Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 Oct 1.
Published in final edited form as: Psychoneuroendocrinology. 2021 Jul 9;132:105355. doi: 10.1016/j.psyneuen.2021.105355

Exercise-induced increases in Anandamide and BDNF during extinction consolidation contribute to reduced threat following reinstatement: Preliminary evidence from a randomized controlled trial

Kevin M Crombie a, Anneliis Sartin-Tarm a, Kyrie Sellnow a, Rachel Ahrenholtz a, Sierra Lee a, Megan Matalamaki a, Neda E Almassi b, Cecilia J Hillard c, Kelli F Koltyn b, Tom G Adams d,e,f, Josh M Cisler g
PMCID: PMC8487992  NIHMSID: NIHMS1725483  PMID: 34280820

Abstract

Introduction:

We recently demonstrated that moderate-intensity aerobic exercise delivered during the consolidation of fear extinction learning reduced threat expectancy during a test of extinction recall among women with posttraumatic stress disorder (PTSD). These findings suggest that exercise may be a potential candidate for improving the efficacy of exposure-based therapies, which are hypothesized to work via the mechanisms of fear extinction learning. The purpose of this secondary analysis was to examine whether exercise-induced increases in circulating concentrations of candidate biomarkers: endocannabinoids (anandamide [AEA]; 2-arachidonoylglycerol [2-AG], brain-derived neurotrophic factor (BDNF), and homovanillic acid (HVA), mediate the effects of exercise on extinction recall.

Methods:

Participants (N=35) completed a 3-day fear acquisition (day 1), extinction (day 2), and extinction recall (day 3) protocol, in which participants were randomly assigned to complete either moderate-intensity aerobic exercise (EX) or a light-intensity control (CON) condition following extinction training (day 2). Blood was obtained prior to and following EX or CON. Threat expectancy ratings during tests of extinction recall (i.e., initial fear recall and fear recall following reinstatement) were obtained 24hrs following EX or CON. Mediation was tested using linear-mixed effects models and bootstrapping of the indirect effect.

Results:

Circulating concentrations of AEA and BDNF (but not 2-AG and HVA) were found to mediate the relationship between moderate-intensity aerobic exercise and reduced threat expectancy ratings following reinstatement (AEA 95% CI:−0.623 to −0.005; BDNF 95% CI:−0.941 to −0.005).

Conclusions:

Exercise-induced increases in peripheral AEA and BDNF appear to play a role in enhancing consolidation of fear extinction learning, thereby leading to reduced threat expectancies following reinstatement among women with PTSD. Future mechanistic research examining these and other biomarkers (e.g., brain-based biomarkers) is warranted.

Keywords: aerobic exercise, endocannabinoids, brain-derived neurotrophic factor, fear extinction, exposure-therapy, Posttraumatic Stress Disorder

1. Introduction

Posttraumatic stress disorder (PTSD) is a highly debilitating mental health disorder that can develop following exposure to one or more traumatic events (e.g., sudden, unexpected death of a loved one, motor-vehicle accidents, physical/sexual assault, and combat exposure). Although PTSD can develop following exposure to a variety of traumatic events, epidemiological evidence suggests that interpersonal violence (IPV) exposure (i.e., physical or sexual assault) is a more potent risk factor compared to other forms of trauma (Resnick et al., 1993). One of the gold standard psychotherapeutic approaches for treating PTSD is prolonged exposure therapy, which is an exposure-based therapy that involves gradual, repeated exposure to trauma-linked cues and stimuli to promote safety learning. Prolonged exposure therapy is hypothesized to work via the mechanisms of fear extinction learning, in which previously learned fearful associations (e.g., all men are dangerous) are inhibited following the acquisition and consolidation of new safety learning (e.g., most men are not dangerous; Rothbaum and Davis, 2003). Although exposure-based cognitive therapies are among the best supported treatments for women with PTSD and are effective for some (50-60% remission rate), a significant number of patients have incomplete responses or fail to sustain extinction learning (i.e., poor extinction recall), and thus, fail to experience a clinically significant reduction in their symptoms (Resick et al., 2002; Schnurr et al., 2007). This occurrence highlights the fact that fear extinction learning is fragile, and thus, the return of fear following successful extinction learning is an unfortunate but common phenomenon (Quirk & Meuller, 2008). For instance, one of the most commonly studied aspects of extinction recall involves the return of fear following exposure to the unconditioned stimulus that instantiated the original fear learning (i.e., reinstatement). Based on these principles, laboratory-based fear extinction paradigms are widely used as models of exposure therapy, under the notion that manipulations that enhance the acquisition, consolidation, or recall of extinction learning in the lab may be promising candidates for improving treatment efficacy in the clinic (Mataix-Cols et al., 2017; Zoellner et al., 2017).

Although most experimental manipulations tested to date have been pharmacological in nature, an emerging body of evidence in preclinical animal models (Tanner et al., 2018; Roquet and Monfils, 2018) suggests that aerobic exercise delivered during or after extinction learning can boost the consolidation of fear extinction memories and improve later extinction recall (i.e., limit return of fear). If similar findings are observed in clinical samples, this would suggest that exercise may be a powerful way to improve the efficacy of exposure-based therapy for PTSD due to its ability to improve extinction recall when administered during the acquisition or consolidation of extinction learning. In order to begin to examine the translational relevance of these findings, we recently conducted a randomized clinical trial to examine whether moderate-intensity aerobic exercise performed shortly after fear extinction training improves the consolidation of fear extinction learning and subsequently prevents the return of fear among a homogenous sample of women with PTSD. Results indicated that moderate-intensity aerobic exercise administered after fear extinction learning (i.e., during the post-extinction consolidation window) enhanced cognitive indices of extinction recall (i.e., decreased threat expectancy ratings during extinction recall test 24 hrs later) compared to a light-intensity aerobic exercise control condition (Crombie et al., 2021). An important next step is to examine potential mechanisms that may mediate these effects. The critical timing aspect of exercise (i.e., exercise delivered after extinction learning), suggests that exercise alters learning processes during the consolidation of the newly formed extinction memory (Tanner et al., 2018). Accordingly, the accompanying mechanisms responsible for enhancing extinction recall and reducing fear relapse likely involve exercise-induced increases in neurochemical events that linger throughout the consolidation window.

Preclinical and clinical investigations attempting to enhance extinction learning processes by targeting several neurophysiological systems (e.g., endocannabinoids, monoamine and amino acid modulators, neurotrophic factors, glucocorticoids, neuropeptides) have significantly contributed to our understanding of the neurophysiological mechanisms of action, and as a result, have demonstrated potential avenues through which exercise may exert its effect. For instance, emerging evidence suggests that the endocannabinoid (eCB) system may be one pathway through which exercise enhances memory consolidation and extinction recall. The eCB system is an expansive neuromodulatory and immunoregulatory system that is widely distributed throughout the central and peripheral nervous systems. The eCB system primarily consists of receptors (cannabinoid type-1, CB1R; cannabinoid type-2, CB2R) and endogenous ligands (eCBs; N-arachidonoylethanolamine, anandamide, [AEA]; and 2-arachidonoylglycerol [2-AG]) that bind to the CB1 and CB2 receptors (Katona and Freund, 2012). Given that eCB signaling (specifically eCB/CB1R signaling) modulates synaptic transmission throughout the limbic system, the eCB system has emerged as playing an important regulatory role in numerous emotional memory and fear-processes including fear extinction learning and recall (Katona & Freund, 2012; Lafenêtre et al., 2007; Lutz et al., 2015). For instance, exogenous pharmacological enhancements of the eCB system have been shown to facilitate extinction learning and improve comorbid mental health outcomes (e.g., PTSD symptoms, anxiety, sleep disturbances) (Bluett et al., 2014; Jetly et al., 2015; Mayo et al., 2020a; Mayo et al., 2020b; Rabinak et al., 2014; Mariscano et al., 2002; Morena et al., 2018). Specifically, administration of CB1R agonists, and inhibitors of enzymes involved in degradation of eCBs (e.g., fatty acid amide hydrolase; FAAH) prior to extinction training enhanced both extinction learning and long-term retention of extinction memory in pre-clinical models of PTSD (Bluett et al., 2014; Mariscano et al., 2002; Morena et al., 2018), and attenuated spontaneous recovery of fear in humans (Rabinak et al., 2014). Importantly, circulating concentrations of AEA are consistently elevated following moderate-intensity aerobic exercise in adults with and without PTSD (Crombie et al., 2018; Crombie et al., 2019; Brellenthin et al., 2017), and thus, it is plausible that aerobic exercise (delivered after extinction training) may exert a similar effect on eCB signaling and extinction processes as pharmacological manipulations to the eCB system.

