Re-experiencing traumatic events is common and often accompanied by emotional distress. While propranolol has been proposed as a treatment for traumatic memories based on reconsolidation theory, evidence remains mixed. This randomized controlled trial aimed to determine the effectiveness of brief propranolol treatment on participants with traumatic memories and associated symptoms (not necessarily meeting diagnostic criteria for PTSD).
MethodsA double-blind, placebo-controlled randomized trial was conducted in an outpatient setting in a Midwestern Canadian city. Seventy-five participants (propranolol: n=38; placebo: n=37) were assessed at baseline, 4 weeks, and 8 weeks post-intervention. Propranolol was administered following memory reactivation, in keeping with reconsolidation theory; the implications of this timing are discussed. Participants completed primary outcome measures including the Clinician-Administered PTSD Scale (CAPS), Impact of Event Scale-Revised (IES-R), and Traumatic Memory Description Measure (TMDM), alongside clinical interviews assessing anxiety, depression, and traumatic memory experiences.
ResultsNo significant differences emerged between groups on primary outcome measures assessing traumatic memories and their consequences. However, anxiety, distress, and trauma-related symptoms decreased two-fold over time across both groups. Clinician and self-ratings of improvement were statistically higher in the propranolol group compared to placebo. The treatment group demonstrated perfect adherence, which was representatively higher than the placebo group.
ConclusionPropranolol did not differ significantly from placebo on primary outcome measures of traumatic memory and associated symptom severity. Both groups showed clinically meaningful within-group reductions over time. Subjective global improvement ratings were higher in the propranolol group, though these secondary findings should be interpreted cautiously. The results do not support propranolol as an efficacious treatment for traumatic memories but highlight the potential therapeutic value of memory reactivation procedures and warrant further controlled investigation. The timing of propranolol administration in relation to memory reactivation, specifically post-reactivation in the current protocol, may be a critical limiting factor.
Re-experiencing of post-traumatic events is common in both the general population and among those in high-risk professions often accompanied by emotional distress and risk for chronic disability.1 With rates as high as 58% in trauma-exposed veterans, these distressfully intrusive symptoms are central to the psychopathology of post-traumatic stress disorder (PTSD).1 While treatment of these symptoms should result in overall improvement and prevention, current treatments adhering to PTSD practice guidelines have yielded unsatisfactory response (53–62%) and remission rates (23–51%) for the disorder.2,3 Several novel strategies for treating these symptoms have been suggested4, many of which are based on the reconsolidation theory of memory formation as supported by findings in animal studies.5–7 When considering treatments, one meta-analysis on the effects of propranolol has stipulated that it may have a role in reducing the recall for negatively valanced memories for both new- and previously learned emotional material in healthy participants.8
Reconsolidation theory and propranololAccording to reconsolidation theory, reactivated memories enter a labile state in which they can either be preserved or altered.5 During these periods, multiple agents impact memory states.4,9–12 Memory mediated through the primary brain regions of the hippocampus and basolateral nuclei of the amygdala is inhibited by propranolol, a beta-adrenergic receptor blocker.6,13 The effects of propranolol on emotion-laden memory in healthy volunteers8,14 and in clinical subjects15 support fewer re-experiences and reduction in quality and quantity of traumatic memory. These results have included decreased emotional arousal, prevention of extinction learning, and the necessity of reactivation for propranolol to have an effect which differentiate from the effects of corticosteroids on neutral, short and long-term memory.16
Although arguments in favor of and against propranolol have been presented17, both sides appear more concerned with its mechanism of action than actual results.18 There is general agreement on the need to study timing of drug administration, and restriction to those with established post-traumatic distress.15,18 This is especially indicated when ‘re-experiencing symptoms’ of the Clinical Assessment of PTSD Scale (CAPS) are present. These symptoms are both treatment-resistant and ominous, as they have been identified as predictors of suicidality in veterans, survivors of combat, and emergency first responders, such as firefighters.2,17
