| Background: The emergency food system is an increasingly important source of food for food insecure households. Meanwhile, there is growing concern among food banks, food bank clients, and public health officials in regard to the quality of foods offered in food banks and whether they support the dietary needs of low-income populations. As agencies begin to implement interventions to improve the healthfulness of food bank offerings by targeting specific nutrients of concern, such as added sugar or sodium, a method of evaluation is required to properly assess the impact of these strategies. However, previous efforts to characterize the overall nutrition profile of foods distributed by the emergency food system have not focused specifically on nutrients of concern, thereby limiting the ability of food bank directors and policymakers to evaluate implemented actions.;Objective: The purpose of this project was to develop and pilot test a methodology to assess the sodium content of food bank offerings. This project was driven by the need to facilitate baseline and ongoing data collection to assess the impacts of a CDC sodium reduction intervention in the emergency food system.;Methods: The data collection team visited several food banks in King County, Washington in order to explore the barriers and opportunities for a methodology that could be used to capture the amount of sodium in food bank offerings. This methodology was pilot tested at the Des Moines Area Food Bank in King County, WA on January, 13, 2017. The food bank organizes similar food items into "bins" and specifies the number of items clients can take from each bin based on their household size. We assessed the mean sodium content in several bin categories per serving and 100kcals. In addition, we calculated the mean amount of sodium distributed per client visit from each bin. Finally, we assessed the smallest change in sodium (mg) needed to detect a difference from mean sodium per serving at follow-up visits by calculating the minimum detectable effect size.;Results: The information for 306 unique items (n=1,395 total) were recorded and analyzed. Fresh prepared meals (e.g. individually packaged sandwiches, salads, etc.) had the highest mean sodium per serving, followed by boxed meals, canned soup/beans, tomato products, and pastry. There was little overlap in the bin categories with highest mean sodium per 100kcals with tomato products being the highest, followed by mixed canned vegetables, canned soup/beans, canned meats, and boxed meals. The standard deviation of both mean sodium per serving and mean sodium per 100kcals indicates a considerable variation in the sodium content of inventoried items in the majority of bin categories. The mean daily sodium contribution from analyzed bin categories was estimated to be 2630mg for a small household, 1937mg for a medium household, and 1937mg for a large household. The percent change in mean sodium per serving needed to obtain a meaningful difference at follow-up ranged from 5--70.7% for bin categories. However, for the top five bin categories with highest mean sodium per serving (fresh prepared meals, boxed meals, canned soup/beans, tomato products, and pastry), 22.3% was the largest percent change needed.;Conclusion: Prior efforts to reduce sodium in a variety of settings achieved reductions of 20--30%.7 This evidence, combined with our results for the required percent change in mean sodium per serving, suggests that it is feasible to implement strategies with a measurable impact in the food bank setting, particularly by focusing on the bin categories with highest mean sodium per serving. Overall, this study contributes novel data on the sodium content of food bank offerings as well as perspectives on the feasibility of data collection in this venue. However, the methodology and pilot test that were developed and conducted in this study have some potential limitations in regard to generalizability. Data collection took place during a single shift, which prevented us from determining how food bank offerings may change across weeks or seasons. In addition, we excluded some bin categories due to feasibility reasons. Future evaluation efforts should work to strengthen this methodology by conducting repeat assessments to determine variation in food item distribution over days, weeks, months, and seasons. Furthermore, it would be useful to assess the differences in feasibility that may arise when applying this methodology in other food banks. |
