Cortisol as a risk biomarker to guide recovery from substance use disorders

Substance use disorders (SUDs) are a devastating and under-treated public health concern worldwide, and are often conceptualized as chronic stress disorders (Koob et al., 2014). Indeed, a notable increase in SUD prevalence over the past few years is likely due to increased stress levels resulting from the ongoing COVID-19 pandemic (Roberts et al., 2021). Importantly, exposure to abused substances themselves, including alcohol, can activate key and integrated components of central brain and neuroendocrine stress systems. In this regard, the hypothalamic-pituitary-adrenal (HPA) axis represents a critical mediator of the adaptive stress response, facilitating production and release of the major glucocorticoid cortisol. As stress exposure serves as a major precipitator of relapse, cortisol has the potential to function as a readily accessible risk biomarker for maladaptive environments and behaviors that trap individuals into the SUD cycle (Edwards et al., 2015).

Several studies have discovered associations between dysregulated HPA axis function and relapse risk across multiple abused substances, including alcohol (Blaine and Sinha, 2017). The term “dysregulation” is frequently used to describe the complex relationships between excessive or diminished cortisol levels and HPA axis responsiveness associated with the manifestation and successful treatment of various psychiatric disorders, including SUDs. Such complications are further compounded in investigations of cortisol as an integrative biomarker in the context of highly co-morbid conditions such as alcohol use disorder (AUD) and post-traumatic stress disorder (Szabo et al., 2020). Nevertheless, given the central role of stress to the establishment and progression of SUDs, additional detailed research is warranted to untangle the most important interactions between cortisol function and key stages of the SUD cycle.

To this end, a recent study (Maddox-Rooper et al., 2022) investigated relationships between salivary cortisol levels, stress exposure, adverse childhood experiences (ACEs), and completion of an abstinence-based residential recovery program. The researchers hypothesized that decreased cortisol levels, less stress, and fewer ACEs would be associated with increased stay in the recovery program or program completion. This hypothesis is consistent with existing literature that recognizes the link between increased stress and risk of relapse, particularly in the context of alcohol use disorder (AUD) (Blaine and Sinha, 2017), but also in other SUDs such as opioid use disorder (OUD). For example, increased ACE scores are associated with an increased risk of relapse in patients undergoing treatment for OUD (Derefinko et al., 2019). Analyzing multiple stress variables (salivary cortisol and stress exposure) alongside ACE scores together in a cohort of men in an abstinence-based SUD program thus provides a unique glimpse into the interactions of different stress variables and relapse propensity.

In line with their hypothesis, Maddox-Rooper and colleagues found that participants who dropped out of the program before 90 days had significantly higher initial cortisol levels compared to participants who remained in the program for ≥ 90 days. These results build on previous findings, which suggest a connection between increased stress and self-reported mental distress and relapse potential (Eddie et al., 2021). Another study of predominately male subjects indicated that alcohol-dependent patients with high cortisol:ACTH ratios (an index of adrenal sensitivity) were twice as likely to relapse faster than patients with low cortisol:ACTH ratios after discharge following inpatient treatment (Sinha, 2011).

Some important limitations and considerations of the Maddox-Rooper study include the method of obtaining cortisol measures, exclusion of female participants, and a small sample size. Although salivary collection is one of the easiest methods for obtaining cortisol measures, it has several potential disadvantages, including daily fluctuations and a diurnal release pattern for cortisol. To account for this, the authors asked participants not to eat prior to testing and took twice-daily samples. However, additional baseline sampling may be advantageous in clarifying the relationship between cortisol and relapse risk. There is evidence that experimental designs measuring cortisol should examine mean results across three days of samples to achieve maximum reliability (Segerstrom et al., 2014). Future investigations might also consider examination of other sources of cortisol, such as hair detection. Examination of free cortisol levels in hair segments is free of circadian confounds and can serve as a longitudinal, retrospective marker of HPA axis activation and potential dysfunction across stages of psychiatric disease, including in direct comparison to recent fluctuations in alcohol use (Price and Nixon, 2021).

It is also important to note that this study exclusively examined parameters in male participants due to restrictions of the treatment program. Sex differences may increase women’s vulnerability to SUD and ovarian hormones may directly influence HPA axis responsiveness. Moreover, sex differences in reward behaviors are well-characterized, with females generally experiencing increased subjective euphoria in response to abused substances, more rapid progression to SUDs, and increased likelihood of relapse during abstinence periods (Becker et al., 2017). Inclusion of female participants in future studies is of the utmost importance to further dissect the contributions of sex to HPA axis function and risk of relapse during SUD recovery.

As relapse risk is perhaps the most fundamental obstacle to long-term SUD therapy, identifying the most salient and perhaps understudied adverse psychosocial factors (e.g., low socioeconomic status, racism) that may predispose individuals to abandoning therapy and returning to substance use is of the utmost importance. Even after identifying these key factors, establishing accessible and viable biomarkers of these complex stimuli and states would be valuable in mitigating prospective and longitudinal relapse risk. Cortisol may be one integrative biometric to gain insight into recovery outcomes, allowing for early or adjunctive pharmacological interventions in patients with heightened state or trait cortisol levels. Indeed, much work has been done in affective disorder research indicating that only patients with high baseline cortisol levels benefit from treatment with cortisol synthesis inhibitors (e.g., metyrapone and ketoconazole; Lombardo et al., 2019). Similar pharmacological strategies are further supported for SUDs by a study that found that cannabidiol (CBD) both decreases salivary cortisol levels and reduces craving and anxiety in abstinent OUD subjects (Hurd et al., 2019). Moreover, mifepristone, a glucocorticoid receptor (GR) antagonist, displays strong promise for preventing relapse in the context of AUD by blocking craving and escalated drinking in human subjects (Vendruscolo et al., 2015). These benefits of mifepristone may also extend to treating OUD-related symptomatology (Carmack et al., 2022).

There may also be potential in expanding upon the functional nature of cortisol physiology to extend its utility as a biomarker of SUD and relapse risk. The phosphorylation status of GRs in circulating peripheral blood mononuclear cells (PBMCs) has been proposed as a biomarker of affective disorder status (Simic et al., 2013). As GRs are transcription factors, assessment of associated epigenetic markers of cortisol function could also be informative (Volkow et al., 2015). Additionally, how cortisol differentially impacts downstream physiological functions of this wide-ranging stress hormone, including immune and metabolic markers, could add combinatorial depth to its predictive utility given the contributions of these physiological systems to AUD risk and recovery (Crews et al., 2017; Wiers et al., 2021). For example, a recent study employed a mediation analysis to understand interactions between metyrapone (cortisol synthesis inhibitor), cortisol, and inflammatory cytokine levels in the context of treatment-resistant depression (Strawbridge et al., 2021). Longitudinal ex vivo assessment of the GR sensitivity of leukocytes has also been employed to track affective disorder outcomes (von Zimmerman et al., 2021).

Based on the considerations above, the complex physiology of SUDs and their often-inextricable co-morbidities has led many to propose the urgent need to develop more precise and personalized therapeutic strategies (Burnette et al., 2022). These directions will require a more comprehensive understanding of how diverse individuals experience stress and are impacted by stress-related pathophysiology. In this regard, there is little doubt that highly relevant and integrative biomarkers such as cortisol will continue to play a central role in future efforts to achieve these ambitious diagnostic and treatment aims. Future work will likely need to consider and take advantage of the multiple dimensions of glucocorticoid synthesis, bioactivity, and physiological function to actualize individualized therapeutic goals.