Lisa Wu, an undergraduate student in agricultural and biological engineering at the University of Illinois, USA, conducts research on detecting food allergens with near-infrared spectroscopy. [Image: College of Agricultural, Consumer and Environmental Sciences, University of Illinois]
Food allergies trigger unpleasant and sometimes even fatal reactions in millions of people worldwide every year. Unexpected traces of allergens in supposedly safe foods can cause everything from gastric upsets to rhinitis and asthma attacks.
Now researchers at a US university have developed a near-infrared spectroscopic technique for detecting small amounts of allergens contaminating a type of gluten-free flour (J. Food. Comp. Anal., doi: 10.1016/j.jfca.2023.105324). The noninvasive method, which pairs NIR spectra with multivariate analysis, detects the contaminants faster than conventional techniques based on DNA analysis and without the need for trained technicians.
Keeping the gluten out
US regulations identify nine food items as major allergens: milk, eggs, fish, shellfish, tree nuts, peanuts, wheat, sesame and soybeans. (Sesame was newly added to the list in January 2023.) Cross-contamination of ingredients can occur during refining, manufacturing and packaging processes if cleaning processes are neglected.
The seeds of the quinoa plant—which are often referred to simply as quinoa—have been growing in popularity as a gluten-free food, high in protein and vitamins. Quinoa by itself does not contain any of the major food allergens. But if allergenic grains or legumes get into the equipment that grinds quinoa into flour, consumers with allergies to those contaminants may suffer.
Spectrometers and algorithms
To test the ability of a compact NIR-spectroscopy setup to detect contaminants, agricultural engineers at the University of Illinois, USA, ground quinoa into flour and mixed some of it with sesame, peanut and wheat flour at ratios that varied from 3% to 98% allergen contamination. The specimen set also included samples of pure allergen flour for comparison.
The engineers found that selecting nine spectral bands from the benchtop system and building a regression model produced slightly more accurate results than the full-spectrum or predefined-wavelength data.
The Illinois researchers scanned the flour samples with two instruments: a benchtop Fourier-transform NIR spectrometer, which scanned the wavelength range of 867 to 2500 nm, and a filter-based NIR spectrometer that scanned at 10 predefined wavelengths between 1680 and 2336 nm. Then the team developed a partial-least-square regression model to analyze the spectral plots for multiple allergens. Additional mathematical preprocessing improved the model’s ability to distinguish among the types of flour.
The engineers found that selecting nine spectral bands from the benchtop system and building a regression model produced slightly more accurate results than the full-spectrum or predefined-wavelength data. They say the technique could lead to the development of an inexpensive filter-based spectroscopic device for detecting food allergens for use in commercial kitchens, industrial settings or perhaps even the home.