LAUSR

Abstract Background: Chronic, cumulative stress has been shown to negatively impact on health and wellbeing. Indigenous Australians often experience multiple stressful events across the life course. While hair samples have been the most often used chronic stress biomarker, limitations exist based on availability, cultural acceptability and possible influence of acute stress on levels. Fingernails samples have recently been explored with promising results. This study examines the correlation between cortisol and cortisone levels in hair and fingernails and their association with blood levels in an at-risk population. Methods: Cross-sectional data (2013-2015) obtained from 459 Indigenous and 117 non-Indigenous young adults aged 21-27. Directly collected height and weight used to calculate body mass index (BMI; kg/m2). Fingernail samples were clipped directly into a plastic bag and scalp hair samples cut from the posterior vertex area. Hair and fingernail samples were fragmented and then twice extracted with 1.5ml methanol as reported by Sharpley et al 2009. Methanol was evaporated under vacuum and then the residue redissolved in 100ul of methanol and analysed by LC/MS/MS using a Shimadzu UPLC and 8050 Mass Spectrometer. Log linear relationship between (i) cortisol and cortisone; (ii) sample methods was assessed by Pearson’s correlation and multivariable regression. Results: Sample numbers varied; blood (Indigenous 410; non-Indigenous 108), hair (Indigenous 333; non-Indigenous 84), and fingernail (Indigenous 189; non-Indigenous 67). Median cortisol levels were lower than cortisone in hair (3.3 vs 5.7 pg/mg) and fingernail (4.3 vs 13.8 pg/mg), but higher in blood (71.6 vs 13.1 mmol/L). Cortisol and cortisone were correlated in all three sample methods. Examination of all available samples showed correlation between blood and hair cortisol (n=402, r=0.106, p=0.034) and between hair and fingernail cortisol (n=222, r=0.182, p=0.007) but not between fingernail and blood (n=242, r=-0.005, p=0.94). However, when limited to those who provided all three samples (n=211), only a correlation between hair and fingernail cortisol remained (r=0.184, p=0.007). The significant association between hair and fingernail was evident on adjustment for sex, Indigeneity and BMI. Similar results were seen in cortisone levels. Conclusions: Collection of fingernails provided an easily conducted, viable alternative to hair in this multicultural cohort. The correlation between cortisol and cortisone levels in hair and fingernail suggests both samples are reflective of chronic stress. Although not seen in the restricted cohort analysis, the possible weak correlation of hair with blood levels supports the need for further exploration of the mechanism of glucocorticoid incorporation in hair, as well in fingernails.