What Is the Herxheimer Reaction? Causes & Symptoms

The Jarisch-Herxheimer reaction is a short-lived inflammatory flare that strikes within hours of the first antibiotic dose for certain bacterial infections, most classically syphilis, Lyme disease, leptospirosis, and relapsing fever. Rather than being a side effect of the drug itself, the reaction is driven by the immune system’s response to a sudden flood of bacterial debris released as the antibiotic kills large numbers of organisms at once. The experience can feel alarming, but in most cases it resolves on its own within a day, and recognizing it for what it is prevents unnecessary panic or, worse, stopping the antibiotic treatment that is actually working.

How the Reaction Was First Identified

The reaction gets its name from two dermatologists working independently in the late nineteenth and early twentieth centuries. Adolf Jarisch, an Austrian physician, documented a pronounced worsening of skin lesions and systemic symptoms in syphilis patients immediately after they were treated with mercury, the standard antisyphilitic therapy of his era.¹ Karl Herxheimer, a German dermatologist, described the same phenomenon around the same time. For decades, clinicians thought this flare was a drug allergy. It was only later, as penicillin replaced mercury for syphilis and the same reaction kept appearing, that researchers realized the drug itself was not the culprit. The reaction happened because the treatment was killing spirochetes, the corkscrew-shaped bacteria responsible for syphilis, and the body was reacting to what those dying organisms left behind.

What Causes It

The core mechanism centers on the immune system’s pattern-recognition machinery. When an antibiotic rapidly kills a large number of bacteria, the organisms fall apart and shed fragments, including lipoproteins and other structural components embedded in their outer membranes. These fragments are powerful immune triggers. Research in both mouse models and human blood samples has shown that receptors on immune cells detect these bacterial remnants and launch an aggressive inflammatory response, effectively a localized cytokine storm.² The specific receptors involved vary somewhat between species, but the end result is the same: an abrupt surge in pro-inflammatory signaling molecules.

Studies measuring blood samples during the reaction have found dramatic spikes in several of these molecules. In patients with relapsing fever treated with penicillin, plasma concentrations of tumor necrosis factor (TNF), interleukin-6, and interleukin-8 rose roughly seven-, six-, and fourfold above their already elevated admission levels, and the timing of those spikes coincided precisely with the worst symptoms.³ Similar cytokine elevations have been documented in the reaction during treatment of Q fever endocarditis and leptospirosis.⁴ These cytokines drive fever, cause blood vessels to dilate (lowering blood pressure), recruit more immune cells, and can even activate the coagulation system, which may explain why rare severe cases involve bleeding complications.⁵

The essential point is that the reaction is not caused by the antibiotic’s chemistry. It is caused by the immune system’s response to bacterial death. Switching to a different antibiotic from the same class would not prevent it, and the same reaction can occur with completely different drug families as long as the bacteria die quickly enough to release their components all at once.

Typical Symptoms and Timing

The hallmark of the Herxheimer reaction is its speed. Symptoms generally appear within the first 24 hours after the initial antibiotic dose. A secondary analysis of a clinical trial involving 249 adults treated for early syphilis with benzathine penicillin found that among those who experienced the reaction, the median time to symptom onset was about five hours, and for the large majority, symptoms began within 12 hours of treatment.⁶ The median duration was roughly 13 hours, though it ranged from a few hours to a full day.

The classic symptoms include shaking chills, a rise in temperature, and intensification of any existing skin rashes.⁷ In the syphilis trial, the most commonly reported individual symptoms were muscle aches (about half of those affected), chills (about 46%), weakness (39%), and feverishness (about 36%).⁶ Many people also report headache, sweating, and general malaise. If you already have a rash from syphilis, it may temporarily flare up or new lesions can appear before resolving.⁸

In most cases, the whole episode passes within a day and leaves no lasting harm. The reaction is self-limiting. But knowing the expected timeline matters: if you develop a fever six hours after your first antibiotic injection for syphilis, that is almost certainly the Herxheimer reaction running its course. A fever that develops three or four days after the injection is a different story and warrants a call to your doctor.

