Hypothalamic-pituitary-adrenal axis activity in post-traumatic stress disorder and cocaine use disorder

Abstract

Post-traumatic stress disorder (PTSD) is often comorbid with cocaine use disorder (CUD), but little is known about hypothalamic-pituitary-adrenal (HPA) axis function in PTSD + CUD. Here we review the clinical and pre-clinical literature of PTSD and CUD with the goal of generating hypotheses about HPA axis activity in comorbid PTSD + CUD. Low glucocorticoid (CORT) levels immediately after trauma exposure are associated with PTSD. CORT administered within 12 h of trauma exposure reduces later PTSD symptoms. Weeks-years after trauma, meta-analyses find lower CORT levels in patients with PTSD relative to never-traumatized controls; the same is found in a pre-clinical model of PTSD. In rodents, reduced basal CORT levels are consistently found after chronic cocaine self-administration. Conversely, increased CORT levels are found in CUD patients during the first 2 weeks of cocaine abstinence. There is evidence for CORT hyper-suppression after dexamethasone, high glucocorticoid receptor (GR) number pre-trauma, and increased GR translocation to the nucleus in PTSD. Hyper-suppression of HPA axis activity after dexamethasone suggests that PTSD individuals may have increased anterior pituitary GR. Given evidence for decreased anterior pituitary GR in rats that self-administer cocaine, PTSD + CUD individuals may have normal GR density and low basal CORT levels during late abstinence. Future studies should aim to reconcile the differences in pre-clinical and clinical basal CORT levels during cocaine and assess HPA axis function in both rodent models of CUD that consider stress-susceptibility and in PTSD + CUD individuals. Although additional studies are necessary, individuals with PTSD + CUD may benefit from behavioral and psychopharmacological treatments to normalize HPA axis activity.

LAY SUMMARY

• Post-traumatic stress disorder (PTSD) is often comorbid with cocaine use disorder (CUD), but little is known about the hypothalamic-pituitary-adrenal (HPA) axis function in PTSD + CUD. The current review provides a synthesis of available clinical and pre-clinical data on PTSD and CUD with the goal of generating hypotheses about HPA axis activity in comorbid PTSD + CUD. While this review finds ample evidence supporting aberrant HPA axis activity in both PTSD and CUD, it suggests that more research is needed to understand the unique changes HPA axis activity in PTSD + CUD, as well as the bidirectional relationship between stress-susceptibility and motivation to seek cocaine.

Keywords:

• Cortisol
• corticosterone
• reinstatement
• relapse
• GR
• dexamethasone

Introduction

Post-traumatic stress disorder (PTSD) is a disabling psychiatric disorder that occurs in a subpopulation of people exposed to a traumatic event. PTSD is associated with increased risk for comorbid psychiatric disorders, impairments in general health, low rates of regular employment, severe legal problems, and high economic costs (American Psychiatric Association, 2013; Breslau et al., 1991; Kessler et al., 1995; Saxon et al., 2001; Solomon & Davidson, 1997). The majority of PTSD cases persist for at least 1 year with many lasting for years (Breslau et al., 1998; Kessler et al., 1995; Mills et al., 2006).

PTSD is often comorbid with substance use disorders. Comorbid PTSD and cocaine use disorder (PTSD + CUD) is especially problematic, with 43% of cocaine dependent individuals meeting criteria for lifetime PTSD (Back et al., 2000). Despite high comorbidity rates, to date little is known about the neurobiological mechanisms of PTSD + CUD. Knowledge of these mechanisms is necessary for the advancement of prevention and treatment of the disorders, as existing methods are not fully efficacious. As such, the present review synthesizes results of human and animal studies of hypothalamic-pituitary-adrenal (HPA) axis function in PTSD and cocaine use disorder (CUD) in order to generate hypotheses about potential adaptations in HPA axis activity in PTSD + CUD. A focus on the HPA axis was chosen due to evidence of HPA axis dysregulation in both PTSD and CUD.

Symptoms of PTSD

PTSD is currently categorized under the new class of trauma- and stressor-related disorders in the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5) (American Psychiatric Association, 2013). DSM-5 diagnosis of PTSD is conditional upon exposure to actual or threatened death, serious injury, or sexual violence that was either 1) directly experienced; 2) witnessed in person as it occurred to another person; 3) learned that it occurred to a close family member or friend; or 4) experienced as repeated or extreme exposure to details, excluding exposure via the media. In addition to trauma exposure, DSM-5 diagnosis of PTSD requires the presence of 1–2 symptoms from each of the following four symptoms clusters: 1) intrusion (e.g. recurrent and involuntary memories of the event); 2) avoidance (e.g. deliberate avoidance of distressing memories or thoughts associated with the event); 3) negative alterations in cognitions and mood (e.g. blaming oneself about the cause or consequence of the event); and 4) arousal (e.g. exaggerated startle response). Similar to the DSM-IV, the DSM-5 requires that the symptoms persist for at least 30 days after exposure to the event (American Psychiatric Association, 2000, 2013). Approximately 60–75% of PTSD cases persist for at least 1 year (Breslau et al., 1998; Mills et al., 2006). For more than 30% of individuals with PTSD, symptoms persist for 10 years or longer (Breslau et al., 1998; Kessler et al., 1995).

PTSD epidemiology

Approximately 40–90% of American adults have been exposed to at least one traumatic event at some point in their lives (Breslau et al., 1998; Kessler et al., 1995; Norris, 1992; Reed et al., 2007; Saxon et al., 2001). The most common types of trauma are witnessing someone being killed or seriously injured, the sudden tragic death of a loved one, being involved in a life-threatening accident, and robbery (Breslau et al., 1991; Kessler et al., 1995; Mills et al., 2006; Najavits et al., 1998; Norris, 1992). Many adults with lifetime trauma-exposure are exposed to multiple traumas (Kessler et al., 1995). Despite the high occurrence of exposure to traumatic events, only 15–25% of individuals exposed to a traumatic event develop PTSD (Breslau, 2001; Breslau et al., 1991, 2003; Mylle & Maes, 2004). The estimated lifetime prevalence rate of PTSD is approximately 8–12% (American Psychiatric Association, 2013; Breslau et al., 1991, 1998; Kessler et al., 1995; Reed et al., 2007).