In addition to the eCB system, neurotrophins, particularly the brain-derived neurotrophic factor (BDNF), may play an important role in the acquisition, consolidation, and recall of fear extinction learning (Lu et al., 2008; Notaras & van den Buuse, 2020). BDNF is a neuropeptide that supports neuronal survival and growth, and enables adaptive plasticity during extinction learning, which highlights the crucial role of BDNF in fear extinction memory consolidation (Notaras & van den Buuse, 2020). Lower BDNF concentrations (in hippocampal afferents to the infralimbic PFC) have been reported in animals that have failed to successfully extinguish learned fear, while augmenting BDNF in these regions has yielded the opposite effect (Peters et al., 2010; Chhatwal et al., 2006). Similar to AEA, increases in circulating concentrations of BDNF have been reported following acute bouts of moderate-intensity aerobic exercise (Kallies et al., 2019; Heyman et al., 2012; Meyer et al., 2016). Moreover, exercise-induced increases in BDNF have been shown to mediate the beneficial effects of exercise on fear extinction consolidation and recall (Keyan & Bryant, 2019).

The dopaminergic system has also been shown to play a critical role in memory consolidation, fear extinction learning, and extinction recall (Abraham et al., 2016; Haaker et al., 2013). Preclinical animal models have revealed that administration of pharmacological agonists (targeting D1/D5 dopamine receptors) prior to or following fear extinction learning, results in decreased fear responding during tests of recall (Abraham et al., 2016; Haaker et al., 2013). In humans, administration of a dopamine precursor (L-DOPA) delivered during the post-extinction learning consolidation window resulted in improved extinction recall in non-clinical samples (Haaker et al., 2013; Gerlicher et al., 2018). Administration of L-DOPA has also been shown to decrease reinstatement (i.e., improved extinction recall) in a sample of adult women with PTSD (Cisler et al., 2020). Moreover, stimulation of substantia nigra dopaminergic neurons (classically associated with movement) replicates the aforementioned effects of acute exercise on fear extinction recall and relapse (Bouchet et al., 2018). Similar to eCBs and BDNF, the collective body of evidence examining the dopaminergic system suggests that enhancing dopaminergic signaling during the post-extinction learning consolidation window may be an effective target for enhancing exposure-based therapies in PTSD. Like eCBs and BDNF, peripheral concentrations of homovanillic acid (HVA; a dopamine metabolite associated with central levels) have also been shown to significantly increase following moderate-intensity aerobic exercise (Kendler et al., 1983).

Therefore, based on the collective body of preclinical research and our recent findings (Crombie et al., 2021) indicating that moderate-intensity aerobic exercise administered after fear extinction learning (i.e., during the post-extinction consolidation window) enhanced cognitive indices of extinction recall (i.e., decreased threat expectancy ratings during extinction recall test 24 hrs later), the current secondary analysis sought to examine potential neurochemical mechanisms that may mediate these effects in women with PTSD. Based on existing literature, and our ability to measure peripheral signals of candidate mechanisms, the current study examined circulating concentrations of eCBs (AEA, 2-AG), BDNF, and HVA prior to and following the acute bout of light-intensity aerobic exercise (control condition) or moderate-intensity aerobic exercise that occurred immediately after extinction training in order to determine if exercise-induced increases in these candidate biomarkers mediates the relationship between aerobic exercise and reduced fear responding following reinstatement. It was hypothesized that circulating concentrations of eCBs, BDNF, and HVA would significantly increase following moderate-intensity aerobic exercise, which in turn would mediate the relationship between aerobic exercise and reduced threat expectancy ratings following reinstatement among women with PTSD.

2. Materials and Methods

All procedures were approved by the Health Sciences Institutional Review Board at the University of Wisconsin - Madison and the work described has been carried out in accordance with the Code of Ethics of the World Medical Association (Declaration of Helsinki) for experiments involving humans. The participants included in the current manuscript involves a subset of participants (i.e., all participants with pre- and post-control/exercise condition blood samples were included) from a recently published manuscript (Crombie et al., 2021) that reported on fear responding (i.e., skin conductance responses and threat expectancy ratings) throughout the 3-day fear acquisition, extinction, and extinction recall protocol, in addition to mood data (i.e., mood states prior to and following control/exercise condition). Accordingly, the biomarker and mediation analyses included in the current manuscript use the same sample of participants but are independent of our previously reported findings (Crombie et al., 2021).

2.1. Study Participants

Inclusion criteria consisted of: females, age 21-50, with a current diagnosis of PTSD where the index event involved assaultive violence exposure (Cisler et al., 2015). This study focused solely on women with IPV-related PTSD, given that: 1) women are more than twice as likely to develop a diagnosis of PTSD following trauma exposure (Blanco et al., 2018; Breslau et al., 1998; Kessler et al., 2005), and 2) the probability of developing PTSD is greatest following IPV exposure, which is more common in women (Kessler et al., 1995; 2005; 2017; Scott et al., 2018; Iverson et al., 2013). Exclusion criteria primarily consisted of any condition that would make it unsafe for participants to exercise (e.g., chest pain); a current or past diagnosis of any psychotic disorder (e.g., schizophrenia); cognitive/intellectual disabilities; developmental disorders; and active substance use disorders. Psychiatric medications were not exclusionary if the medication was stable for at least 4 weeks. Acute sedatives / pain killers (e.g., benzodiazepines, opioids) and prescription stimulants (e.g., methylphenidate, amphetamines) were not permitted for the full day prior to the laboratory sessions or 2 hrs after the laboratory session (i.e., after the initial memory consolidation period). Participants were recruited from general community-based advertising including newspaper and online advertisements (e.g., Craigslist, Facebook), and flyers posted at approved locations on the University of Wisconsin campus and at health clinics throughout the greater Madison, WI area (see Crombie et al., 2021 for detailed description of exclusion criteria and recruitment strategies).

2.2. Randomization

Participants were randomized to one of two groups (moderate-intensity aerobic exercise [exercise] or light-intensity aerobic exercise [control]) using blocked and stratified randomization (using custom code in MATLAB); stratification was based on age (>35 or <= 35), whether they were currently prescribed psychotropic medication (yes or no), and menstrual cycle (regular cycle vs irregular cycle, birth control, or menopause). Participants were not provided information as to which group (exercise or control) they were randomly assigned to until the debriefing session. Randomization was successful in creating groups balanced on key demographic and clinical variables (see Supplementary Table 1).

2.3. Study Overview

This study involved four main stages: intake assessment (day 0), day 1 fear acquisition training, day 2 fear extinction training, and day 3 tests of fear extinction recall. Participants completed these stages across four separate days. Days 1-3 took place on consecutive days, with all visits occurring at 1800 CST. During the intake assessment visit (day 0), written informed consent was first obtained, and then participants were administered a series of structured clinical interviews, questionnaires, and neurocognitive tests to characterize the sample and establish that participants met inclusion/exclusion criteria. Participants must have had a current diagnosis of IPV-related PTSD as assessed via the Clinician-Administered PTSD Scale for DSM-V (CAPS-5; Blevins et al., 2015), which is considered the gold standard in PTSD assessment.

During the day 1 visit, participants first set the intensity for the unconditioned stimulus (US) which consisted of a mild electric shock delivered to the left wrist. The intensity of the shock (US) was determined by each participant with aim of selecting an intensity that was maximally uncomfortable but not painful (i.e., rating of 7/10). It is important to note that: 1) there was no significant (p=.869) group difference in the average selected shock intensity between groups (control = 1.88 ± 1.12 mA; exercise group = 1.95 ± 1.18 mA), and 2) there were no significant differences in shock intensity ratings between groups (ps = .298 to .465) or between sessions (ps = .125 to .507). In other words, the US (shock) was similarly rated as maximally uncomfortable but not painful throughout the task (day 1 acquisition learning shock intensity rating for all participants = 7.00; day 2 extinction learning shock intensity rating for control group = 6.76 ± 0.75 vs exercise group = 6.94 ± 0.64; day 3 initial fear recall shock intensity rating for control group = 6.64 ± 0.70 vs exercise group = 6.83 ± 0.79; day 3 fear recall following reinstatement shock intensity rating for control group = 6.88 ± 0.60 vs exercise group = 7.11 ± 0.60).

After the US was set during the day 1 visit, participants completed a computerized fear acquisition task, in which they were trained to associate a conditioned stimulus (CS+, one of two geometric shapes: triangle or circle) with the occurrence of the US in context A (e.g., yellow background); the remaining geometric shape (CS-) which was also presented in Context A was never paired with the US. At the end of the day 1 visit, participants were reminded to refrain from eating or drinking (other than water and a provided granola bar to be eaten precisely at 1600 CST) for 6 hours prior to the start of the day 2 visit. The dietary restriction was to ensure circulating concentrations of HVA were not altered due to diet (Kendler et al., 1983). During the day 2 visit, participants were administered a fear extinction procedure during which the CS+ and CS− were repeatedly presented in a new context (Context B; e.g., blue background), in absence of the occurrence of the US. After the day 2 extinction task, participants completed acute measures of affect and mood and had their blood drawn before being administered either exercise or control (see below for additional information). Following the exercise or control condition, participants again had their blood drawn and completed acute measures of affect and mood (see Crombie et al., 2021 for data pertaining to acute measures of affect and mood). Participants returned to the lab on day 3 to complete a fear extinction recall task. They first completed an initial recall test in which they were presented with the CS+ and CS− (10 presentations each, alternating between contexts A and B in a pseudorandom order, with no presentations of the US). After this initial recall test, participants then received a single US presentation to promote reinstatement before again completing the identical task implemented in the first run (i.e., fear recall test following reinstatement). The fear acquisition, extinction, and extinction recall task used in this study was modeled after prior investigations examining the impact of a pharmacological manipulation (i.e., L-DOPA) on fear extinction learning among healthy adults (Haaker et al, 2013) and women with PTSD (Cisler et al., 2021), although the current study utilized a 3-day design (as opposed to 2-days) which allowed us to more clearly define the impact of aerobic exercise on extinction consolidation. See Figure 1 for graphical overview of study design.

