Prevalence of inconsistency in trauma reporting

In this study, we explored the prevalence and predictors of the inconsistency of reporting trauma-related questions among HCWs who were exposed to the Beirut blast. The majority of our sample (72.4%) inconsistently answered the trauma-related questions. Most of them either exaggerated (36.43%) their trauma reports or diminished (35.71%) their accounts.

The rate of inconsistency in our sample aligns with the patterns found in the literature. Studies following veterans and trauma survivors have reported inconsistency rates in trauma reporting ranging from 50 to 89%. ^{Reference Fergusson, Horwood and Woodward12,Reference Southwick, Morgan, Nicolaou and Charney21,Reference Krinsley, Gallagher, Weathers, Kutter and Kaloupek29} Moreover, in community studies such as Hepp et al’s, a 63.9% inconsistency rate was revealed when participants were asked the same trauma-related questions after 6 years. ^{Reference Hepp, Gamma, Milos, Eich, Ajdacic-Gross and Rössler6}

Age and the focus on childhood adversities also play a significant role in these findings. Previous studies have shown that older age is a significant risk factor for memory distortion, which contributes to higher inconsistency rates. For example, Burns et al found a lower rate of 16.3% inconsistency in a cohort of children and adolescents, where participants inconsistently reported their trauma after initially endorsing it in their first interview. ^{Reference Burns, Coughlan and Cannon10} This lower rate can be attributed to the fact that the study focused exclusively on childhood maltreatment and involved a younger sample. Similarly, Coleman et al observed an inconsistency rate of 40% after a 12-year follow-up in a community sample. ^{Reference Colman, Kingsbury, Garad, Zeng, Naicker and Patten7} The lower rates compared with our study are likely because their questions were limited to childhood maltreatment alone. The narrower focus and younger age range in these studies help explain the disparity in inconsistency rates compared with our findings.

Predictors of trauma exaggeration

Our findings show that probable PTSD secondary to the Beirut port blast, probable depression and psychological distress are all significantly associated with exaggeration or inflation of trauma endorsement. This is in line with the literature, which states that psychopathologies and mental health outcomes predict memory distortion. There are many explanations for these findings.

PTSD and memory inflation

We know from the literature that PTSD is commonly associated with memory inflation of traumatic events, ^{Reference Roemer, Litz, Orsillo, Ehlich and Friedman20,Reference Southwick, Morgan, Nicolaou and Charney21,Reference Heir, Piatigorsky and Weisæth30,Reference Schwarz, Kowalski and McNally31} and this can be explained in two ways.

First, a key feature of PTSD diagnosis is intrusive memories. ^{Reference Golier and Yehuda18,Reference McNally19} These memories can lead to disrupted and involuntary recollection of traumatic experiences, fostering the inflation of traumatic memories. ^{Reference Roemer, Litz, Orsillo, Ehlich and Friedman20}

Second, among the 13 individuals who developed PTSD at wave 4 and showed memory exaggeration, the diagnosis can be classified as delayed PTSD, occurring two years after the blast. Delayed PTSD has been shown to predict memory inflation, ^{Reference Heir, Hussain, Kristensen and Weisæth32} offering an additional explanation for our findings alongside intrusive symptoms.

In our sample, the variable corresponding to the trend of probable PTSD diagnosis could not fully account for the inflation of traumatic memory endorsement. This limitation stems from the small sample size and the uniform experience among participants who developed PTSD by wave 4. Specifically, all individuals in this subgroup exhibited either memory exaggeration or distortion; none reported no change in their trauma reporting. Consequently, the lack of variability in our outcome variable prevented us from drawing definitive conclusions regarding the influence of PTSD development. However, supplementary analysis revealed that only PTSD at wave 4 was associated with the inflation of traumatic memory reporting. Therefore, it is possible that the observed association is primarily driven by the onset of PTSD at wave 4.