Although the occurrence of exposure to traumatic events is lower for women than men, the lifetime prevalence rate of PTSD is about twice as high in women than in men (Breslau, 2001; Breslau et al., 1991, 1998; Hidalgo & Davidson, 2000; Kessler et al., 1995; Najavits et al., 1998). Kessler et al. (1995) suggest that sex differences in the prevalence rates of PTSD may be due to the higher number of women exposed to severe and devastating traumatic events (e.g. rape). Indeed, rape is among the traumas associated with the highest risk for PTSD (American Psychiatric Association, 2013; Back et al., 2000; Breslau et al., 1998; Kessler et al., 1995; Saxon et al., 2001). However, Breslau and colleagues (1998) found that the sex difference in prevalence rates of PTSD remained nearly the same after controlling for trauma type. Thus, women may have a greater vulnerability to PTSD than men. In addition to sex, risk factors for the development of PTSD include preexisting psychiatric disorders (with substance use disorder being the focus of this review), early separation from parents, neuroticism, family history of anxiety, and a history of prior trauma exposure (American Psychiatric Association, 2013; Breslau, 2001; Breslau et al., 1991; Saxon et al., 2001; van Zuiden et al., 2011; Walsh et al., 2013).

Comorbidity of PTSD and CUD

PTSD is highly comorbid with substance use disorders (SUD). Epidemiological studies show that individuals with PTSD are at two- to sixfold increased risk for the development of SUD relative to individuals without PTSD (Breslau et al., 1991, 2003; Chilcoat & Breslau, 1998; Kessler et al., 1995; Mills et al., 2006; Reed et al., 2007). This results in SUD in approximately 24–52% of civilians with PTSD and 65–79% of combat-exposed military veterans with PTSD (Breslau et al., 1991; Chilcoat & Breslau, 1998; Kessler et al., 1995; Mills et al., 2006; Scherrer et al., 2008; Vujanovic et al., 2016). Comorbid PTSD and SUD is associated with a higher number of co-occurring psychiatric disorders, increased likelihood of having at least one chronic physical health condition, greater number of lifetime admissions to inpatient substance abuse hospitals, and poorer occupational functioning relative to PTSD alone, SUD alone, and neither disorder (Brown et al., 1995; Mills et al., 2006; Najavits et al., 1998).

While alcohol is possibly the substance most frequently abused in PTSD (Gilpin & Weiner, 2017), PTSD + CUD diagnosis is also particularly prevalent and problematic amongst individuals with PTSD. Approximately 43% of CUD patients meet criteria for lifetime PTSD (Back et al., 2000). This rate is alarming given that the estimated lifetime prevalence rate of PTSD in the general population is 8–12%. The prevalence rate of CUD among individuals with PTSD is unknown; however, approximately 34% of individuals exposed to a trauma meet criteria for lifetime CUD, a rate that is approximately twice that found in the general population (Khoury et al., 2010). Finally, one study found that of all patients undergoing treatment for CUD, 21% had comorbid current PTSD (Najavits et al., 1998).

Lifetime CUD has been linked to more severe PTSD symptomology, increased comorbidity with other psychiatric disorders, and worse treatment outcomes (Back et al., 2000; Khoury et al., 2010; Najavits et al., 1998, 2007). However, CUD is especially problematic among some high-risk PTSD subpopulations. For example, among inmates, increased severity of PTSD is associated with heightened severity of CUD, but not with severity of alcohol, marijuana, or heroin dependence (Proctor & Hoffmann, 2012). Incarcerated veterans with PTSD report more years of cocaine use than incarcerated veterans without PTSD (Saxon et al., 2001). Women with PTSD + CUD have signficantly worse social and occupational functioning, as well as more legal problems relative to women with comorbid PTSD and alcohol dependence (Back et al., 2003). Finally, hyperarousal (elevated irritability, angry outbursts and hypervigilance) was found to be greater if subjects diagnosed with comorbid PTSD and alcohol dependence also abused cocaine (Dworkin et al., 2018).

There are four major hypotheses to explain the prevalence of comorbid PTSD and SUD: 1) self-medication; 2) shared diathesis for the two disorders; 3) SUD increases the probability to encounter trauma; and 4) SUD interferes with behavioral coping strategies after trauma (for review see Breslau et al., 2003; Brown & Wolfe, 1994; Chilcoat & Breslau, 1998; Stewart & Conrod, 2008). The “self-medication” hypothesis postulates that individuals with PTSD use drugs to self-medicate their PTSD symptoms (Hruska & Delahanty, 2014). Indeed, individuals with PTSD + CUD report that their cocaine use increases when their PTSD symptoms worsen and decreases when their PTSD symptoms improve (Back et al., 2006). Parallel to these findings, stress and negative mood states are often identified as factors contributing to drug craving and relapse (Childress et al., 1994; Erb et al., 1996; Litt et al., 1990; Sinha et al., 2003). Additionally, approximately 53–84% of individuals with comorbid PTSD and SUD experience their most traumatic events prior to, or at the same time as, the onset of SUD symptoms (Kessler et al., 1995; Mills et al., 2006). The results of a longitudinal study indicate that individuals with PTSD are significantly more likely to develop subsequent SUD relative to individuals without a history of PTSD (Chilcoat & Breslau, 1998). Finally, cross-sectional and longitudinal studies of civilians and military veterans demonstrate that exposure to traumatic events in the absence of PTSD is not associated with CUD or other SUD (Breslau et al., 2003; Chilcoat & Breslau, 1998; Khoury et al., 2010; Reed et al., 2007; Scherrer et al., 2008). This suggests that SUD develops as a result of PTSD rather than trauma exposure alone and, thus, drugs are taken to alleviate PTSD symptoms but not other trauma-related outcomes.

It is possible that PTSD mediates the relationship between trauma and SUD because of shared diathesis. As such, a second hypothesis is that PTSD and SUD share common genetic and/or environmental factors. Indeed, studies of twins found genetic and environmental risk factors common to PTSD and both alcohol and drug dependence (e.g. Scherrer et al., 2008; Xian et al., 2000). Regarding SUD that leads to the subsequent development of PTSD (secondary PTSD), the third hypothesis postulates that drug use heightens risk for exposure to a traumatic event, which consequently increases risk for PTSD. Studies that have tested this hypothesis have yielded inconsistent results (e.g. Chilcoat & Breslau, 1998; Mills et al., 2006). Finally, the fourth hypothesis postulates that drug use may increase susceptibility to PTSD because drug use hinders the ability to effectively cope with the trauma and/or extensive drug abuse alters brain physiology and neurochemical systems (Brown & Wolfe, 1994; Chilcoat & Breslau, 1998).