Following the first report that propranolol differentially impacts emotionally arousing human memories19, ethical issues emerged (i.e., potential implications for later testimony, suppression of positive traumatic growth, and disrupted sense of self).18,20 Findings from studies employing propranolol in secondary prevention of PTSD have been mixed. Results include reductions in PTSD symptom scores and PTSD incidence21–23, reductions in symptomology after 6 months24, initial reductions in symptoms with no difference at three months25, reductions for those with severe symptoms at 3 months26, and no difference.27–29 In those experiencing chronic traumatic memories, propranolol reduced physiological responses to post retrieval reactivation scripts22, but no changes were noted in PTSD symptoms.12 Brunet and colleagues conducted a randomized control trial to evaluate the potential impact of PTSD symptoms using pre-reactivation propranolol therapy and found that those individuals who recalled a traumatic event while under the influence of propranolol for six weeks had a marked decrease in symptom ratings as compared with a placebo control group.30 While Pigeon and colleague’s31 meta-analysis found that reconsolidation impairment under the influence of propranolol reduced recall of aversive materials, the conditioned response, and psychiatric symptoms, when looking specifically at PTSD, Raut and colleagues32 found there was no beneficial effect of propranolol for those with PTSD save for a reduction heart rate after trauma memory recall. Additionally, Steenen et al.33 critiqued Pigeon et al.,31 and re-analyzed the data, including additional studies, and demonstrated no statistically significant effect of propranolol over placebo in treating psychotrauma-related symptoms. They conclude that the current evidence does not fully support the use of propranolol for impairing traumatic memory reconsolidation and call for modesty in interpreting results suggesting complex problems can be solved with medication. Three open label studies with propranolol that have demonstrated clinical benefits,30 alongside case studies,15,34 and the results of Brunet et al24 provided the impetus for this controlled clinical trial. Early enthusiasm for propranolol in this context was also fueled by dramatic anecdotal reports and case descriptions of dramatic symptom relief, which generated considerable clinical interest despite the limited and inconsistent controlled evidence available at the time.15,34 These compelling, if preliminary, accounts underscored the perceived promise of the intervention and contributed to the momentum for this randomized controlled trial. There is also emerging recognition that the timing of propranolol administration relative to memory reactivation may be a critical moderating variable, with pre-reactivation administration potentially offering greater attenuation of adrenergic arousal during the labilization window than post-reactivation dosing.24,35
ObjectiveGiven the conflicting information on the effectiveness of propranolol treatment on traumatic memory reconciliation, the objective of the trial was to determine the effectiveness of brief treatment with propranolol on subjects with traumatic memories and associated symptoms. This study was a randomized controlled trial (RCT), which was a double-blinded, placebo-controlled trial completed in an outpatient setting with clinical trial identifier number NCT01069159. The study was approved by the University of Saskatchewan and certified by Health Canada under the regulations of pharmacological products guidelines. We hypothesized that participants receiving propranolol would experience less emotional distress when the memory is subsequently reactivated and less intact memory for the traumatic experience concerned. Further, it was expected that propranolol treatment would be associated with fewer spontaneous re-experiences.
MethodTrial designThis trial used a double-blind randomized block design where participants were assigned to either the propranolol or placebo condition by a third party who had no contact with participants. Inclusion criteria included having a traumatic memory which produced distress (but not necessarily a clinical diagnosis of PTSD). Pregnancy, breastfeeding, current treatment with a beta-adrenergic blocker, corticosteroid treatment, and contraindications for propranolol such as asthma were considered exclusion criteria for this study. Participants were recruited through posters, the local newspaper, the study website, and contact with potential referral sources of post-traumatic experiences.
ParticipantsA total of 75 participants (Propranolol: n = 38; and Placebo: n = 37) participated in the study between September 2010 and November of 2011. Participants were male and female adults aged 18 to 70 years with a traumatic memory but not necessarily a clinical diagnosis of PTSD. Although formal PTSD diagnosis was not required for inclusion, the sample demonstrated substantial clinical burden at baseline. Over four-fifths of participants reported comorbid depression, three-quarters were receiving psychotropic medication, and mean baseline CAPS scores fell within the severe symptomatology range, collectively supporting characterization of this as a clinically significantly symptomatic sample. Participants were assigned to either the propranolol or placebo condition using a randomized block design by the study statistician, who had no contact with the participants. This individual subsequently supplied the research coordinator with numbered envelopes containing a slip of paper lettered A or B, which corresponded to placebo or propranolol. The medication/placebo was packed and labeled accordingly by the pharmacy department at the cinic where the study took place and delivered to the research coordinator who was blind to the code.
