A method for the rapid assessment of sample size in dietary studies.

Critical readers should be suspicious about the inability of a dietary study to discriminate between the energy intakes of two groups when small sample sizes have been used. The possibility of a false-negative (type II error) should be considered. This problem could be avoided if investigators used adequate sample sizes. A review of 26 dietary studies published in the American Journal of Clinical Nutrition between 1979 and 1981 revealed that the median "SD of energy intakes" was 525 kcal/day. This figure was used to illustrate a simple method for estimating appropriate sample sizes assuming type I and type II error probabilities of 0.05. Prospective use of this method should increase the reproducibility of conclusions drawn from dietary studies.