Reported mean values for urinary 3-methoxy-4-hydroxyphenylglycol (MHPG) for both depressed patients and controls vary substantially between laboratories. The present study describes methodological sources of variance affecting urinary MHPG values. Although intra-assay variance is small (coefficient of variation (CV) = 4.6%), the interassay variance is high (CV = 14.3%) and suggests caution in the interpretation of small group differences. The gas chromatographic (GC) method in this laboratory produces values highly correlated (r = 0.98) with those obtained by a reference gas chromatographic-mass spectrometric (GC-MS) method. The absolute values obtained by GC are 20% higher than those by GC-MS--a problem that may be shared to a varying extent by all GC assays.
In conclusion: methodological differences do contribute to some of the variation in reported values of urinary MHPG.