Figure 1.

Figure 1.

Open in a new tab

Graphical overview of the study design. On Day 1, participants completed a fear acquisition task in which the CS+ was paired with a shock (US) with 50% probability. On Day 2 (24hrs later), participants completed a fear extinction task in a distinct context (i.e., different background color) during which the CS+ was no longer paired with US. Immediately after completing the task, half of the participants were randomly assigned to complete 30-min of moderate-intensity aerobic exercise (EX) on a treadmill or 30-min of very-light intensity walking (CON) on a treadmill. Blood draws (to assess circulating concentrations of AEA, 2-AG, BDNF, and HVA) occurred immediately prior to and following the experimental manipulation on day 2. Participants returned on Day 3 (24 hrs later) for a fear extinction recall test alternating between acquisition and extinction contexts (i.e., initial fear recall). A single US was then presented (i.e., reinstatement procedure), followed by another recall test alternating between acquisition and extinction contexts. Threat expectancy ratings during extinction recall were provided once before the initial fear recall test, and immediately after our reinstatement procedure (i.e., prior to the second extinction recall test).

Participants’ instruction throughout the task was to indicate the identity of the stimulus (by selecting a key corresponding to the shape of the observed stimulus as fast as possible) and were not informed about any specific contingencies between stimuli and shocks. Participants also provided threat expectancy ratings during each test phase, in which they provided a 0 (not at all likely) to 10 (extremely likely) rating of how likely they believed the shock was to follow each of the stimuli. Participants were asked to provide ratings four times (evenly spaced) throughout both the acquisition and extinction phases, along with one set of ratings prior to both the initial fear recall run and the fear recall run following reinstatement on day 3. Participants were compensated $50 for completing the intake visit, and an additional $150 after completing the remaining three visits. Additional details regarding experimental procedures can be found in a previously reported manuscript (Crombie et al., 2021).

2.4. Aerobic exercise procedures

Participants randomly assigned to the exercise condition completed a 5-minute warm-up of light-intensity activity (40-55% age-adjusted maximum heart rate; MHR) on a treadmill followed by 30 minutes of inclined walking or running at a moderate-intensity (70-75% MHR). Participants then cooled down for 5 minutes at a light-intensity. Participants randomly assigned to the control condition completed a 5-minute warm up of light-intensity activity (40-55% MHR) on a treadmill, followed by 30 minutes of walking at the same intensity (40-55% MHR). Ratings of perceived exertion (RPE), and heart rate (HR, assessed via Polar HR chest strap) were continuously monitored and recorded every five minutes throughout the session.

2.5. Sample collection, processing, and assays

Blood draws using standard venipuncture procedures were carried out while participants were seated. Upon collection, plasma samples (BD Vacutainer, K3E EDTA K3 collection tubes; Greiner Bio-One, Monroe, North Carolina) were then centrifuged (4 °C at 3500 RPM) within 2 min of collection, separated into aliquots, and frozen at −80 °C until eCB extractions took place. Serum samples (BD Vacutainer, SST venous blood collection tubes, Model 367988; Greiner Bio- One, Monroe, North Carolina) were allowed to clot for 20-minutes before being centrifuged (4 °C at 3500 RPM). They were then separated into aliquots, and frozen at −80 °C until BDNF and HVA assays took place. Following lipid extraction using a solid phase column, the concentrations of eCBs were quantified using stable isotope-dilution, electrospray ionization liquid chromatography/mass spectrometry (LC-ESI-MS-MS) as previously described (Crombie et al., 2018). Enzyme-linked immunosorbent assay (ELISA) kits were used to measure the concentration of total BDNF in the serum samples according to manufacturer’s instructions (Catalog # DBNT00, R&D Systems, Minneapolis, Minnesota, USA). Samples were run in duplicate and averaged. For HVA assays, the extraction method, calibrators and mobile phase were adapted from Chang et al., 1983. The samples were analyzed on a ThermoFisher Scientific Ultimate 3000 electrochemical detector. The guard cell was set at 200mV, E1 at 500mV and E2 at −150mV. Chromeleon software (version 7.2) was used for data analysis. The intra-assay CV ranged from 2.4-3.5%. The BDNF and HVA analysis was performed at Assay Services, a core lab of the Wisconsin National Primate Research Center.

2.6. Statistical Analyses

2.6.1. Baseline Characteristics and Exercise Session Analyses

Potential group differences in age, education, depressive symptoms, total PTSD symptom severity scores (self-reported and clinical interview), treatment utilization (medication and psychotherapy), consumption of alcohol, cannabis, and cigarettes, along with exercise parameters (e.g., percentage of maximum exercising heart rate during the exercise session, RPE, treadmill speed and incline) were analyzed with independent samples t-tests.

2.6.2. Threat Expectancy Ratings Analyses

The previously reported results from our primary analyses (Crombie et al., 2021) revealed that there were no group differences in threat expectancy ratings during the initial fear extinction recall test, however, following reinstatement, the exercise group reported significantly lower ratings to both the CS+ and CS− compared to the control group. Previously reported results also indicated that the exercise group reported lower ratings during the second recall test (i.e., following reinstatement procedure) compared to the initial recall test. Importantly, the magnitude of group differences in threat expectancy ratings during reinstatement did not significantly vary between the CS− and CS+ (Crombie et al., 2021). As a result of these findings, our primary dependent variable for the current mediation analyses focused on threat expectancy ratings (collapsed across CS+ and CS−) during the second fear recall run (i.e., following reinstatement). Physiological outcomes (i.e., skin conductance responses) were not included as dependent variables in this study based on our previously reported results indicating that moderate-intensity aerobic exercise (compared to light-intensity control) did not significantly affect physiological measures during the day 3 extinction recall tests (see Crombie et al., 2021 for discussion on differential findings between physiological and threat expectancy outcomes).

2.6.3. Mediation Analyses

To test for mediation, linear mixed effects models (LMEMs) were first conducted to estimate regression coefficients for path a (i.e., effect of independent variable [group: moderate-intensity aerobic exercise or light-intensity control] on biomarker variable [post-exercise concentrations of AEA, 2-AG, BDNF, and HVA]) and path b (effect of biomarker variable on dependent variable [threat expectancy ratings following reinstatement] controlling for independent variable). In addition to estimating regression coefficients for path a, exploratory post hoc paired samples t-tests were also conducted to further examine circulating concentrations of AEA, 2-AG, BDNF, and HVA prior to and following the light-intensity control condition or moderate-intensity aerobic exercise (see Figure 2). Next, the indirect effect (path ab) was calculated as the product of regression coefficients for the predictor variables in path a and path b. The indirect effect quantifies the estimated difference in the dependent variable resulting from a one-unit change in the predictor variable through a sequence of causal steps in which the independent variable affects the biomarker variable, which in turn affects the dependent variable (i.e., indirect effect quantifies the difference between the effect of the independent variable on the dependent variable when the biomarker is controlled versus when it is not; Hayes & Rockwood, 2017). A bootstrap percentile confidence interval procedure (Hayes & Rockwood, 2017; Shrout & Bolger, 2002) involving resampling with replacement (10,000 iterations) was used to compute a 95% confidence interval around the indirect effect (using custom code in MATLAB), allowing us to test the statistical significance of the indirect effects; 95% confidence interval values not containing zero were determined to support a claim of mediation (Hayes & Rockwood, 2017). Regression coefficients for the total effect (path c; i.e., effect of independent variable on dependent variable without biomarker variables included in model), and direct effect (path c’; effect of independent variable on dependent variable when controlling for biomarker variable in model) were also estimated. Pre-exercise concentrations of biomarker variables were included as covariates in all respective models (other than the LMEM examining the total effect). Given the small sample size and the preliminary nature of this investigation, the decision to test for mediation using separate mediation models for each candidate biomarker (as opposed to testing for mediation using a multiple mediation model) was made in order to increase our power to detect significant effects (i.e., mediation models would go from having up to 26 degrees of freedom down to 17 degrees of freedom if a multiple mediation model was used).

Figure 2.

Figure 2.

