A new method, the time-varying effect model (TVEM; Tan et al., 2012) can flexibly estimate when disparities are greatest and when crossovers occur. Because TVEM allows for estimation of curves that do not require a specified parametric form, it can identify precise periods of change, such as ages when differences by race/ethnicity are significant. When nationally representative data and weights are used in TVEM, analyses can provide precise estimates of age-varying trends for particular population subgroups. TVEM has been used to understand gender and racial/ethnic differences in substance use among adolescents and young adults (Evans-Polce et al., 2014). The current study applies TVEM to a nationally representative sample of U.S. adults to examine age-varying disparities in SUDs by gender and race/ethnicity across the adult lifespan (ages 18–90).
