ARS | Publication request: Discontinuation of data processing step: Salt adjustment on designated foods likely to be home prepared

Submitted to: Worldwide Web Site: Food Surveys Research Group
Publication Type: Review Article
Publication Acceptance Date: June 7, 2012
Publication Date: July 18, 2012
Repository URL: http://www.ars.usda.gov/Services/docs.htm?docid=18352
Citation: Sebastian, R.S., Enns, C.W., Steinfeldt, L.C., Goldman, J.D., Moshfegh, A.J. 2012. Discontinuation of data processing step: Salt adjustment on designated foods likely to be home prepared. Worldwide Web Site: Food Surveys Research Group. Available: www.ars.usda.gov/Services/docs.htm?docid=18352.

Interpretive Summary: The ability to estimate sodium intake accurately is of public health importance. Methods used in What We Eat In America (WWEIA), the dietary intake interview component of the National Health and Nutrition Examination Survey (NHANES), have evolved over time with constant efforts to improve their ability to reflect the food and nutrient intakes of the U.S. population. Salt adjustment in home-prepared foods, a step in the post-collection processing of survey data, is a procedure that has been used for over 25 years as a way of accounting for respondents’ reports of using salt only occasionally or less often in home food preparation. A number of considerations have called into question the appropriateness and value of continuing the procedure. These considerations include an increase in the purchase of fully-prepared foods to which the consumer is unlikely to add more salt in cooking and results from a study that suggest that the USDA Automated Multiple-Pass Method (AMPM) produces sodium estimates comparable to those estimated by urinary sodium when the salt adjustment procedure is not applied. This report describes (a) the basis for and the process of salt adjustment, (b) the rationale for discontinuing it, and (c) the impact and implications of its discontinuation. In addition, baseline estimates for tracking changes in sodium intakes are provided. This report will be of intense interest to a relatively narrow audience composed mainly of nutrition analysts and policymakers. Ultimately, the information in this report will be of benefit on a national basis, since it provides researchers with appropriate baseline estimates to use in monitoring progress toward reducing sodium intakes in the United States population, a critical public health focus.

Technical Abstract: The objective of this report is to describe (a) the basis for and implementation of a data processing step called salt adjustment that was performed on designated foods in USDA dietary intake surveys from 1985 through 2008, (b) the rationale for discontinuing the step, and (c) the impact and implications of its discontinuation. As implemented in What We Eat In America (WWEIA), the dietary intake interview component of the National Health and Nutrition Examination Survey (NHANES), salt adjustment was a post-collection data processing step that was based on the respondent’s reply to a question about the use of salt in household cooking or food preparation. It was applied only to foods likely to be home prepared, i.e., foods that (a) generally have salt added during preparation and (b) were obtained from the store. For individuals who reported that salt was used occasionally or less often in cooking or food preparation in their household, salt adjustment removed some or all of the salt attributable to home preparation. Reasons for discontinuing this procedure include secular changes that call into question the appropriateness of using store purchase as a proxy indicator of home food preparation; a decrease in the proportion of food obtained from the store and, therefore, eligible for salt adjustment; and research indicating that USDA’s Automated Multiple-Pass Method (AMPM) for the 24-hour dietary recall provides valid sodium intake estimates without application of the salt adjustment step. WWEIA, NHANES 2007-2008 is the final data release to contain sodium data that were salt adjusted; the step was not applied to WWEIA, NHANES 2009-2010 data and will not be performed in future survey cycles. To provide appropriate baseline estimates for tracking the success of initiatives aimed at lowering sodium intakes, this report includes 2007-2008 sodium intake estimates (mean intakes, percentile distributions of usual intakes, and percentages of the population with intakes above tolerable upper intake levels) calculated both with and without the salt adjustment processing step. In addition, files listing the salt adjustment levels for eligible foods in survey cycles from 2003-2008 are provided for use by analysts.