The Hypertension Prevention Trial (HPT) was a randomized unmasked multicenter trial designed to address questions concerning the feasibility and efficacy of dietary intervention in the primary prevention of hypertension. Participants in the diet treatments were given counseling to achieve and sustain changes in calorie, sodium, and/or potassium intake. Diet composition, sodium and potassium excretion, and body weight were assessed for all participants at 6-mo intervals over 3 y. This paper provides information about extent and possible sources of bias in the dietary assessment methodology used in the HPT. Estimates of nutrient intake were derived from food records, urinalysis, and measurement of body weight. Reported potassium intake increased and sodium intake declined to a greater degree during the study than did potassium and sodium excretion. Our results indicate that repeated assessments of diet, which depend upon participant recording, may not accurately represent usual diet in a dietary trial.
