What is hair cortisol analysis and how can it aid endocrinologists?

KEYWORDS:

• HPA axis
• cortisol
• hair
• humans
• saliva
• steroid hormones

The hypothalamus-pituitary-adrenal (HPA) axis serves vital functions across the body and the central nervous system. Its end-product, the steroid hormone cortisol, reaches virtually every single nucleated cell of our organism and is a potent regulator of a myriad of metabolic, immunological, affective, and cognitive functions. It thus comes to no surprise that an over or underactivity of the HPA axis is associated with a wide range of diseases and disorders, including atherosclerosis, sleep apnea, osteopenia, or depression [1]. Recently, low cortisol levels have been reported to be the single most important biomarker for long COVID [2].

Cortisol levels are routinely assessed in blood, urine, or saliva samples. While this is well-accepted for the clinical routine care as well as in research contexts, these spot measures provide the clinician with a rather poor index of longer-term HPA activity. However, since most disorders associated with the HPA axis develop as a consequence of enhanced or reduced cortisol exposure over prolonged periods of time, an index long-term biomarker is needed.

Like spot glucose levels, cortisol measures obtained from blood, urine or salivary samples cannot serve this purpose. They reflect only the acute endocrine situation, covering time intervals from minutes to a few hours. Large variations due to circadian rhythm, physical activity, eating, smoking, stress, and other factors make most cortisol spot measures rather useless for endocrinological diagnoses or treatment checks. Reactivity measures like cortisol levels following CRH or ACTH stimulation tests or after dexamethasone suppression of the HPA axis are much more informative. Yet, such dynamic tests are rather expensive due to their personnel and time-consuming nature. As a consequence, no long(er) term biomarker is currently assessed in hospital settings or in the GP practices.

In analogy to blood sugar regulation, what might be an equivalent of HbA1c levels for HPA axis functioning? Well, maybe cortisol in hair?!

Two decades ago, Raul et al. were the first to show that cortisol (F) and cortisone (E) can be detected in human hair [3]. Building on this, Davenport and colleagues reported that hair cortisol (HF) concentrations correlated well (r = .80) with a cortisol measure derived from multiple-days salivary cortisol assessments in a group of rhesus monkeys [4]. A number of studies followed to show that HF in humans provided meaningful endocrine data. For example, HF levels differ across the course of pregnancy, they are enhanced in Cushing’s patients, endurance athletes, or chronically stressed individuals. On the opposite, significantly lower levels of HF have been reported in patients with Addison’s disease, post-traumatic stress disorder, or generalized anxiety disorder. HF levels can also provide useful prognostic information in studies of animals and humans. For example, the likelihood of survival over a 6-month period was significantly lower for wild lemurs with high HF levels compared to individuals with low HF levels [5].

A first significant advantage of HF over any spot cortisol measure is the fact that HF is an integral measure of cortisol produced over weeks and months. During this time, cortisol (and other molecules) is continuously incorporated into the growing hair, primarily by passive diffusion. Although conclusive experimental data are still missing, it is assumed that the incorporated steroid hormone molecules are entrapped inside the hair matrix and thus do not diffuse along the hair shaft. The incorporated cortisol molecules accumulate, remain within a small segment of hair, and grow out inside the individual hair. HF levels from a given segment can therefore be viewed as an equivalent of an HbA1c value for cortisol.

The second advantage of HF lies in its unique ‘endocrine history book’ feature. As with tree rings, which show the age of a particular part of the tree stem, a given hair segment contains an aggregate amount of cortisol secreted by the body over a certain time interval in the past. Given a growth rate of scalp hair of approximately 1 cm per month, cortisol extracted from a 1 cm hair segment provides an integral of cortisol secretion over a 1-month period [6]. Extraction of the steroid from successive hair segments can therefore provide the endocrinologist with a ‘history book’ of cortisol secretion for a given patient. A stunning example of this unique feature of HF was highlighted by Manenschijn et al.: Cutting a single hair sample into segments representing a period of almost 2 years, they observed that the massively elevated cortisol levels dropped over the course of ketoconazole treatment and finally normalized after surgical removal of the tumor [7]. An accurate diagnosis of endogenous Cushing’s syndrome, including cyclic hypercortisolism, is possible using hair cortisol analysis [8].

Quantifying HF in different hair segments also allows to obtain retrospective information about HPA activity after an (e.g. traumatic) event has occurred or a disease diagnosed. Analyzing hair samples from patients admitted to hospital with acute myocardial infarction showed that the amount of cortisol secreted weeks before the event was the strongest predictor of the forthcoming infarction [9]. Due to the reactive nature of the HPA axis, spot cortisol measures (from blood, saliva, or urine) cannot serve this purpose.

Furthermore, HF can aid the endocrinologist in monitoring hormone levels in patients living in areas with limited or missing medical infrastructure. With minimal instructions and help from a family member or friend, patients can easily self-collect a hair sample and store it for days or weeks before mailing it to a laboratory or hospital. HF is stable at room temperature if stored in the dark for at least six months (unpublished data) leaving ample time to forward the specimen without jeopardizing its steroid hormone content.

While HF has been widely accepted as an excellent tool employed in various research settings, the use of hair steroid hormone measures in clinical contexts is still sparse [10]. This reluctance to use the method may be due to fears that factors, such as hair color, cosmetic hair treatments, or the frequency of hair washes could distort the HF. Fortunately, none of these (and other variables) affect HF to a significant extent that would prohibit its use as a diagnostic tool [11].

One ‘challenge’ for a successful introduction of HF assessments to routine clinical settings are the costs and availability of adequate laboratory methods. In contrast to plasma, saliva, or urine, cortisol needs to be extracted from the hair matrix in a rather time-consuming way with current protocols. The resulting volume of hair extract is rather small (usually less than 0.25 ml), so highly sensitive and specific laboratory methods are needed for quantifying HF. While some commercial immunoassays allow for a reliable HF analysis, mass spectrometry is the preferred analytical method. With unsurpassed specificity, cortisol and a number of other steroid hormones can be easily detected in the same hair extract with high throughput [12]. Quantifying cortisone in addition to cortisol can give endocrinologists a significant edge for the interpretation of HF results. For example, HF results may be false-high due to the use of cortisol-containing ointments – even if they have been used only by a family member but not the patients themselves [13]. In such cases, normal hair cortisone (HE) levels clearly flag cortisol-contaminated samples. HE may eventually turn out to be even more sensitive and specific than HF levels as indicated in a recent study of patients with mild Cushing’s disease [14]. Quantitating the chromatograms also for other steroid hormones, like aldosterone, may add another benefit for the diagnosis and choice of treatment strategy.

In conclusion, hair cortisol is a powerful ‘HbA1c-like’ marker of long-term HPA activity that allows for a unique glimpse into the endocrine past of patients. By introduction into the clinical setting and inclusion in research protocols, it can significantly aid endocrinologists in assessing cortisol secretion patterns over months, even when the patient cannot readily be examined in the clinic or GP practice.

Declaration of interest

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

Reviewer disclosures

Peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Additional information

Funding

This paper was not funded.