Testing the Reliability of a New Diet Tracking App That Mirrors Canada’s Food Guide, iCANPlate

IALH Research Fellow Ryan Rhodes has co-authored a new research article entitled Testing the Reliability of a New Diet Tracking App That Mirrors Canada’s Food Guide, iCANPlate. Collaborating authors include Anisha Sandhu, Angela S. Alberga, Nizar Bouguila, Tamara Cohen, Claudia Faustini, Stephanie L. Fitzpatrick, Jean-Philippe Gouin, Lisa Kakinami, Maryam Kebbe, and Tara Ubovic. The article was published in Current Developments in Nutrition.

Abstract:

Objectives: To assess the accuracy of the iCANPlate app in capturing dietary intake by comparing entries from general users to those of a registered dietitian (RD, the expert)

Methods: An observational study was conducted in April 2025 at the University of British Columbia (Vancouver) with n=59 adults (79.66% women) who tested iCANPlate by recording a preselected lunch provided to them in a laboratory setting. A researcher photographed each meal, and a registered dietitian (RD) used the photos to quantify the food group proportions of each participant’s meal. Since the data consisted of continuous values, inter-rater reliability between participant and RD recordings for each food group category was analyzed using Pearson’s correlation coefficients.

Results: The app exhibited moderate inter-rater reliability for Fruits and Vegetables (r = 0.62, p < 0.0001), Proteins (r = 0.53, p < 0.0001) and Other foods (r = 0.42, p < 0.01). Fair agreement was observed for Grains (r = 0.39, p < 0.01). Beverages demonstrated poor agreement, with no statistical significance (r = 0.22, p < 0.05). Variability in inter-rater reliability may stem from participants’ education levels and socioeconomic backgrounds, influencing their food exposure.

Conclusions: iCANPlate demonstrated moderate inter-rater reliability for documenting most food categories. However, users did not accurately track beverages, highlighting an area for improvement. These findings support iCANPlate’s potential for dietary self-monitoring while emphasizing the need for more research to refine and optimize its reliability.

To read the full article, see https://doi.org/10.1016/j.cdnut.2025.106983