A Prediction Model of PTSD in the Israeli Population in the Aftermath of October 7th, 2023, Terrorist Attack and the Israel-Hamas War

Dana Katsoty, Michal Greidinger, Yuval Neria, Aviv Segev, Ido Lurie
doi: https://doi.org/10.1101/2024.02.25.24303235

ABSTRACT

Background On October 7th, 2023, a mass terror attack was launched by Hamas militants, followed by the Israel-Hamas war. These events are considered a nationwide trauma with paramount ramifications for public mental health.

Aim To develop an evidence-based model for the prediction of posttraumatic stress disorder (PTSD) prevalence related to the terrorist attack and the war.

Methods Firstly, the Israeli population was divided into six groups based on intensity, context, and type of traumatic exposure (direct exposure to terror, close proximity to terror, soldiers in combat and support units, intense exposure to rocket attacks, moderate exposure to rocket attacks, and indirectly affected communities), and the population size of each group was assessed using official national databases. Secondly, an estimation of the expected prevalence of PTSD in each of the exposure groups was based on a literature review of PTSD prevalence following exposure to war and terror, relevant to each group. A random-effects meta-analysis of the prevalence rates was conducted separately per each exposure group. Finally, the predicted number of PTSD causalities in the population was calculated by multiplying the group size and the PTSD prevalence estimation.

Results Based on population size and estimated PTSD prevalence within each exposure category, the model predicts that approximately 5.3% (N=520,278) of the Israeli population (confidence interval, 160,000-881,000), may develop PTSD in the aftermath of the terror attack and the war.

Conclusions The predicted number of PTSD following mass trauma is expected to be considerable. The presented model can assist policymakers, clinicians, and researchers in preparing and devising adequate interventions for the mental health needs of large populations. Moreover, this model can be applied in other instances of mass-trauma exposure.

INTRODUCTION

The terrorist attack conducted by Hamas on October 7th, 2023, is a nationwide mass trauma with critical implications for public mental health across large populations, as it is one of the deadliest terror attacks since 1970 in absolute number of fatalities (Statista, 2023). On that day, Hamas militants infiltrated Israel and targeted Israelis in communities surrounding the Gaza Strip. This assault resulted in a mass tragedy, claiming the lives of over 1,100 Israeli civilians, soldiers, and foreign citizens, with 248 people abducted to the Gaza Strip, including infants, children, women, men, and elderly people. Thousands of civilians were injured, and many were subject to torture and sexual violence. Within the first 4 hours of the attack, 3000 rockets were fired into Israel. The attack initiated an ongoing war between Israel and Hamas that had devastating effects on civilians both in Israel and in Gaza, including mass casualties, injuries, and displacement of civilians (Devi, 2023; Fox and Kolitz, 2023; Gavrielov, 2023; Peled and Jones, 2023; Statista, 2023; Saidel et al., 2024). Many of the atrocities were also filmed and virally spread online through social media (Cortellessa, 2023; Impelli, 2023).

It is crucial to assess the expected psychiatric outcomes of such atrocities, as exposure to war and terror has detrimental implications for the mental health of both individuals and communities (e.g., Neria, et al., 2007; Hoppen, et al., 2021). One of the most prominent outcomes of such exposure is post-traumatic stress disorder (PTSD; (5th ed., text rev.; DSM-5-TR; American Psychiatric Association). The burden of PTSD is substantial, with high direct and indirect healthcare costs (Davis, et al., 2022; Warth, et al., 2020), and even more so when the disorder is untreated (e.g., Pityaratstian et al., 2015). A plethora of research assessed the prevalence of PTSD in the aftermath of exposure to war and terrorism and has demonstrated a notable dose-response relationship between the type and intensity of exposure and the risk for PTSD (e.g., Johnson & Thompson, 2008; May & Wisco, 2016; Neria, et al., 1998; Neria, et al., 2006; North, 2011). Yet, while most terrorist attacks involve a segment of the population, in the October 7th attack and the Israel-Hamas war the entire Israeli population was exposed to the atrocities to some extent, and like other instances we assume immense variability in the context, type, and intensity of exposure across the country.

To systematically estimate the risk for PTSD across Israeli society, we developed an evidence-based predictive model for the prevalence of PTSD across different groups of the Israeli population, depending on the varying levels of exposure to the October 7th terror attacks and the Israel-Hamas war. This prediction model for PTSD prevalence in populations exposed to a large-scale trauma could be applied in future occasions of mass trauma events, with relevant adjustments in terms of exposure type and extent.

METHOD

Procedure

The development of the prediction model consisted of the following steps: 1) Classifying exposure groups across the Israeli population, based on different elements of the traumatic exposure (e.g., the context in which exposure occurred, the intensity of exposure to attacks, and the type of exposure-whether it included direct exposure to a terrorist attack, or other forms of exposure, such as rocket attacks). The size of each of these exposure groups was estimated according to official government-originated information, media publications and national open-access information; 2) Estimating the expected prevalence rates of PTSD in each of the exposure groups. This was done via three steps: First, we conducted a literature review concerning PTSD prevalence following exposure to war and terror. We searched for reviews, systematic reviews, and meta-analyses of the PTSD prevalence in PubMed using the terms: (PTSD OR post-traumatic stress disorder) AND (prevalence OR incidence OR rate*) and (war OR terror), and also relied on Google Scholar for an additional search. For each exposure group, we attempted to identify review papers that reflect the nature of the traumatic exposure of the specific group. Second, we examined all studies within each review and omitted studies that did not meet inclusion criteria (that are addressed under the methodological challenges section). Third, we conducted a random-effects meta-analysis of the prevalence rates separately per each group, based on all studies that met inclusion criteria within each review. The meta-analysis aggregated the proportion of PTSD cases out of each sample, across all studies (i.e., comparison between groups with respect to a dichotomous dependent variable, which is PTSD prevalence); 3) Calculating the predicted number of PTSD causalities in the population by multiplying the group size and the PTSD prevalence estimation.

Classifying Exposure Groups

  1. Direct exposure to the October 7th terror attacks. These attacks included two main areas: the music festivals in Kibbutz Reim, and communities around the Gaza Strip and surrounding cities.