Subjects were assigned random numbered envelopes sequentially. Interviews and treatment were completed by two psychiatrists and two psychiatric residents. The same physician followed the same participants in subsequent sessions to improve test-retest reliability and reduce physician-based variance. During the first visit, the participants were made aware of the purposes of the study, consented voluntarily, had their blood pressure and pulse rate measured, completed a battery of self-report instruments and were interviewed using the measures listed below.
After obtaining consent and completing instruments, the clinician interviewer dispensed randomized pill bottles to participants according to envelope number. Participants were given the immediate release (or placebo) capsule right after reactivation of their traumatic memory. This decision was made to offset the potential sedative effects of propranolol and the potential impact on the reactivation process. They were then instructed to take the slow-release dose two hours after the initial dose. No further doses were administered; the intervention was a single-session pharmacological procedure only. All individuals in the treatment condition received the same dose of propranolol. In the current protocol, participants first completed the memory reactivation procedure, after which propranolol (40mg immediate release) or the placebo was administered. This post-reactivation timing was selected in accordance with reconsolidation theory, which posits that memories enter a labile state following reactivation during which pharmacological interference may disrupt restabilization. Moreover, both groups received the same level of psychological intervention using elements of narrative therapy and CBT. Participants were asked to attend a follow-up review session four weeks later. The RCT was ended once all participants had completed their follow-up and were unblinded. Only after the review was complete was the randomization code revealed. This un-blinding provided an opportunity for placebo-treated participants to provide consent for propranolol treatment. The same procedure of assessment was then followed with a third and final visit four weeks later.
MeasuresMINI International Neuropsychiatric InterviewThe MINI International Neuropsychiatric Interview is a structured diagnostic instrument designed to make accurate psychiatric diagnosis.36 Reliability and validity for the MINI were good to excellent depending on the criterion being examined.37 The reliability and validity when compared to other measurement techniques and expert opinion has been discussed at length (see Hergueta et al.38 for a review). Administration results in a statistical manual of mental disorders (DSM) psychiatric diagnosis. The interview was completed once at the first visit.
The following were completed during each visit:
Clinician assessment of PTSD scale (CAPS) for DSM IVThis 30-item structured interview scale assesses trauma history and the frequency and intensity of the 18 DSM-IV PTSD symptoms39 with good psychometric properties.40 Studies indicate excellent reliability and validity of the CAPS-541 and a recent meta-analysis indicates a global reliability of α = 0.92.42
Zung self –rated anxiety scale (ZAS) & Zung self-rated depression scale (ZDS)The ZAS43 and ZDS44 are 20-item self-rated questionnaires which quantify the anxiety and depression experienced by an individual over a seven-day period, respectively.45-47 Multiple studies emerged from various countries supporting the validity and reliability of the ZAS.48 Reliability and validity information was not available for the ZDS, but the ZDS does correlated strongly with other self- and physician-measures of depression.49
Impact of event scale (IES)-revised (IES-R)The IES is a 22-item self-rated scale that distinguishes symptoms of avoidance, intrusion, and hyper arousal reactions to trauma exposure.50,51 The scale has shown high internal consistency (α = 0.92), acceptable measures of reliability and validity, and research has provided support for its use in the measurement of traumatic stress.52
Additional measuresNumbing-intrusions-hyperarousal (NIH) scaleThis unpublished visual analogue self-report scale quantifies numbing, intrusive thoughts and hyperarousal experiences over a seven-day period. Participants are presented with a visual scale anchored at 0 (not at all) and 10 (extremely) and are asked to indicate how much they have suffered from the characteristic (i.e., numbing, intrusive thoughts, and hyperarousal) in the past week. Ratings of 1 to 3 constitute mild suffering, 4 to 6 represents moderate suffering, and scores of 7 to 9 indicate marked suffering. Information on reliability and validity is not available for the NIH scale.
