A pilot study of the role of the BDNF Val66Met polymorphism in response to exercise-augmented exposure therapy for posttraumatic stress disorder

,

There is abundant evidence that trauma-focused psychotherapies are the frontline treatment for posttraumatic stress disorder (PTSD), and accordingly are recommended by international treatment guidelines (Forbes et al., (2020).The most studied of these treatments is Prolonged Exposure, which involves focusing on trauma memories with the goal of learning that the ongoing threat associated with the memory is no longer evident, and this leads to a reduction in anxiety (Lebois et al., 2019).The prevailing theoretical model to explain this process involves fear conditioning proposals that posit that repeated exposure to reminders of the trauma in a safe context leads to extinction of previously conditioned fear responses to these reminders (Pitman et al., 2012).Despite the utility of these treatments, up to one-half of PTSD patients do not respond to this evidence-based therapy (Loerinc et al., 2015).This has led to calls for greater attention to mechanism-driven treatments for PTSD, and for tailoring treatments to specific profiles of patients (Hofmann and Hayes, 2019).
One of the key mechanisms underpinning extinction learning is brain derived neurotrophic factor (BDNF), a neurotrophin that promotes neural plasticity and long-term potentiation (Lu et al., 2008).BDNF has been found to be critical for extinction learning (Andero and Ressler, 2012).In animals BDNF signaling functions to down-regulate negative memories via facilitating plasticity in the medial prefrontal cortex (Pattwell et al., 2012), and infusion of BDNF into infralimbic and hippocampal regions promotes extinction (Peters et al., 2010).In humans, peripheral assessment of BDNF levels also indicate promotion of plasticity, and has been linked with lower BDNF is linked with psychopathological states, including negative symptoms of schizophrenia and depression (Fang et al., 2019;Han et al., 2021).Human research has also investigated the role of the polymorphisms of the BDNF gene, and particularly a single nucleotide polymorphism that produces a functional Valine (Val) to methionine (Met) substitution at codon 66 (Val66Met) that is associated with reduced hippocampal function and activity-dependent BDNF secretion (Egan et al., 2003).There is human evidence that the Met allele of the BDNF Val66met polymorphism impedes extinction (Soliman et al., 2010), and that Met allele carriers have greater amygdala and reduced ventromedial prefrontal activity during extinction (Lonsdorf et al., 2015).In terms of PTSD, one study found that deficient extinction learning is most evident in Met allele carriers (Felmingham et al., 2018).Further, carriers of the Met allele respond more poorly to exposure therapy for PTSD (Felmingham, 2013).
Convergent evidence indicates that brief bouts of intense exercise can augment BDNF release (Greenwood et al., 2009), such that more intense exercise results in greater levels of BDNF (Winter et al., 2007).Moreover, bouts of intense exercise following extinction learning of conditioned fear consolidates extinction retention in animals (Siette et al., 2014) and humans (Crombie et al., 2021a(Crombie et al., , 2023;;Keyan and Bryant, 2019).There is also evidence that the capacity for exercise to augment extinction in humans is limited by reduced BDNF levels (Crombie et al., 2021b) and in carriers of the BDNF Met allele (Keyan and Bryant, 2019).This convergent evidence led to a controlled treatment study for PTSD that found that combining exposure therapy with brief bouts of intense exercise resulted in greater longer-term reductions in PTSD severity relative to when the exposure was followed by passive stretching (Bryant et al., 2023).
The present study reports a follow-up to the aforementioned treatment study of exercise-augmented PTSD treatment (Bryant et al., 2023) that addresses the issue of the role of the BDNF Val66met polymorphism in moderating response to exposure therapy.Specifically, this study assessed BDNF alleles of participants who were randomized to either exercise-augmented exposure therapy or exposure therapy combined with passive stretching.The primary aim of this research was to replicate one previous finding that carriers of the BDNF Met allele have poorer response to exposure therapy than BDNF Val/Val carriers.Secondarily, on the premise that BDNF Met carriers are less prone to release of BDNF, we aimed to test whether carriers of the Met allele would receive less advantage from the exercise-augmented exposure relative to Val/Val carriers.

