ACSPRI Conferences, RC33 Eighth International Conference on Social Science Methodology

Font Size:  Small  Medium  Large

Conversion of a SAS method for nutrient intake into equivalent R method

Michelle Gosse

Building: Law Building
Room: Breakout 7 - Law Building, Room 028
Date: 2012-07-12 03:30 PM – 05:00 PM
Last modified: 2012-04-20

Abstract


Food Standards Australia New Zealand (FSANZ) is a bi-national government agency that administers the Australia New Zealand Food Standards Code (the Code). FSANZ uses food intake information from population surveys to estimate the distribution of usual intake of specific nutrients, in particular to determine the percentage of the population with a usual daily intake below the Estimated Average Requirement (EAR) and also the percentage with a usual daily intake above the Upper Level of Intake (UL). These percentages are important because the EAR is the daily nutrient level estimated to meet the requirements of half of the healthy individuals in a particular life stage and gender group, and the UL is the highest average daily nutrient intake likely to pose no adverse health effects to almost all individuals in the general population. The nutrient data assists FSANZ in determining whether new food manufacturing/processing techniques that may reduce the nutritional composition of foods should be permitted, as well as identifying nutrients where fortification could be considered in order to lift usual intakes. The National Cancer Institute (NCI) in the USA has developed a set of SAS macros that represent the latest techniques to use for estimating nutrient intake distributions (see http://riskfactor.cancer.gov/diet/usualintakes/macros.html), however FSANZ uses Stata and R. R was identified as the package of choice for the NCI macros conversion, and the initial work has been to convert the method for a single dietary component (http://riskfactor.cancer.gov/diet/usualintakes/macros_single.html). The conversion process will be used as the practical example to cover learning to program in R, when coming from another statistical analysis background, with emphasis on data manipulation. Comparisons between the SAS code and the R code will be given. The main sources of help on this project, which include email lists, books, and websites, will be provided. FSANZ also had the support of the Australian Bureau of Statistics in this project, who provided the SAS datasets produced by the SAS macro, and the importance of this assistance will be described. The project has also required the development of an instruction manual to train nutrition intake specialists in the R code, and a comparison will be given between the level of detail needed in an instruction manual versus that provided in typical code comments. The importance of including the end user in the development of the instruction manual will be briefly covered.