Featured Articles: Two Types of TVEM

Time-varying effect modeling (TVEM) is a flexible approach that can be used to answer different types of questions using different types of data. Two articles in a recent issue of Drug and Alcohol Dependence demonstrate the range of possibilities for TVEM. In one, a group of researchers led by Michael Mason, associate professor of psychiatry and director of the Commonwealth Institute for Child & Family Studies at Virginia Commonwealth University, examined the time-varying effects of a smoking intervention using ecological momentary assessment (EMA). In the other, Megan Schuler, Marshall J. Seidman Fellow in Health Care Policy at Harvard Medical School, and her co-authors use data from Add Health to examine how the relationship between depression and substance use changes across adolescence and young adulthood. Both articles use The Methodology Center’s TVEM SAS macro for analyses.

In the first article, “Time-Varying Effects of a Text-Based Smoking Cessation Intervention for Urban Adolescents,” the authors examined EMA data from a six-month intervention that sent text messages to 200 urban adolescents who were trying to quit smoking. The authors used TVEM to examine whether the intervention had an impact on the association between stress and craving over time. They found that, during the second and third months, the association between stress and craving was weaker for the intervention group than it was for the control group. They also found that the intervention helped steadily reduce craving over time.  The authors of the article include Methodology Center Investigators Stephanie Lanza and Michael Russell.

Open the “smoking cessation intervention” article. (Journal access required.)

In the second article, “Age-Varying Associations Between Substance Use Behaviors And Depressive Symptoms During Adolescence And Young Adulthood,” the authors used TVEM to examine how heavy episodic drinking (HED), daily smoking, and marijuana use related to depressive symptoms from ages 12 to age 31. HED was associated with depressive symptoms during adolescence, but not afterwards. Both marijuana use and daily smoking were associated with depressive symptoms at most ages from 12 to 31. Megan Schuler, a former postdoctoral fellow in the Prevention and Methodology Training (PAMT) program, worked with Methodology Center Investigators Sara Vasilenko and Stephanie Lanza on the article.

Open the “age-varying associations” article. (Journal access required.)


Mason, M., Mennis, J., Way, T., Lanza, S., Russell, M., & Zaharakis, N. (2015). Time-varying effects of a text-based smoking cessation intervention for urban adolescents. Drug and Alcohol Dependence157, 99-105. doi: 10.1016/j.drugalcdep.2015.10.016

Schuler, M. S., Vasilenko, S. A., & Lanza, S. T. (2015). Age-varying associations between substance use behaviors and depressive symptoms during adolescence and young adulthood. Drug and Alcohol Dependence,157, 75-82. doi: 10.1016/j.drugalcdep.2015.10.005

Bookmark the permalink.