Animal models of PTSD and CUD

Animal models of PTSD typically capture some, but not all, clinical symptoms. For example, exaggerated startle response, generalized anxiety-like behavior, contextual fear and failure to extinguish cue + trauma associations can be directly assessed in animal models of PTSD, while recapitulating cognitive changes (e.g. distorted cognitions about the cause of the traumatic event) in animals is challenging (or impossible). Likewise, animal models of CUD are limited to modeling behavioral symptoms of the disorder (e.g. cocaine-seeking), but can only approximate more complex, cognitive features of CUD such as craving or the desire to control cocaine use. Despite this limitation, animal models have advantages and are necessary for the advancement of research pertaining to these disorders. Most notably, animal models enable researchers to conduct prospective studies under very controlled conditions, assess brain areas and other biological aspects related to disease, and/or test medications for treatment. Animal models of PTSD and CUD have yielded invaluable knowledge about neurobiological mechanisms underlying these disorders. A brief review of these models is provided below.

Animal models of PTSD

Several animal paradigms capture characteristics of PTSD (for review see Whitaker et al., 2014). However, although only a subset of individuals exposed to trauma develops PTSD, most animal models of PTSD include the entire stress-exposed group as the study population. Additionally, although animal models of PTSD should simulate real-world situations, some paradigms place the animals in situations that would not be encountered in the wild such as electric shock. Animal models of PTSD should be capable of inducing long-term behavioral and biological changes in the animal upon exposure to a brief stressor (Yehuda & Antelman, 1993), but many PTSD paradigms expose animals to long-term stressors. To the authors’ knowledge, only one model satisfies these conditions: the “cut-off behavioral criteria” (CBC) model.

The “CBC” model is an animal model of PTSD developed by Cohen et al. that has been extensively used to identify physiological differences between stress-susceptible and resilient animals (Cohen & Zohar, 2004; Cohen et al., 2003, 2005, 2006, 2012). In this model, rats are exposed to a single 10- to 15-min predator scent stressor (PSS). Based on cutoff behavioral criteria, rats are categorized as displaying extreme behavioral responses (EBR; PTSD-like) or minimal behavioral responses (MBR; resilient) to tests of unconditioned anxiety-like behavior 7 days post-exposure, which is equivalent to 1-month post-exposure in humans (Cohen et al., 2012). “PTSD-like” behaviors are characterized by anxiety-like behavior in the elevated plus maze and arousal in the acoustic startle response paradigm. Notably, one characteristic of PTSD is hyper-reactivity to unanticipated stimuli, such as loud noises (American Psychiatric Association, 2013). “Resilient” behaviors are characterized by low levels of anxiety-like behavior and fear in response to the tests, as evidenced by 1) little time spent in the closed arms, and several entries into the open arms, of the elevated plus maze and 2) low mean startle amplitude and habituation to the acoustic startle test. Studies using the “CBC” model have found that approximately 25% of rats show “PTSD-like” behaviors (Cohen & Zohar, 2004; Cohen et al., 2003, 2012; Schwendt et al., 2018), an incidence rate similar to PTSD in trauma-exposed humans. Moreover, similar to human behavior following trauma, “PTSD-like” rats consistently show long-lasting symptoms of anxiety-like behavior and hyper-arousal (Cohen & Zohar, 2004; Cohen et al., 2012) and the CBC criteria predicts which rats will show anxiety-like behavior in another test of unconditioned anxiety – the light-dark box (Shallcross et al., 2019) and conditioned fear in the PSS environment (Schwendt et al., 2018; Shallcross et al., 2019).

Animal models of CUD

Various animal paradigms model characteristics of CUD (for review see Epstein et al., 2006; Lynch et al., 2010). In the “operant cocaine self-administration” model, animals are placed in an operant chamber where responses (e.g. lever-pressing) result in the delivery of cocaine. Responses are often paired with drug-paired cues (such as, a light and a tone). Responding for cocaine must be maintained above the responding rate in control conditions (such as saline administration). Notably, this model is found to be a valid animal model of cocaine use and has provided invaluable knowledge about brain regions mediating cocaine-seeking (Lynch et al., 2010). A modification of the self-administration model is the “cocaine extended access” model, which gives some rats short-access to cocaine (ShA) and other rats long-access to cocaine (LgA). LgA rats progressively increase cocaine consumption and self-administer more cocaine than ShA rats (Ahmed & Koob, 1998; Mantsch et al., 2003). As such, this model is thought to capture the transition from controlled cocaine use to cocaine addiction (Ahmed & Koob, 1998). Relapse to cocaine use after a period of abstinence is also an important characteristic of CUD. In the “cocaine reinstatement” model, following a period of cocaine self-administration, animals undergo extinction training wherein cocaine is no longer delivered upon responses (e.g. lever-pressing), thus resulting in a decline in responses. Next, animals undergo reinstatement testing wherein stimuli (e.g. drug-paired cues) are delivered. If the stimulus results in increased responding relative to responding during extinction, “reinstatement” of the drug-seeking response has occurred. This model is found to be a valid animal model of cocaine relapse (Epstein et al., 2006; Lynch et al., 2010). Other animal paradigms are also useful for the study of CUD (e.g. the “conditioned place preference” model), but they have not been utilized to examine HPA axis activity addressed here and will not be discussed further.

Criteria for inclusion in the present review

The goal of the present review is to synthesize the clinical and preclinical PTSD and CUD literature regarding HPA axis activity to identify potential common treatment targets or features that predispose the development of both disorders. To do so, we restrict our discussion of preclinical results to those using animal models of PTSD that separate stressed animals into “susceptible” and “resilient” groups with distinct long-lasting fear and anxiety-like behavior. Studies that examine the effects of stress on HPA axis using the entire stress-exposed group as the study population are not discussed, as they do not advance our understanding of how the HPA axis could influence the development of PTSD in a subset of trauma-exposed individuals. For animal studies of CUD, the present review includes results obtained by the animal models described above. Animal studies in which animals do not voluntarily self-administer cocaine are not discussed here. To keep the focus on CUD, clinical studies in which participants had comorbid dependence on substances other than alcohol or nicotine are not discussed.

