posted on 2019-07-12, 05:00authored byEna Huseinovic, Agneta Hörnell, Ingegerd Johansson, Anders Esberg, Bernt Lindahl, Anna Winkvist
Principal component analysis model displaying separation for women during 2000-2007 related to belonging to the four clusters from Latent Class Analysis (Cluster 1, red. High-fat dairy, white bread, sugar/jam and cookies; Cluster 2, dark green. Fruit, low-fat milk and high-fiber bread; Cluster 3, light green. Bacon/sausage and fast food, and Cluster 4, blue. Pulses and tea). Panel A: Scores for components 1 versus 2. Panel B: Scores for components 1 versus 3. Further, loading plots corresponding to the score plots based on the 40 food groups responsible for the variation among the women. Location of the four clusters from Latent Class Analysis are indicated. Panel C: Loadings for components 1 vs 2. Panel D: Loadings for components 1 vs 3. (TIF 3129 kb)