Adherence to Self-monitoring of Dietary Intake During a Weight Loss Intervention: Does a Personalized Approach Maintain Adherence? (FS11-04-19)

We compared self-monitoring adherence in participants randomized to two weight loss programs: a STANDARDIZED, one-size-fits-all, low-fat diet, or a diet PERSONALIZED to minimize the postprandial glycemic response. Participants were adults with pre-diabetes or type 2 diabetes, and a BMI >27 k/m². Both groups were instructed to restrict total calories, monitor dietary intake with the Personal Nutrition Program (PNP) smartphone app, and attend videoconference behavioral counseling sessions on the same intervention schedule. STANDARDIZED (n = 12) received app feedback about intake of total calories and dietary fat. PERSONALIZED (n = 20) received app feedback about intake of total calories plus a meal-specific predicted glycemic score. Total meal entries were measured at 1, 2 and 3 months. Self-monitoring adherence was defined as logging >50% of expected meals each month into the PNP app, assuming 3 meals/day. Session attendance was also measured. Repeated measures binomial logistic regression analysis was used to assess change in adherence due to treatment group, time (i.e., months), and the interaction between treatment and time, adjusting for age, gender and BMI. Proportion adherent was 75.0%, 41.7% and 8.3% in the STANDARDIZED group and 85.0%, 80.0% and 75.0% in the PERSONALIZED group during months 1, 2 and 3, respectively. The repeated measures model demonstrated a significant effect of month (P < 0.001) and a treatment*month interaction (P = 0.011). After adjusting for covariates, these effects remained significant, showing a significant reduction in odds of adherence by month (OR [95%CI]: 0.13 [0.05, 0.37]; P < 0.001). Moreover, compared to the STANDARDIZED, PERSONALIZED participants had greater odds of adherence over time (OR [95%CI]: 5.12 [1.49, 17.6]; P = 0.009). Higher BMI was significantly associated with lower adherence (OR [95%CI]: 0.92 [0.87, 0.98]; P = 0.006). The proportion of attendance at videoconference sessions was similar between groups (STANDARDIZED: 77.1%; PERSONALIZED: 77.5%). Two weight loss programs having similar calorie targets, behavioral approach, and contact schedule resulted in similar session attendance. However, adherence to self-monitoring was better when feedback was personalized. American Heart Association.