A proposed framework to develop nutrient profiling algorithms for assessments of sustainable food: the metrics and their assumptions matter

Abstract

Abstract Purpose To holistically assess the sustainability of our global and local food systems, we need methods that combine environmental and nutritional/health dimensions. One option is nutritional life cycle assessment in which a nutrient or health metric is incorporated into standard environmental life cycle assessment. Measuring nutritional and environmental outcomes in tandem can help elucidate new results that can aid farmers, policymakers, industry, and consumers in transitioning to a food system that is more beneficial for our planet and our health. However, the development of nutrient profiling metrics is still ongoing; thus, we develop and test, with case studies, the “points of differentiation” framework to guide the use of these algorithms in sustainability analyses. Methods For each of these “points” in the framework, we provide the current state-of-the-art based on the literature and then offer new insights and recommendations for their use; we do this for different food levels (i.e., production systems, food items, and diets/food supply). We provide an explicit framework for nutrient adequacy metrics, while discussing the framework’s relevance to nutrient diversity and quality metrics. We also test the “points” in this framework using case studies that examine select foods representative of a diet. Results Based on our review, the “points” in this framework to consider include capping, weighting, energy standardization, across-the-board versus group-specific, dietary- and/or context-specific, validation, disqualifying nutrients (e.g., saturated fat), reference amount, processing quality, selection of nutrients/ingredients, interpretation, and data quality. Based on changes in nutrient density scores and rankings, Spearman rank correlations, and Wilcoxon signed rank p-values, the “points” that have the largest effect are energy standardization, dietary specificities when assessing nutrient indices in isolation, as well as capping and disqualifying nutrients, depending on how these are applied. We hope our proposed “points of differentiation” framework can provide new insights into this quickly evolving space by offering structure to improve standardization and transparency. Conclusion The framework provides recommendations for select “points;” nevertheless, there are still many open areas in nutritional life cycle assessment, including food functionality (e.g., processing effects, interaction factors, and bioavailability) and methodological questions related to the use of disqualifying nutrients and interpretation. Moreover, future work should focus on the role of “points” with respect to validation and differences within food groups.