Trial Design
The trial was approved by the UNSW Human Research Ethics Committee, and was prospectively registered on the Australian New Zealand Clinical Trials Registry (ACTRN12612000185864).Full details are reported in the primary trial report (Bryant et al., 2023).This study involves an analysis of a subset of participants who met criteria for PTSD and were enrolled in a randomised, parallel, controlled trial in which they were randomly assigned to either exposure therapy in combination within exercise (ET/Exer) or standard exposure therapy (ET) on a 1:1 basis.Participants were genotyped for BDNF alleles, and were assessed at baseline and following treatment.The outcome of interest was reduction in PTSD severity in relation to BDNF allele.

Participants
The power analysis was based on the predicted outcome of the primary clinical outcome of the treatment trial rather than the moderating influence of the BDNF alleles.The initial power analysis of the trial was based on detecting moderate between conditions effect size (d = 0.6), resulting in a required sample size of 130 participants (power = 0.80, α =.05), allowing for 30 % attrition at the follow-up assessment.Participants in Australia were enrolled between December 12, 2012 and July 25, 2018, and met DSM-IV criteria for PTSD as defined by clinical interview based on the Clinician Administered PTSD Scale (D. D. Blake et al., 1995).Participants were randomized to receive exercise-augmented exposure therapy (ET/Exer) or standard exposure therapy combined with passive stretching (ET); there were 65 participants assigned to each treatment arm.The inclusion criteria comprised (a) aged ≥ 18 years, and (b) met criteria for DSM-IV defined PTSD, and exclusion criteria included (a) aged ≥ 70 years scores, (b) imminent suicidal risk, (c) presence of psychosis or substance dependence, (d) history of moderate/severe traumatic brain injury, or (d) presence of a physical condition that may be exacerbated by aerobic exercise.As investigation of the role of BDNF alleles on treatment outcome was a secondary analysis, no power analysis was conducted for this effect.

Procedure
Following written informed consent, participants underwent a baseline assessment that comprised a range of measures.Relevant to this study was the administration of Clinical Administered PTSD Scale -2 (CAPS-2) (D.D. Blake et al., 1995), which is a structured clinical interview that indexes the 17 symptoms described by the DSM-IV PTSD criteria; each symptom is rated on a 5-point scale in terms of severity and frequency of the symptoms in the past month score (range, 0-136; higher scores indicate greater PTSD severity).We also assessed depression via the Beck Depression Inventory-2 (BDI-2) (Beck et al., 1996), which is a 21-item self-report measure of depression in the past two weeks (range, 0-63; higher scores indicate more severe depression).Participants then received 9 ×90-minute weekly sessions of exposure therapy for PTSD combined with aerobic exercise (ET/Exer) conducted for 10 minutes or to exposure therapy combined with passive stretching (ET).The ET sessions comprised an initial session of psychoeducation (Session 1), cognitive reframing of maladaptive appraisals (Sessions 2-9), exposure therapy of 20 minutes that involved reliving the trauma memory (Sessions 3-8), in vivo exposure to feared situations (Sessions 3-8), and relapse prevention (Session 9).Following each exposure exercise when participants relived the trauma memory, participants engaged in aerobic exercise for 20-25 minutes.This period of exercise comprised 10-15 minutes of exercising until they reached their personal aerobic level heart rate and then maintained this intensity of exercise for 10 minutes.Engaging in 10 minutes of exercise that involved one's aerobic level was adopted because this has been shown to be sufficient to promote BDNF levels (Winter et al., 2007), and 10 minutes of aerobic exercise can facilitate extinction learning (Keyan and Bryant, 2019).Before treatment participants were assessed for their personal aerobic target zone.Adopting the guidelines of the (Center for Disease Control, 2022), one's aerobic heart rate was determined as 65 %-85 % of maximum heart rate (defined as 220 -age in years).Heart rate was measured with a Garmin FR70 heart rate watch and chest belt, which consisted of two smart fabric sensors that acquired cardiac activity.Heart rate data was sampled at a rate of 2.4 GHz, and stored as average beats per minute over a 10-min period.Following each exposure exercise, participants were instructed to run on a 15-cm high stepper exercise platform whilst having their cardiac activity recorded.Once they reached their target heart rate (which typically required 10 minutes of exercise prior to reaching their aerobic zone), they were instructed to maintain this exercise level for the 10-minute aerobic exercise period.In the ET arm, participants received the same treatment components except that following their exposure sessions they lay on an exercise mat and underwent 20 minutes of supervised stretching exercises, which comprised non-strenuous stretching of limbs.

