JMIR Publications

ABSTRACT

Background:

Self-reported dietary intake data is crucial in nutrition and health research, yet its accuracy is compromised by challenges such as portion size estimations, food identification, memory-related bias, social desirability bias, and reactivity bias. Dietary assessment in adolescents is particularly challenging due to irregular eating habits, meal skipping, and parent or peer influences resulting in potential misreporting. Leveraging adolescents’ receptiveness to technology, we investigated the use of an innovative smartphone app (i.e., Traqq®) that facilitates dietary assessment using repeated short recalls instead of traditional 24-hour recalls. Evaluation studies of Traqq® in Dutch adults have shown successful results, but its suitability for other target populations, such as adolescents, requires further investigation.

Objective:

We designed a comprehensive, three-phase study to evaluate Traqq®’s accuracy using repeated short recalls, usability, and user perspectives among Dutch adolescents aged 12-18 years. This manuscript details the study setup, research methods, and basic characteristics in phase 1 and 2.

Methods:

In phase 1, adolescents (12-18 years) downloaded Traqq® and completed a demographic questionnaire. Traqq® was used on four random schooldays over four weeks, using 2-hour recalls on two days and 4-hour recalls on two days. A Food Frequency Questionnaire and two interviewer-administered 24-hour recalls served as dietary reference methods to assess Traqq®’s accuracy. In addition, usability was evaluated using the System Usability Scale and an experience questionnaire. In phase 2, user experiences were further explored through semi-structured interviews within a subsample of 24 adolescents. These first two phases of this mixed-methods study are now finalized for data collection. Phase 3 will focus on collecting user insights to inform app customization through co-creation sessions.

Results:

Recruitment concluded in September 2022 with 102 adolescents; 98 provided dietary data via Traqq®, and 79 completed the evaluation questionnaire. Adolescents had a mean age of 15±2 years, a BMI of 19.9±3 kg/m², 63% (n=64) were girls, 85% (n=81) attended high school, and 92% (n=88) were born in the Netherlands. Interviews were held with 6 (25%) boys and 18 (75%) girls. Co-creation sessions are planned for 2025.

Conclusions:

In this holistic study we combine quantitative and qualitative methods to evaluate the dietary assessment performance among adolescents of the Traqq® app, initially designed for adults. Specifically, next to quantitative comparisons of Traqq®’s dietary assessment potential as compared to widely accepted reference methods, we conducted semi-structured interviews, and we will carry out co-creation sessions to further guide customization for adolescents. With this user-centered, synergistic approach, we aim to establish a list of requirements for a dietary assessment app for adolescents, resulting in more efficient assessments, improved compliance, and enhanced overall accuracy in this population. Clinical Trial: ISRCTN Registry ISRCTN46230386; https://www.isrctn.com/ISRCTN46230386