Open in a new tab

Circulating concentrations of AEA (A), 2-AG (B), BDNF (C), and HVA (D) prior to and following the light-intensity control condition and moderate-intensity aerobic exercise. In addition to estimating regression coefficients for path a, exploratory post hoc paired samples t-tests were also conducted to further examine circulating concentrations of AEA, 2-AG, BDNF, and HVA prior to and following light-intensity (CON) or moderate-intensity aerobic exercise (EX). There was a significant increase (denoted *) in circulating concentrations of AEA following EX (t(13)=4.21, p = .001, d = 0.58), but not CON (t(13)=−0.66, p = .523, d = −0.16; see figure 2a); along with a significant increase in circulating concentrations of BDNF following EX (t(12)=2.39, p = .034, d = 0.92), but not CON (t(11)=−0.71, p = .494, d = −0.18; see figure 2c); and a significant increase in circulating concentrations of HVA following EX (t(12)=3.71, p = .003, d = 0.94), but not CON (t(11)=1.39, p = .191, d = 0.33; see figure 2d). There was not a significant increase in circulating concentrations of 2-AG following EX (t(13)=−.157, p =.878, d = −0.01) or CON (t(13)=0.922, p=.373, d = 0.20; see figure 2b).

3. Results

3.1. Participant characteristics

The overall study sample consisted of 35 adult women (control n=17, exercise n=18). The current study sample included all participants with blood samples available at both collection time points (before and after exercise or control), which resulted in twenty-eight participants (control n=14, exercise n=14) for the eCB (AEA, 2-AG) mediation analyses, and twenty-five participants (control n=12, exercise n=13) for both the BDNF and HVA mediation analyses (see Supplementary Figure 1). Like the overall sample, there were no significant (ps >.05) differences between groups (exercise and control) in the mediation samples for age (control M = 29.64, SD = 5.92; exercise M = 31.78, SD = 7.93), education, depressive symptoms, total PTSD symptom severity (self-reported and clinical interview), treatment utilization (medication and psychotherapy), or consumption of alcohol, cannabis, and cigarettes (see Supplementary Table 1 for means and standard deviations). As expected, based on experimental design, there was a significant group difference in the percentage of maximum exercising heart-rate, RPE, exercising heart-rate, treadmill speed and incline during the exercise session (see Table 1 for means and standard deviations).

Table 1.

Group means and standard deviations for exercise session variables.

Variable Control Exercise t-statistic P-value
Ratings of perceived exertion (RPE) 8.49 ± 1.01 12.84 ± 1.54 −8.61 < .001**
Exercising heart rate (bpm) 108.23 ± 11.19 138.96 ± 8.84 −7.88 < .001**
Heart rate percentage (% MHR) 56.96 ± 6.54 73.88 ± 1.82 −9.32 < .001**
Treadmill speed (km/hr) 3.57 ± 0.61 5.88 ± 0.89 −8.04 < .001**
Treadmill incline (% grade) 0.23 ± 0.42 1.80 ± 1.60 −3.54 .002*
Open in a new tab

Note. Bpm = beats per minute; MHR= maximum heart rate; km/h = kilometers per hour; HR = heart rate.

*

significant group difference, p <.01;

**

indicates significant group difference, p < .001. RPE ratings, exercising HR (bpm and % of MHR), treadmill speed and incline were significantly greater for the exercise group compared to the control group.

3.2. Effect of Aerobic Exercise on Threat Expectancy Ratings Following Reinstatement

Prior to examining whether post-exercise concentrations of our peripheral biomarkers mediate the relationship between our independent variable (group: moderate-intensity aerobic exercise vs light-intensity control) and dependent variable (threat expectancy ratings following reinstatement) by examining the indirect effect bootstrapped confidence intervals, we first examined the effect of independent variable on dependent variable without biomarkers included in the model (i.e., total effect, path c). As expected for the overall sample (n=35), our independent variable significantly predicted our dependent variable, as the moderate-intensity exercise group exhibited lower threat expectancy ratings following reinstatement (t(33)=−2.233, p = .032) compared to the light-intensity control group (between group d = 0.75). Although insignificant, the effect was in the same direction (and of similar magnitude as overall sample) for the eCB (t(26)=−1.50, p = .145) and BDNF/HVA (t(23)=−1.24, p = .227) mediation samples (n=28 and 25, respectively), as the moderate-intensity exercise group exhibited lower threat expectancy ratings following reinstatement (see path c of Figure 3a, 3b, 4a, and 4c) compared to the light-intensity control group (between group ds = 0.85 and 0.60, for eCB and BDNF/HVA samples, respectively).

Figure 3.

Figure 3.

Open in a new tab

Mediation Models for AEA and 2-AG. Mediation model for AEA (A) along with corresponding histogram of bootstrapped indirect effects and 95% CI (B) indicated that circulating concentrations of AEA mediated the relationship between aerobic exercise and reduced threat expectancy ratings following reinstatement. Mediation model for 2-AG (C), along with corresponding histogram of bootstrapped indirect effects and 95% CI (D) indicated that circulating concentrations of 2-AG did not mediate the relationship between aerobic exercise and reduced threat expectancy ratings following reinstatement. Values listed are standardized regression coefficients for path a (effect of independent variable on biomarker variable), path b (effect of biomarker variable on dependent variable), the indirect effect (product of regression coefficients for predictor variables in path a and b), path c (effect of independent variable on dependent variable without biomarker variable included in model), and path c’ (difference between direct and indirect effect). 95% CI for the indirect effect (used to determine if mediation occurred; denoted * for 95% CIs not containing zero) was obtained from bootstrap procedure involving resampling with replacement and 10,000 iterations.

Figure 4.

Figure 4.

Open in a new tab

Mediation Models for BDNF and HVA. Mediation model for BDNF (A) along with corresponding histogram of bootstrapped indirect effects and 95% CI (B) indicated that circulating concentrations of BDNF mediated the relationship between aerobic exercise and reduced threat expectancy ratings following reinstatement. Mediation model for HVA (C), along with corresponding histogram of bootstrapped indirect effects and 95% CI (D) indicated that circulating concentrations of HVA did not mediate the relationship between aerobic exercise and reduced threat expectancy ratings following reinstatement. Values listed are standardized regression coefficients for path a (effect of independent variable on biomarker variable), path b (effect of biomarker variable on dependent variable), the indirect effect (product of regression coefficients for predictor variables in path a and b), path c (effect of independent variable on dependent variable without biomarker variable included in model), and path c’ (difference between direct and indirect effect). 95% CI for the indirect effect (used to determine if mediation occurred; denoted * for 95% CIs not containing zero) was obtained from bootstrap procedure involving resampling with replacement and 10,000 iterations.

3.3. AEA Mediation Analyses

There was a trend for the moderate-intensity exercise group to exhibit greater post-exercise levels of AEA compared to the light-intensity control group (t(25)= 2.021 p = .054, η2 = .127, 95% CI: [.000, .352]) after controlling for pre-exercise concentrations of AEA (see path a of Figure 3a for standardized beta coefficient). Additionally, there was a trend for greater post-exercise concentrations of AEA to predict lower threat expectancy ratings following reinstatement (t(25)= −1.907, p = .068, η2 = .115, 95% CI: [.000, .338]) after controlling for group and pre-exercise concentrations of AEA (see path b of Figure 3a for standardized beta coefficients). Results from the bootstrapping procedure revealed a 95% confidence interval of −0.623 to −0.005 for the indirect effect (path ab), indicating that circulating concentrations of AEA following exercise mediated the relationship between moderate-intensity aerobic exercise and reduced threat expectancy ratings following reinstatement (see ab path of Figure 3a and Figure 3b).

3.4. 2-AG Mediation Analyses

Moderate-intensity aerobic exercise did not significantly predict post-exercise levels of 2-AG compared to the light-intensity control group (t(25)= −0.930, p = .361, η2 = .030, 95% CI: [.000, .218]) after controlling for pre-exercise concentrations of 2-AG (see path a of Figure 3c for standardized beta coefficient). Additionally, post-exercise concentrations of 2-AG did not significantly predict lower threat expectancy ratings following reinstatement (t(25)= −0.546, p = .590, η2 = .011, 95% CI: [.000, .171]) after group and pre-exercise concentrations of 2-AG (see path b of Figure 3c for standardized beta coefficients). Results from the bootstrapping procedure revealed a 95% confidence interval of −0.050 to 0.210 for the indirect effect (path ab), indicating that circulating concentrations of 2-AG following exercise were not found to mediate the relationship between moderate-intensity aerobic exercise and reduced threat expectancy ratings following reinstatement (see Figure 3c and 3d).

3.5. BDNF Mediation Analyses

The moderate-intensity exercise group exhibited significantly greater post-exercise levels of BDNF compared to the light-intensity control group (t(22)=2.601, p = .016, η2 = .213, 95% CI: [.007, .446]) after controlling for pre-exercise concentrations of BDNF (see path a of Figure 4a for standardized beta coefficient). Additionally, there was a trend for greater post-exercise concentrations of BDNF to predict lower threat expectancy ratings following reinstatement (t(22)=−2.067. p = .051, η2 = .146, 95% CI: [.000, .383]) after controlling for group and pre-exercise concentrations of BDNF (see path b of Figure 4a for standardized beta coefficients). Results from the bootstrapping procedure revealed a 95% confidence interval of −0.941 to −0.005 for the indirect effect (path ab), indicating that circulating concentrations of BDNF following exercise mediated the relationship between moderate-intensity aerobic exercise and reduced threat expectancy ratings following reinstatement (see Figure 4a and 4b).

