Content uploaded by Charline Vincent
Author content
All content in this area was uploaded by Charline Vincent on Nov 15, 2023
Content may be subject to copyright.
HAL Id: hal-04088838
https://hal.science/hal-04088838
Submitted on 4 May 2023
HAL is a multi-disciplinary open access
archive for the deposit and dissemination of sci-
entic research documents, whether they are pub-
lished or not. The documents may come from
teaching and research institutions in France or
abroad, or from public or private research centers.
L’archive ouverte pluridisciplinaire HAL, est
destinée au dépôt et à la diusion de documents
scientiques de niveau recherche, publiés ou non,
émanant des établissements d’enseignement et de
recherche français ou étrangers, des laboratoires
publics ou privés.
Distributed under a Creative Commons Attribution| 4.0 International License
Post-traumatic stress disorder among civilians 6 and 18
months after the January 2015 terrorist attacks in the
Paris region
Charline Vincent, Philippe Pirard, Yvon Motre, Leticia Bertuzzi, Stéphanie
Vandentorren, Cécile Vuillermoz
To cite this version:
Charline Vincent, Philippe Pirard, Yvon Motre, Leticia Bertuzzi, Stéphanie Vandentorren, et al..
Post-traumatic stress disorder among civilians 6 and 18 months after the January 2015 terrorist attacks
in the Paris region. Psychiatry Research, 2023, 322, pp.115137. �10.1016/j.psychres.2023.115137�.
�hal-04088838�
PTSD 6 and 18 months after terrorist attacks
Post-traumatic stress disorder among civilians 6 and 18 months after the January 2015
terrorist attacks in the Paris region
Vincent Charline*a, Philippe Pirardb, Yvon Motreffb, Leticia Bertuzzia, Vandentorren
Stéphanie*c,d, Vuillermoz Cécile*a
a: INSERM, Sorbonne Université, Institut Pierre Louis d’Epidémiologie et de Santé Publique
(IPLESP), Department of social epidemiology, F75012 Paris, France
b: Santé publique France, Direction des maladies non transmissibles et traumatismes, F94415
Saint-Maurice, France
c: Santé publique France, Direction scientifique et internationale, F94415 Saint-Maurice,
France
d: University of Bordeaux, INSERM, Bordeaux Population Health Research Center, U1219, F-
33000 Bordeaux, France.
*: co-last authors
Information about the author
Charline VINCENT
Email address: charlinevincent4@gmail.com
Address: INSERM, Sorbonne University - Faculty of Medicine
Pierre Louis Institute of Epidemiology and Public Health
2
Saint-Antoine site - UMR-S 1136
27 rue Chaligny, 75012 PARIS
Phone number: +33 (0)7 68 94 94 20
Corresponding Authors
Philippe Pirard: philippe.pirard@santepubliquefrance.fr
Yvon Motreff: yvon.motreff@santepubliquefrance.fr
Leticia Bertuzzi: leticia.bertuzzi@iplesp.upmc.fr
Vandentorren Stéphanie: stephanie.vandentorren@santepubliquefrance.fr
Vuillermoz Cécile: cecile.vuillermoz@inserm.fr
3
Post-traumatic stress disorder among civilians 6 and 18 months after the
January 2015 terrorist attacks in the Paris region
Abstract
In the literature, the association between medium- and long-term PTSD (post-traumatic stress
disorder) after terrorist attack has rarely been described. The objective of our study was to
identify the factors associated with PTSD in the medium and longer term among people exposed
to a terrorist attack in France.
We used data from a longitudinal survey of 123 terror-exposed people interviewed 6-10
(medium term) and 18-22 (long term) months after. Mental health was assessed by the Mini
Neuropsychiatric Interview.
PTSD in the medium term was associated with history of traumatic events, low levels of social
support and severe peri-traumatic reactions, which were in turn associated with high levels of
terror exposure. PTSD in the medium term was linked in turn to the presence of anxiety and
depressive disorders, which was also linked to PTSD in the longer term.
The factors leading to PTSD are different in the medium and long term. In order to improve
future support for people exposed to distressing events, it is important to follow up people with
intense peri-traumatic reactions, high levels of anxiety and depression and to measure reactions.
Keywords: post-traumatic stress disorder, January 2015 terrorist attacks in France, exposure,
longitudinal study, structural equation model.
