MINNEAPOLIS — Research into the impact that prior antibiotics exposure may have on the risk for a new diagnosis of juvenile idiopathic arthritis (JIA) indicates that while antibiotics are associated with significantly increased risk, the relationship is even stronger with exposure to nonbacterial antimicrobials or to infections in general, according to new data presented at the American College of Rheumatology’s Pediatric Rheumatology Symposium (PRSYM) 2026.

“While we found dose- and time-dependent associations between antibiotic exposure and subsequent JIA diagnoses, we concluded that this seemed most likely to be markers of risk of infections before JIA diagnosis, either due to underlying, preceding immune dysfunction or infections that triggered subsequent disease,” Daniel Horton, MD, an associate professor of pediatrics at Rutgers Robert Wood Johnson Medical School, New Brunswick, New Jersey, told attendees.

“Antibiotics are commonly given to children early in life and have been associated with disruption of the gut microbiome, which, in turn, is linked to disruption of the normal process of immune system maturation and function,” Horton explained. While his team and others have previously shown links between early childhood antibiotic exposure and a subsequent diagnosis of JIA, “there are still questions about whether these associations are causal or merely noncausal associations.”

They therefore used two large insurance databases — a public one from 2001 to 2018 and a commercial one from 2006 to 2023 — to conduct a nested case-control study with two cohorts. A new-user cohort included patients aged 0-17 years who had more than 9 months of antibiotic-free baseline time, and a birth inception cohort included those in the public insurance dataset who were enrolled within 3 months after birth.

They identified 5175 cases who were aged 1-17 years and received a new JIA diagnosis based on either a rheumatologist’s diagnosis with more than four lab orders or more than five outpatient JIA encounters within 2 months to a year along with any JIA treatment. The researchers identified 44,309 control participants — those with no prior JIA diagnosis or history of immunosuppressive drug use — and matched 10 control participants to each case of JIA based on JIA diagnosis/index date, sex, state, and year and quarter of birth, and enrollment date.

The population was 68.1% female, and most (86.9%) were publicly insured. Both cohorts were racially and ethnically diverse and well balanced, though there were more Hispanic participants from the public insurance database. Approximately 1 in 5 participants had a baseline asthma or allergy diagnosis.

The researchers compared the proportions of children who had been exposed to any antibiotic within the 10 months before the index date, excluding the 3 months before the JIA diagnosis. They also looked at how many had been exposed to nonbacterial antimicrobials as a negative control.

In the new-user cohort, antibiotic exposure was associated with a significant 24% higher likelihood of a JIA diagnosis (adjusted odds ratio [aOR], 1.24). The association strengthened with the number of courses of antibiotics such that exposure to four or more courses more than doubled the odds of a JIA diagnosis (aOR, 2.48).

Few Differences in Risks Observed for Different Types of Antibiotics

“Also as hypothesized, we saw that the more recently that an antibiotic was received, the higher the risk of JIA was,” Horton said. When they looked at anticipated effects of different antibiotics on the microbiome, however, “we found very little differences across different types of antibiotics and their association with new JIA diagnosis,” contrary to what they expected, he said.

The association between antibiotic exposure and a JIA diagnosis was slightly stronger for the most anti-commensal antibiotics, which tend to be more disruptive to the immune system, but the associations were similar for penicillin vs clindamycin or macrolides and for anti-anaerobic vs non-anti-anaerobic antibiotics.

Associations were similar and significant in the birth inception cohort, including a greater association between four or more antibiotics and a JIA diagnosis (aOR, 1.78) than for any antibiotic (aOR, 1.17). The associations in this cohort followed similar patterns for how anti-commensal the antibiotics were, whether they were penicillin vs clindamycin and macrolides, and whether they were anti-anaerobic or non-anti-anaerobic.

However, the significant association between nonbacterial antimicrobials, such as antivirals and antifungals, and JIA diagnosis was actually stronger than that for any antibiotics exposure, both for the new-user cohort (aOR, 1.49) and the birth inception cohort (aOR, 1.26). The association was even stronger for JIA diagnosis and any infection in both the new-user (aOR, 2.05) and birth inception (aOR, 1.73) cohorts, “suggesting that maybe the infections are actually the more important exposure here,” Horton said.

