Dietary Patterns Track from Infancy to Preschool Age: Cross-Sectional and

Background: Although it has been suggested that dietary patterns emerge early in life, less is known about the extent to which they track through the toddler and preschool ages. Objectives: The objectives of this study were to derive cross-sectional dietary patterns at 2, 3, and 5 y of age and assess their correlations and to derive multi–time point dietary patterns from ages 2–5 y and assess their associations with sociodemographic factors and infant feeding patterns. Methods: Depending on the age considered, analyses included 989–1422 children from the EDEN (Étude des Déterminants pré- et postnatals précoces du développement et de la sante ́ de lENfant) mother–child cohort. Dietary intake was collectedwith the use of food-frequency questionnaires at 2, 3, and 5 y of age. Principal component analyses were applied to these data, first cross-sectionally at each age, then longitudinally accounting for the data collected at all 3 ages. Tracking between patterns was estimated by Spearman correlation coefficients and associations with either the infant feeding patterns or the demographic and socioeconomic factors were assessed with the use of multivariable linear regression analyses. Results: Overall, we derived 2 main cross-sectional patterns labeled ‘‘Processed and fast foods’ ’ and ‘‘Guidelines,’ ’ the latter being characterized by intakes approximating age-specific dietary guidelines; and 2 multi–time point dietary patterns that corresponded to consistent exposures to similar foods across the 3 ages. The first, labeled ‘‘Processed and fast foods at 2, 3, and 5 y,’ ’ was inversel