3.6. HVA Mediation Analyses

Moderate-intensity aerobic exercise did not significantly predict post-exercise levels of HVA compared to the light-intensity control group (t(22)=0.471, p = .642, η2 = .008, 95% CI: [.000, .175]) after controlling for pre-exercise concentrations of HVA (see path a of Figure 4c for standardized beta coefficient). Additionally, post-exercise concentrations of HVA did not significantly predict lower threat expectancy ratings following reinstatement (t(22)=−0.721, p = .670, η2 = .020, 95% CI: [.000, .209]) after controlling for group and pre-exercise concentrations of HVA (see path b of Figure 4c for standardized beta coefficients). Results from the bootstrapping procedure revealed a 95% confidence interval of −0.190 to 0.134 for the indirect effect (path ab), indicating that circulating concentrations of HVA following exercise were not found to mediate the relationship between moderate-intensity aerobic exercise and reduced threat expectancy ratings following reinstatement (see Figure 4c and 4d).

4. Discussion

We previously reported that moderate-intensity aerobic exercise significantly reduced threat expectancy ratings (fear responding) to both the CS+ and CS− following reinstatement among women with PTSD (Crombie et al., 2021). The current analyses expanded upon these findings by examining potential neurochemical biomarkers contributing to this effect. Results indicated that exercise-induced increases in circulating concentrations of AEA and BDNF mediate the relationship between aerobic exercise and reduced threat expectancy ratings following reinstatement. In total, the results of this study suggest that exercise increases circulating levels of AEA and BDNF during the critical acute consolidation window following fear extinction learning, which in turn facilitates reduced threat expectancies following fear reinstatement.

Our AEA findings are consistent with several other pharmacological investigations. For instance, pharmacological elevations to AEA via inhibition of the catabolic enzyme responsible for AEA degradation (i.e., FAAH) have resulted in enhanced extinction recall in animal models (Mayo et al., 2020b; Dincheva et al., 2015; Gundux-Cinar et al., 2013) and a small sample of healthy adults (Mayo et al., 2020a). Similarly, enhanced memory consolidation and greater ventromedial prefrontal cortex and hippocampus activation during extinction recall in healthy adults (Rabinak et al., 2014) has been demonstrated following an acute dose of dronabinol (synthetic delta-9 tetrahydrocannabinol [THC]) which is a CB1 receptor agonist, and thus binds to the same receptors as AEA. An acute oral dose of THC in adults with PTSD also lowered threat-related amygdala reactivity and increased functional coupling of critical fear extinction structures (medial prefrontal cortex and amygdala) during a social threat processing task (Rabinak et al., 2020). It has also been demonstrated that healthy adults with elevated AEA exhibit enhanced activation of fear extinction neurocircuitry (i.e., enhanced regulation of amygdala by ventromedial prefrontal cortex; Dincheva et al., 2015). Collectively, it is plausible that enhancing CB1 receptor signaling (through increases in AEA) enhances extinction memory consolidation and recall possibly via correcting neurocircuitry deficits critical for such processes. Our findings suggest that aerobic exercise may exert a similar effect on neurochemical biomarkers of extinction learning and memory as some of the pharmacological manipulations to the eCB system, although additional research is needed to determine brain-specific effects as the current study did not include a neuroimaging component.

The role of exercise-induced increases of BDNF in reducing fear responding is consistent with a few other recent investigations. For instance, a clinically healthy sample of adults that engaged in an acute bout of moderate-intensity aerobic exercise exhibited lower fear responding (fear-potentiated startle) during a test of recall (24 hrs later) compared to a light-intensity exercise control condition, with genetic moderation analyses indicating that the effect of exercise was greater in individuals with a BDNF Va166Met polymorphism (Keyan and Bryant, 2019). Additionally, moderate-intensity aerobic exercise in conjunction with prolonged exposure therapy resulted in greater reductions in PTSD symptoms and increased BDNF concentrations (from baseline) following a 12-week intervention compared to a prolonged exposure therapy-only group (Powers et al., 2015).

The current study’s findings suggesting that circulating concentrations of HVA did not mediate the relationship between moderate-intensity aerobic exercise and reduced threat expectancy ratings following reinstatement were unexpected given the effects of exercise on peripheral HVA and dopaminergic signaling (Kendler et al., 1983; McMorris et al., 2008), and the extensive role that dopaminergic signaling plays in memory consolidation, including the consolidation of extinction learning (Abraham et al., 2016; Haaker et al., 2013; Gerlicher et al., 2018; Cisler et al., 2020). In fact, a recent placebo-controlled randomized trial among a similar population of adult women with PTSD stemming from interpersonal violence reported that post-extinction learning administration of a dopamine precursor (L-DOPA) similarly had no effect on initial extinction recall, but reduced fear responding following reinstatement (Cisler et al., 2021). While the zero-order effects of exercise on HVA were significant (see Figure 2), the effect was reduced to non-significance when controlling for baseline HVA levels. Nonetheless, building evidence suggests that dopaminergic neurotransmission may be altered in PTSD (Sherin and Nemeroff, 2011; Torrisi et al., 2019), which may explain the current study’s inability to demonstrate a significant HVA increase following exercise. It is possible, however, that exercise did influence dopaminergic signaling in specific parts of the brain even if peripheral levels of HVA were not significantly affected. Imaging research (e.g., functional magnetic resonance spectroscopy [fMRS] or positron emissions tomography [PET]) are warranted to explore effects within the brain. Similarly, although there is some preclinical evidence to suggest that 2-AG plays a role in fear extinction processes (Cavener et al., 2018; Jenniches et al., 2016), the current study’s findings suggest that circulating concentrations of 2-AG following exercise do not drive the relationship between aerobic exercise and reduced threat expectancies following reinstatement. However, this finding is not entirely unexpected as 2-AG levels were unchanged following both light- and moderate-intensity exercise.

Given that moderate-intensity aerobic exercise has been shown to increase circulating concentrations of AEA, BDNF, and HVA (in addition to influencing numerous other neurophysiological systems), it is plausible that exercise exerts cognitive enhancing effects through a synergistic process involving multiple systems working in concert. In fact, recent research has provided initial support for cross-talk between BDNF and eCB signaling (Zhao et al., 2015). For instance, acute intravenous administration of THC resulted in increased serum BDNF concentrations in healthy humans (D’Souza et al., 2009), while a preclinical investigation reported on an up-regulation of BDNF in several brain regions following one-week of treatment with a peripheral CB1 receptor agonist (Butovsky et al., 2005). Additionally, exercise-induced increases in AEA and BDNF were positively correlated immediately following and 15-minutes after an acute bout of aerobic exercise, which led the investigators to conclude that the increase in AEA following exercise may be at least partially responsible for driving the exercise-induced increase in peripheral BDNF levels (Heyman et al., 2012).

Moreover, it has been well-established that an intricate synergistic relationship exists between the eCB and dopaminergic systems, including among those with PTSD (Ney et al., 2020). For example, administration of CB1 receptor agonists have been shown to increase dopamine release in the nucleus accumbens (Cheer et al., 2004). Dopaminergic neurons have also been shown to synthesize and release eCBs (Lupica and Riegel, 2005), which can in turn influence the dopamine system. As an example of this dynamic relationship, eCBs released within the mesolimbic dopamine system can lead to an overall activation of dopaminergic cells as eCB/CB1 receptor binding relieves the inhibitory influence of CB1-expressing GABAergic terminals onto dopamine neurons, thereby enhancing dopamine release in regions such as the nucleus accumbens (Lupica & Riegel, 2005). Clearly, due to the sophisticated nature of these interactions, future research is needed to continue examining synergistic mechanisms responsible for the pro-cognitive effects of exercise, especially as it pertains to fear extinction learning and recall.

It is also plausible that other mechanisms are at least partially responsible for the observed effects reported in this manuscript. One particularly interesting possibility involves the behavioral tagging phenomenon (de Carvalho et al., 2013; de Carvalho et al., 2014; Nachtigall et al., 2019) which is an extension of the synaptic tagging and capture process (Frey & Morris, 1997) that suggests that “weak” learning experiences results in a tag being set at synapses involved in the initial learning, which can later capture proteins synthesized by another behavioral experience (performed in close proximity to when the extinction learning occurred), resulting in enhanced long-term memory formation. Support for the behavioral tagging phenomenon was initially based on pre-clinical findings of enhanced extinction learning following novel open-field exploration immediately following the extinction learning task (de Carvalho et al., 2013; de Carvalho et al., 2014). Therefore, it may be the case that our extinction learning task resulted in tags at weak early long-term hippocampal synapses. Administration of moderate-intensity aerobic exercise immediately after extinction learning may have then resulted in the synthesis of plasticity-related proteins (e.g., BDNF, AEA, dopamine) that were captured by the tags, thereby leading to enhanced long-term potentiation and enhanced memory consolidation, along with greater recall when tested 24hrs later. While this hypothesis warrants further consideration, it may also be the case that other relevant neurophysiological mechanisms not assessed (e.g., serotonin, cortisol, Neuropeptide Y, cytokines) may be contributing to the effects of exercise on extinction consolidation.

