Longitudinal cohort data are often analyzed with cross-sectional methods that obscure their complexity, introducing bias and missing key time-dependent relationships. Too often, such data are treated as fixed across the entire study period, rather than capturing how exposures, outcomes, and covariates evolve in their definitions, distributions, and influence on associations over time. Measurement tools may change, biological states transition within individuals, and biomarkers that reflect the impact of environmental stressors also vary across age. Recognizing these nuances is essential to avoid bias and ensure longitudinal methods yield valid evidence to inform policy and improve outcomes.
This symposium will highlight strategies to address these challenges by integrating work from environmental exposures, substance abuse disorders, epigenetics, and cancer research.