HPA axis function in healthy individuals

The HPA axis, a neuroendocrine system, responds to a variety of stimuli. One such stimulus is stress, which is defined as a circumstance that threatens homeostasis (de Kloet et al., 1998). Upon activation of the HPA axis, the parvocellular cells of the hypothalamic paraventricular nucleus synthesize and release corticotropin-releasing hormone (CRH; also known as corticotropin-releasing factor or CRF) and its co-secretagogue vasopressin into the primary plexus of the median eminence. From there, CRH and vasopressin travel along the hypophyseal portal system, to reach the secondary plexus of the anterior pituitary where they act synergistically to stimulate corticotroph cells to produce the polypeptide precursor proopiomelanocortin (POMC). POMC is then cleaved into several active peptides, including adrenocorticotropic hormone (ACTH; also known as corticotropin). In turn, ACTH circulates in the blood of the secondary plexus and is carried to the adrenal cortex. Binding of ACTH to its receptors within the zona fasciculata of the adrenal cortex promotes the local synthesis and release of glucocorticoids into the bloodstream. In humans and other primates, the main glucocorticoid secreted by the adrenal cortex is cortisol, while in many other species (including rodents and birds) it is corticosterone. Both glucocorticoids are abbreviated here and elsewhere as “CORT”. CRH and ACTH are secreted in bursts, with the frequency of bursts showing a circadian rhythm (Baum & Grunberg, 1995). As a result, under basal conditions, human CORT levels peak at approximately 8AM and decline throughout the day to reach the lowest levels around midnight (Baum & Grunberg, 1995). Under conditions of stress, CORT reaches its peak levels in the blood 15–30 min after the onset of the stressor and then declines to pre-stress levels after 60–90 min (Baum & Grunberg, 1995; de Kloet et al., 1998, 2005).

Once CORT is released into the bloodstream, most of it binds immediately with protective carrier proteins, such as cortisol binding globulin (CBG, also known as Transcortin). CBG facilitates CORT transportation throughout the body and prevents it from being metabolized. It should be noted that CBG production is stimulated by estradiol (Feldman et al., 1979), indicating the potential for sex-differences in CORT transport and metabolism. In order for CORT to bind to its target receptor, it must uncouple from its carrier proteins. Thus, CORT that is free, or unbound to binding proteins, is thought to be the biologically active portion of the hormone. Since CORT is small and has a high affinity for lipids, free CORT can passively diffuse into saliva (Baum & Grunberg, 1995). Thus, assaying salivary CORT is a useful and convenient method for measuring free CORT. Whereas assaying salivary and blood CORT allows for the analysis of CORT levels during relatively short intervals (several minutes to 2 h), assaying urinary CORT provides a measure of CORT secretion over a longer period of time (Baum & Grunberg, 1995).

The HPA axis is regulated by a third-order negative feedback system, which involves three levels of negative feedback: long, short, and ultra-short. In the long feedback loop, CORT inhibits both CRH release from the hypothalamus and ACTH release from the anterior pituitary. In the short feedback loop, ACTH provides negative feedback to the hypothalamus to inhibit CRH release. In the ultra-short feedback loop, CRH regulates its own release from the hypothalamus.

CORT exerts its actions primarily via its two receptors: mineralocorticoid (MR, also known as Type I) and glucocorticoid (GR, also known as Type II). MR is found throughout the limbic system, especially in the hippocampus, whereas GR, although ubiquitous in the brain, is most abundant in hypothalamic CRH neurons and pituitary corticotroph cells (de Kloet et al., 1998; Reul & de Kloet, 1985). There is evidence of MR and GR co-expression in the hippocampus, amygdala, dentate gyrus, lateral septal nuclei, and some portions of the cortex (de Kloet et al., 2005). CORT has a 6- to 10-fold higher affinity to MR than GR (Reul & de Kloet, 1985). As such, approximately 90% of MR are occupied at basal levels of CORT (de Kloet et al., 1999; Reul & de Kloet, 1985). MR regulates basal CORT secretion across the circadiam rhythm and is implicated in the appraisal of, and the intitial response to, stress (de Kloet et al., 1998, 1999, 2005; Herman et al., 2012). In contrast to MR, GR are only occupied after exposure to stress, or at the diurnal peak (de Kloet et al., 1999; Reul & de Kloet, 1985). Stimulation of GR mobilizes energy resources, promotes the synthesis of newly learned information, and mediates the termination of the HPA axis via negative feedback (de Kloet et al., 1998, 1999, 2005; Herman et al., 2012). As such, the MR-GR balance is vital to homeostasis (de Kloet et al., 1998).

HPA axis function in PTSD and CUD

Although HPA axis activation is adaptive and plays a vital role in regulating homeostasis, chronic or insufficient activation of the HPA axis is maladaptive. For example, chronic HPA axis activity can reduce GR expression in the hippocampus, which decreases the system’s ability to terminate the HPA axis response (Herman et al., 2012). In addition, GR dimers stimulate CRH-producing cells in the amygdala (de Kloet et al., 2005). Thus, in a feed-forward manner, frequent and prolonged activation of the HPA axis impacts the amygdala, resulting in increased anxiety and fear (Herman et al., 2012). This is one example of how HPA axis homeostasis can influence psychopathology, such as PTSD and CUD. Here, we review the literature regarding HPA axis activity in PTSD, comparing it to that in CUD, with a focus on CORT levels, the effect of exogenous CORT administration, the dexamethasone suppression test, GR number, and GR translocation to the nucleus.

HPA axis function in PTSD – clinical data

Low urinary and salivary CORT levels measured within 15 h of trauma exposure predict PTSD diagnosis and increased PTSD symptom severity 1–6 months post-trauma (Delahanty et al., 2000; Ehring et al., 2008). Motor vehicle accident victims who develop PTSD 6 months post-accident show significantly lower serum CORT levels immediately upon hospital admission than victims with major depression 6 months post-accident (McFarlane et al., 1997). Of interest, patients who receive high doses of exogenous CORT immediately prior to ICU treatment for cardiac surgery or septic shock develop significantly less PTSD symptoms and have lower PTSD prevalence rates 6–31 months post-treatment relative to patients who receive placebo (Schelling et al., 1999, 2001, 2004). Likewise, a single IV administration of high-dose exogenous CORT within 6 h of trauma exposure (Zohar et al., 2011) or a course of low dose exogenous CORT for 10 days beginning within 12 h of trauma exposure (Delahanty et al., 2013) significantly lowers PTSD symptom severity 2 weeks to 3 months post-trauma.