As previously highlighted, the current study is not without limitations (see Crombie et al., 2021). The small sample size should be noted, and it is important for future studies with larger sample sizes to be conducted to attempt replication of these preliminary findings. In fact, the current study examining whether our candidate biomarkers significantly mediated the relationship between moderate-intensity aerobic exercise and threat expectancy ratings following reinstatement was underpowered (see supplementary material). Although a conservative bootstrap percentile confidence interval procedure involving 10,000 iterations was implemented, given that the results of bootstrapping procedures are inherently dependent on the number of original data points, and given that the current study tested a small sample (resulting in a small number of data points for the bootstrapping procedure), additional research with larger sample sizes is warranted in order to improve our understanding of whether exercise-induced increases in circulating concentrations of candidate biomarkers mediate the effects of exercise on extinction recall. These findings should also be extended to other samples (e.g., men, combat-related PTSD, or other anxious psychopathology). In addition, future research should utilize functional neuroimaging to examine the effects of exercise on fear extinction neurocircuitry and memory consolidation processes, rather than solely relying on peripheral circulating concentrations. For instance, the aforementioned functional neuroimaging investigation examining post-extinction learning administration of L-DOPA in adult women with PTSD resulted in reduced reinstatement possibly due to reactivation of amygdala extinction encoding patterns (Cisler et al., 2020). Determining whether exercise exhibits similar effects is a potential avenue for future research.

Overall, the obtained results support a role for moderate-intensity aerobic exercise during the consolidation window in reducing threat expectancy ratings following reinstatement in women with PTSD, potentially due to exercise-induced increases in circulating concentrations of AEA and BDNF. Continued research examining exercise as a potentially important component of treating PTSD is warranted.

Supplementary Material

1

Supplementary Figure 1. Consort diagram.

Highlights:

  • Exposure-based therapies are hypothesized to work via the mechanisms of fear extinction learning.

  • Aerobic exercise delivered after extinction learning enhanced consolidation of fear extinction learning.

  • Exercise-induced increases in AEA and BDNF were found to mediate this relationship.

  • Aerobic exercise may be a promising candidate for improving exposure-based therapies.

Acknowledgments:

This study was supported by the Virginia Home Henry Fund at the University of Wisconsin – Madison. The funding source had no role in the collection, analysis, and interpretation of data; in writing the report, or in the decision to submit the article for publication. Research reported in this publication (i.e., BDNF and HVA results) was also supported in part by the Office of the Director, National Institutes of Health under Award Number P51OD011106 to the Wisconsin National Primate Research Center, University of Wisconsin-Madison. TGA was supported by the National Institute of Health under award numbers K23MH111977 and F30MH111037. The content is solely the responsibility of the authors. We would like to express many thanks to Delaney Dvorak for assisting with blood processing, Garrett Sauber for conducting the endocannabinoid assays, and all of the participants.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Conflicts of Interest: CJH is a member of the Scientific Advisory Board of Phytecs, Inc. and has an equity position in Formulate Biosciences. KMC, AS-T, KS, RA, SL, MM, NEA, KFK, TGA, and JMC have no conflicts of interest.

