Tuesday, February 16, 4:10pm, MSB 1147 (Colloquium Room)
Speaker: Yuqing Zhang, Professor of Medicine & Public Health, Boston University
Title: What effect is really being measured? Challenges in studying risk factors for disease sequelae.
Abstract: Many observational studies have found that risk factors for incidence of a disease are not associated, or even inversely associated, with the risk of recurrence or progression of the disease. For example, overweight/obesity increases the risk of mortality; however, it improves survival among patients with cardiovascular disease (i.e., obesity paradox). We proposed that one of the explanations for such paradoxical phenomena might be owing to lack of clarity in the research question; thus, the effect estimate obtained from the study (e.g., often biased direct effect) is not what an investigator attempts to obtain (i.e., total effect). Using the causal diagram and mediation analysis we illustrate how an observational study can lead to incorrect inference in the context of an ill-defined research question. We offer a few suggestions to avoid the occurrence of paradoxical phenomenon.