Hence, despite being obscured in the main analyses due to limited sample size and the lack of variability in the variable under study, our results show that developing PTSD by wave 4 predicts memory exaggeration. This can partially be explained by Ehlers et al’s cognitive model of PTSD which states that the persistence of the disorder is often associated with how serious and current the threat of the trauma is processed. ^{Reference Ehlers and Clark33} This in part is explained by the disturbance of autobiographical memory, including selective attention to threat and memory amplification which is evident in our sample.

A further mechanism worth considering is priming, which the literature describes as the activation of trauma-related cognitive networks in response to reminders of the event. ^{Reference Ehlers and Clark33,Reference Ehring and Ehlers34} Empirical studies show that individuals with PTSD exhibit heightened sensitivity to trauma cues, and such cues can trigger symptom reactivation and influence memory reconstruction. Because our study assessed trauma repeatedly across waves, within a context where the Beirut blast remained highly salient, participants may have been repeatedly primed through the survey itself. This repeated activation could bias the consolidation and retrieval of trauma memories towards inflation. Moreover, the directionality between memory and PTSD is unlikely to be linear; PTSD may increase the likelihood of exaggerated recall, while exaggerated recall may reinforce trauma-related threat appraisals, thereby maintaining or intensifying PTSD symptoms. Integrating a priming perspective therefore provides a more comprehensive understanding of the cognitive processes that may underlie, partially, the inconsistency and inflation observed in our sample.

Depression, psychological distress and trauma memory inflation

We have shown before in a nationally representative sample of Lebanese participants that the multiple stressors that the country has witnessed (including the financial meltdown and the global pandemic) were the main drivers of the observed high rates of mental health disorders in the sample. ^{Reference Karam, El-Jamal, Osman, Toukan, Mouawad and Al Barathie35} This is relevant to our findings because the majority of our sample had psychological stress in either wave 3 or 4. We believe that this increased rate of distress alongside the observed rates of depression also contributed to the exaggerated trauma responses in our sample.

Research shows that both negative mood and clinical depression bias the effect of memory recollection – this is known as the mood congruency theory. ^{Reference Dalgleish and Watts11} Particularly, psychological distress and depression induce the recollection of sad memories. Depressed individuals are more likely to experience recall bias, particularly for negative information, which influences the increased reporting of trauma-related memories. ^{Reference Beck, Rush, Shaw and Emery36}

Our findings add to the literature on trauma-related cognitive processes by suggesting that trauma memory may be influenced by the presence and severity of PTSD symptoms. These processes are not fully captured by general mood-congruent memory effects, but are instead grounded in trauma-specific schemas that may bias the encoding and retrieval of prior adverse experiences, particularly when contextual cues resemble aspects of the index trauma. ^{Reference Moore and Zoellner37}

Also, the directionality of the observed associations warrants careful consideration. While our cross-sectional findings suggest that PTSD symptoms may have an impact on the recall, it is also plausible that memory inflation or reinterpretation over time contributes to the persistence or intensification of PTSD symptoms. The relationship is likely bidirectional, and future longitudinal or experimental studies would be required to disentangle these pathways. ^{Reference Rubin, Berntsen and Bohni38}

This aligns with our findings. Our sample demonstrated that individuals who developed depression by wave 4 were at a higher risk of inflating their trauma-related memories. This finding is similar to those of Coleman et al, ^{Reference Colman, Kingsbury, Garad, Zeng, Naicker and Patten7} and others who have shown that the development or worsening of depressive symptoms over time predicts the exaggeration of traumatic memories. ^{Reference Fogarty and Hemsley16}

Generalisation and psychopathology

Another explanation for the observed inflation of traumatic events in our sample is the overgeneralisability associated with psychopathologies. ^{Reference Beck, Rush, Shaw and Emery36} According to the literature, individuals with psychopathologies are more likely to generalise their reports of traumatic events. This tendency to overgeneralise could also be contributing to the inflated reports observed in our study.