  2. Close proximity to the October 7th terror attacks. Communities and areas within cities in this group were not infiltrated by Hamas perpetrators, but their residents were under severe perceived life threat due to the notion of the invasion to nearby areas, and the high probability of the perpetrators invading their homes. Many of these civilians were hiding in shelters for hours or even days during and after the attack (Awad & Lubell, 2023).

  3. Soldiers in combat and support units involved in the war. This group includes active duty and reserve soldiers. Both infantry units and soldiers in support units, which were not directly involved in combat, were included in this group.

  4. Civilians under intense exposure to rocket attacks (living up to 40 kilometers (25 miles) from the Gaza Strip). Civilians who live in the cities and communities located up to 40 km from the Gaza-Israel border and have been under recurring rocket attacks. The delineation of 40 km, as utilized in various studies (Besser et al., 2013; Besser et al., 2014; Besser et al., 2015; Gil et al., 2015; Gil et al., 2016; Weinberg et al., 2015), was selected based on the distinct characteristics of this area, which involves a brief window of time to reach shelters immediately after the rocket siren is activated (up to 1 minute; Israel Home Front Command, 2023). This group was also exposed to the largest number of rocket attacks (Israel Home Front Command, 2023).

  5. Civilians under moderate exposure to rocket attacks (living 40-80 km (25-50 miles) from the Gaza Strip). Civilians living between 40-80 kilometers from the Israel-Gaza border that have been exposed to rocket attacks to a substantial degree, but to a smaller extent compared to the previous group, and with a longer duration of time to enter shelters (1 to 1.5 minutes; Israel Home Front Command, 2023).

  6. Indirectly affected communities (living more than 80 km (50 miles) from the Gaza Strip). Civilians who live in communities and cities located more than 80 km from the Israel-Gaza border. Some of them have suffered rocket attacks, but to a substantially smaller degree, which we considered less threatening and interfering with everyday life. Additionally, they have been exposed to the traumatic events through media coverage. To note, this group includes the rest of the Israeli population that was not included in the previous categories.

Addressing Methodological challenges

The development of the predictive model required addressing several significant methodological and theoretical challenges, each demanding a distinctive solution:

Variability in PTSD prevalence estimates

Past research has demonstrated striking differences across studies assessing PTSD prevalence. This variability was attributed to methodological differences among the studies, such as differences in the method of PTSD measurement (i.e., clinical assessments vs. self-report questionnaires), time of measurement after the trauma, and specific characteristics of the exposure (e.g., Sundin, et al., 2010; Hoffman, et al., 2011).

Method of PTSD measurement

Two methods of measurement are used to determine the occurrence of PTSD. The first is clinical assessments, which include validated diagnostic interviews with explicit PTSD diagnostic criteria based on the Diagnostic and Statistical Manual of Mental Disorders (5th ed., text rev.; DSM-5-TR; American Psychiatric Association) or the International Disease Classification (11th ed.; ICD-11). The second method is self-report questionnaires or symptom checklists, wherein individuals are classified within a category of probable PTSD when exceeding a specific questionnaire cutoff (e.g., Neria, et al., 2011). These two methods of measurement have been shown to yield significantly different PTSD prevalence estimates, and self-report questionnaires have been argued to overestimate the prevalence of PTSD (Turner et al., 2003). More generally, as self-report questionnaires or symptom checklists do not properly assess the severity, frequency, duration, covariation, and degree of interference (McNally, et al., 2003), and since clinical assessments are generally considered the gold standard for PTSD diagnosis, we preferred clinical assessments in the literature search process. In case a review included both clinical assessments and questionnaires, we omitted the questionnaire estimates, and included only clinical assessments in the meta-analysis. In instances where no studies using clinical assessment were included in a review, we implemented a mitigating correction factor in order to maintain the most conservative estimates. A factor of 0.71 was chosen as self-rating and symptoms checklists were found to overestimate the prevalence of PTSD by 41% (Steele, et al., 2009).

Time of measurement after trauma exposure

Although PTSD can only be diagnosed one month or more after the traumatic exposure, several studies have examined PTSD symptoms within the time frame of the traumatic exposure, or less than one month after the exposure. These assessments reflect acute stress disorder (ASD) symptomatology and not PTSD. Therefore, studies that examined post-traumatic symptoms less than one month after traumatic exposure were omitted from analyses. Additionally, in case a specific study included more than one measurement, we included only the highest prevalence estimate (as long as that measurement was conducted one month or more after the traumatic exposure). Importantly, we only included current, point-prevalence estimates, as opposed to period prevalence, to avoid inflation of estimates.

The unique nature of exposure and implications for the literature review process

The events of the war were characterized by unique types of exposure to terror and violence. Therefore, for some of the exposure groups, we were not able to find a review that reflected the nature of exposure. This was the case for two groups: group 2 (close proximity to the October 7th terror attacks), and group 5 (moderate exposure to rocket attacks). To derive estimates for these two groups, we relied on estimates from the other exposure groups, and addressed the specific manner in which the estimate was calculated.

Prevalence rates in children vs. adults

Children may respond to trauma differently compared to adults (Shaw, 2003). The majority of studies conducted regarding PTSD prevalence focused on adults, and studies that compared PTSD prevalence in adults and children demonstrated conflicting results (Tang, et al., 2017, Fairbank & Fairbank, 2009, Pat-Horenczyk, et al., 2014). Due to the ambiguity of these estimations, we did not distinguish between children and adults within each exposure group.

Other issues

Notably, many individuals in each group had more than one form of exposure (e.g., survivors of the October 7th attack were also subject to rocket attacks). In cases of multifaceted exposure, we chose the highest prevalence rate. Additionally, we omitted studies that examined only populations seeking mental health services and studies conducted in clinical settings, as these are expected to include higher PTSD prevalence compared to general samples of people who were exposed to trauma.

Statistical analysis

Meta-analyses were conducted using R package metafor (Viechtbauer, 2010).

RESULTS

We present the identified reviews and address the number of studies that were omitted within each review (if such omission was conducted). The supplementary material includes the full list of excluded studies within each review. We then report the results of the meta-analyses for groups in which estimates were derived from meta-analyses (including prevalence estimates, confidence intervals, and heterogeneity estimates). The supplementary material includes forest plots of summary effects for each of the meta-analyses. We also describe how group sizes were determined (with a full elaboration in appendix 1) and present the number (and range) of people within each group that are expected to develop PTSD, calculated by multiplying the estimate and confidence interval by the group size. For groups in which estimates were not derived from meta-analyses, we address the manner in which the estimate was determined, and address determination of group size, estimate and range.

