Changes in the attenuation of near infrared spectra by the healthy adult brain during hypoxaemia cannot be accounted for solely by changes in the concentrations of oxy- and deoxy-haemoglobin.

It has been suggested that changes in oxidised cytochrome c oxidase concentration ([oxCCO]) measured using cerebral near infrared spectroscopy (NIRS) may be algorithm artefacts. We examine the change in near infrared (NIR) attenuation by the healthy adult brain (n = 10) during hypoxaemia. Broadband spectroscopic data were collected during normoxia, and hypoxaemia. The UCLn algorithm was used to fit (a) oxy- (HbO2) and deoxy-haemoglobin (HHb) spectra (2 component fit), and (b) HbO2, HHb and oxidised-reduced cytochrome c oxidase difference spectra (3 component fit) to the mean change in NIR attenuation between baseline and hypoxaemia. The sum of squares of the residuals was 100 x 10(-7) OD2 for the 2 component fit and 8 x 10(-7) OD2 for the 3 component fit, and the two sets of residuals differed from each other (p = 0.0003). We compare experimental and simulated data and suggest that the 2 component residuals indicate a change in [oxCCO]. Changes in near infrared attenuation by the healthy adult brain during hypoxaemia cannot be accounted for solely by changes in oxy- and deoxy-haemoglobin concentrations. Including [oxCCO] in the algorithm improves its fit quality. These data suggest that changes in cerebral cytochrome c oxidase redox occur during hypoxaemia and that they can be detected using NIRS.