Traumatic memory description measure (TMDM)This scale was designed to record the qualitative and quantitative aspects of traumatic memories. The qualitative component records descriptions of traumatic memories and doubles as a reactivation probe. It places the individual cognitively at the very site and time of trauma and encourages reliving the moment and emotional reactions involved with the trauma. This scale has a range of 10- to 50-points and includes an assessment of level of distress, recollection of sensations, and traumatic event imagery. The higher the score the more integrated and distressing the memory. The scale was used successfully in a case series of 36 patients. A copy of this scale is available on request. Reliability and validity information is not available for the TMDM.
Patient global improvement scale (PGIS)This was a visual analogue scale commonly used to rate general patient improvement or lack thereof with a range of (-10 to +10). This is completed only at treatment visits. Reliability and validity information is not available for the PGIS, but it has been used in several studies on general studies of psychological/psychiatric symptomology and propranolol more specifically.15
Clinician global improvement scale (CGIS)This was visual analogue scale that clinician interviewers completed on their evaluation of patient responses to treatment, either placebo or propranolol. It has a range of (-10 to +10) and was completed only in the first post-treatment visit by interviewers blinded to the composition of the assigned groups. The CGIS demonstrates acceptable levels of reliability and validity with acceptable correlations and ICCs of 0.9.53
Analytic strategyDescriptive statistics were used to summarize the characteristics of the study participants. Means and standard deviations for continuous variables and percentages for categorized variables were presented. For group comparison of continuous variables t-tests were used. The Chi-square test or Fisher's Exact test was used for group comparison of categorical variables. All reported p-values were two-tailed, and α = 0.05 was set for statistical significance. Statistical analysis was performed using SAS software. Cohen’s d was calculated and interpreted based on the guidelines reported by Cohen54 where 0.2 represents a small effect, 0.5 represents a medium effect, and 0.8 represents a large effect.
ResultsAll participants revealed a traumatic event that precipitated traumatic memories. There was no significant mean difference in time of onset of traumatic memories between the placebo group (167.36 months) and the treatment group (163.63 months), p = 0.921. The two groups were very similar in terms of demographic and clinical characteristics (See Table 1), although a greater number of subjects in the placebo group were diagnosed with alcohol dependence (p = 0.033).
Demographic and clinical characteristics of the two groups.
| Characteristic | Propranolol(n=38) | Placebo(n=33) | p-value |
|---|---|---|---|
| Mean Age (SD) | 40.69 (±12.89) | 45.34 (±12.24) | 0.118 |
| Female (%) | 20 (52.63%) | 23 (63.89%) | 0.327 |
| Depression (%) | 31 (81.58%) | 29(80.56%) | 0.911 |
| Psychotic history prior*(%) | 11 (36.67%) | 9 (26.47%) | 0.380 |
| Alcohol dependence (%) | 3 (7.89) | 10 (27.78) | 0.033 |
| Employed (%) | 20 (64.52) | 18 (52.94) | 0.344 |
| Psychotropic medications (%) | 31 (81.58%) | 27 (75.00%) | 0.492 |
| Married | 16 (50%) | 18 (52.94%) | 0.811 |
| Mean age in months of traumatic memory (SD) | 163.63 (170.85) | 167.36 (149.85) | 0.911 |
Pre-and post-mean blood pressure and pulse rate measurements during both visits were not statistically different between placebo and treatment groups. Over four-fifths of participants reported having depression and three-quarters of participants were on psychotropic medications. All treatment group participants completed the study but four (11%) of the placebo group participants dropped out, ultimately, the placebo group reflect a total of 33 subjects (see Fig. 1).
Treatment comparisonsPrimary measures (CAPS, IES-R and TMDM)
Clinician and self-ratings of improvement were statistically higher in the propranolol group as compared with placebo: Self-rating using the PGIS was 4.63 (± 3.40) vs. 2.81 (± 3.97) and clinician rating (CGIS) was 4.34 (± 3.54) vs. 2.38 (± 3.96) while the propranolol group were almost double those of placebo at 4-week follow-up. After completion of the trial and unblinding, 70% of those on placebo that accepted treatment with propranolol showed an improvement in primary outcome measures. No statistically significant difference existed between the propranolol group and the placebo group in mean values for CAPS, IES-R and TMDM scores.
