Quantitation of metabolites in human skeletal muscle during rest and exercise: a comparison of methods

Quantitation of muscle metabolites in vastus lateralis needle biopsies from human subjects at rest and following exercise is evaluated using three different methods of normalization of the data, i.e., total protein, total creatine, and NAD+. Coefficients of variation between patient analyses are similar when the metabolite data are normalized by any of the descriptors in resting muscle and in muscle samples obtained following vigorous exercise. When multiple samples are taken from individual patients the intrapatient coefficient of variation is less than that for interpatient comparisons. Normalization to total creatine or NAD+ yields significantly smaller coefficients of variation than does normalization to total protein in these intrapatient comparisons (P less than 0.05). Based on these comparisons, we conclude that any of these methods evaluated for normalizing metabolite content is adequate for studying human subjects, but the total creatine and NAD+ methods may offer some advantage because of the lower variability obtained with these analyses.