Human papillomaviruses (HPVs) are double-stranded DNA viruses associated with a wide spectrum of human diseases, including cervical cancer and autoimmune conditions (1, 2). Prophylactic HPV vaccination, particularly the quadrivalent HPV4 vaccine, has demonstrated robust efficacy in preventing HPV-related infections and malignancies (3, 4). Despite its favorable safety profile, concerns remain regarding its potential role in triggering autoimmune disorders such as juvenile idiopathic arthritis (JIA).

JIA is the most prevalent autoimmune rheumatic disease in children, characterized by persistent joint inflammation without a known etiology. It is believed to result from complex interactions between genetic susceptibility and environmental triggers, including viral infections (5). Molecular mimicry and immune dysregulation have been proposed as key mechanisms by which pathogens may initiate or exacerbate autoimmunity (6, 7).

Previous studies have yielded conflicting findings on the association between HPV4 vaccination and the development of autoimmune arthritis. A U.S.-based analysis of the Vaccine Adverse Event Reporting System (VAERS) reported an elevated odds ratio for rheumatoid arthritis following HPV4 vaccination (8). Similarly, a Colombian study employing inverse probability of treatment weighting also observed increased arthritis risk (9). However, both relied on self-reported adverse events, potentially introducing reporting bias. Conversely, a retrospective study using electronic medical records from Kaiser Permanente found no significant increase in JIA incidence within 180 days post-vaccination (10).

To address these inconsistencies, we conducted a large-scale retrospective cohort study using the TriNetX U.S. Collaborative Network. This federated platform offers access to de-identified electronic health records across multiple healthcare organizations, enabling robust longitudinal analyses (11). We aimed to evaluate whether HPV4 vaccination is associated with an increased or decreased risk of new-onset JIA in adolescent females. Three complementary analyses were performed to examine the effect of vaccination status, antinuclear antibody (ANA) positivity, and HPV4 dose number on subsequent JIA risk.

This large-scale, multicenter retrospective cohort study found no evidence of increased risk of juvenile idiopathic arthritis (JIA) following quadrivalent HPV vaccination. In contrast, our findings consistently demonstrated a protective association between HPV4 vaccination and new-onset JIA, particularly from 12 months to 36 months post-vaccination. These results held true across two distinct analyses, including one that excluded ANA-positive individuals, thus strengthening the validity of the observed protective trend.

The absence of early-onset autoimmune flares (i.e., within 42 days) aligns with current immunological understanding, which suggests that an adaptive autoimmune response generally takes longer to manifest (12). The 42-day threshold is also the standard time frame used by the VAERS system for defining vaccine-associated arthritis (13). Therefore, our observation that JIA onset was significantly lower in the vaccinated cohort beginning at 6 to 12 months post-vaccination is biologically plausible and statistically robust.

Our findings corroborate prior research suggesting a potential autoimmune link between HPV infection and rheumatic diseases, including RA and systemic lupus erythematosus (14, 15). One tertiary center study showed that 48.2% of HPV DNA+ patients had autoantibody or autoimmune disease (16). The mechanism for their condition was generally considered to be via molecular mimicry and cross reactions between self-antigens and viral proteins (17). HPV-47, in particular, contains sequences with molecular mimicry to citrullinated autoantigens involved in RA pathogenesis (18). Moreover, HPV infection has been associated with altered immune cell activation, including Th17 cell expansion and elevated IL-17 levels, which may potentiate local and systemic inflammation (19).

One possible explanation for our findings is that HPV4 vaccination may reduce the systemic immune burden associated with persistent HPV infection, thereby lowering the probability of autoimmune activation in genetically predisposed individuals. This hypothesis is supported by recent NHANES data showing that HPV4 vaccination was associated with reduced RA incidence among adults aged 18–59 years (20).

Although the HPV4 vaccine is not classified as a tolerogenic vaccine, its nature as a non-live virus-like particle (VLP) formulation typically administered in multiple doses may favor regulatory T cell (Treg) induction under low-inflammatory conditions. Previous studies have suggested that repeated antigen exposure from inactivated or subunit vaccines can promote tolerogenic responses by providing persistent but non-replicative stimulation, which facilitates Treg differentiation via IL-10 production, reduced co-stimulation, and antigen presentation in tolerogenic dendritic cell contexts (21). However, in our study, no significant difference in JIA incidence was observed between individuals who received only one dose versus those who completed the full vaccination schedule. This may reflect the relatively small sample size in the single-dose subgroup, limiting the power to detect potential dose-dependent effects. Nonetheless, the possibility that even a single dose of HPV4 vaccine provides sufficient antigen exposure to initiate protective immune modulation cannot be excluded.

Our results are consistent with the known epidemiology of JIA in the United States. Previously reported incidence rates in girls aged 11–15 years range from 13.2 to 30.2 per 100,000 (22). In our study, the observed JIA incidence—based on both diagnostic coding and medication use—fell within or slightly above this range, affirming the external validity of our cohort. Importantly, using a more stringent diagnostic criterion (M08.xx plus drug use) revealed a delayed but more robust protective effect beginning at 6–12 months, likely reflecting improved specificity for true JIA cases. Taken together, the consistency between our observations and prior evidence—including NHANES epidemiologic data, molecular mimicry mechanisms, and the Treg-promoting potential of inactivated vaccines—provides additional support for the biological plausibility of a protective association.

This study has several limitations. First, the TriNetX dataset is predominantly U.S.-based and may not generalize to other populations, particularly Asian cohorts where sample sizes were limited. Second, temporal variation in vaccine administration may confound short-term effects, although this is unlikely to influence long-term outcomes. Third, while the use of both diagnostic codes and medication prescriptions enhances case validation, misclassification remains possible. Not all individuals with an M08.xx code received antirheumatic therapy, suggesting possible overcoding or transient arthropathy. As with all EHR-based studies, the potential for diagnostic misclassification exists; however, our strict case definitions and exclusion of pre-index autoimmune diagnoses or treatments help reduce such risk. Additionally, although extensive matching was performed, residual confounding from unmeasured variables—such as family history, HLA genotype, environmental exposures, and socioeconomic status —cannot be excluded. Although vaccination histories other than HPV were not available for analysis, routine immunizations are limited between ages 9 and 13, with seasonal influenza vaccine being the only widely administered immunization during this period.

In conclusion, this large-scale TriNetX study showed the protective effect which at least one dose of HPV4 vaccination from 12 months to 36 months has on developing JIA. Single or multiple doses of the vaccine showed similar rates of new onset JIA from 42 days to 3 years after HPV vaccination.