A study of repeatability of dietary data over a seven-year period.

A repeatability study of dietary intake has been carried out using 94 control subjects from a previously reported case-control study of diet and colorectal cancer. Dietary histories readministered on average seven years after an initial interview were used to estimate intake of various macro- and microdietary components. Comparisons of current intake with intake of seven years previously yielded good reliability for macrocomponents among subjects reporting no dietary change during the time period. Reliability was poorer for microcomponents and for subjects reporting a dietary change. Comparisons were also made between diet measured seven years previously with subjects' recall of that diet. It is concluded that current diet is at least as reliable a measure of past diet as that assessed by a questionnaire addressed to the past period.