A phone prompt after dinner did not improve glucose overall in 56 adults with type 2 diabetes — and its early effect faded

After-dinner snacking is a stubborn habit: by the time someone remembers their plan, the cupboard door may already be open. Researchers in Adelaide tested whether a phone could intervene in that small gap. Their study, published on 20 October, used continuous glucose monitors to identify people with type 2 diabetes who regularly snacked after dinner, then sent a picture-based prompt on selected evenings.

Sixty-six adults completed a two-week lead-in period. Fifty-six then entered the two-week micro-randomized phase, in which each person could receive a prompt or no prompt on a given evening. This design means participants served as their own comparison across many nights, rather than being assigned permanently to one group. The team collected sensor data from 780 nights during that phase and sent 410 prompts.

The main result was not a win: across the full two weeks, prompts did not change the glucose measure the study had set as its primary outcome, the rise in sensor glucose from 90 minutes after dinner until 4 am. The authors also found no overall improvement in reported snacking or glucose time in range. Secondary analyses suggested that prompts may have helped at first, then lost their effect as the study went on.

That pattern is interesting but preliminary. The trial was small and brief, and the early effect was a secondary finding rather than the main endpoint. It cannot show whether a different message, a less frequent schedule, practical support at home, or a longer programme could make a difference. It does underline a sensible point for anyone using an app around food: a notification can be a cue, but it is not a substitute for an eating plan or diabetes care agreed with a clinician.