A systematic review of hair cortisol in healthy adults measured using immunoassays: Methodological considerations and proposed reference values for research.

Hair cortisol concentration (HCC) has shown remarkable promise as a stable, non‐invasive measure of systemic cortisol; however, despite methodological advances, the value that would typically be seen in healthy adults has not been established. Therefore, we sought to review the relevant literature to determine a reference value for HCC in healthy (i.e., non‐clinical) adults. To this end, we conducted a systematic review of the PubMed, Scopus, and CINAHL databases for studies that measured healthy adult HCC using immunoassay methods, given that these are the most widely accessible analytical tools. To be eligible, studies were required to have been published in English, to have provided relevant descriptive statistics (i.e., means and standard deviations), and to have used a healthy adult human sample. We found 17 studies that met our inclusion criteria; the reports involved 1348 participants with a mean age of about 38 years. Since we identified a large amount of between‐study heterogeneity, we completed a random‐effect meta‐regression analysis and found that test kit vendor was the only significant variable of the model. As a result, when using methodologies from traditional finite mixture distributions to determine reference values for mean and elevated HCC in individual healthy adults, we calculated these estimates for each of the major test kit vendors. Future work will need to determine whether our estimated reference values need to be modified, and these efforts will be greatly assisted by studies that account for potential moderating factors, such as age, sex, and ethnicity. Due to ease and cost, immunoassays are widely used for hair cortisol analysis. However, guidelines for typical hair cortisol concentration (HCC) in healthy adults are lacking. Therefore, we completed a systematic review of relevant studies and used methodologies from traditional finite mixture distributions to determine reference values for mean and elevated HCC in individual healthy adults. Given that our random‐effect meta‐regression analysis found test kit vendor to be the only significant variable of the model, we prepared estimates for each of the major test kit vendors. [ABSTRACT FROM AUTHOR]