Cover image: Thirteen calibration curves from the method development. Created by graphic designer Maria Jacobsson at LiU-tryck.

No abstract

Background Testosterone is thought to be incorporated in growing hair strands so that specific hair segments reflect average free hormone concentrations from the corresponding time period. However, the exact mechanisms of hormone integration in scalp hair have not yet been established and it is not known how testosterone is stored in the hair segments over time. The aim of this study was to investigate the stability of testosterone concentrations in hair as it grows and to determine if segmental hair analysis can be used as a retrospective testosterone diary. Methods Thirty men and 40 women provided two hair samples and 16 saliva samples during a period of three months. Hair growth between the two samplings was measured. Hair samples were cut into 10 mm segments resulting in three segments from the first sampling and six segments from the second sampling. Hair samples were pulverised and extracted with methanol. Hair testosterone concentrations were analysed using an in-house radioimmunoassay. Salivary testosterone was analysed using a commercial enzyme-linked immunosorbent assay (Demeditec). Results The results demonstrated that there is a degree of segmental hormone conservation over time (rho = 0.405–0.461, p < 0.001, n = 66–67), but also highlighted three potential confounders. Firstly, testosterone concentrations were higher in distal hair segments (mean concentration ratio most distal by most scalp-near was 1.55, SD 0.70), which may be due to continuous hormone integration from sebum and changes in hair matrix composition. Secondly, more frequent hair washing stunted the increase in testosterone concentrations in distal segments (rho=-0.404, p = < 0.001, n = 66). And lastly, intra- and inter- individual variability in hair growth rate influenced the temporal resolution along the hair, although mean growth rate was indeed 30.0 mm for three months. In a multiple regression model the biological sex, natural hair colour, and relationship status were significant explanatory variables to hair testosterone concentrations. Conclusions The current results indicate that repeated hair sampling near the hair roots during a study may be preferable to analysing concentration changes between proximal and distal segments within the same hair sample. Also, hair testosterone analysis needs to be adjusted for sex and the natural hair colour.