CORT levels measured more than 15 h after the trauma are not consistently linked to the severity of PTSD symptoms in individual studies (Table 1). Morning and evening salivary CORT levels on the day following a motor vehicle accident are not related to later PTSD symptom severity (Ehring et al., 2008). However, the same study found that 6 months post-accident, high evening salivary CORT levels predict increased PTSD symptom severity (Ehring et al., 2008). Low morning and high afternoon salivary CORT levels 2 days post-accident (McFarlane et al., 2011), and low serum CORT levels within 3 days of rape among women with a history of prior sexual assault (Walsh et al., 2013) predict PTSD severity 6 weeks to 6 months post-trauma. Years after trauma exposure, both higher (e.g. Lindley et al., 2004) and lower (e.g. Griffin et al., 2005; Labonte et al., 2014; Wahbeh & Oken, 2013) overall daily salivary CORT levels have been found among PTSD individuals relative to individuals without PTSD and/or a history of trauma.

Table 1. Changes in CORT levels following trauma exposure.

Type of trauma	Time of CORT assessment (after trauma)	CORT change in PTSD population	Citation(s)
Motor vehicle accident	Within 15 h	Reduced (urinary and salivary)	Delahanty et al. (2000); Ehring et al. (2008); McFarlane et al. (1997)
Motor vehicle accident	24 h	No change morning (salivary) No change evening (salivary)	Ehring et al. (2008)
Motor vehicle accident	2 days	Low morning (salivary) High afternoon (salivary)	McFarlane et al. (2011)
Rape	Within 3 days	Low (serum)	Walsh et al. (2013)
Motor vehicle accident	6 months	High evening (salivary) No change morning (salivary)	Ehring et al. (2008)

While other studies investigate CORT levels after trauma exposure, they do not specify the time of assessment relative to trauma and are not summarized here. Meta-analyses have been conducted that do consider such studies (Table 2). For example, one meta-analysis concluded that overall daily CORT levels do not differ between PTSD patients and trauma-exposed controls (Klaassens et al., 2012). In contrast, other meta-analyses have found lower afternoon/evening and overall daily CORT levels in individuals with PTSD relative to control subjects without a history of trauma (“never traumatized controls”; Meewisse et al., 2007; Morris et al., 2012; Pan et al., 2020). Morning CORT levels have been found to be lower (Morris et al., 2012; Pan et al., 2018) or no different (Meewisse et al., 2007) in PTSD individuals when the control groups included subjects both with and without a history of trauma. Interestingly, Pan et al. (2018) reported that lower morning salivary CORT associations with PTSD were stronger in publications after 2007, potentially due to improved CORT assay methods. Furthermore, the same group found in a later meta-analysis of 24 h urinary CORT findings, that CORT was only lower in PTSD relative to never traumatized controls in subgroup analyses of studies employing the newer and more accurate radioimmunoassay for CORT levels (Pan et al., 2020). Data heterogeneity was reduced from 92% to 33% when considering only publications after 2007 when such assays were used. Thus, meta-analyses consistently find that overall afternoon/evening and morning CORT levels in patients with PTSD are lower than those of never traumatized controls. Some evidence indicates that trauma-exposed controls may also show such reductions, potentially dissociating CORT levels from the severity of PTSD symptoms. However, high dose exogenous CORT administered immediately after reactivating traumatic memories in combat veterans with PTSD results in reduced DSM-IV avoidant/numbing symptoms 1 week after treatment (Suris et al., 2010). Additional studies are needed to understand the limitations of CORT treatment days and weeks after trauma for reducing PTSD symptoms.

Table 2. Results of CORT – PTSD meta-analyses.

Control group	Type of CORT assessment	CORT change in PTSD group	Citation
Trauma-exposed	Overall daily	No change	Klaassens et al. (2012)
		Reduced	Pan et al. (2020)
Trauma-exposed	Afternoon/evening	No change	Morris et al. (2012)
Trauma-exposed	Morning	No change	Morris et al. (2012)
Never-traumatized controls	Overall daily	Reduced	Meewisse et al. (2007); Morris et al. (2012); Pan et al. (2018)
Never-traumatized controls	Afternoon/evening	Reduced	Meewisse et al. (2007); Morris et al. (2012)
Never-traumatized controls	Morning	No Change	Meewisse et al. (2007)
		Reduced	Morris et al. (2012)

Dexamethasone (DEX) is a synthetic steroid that through its binding to GR in the anterior pituitary, suppresses ACTH secretion and terminating HPA axis activity. The CORT response to DEX serves as an indicator of GR-mediated negative feedback inhibition and is a reliable indicator of HPA axis function. In PTSD, CORT levels in response to low-dose DEX 2 days after a motor vehicle accident do not differ between individuals who do and do not subsequently develop PTSD (McFarlane et al., 2011). However, the same study found that 1 month post-accident, victims with PTSD displayed a greater CORT suppression response to DEX than victims who did not develop PTSD. There was a non-significant trend in the same direction 6 months post-accident, with non-significant effects possibly due to the substantial decline in the number of individuals with PTSD 6 months post-accident. Combat veterans, terror-related bereaved spouses, and domestic violence survivors with PTSD show enhanced CORT suppression to DEX relative to their counterparts and healthy controls without PTSD (Griffin et al., 2005; Pfeffer et al., 2009; Yehuda et al., 1993, 1995). There is also evidence for no differences in CORT suppression following DEX in PTSD and trauma-exposed controls (Klaassens et al., 2012; Lindley et al., 2004). However, these studies included individuals with comorbid depression in the PTSD study populations. Individuals with comorbid PTSD and major depression show non-suppression in response to DEX (Griffin et al., 2005; Kudler et al., 1987). Taken together, CORT hyper-suppression to DEX is evident in PTSD not comorbid with major depression.

Given that GR mediates HPA axis negative feedback, and GR translocation to the nucleus results in gene transcription, GR number and translocation serve as indicators of recent HPA axis activity. Increased GR in peripheral mononuclear cells (PBMCs) before military deployment significantly predicts the development of PTSD post-deployment, with a 7.5-fold increased risk for post-deployment PTSD per pre-deployment GR increase of 1,000 (van Zuiden et al., 2011). When assessed only after trauma exposure, GR numbers in PTSD are inconsistent. Specifically, veterans and civilians with PTSD are found to have more (van Zuiden et al., 2011; Yehuda et al., 1991, 1995), similar (de Kloet et al., 2007; Yehuda et al., 1995), or less (de Kloet et al., 2007; Matić et al., 2013) GRs in PBMCs relative to veterans and civilians without PTSD. Two studies assessed GR translocation to the nucleus in PTSD and found that, unlike healthy controls and combat veterans without PTSD, combat veterans with PTSD show a reduction in cytosolic PBMC GR after 0.5 mg (but not 0.25 mg) DEX compared with baseline GR (Yehuda et al., 1995). Taken together, although more research is necessary, these results suggest increased GR translocation to the nucleus in PTSD. Assessment of GR translocation immediately after trauma exposure among individuals who subsequently develop PTSD warrants attention.

