The application of infrared spectroscopy to breath CO2 isotope ratio measurements and the risk of spurious results