you should care about depends on what you are being treated for.

Side effects when treating strongyloidiasis

These are the reactions reported as possibly, probably or definitely related to ivermectin across four clinical studies covering 109 patients given one or two doses of 170 to 200 mcg/kg.

Side effect	Rate	Category
Dizziness	2.8%	Nervous system
Pruritus (itching)	2.8%	Skin
Diarrhoea	1.8%	Gastrointestinal
Nausea	1.8%	Gastrointestinal
Fatigue or weakness	0.9%	Whole body
Abdominal pain	0.9%	Whole body
Loss of appetite	0.9%	Gastrointestinal
Constipation	0.9%	Gastrointestinal
Vomiting	0.9%	Gastrointestinal
Drowsiness	0.9%	Nervous system
Vertigo	0.9%	Nervous system
Tremor	0.9%	Nervous system
Rash	0.9%	Skin
Urticaria (hives)	0.9%	Skin

Nothing on that list reaches 3%. With 109 patients, 0.9% is a single person. For context, the label notes that in comparative trials patients on ivermectin reported more abdominal distension and chest discomfort than patients on albendazole, but tolerated ivermectin better than thiabendazole.

Laboratory findings in the same 109 patients, regardless of whether the drug caused them: raised ALT or AST in 2%, and a fall in white cell count in 3%. One patient had both leukopenia and anaemia.

The label is explicit that the Mazzotti-type and eye reactions seen in onchocerciasis would not be expected in strongyloidiasis patients. If you are being treated for an intestinal worm, this table is your table.

Side effects when treating onchocerciasis

These are the Mazzotti reactions recorded during the first four days after treatment, in trials covering 963 adults given 100 to 200 mcg/kg.

Reaction	Rate
Pruritus (itching)	27.5%
Skin involvement: swelling, papular or pustular rash, hives	22.7%
Fever	22.6%
Inguinal (groin) lymph node tenderness	13.9%
Inguinal lymph node enlargement	12.6%
Axillary (armpit) lymph node enlargement	11.0%
Joint pain or synovitis	9.3%
Cervical (neck) lymph node enlargement	5.3%
Axillary lymph node tenderness	4.4%
Other lymph node enlargement	3.0%
Other lymph node tenderness	1.9%
Cervical lymph node tenderness	1.2%

Separately, reactions judged related to the drug itself in at least 1% of those 963 patients were: tachycardia 3.5%, peripheral oedema 3.2%, facial oedema 1.2% and orthostatic hypotension 1.1%.

The label draws a careful distinction on headache and muscle pain. As drug-related events they were rare, 0.2% and 0.4%. As events reported during the trials regardless of cause, they were the most common of all, 22.3% and 19.7%. Same trials, same patients, two very different numbers depending on whether you ask "did this happen?" or "was the drug responsible?"

Laboratory findings: eosinophilia 3%, raised haemoglobin 1%. The label also reports a similar safety profile in an open study of children aged 6 to 13.

The two uses next to each other

Reaction	Strongyloidiasis (n=109)	Onchocerciasis (n=963)
Itching	2.8%	27.5%
Rash or skin involvement	0.9% rash, 0.9% hives	22.7% (swelling, rash or hives)
Fever	Not on the list	22.6%
Lymph node enlargement	Not on the list	Up to 12.6% at one site
Joint pain	Not on the list	9.3%

Read that as a measure of parasite burden rather than of drug toxicity. Ivermectin is not more toxic in onchocerciasis. It is killing far more organisms at once, and the body is responding to the debris. The two sets of figures come from separate trials in different infections, so they show the size of the difference rather than an exact ratio.

How severe Mazzotti reactions are managed

The label is candid that the treatment of severe Mazzotti reactions has never been tested in controlled trials. What it records is what has been used: oral fluids, lying down, intravenous saline and parenteral corticosteroids for postural hypotension, and antihistamines or aspirin for most mild to moderate cases.

It also flags one group at higher risk. People with hyperreactive onchodermatitis, known as sowda, may be more likely than others to have severe reactions after microfilaricidal treatment, especially swelling and a flare of the skin disease.

Eye effects, and a result worth reading carefully

Onchocerciasis affects the eyes, so eye findings were examined in all 963 patients before treatment, at day 3, and at months 3 and 6. The label reports that changes were mainly a deterioration from baseline at day 3, and that most either returned to baseline or improved beyond it by months 3 and 6.

Finding	Day 3	Month 3	Month 6
Limbitis, ivermectin	5.5%	4.8%	3.5%
Limbitis, placebo	6.2%	9.9%	9.4%
Punctate opacity, ivermectin	1.8%	1.8%	1.4%
Punctate opacity, placebo	2.0%	6.4%	7.2%

Look at the placebo rows. Eye findings in the untreated group got worse over six months while the treated group improved. At month 6, limbitis affected 3.5% of treated patients against 9.4% of untreated ones.