In summary, substantial clinical evidence exists supporting HPA axis alterations in PTSD. These changes exhibit circadian variation and are impacted by the time since trauma. There is also evidence for HPA axis alterations in trauma-exposed individuals who do not go on to develop PTSD symptoms, potentially dissociating such changes from symptom severity.

HPA axis function in PTSD – preclinical data

Preclinical studies exposing rodents to a predator scent stressor and then employing a “CBC” model to segregate rats into stress-Susceptible (“PTSD-like”) and Resilient have assessed CORT levels at different times post-PSS and at different times of the day. Late waking CORT assessed 7–8 days after PSS is higher in Susceptible Sprague-Dawley rats relative to Resilient rats and unexposed controls (Brodnik et al., 2017; Cohen et al., 2007; Kozlovsky et al., 2009). Three weeks after PSS, early waking CORT was found to be decreased in both Susceptible and Resilient rats relative to Controls (Schwendt et al., 2018), consistent with some human data comparing PTSD to never traumatized and trauma-exposed controls (Klaassens et al., 2012; Meewisse et al., 2007, Morris et al., 2012).

A strain of rats with greater susceptibility to “PTSD-like” behaviors (Lewis rats) show no significant elevations in late-waking plasma CORT levels 7 days after PSS relative to pre-exposure levels, whereas rats less susceptible to stress (Sprague-Dawley and Fischer F344 rats) show a significant increase in plasma CORT levels from pre-exposure to 7 days post-exposure (Cohen et al., 2006). Similar to the effects seen in the human population, in all strains of rats tested, administration of high-dose exogenous CORT 1 h before (Cohen et al., 2006), immediately after (Cohen et al., 2008) or 1 h after (Daskalakis et al., 2014; Zohar et al., 2011) predator exposure reduces “PTSD-like” behaviors and the prevalence rate of the “PTSD-like” phenotype 7 days post-exposure relative to placebo. Notably, these effects are long-lasting, as evidenced by the significant attenuation of freezing responses to a trauma cue 31 days post-treatment (Cohen et al., 2008). Using the “CBC” model, Kozlovsky et al. (2009) found that “PTSD-like” rats have higher nuclear GR in the dentate gyrus and the CA1 and CA3 sub-regions of the hippocampus (but not the frontal cortex) 7 days post-exposure relative to “resilient” rats and unexposed-controls. Thus, like humans, rodents with increased long-term anxiety-like responses to a single stressor display altered HPA axis negative feedback, and exogenous CORT reduces PTSD symptoms.

HPA axis function in CUD – clinical data

In cocaine users, acute IV cocaine injections results in dose-dependent increases in plasma CORT levels (Elman et al., 1999). Basal CORT levels in CUD patients are higher than those of non-CUD controls throughout the day during early abstinence and normalize to control levels after 2–3 weeks of inpatient abstinence (Contoreggi et al., 2003; Fox et al., 2009; Vescovi et al., 1992). This suggests that CORT levels re-regulate during prolonged periods of abstinence. There may be sex differences in the re-regulation of HPA axis activity during abstinence, as evidenced by higher levels of salivary CORT throughout 28 days of abstinence in treatment-seeking CUD women relative to healthy controls (Fox et al., 2008). Basal early waking CORT levels in inpatient, abstinent CUD patients are positively correlated with the frequency of cocaine use prior to treatment (Buydens-Branchey et al., 2002). Thus, while CORT levels are consistently found to be increased in early abstinence from cocaine in CUD, the role of HPA axis negative feedback on such levels is currently unknown as the DEX test has not been conducted in this population.

Pertinent to self-medicating PTSD symptoms with cocaine, personalized stress-cue imagery results in concurrent increases in cocaine craving and heightened CORT levels among treatment-seeking patients with CUD (Sinha et al., 1999, 2003). Increased CORT levels in response to stress-imagery are associated with an increased amount of cocaine used per occasion during relapse (Sinha et al., 2006).

HPA axis function in CUD – preclinical data

In agreement with the dose-dependent increases in plasma CORT levels produced by IV cocaine injections in cocaine users (Elman et al., 1999), cocaine self-administration in rats (Galici et al., 2000; Mantsch & Katz, 2007; Mantsch et al., 2003) and monkeys (Broadbear et al., 1999a, 1999b) increases acute CORT levels when assessed after the self-administration session. Basal early waking CORT levels are reduced in rats that chronically self-administer high dose cocaine as early as 24 h after cessation of self-administration and remain below those of cocaine-naïve controls for at least 24 days (Hadad et al., 2016; Mantsch & Goeders, 2000; Mantsch et al., 2003, 2007), indicating long-lasting HPA axis hypoactivity as a result of chronic cocaine use. Rats that undergo LgA cocaine self-administration also display reduced basal CORT in the late waking and early evening phases, while ShA rats only display reductions in the early waking period (Mantsch et al., 2003). Thus, while CORT levels are consistently increased in early abstinence (1–2 weeks) in human cocaine users, during the same time frame, CORT levels are consistently found to be reduced in cocaine abstinent rats. At this time, it is not clear what underlies this lack of agreement between clinical and preclinical results. Given that patterns of cocaine use in CUD individuals vary from chronic daily use to episodic use (American Psychiatric Association, 2013), potential differences in basal CORT levels based on variable patterns of cocaine use in CUD could be explored in future studies.

