JMIR Publications

ABSTRACT

Background:

Obesity remains a critical public health challenge, requiring innovative strategies to improve weight loss. Although preliminary evidence indicates that tailoring interventions to both cognitive and behavioral factors may enhance engagement and effectiveness, most research has focused on these elements in isolation. With the increasing scalability and convenience of digital weight loss programs, developing reliable methods for automated personalization has become essential. To address this, we constructed a questionnaire that assigns patients to cognitive-behavioral phenotypes, to enable tailored advice. This study evaluates the effectiveness of this approach in improving patient engagement and weight loss within a digital intervention.

Objective:

To assess whether sending personalized weight loss advice can increase program engagement and lead to greater weight loss.

Methods:

A quasi-experimental design was employed. Patients on a weight loss program were sent a 17-item questionnaire which matched them to one of four cognitive-behavioral profiles (phenotypes). Those who completed the questionnaire were sent phenotype-tailored weight loss advice once per week. As part of their weight loss program, patients used an app to track their meals and activities. Number of in-app events (i.e. engagements) was used as a proxy measure of program engagement. Self-reported weight was submitted as part of the structured weight loss program. Outcomes were compared to a historical cohort of patients who participated in the same weight loss program one year earlier, and those who were sent an invitation to complete the questionnaire but did not (i.e. non-responders).

Results:

Those who received phenotype-tailored advice generated significantly more in-app engagements (M=257, SD=232), than those in the historical cohort (M=159, SD=187; P<.001), and compared to non-responders (M=135, SD=198; P<.001). There was a trend towards greater weight loss for those who received the tailored advice (M=-2.23kg, SD=7.97), compared to the historical cohort (M=-1.6kg, SD=5.39), and the non-responders (M=-0.69kg, SD=13.23) but these differences were not significant, (P=.29 and P=.23, respectively).

Conclusions:

This study provides preliminary evidence that receiving weight loss advice tailored to an individual’s cognitive-behavioral phenotype can increase engagement on a structured weight loss program, which could in turn support greater body mass reduction.