There were significant associations in both cohorts between JIA diagnosis and bacterial infections, a viral infection, and any skin infection, with the strongest association for two or more bacterial infections. All of these associations between JIA diagnosis and either nonbacterial antimicrobials or various infections remained statistically significant in a sensitivity analysis with those who had never received antibiotics.

When the researchers looked at different types of JIA, they found much stronger associations for antibiotic exposure with spondyloarthritis, but that association was still weaker than the association of spondyloarthritis with nonbacterial antimicrobials and with various types of infections.

‘Is It the Immune System or the Infection’ That Increases Risk for JIA?

Randy Cron, MD, PhD, a professor of medicine and director of pediatric rheumatology at The University of Alabama Heersink School of Medicine in Birmingham, Alabama, found the results interesting, particularly the higher risk for JIA diagnosis seen in children with an infection who did not get treated with antibiotics. He agreed with Horton that it does not appear that the antibiotics themselves are causally related to developing JIA.

“What’s unclear is, is it the infection, or is it the host immune response that makes you more susceptible to infection that puts you at risk for developing JIA?” Cron told Medscape Medical News. Knowing the answer will not necessarily help clinicians make a diagnosis, “but it may be interesting to know, ultimately, is it the immune system or the infection?”

A Closer Look at Early-Life Exposures and Risk for Juvenile Spondyloarthritis, Psoriatic Arthritis

Another research poster at the meeting also investigated various early-life exposures that may be associated with gut microbial diversity and risk for spondyloarthritis. In that case-control study, led by Ana L. Altaffer, MD, who was a pediatric rheumatology fellow at Baylor College of Medicine, Houston, at the time of the study, the researchers used data from various Danish population registries from 1997 to 2024 to match 28,000 control participants with 560 cases of patients born in 1997-2023 and diagnosed with spondyloarthritis or inflammatory bowel disease (IBD)-associated arthritis between ages 6 and 21 or psoriatic arthritis between ages 1 and 21.

Most of the cases (59%) were spondyloarthritis, whereas 28% were psoriatic arthritis and 13% were IBD-associated arthritis. Specific odds ratios were not provided, but forest plots showed the expected significant association between any systemic antimicrobial agent and both spondyloarthritis and psoriatic arthritis, after adjustment for maternal smoking, maternal age, maternal antimicrobial use in the third trimester, vaginal or cesarean birth, and gestational age. There were no significant associations, however, for any penicillin or any macrolide.

As in the study by Horton and colleagues, the associations were dose-dependent, strengthening with the number of antimicrobial courses. While neither cesarean delivery nor systemic nonbacterial antimicrobials were significantly associated with any form of arthritis, upper respiratory tract infections were significantly associated with spondyloarthritis overall and with psoriatic arthritis, appearing to complement the findings in Horton’s study.

The research by Horton and colleagues was funded by the National Institute of Arthritis and Musculoskeletal and Skin Diseases and the National Center for Advancing Translational Sciences. The research by Altaffer and colleagues was funded by the Spondylitis Association of America and the National Institutes of Health.

Horton reported receiving research fund from Danisco USA Inc. Other authors reported disclosures of various types with Alkermes, Eisai, Intra-Cellular Therapies, Lundbeck, Pharmacosmos, Bristol Myers Squibb, Novartis, Pfizer, Teva, AbbVie, and Johnson & Johnson, and one author worked at Carelon Research, a subsidiary of Elevance Health. Altaffer reported being currently an employee of Vertex Pharmaceuticals. Cron reported receiving speaker, advising, committee work, and/or consulting fees and grant support from Sobi, Sironax, CareerPhysician, Neurogene, AS2 Biotherapeutics, Novartis, Lilly, and AbbVie, and Pfizer.

Tara Haelle is a science/health journalist based in Dallas.