References

  1. Abraham AD, Neve KA, Lattal KM, 2016. Activation of D1/5 dopamine receptors: a common mechanism for enhancing extinction of fear and reward-seeking behaviors. Neuropsychopharmacology 41, 2072–2081. 10.1038/npp.2016.5 [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Blanco C, Hoertel N, Wall MM, Franco S, Peyre H, Neria Y, Helpman L, Limosin F, 2018. Toward Understanding Sex Differences in the Prevalence of Posttraumatic Stress Disorder: Results From the National Epidemiologic Survey on Alcohol and Related Conditions. J Clin Psychiatry 79. 10.4088/JCP.16m11364 [DOI] [PubMed] [Google Scholar]
  3. Blevins CA, Weathers FW, Davis MT, Witte TK, Domino JL, 2015. The posttraumatic stress disorder checklist for DSM-5 (PCL-5): development and initial psychometric evaluation. J Trauma Stress 28, 489–498. 10.1002/jts.22059 [DOI] [PubMed] [Google Scholar]
  4. Bluett RJ, Gamble-George JC, Hermanson DJ, Hartley ND, Marnett LJ, Patel S, 2014. Central anandamide deficiency predicts stress-induced anxiety: behavioral reversal through endocannabinoid augmentation. Transl Psychiatry 4, e408. 10.1038/tp.2014.53 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Bouchet CA, Miner MA, Loetz EC, Rosberg AJ, Hake HS, Farmer CE, Ostrovskyy M, Gray N, Greenwood BN, 2018. Activation of Nigrostriatal Dopamine Neurons during Fear Extinction Prevents the Renewal of Fear. Neuropsychopharmacology 43, 665–672. 10.1038/npp.2017.235 [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Brellenthin AG, Crombie KM, Hillard CJ, Koltyn KF, 2017. Endocannabinoid and mood responses to exercise in adults with varying physical activity levels. Med Sci Sports Exer, 49, 1688–1696. 10.1249/MSS.0000000000001276 [DOI] [PubMed] [Google Scholar]
  7. Breslau N, Kessler RC, Chilcoat HD, Schultz LR, Davis GC, Andreski P, 1998. Trauma and posttraumatic stress disorder in the community: the 1996 Detroit Area Survey of Trauma. Arch Gen Psychiatry 55, 626–632. 10.1001/archpsyc.55.7.626 [DOI] [PubMed] [Google Scholar]
  8. Butovsky E, Juknat A, Goncharov I, Elbaz J, Eilam R, Zangen A, Vogel Z, 2005. In vivo up-regulation of brain-derived neurotrophic factor in specific brain areas by chronic exposure to Delta-tetrahydrocannabinol. J Neurochem 93, 802–811. 10.1111/j.1471-4159.2005.03074.x [DOI] [PubMed] [Google Scholar]
  9. Cavener VS, Gaulden A, Pennipede D, Jagasia P, Uddin J, Marnett LJ, Patel S, 2018. Inhibition of diacylglycerol lipase impairs fear extinction in mice. Front Neurosci 12, 479. 10.3389/fnins.2018.00479 [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Chang WH, Scheinin M, Burns RS, Linnoila M, 1983. Rapid and simple determination of homovanillic acid in plasma using high performance liquid chromatography with electrochemical detection. Acta Pharmacol Toxicol (Copenh) 53, 275–279. 10.1111/j.1600-0773.1983.tb03423.x [DOI] [PubMed] [Google Scholar]
  11. Cheer JF, Wassum KM, Heien MLAV, Phillips PEM, Wightman RM, 2004. Cannabinoids enhance subsecond dopamine release in the nucleus accumbens of awake rats. J Neurosci 24, 4393–4400. 10.1523/JNEUROSCI.0529-04.2004 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Chhatwal JP, Stanek-Rattiner L, Davis M, Ressler KJ, 2006. Amygdala BDNF signaling is required for consolidation but not encoding of extinction. Nat Neurosci 9, 870–872. 10.1038/nn1718 [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Cisler JM, Bush K, Scott Steele J, Lenow JK, Smitherman S, Kilts CD, 2015. Brain and behavioral evidence for altered social learning mechanisms among women with assault-related posttraumatic stress disorder. J Psychiatr Res 63, 75–83. 10.1016/j.jpsychires.2015.02.014 [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Cisler JM, Privratsky AA, Sartin-Tarm A, Sellnow K, Ross M, Weaver S, Hahn E, Herringa RJ, James GA, Kilts CD, 2020. L-DOPA and consolidation of fear extinction learning among women with posttraumatic stress disorder. Transl Psychiatry 10, 287. 10.1038/s41398-020-00975-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Crombie KM, Brellenthin AG, Hillard CJ, Koltyn KF, 2018. Psychobiological responses to aerobic exercise in individuals with posttraumatic stress disorder. J Trauma Stress 31, 134–145. 10.1002/jts.22253 [DOI] [PubMed] [Google Scholar]
  16. Crombie KM, Leitzelar BN, Brellenthin AG, Hillard CJ, & Koltyn KF, 2019. Loss of exercise- and stress-induced increases in circulating 2-arachidonoylglycerol concentrations in adults with chronic PTSD. Biol Psychol 145, 1–7. 10.1016/j.biopsycho.2019.04.002 [DOI] [PubMed] [Google Scholar]
  17. Crombie KM, Sartin-Tarm A, Sellnow K, Ahrenholtz R, Lee S, Matalamaki M, Adams TG, Cisler JM, 2021. Aerobic exercise and consolidation of fear extinction learning among women with posttraumatic stress disorder. Behav Res Ther 142, 10.1016/j.brat.2021.103867 [DOI] [PubMed] [Google Scholar]
  18. de Carvalho Myskiw J, Benetti F, Izquierdo I, 2013. Behavioral tagging of extinction learning. Proc Natl Acad Sci U S A 110, 1071–1076. 10.1073/pnas.1220875110 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. de Carvalho Myskiw J, Furini CRG, Benetti F, Izquierdo I, 2014. Hippocampal molecular mechanisms involved in the enhancement of fear extinction caused by exposure to novelty. Proc Natl Acad Sci U S A 111, 4572–4577. 10.1073/pnas.1400423111 [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Dincheva I, Drysdale AT, Hartley CA, Johnson DC, Jing D, King EC, Ra S, Gray JM, Yang R, DeGruccio AM, Huang C, Cravatt BF, Glatt CE, Hill MN, Casey BJ, Lee FS, 2015. FAAH genetic variation enhances fronto-amygdala function in mouse and human. Nat Commun 6, 6395. 10.1038/ncomms7395 [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. D’Souza DC, Pittman B, Perry E, Simen A, 2009. Preliminary evidence of cannabinoid effects on brain-derived neurotrophic factor (BDNF) levels in humans. Psychopharmacology (Berl) 202, 569–578. 10.1007/s00213-008-1333-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Frey U, Morris RG, 1997. Synaptic tagging and long-term potentiation. Nature 385, 533–536. 10.1038/385533a0 [DOI] [PubMed] [Google Scholar]
  23. Gerlicher AMV, Tüscher O, Kalisch R, 2018. Dopamine-dependent prefrontal reactivations explain long-term benefit of fear extinction. Nat Commun 9, 4294. 10.1038/s41467-018-06785-y [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Gunduz-Cinar O, MacPherson KP, Cinar R, Gamble-George J, Sugden K, Williams B, Godlewski G, Ramikie TS, Gorka AX, Alapafuja SO, Nikas SP, Makriyannis A, Poulton R, Patel S, Hariri AR, Caspi A, Moffitt TE, Kunos G, Holmes A, 2013. Convergent translational evidence of a role for anandamide in amygdala-mediated fear extinction, threat processing and stress-reactivity. Mol Psychiatry 18, 813–823. 10.1038/mp.2012.72 [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Haaker J, Gaburro S, Sah A, Gartmann N, Lonsdorf TB, Meier K, Singewald N, Pape H-C, Morellini F, Kalisch R, 2013. Single dose of L-dopa makes extinction memories context-independent and prevents the return of fear. Proc. Natl. Acad. Sci. U.S.A 110, E2428–2436. 10.1073/pnas.1303061110 [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Hayes AF, Rockwood NJ, 2017. Regression-based statistical mediation and moderation analysis in clinical research: Observations, recommendations, and implementation. Behav Res Ther 98, 39–57. 10.1016/j.brat.2016.11.001 [DOI] [PubMed] [Google Scholar]
  27. Heyman E, Gamelin F-X, Goekint M, Piscitelli F, Roelands B, Leclair E, Di Marzo V, Meeusen R, 2012Intense exercise increases circulating endocannabinoid and BDNF levels in humans--possible implications for reward and depression. Psychoneuroendocrinology 37, 844–851. 10.1016/j.psyneuen.2011.09.017 [DOI] [PubMed] [Google Scholar]
  28. Iverson KM, McLaughlin KA, Gerber MR, Dick A, Smith BN, Bell ME, Cook N, Mitchell KS,2013. Exposure to Interpersonal Violence and Its Associations With Psychiatric Morbidity in a U.S. National Sample: A Gender Comparison. Psychol Violence 3, 273–287. 10.1037/a0030956 [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Jenniches I, Ternes S, Albayram O, Otte DM, Bach K, Bindila L, Michel K, Lutz B, Bilkei-Gorzo A, Zimmer A, 2016. Anxiety, stress, and fear response in mice with reduced endocannabinoid levels. Biol Psychiatry 79, 858–868. 10.1016/j.biopsych.2015.03.033 [DOI] [PubMed] [Google Scholar]
  30. Jetly R, Heber A, Fraser G, Boisvert D, 2015. The efficacy of nabilone, a synthetic cannabinoid, in the treatment of PTSD-associated nightmares: A preliminary randomized, double-blind, placebo-controlled cross-over design study. Psychoneuroendocrinology 51, 585–588. 10.1016/j.psyneuen.2014.11.002 [DOI] [PubMed] [Google Scholar]
  31. Kallies G, Rapp MA, Fydrich T, Fehm L, Tschorn M, Terán C, Schwefel M, Pietrek A, Henze R, Hellweg R, Ströhle A, Heinzel S, Heissel A, 2019. Serum brain-derived neurotrophic factor (BDNF) at rest and after acute aerobic exercise in major depressive disorder. Psychoneuroendocrinology 102, 212–215. 10.1016/j.psyneuen.2018.12.015 [DOI] [PubMed] [Google Scholar]
  32. Katona I, Freund TF, 2012. Multiple functions of endocannabinoid signaling in the brain. Annu. Rev. Neurosci 35, 529–558. 10.1146/annurev-neuro-062111-150420 [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Kendler KS, Mohs RC, Davis KL, 1983. The effects of diet and physical activity on plasma homovanillic acid in normal human subjects. Psychiatry Res 8, 215–223. 10.1016/0165-1781(83)90065-3 [DOI] [PubMed] [Google Scholar]
  34. Kessler RC, Aguilar-Gaxiola S, Alonso J, Benjet C, Bromet EJ, Cardoso G, Degenhardt L, de Girolamo G, Dinolova RV, Ferry F, Florescu S, Gureje O, Haro JM, Huang Y, Karam EG, Kawakami N, Lee S, Lepine J-P, Levinson D, Navarro-Mateu F, Pennell B-E, Piazza M, Posada-Villa J, Scott KM, Stein DJ, Ten Have M, Torres Y, Viana MC, Petukhova MV, Sampson NA, Zaslavsky AM, Koenen KC, 2017. Trauma and PTSD in the WHO World Mental Health Surveys. Eur J Psychotraumatol 8, 1353383. 10.1080/20008198.2017.1353383 [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Kessler RC, Chiu WT, Demler O, Merikangas KR, Walters EE, 2005. Prevalence, severity, and comorbidity of 12-month DSM-IV disorders in the National Comorbidity Survey Replication. Arch Gen Psychiatry 62, 617–627. 10.1001/archpsyc.62.6.617 [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Kessler RC, Sonnega A, Bromet E, Hughes M, Nelson CB, 1995. Posttraumatic stress disorder in the National Comorbidity Survey. Arch Gen Psychiatry 52, 1048–1060. 10.1001/archpsyc.1995.03950240066012 [DOI] [PubMed] [Google Scholar]
