Proton nuclear magnetic resonance spectroscopy is a unique method to monitor noninvasively the concentrations of cerebral metabolites. N-Acetyl-L-aspartate, the concentration of which is assumed to be stable during hypoxia, has been used to form ratios with lactate. To determine the stability of the signal from N-acetyl-L-aspartate, we used a model of graded hypoxia in rats to monitor the percentage changes from baseline of the peak heights for lactate, lipids, and N-acetyl-L-aspartate. Anesthetized adult rats were exposed sequentially to 15% and 10% O2 while proton nuclear magnetic resonance spectra were collected with a surface coil in a 7-T 89-mm-bore spectrometer. Brain lactate concentration was either increased by feeding or infusion of glucose (n = 9) or lowered by fasting (n = 7). After death the brains were removed and frozen, and the water- and lipid-soluble compounds were extracted to identify the origin of the signals. We analyzed the data both as the percentage change from baseline for heights of the lactate (1.33 ppm), lipids (1.5 ppm), and N-acetyl-L-aspartate (2.02 ppm) peaks and as the ratios of heights of the 1.33 and 2.02 and the 1.5 and 2.02 ppm peaks. Both hypoxic episodes caused a 45% decrease from baseline in the 2.02 ppm peak. During the second hypoxic episode, the 1.33:2.02 ppm peak height ratio increased significantly in hyperglycemic rats (p less than 0.05) but was unchanged in hypoglycemic rats.(ABSTRACT TRUNCATED AT 250 WORDS)