Secondary measuresBoth groups reported a reduction in depression levels as measured by the ZDS, with the treatment group experiencing a decrease in the mean score from 48.00 to 43.47 and from 52.28 to 40.86 in the placebo group. There was also a significant group by treatment interaction (p = 0.019) for depression (see Table 2). The significant group-by-time interaction for depression (p = 0.019) reflected greater reduction in depressive symptoms in the placebo group (Δ = −11.42) compared to the propranolol group (Δ = −4.53). This pattern is consistent with the known risk of depressive symptoms associated with beta-adrenergic blockade, which may have attenuated propranolol participants' improvement on this measure relative to placebo. There was no difference in the change in ZAS scores between propranolol and placebo groups.
Primary (CAPS, IES-R and TMDM) and secondary (NIH, ZAS, ZDS) variables before and after treatment mean scores.
Means for all study variables are listed in Table 2. Scores on the CAPS, IES-R, TMDM, ZAS, and ZDS decreased significantly across time in both groups. In the propranolol group, for instance, the reduction in anxiety, distress and trauma-related symptoms as measured by IES-R, CAPS and ZAS respectively were 46%, 40% and 20%. These are also clinically significant reductions in symptoms. Results are depicted graphically in Figs. 2–4. Outcomes for placebo participants who subsequently decided to accept the propranolol after unblinding are depicted in Figs. 3 and 4 (asterisked data points).
Significant effects of time were observed across all memory-related outcome measures; total scores for the IES-R, TMDM, and CAPS all decreased significantly across assessment points in both groups (see Table 2). No significant between-group differences were observed for these measures across time points.
Unblinding phaseOf the 33 placebo completers, 12 consented to receive propranolol following unblinding. Of these, 70% demonstrated improvement in primary outcome measures at the final assessment (week 8). These findings should be interpreted cautiously given the absence of a control comparison at this phase.
Adverse effectsHeadaches were reported in 5% of the patients in the propranolol group.
DiscussionThis RCT found no significant difference between propranolol and placebo on the primary outcome measures of traumatic memory and symptom severity. This null result is consistent with several prior controlled trials and meta-analyses suggesting limited evidentiary support for propranolol in this context.32,33 While clinician and self-rated global improvement were higher in the propranolol group, these subjective ratings should be interpreted cautiously as secondary, exploratory findings that are not corroborated by the objective primary measures. While these trends did not reach the level of statistical significance, there seemed to be several benefits of treatment with propranolol including clinically relevant decreases in the level and severity of anxiety and distresss. The discrepancy between null results on objective symptom measures and higher subjective global improvement ratings in the propranolol group warrants careful consideration. Several explanations are plausible. First, differential unblinding cannot be ruled out; propranolol's physiological effects, including reduced heart rate and blood pressure, may have allowed participants or clinicians to infer group assignment, introducing expectancy bias into subjective ratings. Second, the PGIS and CGIS may be capturing dimensions of functional or experiential improvement, such as reduced anticipatory anxiety or improved sense of control, that are not adequately indexed by the CAPS, IES-R, or TMDM. Third, the primary measures may have lacked sufficient sensitivity to detect clinically meaningful changes in this specific population, particularly given the sample size and the absence of a pure PTSD diagnosis requirement. These competing explanations cannot be disentangled with the current data and underscore the need for larger trials with objective physiological endpoints and blinding verification procedures.
While both groups showed reductions in depression over time, the significant group-by-time interaction (p = .019) indicated that this reduction was greater in the placebo group, a pattern consistent with the potential depressogenic effects of beta-adrenergic blockade potentially attenuating improvement in the propranolol group although the relationship between beta-adrenergic blockade and depressive symptoms remains debated.64 For both groups, the post-treatment mean scores for the primary outcome measures were significantly lower than pre-treatment mean scores suggesting a decrease in the main features of post-traumatic stress, anxiety, depression, intrusive, numbing, hypervigilance and traumatic recall. This assertion is supported in the propranolol group by the two-fold improvement in the severity and level of symptoms and the higher adherence rate of 100% in the propranolol group compared to 89% in the placebo group. Most relevantly, both patient and clinician-rated global clinical impressions were higher in the propranolol group compared to the placebo group. While propranolol may best be used clinically with more longitudinal administration, it is worth noting that the majority of reports on propranolol for anxiety-related conditions have similarly employed single-dose protocols. A notable exception is the emerging use of serial dosing in autism spectrum disorder trials, where repeated administration has shown promise for anxiety and cognitive flexibility.65 Whether a comparable serial dosing approach might yield more robust outcomes in traumatic memory populations warrants investigation.
