MS-Nutrition, 27 bld Jean Moulin Faculté de Médecine la Timone, Laboratoire C2VN, 13385 Marseille CEDEX 5, France.
Albert Einstein College of Medicine, Montefiore Medical Center, New York, NY 10467, USA.
Nutrients. 2019 Jun 19;11(6):1374. doi: 10.3390/nu11061374.
The quality of dietary patterns can be optimized using a mathematical technique known as linear programming (LP). LP methods have rarely been applied to individual meals. The present LP models optimized the breakfast meal for those participants in the nationally representative National Health and Nutrition Examination Survey 2011-2014 who ate breakfast ( = 11,565). The Nutrient Rich Food Index (NRF9.3) was a measure of diet quality. Breakfasts in the bottom tertile of NRF9.3 scores (T1) were LP-modeled to meet nutrient requirements without deviating too much from current eating habits. Separate LP models were run for children and for adults. The LP-modeled breakfasts resembled the existing ones in the top tertile of NRF9.3 scores (T3), but were more nutrient-rich. Favoring fruit, cereals, and dairy, the LP-modeled breakfasts had less meat, added sugars and fats, but more whole fruit and 100% juices, more whole grains, and more milk and yogurt. LP modeling methods can build on existing dietary patterns to construct food-based dietary guidelines and identify individual meals and/or snacks that need improvement.
线性规划（LP）是一种优化饮食模式质量的数学技术。LP 方法很少应用于个体餐食。本 LP 模型优化了 2011-2014 年全国健康与营养调查中吃早餐的参与者（n=11565）的早餐。营养丰富食品指数（NRF9.3）是衡量饮食质量的指标。NRF9.3 得分最低的三分之一（T1）的早餐，通过 LP 模型来满足营养需求，同时尽量不偏离当前的饮食习惯。为儿童和成人分别运行 LP 模型。LP 模型构建的早餐与 NRF9.3 得分最高的三分之一（T3）的现有早餐相似，但更具营养。LP 模型构建的早餐偏好水果、谷物和乳制品，肉类、添加糖和脂肪更少，而更多的是完整水果和 100%果汁、更多的全谷物和更多的牛奶和酸奶。LP 建模方法可以在现有饮食模式的基础上构建基于食物的膳食指南，并确定需要改进的个别餐食和/或零食。
