Methods: Data were drawn from the Social and Emotional Contexts of Adolescent Smoking Patterns (SECASP) Study, a longitudinal study oversampling light and novice adolescent smokers. Initial ND was time-invariant and drawn from the baseline wave, and the outcome of smoking frequency was time-varying over 6 years. A number of control variables were also included. The predictive validity of baseline ND for smoking frequency, and the potential changes in predictive validity over time, were examined using time-varying effect models.
Results: ND had a time-varying effect on future smoking frequency, such that its predictive validity was highest initially and declined over time. However, baseline ND remained significantly associated with smoking frequency at all timepoints from baseline through 6 years.
Conclusions: These findings support previous literature showing that early-emerging ND is a strong risk factor for future smoking, and add to current knowledge about how the predictive validity of initial ND changes over time. Though baseline ND significantly decreased over time in predictive validity for smoking behavior, it remained a significant predictor 6 years into the future. These findings highlight the need for early interventions targeted at youth who experience ND symptoms at low levels of smoking.