You decide to conduct a cohort study to examine whether marijuana use leads to depression in young adults. You enroll a cohort of non-depressed young adults, 200 of whom use marijuana and 300 who do not, and follow them two years to measure incident depression. Over the course of the study, 50 participants were lost to follow-up (LTF).
1. What is the incidence proportion of depression in this study?
2. Calculate the appropriate measure of association for this study based on the information provided and interpret your finding in a sentence (NOTE: you do not need to calculate the 95%CI or interpret statistical significance).
3. As noted, 50 participants were lost to follow-up. You observe that most of the participants who were LTF were marijuana users. You think that those who were LTF might have been depressed and that’s why they dropped out of the study. What type of bias would you be concerned about?
4. Explain how this bias could affect the measure of association (i.e. bias towards or away from the null)?