4
1. INTRODUCTION
The terrorist attacks perpetrated on January 7, 8 and 9, 2015 in the Paris region caused the death
of 17 people and physically injured 20. The literature has shown that an exposed population
can suffer from PTSD, anxiety disorders and depression (Bonanno et al., 2010). For intentional
traumatic events, the prevalence of PTSD is thought to decrease more slowly than for
unintentional events (Santiago et al., 2013). According to a 2013 literature review of 35 studies
on trauma survivors, the prevalence of PTSD ranged from 28.8% at 1 month to 17.0% at 12
months (Santiago et al., 2013).
In studies on terrorist attacks, authors have suggested that the type of exposure (being a witness,
being close to the perpetrators, or being injured) is a factor that could be associated with the
occurrence of PTSD (Brewin et al., 2000; Glad et al., 2016; Ozer et al., 2003). Several authors
have observed that the intensity of the experience of the event is associated with the occurrence
of PTSD and it depends in particular on the peri-traumatic reactions, mostly in the form of a
disconnection from reality during the events (Brewin et al., 2000; Rouillon et al., 2001). The
onset of PTSD can also be associated with a history of mental health disorder (Brewin et al.,
2000; Ozer et al., 2003; Perlman et al., 2011), the use of psychotropic drugs in the past and life
difficulties in the year preceding the events (Brewin et al., 2000; Ozer et al., 2003)as well as
the perceived absence of social support after the event (Brewin et al., 2000; Dyb et al., 2014;
Ozer et al., 2003; Silver et al., 2002; Simard, 2018). The aforementioned studies have only
focused on factors associated with a single disorder. However, approximately 80% of people
with PTSD are reported to have concomitant psychiatric disorders (Jolly, 2000). Anxiety
disorders and depression are risk factors for the development of post-traumatic stress disorder
(Adams et al., 2019; Bugge et al., 2017; Vázquez et al., 2006; Vlahov et al., 2002).
5
In the literature, most studies that have been conducted on mental health at the time of terror
attacks are cross-sectional and use simple regression models to identify factors associated with
mental health disorders. However, these designs and models do not enable the complex
relationships between the different disorders to be studied, nor do they enable assessment of the
evolution of mental health over time.
Considering the various aspects mentioned above, the aim of this study was to identify factors
associated with post-traumatic stress disorder among civilians, 6 to 18 months after the January
2015 attacks in France, using structural equation modelling in order to take the complex
relationships between these factors into account and to assess the evolution of mental health
over time.
2. MATERIALS AND METHODS
2.1.Participants and procedure
The study was based on data collected from the IMPACTS survey (Investigation of Post-Attack
Traumatic Manifestations and Therapeutic and Supportive Care), led by Santé Publique France
(the French national public health agency), the Ile-de-France Regional Health Agency and
INSERM (National Institute of Health and Medical Research). The survey population consisted
of people who had been exposed to the January 2015 terrorist attacks perpetrated in the Paris
region. The people exposed to the attacks were defined in the IMPACTS survey as those who
had been exposed according to "criterion A" of the DSM-V definition of post-traumatic stress
disorder. They included the following: physically wounded individuals, hostages, witnesses
directly threatened or physically present at the scene of the events (directly threatened subjects);
people who had to take shelter during the terrorist attacks and whose lives were indirectly
threatened (indirectly threatened subjects); and family members and close relatives of the
6
victims, people living or working in the immediate proximity of the scene of the events and
present at the time of the attacks (indirect witnesses).
The aims of the IMPACTS survey were to study the impact of these terrorist attacks on civilians'
and rescue workers’ mental health and social functioning, as well as to assess perceived social
support and mental health care received in medium and long-term care. The first round of the
investigation was conducted from June to October 2015 (medium term), and the second round
from June to October 2016 (long term). Details of the survey design and sampling process have
been provided in another article (Vandentorren et al., 2018; Vuillermoz et al., 2020). In this
study, we considered the 123 civilians that were interviewed 6-10 (medium term) and 18-22
months (long term) after the attacks.
2.2.Measures
The survey questionnaires were administered face-to-face by investigators trained in trauma
management, and the following were collected: socio-demographic characteristics, the level of
exposure, peri-traumatic reactions, perceived social support received after the events,
psychological support received after the events, history of medical and psychological care,
history of potentially traumatic situations and mental health.
2.3.Structural equation modelling
In the context of the complex relationships between PTSD, anxiety disorders, depression and
factors associated with these disorders, we considered that Structural Equation Modelling
(SEM) models were the best suited to studying these relationships (Bachet, 1999; Mansiaux,
2014).
2.3.1. The hypothesised measurement models
In view of the literature and the data collected from the IMPACTS survey, we made
assumptions that were placed in four categories (Fig 1).
