Interaction- and Communication-based Systems, Institute of Computer Science, University of St. Gallen, 9000 St. Gallen, Switzerland.
ETH AI Center, ETH Zurich, Swiss Federal Institute of Technology in Zurich, 8092 Zurich, Switzerland.
Nutrients. 2021 Dec 29;14(1):159. doi: 10.3390/nu14010159.
In light of the globally increasing prevalence of diet-related chronic diseases, new scalable and non-invasive dietary monitoring techniques are urgently needed. Automatically collected digital receipts from loyalty cards hereby promise to serve as an objective and automatically traceable digital marker for individual food choice behavior and do not require users to manually log individual meal items. With the introduction of the General Data Privacy Regulation in the European Union, millions of consumers gained the right to access their shopping data in a machine-readable form, representing a historic chance to leverage shopping data for scalable monitoring of food choices. Multiple quantitative indicators for evaluating the nutritional quality of food shopping have been suggested, but so far, no comparison has validated the potential of these alternative indicators within a comparative setting. This manuscript thus represents the first study to compare the calibration capacity and to validate the discrimination potential of previously suggested food shopping quality indicators for the nutritional quality of shopped groceries, including the Food Standards Agency Nutrient Profiling System Dietary Index (FSA-NPS DI), Grocery Purchase Quality Index-2016 (GPQI), Healthy Eating Index-2015 (HEI-2015), Healthy Trolley Index (HETI) and Healthy Purchase Index (HPI), checking if any of them performs differently from the others. The hypothesis is that some food shopping quality indicators outperform the others in calibrating and discriminating individual actual dietary intake. To assess the indicators' potentials, 89 eligible participants completed a validated food frequency questionnaire (FFQ) and donated their digital receipts from the loyalty card programs of the two leading Swiss grocery retailers, which represent 70% of the national grocery market. Compared to food and nutrient intake, correlations between density-based food and nutrient intake and food shopping data are stronger. The FSA-NPS DI has the best calibration and discrimination performance in classifying participants' consumption of nutrients and food groups, and seems to be a superior indicator to estimate nutritional quality of a user's diet based on digital receipts from grocery shopping in Switzerland.
鉴于与饮食相关的慢性疾病在全球范围内的患病率不断上升，我们迫切需要新的可扩展且非侵入性的饮食监测技术。从会员卡自动收集的数字收据有望成为个体食物选择行为的客观且可自动追踪的数字标记，且无需用户手动记录每餐的食物。随着欧盟《通用数据保护条例》的引入，数以百万计的消费者有权以机器可读的形式访问其购物数据，这代表了利用购物数据进行可扩展的食物选择监测的历史性机会。已经提出了多种用于评估食物购买的营养质量的定量指标，但迄今为止，尚无研究在比较环境中验证这些替代指标的潜力。因此，本研究首次比较了先前提出的食物购买质量指标对于所购买杂货的营养质量的校准能力和验证潜力，这些指标包括食品标准局营养成分分析系统饮食指数（FSA-NPS DI）、2016 年杂货购买质量指数（GPQI）、2015 年健康饮食指数（HEI-2015）、健康手推车指数（HETI）和健康购买指数（HPI），以检验它们中的任何一个是否与其他指标有所不同。该假设是，某些食物购买质量指标在校准和区分个体实际饮食摄入方面优于其他指标。为了评估这些指标的潜力，89 名符合条件的参与者完成了一份经过验证的食物频率问卷（FFQ），并向瑞士两家领先的杂货零售商的会员卡计划捐赠了他们的数字收据，这两家零售商占据了全国杂货市场的 70%。与基于食物和营养素摄入量的相关性相比，基于食物密度和营养素摄入量与食物购买数据的相关性更强。FSA-NPS DI 在分类参与者对营养素和食物组的消耗方面具有最佳的校准和区分性能，并且似乎是一种优于其他指标的指标，可以根据瑞士的杂货购物数字收据来估计用户饮食的营养质量。