  37. Keyan D, Bryant RA, 2019. Acute exercise-induced enhancement of fear inhibition is moderated by BDNF Va166Met polymorphism. Transl Psychiatry 9, 131. 10.1038/s41398-019-0464-z [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Lafenêtre P, Chaouloff F, Marsicano G, 2007. The endocannabinoid system in the processing of anxiety and fear and how CB1 receptors may modulate fear extinction. Pharmacol. Res 56, 367–381. 10.1016/j.phrs.2007.09.006 [DOI] [PubMed] [Google Scholar]
  39. Lu Y, Christian K, Lu B, 2008. BDNF: a key regulator for protein synthesis-dependent LTP and long-term memory? Neurobiol Learn Mem 89, 312–323. 10.1016/j.nlm.2007.08.018 [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Lupica CR, Riegel AC, 2005. Endocannabinoid release from midbrain dopamine neurons: a potential substrate for cannabinoid receptor antagonist treatment of addiction. Neuropharmacology 48, 1105–1116. 10.1016/j.neuropharm.2005.03.016 [DOI] [PubMed] [Google Scholar]
  41. Lutz B, Marsicano G, Maldonado R, Hillard CJ, 2015. The endocannabinoid system in guarding against fear, anxiety and stress. Nat. Rev. Neurosci 16, 705–718. 10.1038/nrn4036 [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Marsicano G, Wotjak CT, Azad SC, Bisogno T, Rammes G, Cascio MG, Hermann H, Tang J, Hofmann C, Zieglgänsberger W, Di Marzo V, Lutz B, 2002. The endogenous cannabinoid system controls extinction of aversive memories. Nature 418, 530–534. 10.1038/nature00839 [DOI] [PubMed] [Google Scholar]
  43. Mataix-Cols D, Fernández de la Cruz L, Monzani B, Rosenfield D, Andersson E, Pérez-Vigil A, Frumento P, de Kleine RA, Difede J, Dunlop BW, Farrell LJ, Geller D, Gerardi M, Guastella AJ, Hofmann SG, Hendriks G-J, Kushner MG, Lee FS, Lenze EJ, Levinson CA, McConnell H, Otto MW, Plag J, Pollack MH, Ressler KJ, Rodebaugh TL, Rothbaum BO, Scheeringa MS, Siewert-Siegmund A, Smits JAJ, Storch EA, Ströhle A, Tart CD, Tolin DF, van Minnen A, Waters AM, Weems CF, Wilhelm S, Wyka K, Davis M, Rück C, the DCS Anxiety Consortium, Altemus M, Anderson P, Cukor J, Finck C, Geffken GR, Golfels F, Goodman WK, Gutner C, Heyman F, Jovanovic T, Lewin AB, McNamara JP, Murphy TK, Norrholm S, Thuras P, 2017. D-Cycloserine augmentation of exposure-based cognitive behavior therapy for Anxiety, obsessive-compulsive, and posttraumatic stress disorders: a systematic review and meta-analysis of individual participant data. JAMA Psychiatry 74, 501–510. 10.1001/jamapsychiatry.2016.3955 [DOI] [PubMed] [Google Scholar]
  44. Mayo LM, Asratian A, Lindé J, Holm L, Nätt D, Augier G, Stensson N, Vecchiarelli HA, Balsevich G, Aukema RJ, Ghafouri B, Spagnolo PA, Lee FS, Hill MN, Heilig M, 2020a. Protective effects of elevated anandamide on stress and fear-related behaviors: translational evidence from humans and mice. Mol. Psychiatry 25, 993–1005. 10.1038/s41380-018-0215-1 [DOI] [PubMed] [Google Scholar]
  45. Mayo LM, Asratian A, Lindé J, Morena M, Haataja R, Hammar V, Augier G, Hill MN, Heilig M, 2020b. Elevated anandamide, enhanced recall of fear extinction, and attenuated stress responses following inhibition of fatty acid amide hydrolase: A randomized, controlled experimental medicine trial. Biol. Psychiatry 87, 538–547. 10.1016/j.biopsych.2019.07.034 [DOI] [PubMed] [Google Scholar]
  46. McMorris T, Collard K, Corbett J, Dicks M, Swain JP, 2008. A test of the catecholamines hypothesis for an acute exercise-cognition interaction. Pharmacol Biochem Behav 89, 106–115. 10.1016/j.pbb.2007.11.007 [DOI] [PubMed] [Google Scholar]
  47. Meyer JD, Koltyn KF, Stegner AJ, Kim J-S, Cook DB, 2016. Relationships between serum BDNF and the antidepressant effect of acute exercise in depressed women. Psychoneuroendocrinology 74, 286–294. 10.1016/j.psyneuen.2016.09.022 [DOI] [PubMed] [Google Scholar]
  48. Morena M, Berardi A, Colucci P, Palmery M, Trezza V, Hill MN, Campolongo P, 2018. Enhancing Endocannabinoid neurotransmission augments the efficacy of extinction training and ameliorates traumatic stress-induced behavioral alterations in rats. Neuropsychopharmacology 43, 1284–1296. 10.1038/npp.2017.305 [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Nachtigall EG, Furini CRG, Behling J. a. K., Farias CP, Izquierdo E, Myskiw J. de C, 2019. Facilitation of fear extinction by novelty is modulated by β-adrenergic and 5-HT1A serotoninergic receptors in hippocampus. Neurobiol Leam Mem 166, 107101. 10.1016/j.nlm.2019.107101 [DOI] [PubMed] [Google Scholar]
  50. Ney LJ, Akhurst J, Bruno R, Laing PAF, Matthews A, Felmingham KL, 2020. Dopamine, endocannabinoids and their interaction in fear extinction and negative affect in PTSD. Prog Neuropsychopharmacol Biol Psychiatry 105, 110118. 10.1016/j.pnpbp.2020.110118 [DOI] [PubMed] [Google Scholar]
  51. Notaras M, van den Buuse M, 2020. Neurobiology of BDNF in fear memory, sensitivity to stress, and stress-related disorders. Mol Psychiatry 25, 2251–2274. 10.1038/s41380-019-0639-2 [DOI] [PubMed] [Google Scholar]
  52. Peters J, Dieppa-Perea LM, Melendez LM, Quirk GJ, 2010. Induction of fear extinction with hippocampal-infralimbic BDNF. Science 328, 1288–1290. 10.1126/science.1186909 [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Powers MB, Medina JL, Burns S, Kauffman BY, Monfils M, Asmundson GJG, Diamond A, McIntyre C, Smits JAJ, 2015. Exercise augmentation of exposure therapy for PTSD: rationale and pilot efficacy data. Cogn Behav Ther 44, 314–327. 10.1080/16506073.2015.1012740 [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Quirk GJ, Mueller D. Neural mechanisms of extinction learning and retrieval. Neuropsychopharmacology. 2008;33(1):56–72. 10.1038/sj.npp.1301555 [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Rabinak CA, Angstadt M, Lyons M, Mori S, Milad MR, Liberzon I, Phan KL, 2014. Cannabinoid modulation of prefrontal-limbic activation during fear extinction learning and recall in humans. Neurobiol Learn Mem 113, 125–134. 10.1016/j.nlm.2013.09.009 [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Rabinak CA, Blanchette A, Zabik NL, Peters C, Marusak HA, Iadipaolo A, Elrahal F, 2020. Cannabinoid modulation of corticolimbic activation to threat in trauma-exposed adults: a preliminary study. Psychopharmacology (Berl) 237, 1813–1826. 10.1007/s00213-020-05499-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Resick PA, Nishith P, Weaver TL, Astin MC, Feuer CA, 2002. A comparison of cognitive-processing therapy with prolonged exposure and a waiting condition for the treatment of chronic posttraumatic stress disorder in female rape victims. J Consult Clin Psychol 70, 867–879. 10.1037//0022-006x.70.4.867 [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Resnick HS, Kilpatrick DG, Dansky BS, Saunders BE, Best CL, 1993. Prevalence of civilian trauma and posttraumatic stress disorder in a representative national sample of women. J Consult Clin Psychol 61, 984–991. 10.1037//0022-006x.61.6.984 [DOI] [PubMed] [Google Scholar]
  59. Rothbaum BO, Davis M, 2003. Applying learning principles to the treatment of post-trauma reactions. Ann N Y Acad Sci. 1008, 112–21. [DOI] [PubMed] [Google Scholar]
  60. Roquet RF, Monfils M-H, 2018. Does exercise augment operant and Pavlovian extinction: A meta-analysis. J Psychiatr Res 96, 73–93. 10.1016/j.jpsychires.2017.09.018 [DOI] [PubMed] [Google Scholar]
  61. Schnurr PP, Friedman MJ, Engel CC, Foa EB, Shea MT, Chow BK, Resick PA, Thurston V, Orsillo SM, Haug R, Turner C, Bernardy N, 2007. Cognitive behavioral therapy for posttraumatic stress disorder in women: a randomized controlled trial. JAMA 297, 820–830. 10.1001/jama.297.8.820 [DOI] [PubMed] [Google Scholar]
  62. Scott KM, Koenen KC, King A, Petukhova MV, Alonso J, Bromet EJ, Bruffaerts R, Bunting B, de Jonge P, Haro JM, Karam EG, Lee S, Medina-Mora ME, Navarro-Mateu F, Sampson NA, Shahly V, Stein DJ, Torres Y, Zaslavsky AM, Kessler RC, 2018. Post-traumatic stress disorder associated with sexual assault among women in the WHO World Mental Health Surveys. Psychol Med 48, 155–167. 10.1017/S0033291717001593 [DOI] [PMC free article] [PubMed] [Google Scholar]
  63. Sherin JE, Nemeroff CB, 2011. Post-traumatic stress disorder: the neurobiological impact of psychological trauma. Dialogues Clin Neurosci 13, 263–278. 10.31887/DCNS.2011.13.2/jsherin [DOI] [PMC free article] [PubMed] [Google Scholar]
  64. Shrout PE, Bolger N, 2002. Mediation in experimental and nonexperimental studies: new procedures and recommendations. Psychol Methods 7, 422–445. 10.1037/1082-989X.7.4.422 [DOI] [PubMed] [Google Scholar]
  65. Tanner MK, Hake HS, Bouchet CA, Greenwood BN, 2018. Running from fear: exercise modulation of fear extinction. Neurobiol Learn Mem 151, 28–34. 10.1016/j.nlm.2018.03.021 [DOI] [PMC free article] [PubMed] [Google Scholar]
  66. Torrisi SA, Leggio GM, Drago F, Salomone S, 2019. Therapeutic challenges of post-traumatic stress disorder: focus on the dopaminergic system. Front Pharmacol 10, 404. 10.3389/fphar.2019.00404 [DOI] [PMC free article] [PubMed] [Google Scholar]
  67. Zhao L, Yeh MF-W, Fevine ES, 2015. Role for endogenous BDNF in endocannabinoid-mediated long-term depression at neocortical inhibitory synapses. eNeuro 2. 10.1523/ENEURO.0029-14.2015 [DOI] [PMC free article] [PubMed] [Google Scholar]
  68. Zoellner LA, Telch M, Foa EB, Farach FJ, McLean CP, Gallop R, Bluett EJ, Cobb A, Gonzalez-Lima F, 2017. Enhancing extinction learning in posttraumatic stress disorder with brief daily imaginal exposure and methylene blue: a randomized controlled trial. J Clin Psychiatry 78, e782–e789. 10.4088/JCP.16m10936 [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1

Supplementary Figure 1. Consort diagram.

NIHMS1725483-supplement-1.pdf (354.6KB, pdf)

RESOURCES