Background: Low levels of serum testosterone, as typically found in women and children, cannot be measured reliably by immunoassays. Our aim was to develop a sensitive assay to quantitate low serum testosterone concentrations using liquid chromatography-tandem mass spectrometry (LC-MS/MS). The results were compared to those obtained with various immunoassay techniques.
Methods: Serum testosterone levels in 70 women and children were measured using LC-MS/MS and compared with two automated, non-isotopic immunoassays, and three manual, isotopic immunoassays. Serum extraction was required only for LC-MS/MS and one of the isotopic methods.
Results: Deming regression analysis was used for comparison: the correlation coefficients were between 0.772 and 0.870, and the slopes between 0.972 and 1.365. Using Bland and Altman analysis, all the 5 immunoassays showed a positive mean difference compared with LC-MS/MS: all overestimated the testosterone levels in women and children.
Conclusion: None of the immunoassays tested proved sufficiently reliable when low testosterone concentrations (< or =3.47 nmol/L) were measured. In contrast to conventional isotopic and non-isotopic immunoassay techniques, LC-MS/MS allows the precise determination of low testosterone levels. It has adequate sensitivity and is not subject to interference from other steroids that were tested.