Genetic Acquisition and Analysis
Genomic DNA was extracted from saliva samples obtained from samples using Oragene-DNA OG-500 saliva kits (DNA Genotek Inc., 2021).Participants provided at least one ml of saliva via the passive drool method., then post the sample to Westmead Hospital via a pre-paid  (Felmingham, 2013;Soliman et al., 2010).This grouping resulted in 52 participants being categorized as Val/Val carriers (60.7 %) and 33 as Met carriers (39.3 %).

Statistical Analyses
To determine the contributions of BDNF alleles to reduction of PTSD severity following treatment, we conducted a repeated measures of variance (ANOVA) on total CAPS scores with BDNF allele (Val/Val, Met) and the treatment arm as the between factor, and Time (baseline, posttreatment) as the within factor.This analytic approach required us to focus on those who completed the posttreatment assessment rather than adopt an intent-to-treat approach (as the purpose of this study was not to determine the efficacy of the treatment).All analyses used p <.05 as the alpha value.

Participant demographics
Although there 130 participants randomized to either treatment arm in this trial, only 85 participants provided saliva samples (this occurred because either participants declined to provide samples or collection was not possible during the COVID-19 pandemic when lockdown and safety issues precluded acquisition of samples).More participants in the ET/Exer condition (48; 73.8 %) provided saliva samples relative to those in the ET condition (37; 56.9 %), χ 2 = 4.1, p =.04.In terms of allele distribution in each treatment arm, the ET/Exer arm comprised 25 (52.1 %) Val/Val carriers and 23 (47.9 %) Met carriers, and the ET arm comprised 27 (73.0%) Val/Val carriers and 10 (27.0 %) Met carriers; there were marginally more Met carriers in the ET/Exer arm than the ET arm, χ 2 = 3.8, p =.05.Table 1 presents the participant characteristics of Val/Val and Met carriers.There were no differences between allele groups on age, PTSD severity, time since trauma, or severity of depression.All participants who provided saliva samples were retained at the posttreatment assessment.

Treatment Response
Table 1 presents the mean PTSD severity levels (CAPS total scores) at baseline and at the posttreatment assessment.A 2 (Time) x 2 (Treatment Arm) x 2 (BDNF Allele) repeated measures analysis of variance (ANOVA) indicated a significant main effect for Time In recognition that there are ethnic differences in terms of the BDNF alleles (Luo et al., 2019;Pivac et al., 2009), these analyses were repeated with Caucasian participants only (40 in ET/Exer, 29 in ET).These analyses also indicated that Caucasian carriers of the Val/Val allele achieved better symptom reduction in the ET/Exer condition than Met carriers, and that this difference was not apparent in the ET condition (see supplementary materials).

