Separating causal questions from statistics with application to dietary interventions in the Project Viva cohort offspring
Virtual Advanced Methods Workshop
Tuesday, April 28, 2026
12:00 PM – 2:00 PM EASTERN
Workshop summary: Causal effects that compare interventions under which everyone in the study population strictly receives one treatment level versus another often have limited clinical or public health relevance. In this workshop, we will introduce reasoning and methods for estimating causal effects of generalized interventions that allow dependence on so-called natural treatment values, the treatment value an individual would receive in the absence of intervention. We illustrate differences between these “strict” versus “pragmatic” effect notions and their corresponding statistical methods in an application where interest is in effects of improving specific eating behaviors on insulin resistance in adolescents.
Jessica G. Young, Department of Population Medicine, Harvard Pilgrim Health Care Institute
Jessica Young is an Associate Professor in the Department of Population Medicine. She holds a secondary appointment in the Department of Epidemiology at the Harvard T.H. Chan School of Public Health. Dr. Young’s research focuses on the development and application of statistical methods for estimating causal effects of treatment strategies, including strategies that depend on natural treatment values, in real-world data with complexities that include time-varying confounding affected by past treatment, censoring, competing events, and truncation by death. She is a collaborating biostatistician and co-investigator of Project Viva, a longitudinal pre-birth cohort that began in 1999 with participants now in young adulthood.
Dr. Lan Wen is an Assistant Professor in the Department of Statistics and Actuarial Science at the University of Waterloo. Her research focuses on the development and application of statistical methods for causal inference, particularly in observational studies. These settings pose challenges such as model misspecification, time-varying confounding, and censoring or missing data. Dr. Wen is especially interested in methodological issues that arise when drawing causal conclusions without randomized treatment assignment and with incomplete follow-up. Guided by the principle that new methods should be motivated by real-world problems, her work is grounded in applications from public health and medicine, with the goal of strengthening causal inference in complex data settings to support more reliable decision-making.
Soren Harnois-Leblanc, Department of Population Medicine, Harvard Pilgrim Health Care Institute and Harvard Medical School
Soren Harnois-Leblanc is a registered dietitian, epidemiologist, and postdoctoral fellow at the Department of Population Medicine from the Harvard Medical School and the Harvard Pilgrim Health Care Institute. During her postdoctoral studies, she is investigating how lifestyle habits and genetic susceptibility influence the risk of diabetes from early childhood to late adolescence with causal inference methods through the Project Viva mother-child cohort in Massachusetts. More broadly, her research focuses on the prevention of obesity, type 2 diabetes, and future cardiovascular disease in the pediatric population.