Determination, intercorrelation and intraindividual stability of five steroids in hair, saliva and urine among chinese college students

Hair steroids and their ratios are believed to be reliable biomarkers reflecting the long-term exposure of circulating steroids. Hereinto, two underlying assumptions are that hair biomarkers have consistency with traditional biomarkers in saliva or urine, and good long-term intraindividual stability across a long time. However, these two assumptions have not been well verified for most of hair biomarkers except for hair cortisol. Thus, this study aimed to verify the two issues on eight hair biomarkers: cortisol, cortisone, dehydroepiandrosterone (DHEA), testosterone, progesterone, the ratios of cortisol to cortisone, DHEA and testosterone. The five steroids in hair, saliva and urine were measured with high performance chromatography tandem mass spectrometry. The results revealed that the hair biomarkers had significant correlations with the salivary biomarkers calculated by the mean area under curve (AUCg) in a matched time span (ps < 0.05) where the coefficients of correlations (r) were > 0.3 (r = 0.322-0.616) except cortisone and progesterone (r = 0.177 and 0.212, respectively). It indicated that hair biomarkers had weak to moderate consistency with salivary ones. But only three biomarkers showed the consistency between hair and urine, such as testosterone (r = 0.352, p < 0.01), progesterone (r = 0.228, p < 0.05) and the ratio of cortisol to testosterone (r = 0.502, p < 0.01). Hair biomarkers showed no absolute stability, but moderate to high long-term relative stability across 12 months where interclass correlation coefficients ranged between 0.356 and 0.678 (ps < 0.01). These results implied that the eight biomarkers in hair could retrospectively reflect their cumulative exposure in vivo. Therefore, the hair biomarkers would be considerable reliable long-term biomarkers for psychological and physiological research.