In the context of epidemiology, attrition refers to the loss of participants over time in a longitudinal study. Attrition can occur due to various reasons such as participants dropping out, being lost to follow-up, or passing away. This loss can significantly impact the validity and reliability of the study's findings.