Fluoroquinolones showed no extra peripheral neuropathy versus cephalosporins in 3.9 million UK adults

Fluoroquinolones, the antibiotic family that includes ciprofloxacin and moxifloxacin, come with a long list of safety warnings. Regulators in the United States and Europe have cautioned for more than a decade that they can damage nerves. A UK study of 3.9 million adults has now tested that risk against a closer comparison, another antibiotic, and did not find a clear increase.

Researchers at the London School of Hygiene and Tropical Medicine compared people prescribed a fluoroquinolone with people prescribed a cephalosporin, another common class of antibiotics, and counted new diagnoses of peripheral neuropathy, nerve damage usually felt in the hands and feet, over the next two months. The rates were close to identical. The same analysis did pick up the expected rise in torn tendons, a known fluoroquinolone harm, which suggests the method could see a real problem when one was there. The paper, on systemic fluoroquinolones and risk of peripheral neuropathy, was led by Jeremy Brown with Ian Douglas as senior author and published on October 20 in Neurology.

Why fluoroquinolones carry a warning about nerve damage

Fluoroquinolones are broad-spectrum antibiotics, used for infections of the urinary tract, chest, gut, skin and joints. Over the years, case reports and observational studies tied them to an increased risk of rare adverse effects. Two of them matter here. One is tendon rupture; the other is peripheral neuropathy, damage to the nerves outside the brain and spinal cord that shows up as numbness or pain.

Regulators acted on that evidence. A German team describes repeated warnings from the European and US drug agencies, followed in 2019 by European restrictions with advice to use them only in specific cases such as complicated urinary tract infection. Australia goes further still; there, a review in Australian Prescriber notes, prescribing is restricted to infections resistant to all other recommended drugs.

The case on nerves has always rested on less than the case on tendons. The new paper’s authors note that the link was first seen in case reports, and that there have been few noninterventional studies validating it, meaning studies of real-world patients rather than single cases. They add that causality is unclear.

Two of those real-world studies did find a signal. A 2014 study of US men aged 45 to 80 found that current users of fluoroquinolones were at a higher risk of developing peripheral neuropathy, at nearly twice the rate of men not taking them. A 2019 study of UK general practice records, covering 1,338,900 adults without diabetes, found a rate about 47% higher during current use than with no antibiotic at all. Even that study judged the absolute risk to be low, at 2.4 cases per 10,000 patients per year of current use, and its authors calculated that about 152,000 ten-day courses would be needed to cause one extra case.

How the study compared fluoroquinolones with cephalosporins

Those earlier studies mostly set people taking a fluoroquinolone against people not taking one. That comparison is hard to make fair. Someone ill enough to need an antibiotic is not the same as someone who is well, and the infection, or what comes with it, can bring on symptoms of its own.

Brown and colleagues instead compared fluoroquinolone users with people given a different antibiotic, an approach epidemiologists call an active comparator. Both groups were sick enough to be prescribed something, so both had an infection to treat. They worked from 2 databases of UK electronic primary care records, known as CPRD Aurum and GOLD, covering April 1997 to the end of 2019. Everyone included was an adult without diabetes or previously recorded peripheral neuropathy, since diabetes is itself a leading cause of nerve damage. The outcome was a new diagnosis of peripheral neuropathy within 60 days of the prescription.

The numbers were large. Aurum contributed 2,933,982 individuals prescribed fluoroquinolones or cephalosporins and GOLD another 995,457. Close to two thirds were women, and the median age was around 50. At their first course, about a third had been given a fluoroquinolone: 1,002,004 people in one database and 381,734 in the other. A statistical weighting method balanced the measured characteristics of the two groups. The team then ran the analysis separately in each database and pooled the results.

The design had one more safeguard. Alongside nerve damage, the researchers tracked a positive control outcome of tendon rupture, a side effect that fluoroquinolones are well established to cause, with earlier work finding a two- to fourfold increased risk. If the method could not find more torn tendons, its silence on nerves would mean little. It works like testing a smoke alarm on a known fire before trusting it in a quiet room.

What the fluoroquinolone comparison found

The alarm went off where it should. Tendon ruptures were about twice as common in the 60 days after a fluoroquinolone as after a cephalosporin, both at a person’s first course and across all their courses.

For nerves, it stayed quiet. The estimated association was close to the null at first treatment episode, the authors write, meaning the two groups had nearly the same rate, and the same held across every course people received. Their conclusion is that the results “were not indicative of a substantive difference in risk of peripheral neuropathy” between the two antibiotic classes.

The size of the study gives that null some weight. The likely range around the main estimate runs from about a fifth lower to about a fifth higher. Across all courses it runs from slightly lower to a fifth higher. A doubling, the kind of increase the 2014 US study reported, sits well outside those ranges, and so does the 47% rise in the 2019 UK study.

The same team got a similar answer for the eye. In 2024, drawing on the same databases and on adults prescribed a fluoroquinolone or a cephalosporin, they found no clear link with uveitis, inflammation inside the eye, and an uncertain result for retinal detachment.

Why earlier fluoroquinolone studies and this one disagree

The choice of comparison seems to matter. In the 2019 UK study, no increased risk was observed with oral amoxicillin-clavulanate, a different antibiotic. That hints that part of the earlier signal may come from what goes with being treated for an infection, rather than from the fluoroquinolone itself.

A large German study published this year makes the same point from the other side. Working from health insurance claims, it found more nerve disorders after fluoroquinolones than after other antibiotics, an excess the authors say was mainly comprised of drug-induced polyneuropathy, a form of nerve damage. But the result shifted with the comparison. The excess appeared in comparison to amoxicillin and azithromycin and not against amoxicillin-clavulanic acid, and the German team concluded that results for each outcome varied by choice of active comparator.

None of this settles who is right. A comparator antibiotic can carry a small nerve risk of its own, which would hide a fluoroquinolone effect, and the infections treated with each class are not identical. What the new study adds is a very large comparison against one other commonly prescribed class of antibiotics, with a working check built in.

What the neuropathy result cannot rule out

The authors flag the main weakness themselves: potential residual confounding is a study limitation. That means differences between the two groups that the weighting could not capture, such as differences in the infections each drug was used to treat.

The study looked only at the 60 days after each prescription, so nerve symptoms that begin later, or that build up with repeated courses, fall outside it. The 2019 UK study reported raised risk lasting up to 180 days after a prescription. People with diabetes, who carry the highest background risk of neuropathy, were left out, so the result does not speak to them. The comparison is also relative. If cephalosporins carried a similar nerve risk, the two classes would look alike. And a study built on diagnoses recorded in general practice is poorly placed to count rare, severe, long-lasting cases of the kind described in case reports. One co-author works in research and development at the drug company GSK (GlaxoSmithKline).

What the study means for people prescribed fluoroquinolones

The researchers write that their results “provide some reassurance to clinicians regarding risk of peripheral neuropathy” with fluoroquinolones, relative to another antibiotic. That reassurance is narrow. It concerns a common, short-term increase large enough to show up in millions of records, not rare permanent harm. It leaves the other reasons for caution untouched: tendon rupture, which this study confirmed, and the further side effects behind the restrictions. The Australian review lists tendon problems, damage to the aorta, heart rhythm problems and swings in blood sugar among them.

The warnings on fluoroquinolone labels remain in place. Do not stop or change a prescribed medicine without talking to your doctor. Which antibiotic suits a particular infection is decided by the prescriber.

Against another common antibiotic, and within two months, fluoroquinolones showed no clear sign of extra nerve damage in 3.9 million British adults, a result that shrinks the likely size of the short-term risk without erasing the warning.