Discussion
The major finding from this study was that BDNF Val/Val allele carriers allele reported greater reduction of PTSD symptoms following exposure therapy relative to BDNF Met allele carriers.This finding replicates one previous study that also reported that BDNF Met carriers had poorer response to exposure therapy for PTSD than Met carriers (Felmingham et al., 2013).These observations are consistent with the conceptualization of exposure therapy as a form of extinction learning, and that extinction is facilitated by release of BDNF (Peters et al., 2010).On the premise that the BDNF Val allele promotes BDNF release (Egan et al., 2003) and the Met allele is linked to poorer extinction learning (Lonsdorf et al., 2010), a likely inference is that BDNF Val carriers will  Note.SEB = standard error of B.
respond more effectively to the extinction-based exposure therapy than Met carriers because they are able to engage in greater extinction learning.This interpretation can be considered in the context of evidence that there is reduced BDNF levels in hippocampal and ventromedial regions in Met carriers (Soliman et al., 2010), and this can result in diminished protein synthesis and neural plasticity in key regions that are also implicated in successful response to exposure therapy (Bryant et al., 2008;Fonzo et al., 2017).
The other finding to emerge from this study was that exerciseaugmented exposure therapy was particularly associated with stronger treatment gains in BDNF Val/Val carriers than Met carriers.We premise this conclusion by noting that only 65 % of the potential sample provided saliva samples, and that fewer participants in the exposure therapy arm provided saliva than those in the exercise-augmented treatment condition.This factor renders any conclusions about the role of BDNF allele and exercise-augmented treatment as tentative because the sample size is limited.There is convergent evidence that intense exercise promotes BDNF release (Winter et al., 2007), and this is associated with facilitated extinction retention in animals (Siette et al., 2014) and humans (Keyan and Bryant, 2019).In the context of findings that memory for emotional events is stronger in BDNF Val carriers following intense exercise (Keyan and Bryant, 2017), it is possible that the exercise-augmented exposure therapy achieved greater PTSD reduction in Val carriers because they were able to learn and retain the new memories of safety and mastery achieved during exposure therapy.We emphasize again that this interpretation remains speculative until more definitive studies are conducted with larger samples sizes.
There are other possible explanations for the role of BDNF Val carriers responding more effectively to exposure therapy than Met carriers.It is possible that BDNF interacts with specific mechanisms that underpin a subtype of PTSD, and that this subtype is distinctly responsive to exposure therapy (Felmingham et al., 2013).This possibility is worth further investigation considering increasing evidence of subtypes of PTSD (Forbes et al., 2010), and that there are distinct neural predictors of response to exposure therapy for these subtypes (Bryant et al., 2020(Bryant et al., , 2021)).This possibility is worth investigating by future adequately powered studies that allow investigation of known subtypes of PTSD, such as dissociative or dysphoric presentations because they purportedly have less arousal responses (Bryant et al., 2020;Lanius et al., 2021) and therefore the role of the BDNF allele may have distinct influences on these subtypes following exposure-augmented therapy.We also recognize that exercise can augment extinction via other mechanisms associated with BDNF, including endocannabinoids (Crombie et al., 2021b;Heyman et al., 2012) and the mammalian target of rapamycin (Moya et al., 2020).
We note a number of methodological issues in this study.First, we reiterate that the sample size for this study was lower than expected and we therefore can only draw tentative conclusions from the findings.There is a need to replicate the study with a significantly larger sample size.Second, to more accurately understand the role of BDNF in treatment response for exposure therapy, and particularly exerciseaugmented therapy, obtaining BDNF levels via plasma should be undertaken prior to and following treatment in future trials.Third, we recognize that the BDNF polymorphism may interact with other genotypes that are associated with PTSD and extinction learning, including the 5HTT (implicated in serotonergic activity), FKBP5 (glucocorticoid receptor), and the COMT (catecholaminergic activity) genotypes (Bauer, 2015;Binder et al., 2008;Lonsdorf et al., 2009).The importance of considering multiple genetic influences is particularly indicated by findings that serotonin transporter genotype is predictive of response to exposure therapy for PTSD (Bryant et al., 2010).Fourth, our sample was not adequately representative of different ethnicities, which limits the generalizability of the findings regarding the role of genotypes in treatment response in different ethnic populations.
In conclusion, this study provides further evidence exposure therapy for PTSD has greater benefit for people with the BDNF Val allele, and contributes to the growing knowledge of the role of BDNF in extinction processes in humans.The study also suggests that augmenting exposure therapy for PTSD with aerobic exercise may be particularly useful for Val/Val carriers, which also accords with existing evidence that exercise promotes extinction learning in BDNF carriers.Further investigations of these factors may advance our understanding of how to optimize exposure therapy for PTSD, and potentially point to strategies for tailoring interventions for PTSD patients with specific genotypic profiles.

Table 1
Baseline Participant Characteristics.Godin Leisure-Time Questionnaire (total score range: 0-119; higher scores indicate greater engagement in physical activity).Percentages appear in parentheses.Standard deviations appear in ±.PTSD severity = Clinician Administered PTSD Scale total score.Depression severity = Black Depression Inventory total score.

Table 2
Summary of hierarchical regression models for posttreatment PTSD severity.