OSU Plate Waste Study

The objective of this project is to pilot test the Remote Food Photography Method (RFPM) and SmartIntake smartphone app as an approach for measuring food waste in controlled and free-living environments. The RFPM and SmartIntakeTM app constitutes a distinct improvement over self-report FI methods, such as pen-and-paper surveys and dietary recall methods that rely on subjects' ability to accurately recall the types and portion sizes of foods consumed. After training, people use the SmartIntakeTM app to capture images of their foods both before and after each eating occasion. Photos are automatically uploaded from the user's smartphone to a secure server at Pennington Biomedical. The images are analyzed to quantify food selection, food waste, and food intake using validated methods and a computer program (the Food Photography Application©) built by Pennington Biomedical. The method has been found to accurately quantify food selection, food waste, and food intake.

Importantly, the RFPM and SmartIntakeTM app eliminate one of the largest sources of error associated with self-report methods - subjects' inability to accurately estimate portion size, which accounts for ~50% of the error in self-report methods11, 13-15. FI estimation studies have observed that overweight subjects react to monitoring by underestimating the self-reported amounts of FI or eat less when monitored16, 17. Similarly, FW behavioral studies have observed that subjects start self-reporting smaller amounts of FW, however it is not clear if this is the result of increased awareness and less generation, or if it due to not appearing as wasteful individuals18. However, people respond less to the RFPM/SmartIntake app and experiment results show that the underestimation from RFPM does not vary with energy intake level12.

Many shortcomings that plague FI measurement likely hamper FW measurement. Biases in the opposite direction (e.g., cleaning your plate to show less FW) likely emerge when measuring FW and may be similarly mitigated with the proposed app. We believe using RFPM and SmartIntakeTM app as the base for a FW estimation app holds great promise to transform FW measurement. First, we note that FW associated with dining situations (i.e., plate waste) is directly estimated by the RFPM, hence only a simple transform of current outputs is required. To fully adapt RFPM for FW measurement, the procedures will be modified to estimate FW created during food preparation and during stored food purges (e.g., cleaning the fridge). The RFPM method already works in dine-out settings, which are currently omitted by some existing FW measurement methods. Additional functionality, such as estimating edible vs inedible FW (and hence calories wasted), estimating waste rate among food categories, and estimating rates of diversion from landfills, will also be explored so that multiple policy relevant metrics can be instantly calculated and stored.