Do follicles matter? Testing the effect of follicles on hair cortisol levels

Abstract

Cortisol concentrations in hair are used increasingly as a biomarker of long-term stress in free-ranging wildlife. Cortisol is believed to be integrated into hair primarily during its active growth phase, typically occurring over weeks to months or longer periods, depending on latitude. Cortisol concentrations in hair thus reflect the activity of the hypothalamic-pituitary-adrenal axis over this time. However, local, independent cortisol secretion within the skin, which includes hair follicles, may also contribute to cortisol levels in growing hair. Methodological differences between studies include the measurement of cortisol in only the hair shaft (i.e. follicle absent, as with shaved hair) versus the whole hair (i.e. follicle present, as with plucked hair). If the concentration of cortisol in the follicle is high enough to influence the overall hair cortisol concentration (HCC), this could confound comparisons between studies using different types of hair samples (hair shafts vs. whole hair) and collection methods. Here, we test the hypothesis that cortisol present in follicles influences HCC. We compared HCC in paired subsamples of hair with and without follicles from 30 free-ranging Scandinavian brown bears (Ursus arctos) and observed significantly greater HCC in samples with follicles present. The effect of follicles remained significant also with sex and age of sampled bears taken into account in a linear mixed model. Finally, we provide an overview of collection methods and types of hair samples used for HCC analysis in 77 studies dealing with stress in wild mammal species. Our findings highlight the need to unify methods of hair collection and preparation to allow for valid comparisons, and to optimize labour input in ecophysiological studies.

Keywords: Hair cortisol extraction methods; hair sampling; physiological ecology; stress.

Figure 1

Hair cortisol concentrations (HCC) in paired subsamples of 30 brown bears, collected in Sweden from 2002 to 2006, extracted with hair follicles removed compared to subsamples extracted with hair follicles intact (left), with histogram of pairwise HCC difference (follicles present–absent; right)

Figure 2

Hair cortisol concentration in samples extracted with follicles present (filled bullets) and absent (white bullets), broken down by sex, against individual age (on log-scaled x-axis). Lines show predictions of the linear mixed effect model (see Methods)—mean ± 95% confidence intervals (black solid and dotted lines for follicles present, and white line and gray shading for follicles absent)

Figure 3

The number of peer-reviewed studies reporting hair cortisol analysis in wild animal species from January 2002 to December 2018, broken down by type of hair sample used for cortisol extraction (follicles absent, present or lacking those details)