Table 1 summarizes the results of the prediction model seperately for each exposure group described above –the population size estimate, the probability of developing PTSD (prevalence estimate and range), and the expected number of people with PTSD (the number of people within each group multiplied by the prevalence estimate).

View this table:
Table 1: Results of the prediction model

Prevalence estimates derived from systematic reviews and meta-analyses

Group 1: Direct exposure to the October 7th terror attacks

We relied on a systematic review by Paz García-Vera, et al. (2016), which examined several types of exposure to terrorism, one of which was direct exposure to terror attacks. All studies presented for in this review included clinical assessments.

Omission of studies

One study was omitted as it included help-seeking population. As mentioned earlier, in longitudinal studies we included measurement with the highest prevalence.

Meta-analysis of included studies

Nine studies were included int the meta-analysis. The mean prevalence estimate was μ^ = 0.31 (95% CI[0.21-0.41]). The Q-test indicated that outcomes are heterogeneous, suggesting that different studies vary in the observed prevalence estimate. Moreover, the I2 statistic indicated that a meaningful portion of the variance could be explained by true heterogeneity rather than chance (Q(9) = 219.79, p < 0.001, I2 = 93.71%, τ = .15).

Group size and numerical estimate and range

We included communities and areas within cities that were infiltrated by Hamas perpetrators, a total of 39,664 people (Israel Central Bureau of Statistics, 2022; Almog, 2024; Tzalah, 2023). After multiplying the estimate and CI by group size, we predict that 12,366 (a range of 8,340 to 16,392) people are expected to develop PTSD.

Group 3: Soldiers in combat and support units involved in the war

We identified two literature reviews that were relevant for this exposure group. Both reviews examined combat-PTSD after war exposure (Richardson, et al, 2010; Xue, et al., 2015). These reviews included both clinical assessments and questionnaire assessments.

Omission of studies

In the review by Richardson, et al. (2010) we omitted studies for the following reasons: 18 studies included questionnaire assessments or non-clinical assessments; two studies re-analyzed existing data from another sample; one study included help-seeking clinical population; and one study that did not examine current PTSD prevalence. In the review by Xue, et al., (2015), we omitted studies for the following reasons: 23 studies included questionnaire assessments or non-clinical assessments; three studies examined PTSD diagnosis in medical files without specific reference to the war exposure; and three studies did not examine current PTSD prevalence.

Meta-analysis of included studies

Four studies were included in the meta-analysis. The mean prevalence estimate was μ^ = 0.08 (95% CI[0.01-0.14]). The Q-test and I2 statistic indicated that outcomes are heterogeneous and that a meaningful portion of the variance could be explained by true heterogeneity rather than chance (Q(3) = 226.18, p < 0.001, I2 = 99.38%, τ = .06).

Group size and numerical estimate and range

An estimate for the group size was made based on reports in the media of 10 divisions operating in the escalation areas: the Gaza Strip, the Israel-Lebanon border, and the West Bank (Israel Defense Forces, 2023; Israel Defense forces, 2024; Zeitoon, 2023). The size of a division ranges from 13,000 to 20,000 soldiers, with an average of 16,500, and a total of 165,000 people (Hamichlol, 2023). Importantly, the 20,773 soldiers that were subtracted from the soldiers groups were classified under groups with higher prevalence rates, since in cases of multifaceted exposure, individuals were classified under the higher-prevalence group (see appendix 1 for an elaboration). After this subtraction, the group size was 144,227 people. After multiplying the estimate and CI by the group size, we predict that 11,021 (a range of 1,494 to 20,549) people are expected to develop PTSD.

Group 4: Civilians under intense exposure to rocket attacks (living up to 40 km from the Gaza Strip)

We relied on a review by Greene et al (2018), which examined PTSD rates in residents of the south of Israel that were under exposure to continuous rocket attacks. This review only included questionnaire-based assessments, and therefore all estimates included in the review were multiplied by the correction factor (of 0.71).

Omission of studies

Twelve of the studies presented in the review included PTSD prevalence estimates. Out of these studies, we omitted studies for the following reasons: two studies examined PTSD within the timeframe of war exposure; one study did not specify whether measurement was conducted during or after war exposure; and four studies did not clarify whether they assessed symptoms that related specifically to rocket attacks.

Meta-analysis of included studies

Five studies (and six samples) were included in the meta-analysis. The mean prevalence estimate was μ^ = 0.10 (95% CI[0.03-0.17]). The Q-test and I2 statistic indicated that outcomes are heterogeneous and that a meaningful portion of the variance could be explained by true heterogeneity rather than chance (Q(5) = 143.52, p < 0.001, I2 = 97.34%, τ = .09).

Group size and numerical estimate and range

This group amounted to a total of 1,069,011 people (Israel Central Bureau of Statistics, 2023). Groups 1 and 2, which are also located within the 40 km from Gaza, were subtracted from this group since they were classified under the higher prevalence groups. After multiplying the estimate and CI by the group size, we predict that 109,249 (a range of 32,917 to 185,582) people are expected to develop PTSD.

Group 6: Indirectly affected communities (living more than 80 km from the Gaza strip)

We identified two literature reviews relevant to this group. The first is the aforementioned review of Paz García-Vera, et al. (2016), which also examined PTSD prevalence in groups that were indirectly exposed to terror attacks (e.g., residents of cities that suffered terrorist attacks). We also relied on the review by Greene et al., (2018), specifically on studies that included comparison groups to the rocket exposure groups. The comparison groups included communities and cities within Israel that were exposed to low or no rocket attacks and, thus qualifying as indirectly affected.

Omission of studies

In the review by Paz García-Vera (2016), we omitted studies for the following reasons: six studies did not distinguish between people who were directly affected by the terror attack and those who were not; one study included help-seeking population; one study included a community not affected by the terror attack; and one study was conducted within a period of marked threat of terrorism. This resulted in only one study from this review being included in the meta-analysis. In the review by Green et al., (2018), 3 studies included a comparison group (i.e., low to no exposure to rocket attacks). All other studies did not include data of comparison groups and therefore were not considered for the current meta-analysis.

