Longitudinal nutrient intake patterns of US adolescent women: The Penn State Young Women's Health Study

Purpose: To use longitudinal nutrient intake data to determine whether diet ary patterns remain consistent (or "track") as U.S. females progress from a ge 12 to 18 years. Methods: Three-day diet records were collected at regular intervals over 6 years from participants in the Penn State Young Women's Health Study. Eight y-one subjects remained in the cohort during the study period. Tracking in body weight, in dietary intake of fat, sugar, iron, vitamin C, and in a tot al dietary score (TDS) was assessed using quartile-ranking analysis, year-t o-year Pearson correlation analysis, and longitudinal linear analysis. Results: Rank analysis revealed that subjects maintained their relative qua rtile positions for body weight throughout the study period, and year-to-ye ar correlation coefficients for this variable were .93-.94. In contrast, ra nk and correlation analyses showed that the subjects did not track strongly with respect to any nutrient variable. Age 12 to 28 years correlation coef ficients ranged from r = .04 for fat intake to r = .15 for the TDS. In long itudinal linear models, slopes differed in direction and significance acros s the original quartiles for nutrient intake, indicating varying dietary tr ends over time within the study population. Conclusions: Nutrient intake patterns do not track strongly throughout adol escence among U.S. females. (C) Society for Adolescent Medicine, 2000.