An adequate iodine intake is essential to ensure good thyroid hormone production, prevent multinodular goitre and guarantee adequate brain development during foetal life and infancy. Milk and dairy products are valuable iodine sources in the human diet. The aim of the present study was to investigate the variation of iodine content in individual bovine milk samples and to evaluate the feasibility of using mid-infrared spectroscopy to predict levels of this mineral. Iodine was measured in milk samples of 147 Holstein Friesian cows, farmed in 4 commercial herds, through inductively coupled plasma mass spectrometry. Mean iodine level was 156.6 μg kg−1, was positively correlated with lactose content, and increased in early lactation. Mid-infrared spectroscopy prediction models developed using backward interval partial least squares approach exhibited a coefficient of determination in external validation of 0.57, being slightly more accurate than models developed through partial least squares algorithm.
Variation of iodine content in bovine milk and predictive ability of mid-infrared spectroscopy
Niero G.;Franzoi M.;Vigolo V.
;Cassandro M.;De Marchi M.
2020
Abstract
An adequate iodine intake is essential to ensure good thyroid hormone production, prevent multinodular goitre and guarantee adequate brain development during foetal life and infancy. Milk and dairy products are valuable iodine sources in the human diet. The aim of the present study was to investigate the variation of iodine content in individual bovine milk samples and to evaluate the feasibility of using mid-infrared spectroscopy to predict levels of this mineral. Iodine was measured in milk samples of 147 Holstein Friesian cows, farmed in 4 commercial herds, through inductively coupled plasma mass spectrometry. Mean iodine level was 156.6 μg kg−1, was positively correlated with lactose content, and increased in early lactation. Mid-infrared spectroscopy prediction models developed using backward interval partial least squares approach exhibited a coefficient of determination in external validation of 0.57, being slightly more accurate than models developed through partial least squares algorithm.Pubblicazioni consigliate
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.