Meta-analysis of included studies

Four studies (with 5 samples) were included in the meta-analysis. The mean prevalence estimate was μ^ = 0.02 (95% CI [0.01-0.04]). The Q-test and I2 statistic indicated that outcomes are heterogeneous and that a meaningful portion of the variance could be explained by true heterogeneity rather than chance (Q(5) = 143.52, p < 0.001, I2 = 97.34%, τ = .09).

Group size and numerical estimate and range

As the current population of the State of Israel is 9,767,718 (Israel Central Bureau of Statistics, 2023), this group was calculated to be 3,433,286 (after subtraction of all other groups). After multiplying the estimate and CI by the group size, we predict that 70,714 (a range 21,410 to 120,019) people are expected to develop PTSD.

Prevalence estimates which were not based on reviews and meta-analyses

Group 2: Close proximity to the October 7th terror attacks

As mentioned, for two of the exposure groups we were not able to identify a relevant review. One of these groups was group 2 (close proximity to the October 7th terror attacks), which was under severe threat of invasion by Hamas perpetrators. We searched for reviews specifically addressing a population that was not directly subject to a terror attack but was under perceived life threat due to close proximity to a terror event. We did not find any review that reflected such exposure. One relevant study was identified from the review by Paz García-Vera, et al., (2016) that was utilized for other groups. This study, conducted by North, et al., (2011), examined PTSD prevalence in direct survivors of terror attacks, as well as in employees of New York City companies affected by the events of September 11. The latter group was in close proximity to the terror attacks and experienced perceived life threat but did not personally experience the acts of terror. The prevalence estimate of this group was 0.21 (0.15 after adjustment with the correction factor). However, the size of this specific exposure group within the study was very small (N=19). Including only one study with such a small sample size can result in a biased estimate. Therefore, we chose to not rely on the estimate derived from this study. In the absence of other relevant studies, we used the estimates of group 4 (intensive rocket exposure), as civilians who were under threat of invasion by Hamas perpetrators were also under intense rocket attacks. We therefore relied on the estimate and confidence interval of group 4 (μ^= 0.10 (95% CI [0.03-0.17]).

Group size and numerical estimate and range

This group amounted to a total of 121,061 people (Israel Central Bureau of Statistics, 2023; after subtraction of group 1). After multiplying the aforementioned estimate and CI by the group size, we predict that 12,372 people (with a range of 3,728 to 21,016) are expected to develop PTSD.

Group 5: Civilians under moderate exposure to rocket attacks (living 40-80 km from the Gaza Strip)

A second exposure group for which we could not identify a review was group 5 (moderate exposure to rocket attacks). The majority of studies regarding the exposure of a civilian population to rocket attacks was conducted in Israel, and included in the review by Greene et al., (2018) presented for group 4 (under intensive rocket attacks). However, this group shares similarities with both the intensive rocket exposure group, and the indirectly affected communities, that are characterized by low to no rocket exposure. Therefore, we addressed this group as representing an intermediate level of exposure between groups 4 and 6. Accordingly, to create the estimate for this group, we averaged the prevalence rates of groups 4 and 6. To create the range, we averaged the lower and upper bounds of the confidence intervals of groups 4 and 6. The computed estimate for this group was 0.06, and the range was 0.02 to 0.10.

Group size and numerical estimate and range

This group amounted to a total of 4,960,469 people (Israel Central Bureau of Statistics, 2023). After multiplying the estimate and the range by the group size, we predict that 304,556 (a range of 91,838 to 517,275) people are expected to develop PTSD.

DISCUSSION

We developed a prediction model of PTSD prevalence following the October 7th mass terror attack and the Israel-Hamas war for the varying levels of exposure across the Israeli population. Relying on meta-analyses and review papers, prevalence estimates ranged between 0.02 for the general population, defined as indirectly exposed to the events of the war, to 0.31 for those directly exposed to the October 7th terror atrocities and the ensuant war. Considering the size of each exposure group, the model predicts that approximately 520,278 people (a range of 159,727 to 880,833) which constitute 5.3% of the Israeli population (a range of 1.6% to 9%), are expected to develop PTSD in the aftermath of the traumatic events.

A recent meta-analysis of PTSD prevalence in populations exposed to war and violent conflict, that included both clinical assessments and questionnaire-based assessments, indicated 23.5% prevalence estimate in civilian populations (Lim, et al., 2022). Another meta-analysis of adult civilian war survivors who live in war-afflicted regions indicated a 24% prevalence estimate (Hoppen, et al., 2019). Our results present considerably smaller prevalence estimates compared to these studies. Importantly, studies regarding the mental health of the Israeli population following the current events have already indicated a high prevalence rate of PTSD and other post-traumatic sequelae of traumatic exposure. For example, a longitudinal survey in a representative sample (N= 710) of the adult population in Israel (aged 18-85), of whom 30 participants (4.3%) were directly exposed to the terror attacks, indicated an increase in the probable rate of PTSD, from 16% in August 2023 to 30% in November 2023, one month after the terror attacks (Levi-Belz, 2024).

While this review focused on the Israeli population, it is imperative to note that the implications for Palestinians living in Gaza have been widespread and severe, affecting both those directly involved and those uninvolved, with mass casualties and displacement of the civilian population. It is important for future research to also estimate the prevalence of PTSD among Palestinian civilians in Gaza, which requires access to relevant sources of information.

Limitations

While this model focused on predicting PTSD, other forms of psychopathology are also highly prevalent after exposure to war and terror, particularly depression and anxiety disorders (e.g., Hoppen, et al., 2021). Moreover, war exposure is associated with additional stressors that do not include direct exposure to violence, such as financial, occupational, and interpersonal challenges. These are also expected to be associated with the extent of distress and mental health outcomes (even if not specifically to PTSD).

In addition, given the ecological, wide-population approach, we could not address specific personal and sub-population risk and protective factors that were associated with the development of PTSD following exposure to terror attacks and war atrocities, among others: prior psychopathology alcohol/substance abuse disorders, gender, socio-economic and marital status (Shalev, et al., 2019; Sayed, et al., 2015). Nonetheless, the vast majority of research presented in the reviews included population-based samples, which are considered to be representative of the general population and do not exclude individuals who have prior risk factors. Therefore, we assume that this issue does not substantially compromise the validity of our findings.

