Garsetti Marcella, de Vries Jan, Smith Maurice, Amosse Amélie, Rolf-Pedersen Nathalie
Danone Research, Palaiseau, France.
Eur J Nutr. 2007 Dec;46 Suppl 2:15-28. doi: 10.1007/s00394-007-2002-7.
Nutrient profiling is a discipline aimed at classifying foods based on their nutritional composition. So far, several profiling schemes have been proposed for varied purposes world-wide. Primary aim to inventory the main profiling schemes that have been developed so far (both applied and not) and to summarise their main aspects. Secondary aim to critically review a selection of them, to test their "performance" and to evaluate their strengths and weaknesses. Scientific and popular search engines were used for identifying profiling schemes. Schemes were described concisely by providing details on four main "Building Blocks" or factors: (1) Food category declination: category-wise or "across the board"; (2) Reference amount: 100 g, 100 kcal; serving; (3) Cut-off use: thresholds or scores; (4) Nutrients Selection: balance between positive and negative nutrients and number of them. The "performance" analysis was done by testing how the selected schemes classify a sample of food. Profiling schemes display considerable variation based on the underlying approach, format and content. Moreover, the rationale of the schemes largely varies and seems to be inspired by either nutrient recommendations or regulations figures. When tested for "performance", the five selected schemes classify in the same way foods having either a very "positive" or a very "negative" nutrient profile, whereas they give inconsistent results for food products with intermediate characteristics. Strengths and weaknesses analysis shows the difficulty of finding schemes combining qualities such as simplicity, scientific relevance, ability to cope with changes in nutrient recommendations. Current proposed profiling schemes exhibit a wide range of differences both in terms of approaches and "performance". Nutrition scientists have now the challenge to develop the "ideal scheme" that, in our view, will have to be strict enough to ensure consumer protection but also flexible enough to encourage food industry innovation and to promote a "healthy" competitive market.
营养成分剖析是一门旨在根据食物的营养成分对其进行分类的学科。到目前为止，全球已针对不同目的提出了多种剖析方案。主要目的是梳理迄今为止已开发的主要剖析方案（包括已应用的和未应用的）并总结其主要方面。次要目的是对其中一些方案进行批判性审查，测试它们的“性能”并评估其优缺点。使用科学和大众搜索引擎来识别剖析方案。通过提供有关四个主要“构建模块”或因素的详细信息来简要描述这些方案：(1) 食物类别划分：按类别或“全面”划分；(2) 参考量：100克、100千卡；一份；(3) 临界值用途：阈值或分数；(4) 营养素选择：正负营养素之间的平衡及其数量。“性能”分析是通过测试所选方案如何对食物样本进行分类来完成的。剖析方案因基础方法、形式和内容的不同而存在很大差异。此外，这些方案的基本原理差异很大，似乎受到营养建议或法规数据的启发。在测试“性能”时，五个所选方案对营养成分非常“正面”或非常“负面”的食物分类方式相同，而对于具有中等特征的食品则给出不一致的结果。优缺点分析表明，很难找到兼具简单性、科学相关性、适应营养建议变化能力等品质的方案。当前提出的剖析方案在方法和“性能”方面都存在广泛差异。营养科学家现在面临的挑战是开发出我们认为必须足够严格以确保消费者保护，但又足够灵活以鼓励食品行业创新并促进“健康”竞争市场的“理想方案”。