These findings corroborate existing literature and suggest that a reduction in distress is specifically related to clinically relevant outcomes rather than physiological changes alone.23,24,30 The 29-point reduction in the total CAPS score for the propranolol group equates to a 40% reduction of mean post-traumatic symptoms (72 to 44)—a twofold improvement, when using a reduction of 15-points as evidence for clinical significance (see Fig. 2). Notably, this reduction moves the mean score from ‘severe’ PTSD symptomatology to ‘moderate’ symptomatology.41 The reduction in CAPS score is consistent with aggregate findings of improvement (10-63%) reported in a systematic review of PTSD treatment studies.41 The effect of propranolol in reducing traumatic symptoms is made more robust by clinical population studies25,55 and a previously published RCT which demonstrated similar improvements.24 An important difference in our protocol was that it allowed for one time administration of propranolol as compared to weekly for six weeks in other studies.24,25,30 A potentially critical methodological consideration concerns the timing of propranolol administration relative to memory reactivation. In the current protocol, propranolol was administered following memory reactivation, consistent with a reconsolidation interference model. However, an alternative and arguably more pharmacologically intuitive approach would be to administer propranolol prior to reactivation, allowing peak plasma concentrations to be present during the reactivation procedure itself. Propranolol is well established as an anxiolytic agent used widely for performance and test anxiety, where pre-event administration attenuates the adrenergic arousal that would otherwise consolidate or strengthen aversive memories.66 In the trauma context, pre-reactivation dosing would blunt sympathetic activation during the reactivation session, potentially producing a weaker and less emotionally charged reconsolidated trace. The post-reactivation timing used here may have been insufficient to interfere meaningfully with reconsolidation if the reactivation-induced labile window had already begun closing by the time drug plasma levels were adequate. Notably, Brunet and colleagues' positive RCT24 employed pre-reactivation propranolol, which may partly account for the discrepancy between their findings and the null primary results obtained here. The issue of timing is not simply a methodological footnote but may constitute a critical variable that future trials should address by systematically comparing pre- and post-reactivation administration.
Although a critical review of PTSD outcome studies identified propranolol as the most promising intervention agent, our results are measured on global outcomes.11 Other agents (i.e., morphine, ketamine) have been investigated as proposed treatments.4,9,11 We chose propranolol because of increasing theoretical evidence from the role of limbic structures that control reconsolidation,56 retrospective inquiry,28 a clinical case series with excellent results from our center,15 double blind studies,25 and a meta-analysis.8 showing categorically positive outcomes—although the results of more recent meta-analyses have been inconsistent.31,33 We used a specific memory tool, the TDMD, to access memory recall difficulties and gained more informed clinical outcome measures. This approach was suggested as a promising pathophysiological-based, theory-driven method for elucidating the diverse psychobiological complexity common in overall PTSD. Such an approach is needed to guide rational pharmacotherapy.57 Scale coherence revealed no significant differences between groups but rather clinically relevance reductions in nightmares, dreams, and relived experiences. However, two IES-R features were lower among the propranolol group. These features are intuitively linked to traumatic memories. Since there is no known memory-specific scale, more work is necessary to develop a new scale or refine the current TMDM. Given the lack of statistical significance, these findings seem to be more in line with meta-analyses by Steenen and colleagues33,58 which suggest that the evidentiary support for the use of propranolol is limited.
