Each cross-sectional curve was expressed relative to the same age used for its corresponding longitudinal group.
Cross-sectional and longitudinal models differed in both magnitude and, at younger ages, direction. At age 20, males in the classic event were 17.1% faster cross-sectionally versus 4.7% slower longitudinally relative to the age-46 reference. The contrast was smaller among males in the freestyle event than the classic event, but still quite dramatic.
Similar trends were observed for females in the freestyle event.
At older ages, longitudinal estimates generally indicated greater deterioration in performatance as compared to cross-sectional estimates. These contrasts show that cross-sectional analyses may not reliably represent within-person age changes in performance.