Adjusting diet-outcome associations for random error: comparison of associations based on observed and estimated usual intakes.

Adjusting diet-outcome associations for random error: comparison of associations based on observed and estimated usual intakes.