The incidence of uveitis associated with juvenile idiopathic arthritis (JIA) varies around the world. Our aim was to investigate the incidence and predictors of uveitis in a Nordic population-based cohort.

Consecutive JIA cases from defined geographical areas in Denmark, Finland, Sweden and Norway with disease onset between January 1997 to June 2000 were followed for median 98 months in this prospective longitudinal cohort study. Potential clinical and immunological predictors of uveitis were identified with logistic regression analysis.

Uveitis occurred in 89 (20.5%) of the 435 children with regular ophtalmologic follow-up among the 500 included. Chronic asymptomatic uveitis developed in 80 and acute symptomatic uveitis in 9 children. Uveitis developed at a median interval of 0.8 (range − 4.7 to 9.4) years after onset of arthritis. Predictors of uveitis were age < 7 years at JIA onset (Odds ratio (OR) 2.1, 95% confidence interval (CI) 1.3 to 3.5), presence of antihistone antibodies (AHA) > 15 U/ml (OR 4.8 (1.8 to 13.4)) and antinuclear antibodies (ANA) (OR 2.4 (1.5 to 4.0)). Mean combined IgM/IgG AHA was significantly higher in the uveitis group (19.2 U/ml) than in the non-uveitis group (10.2 U/ml) (p = 0.002). Young age at JIA onset predicted uveitis in girls (p < 0.001), but not in boys (p = 0.390).

Early-onset arthritis and presence of AHA in girls, as well as presence of ANA in both genders, were significant predictors of chronic uveitis. The high incidence of uveitis in this long-term Nordic JIA cohort may have severe implications in a lifelong perspective.