So the eye reactions after treatment are short-term and the untreated disease is what damages sight. This is one of the few places in a drug label where the placebo column tells the more important story.

The label also lists eye effects that occur because of the disease itself but have been reported after treatment too: abnormal sensation in the eyes, eyelid swelling, anterior uveitis, conjunctivitis, limbitis, keratitis and chorioretinitis or choroiditis. These have rarely been severe or associated with vision loss, and have generally resolved without corticosteroid treatment.

The highest-risk group: Loa loa

This is the part of the label that carries the most weight.

People with onchocerciasis who are also heavily infected with Loa loa can rarely develop a serious or even fatal encephalopathy, either spontaneously or after treatment with an effective microfilaricide. The label lists the signs reported in these patients: pain including neck and back pain, red eye, conjunctival haemorrhage, breathlessness, loss of bladder or bowel control, difficulty standing or walking, mental status changes, confusion, lethargy, stupor, seizures or coma. It describes the syndrome as seen very rarely after ivermectin.

The label's instruction is specific. Anyone who warrants treatment with ivermectin for any reason and has had significant exposure to Loa loa endemic areas of West or Central Africa should be assessed for loiasis before treatment and followed up carefully afterwards.

If you are being treated for an intestinal worm infection in the United States, the UK or Australia and have no history of living in or travelling to those regions, this warning does not describe your situation. If you do have that history, tell your doctor before taking ivermectin for anything.

Neurotoxicity without Loa loa

Merck's current prescribing information (revised September 2024) adds a warning that earlier versions did not carry. Neurotoxicity with ivermectin, including altered consciousness of varying severity (drowsiness, stupor and coma), confusion, disorientation and death, has been reported in people without onchocerciasis and in people with onchocerciasis who did not have Loa loa. The label says these reactions generally resolved with supportive care and stopping ivermectin, and that such cases have been reported at the recommended dosage as well as in overdose.

These come from reports rather than trials, so the label gives no rate. The practical point is simple: confusion, unusual drowsiness or any change in alertness after taking ivermectin is a reason to stop it and get medical help, wherever you live and whatever it was taken for.

Post-marketing reports

These come from voluntary reports after approval, so no rate can be calculated and a causal link is not established. For all indications the label lists hypotension (mainly orthostatic), worsening of bronchial asthma, toxic epidermal necrolysis, Stevens-Johnson syndrome, seizures, hepatitis, raised liver enzymes, raised bilirubin and neurotoxicity. For onchocerciasis it adds conjunctival haemorrhage.

Toxic epidermal necrolysis and Stevens-Johnson syndrome are severe reactions of the skin and mucous membranes. Any spreading rash with blistering, or sores in the mouth or eyes, needs urgent medical attention whatever medicine you are taking. Anyone with asthma should know that worsening of it has been reported.

Dosing by weight, and why it matters for side effects

The approved dose differs by indication, which is part of why the side effect profiles differ. Both are single oral doses of 3 mg tablets, calculated by body weight.

Strongyloidiasis: about 200 mcg/kg.

Body weight	3 mg tablets
15 to 24 kg	1
25 to 35 kg	2
36 to 50 kg	3
51 to 65 kg	4
66 to 79 kg	5
80 kg and over	Calculated at 200 mcg/kg

Onchocerciasis: about 150 mcg/kg.

Body weight	3 mg tablets
15 to 25 kg	1
26 to 44 kg	2
45 to 64 kg	3
65 to 84 kg	4
85 kg and over	Calculated at 150 mcg/kg

The weight bands differ between the two indications, so use the table that matches your treatment. For strongyloidiasis, additional doses are generally not needed, but follow-up stool tests are. For onchocerciasis, ivermectin does not kill the adult worms, so retreatment is usually required; the label notes mass treatment programmes most often use 12 month intervals and that individual patients may be retreated at intervals as short as 3 months.

An underdose leaves parasites alive and an overdose raises the side effect rate without improving the kill. Our ivermectin dosage guide works through the calculation, and the high dose guide covers doses beyond the label.

Food: why the label says take it on an empty stomach

The label tells patients to take ivermectin on an empty stomach with water. The reason is in its pharmacokinetics section. In a study of healthy volunteers, a 30 mg dose taken after a high fat meal containing 48.6 g of fat produced roughly 2.5 times the bioavailability of the same dose taken fasting.

That is the opposite of how food affects most of the drugs people compare it with. More absorption sounds useful, but the dosing tables and the side effect data were built on fasting administration. Taking the tablet with a fatty meal means taking, in effect, a larger dose than the one that was studied.

How long ivermectin stays in the body

Measure	From the label
Time to peak blood level	About 4 hours after a 12 mg dose, fasting
Plasma half-life	About 18 hours
Metabolism	In the liver, mainly by CYP3A4
Excretion	Almost entirely in the faeces over about 12 days; under 1% in urine

The practical point is that most side effects, where they occur, cluster in the first few days. The Mazzotti reactions in the onchocerciasis trials were recorded in the first four days after treatment.

