Digital microsteps raise short-term intent to adopt healthy behaviors during GLP‑1 treatment

What happened

An online randomized trial tested whether brief digital prompts could encourage healthier routines among adults taking GLP-1 receptor agonists. Between June 19 and July 1, 2025, researchers recruited English-speaking adults through Prolific Academic and collected responses using an online Stanford Medicine platform. The analysis included 5,054 participants, whose average age was 38.8 years; 66.5% were women. About 68% lived in the United States or United Kingdom, and the remainder lived in other countries.

Participants were assigned to one of three groups. One received written “microsteps” paired with an approximately two-minute storytelling video. A second received the same written prompts with a didactic video. A control group received no intervention. The suggested small actions addressed diet, movement, stress management, and sleep.

The primary outcome was participants’ expectation that they would adopt the suggested behaviors, measured immediately after the single exposure and again two weeks later. Both intervention versions improved those expectations compared with control. Immediately after viewing, standardized effects across eight expectation measures ranged from 0.30 to 0.72 in the storytelling group and 0.26 to 0.77 in the didactic group. Effects were smaller at the two-week follow-up.

The storytelling format performed better than the didactic version on seven of the eight expectation measures. Both videos also produced immediate increases in reported hope and happiness, but those mood differences had faded by two weeks. Of the three self-reported diet and exercise behaviors assessed over follow-up, one differed: the storytelling group reported consuming fewer sugar-sweetened drinks than the control group.

What it means

People using GLP-1 medicines may be asked to work on several routines alongside pharmacotherapy. This trial suggests that a short, inexpensive digital message can make small actions seem more likely in the near term. The storytelling result also indicates that how a message is presented may influence its immediate effect, not only which action it recommends.

The distinction between expectation and action is central. The trial mainly measured what participants thought they would do, not sustained behavior observed by researchers. A change in expectation can be relevant to designing patient support, but it does not establish improved nutrition, greater fitness, better medication use, weight change, or another clinical benefit. The only reported behavioral difference at two weeks was based on participants’ own accounts of sugary-drink consumption.

Other limits narrow the conclusion. Participants saw the material once, follow-up lasted two weeks, and recruitment occurred on an online research platform. People willing and able to join an internet trial may not represent everyone receiving GLP-1 care. The abstract also does not establish whether effects differed by country, age, medicine, reason for treatment, or access to clinical support.

What it does not mean