Weight in autoantibody-positive children is not just something to record

Design

Prospective cohort, secondary analysis of the TrialNet Pathway to Prevention study using Cox models

Population

833 autoantibody-positive participants with overweight or obesity at baseline; 421 youth and 412 adults

Primary outcome

Progression to stage 3 (clinical) type 1 diabetes

Effect

BMI normalised in 26.8%; hazard ratio 0.516 overall (P < 0.001) over median 4.1 years; youth hazard ratio 0.497 (P = 0.001); adults not significant (P = 0.130)

Raised BMI is already known to sit alongside faster progression to clinical type 1 diabetes in people who screen autoantibody-positive. This analysis of the TrialNet Pathway to Prevention study asked the next question: among those who start with overweight or obesity, does BMI coming back to normal go with a slower course?

In 833 autoantibody-positive participants, BMI normalised in 26.8% during follow-up. Those participants had about half the risk of reaching stage 3 disease over a median 4.1 years (hazard ratio 0.516). The signal was carried by the young — hazard ratio 0.497 in 421 youth, and not statistically significant in the 412 adults. It held for the transitions from stage 1 to stage 2 and stage 2 to stage 3, but not clearly for the earliest transition. Adjusting for insulin sensitivity weakened the association, which fits insulin resistance being part of the mechanism rather than a confounder to be removed.

Nobody was assigned to lose weight, so this does not establish that a weight intervention delays type 1 diabetes. What it does is make BMI worth acting on in a group where the monitoring visit is often only about antibodies and glucose tolerance. That is a conversation available in any clinic, at no cost, with a plausible mechanism behind it.

• Record height, weight and BMI centile at every monitoring visit for an autoantibody-positive child
• Raise weight as a modifiable factor with the family rather than only as an observation
• Note the stage — the association was clearest across stage 1 to 2 and stage 2 to 3
• Do not extend the advice to autoantibody-positive adults on this evidence
• Keep the antibody and oral glucose tolerance monitoring schedule unchanged

Why it matters

Weight has been treated as a risk marker in autoantibody-positive monitoring; this puts it among the things that might be modifiable.

Don't overread it

This was observational: BMI normalisation was not assigned, so it cannot show that deliberate weight loss delays type 1 diabetes.

The statistics, in plain English

A hazard ratio of 0.516 means roughly half the rate of progression over the follow-up, not that half of those children avoid diabetes. The adult result was not a contradicting finding — it was a non-significant one in a smaller group, which may be absence of effect or absence of power. The subgroup breakdowns by age and sex sit on a few dozen events each and should be read as pattern, not as separate results.

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.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.