- Design
- systematic review and meta-analysis of 15 observational studies
- Population
- 19,015 children with juvenile idiopathic arthritis, weighted mean age 8.0 years
- Primary outcome
- prevalence of antithyroid antibody positivity and thyroid dysfunction, and associated risk factors
- Effect
- prevalence 4% (95% CI 3 to 5); risk difference vs controls 0.08 (0.02 to 0.14); family history OR 3.91 (1.86 to 8.25)
A meta-analysis pooled 15 studies covering 19,015 children with juvenile idiopathic arthritis, 69.6% female, weighted mean age 8.0 years, reporting antithyroid antibody positivity and thyroid dysfunction.
Pooled prevalence was 4% for antithyroid antibody positivity (95% CI 3% to 5%) and 4% for thyroid dysfunction (3% to 8%). Against controls, the risk difference for antibody positivity was 0.08 (0.02 to 0.14). Three characteristics predicted it: a family history of autoimmune thyroid disease (OR 3.91, 95% CI 1.86 to 8.25), concurrent antinuclear antibody positivity (OR 1.62, 1.13 to 2.31) and older age at onset (mean difference 2.10 years, 1.32 to 2.89). The specific juvenile arthritis subtype made no difference.
This is the shape of finding that supports targeted rather than universal screening. A 4% prevalence does not justify testing thyroid function and antibodies in every child at every visit, but a family history nearly quadrupling the odds is a concrete trigger - as is antinuclear antibody positivity, which most of these children have had tested anyway for uveitis risk. Take a family history of thyroid disease at diagnosis and record it, then check thyroid function in the children it flags, in the antinuclear-positive, and in those presenting closer to adolescence. And remember that fatigue and growth faltering in a child with arthritis have more than one explanation.
- Take and record a family history of autoimmune thyroid disease at diagnosis
- Check thyroid function where there is a family history, antinuclear antibody positivity or older onset age
- Do not screen every child at every visit - pooled prevalence was about 4%
- Consider thyroid disease in a child with arthritis and unexplained fatigue or growth faltering
- Do not use the arthritis subtype to decide - it did not predict thyroid involvement
Why it matters
It converts a vague association between juvenile arthritis and thyroid autoimmunity into three specific reasons to send the test.
Don't overread it
Pooled observational prevalence with substantial heterogeneity; it identifies who is more likely to test positive, not that screening improves any outcome.
The statistics, in plain English
A risk difference of 0.08 sits oddly beside a pooled prevalence of 0.04, which signals real heterogeneity between the studies contributing each estimate rather than an error to resolve; the direction is consistent but the magnitude is not settled. The family history odds ratio interval, 1.86 to 8.25, is wide - the association is reliable, its size is not. Pooling 15 observational studies with differing antibody assays and dysfunction definitions limits what these numbers can mean in any one clinic.
Read the rest in the app
You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.
Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for rheumatology, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free