Exogenous CORT administration in rats with intact adrenal glands enhances the acquisition of cocaine self-administration in rats, but has no effect on the maintenance of self-administration (Mantsch et al., 1998; Mantsch & Katz, 2007). Adrenalectomized (ADX) rats treated with diurnal CORT replacement display similar intake during cocaine self-administration as intact rats, indicating that acute CORT increases produced by cocaine are not necessary for ongoing self-administration (Mantsch & Katz, 2007). However, ADX rats not given diurnal CORT replacement reduce ongoing cocaine self-administration and show attenuated acquisition of cocaine-seeking, indicating that circulating CORT levels influence the motivation to seek cocaine (Goeders & Guerin, 1996; Deroche et al., 1997). Metyrapone, a CORT synthesis inhibitor, decreases ongoing cocaine self-administration (Goeders & Guerin, 1996) and cocaine seeking during a relapse test (Piazza et al., 1994) in rats. When rats were adrenalectomized (or received sham surgery) after completing 10–14 days of cocaine self-administration, ADX had no effect on cocaine-primed reinstatement of cocaine seeking but prevented stress-primed reinstatement (Erb et al., 1998). Thus, CORT is necessary for the acquisition and stress-primed reinstatement of cocaine self-administration, but not for established cocaine self-administration or cocaine-primed reinstatement of cocaine-seeking.

While DEX-induced changes in CORT among CUD individuals have not been assessed, DEX attenuates CORT secretion induced by cocaine self-administration in rats (Mantsch et al., 1998) and monkeys (Broadbear et al., 1999a). High-dose DEX fully attenuates the CORT response to cocaine self-administration, reduces basal CORT to undetectable levels, and inhibits the acquisition of cocaine self-administration (Mantsch et al., 1998), providing further evidence that CORT is necessary for the acquisition of cocaine self-administration. However, although a lower dose of DEX similarly reduces basal CORT to undetectable levels, it does not alter ongoing cocaine self-administration in monkeys (Broadbear et al., 1999b), in agreement with a lack of effect of exogenous CORT administration to intact rats on established cocaine self-administration. Thus, acute CORT manipulations do not alter ongoing cocaine self-administration in rats or monkeys; however after ADX and chronic CORT deprivation, CORT is necessary for the maintenance of cocaine self-administration. Given that DEX acts on GR at the pituitary, it is possible that stimulation of GR outside of the pituitary and/or elevated CRH (and not CORT per se) is necessary for the maintenance of cocaine self-administration. Indeed there is evidence that extrahypothalamic CRH is necessary for escalation of cocaine intake (Boyson et al., 2014) and stress-primed reinstatement of cocaine seeking (Vranjkovic et al., 2018).

GR number has not been assessed in humans with CUD. In rats, long sessions of chronic high-dose cocaine self-administration results in decreased GR mRNA expression in the anterior pituitary relative to rats with short-access to cocaine 10 days after cessation of self-administration (Mantsch et al., 2003). Long-access cocaine self-administration also results in decreased GR protein expression in the dorsomedial hypothalamus, including the paraventricular nucleus, but not in the pituitary gland, ventromedial hypothalamus, dorsal hippocampus, ventral subiculum, medial prefrontal cortex or amygdala (Mantsch et al., 2007). Since GR expression in the hippocampus is similar in rats given short- or long-access to cocaine (Mantsch et al., 2003), it is likely that short-access to high-dose cocaine self-administration also increases hippocampal GR expression relative to saline self-administration. There are no known human or animal studies of GR translocation in CUD.

Taken together, in early abstinence from cocaine (1–2 weeks), CUD patients display increased basal CORT levels throughout the day. Conversely, there is consistent evidence for reduced basal CORT levels in rodents during the first 2 weeks of abstinence from cocaine self-administration. Rats that self-administered cocaine for longer periods per day exhibit reduced CORT throughout the day, indicating a dose or time-dependent effect of cocaine on CORT levels in the rodent. In humans, greater cocaine intake predicts greater increases in CORT levels in cocaine abstinence. Thus, while rodents and humans differ in the directionality of CORT changes after cocaine, both exhibit dose dependent changes. DEX attenuates cocaine-induced CORT secretion in rats, indicating that GR feedback is intact despite changes in protein expression after cocaine self-administration. While DEX-induced changes in CORT among CUD individuals have not been assessed, it would be hypothesized that such negative feedback is not effectively dampening CORT levels. This should be explicitly tested in CUD individuals. Just as individual differences in stress susceptibility has been studied in relation to CORT levels in rats, the same should be done in regards to individual differences in the motivation to seek cocaine.

Comorbid PTSD + CUD

While studies characterizing the epidemiology and clinical symptomology of comorbid PTSD + CUD have been published, none have investigated HPA axis adaptations. While many preclinical studies have been conducted on the interaction between stress and cocaine-seeking, few have segregated rodents into stress-Susceptible (“PTSD-like”) or Resilient phenotypes. We have found that stress-susceptibility alters CORT levels in rats with a history of high dose cocaine self-administration (Hadad et al., 2016). In this study, rats were exposed to PSS, then assessed for anxiety-like behavior in a CBC model. Rats then self-administered cocaine for 17 days followed by 10–11 days of abstinence and a cue-primed reinstatement test. We found that, relative to cocaine-naïve rats, unstressed cocaine-experienced rats exhibited reduced CORT levels. However, PSS-exposed rats that self-administered cocaine displayed CORT levels that were no different than cocaine-naïve controls, regardless of susceptibility to anxiety-like behavior (Hadad et al., 2016). Thus, the re-exposure to the cocaine-associated context and cues selectively increased CORT levels only in stress-exposed rats. We have also found that only stress-Susceptible rats display enhanced cue-primed reinstatement of cocaine-seeking, while stress-Resilient rats display levels of seeking no different than controls (Schwendt et al., 2018). It remains to be studied whether CORT manipulations in stress-Susceptible or Resilient rats influence cocaine intake or the reinstatement of cocaine seeking. Based on the literature summarized here, we hypothesize that the motivation to seek cocaine (which acutely increases CORT levels) is greater following stress/trauma due to reductions in basal CORT. To test this hypothesis, future studies should administer exogenous CORT to rats with a history of stress and assess cocaine intake and seeking, as has been done in stress-naïve rodents.