Similarly, some specific subgroups were not included in this model (e.g., aid teams and individuals who have a relative who was killed, injured, or abducted during the war). Studies have shown that these types of exposure are associated with an increased risk for PTSD (Chiu et al, 2011; Cukor, et al., 2011, for aid teams; North, et al., 2011; Miguel-Tobal, et al., 2005, for relatives of victims). However, it was unfeasible to assess group size and estimate the unique prevalence rates of such forms of exposure; therefore, they were not included in the final model.

This study was conducted while the Israel-Hamas war was still going on. Although we reviewed research that examined PTSD prevalence in the aftermath of terror attacks and war exposures, large groups of the Israeli population are still experiencing traumatic events, and the number of casualties and the death toll is rising and undetermined. We relied on the most validated information within our reach, with a preference for government-originated information, yet true group sizes might change to some degree.

Although we attempted to identify relevant reviews for each exposure group in the most nuanced manner possible, the singular characteristics of the October 7th terror attack and the Israel-Hamas might compromise our ability to rely on and extrapolate from previous research for the purpose of predicting PTSD in the current situation. Hoffman’s et al., (2011) study regarding the unique elements of ongoing trauma is particularly relevant to war and terror exposure in the Israeli population. Israeli citizens have been exposed to various forms of terror and war ever since the establishment of the state. Many of the studies presented in this review did not address such unique long-term exposure and have referred to traumatic events that include a specific, often single-time terror exposure (e.g., Paz Garcia-Vera, et al., 2015).

Along with this notion, we have several reasons to assume that this model can be considered conservative with regards to its prediction of PTSD prevalence. Firstly, we did not account for some risk factors that do not qualify as direct exposure to terror. One such risk factor is the intense exposure to viral media coverage of the atrocities, which included videos and testimonies of extreme violence, torture, and sexual assault. Indirect exposure to trauma through the media has been shown to relate to post-traumatic symptoms (e.g., Neria & Sulliven, 2011). In the absence of specific data regarding the extent of media exposure across the population, and due to the challenge of assessing the effects of such exposure on PTSD prevalence, this factor was not accounted for in the literature review process. Nonetheless, media exposure is considered to be widespread across all exposure groups, and there is reason to expect that prevalence rates might be higher due to this form of exposure.

Another limitation of our model is that we did not account for a possible additive risk for PTSD due to more than one form of exposure. As mentioned, in cases of multifaceted exposure, we chose the highest prevalence rate. This decision does not account for a possible additive component of several types of terror exposure. Additionally, for group 2 (close proximity to the October 7th attack), we relied on less severe exposure (rocket attacks) compared to the actual exposure.

Importantly, approximately 200,000 Israelis were evacuated from their homes (the majority of them from the southern area, and others from the communities close to the Israel-Lebanon border; Peleg & Gendelman, 2023). We did not address displacement as exposure to terror since some of the displacements occurred for preventive purposes. However, internal displacement has been shown to relate to detrimental mental health outcomes (Husain, et al., 2011). Moreover, many of those displaced belong to exposure groups 1, 2, and 4, which were all subject to direct terror or rocket attacks. Such a dramatic and adverse event such as displacement might result in an increased risk for PTSD.

We presented several reasons to assume our model might underestimate the prevalence of PTSD, yet some of the methodological decisions that we made might result in overestimation of prevalence rates. Specifically, in longitudinal studies that included more than one prevalence estimate we chose the highest prevalence estimate. This is due to the fact that we wanted to capture the extent to which individuals within each group are expected to develop PTSD, without addressing a specific time frame. As most of the studies presented in the review showed a general trend of decline in PTSD prevalence in longitudinal examinations, this might result in an overestimation of the PTSD prevalence, and not account for some of the natural processes of decline in the magnitude of post-traumatic symptoms. It is also important to note that the high heterogeneity in the timing of PTSD measurement within the reviewed studies compromises the precision of the model in terms of time-specific prediction and does not allow for predicting PTSD prevalence at a specific time period after the war exposure.

There are two additional limitations that might not over or underestimate the prevalence estimates but are important to mention. The first is the considerable heterogeneity within exposure groups, particularly in the exposure to rockets groups. Within the same exposure groups, some civilians were much more exposed to rocket attacks than others, even when distances between their areas of residency were small. The second is the fact that we did not differentiate between children and adults, due to the ambiguity of findings regarding difference between children and adults in PTSD prevalence after exposure to trauma.

Another limitation that is not expected to significantly change prevalence estimates but has important clinical implications is the fact that subgroups of exposure groups 1 (civilian direct victims) and 3 (soldiers) were subject to extremely severe traumas, including abduction, torture, or sexual assault. These subgroups are small in magnitude compared to the groups in which they are included. However, these subgroups are at much higher risk for PTSD than the estimates of their respective groups. For example, the review by Johnson and Thompson (2008) regarding torture survivors presented several studies reporting more than 50% PTSD prevalence.

Clinical Implications

Given the prediction for hundreds of thousands of people suffering from PTSD in Israel, stepped, multi-layered, monitored, and cost-effective mental health services should be planned to reach the maximum number of patients suffering from PTSD or other mental disorders. Although most of the exposed population will recover from PTSD without need for psychological or psychiatric intervention (Rosellini, et al., 2018), immediate and comprehensive interventions are required to prevent chronic and severe morbidity for the wide population at risk. This model can help policymakers, philanthropists, researchers, and clinicians in planning future steps and preparing adequate short and long-term responses for the mental health needs of the population exposed to mass trauma. Additionally, underpinning of the model can be utilized, with relevant, necessary modifications, in other occasions of mass-trauma exposure.

Data Availability

No original data was included in this study. All articles included in meta-analyses are addressed in the manuscript and supplementary material.

ACKNOWLEDGMENTS

The authors wish to thank Ariel Knafo-Noam and Yaakov Greenwald for their helpful comments regarding the manuscript, Gidi Kadosh for his help in conceptualization of the project, Chen Pizanti for her assistance in collection of data, and Effective Altruism Israel for the support of the project.