Drug interactions

The label lists one interaction from post-marketing reports: raised INR, meaning blood that takes longer to clot, has rarely been reported when ivermectin was taken with warfarin. Anyone on warfarin should have their INR checked around a course of ivermectin.

Because ivermectin is metabolised mainly by CYP3A4, drugs that strongly affect that enzyme could in principle change its levels. The label reports laboratory studies showing ivermectin itself does not significantly inhibit CYP3A4, CYP2D6, CYP2C9, CYP1A2 or CYP2E1 at clinically relevant concentrations, so it is unlikely to raise the levels of other drugs through those routes.

Pregnancy, breastfeeding, children and older adults

Pregnancy. The label says ivermectin should not be used during pregnancy, because safety in pregnancy has not been established. It was teratogenic in mice, rats and rabbits, at doses at or near those toxic to the mother, and there are no adequate and well-controlled studies in pregnant women. It is pregnancy category C on this label.

Breastfeeding. Ivermectin passes into breast milk in low concentrations. The label says mothers who intend to breastfeed should be treated only when the risk of delaying their treatment outweighs the possible risk to the baby.

Children. Safety and effectiveness have not been established below 15 kg, which is why the dosing tables start there.

Older adults. The trials did not include enough people over 65 to tell whether they respond differently, and the label advises caution because liver, kidney and heart function decline with age.

Weakened immune systems. In immunocompromised patients, including people with HIV, strongyloidiasis may need repeated courses, sometimes at 2 week intervals, and cure may not be achievable.

Veterinary ivermectin: what the label records

The overdose section of the human label is built partly on accidental exposure to veterinary products. After ingestion, inhalation, injection or skin exposure to unknown amounts of veterinary ivermectin, the most commonly reported effects were rash, swelling, headache, dizziness, weakness, nausea, vomiting and diarrhoea. Seizure, loss of coordination, breathlessness, abdominal pain, pins and needles, hives and contact dermatitis were also reported.

Veterinary formulations differ in concentration and ingredients from human tablets, which is what makes the dose unknowable. Our guide to horse paste explains the difference.

What raises your risk of side effects

• A heavy parasite burden. The strongest predictor of a Mazzotti reaction is how many organisms are dying at once.
• Loa loa co-infection. The highest-risk group, covered above.
• Hyperreactive onchodermatitis (sowda). Flagged by the label as more likely to produce severe reactions.
• Taking it with a high fat meal. About 2.5 times the absorption.
• Liver disease. Ivermectin is metabolised in the liver, and the label records raised liver enzymes in 2% of strongyloidiasis patients.
• Warfarin. Rare reports of raised INR.
• Higher than labelled doses. Side effects rise; effectiveness against the target parasite does not.
• Pregnancy. The label advises against use because safety has not been established.

When to seek medical help

Most reactions to ivermectin are mild and settle within a few days. Seek care promptly for any of the following.

• Confusion, severe drowsiness, seizures, difficulty standing or any change in mental state. This applies to everyone, and particularly to anyone who has lived in or travelled to West or Central Africa.
• A spreading rash with blistering, or sores in the mouth or eyes.
• Yellowing of the skin or eyes, dark urine, or pain under the right ribs.
• Sudden vision change, red eye or eye pain.
• Difficulty breathing, worsening asthma, or swelling of the face, lips or throat.
• Fainting or dizziness on standing that does not settle.
• Fever that keeps climbing rather than settling after a few days.

What this label does not cover

The figures above come from trials in strongyloidiasis and onchocerciasis, the two indications this label approves. They do not describe what happens at doses well above 200 mcg/kg, in repeated courses over long periods, or in conditions outside those two.

We say that plainly because side effect rates do not transfer between uses. If you are reading about ivermectin for something this label does not cover, these percentages are not the right guide to what you should expect, and no reliable table exists for a use that has not been studied that way.

Related reading

• What ivermectin is and how it works
• Ivermectin dosage by weight
• Ivermectin for scabies
• Ivermectin compared with mebendazole
• Intestinal parasites: symptoms and treatment

Sources

• STROMECTOL (ivermectin) tablets, US prescribing information: Clinical Pharmacology, Warnings, Precautions, Adverse Reactions, Overdosage, and Dosage and Administration. DailyMed
• Strongyloidiasis figures: four clinical studies, 109 patients, one or two doses of 170 to 200 mcg/kg.
• Onchocerciasis figures: clinical trials in 963 adults, 100 to 200 mcg/kg, Mazzotti reactions recorded during the first 4 days after treatment; eye findings at baseline, day 3, month 3 and month 6.

This article is for information only and is not medical advice. Ivermectin is a prescription medicine. Talk to a doctor or pharmacist before starting, stopping or changing any treatment.