Treatment implications

Among individuals with PTSD + CUD, approximately 41% report a preference for treatment that concurrently addresses both PTSD and CUD symptoms, but treatment efforts often focus on only one of the two disorders (Back et al., 2006). Discrete stress-paired cues facilitate the acquisition and relapse of cocaine seeking in rats (Goeders & Guerin, 1996; Mantsch & Katz, 2007; Mantsch et al., 1998; Sinha et al., 1999, 2003, 2006). Thus, it is possible that if trauma cues or context similarly increase CORT levels in individuals with PTSD + CUD, psychotherapy and pharmacotherapy aimed at reducing trauma-cue-associated increases in cocaine craving and CORT levels have the potential to reduce the acquisition of, and relapse to, cocaine self-administration. However, this preclinical work was done in rats that were not previously stressed (CUD only), and our work shas shown that re-exposure to the stress context decreases cue-primed reinstatement when the re-exposure occurs immediately prior to the test. Thus, more work needs to be done to characterize the cocaine and CORT relationship in CUD + PTSD. Only one study (Liberzon et al., 1999) is known to assess CORT levels after exposure to a trauma-related discrete cue, but CORT levels were assessed within a few minutes of the onset of the trauma-cue, which is not the appropriate time to capture changes in circulating CORT levels. Thus, studies that assess the effect of trauma-cue exposure on CORT levels are warranted. In the classic fear conditioning model, exposure to a cue previously paired with shock increases CORT levels in rats (Sullivan et al., 2004). Because DEX attenuates cocaine-induced CORT levels and inhibits cocaine self-administration, pharmacotherapies that stimulate pituitary GR may prevent CUD in PTSD individuals. However, given that hyper-suppression in response to DEX is evident in PTSD, future research on DEX administration as a treatment for PTSD + CUD is necessary. Finally, administration of exogenous CORT within 15 h of trauma exposure is promising for the prevention of PTSD and, thus, PTSD + CUD. However, animal work indicates that the administration of CORT to someone in early abstinence from CUD may increase cocaine use. Exogenous CORT may be more helpful as a prevention for CUD in PTSD patients experiencing chronic reduced basal CORT levels, in that those with low CORT levels may seek out stimuli such as cocaine to increase CORT.

Conclusions

This review synthesized results from human and animal studies of PTSD and CUD and found HPA axis alterations in both disorders. Specifically, the results reviewed here indicate that CORT levels are low immediately after trauma in individuals who subsequently develop PTSD, and administration of high dose exogenous CORT during this time buffers against PTSD. While individual studies characterizing CORT levels more than 15 h after trauma report inconsistent findings, meta-analyses of such studies find that when PTSD patients are compared to non-trauma exposed controls, and not to trauma-exposed controls, lower basal CORT levels are found in both morning and evening. This data is in agreement with that of a single rodent study reporting that early waking CORT is decreased in both Susceptible and Resilient rats relative to unstressed Controls, 3 weeks after PSS exposures (Schwendt et al., 2018).

Acute cocaine administration increases CORT in both humans and rodents, and CORT is necessary for the acquisition of cocaine self-administration in rats. CORT alone has not been found to be sufficient for the maintenance of cocaine self-administration. However, this work was done in a rodent model that did not consider comorbidity with PTSD (e.g. tested unstressed rats). In a rodent model of PTSD + CUD, basal CORT levels are decreased 10 days after cessation of cocaine self-administration only in unstressed rats, but are increased in stress-susceptible rats that had self-administered cocaine (Schwendt et al., 2018). Future studies should aim to characterize basal and cocaine-seeking associated CORT levels during self-administration, abstinence, and reinstatement. Studies should also test whether manipulations of CORT levels influence cocaine seeking in stress-susceptible rodents as they do in unstressed rodents as reviewed here. Human studies of CUD show increased basal CORT during abstinence while rodent studies consistently show decreased basal CORT days-weeks after cocaine self-administration. More studies should be conducted in CUD to understand whether this is due to potential co-morbidities such as depression in abstinent cocaine users, as well as long-lasting changes in basal CORT during cocaine abstinence (post 3 weeks).

PTSD individuals without comorbid major depression show hyper-suppression of CORT after DEX, potentially explaining persistent CORT decreases in this population. Such an effect has not been investigated in rodent models that also show reduced basal CORT levels weeks after the stressor. DEX attenuates cocaine-induced CORT increases in rodents and primates, indicating that GR-mediated negative feedback is intact. CORT levels in response to DEX among CUD individuals are not known, and this should be investigated to determine whether increased CORT levels in the weeks following initiation of cocaine abstinence are due to a failure of negative feedback in the HPA axis.

Although more research is necessary, high GR number may be a preexisting vulnerability factor for subsequent PTSD. The results regarding post-trauma GR number in individuals with PTSD are inconsistent. Decreased GR expression in the anterior pituitary and trend-level increases in hippocampal GR number is evident in rats that chronically self-administer high-doses of cocaine. GR number has not yet been assessed in early abstinence from cocaine use in CUD, but should be in order to understand the elevated CORT levels at this time. There is support for increased GR translocation to the nucleus in PTSD, but GR translocation in CUD has not been studied.

The results presented here indicate that altered HPA axis activity is consistently found in both clinical and pre-clinical studies of PTSD and CUD. However, no studies have yet been done in a comorbid patient population, and limited work has been done using animal models of PTSD + SUD. The existing animal studies reveal that stress-susceptibility interacts with cocaine to produce unique changes in CORT levels and enhanced cocaine-seeking. However, pre-clinical studies have thus far been limited to investigating how individual differences in stress-susceptibility influence later cocaine-related behavior. The converse should also be investigated: does cocaine-related behavior (e.g. dose attained, motivation to seek) predict susceptibility to later stressor? As long-term decreased CORT levels are observed in PTSD, and acute cocaine increases CORT, this supports the self-medication hypothesis and indicates that psychotherapy and pharmacotherapy that aim to normalize HPA axis activity may be effective in preventing and treating PTSD + CUD.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by a pilot grant from the Center for OCD, Anxiety, and Related Disorders (COARD) at University of Florida.

Notes on contributors

Natalie A. Hadad

Natalie Aviva Hadad earned a M.A. in Developmental Psychology with Summa Cum Laude Honors from Teachers College Columbia University. She earned her Ph.D. in Psychology from the University of Florida in August 2016. She is currently an Assistant Professor of Psychology at Santa Fe College in Gainesville, FL.

Marek Schwendt

Marek Schwendt earned his Ph.D. from P.J. Safarik University, Slovakia, followed by post-doctoral training in neuroscience at the Medical University of South Carolina. He is currently an Assistant Professor of Psychology at the University of Florida, where his research focuses on the neurobiology of cognitive, emotional, and motivational deficits produced by stress and drugs of abuse using animal models.

Lori A. Knackstedt

Lori A. Knackstedt earned her Ph.D. in Psychology at the University of California, Santa Barbara. She next conducted a post-doctoral fellowship in the Neuroscience Department of the Medical University of South Carolina. She is now an Associate Professor of Psychology at the University of Florida. Her work focuses on long term adaptations to the glutamate and dopamine neurotransmitter systems in response to stress and/or substance use.