Task 1: Key Concepts about Estimating Prevalence of Supplement Use Using Proportions

Prevalence estimates (percent of population who reports supplement consumption) are of interest in nutritional epidemiology and surveillance. Using a national cross-sectional survey such as NHANES, prevalence estimates of supplement use among the U.S. population and among population supgroups can be generated.

To calculate proportions and standard errors, it is necessary to use software that takes into account the complex survey design of NHANES when determining variance estimates. If the standard errors are not needed, you simply could use a SAS procedure, i.e., proc freq with the weight statement.

Supplement data are collected from the household interview.  For further information on developing a dataset with supplement data, see Module 3 (Dietary Data Structure & Contents), Task 4, Key Concepts About Dietary Supplement Files.  It is important to note that beginning in 2007-08, dietary supplement data were collected on the 24-hour recall.  Different statistical techniques may be used with these data, other than those described in this tutorial.