JMIR Publications

ABSTRACT

Background:

Globally, many adolescents fail to meet dietary guidelines and engage in unhealthy eating behaviours. Dynamically tailored mobile health (mHealth) interventions are interventions that adjust the timing and type of content delivered based on repeated assessments of an individual’s behaviour, context, or preference. These interventions may offer a promising platform for improving adolescent diet, as smartphone ownership among youth has surpassed 90% in most middle to high income countries. These interventions can deliver timely, personalized information to enhance relevance, and behavioural impact.

Objective:

This systematic review explored dynamically tailored mHealth interventions aimed at improving diet-related behaviours among adolescents, by examining behaviour change models underpinning app development, key intervention components, intervention attrition, feasibility and acceptability, and effectiveness in diet and food-related behaviours.

Methods:

Four electronic databases were searched using keywords related to adolescents, mHealth interventions, and diet-related behaviours. Study quality was assessed using the mHealth Evidence Reporting and Assessment Checklist (mERA) and Cochrane’s Risk of Bias Tool for Randomized Trials (ROB-2), as applicable. Findings were narratively synthesized to summarize intervention characteristics, quality, and effectiveness.

Results:

Eleven studies involving 3,674 adolescents (mean age 12.7-18.3 years) with average sample sizes of 20-1,111 participants were included. Interventions targeted various diet-related behaviours, including general healthy eating habits (e.g., fruit and vegetable consumption), sugar consumption, iron intake, eating rate and disordered eating. Five of eleven studies were informed by a behaviour change theory. Common app components included feedback mechanisms (automated messages, human-delivered feedback, or gamification), self-monitoring, social supports, and cognitive tools. Seven studies reported improvement in at least one dietary, psychosocial, or anthropometric outcome, however effects were often not sustained. Feasibility and acceptability were reported in three of the eleven mHealth interventions and results were mixed. Intervention attrition was a common limitation, with 7 studies reporting rates >29%. While just over half of the studies (n=6) detailed intervention delivery and content, limited information was provided on the details of app cost, adaptability, or regulatory compliance. Two out of the five studies assessed for risk of bias using the RoB-2 were considered to have some concerns.

Conclusions:

Dynamically tailored mHealth interventions show emerging potential to support diet-related behaviours among adolescents. Although several studies demonstrated positive outcomes, effectiveness remain limited by high attrition rates and lack of sustained impact.