REFRENCES

  1. ↵
    Awad, A. and Lubell, M. (2023, October 7th). Gunmen in homes, captives abducted leave Israelis in shock. Reuters. https://www.reuters.com/world/middle-east/gunmen-homes-captives-abducted-gaza-leave-israelis-shock-2023-10-07/
  2. ↵
    Almog, T. (2024, January 1st). Victims of the Nova Party demand hundreds of millions from the state. Kan.
    Pat-Horenczyk, R., Cohen, S., Ziv, Y., Achituv, M., Asulin-Peretz, L., Blanchard, T. R., … & Brom, D. (2015). Emotion regulation in mothers and young children faced with trauma. Infant mental health journal, 36(3), 337–348.
    OpenUrl
  3. Ben Gigi, N. (2024, January 10th). 100 Days of war on terr–r - The data behind the activity in Judea and Samaria revealed. Israel Defense Forces. https://www.idf.il/171211
  4. Besser, A., & Neria, Y. (2009). PTSD symptoms, satisfaction with life, and prejudicial attitudes toward the adversary among Israeli civilians exposed to ongoing missile attacks. Journal of Traumatic Stress: Official Publication of The International Society for Traumatic Stress Studies, 22(4), 268–275.
    OpenUrl
  5. ↵
    Besser, A., Weinberg, M., Zeigler-Hill, V., & Neria, Y. (2014). Acute symptoms of posttraumatic stress and dissociative experiences among female Israeli civilians exposed to war: The roles of intrapersonal and interpersonal sources of resilience. Journal of Clinical Psychology, 70, 1227–1239. doi:10.1002/jclp.22083
    OpenUrlCrossRef
  6. ↵
    Besser, A., Zeigler-Hill, V., Pincus, A. L., & Neria, Y. (2013). Pathological narcissism and acute anxiety symptoms after trauma: A study of Israeli civilians exposed to war. Psychiatry: Interpersonal and Biological Processes, 76, 381–397. doi:10.1521/psyc.2013.76.4.381
    OpenUrlCrossRef
  7. ↵
    Besser, A., Zeigler-Hill, V., Weinberg, M., Pincus, A., & Neria, Y. (2015). Intrapersonal resilience moderates the association between exposure severity and PTSD symptoms among civilians exposed to the 2014 Israel–Gaza conflict. Self and Identity, 14, 1–15. doi:10.1080/15298868.2014.966143
    OpenUrlCrossRef
  8. ↵
    Cukor, J., Wyka, K., Jayasinghe, N., Weathers, F., Giosan, C., Leck, P., Roberts, J., Spielman, L., Crane, M. & Difede, J. (2011). Prevalence and predictors of posttraumatic stress symptoms in utility workers deployed to the World Trade Center following the attacks of September 11, 2001. Depression and anxiety, 28(3), 210–217.
    OpenUrlCrossRefPubMed
  9. Data N12 (2023). All the settlements, all the stories: this is what the days of fighting in the Gaza Envelope looked like. Mako. https://www.mako.co.il/news-specials/data_n12/Article-3458498d1cd3b81027.htm
  10. ↵
    Devi, S. (2023). Israeli and Palestinian doctors speak out. The Lancet, 402(10411), 1406–1407.
    OpenUrl
  11. Division. (2024, January 2). In HaMichlol. https://www.hamichlol.org.il/%D7%90%D7%95%D7%92%D7%93%D7%94
  12. ↵
    Fox, M.A., Kolitz, T. 2023. Victims of gender-based violence on Oct 7 must be given a voice. Lancet. November. DOI:10.1016/s0140-6736(23)02654-5
    OpenUrlCrossRef
  13. Engelhard, I. M., Van Den Hout, M. A., Weerts, J. O. S., Arntz, A., Hox, J. J., & McNally, R. J. (2007). Deployment-related stress and trauma in Dutch soldiers returning from Iraq: Prospective study. The British Journal of Psychiatry, 191(2), 140–145.
    OpenUrlAbstract/FREE Full Text
  14. ↵
    Fairbank, J. A., & Fairbank, D. W. (2009). Epidemiology of child traumatic stress. Current psychiatry reports, 11(4), 289–295.
    OpenUrlCrossRefPubMedWeb of Science
  15. Galea, S., Vlahov, D., Resnick, H., Ahern, J., Susser, E., Gold, J., Bucuvalas, M. & Kilpatrick, D. (2003). Trends of probable post-traumatic stress disorder in New York City after the September 11 terrorist attacks. American journal of epidemiology, 158(6), 514–524.
    OpenUrlCrossRefPubMedWeb of Science
  16. ↵
    Cortellessa, E. (2023, October 31th). The Oct. 7 Massacre revealed a new Hamas social media strategy. Time. https://time.com/6330005/the-oct-7-massacre-revealed-a-new-hamas-social-media-strategy/
  17. ↵
    Gavrielov, N. (2023, Dec 27). Hamas and other militant groups are firing rockets into Israel every day. The New York Times. https://www.nytimes.com/2023/12/27/world/middleeast/israel-hamas-gaza-rockets.html?smid=url-share
  18. Gelkopf, M., Berger, R., Bleich, A., & Silver, R. C. (2012). Protective factors and predictors of vulnerability to chronic stress: A comparative study of 4 communities after 7 years of continuous rocket fire. Social science & medicine, 74(5), 757–766.
    OpenUrl
  19. ↵
    Gil, S., Weinberg, M., Or-Chen, K., & Harel, H. (2015). Risk factors for DSM–5 PTSD symptoms in Israeli civilians during the Gaza war. Brain and Behavior, 5, e00316. doi:10.1002/brb3.316
    OpenUrlCrossRef
  20. ↵
    Gil, S., Weinberg, M., Shamai, M., Ron, P., Harel, H., & Or-Chen, K. (2016). Risk factors for DSM–5 posttraumatic stress symptoms (PTSS) among Israeli civilians during the 2014 Israel-Hamas war. Psychological Trauma: Theory, Research, Practice, and Policy, 8, 49
    OpenUrl
  21. ↵
    Greene, T., Itzhaky, L., Bronstein, I., & Solomon, Z. (2018). Psychopathology, risk, and resilience under exposure to continuous traumatic stress: A systematic review of studies among adults living in southern Israel. Traumatology, 24(2), 83.
    OpenUrl
  22. ↵