Our placebo group, by receiving elements of ‘psychological treatment,’ was not strictly a control group. Narrative exposure therapy incorporates similar reactivation of traumatic memories59 and brief cognitive behavior therapy (CBT) for PTSD is premised on similar principles.59,60 Cognitive behavior therapy approaches such as prolonged exposure and cognitive processing therapies are more effective for PTSD than pharmacotherapy61 and positive outcomes following a single session have been obtained among traumatized earthquake victims.62 The placebo group is therefore best conceptualized as a parallel treatment group involved in a head-to-head psychological and pharmacological design. Repeating the narrative exposure in written and verbal form resembles procedures such as guarded flooding to traumatic images, desensitization by recounting after 4 to 8 weeks, and successful psychologically based treatments like CBT and eye movement desensitization and reprocessing (EMDR).63 Importantly, the placebo group reactivates traumatic memories, undergoes emotional processing and achieves a sense of control over their fears.59,60 However, the higher dropout rates in the placebo group and 70% improvement in the placebo arm that received propranolol support tolerability and potential usefulness of propranolol. The amnestic effect of propranolol may reduce links between memories and reliving of experiences.
The substantial within-group improvement observed in the placebo group warrants careful consideration and may in fact represent one of the more clinically meaningful findings of this trial. Both groups underwent an identical memory reactivation procedure (i.e., writing and verbally recounting their traumatic experience in the presence of a clinician) which closely mirrors the core mechanism of established exposure-based therapies including prolonged exposure, narrative exposure therapy, and EMDR. It is therefore plausible that the reactivation procedure itself functioned as an implicit form of exposure therapy for both groups, producing genuine symptom reduction independent of pharmacological effects. This interpretation has important implications for interpreting the null between-group result. If the reactivation procedure produced a strong therapeutic effect in both arms, a ceiling effect on symptom improvement may have obscured any additional pharmacological benefit of propranolol — that is, both groups may have improved to a degree that left limited remaining variance for a drug effect to be detected. This is distinct from a floor effect at baseline; rather, it reflects the possibility that the behavioral component of the intervention was sufficiently potent to drive improvement in both groups toward a similar endpoint, compressing between-group differences on the primary outcome measures. This interpretation is consistent with findings that cognitive behavioral approaches are generally more effective for PTSD than pharmacotherapy alone.61 Critically, this ceiling effect hypothesis cannot be disentangled from a blinding integrity explanation with the current data — if propranolol participants inferred their group assignment through somatic cues, expectancy effects may have additively contributed to their subjective improvement ratings. The convergence of these two mechanisms, a potent behavioral intervention in both arms and potential expectancy amplification in the active arm, provides a parsimonious account of the observed pattern: substantial within-group improvement, null between-group differences on structured measures, and higher global impression ratings in the propranolol group. Future studies seeking to isolate pharmacological effects of propranolol on traumatic memory should carefully consider whether the reactivation procedure itself constitutes an active treatment component. Designs employing a minimal-contact control condition, active placebo, or dismantling methodology would help clarify the relative contributions of pharmacological and behavioral mechanisms.
Limitations and future directionsWhile this research has advanced our understanding of specific memory items as treatment targets, we recognize the following limitations to our results. First, while a specialized population, the sample included in this study was quite small and potentially not representative of the population. Given the small sample size employed in this study, it is quite possible that many of these statistical tests were underpowered, potentially leading to null findings. Additionally, due to multiple comparisons with t-tests, chi squares, and Fisher’s exact tests, there is an increased risk of error in the results. Future studies should strive to replicate these findings in a large, representative sample, using more complex, robust methods capable of determining cause and effect. Second, the psychometric properties of the TMDM need to be established in larger validation studies to improve scale utility. Given the focus on amnesia and memory recall, the available scales (CAPS, IES-R, NIH) used were not developed for those purposes. Third, there remains a need for standardization of dose and administration of propranolol. Additionally, the timing of memory reactivation relative to drug administration is theoretically critical, as propranolol must be present during the reconsolidation window to interfere with memory restabilization. Variations in this timing (i.e., whether reactivation occurred before peak plasma concentration was achieved or after the window had closed) may have influenced the robustness of our findings. Future studies should employ stricter pharmacokinetic monitoring to optimize the alignment of reactivation procedures with peak drug availability. Fourth, the brief encounter with study participants may limit the recounting of traumatic