Self-Monitoring of Weight Loss Over Time: Secondary Data Analysis of a Randomized Controlled Trial

Behavior change support systems aim to shape, modify, or strengthen attitudes or behaviors without using coercion or deception. One of the main software features of persuasive system design is self-monitoring, which provides the means for users to continuously track their own performance or status, thereby facilitating goal attainment.

The aim of this study is to examine whether self-input of weight (self-monitoring frequency) and its interaction with time influence weight loss in adults using a mobile health behavior change support system (mHBCSS). We hypothesized that higher self-monitoring frequency would be associated with greater weight loss, with effects varying across the intervention period.

This secondary analysis used data from the intervention group of a randomized, open, waitlist-controlled trial in adults with obesity (BMI 30-40 kg/m²). Participants used the mHBCSS for 12 months, and analyses included only participants who maintained self-monitoring for at least 6 months (N=75). Weight changes were analyzed across 9 time periods. Quantile regression (QR) was applied to examine effects on the 25th, 50th (median), and 75th percentiles of weight loss. The models included self-monitoring frequency, time periods, and their interaction. A sensitivity analysis using a multiple imputation procedure was performed to assess the robustness of the QR.

The time period variable was significant at the 25th weight loss percentile (QR coefficient: -1.164, 95% CI -1.453 to -0.756) and at the 50th weight loss percentile (QR coefficient: -0.603, 95% CI -0.733 to -0.493). The interaction variable was significant at the 50th (QR coefficient: -0.018, 95% CI -0.047 to -0.003) and 75th (QR coefficient: -0.036, 95% CI -0.051 to -0.027) weight loss percentiles. Self-monitoring frequency alone was not statistically significant.

The study demonstrates that the effect of self-monitoring on weight loss is time-dependent. While a higher frequency of self-monitoring is associated with greater weight loss early in the intervention, its influence decreases as time progresses. These findings emphasize the importance of sustained engagement with self-monitoring rather than focusing solely on frequency, suggesting that interventions should incorporate strategies to maintain consistent self-monitoring use throughout the behavior change process.