Which Infections Trigger It

The Herxheimer reaction is best studied and most common in spirochetal infections, a group of diseases caused by spiral-shaped bacteria. Syphilis is the classic example. In the early-syphilis trial cited above, about 24% of participants experienced one or more symptoms.⁶ Rates are higher in secondary syphilis (when the bacterial load is typically at its peak) and lower in late-stage disease.

Lyme disease, caused by the spirochete Borrelia burgdorferi, can also provoke the reaction, though it is generally thought to be less common than in syphilis. Case reports describe it appearing during doxycycline treatment, sometimes with a delayed or prolonged course. One reported case involved a patient whose symptoms worsened markedly on the seventh day of doxycycline therapy for Lyme arthritis, with severe joint pain, muscle aches, and elevated inflammatory markers.⁹ That later onset is unusual and underscores that the reaction does not always follow the textbook timeline seen in syphilis.

Leptospirosis is another well-documented trigger. A retrospective study of 262 confirmed leptospirosis patients treated with amoxicillin found an overall Herxheimer reaction rate of about 21%, with striking variation between populations: 12% in New Caledonia versus 44% in Futuna.¹⁰ Relapsing fever, caused by various Borrelia species transmitted by ticks or lice, has the highest reported rates of all, with some older studies documenting the reaction in the majority of treated patients.¹¹

Beyond spirochetal infections, case reports have documented the reaction in Q fever and a handful of other bacterial diseases, though the evidence base is much thinner for these conditions.

Risk Factors That Make It More Likely

Not everyone treated for these infections experiences a Herxheimer reaction, so researchers have tried to pin down who is most vulnerable. Two patterns stand out across the literature.

The first is bacterial burden. The reaction is more common in stages of disease where there are more organisms circulating. In syphilis, having secondary-stage disease (the stage with the most widespread spirochete dissemination) nearly tripled the odds of a reaction compared to primary syphilis.⁶ In leptospirosis, starting antibiotics early (within three days of symptom onset, when bacterial loads are presumably still high) roughly doubled the risk compared to later initiation.¹⁰ The infecting bacterial strain also mattered in the leptospirosis study: infection with Leptospira interrogans serogroup Australis carried a higher risk than other strains.

The second pattern involves immune status. In the syphilis trial, living with HIV was associated with roughly half the odds of experiencing the reaction, suggesting that an intact and vigorous immune system may actually predispose someone to a stronger inflammatory response when bacterial components flood the bloodstream.⁶ This makes immunological sense: if the cytokine storm is what drives the reaction, a suppressed immune system might simply be unable to mount as large a response.

How Antibiotic Choice Affects the Reaction

Since the reaction depends on rapid bacterial killing, the pharmacology of the antibiotic matters. Bactericidal drugs, which actively destroy bacteria, are more likely to trigger it than bacteriostatic drugs, which merely stop bacteria from multiplying and leave the immune system to mop them up gradually. Laboratory experiments with leptospires showed that the beta-lactam antibiotics amoxicillin and ceftriaxone killed all the organisms within three hours, a time frame that matches when the reaction is typically observed in patients. By contrast, azithromycin and doxycycline failed to kill the bacteria within three or six hours, though they were effective at 24 hours.²

This does not mean slower-acting antibiotics prevent the reaction entirely. Doxycycline is the standard treatment for Lyme disease, and Herxheimer reactions still occur during doxycycline therapy, as the Lyme arthritis case report illustrates.⁹ But the speed and magnitude of the kill likely influence how dramatic the reaction is. A fast-acting bactericidal drug releases a larger bolus of bacterial fragments in a shorter window, which could explain why the reaction tends to be more predictable and intense with penicillin for syphilis than with doxycycline for Lyme.

When the Reaction Gets Dangerous

For most people, the Herxheimer reaction is uncomfortable but harmless. The scenario changes in certain clinical situations.

