Behavioral Lifestyle Intervention Program Using Mobile Application Improves Diet Quality in Adults With Prediabetes (D'LITE Study): A Randomized Controlled Trial

Background Mobile health applications (mHealth apps) are increasingly being used in weight loss interventions. However, evidence on the effects of such interventions on diet quality and their correlation with weight loss is lacking. Objective The objective of this study was to examine whether changes in the diet quality of adults with prediabetes followed the use of an mHealth-enabled lifestyle intervention, compared with those who did not, and whether these changes correlated with weight loss. Design A secondary analysis of a 6-month randomized controlled trial Diabetes Lifestyle Intervention using Technology Empowerment (D'LITE) was conducted, with participants recruited from October 2017 to September 2019. Participants/setting Community-dwelling adults (n = 148) in Singapore diagnosed with prediabetes and body mass index (BMI) >= 23 were included in this study. Intervention Participants were randomized to receive either a 6-month mHealthenabled lifestyle intervention program (diet and physical activity) or standard care dietary advice. Main outcome measures Dietary data were collected in the form of 2-day food records at baseline, 3, and 6 months. Changes in Alternate Healthy Eating Index-2010 (AHEI2010) scores and food groups (servings/day), calculated from the dietary data, and correlation between changes in AHEI-2010 and weight loss at 3 and 6 months, were examined. Statistical Analyses Between-group comparisons of continuous variables and withinparticipants variation were performed using longitudinal mixed-effect models, intention-to-treat principles. The models included treatment groups, time (baseline, 3 months, and 6 months), and covariates (age, sex, and BMI), as well as the group x time interactions, as fixed variables and within-participant variation in outcome values as random variable. The random intercept for participants accounted for the dependence of repeated measures. A likelihood ratio test was also conducted to test random effect variance. Spearman correlation test was used to examine correlation between changes in AHEI-2010 scores and weight loss. Results There was a significant improvement in overall diet quality as ascertained by the AHEI-2010, by 6.2 points (95% confidence interval [CI], 3.8e8.7; P < 0.001) in the intervention group as compared with the control. The participants in the intervention group had a significantly greater reduction in intake of sugar-sweetened beverages (SSB) by 0.5 servings/day (95% CI, -0.8, -0.2; P < 0.001) and sodium by 726 mg/day (95% CI, -983, -468; P < .001), compared with those receiving standard care. At 3 and 6 months, a significant decrease in SSB (0.8 servings/day; 0.7 servings/day, respectively) and sodium (297 mg/day; 296 mg/day, respectively) intakes were reported compared with baseline intakes. Small positive correlations (r = 0.2; P < 0.05) were observed between changes in AHEI-2010 scores from baseline and percentage weight loss at 3 and 6 months. Conclusion For adults with prediabetes in Singapore, diet quality can be improved with an mHealth-enabled lifestyle intervention program. A small positive correlation exists between AHEI-2010 scores and weight loss.