Nutrition studies often groups "red and processed meats" together, limiting researchers' ability to examine meat subtype-specific dietary intake. A new open-source software package called PFclassify gives researchers a standardized workflow to classify and quantify unprocessed and processed meats into separate subtypes of pork, beef, chicken, turkey and other meat and poultry.
The software is described in a peer-reviewed study published in the Journal of Food Composition and Analysis . Funding was provided by the National Pork Board and the National Cattlemen's Beef Association, a contractor to the Beef Checkoff.
Previously, the National Cancer Institute developed a standardized method to separate processed meat intake from processed red meat and processed poultry. PFclassify builds on this by further separating the processed red meat category into processed pork, processed beef, other processed red meat, and the processed poultry category into processed chicken, processed turkey or other processed poultry.
"Pork and beef are frequently studied as part of a broader red meat category," said Dr. Taylor Wallace, corresponding author on the paper and CEO of Think Healthy Group, LLC. "PFclassify gives researchers a standardized way to examine unprocessed pork, unprocessed beef, processed pork, and processed beef independently, which can support more transparent and higher quality dietary research."
How researchers evaluated the software
Researchers demonstrated PFclassify using nationally representative dietary data from the National Health and Nutrition Examination Survey, or NHANES. The study included 11,816 people age 2 or older from the 2017-2020 pre-pandemic NHANES cycle for descriptive analyses.
At the broader-category level, PFclassify produced estimates of processed meat intake comparable with the classification method established by the National Cancer Institute , with the ability to separate further into processed meat subtypes. Differences between the approaches were less than 5% for meat-category composition and less than 0.5 ounce equivalents per day for absolute intake estimates.
Supporting future research
PFclassify provides an automated and reproducible workflow for red meat (e.g., pork, beef) and poultry (e.g., chicken, turkey) intake, to help improve classification accuracy of meat subtypes when researchers assess dietary intake data from NHANES and other datasets that use USDA food codes.
"By giving scientists an open framework they can apply, test and refine, PFclassify can make it easier to compare findings across studies and build a more detailed evidence base for nutrition research," said Dr. Kristen Hicks-Roof, Senior Director of Human Nutrition, National Pork Board. She adds, "Ultimately, this work untethers foods that have fundamentally different nutritional values. In turn, this can help nutrition authorities and health professionals offer dietary guidance with far more precision than ever before."
Software Availability
PFclassify, its documentation, example data, and the analytical code used in the study are available through the public PFclassify GitHub repository .
About the National Pork Board