Pregnancy is the most well-recognized concern. A study of pregnant women treated for syphilis found that the most common Herxheimer symptoms were fever (73%), uterine contractions (67%), and decreased fetal movement (67%). Transient worrisome fetal heart-rate patterns were detected in some monitored patients. In a pregnancy where the fetus is already affected by congenital syphilis, the reaction can provoke preterm labor, preterm delivery, or even fetal death.¹² This does not mean pregnant women should skip treatment; untreated syphilis in pregnancy carries its own devastating risks. But it does mean treatment should be given in a monitored setting where the fetus can be watched closely.

Outside of pregnancy, case reports describe rare severe presentations including drops in blood pressure, worsening of liver and kidney function, acute respiratory distress, injury to the heart muscle, meningitis, altered consciousness, seizures, and strokes.⁷ In leptospirosis, the coagulation-system activation driven by the cytokine surge may contribute to pulmonary hemorrhage, one of the most feared complications of severe disease.⁵ These severe outcomes are uncommon, but they are the reason why patients with high-risk infections, particularly leptospirosis with organ involvement or syphilis in pregnancy, are often started on antibiotics in a hospital.

Telling It Apart From a Drug Allergy

One of the most common clinical mistakes with the Herxheimer reaction is misidentifying it as an allergic reaction to the antibiotic. The confusion is understandable: both can produce fever, skin changes, and general distress shortly after a drug is given. But the distinction has real consequences. If a clinician mistakes the Herxheimer reaction for penicillin allergy, the patient may be switched to a less effective antibiotic or have penicillin permanently and incorrectly added to their allergy list. Case reports have documented exactly this scenario.¹³

A few features help distinguish the two. The Herxheimer reaction tends to produce flu-like symptoms (chills, muscle aches, fever, malaise) and a flare of existing skin lesions, while a true drug allergy more often involves hives, itching, swelling, and in severe cases, difficulty breathing or anaphylaxis. The Herxheimer reaction resolves within hours to a day even if the antibiotic is continued, whereas an allergic reaction typically worsens with continued exposure. The clinical context matters too: a patient being treated for their first dose of a spirochetal infection is a prime candidate for a Herxheimer reaction, and clinicians who are aware of this possibility can counsel patients in advance.

Can You Prevent It

Given that the reaction is driven by inflammatory cytokines, researchers have explored whether anti-inflammatory pretreatment could blunt it. Corticosteroids (like prednisone) are sometimes used in practice, particularly before treating neurosyphilis or in pregnant patients, to try to dampen the inflammatory surge. However, the evidence that this actually prevents the reaction is limited and inconsistent. In relapsing fever, experimental work with anti-TNF antibodies showed some suppression of the cytokine peaks, which is consistent with the idea that TNF is a central driver.⁷ But anti-TNF therapy is not a practical preventive measure for routine clinical use.

In practice, management is usually supportive. Patients are warned ahead of time that they may feel worse before they feel better. Over-the-counter fever reducers and adequate fluids get most people through the episode. The critical message that clinicians emphasize is: do not stop the antibiotic. The reaction means the drug is doing its job, and halting treatment leaves the underlying infection in place.

The Herxheimer Reaction in Alternative Medicine

If you search online for “Herxheimer reaction” or “herxing,” you will quickly encounter claims far beyond what the medical literature supports. In some alternative and complementary health communities, the term has been adopted to describe any worsening of symptoms during various detox programs, herbal protocols, or dietary changes. The reasoning usually goes: if you feel worse, it means toxins are being released as your body heals, and you should push through.

This is a significant misapplication of the concept. The real Herxheimer reaction is specific to the killing of certain microorganisms by antimicrobial therapy, documented by measurable spikes in inflammatory cytokines, and resolves within hours to a day. It has a defined mechanism, occurs in defined clinical contexts, and has been studied for over a century. Feeling generally unwell during a juice cleanse or a change in supplements is not the same phenomenon, and the logic of “feeling worse means getting better” can be genuinely dangerous if it leads someone to ignore symptoms that actually need medical attention. A real Herxheimer reaction is a recognized medical event managed under clinical supervision; co-opting the term to validate any negative symptom as a healing crisis strips it of meaning and can delay proper care.