Riedl Anna, Hillesheim Elaine, Wawro Nina, Meisinger Christa, Peters Annette, Roden Michael, Kronenberg Florian, Herder Christian, Rathmann Wolfgang, Völzke Henry, Reincke Martin, Koenig Wolfgang, Wallaschofski Henri, Daniel Hannelore, Hauner Hans, Brennan Lorraine, Linseisen Jakob
Independent Research Group Clinical Epidemiology, Helmholtz Zentrum München, German Research Center for Environmental Health (GmbH), Ingolstädter Landstr. 1, 85764, Neuherberg, Germany.
Chair of Epidemiology, Ludwig-Maximilians-Universität München, at UNIKA-T, Neusässer Str. 47, 86156, Augsburg, Germany.
Mol Nutr Food Res. 2020 Apr;64(8):e1900918. doi: 10.1002/mnfr.201900918. Epub 2020 Feb 20.
Previous work identified three metabolically homogeneous subgroups of individuals ("metabotypes") using k-means cluster analysis based on fasting serum levels of triacylglycerol, total cholesterol, HDL cholesterol, and glucose. The aim is to reproduce these findings and describe metabotype groups by dietary habits and by incident disease occurrence.
1744 participants from the KORA F4 study and 2221 participants from the KORA FF4 study are assigned to the three metabotype clusters previously identified by minimizing the Euclidean distances. In both KORA studies, the assignment of participants results in three metabolically distinct clusters, with cluster 3 representing the group of participants with the most unfavorable metabolic characteristics. Individuals of cluster 3 are further characterized by the highest incident disease occurrence during follow-up; they also reveal the most unfavorable diet with significantly lowest intakes of vegetables, dairy products, and fibers, and highest intakes of total, red, and processed meat.
The three metabotypes originally identified in an Irish population are successfully reproduced. In addition to this validation approach, the observed differences in disease incidence across metabotypes represent an important new finding that strongly supports the metabotyping approach as a tool for risk stratification.
先前的研究基于空腹血清中甘油三酯、总胆固醇、高密度脂蛋白胆固醇和葡萄糖水平，采用k均值聚类分析确定了三个代谢同质的个体亚组（“代谢型”）。目的是重现这些发现，并通过饮食习惯和疾病发生率来描述代谢型组。
将KORA F4研究中的1744名参与者和KORA FF4研究中的2221名参与者，通过最小化欧氏距离分配到先前确定的三个代谢型聚类中。在两项KORA研究中，参与者的分配均产生了三个代谢不同的聚类，聚类3代表代谢特征最不利的参与者组。聚类3的个体在随访期间疾病发生率最高，这进一步表明了其特征；他们还呈现出最不健康的饮食，蔬菜、乳制品和纤维的摄入量显著最低，而总肉类、红肉和加工肉类的摄入量最高。
最初在爱尔兰人群中确定的三种代谢型成功重现。除了这种验证方法外，不同代谢型之间疾病发病率的差异是一项重要的新发现，有力地支持了代谢型分类方法作为风险分层工具的作用。