    Hoffman, Y. S., Diamond, G. M., & Lipsitz, J. D. (2011). The challenge of estimating PTSD prevalence in the context of ongoing trauma: The example of Israel during the Second Intifada. Journal of Anxiety Disorders, 25(6), 788–793
    OpenUrlCrossRefPubMed
  23. Hopwood, T. L., & Schutte, N. S. (2017). Psychological outcomes in reaction to media exposure to disasters and large-scale violence: A meta-analysis. Psychology of violence, 7(2), 316.
    OpenUrl
  24. ↵
    Hoppen, T. H., Priebe, S., Vetter, I., & Morina, N. (2021). Global burden of post-traumatic stress disorder and major depression in countries affected by war between 1989 and 2019: a systematic review and meta-analysis. BMJ global health, 6(7), e006303.
    OpenUrlAbstract/FREE Full Text
  25. ↵
    Husain, F., Anderson, M., Cardozo, B. L., Becknell, K., Blanton, C., Araki, D., & Vithana, E. K. (2011). Prevalence of war-related mental health conditions and association with displacement status in postwar Jaffna District, Sri Lanka. Jama, 306(5), 522–531.
    OpenUrlCrossRefPubMedWeb of Science
  26. Israel Central Bureau of Statistics (2024, January 31th). Population. https://www.cbs.gov.il/he/publications/LochutTlushim/2020/%D7%90%D7%95%D7%9B%D7%9C%D7%95%D7%A1%D7%99%D7%99%D7%942020.xlsx
  27. ↵
    Israel Central Bureau of Statistics (2022). Regional statistics. https://www.cbs.gov.il/he/publications/doclib/2019/ishuvim/bycode2022.xlsx
  28. ↵
    Johnson, H., & Thompson, A. (2008). The development and maintenance of post-traumatic stress disorder (PTSD) in civilian adult survivors of war trauma and torture: A review. Clinical psychology review, 28(1), 36–47.
    OpenUrlCrossRefPubMedWeb of Science
  29. Jones, M., Rona, R. J., Hooper, R., & Wesseley, S. (2006). The burden of psychological symptoms in UK Armed Forces. Occupational medicine, 56(5), 322–328.
    OpenUrlCrossRefPubMed
  30. ↵
    Levi-Belz, Y., Groweiss, Y., Blank, C., & Neria, Y. (2024). PTSD, depression, and anxiety after the October 7, 2023 attack in Israel: a nationwide prospective study. eClinicalMedicine.
  31. ↵
    May, C. L., & Wisco, B. E. (2016). Defining trauma: How level of exposure and proximity affect risk for posttraumatic stress disorder. Psychological trauma: theory, research, practice, and policy, 8(2), 233
    OpenUrl
  32. ↵
    McNally, R. J., Bryant, R. A., & Ehlers, A. (2003). Does early psychological intervention promote recovery from posttraumatic stress?. Psychological science in the public interest, 4(2), 45–79.
    OpenUrlCrossRefPubMed
  33. ↵
    Miguel-Tobal, J. J., Cano-Vindel, A., Iruarrizaga, I., González-Ordi, H., & Galea, S. (2005). Psychopathological repercussions of the March 11 terrorist attacks in Madrid. Psychology in Spain, 9, 75–80.
    OpenUrl
  34. Neria, Y., Besser, A., Kiper, D., & Westphal, M. (2010). A longitudinal study of posttraumatic stress disorder, depression, and generalized anxiety disorder in Israeli civilians exposed to war trauma. Journal of traumatic stress, 23(3), 322–330
    OpenUrlCrossRefPubMedWeb of Science
  35. ↵
    Neria, Y., DiGrande, L., & Adams, B. G. (2011). Posttraumatic stress disorder following the September 11, 2001, terrorist attacks: A review of the literature among highly exposed populations. American Psychologist, 66(6), 429–446. doi:10.1037/a0024791
    OpenUrlCrossRefPubMedWeb of Science
  36. ↵
    Neria, Y., Gross, R., Olfson, M., Gameroff, M. J., Wickramaratne, P., Das, A., … & Weissman, M. M. (2006). Posttraumatic stress disorder in primary care one year after the 9/11 attacks. General hospital psychiatry, 28(3), 213–222.
    OpenUrlCrossRefPubMedWeb of Science
  37. ↵
    Neria, Y., Solomon, Z., & Dekel, R. (1998). An eighteen-year follow-up study of Israeli prisoners of war and combat veterans. The Journal of Nervous and Mental Disease, 186(3), 174–182.
    OpenUrlCrossRefPubMed
  38. ↵
    Neria, Y., & Sullivan, G. M. (2011). Understanding the mental health effects of indirect exposure to mass trauma through the media. Jama, 306(12), 1374–1375.
    OpenUrlCrossRefPubMedWeb of Science
  39. ↵
    North, C. S., Pollio, D. E., Smith, R. P., King, R. V., Pandya, A., Suris, A. M., Hong, B. A., Dean, D. J., Wallace, N. E., Herman, D. B., Conover, S., Susser, E. & Pfefferbaum, B. (2011). Trauma exposure and posttraumatic stress disorder among employees of New York City companies affected by the September 11, 2001 attacks on the World Trade Center. Disaster medicine and public health preparedness, 5(S2), S205–S213
    OpenUrlCrossRefPubMed
  40. Palmieri, P. A., Chipman, K. J., Canetti, D., Johnson, R. J., & Hobfoll, S. E. (2010). Prevalence and correlates of sleep problems in adult Israeli Jews exposed to actual or threatened terrorist or rocket attacks. Journal of Clinical Sleep Medicine, 6(6), 557–564.
    OpenUrl
  41. ↵
    Paz García-Vera, M., Sanz, J., & Gutierrez, S. (2016). A systematic review of the literature on posttraumatic stress disorder in victims of terrorist attacks. Psychological reports, 119(1), 328–359.
    OpenUrlCrossRefPubMed
  42. ↵
    Peleg, O., & Gendelman, L. (2023). Internally displaced people amidst war: the Israeli narrative. The Lancet, 402(10417), 2071–2072.
    OpenUrl
  43. ↵
    Peled, A. and Jones, R. (2023, Dec 31). Israel’s ‘Black Sabbath’: Murder, sexual violence and torture on Oct. 7. The Wall Street Journal. https://www.wsj.com/world/middle-east/israel-hamas-oct-7-murder-sexual-violence-torture-45aab439