Lack of predictors for memory diminishment

We should point out that in our study, none of the predictors explained memory diminishment. This is not an unusual finding since many have also found that after being exposed to a major stressful event – which in our case is the Beirut port blast – individuals tend to exaggerate their stressful events as time passes. ^{Reference Roemer, Litz, Orsillo, Ehlich and Friedman20,Reference Southwick, Morgan, Nicolaou and Charney21,Reference Heir, Piatigorsky and Weisæth30,Reference Schwarz, Kowalski and McNally31} This can explain why none of our variables predicted memory diminishment.

Study limitations

Our study has several limitations that should be acknowledged. First, we did not collect data on the age at which participants experienced their trauma. Consequently, we could not assess how the timing of the traumatic event influenced the consistency of recall. Second, the trauma-related memories reported by HCWs in our study were self-reported, and we did not have an assessment to verify these reports. This is a significant limitation because research indicates that discrepancies between objective and subjective reports of maltreatment and trauma are common. ^{Reference Baldwin, Reuben, Newbury and Danese8,Reference Francis, Tsaligopoulou, Stock, Pingault and Baldwin9,Reference Fergusson, Horwood and Woodward12} Although classifying a memory as definitively ‘true’ is not always accessible, we encourage objective verification of reported incidents to reduce inconsistency in trauma reporting. This is especially true for self-reported measures that can suffer due to the participant’s potential inattention during the administration of the survey. Hence, without objective verification, we cannot determine the accuracy of the reported traumas, which has an impact on the validity of our findings regarding memory diminishment and exaggeration. Third, while we included a question assessing whether participants were physically injured in the explosion (as an objective indicator of threat), we did not assess their subjective perception of life threat at the time of the event. This may influence the encoding and recall of traumatic memories. Lastly, we did not control for several behavioural factors that can influence the consistency of trauma reporting. These include self-esteem, chronic stress, mastery, psychotic episodes in addition to treatment-seeking options. ^{Reference Colman, Kingsbury, Garad, Zeng, Naicker and Patten7,Reference Burns, Coughlan and Cannon10} The lack of control for these variables may have affected the results, as these factors can also contribute to variations in trauma recall.

Strengths and recommendations

One of the strengths of our study is that we controlled for age. This is particularly important as older age is associated with memory diminishment. By accounting for age, we minimised its potential confounding effect on the consistency of trauma recall. Moreover, we also controlled for social desirability bias by ensuring that the responses were anonymous. Participants were assured that their identities would not be disclosed since the surveys were sent via email. This approach likely encouraged more honest and accurate reporting of trauma-related memories.

For future studies, it is recommended to ask about the severity and onset of the trauma. Collecting this data would allow researchers to objectively trace the trauma over time and better understand its impact on memory consistency. Moreover, to address the limitation of relying on self-reported trauma, future research should incorporate objective assessments to verify the reported traumas. This could involve cross-referencing self-reports with medical records or using validated trauma assessment tools.

Research implications

Our study has important implications for trauma research among disaster survivors such as military personnel, war refugees or emergency responders: inconsistencies in self-reported trauma may reflect dynamic psychological processes rather than measurement error alone. Researchers should therefore interpret changes in trauma narratives with caution, consider repeated assessments over time and incorporate mental health indicators when analysing self-reported trauma data. This approach can improve the validity of trauma research findings and ensure more accurate relationships between trauma exposure and mental health outcomes.

In conclusion, this study provides insight into the variability of trauma reporting among HCWs following the Beirut blast, where over 70% of participants exhibited inconsistency in their accounts. The strong correlation between trauma exaggeration and mental health conditions like PTSD and depression suggests that these psychological factors may distort trauma memories, leading to inflated reports over time. Our findings align with existing literature, reaffirming the need for clinicians and researchers to be aware of the potential biases in self-reported trauma, particularly in individuals with underlying psychopathologies.