The stability of arsenobetaine in baby foods under different experimental conditions is evaluated. Total arsenic was analyzed by electrothermal atomic absorption spectrometry, and the speciation of arsenicals was carried out by coupling liquid chromatography to hydride generation–atomic absorption spectrometry. The highest arsenic levels in the analyzed baby foods corresponded to those containing plaice (2 to 3 μg/g). The speciation data indicated that arsenobetaine, a nontoxic species, was the only arsenical present in the baby foods analyzed at levels between 0.2 and 3 μg/g. Two different procedures for extracting arsenicals from baby foods, involving a water-methanol-chloroform mixture and enzymatic hydrolysis, were tested, and similar results were obtained. Furthermore, the arsenobetaine levels remained unchanged when the baby foods were stored for different times or when the samples were freeze-dried, thus confirming the stability of arsenobetaine and the nonappearance of other arsenic species by interconversion. The reliability of the procedure was checked by analyzing a certified reference material.
