JMIR Publications

ABSTRACT

Background:

People commonly eat while using screens, which divide attention and can weaken awareness of and memory for the meal—processes involved in intake regulation. Existing mindful-eating technologies often require changing eating behavior, asking diners to slow their eating or redirect attention toward food, demands that may be difficult to sustain during engaging screen activity.

Objective:

We explore consumption awareness—a diner's sense of how much they have consumed during a meal—as a lower-burden intervention target. We designed Digital Bites, a browser extension that uses webcam video to infer chewing activity and overlays two peripheral cues: a cumulative consumption bar with a food-icon counter and reset chime, and an ambient edge gradient reflecting momentary chewing. We evaluated whether this feedback would reduce intake and increase attention to and memory of the meal without compromising hunger, fullness, or palatability.

Methods:

We conducted a formative randomized, counterbalanced crossover laboratory study with 21 adults. Each participant ate a standardized pizza meal while reading self-selected articles in two sessions approximately one week apart: once with Digital Bites feedback and once without. Food intake was measured by plate weight; attention to food, memory of the meal, hunger, fullness, and palatability were self-reported. Outcomes were compared using 2-tailed paired-samples t tests, with two one-sided tests assessing equivalence in hunger, fullness, and palatability. Interviews conducted after the intervention condition were analyzed using deductive thematic analysis.

Results:

Participants consumed 16% less food with Digital Bites than in the control condition (255 g vs 305 g; a reduction of 49 g, 95% CI 15–83 g, approximately 128 kcal; Cohen dz=0.66; P=.007); 16 of 21 participants (76%) reduced their intake. Self-reported attention to food (P=.03; dz=0.53) and memory of the meal (P=.02; dz=0.57) were significantly higher with the feedback. Hunger change (P=.08), fullness change (P=.79), and palatability (P=.11) did not differ significantly; fullness changes were statistically equivalent between conditions (TOST P=.03). In interviews, 17 of 21 participants (81%) reported heightened awareness of their eating. Participants described using the feedback to monitor consumption, reconsider automatic continuation, and form personal consumption rules. Four of 21 participants (19%) described or imagined goals beyond intake reduction, and 13 of 21 (62%) described the overlay as natural or unobtrusive.

Conclusions:

In this formative laboratory study, the on-screen feedback reduced intake and increased self-reported attention to and memory of the meal. Participants generally described the feedback as compatible with ongoing screen activity and interpreted it according to differing personal goals. These findings provide preliminary evidence that consumption awareness may offer a flexible, lower-burden target for eating-regulation technology. Field deployments across foods, screen activities, and repeated meals, together with studies isolating the two feedback components, are natural next steps. Clinical Trial: This study was not prospectively registered because it was designed as a small formative evaluation of a prototype intervention, intended to generate preliminary evidence and inform further development rather than serve as a definitive efficacy trial.