How women with anorexia sense their gut predicted relapse

Design

Single-blind, within-participant, block-order randomised experimental study with 6-month follow-up

Population

62 weight-restored women with restrictive anorexia nervosa and 57 matched controls

Primary outcome

Gut interoceptive accuracy; relapse at 6 months

Effect

Accuracy d −0.98 (−1.51 to −0.44); stomach unpleasantness predicted relapse, OR 5.73 (1.38 to 33.5)

This JAMA Psychiatry study, published on 1 September, gave 62 women with weight-restored restrictive anorexia nervosa and 57 matched controls a vibrating capsule to swallow, then measured how well they detected the gut sensations it produced. The order of stimulation blocks was randomised, but the comparison is between groups, not treatments.

Women with anorexia detected the vibrations less accurately (Cohen d −0.98) and missed more of them. Stimulation raised hunger more in the anorexia group. At six months, among 54 followed up, a stronger prior expectation that no sensation was present, a biased response pattern and unpleasant stomach sensations each predicted relapse, with odds ratios between about 4 and 6 and very wide intervals.

This is a research tool, not a clinical test. What it adds for clinicians is a reason to take gut complaints seriously after weight restoration: discomfort and misread hunger signals may be part of the relapse pathway, not only a side effect of refeeding.

• After weight restoration, ask specifically about fullness, bloating and stomach discomfort.
• Treat persistent gut discomfort as a possible relapse risk, not only a refeeding side effect.
• Consider including interoceptive work in relapse-prevention therapy where it is available.
• Do not use the capsule measures clinically; they are experimental.

Why it matters

Relapse after weight restoration may partly be about misread body signals, which routine follow-up rarely asks about.

Don't overread it

With 54 women followed up, the relapse predictors are imprecise and need replication.

The statistics, in plain English

A Cohen d near 1 means a large difference in how accurately the two groups detected gut signals. The relapse odds ratios (for example 5.73, 95% CI 1.38 to 33.5) are large but very uncertain; an interval that wide means the true effect could be small.

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.