  44. ↵
    Pityaratstian, N., Piyasil, V., Ketumarn, P., Sitdhiraksa, N., Ularntinon, S., & Pariwatcharakul, P. (2015). Randomized controlled trial of group cognitive behavioural therapy for post-traumatic stress disorder in children and adolescents exposed to tsunami in Thailand. Behavioural and cognitive psychotherapy, 43(5), 549–561
    OpenUrl
  45. ↵
    Richardson, L. K., Frueh, B. C., & Acierno, R. (2010). Prevalence estimates of combat-related post-traumatic stress disorder: critical review. Australian & New Zealand Journal of Psychiatry, 44(1), 4–19.
    OpenUrl
  46. ↵
    Saidel, P., Said, S. and Peled, A. (2024, Jan 24). Hamas took more than 200 hostages grom Israel. Here’s what we know. The Wall Street Journal. https://www.wsj.com/world/middle-east/hamas-hostages-israel-gaza-41432124
  47. ↵
    Sayed, S., Iacoviello, B. M., & Charney, D. S. (2015). Risk factors for the development of psychopathology following trauma. Current psychiatry reports, 17, 1–7.
    OpenUrlCrossRefPubMed
  48. ↵
    Shalev, A. Y., Gevonden, M., Ratanatharathorn, A., Laska, E., Van Der Mei, W. F., Qi, W., Lowe, S., Lai, B. S., Bryant, R. A., Delahanty, D., Matsuoka, Y. J., Olff., M., Schnyder, U., Seedat, S., deRoon-Cassini, T. A., Kessler, R. C., Koenen, K. C. & International Consortium to Predict PTSD (2019). Estimating the risk of PTSD in recent trauma survivors: results of the International Consortium to Predict PTSD (ICPP). World Psychiatry, 18(1), 77–87.
    OpenUrl
  49. ↵
    Shaw, J. A. (2003). Children exposed to war/terrorism. Clinical child and family psychology review, 6, 237–246.
    OpenUrlCrossRefPubMedWeb of Science
  50. Smith, T. C., Ryan, M. A., Wingard, D. L., Slymen, D. J., Sallis, J. F., & Kritz-Silverstein, D. (2008). New onset and persistent symptoms of post-traumatic stress disorder self reported after deployment and combat exposures: prospective population based US military cohort study. Bmj, 336(7640), 366–371.
    OpenUrlAbstract/FREE Full Text
  51. ↵
    Statista (2023). Deadliest terrorist attacks worldwide from 1970 to 2023, by number of fatalities and perpetrator. https://www.statista.com/statistics/1330395/deadliest-terrorist-attacks-worldwide-fatalities/
  52. Steel, Z., Chey, T., Silove, D., Marnane, C., Bryant, R. A., & Van Ommeren, M. (2009). Association of torture and other potentially traumatic events with mental health outcomes among populations exposed to mass conflict and displacement: a systematic review and meta-analysis. Jama, 302(5), 537–549.
    OpenUrlCrossRefPubMedWeb of Science
  53. ↵
    Sundin, J., Fear, N. T., Iversen, A., Rona, R. J., & Wessely, S. (2010). PTSD after deployment to Iraq: conflicting rates, conflicting claims. Psychological medicine, 40(3), 367–382
    OpenUrlCrossRefPubMedWeb of Science
  54. ↵
    Tang, B., Deng, Q., Glik, D., Dong, J., & Zhang, L. (2017). A meta-analysis of risk factors for post-traumatic stress disorder (PTSD) in adults and children after earthquakes. International journal of environmental research and public health, 14(12), 1537.
    OpenUrl
  55. The American Psychiatric Association’s (2022) Diagnostic and Statistical Manual of Mental Disorders (5th ed., text rev.; DSM-5-TR) is a widely used diagnostic manual in the United States.
  56. ↵
    Turner, S. W., Bowie, C., Dunn, G., Shapo, L., & Yule, W. (2003). Mental health of Kosovan Albanian refugees in the UK. The British Journal of Psychiatry, 182(5), 444–448
    OpenUrlAbstract/FREE Full Text
  57. ↵
    Tzalah, Z. (2023, October 10th). 100 People participated in ‘Psyduck’, a small party between Nirim and Nir Oz. 10 of them are missing. Mako. https://www.mako.co.il/music-news/Article-551663c5e1a1b81026.htm
  58. ↵
    Viechtbauer, W. (2010). Conducting meta-analyses in R with the metafor package. Journal of statistical software, 36, 1–48.
    OpenUrl
  59. ↵
    Weinberg, M., Besser, A., Zeigler-Hill, V., & Neria, Y. (2015). Dispositional optimism and self– esteem as competing predictors of acute symptoms of generalized anxiety disorders and dissociative experiences among civilians exposed to war trauma. Psychological Trauma: Theory, Research, Practice, and Policy, 7, 34.
    OpenUrl
  60. World Health Organization (2021). International statistical classification of diseases and related health problems (11th ed.). https://icd.who.int/
  61. ↵
    Xue, C., Ge, Y., Tang, B., Liu, Y., Kang, P., Wang, M., & Zhang, L. (2015). A meta-analysis of risk factors for combat-related PTSD among military personnel and veterans. PloS one, 10(3), e0120270.
    OpenUrlCrossRefPubMed
  62. ↵
    Zeitoon, Y. (2023, October 12th). Gaza brigade commander in first statement: ‘We stopped the attack. We did everything to fight’. Ynet. https://www.ynet.co.il/news/article/bjjhjssza
Back to top
Posted February 27, 2024.
Download PDF

Supplementary Material

Data/Code
Email
A Prediction Model of PTSD in the Israeli Population in the Aftermath of October 7th, 2023, Terrorist Attack and the Israel-Hamas War
Dana Katsoty, Michal Greidinger, Yuval Neria, Aviv Segev, Ido Lurie
medRxiv 2024.02.25.24303235; doi: https://doi.org/10.1101/2024.02.25.24303235
Twitter logo Facebook logo LinkedIn logo Mendeley logo
Citation Tools

Subject Area