events. Propranolol may best be used clinically with more longitudinal administration as seen recently with autism trials.65 Fifth, the placebo group in this study was not a strict placebo group but, rather, a group that received an established treatment for PTSD without the addition of propranolol. While this was unavoidable given the design of the study, it does complicate the interpretation of the results and could have potentially obscured some potential significant findings. Moreover, the fact that two-thirds of the sample were on psychotropic medications could have presented a confound and further complicated interpretation of the data. Concomitant psychotropic medication dosages were not recorded, precluding examination of potential pharmacodynamic interactions with propranolol or interference with memory reconsolidation mechanisms. Given that approximately two-thirds of participants were receiving psychotropic medications at the time of the trial, this represents a meaningful interpretive limitation. Future studies should systematically document medication type, dosage, and duration as covariates, as these factors may modulate both the pharmacological and behavioral components of the intervention. Furthermore, there were significantly more individuals who reported having an alcohol dependency issue in the placebo group which may introduce an additional confound. Another notable limitation of this trial concerns blinding integrity and its potential impact on outcomes. The treatment arm demonstrated perfect retention (100%) while 11% of placebo participants dropped out — a pattern that is atypical for pharmacological trials, where active treatment arms more commonly experience higher attrition due to side effects. Propranolol produces readily detectable somatic effects including reduced heart rate and blood pressure, which may have allowed participants to infer their group assignment. Formal blinding verification was not conducted, which represents a methodological gap. If unblinding occurred, expectancy effects could have inflated subjective global improvement ratings in the propranolol group, potentially accounting for the discrepancy between higher PGIS and CGIS scores on the one hand and null results on structured symptom severity measures on the other. Future trials of propranolol should incorporate formal blinding checks, physiological monitoring to detect expectancy-related responding, and active placebos that better mimic the somatic profile of propranolol to minimize differential unblinding. These limitations can be addressed as we plan to review the qualitative aspects of the narratives written by participants. We also intend to obtain and analyze follow-up data to determine the stability of improvement. The strengths of this study include the inclusion of a representatively clinically ill population and the use of clinically relevant outcome measures.
ConclusionThis RCT did not demonstrate a statistically significant benefit of propranolol over placebo on primary measures of traumatic memory symptom severity. These findings are consistent with the broader literature suggesting limited and inconsistent efficacy of propranolol for post-traumatic presentations. The within-group reductions observed in both groups may reflect the therapeutic effects of structured memory reactivation in a clinical context, a finding with implications for understanding the active components of trauma-focused interventions. Secondary subjective improvement ratings favored propranolol and may warrant further investigation in adequately powered trials with verified blinding and objective physiological endpoints. Based on the current evidence, propranolol alone cannot be recommended for inclusion in clinical guidelines for traumatic memory treatment. A key direction for future research is the systematic examination of propranolol administration timing relative to memory reactivation. The current protocol's post-reactivation approach may have limited pharmacological efficacy; pre-reactivation administration, consistent with propranolol's established use in performance anxiety contexts, may more effectively attenuate adrenergic arousal during memory labilization and warrants investigation in adequately powered trials.
Availability of data and materialThe datasets generated and analysed during the current study are available from the corresponding author on reasonable request.
Authors' contributionsConceptualization– Robin, Mansfield, Tamara, Curtis; Data curation – Lim; Formal analysis – Lim; Methodology – Robin, Tamara; Project administration – Claudia, Alexandra; Resources – Robin; Supervision – Mansfield; Writing – original draft – all except Robin; Writing – review & editing – all except Robin
Ethical ConsiderationsThe study was approved by the University of Saskatchewan and certified by Health Canada under the regulations of pharmacological products guidelines. All participants provided their informed consent to participate. The manuscript has been read and approved by all named authors. The randomized control trial and research work stated in this manuscript continued after the death of Dr. Robin Menzies.
FundingNo funding was provided for this research
The authors have no conflicts of interest to declare.
The authors appreciate the assistance of the clinic staff for their assistance in maintaining compliance with regulatory authorities and Health Canada. A special thanks to Dr. L Balbuena for his critical review of and contribution to the manuscript.
ClinicalTrials.gov (Randomized controlled trial number-NCT01069159).
https://clinicaltrials.gov/ct2/show/NCT01069159?term=propranolol+treatment+of+traumatic+memories&rank=1
The randomized control trial and research work stated in this manuscript continued after the death of Dr. Robin Menzies.
articles
