Ran Xu

Crowdsourced online food images, when combined with food image recognition technologies, have the potential to offer a cost-effective and scalable solution for the assessment of the restaurant nutrition environment. While previous research has explored this approach and validated the accuracy of food image recognition technologies, much remains unknown about the validity of crowdsourced food images as the primary data source for large-scale assessments. In this paper, we collect data from multiple sources and comprehensively examine the validity of using crowdsourced food images for assessing the restaurant nutrition environment in the Greater Hartford region. Our results indicate that while crowdsourced food images are useful in terms of the initial assessment of restaurant nutrition quality and the identification of popular food items, they are subject to selection bias on multiple levels and do not fully represent the restaurant nutrition quality or customers’ dietary behaviors. If employed, the food image data must be supplemented with alternative data sources, such as field surveys, store audits, and commercial data, to offer a more representative assessment of the restaurant nutrition environment.

Objective The aim of this study was to identify the levels of food‐tracking adherence that best predict ≥3%, ≥5%, and ≥10% weight loss at 6 months and to identify distinctive food‐tracking trajectories among participants in a 6‐month, commercial digital weight‐management program. Methods This study used data from 153 participants of a single‐arm trial of a commercial digital weight‐management program, WeightWatchers (WW). Receiver operating characteristic (ROC) curve analysis was conducted to identify the optimum thresholds of food tracking that can predict ≥3%, ≥5%, and ≥10% weight loss at 6 months. Time series clustering analysis was performed on weekly food‐tracking data to identify trajectories and compare trajectories on weight loss at 6 months. Results ROC analyses showed that the optimum thresholds of food tracking were 28.5% of the intervention days to achieve ≥3% weight loss (area under the curve [AUC] = 0.820, p < 0.001), 39.4% to achieve ≥5% weight loss (AUC = 0.744, p < 0.001), and 67.1% to achieve 10% weight loss (AUC = 0.712, p = 0.002). Time series clustering analyses found three food‐tracking trajectories. Trajectories differed significantly in weight loss at 6 months ( F = 14.1, p < 0.001). Conclusions Results suggest that perfect food‐tracking adherence is not necessary to achieve clinically significant weight loss.