Which potential bias is most important to address when interpreting a study that reports people who consume many phytochemical-rich foods have lower rates of several chronic diseases?

• A. Lead-time bias that affects nutritional cohort studies more than cancer screening studies and always causes overestimation of benefit for dietary exposures.
• B. Healthy user confounding, because people who eat phytochemical-rich diets often engage in other health-promoting behaviors (exercise, non-smoking), so the apparent protective effect may not be solely attributable to phytochemical intake.
• C. Differential recall bias that only occurs in randomized controlled trials where participants forget assigned diets at random, which artificially inflates associations in cohort studies.
• D. Attrition bias where participants who eat more phytochemical-rich foods are more likely to drop out, thereby strengthening the observed association in favor of no effect.

Answer: (B)