7
Fig 1: Hypothesised model for relationships between latent and observed variables and post-
traumatic stress disorder in the medium and the long term after the attacks, IMPACTS 2015
2016 survey. Circles: latent variables; Boxes: observed variables
Firstly, we hypothesised that PTSD in the medium term was related more to the situation before
the attacks (i.e. psychological and psychiatric history) (Ozer et al., 2003; Perlman et al., 2011)
and especially to a high level of exposure to the attacks (i.e. directly or indirectly threatened)
(Glad et al., 2016; Monfort & Afzali, 2017; Ozer et al., 2003) and to severe peri-traumatic
reactions experienced around the event (Ozer et al., 2003; Rouillon et al., 2001). We also
thought that female gender (Carragher et al., 2016; Olff et al., 2007) and the absence of social
support in the medium term (Motreff et al., 2020) would be associated with PTSD in the
medium term.
Secondly, we hypothesised that PTSD in the longer term would be mainly linked to the presence
of PTSD in the medium term and the presence of comorbidities (depressive episode or anxiety
disorders) between the two rounds of the survey (Adams et al., 2019; Simard, 2018), which
8
would otherwise have led to chronic PTSD or increased the symptoms. Furthermore, we
hypothesised that secondary exposure to the attacks in November 2015 and having been
interviewed in the second round of the survey before the July 2016 attacks in Nice (Amoretti,
2018; Solla et al., 2018) could be linked to the presence of PTSD in the longer term. As for the
medium term, we hypothesised that PTSD in the long term would also be related to female
gender, a history of trauma, psychological et psychiatric history, severe peri-traumatic
reactions, high level of exposure to the attacks and the absence of perceived social support in
the medium and long term.
Thirdly, we hypothesised that long-term co-morbidities would be linked to the presence of
PTSD in the medium term (Bugge et al., 2017; Vázquez et al., 2008), to a high level of exposure
during the attack (Glad et al., 2016; Monfort & Afzali, 2017; Vázquez et al., 2008) to severe
peri-traumatic reactions experienced around the attack (Henriksen et al., 2010), to the absence
of perceived social support in the medium and long term (Fekih-Romdhane et al., 2017; Wood
et al., 2013), to psychological and psychiatric histories, to female gender (Dyb et al., 2014;
Hales et al., 2014; Salguero et al., 2011) and secondary exposure to the attacks in November
2015.As the second round of the IMPACTS survey was conducted from June to October 2016,
people participated in the survey both before and after the attacks in Nice in 2016. We
hypothesised that among people who were interviewed after the Nice 2016 attacks, PTSD
would be more frequent as a result of a potential effect of secondary exposure to trauma.
Fourthly, we made some hypotheses concerning explanatory factors. One of them was that
gender would be linked to peri-traumatic reactions experienced after the event (Elhai et al.,
2006; Hamama-Raz et al., 2015; Maguen et al., 2012) and the presence of psychological and
traumatic history. In fact, according to the literature, women face more violence than men and
therefore have a potentially more traumatic life history (Morin et al., 2013). We also
hypothesised that the level of exposure (being close to the aggression, being directly threatened)
9
could have influenced the perceptions of social support, and that the most exposed people had
the strongest peri-traumatic reactions.
2.3.2. Variables used
Outcomes. The two variables of interest in our study, the presence of PTSD in the medium (6-
10 months) and in the long term (18-22 months), were constructed from the presence of PTSD
in the last month measured in each round (T1 and T2) of the survey using the MINI (Mini-
International Neuropsychiatric Interview v6 DSM IV, (Sheehan et al., 1998)). The MINI is a
structured interview developed jointly by psychiatrists and clinicians in the United States and
Europe, for the diagnosis of 17 psychiatric disorders aministrated in face-to-face by non-
specialsed interviewer, it is a validated gold standard (Spoont et al., 2013).
These variables were binary (1 in presence of PTSD and 0 in absence of PTSD).
Latent variables. All other variables used in this model were ordinal and categorical; they
were all recoded to go in the same direction as a latent variable to reflect the same concept.
Comorbidities were constructed using the two items: (i) the presence of major depressive
episode in the last two weeks in the medium and long term (at least once at T1 or T2/neither
in T1 nor T2) measured with the MINI and (ii) the presence of at least one of these anxiety
disorders: generalised anxiety, panic disorder, agoraphobia and social phobia, in the medium
and long term (at least once at T1 or T2/ neither in T1 nor T2) measured with the MINI
(Pettersson et al., 2018).
A high level of exposure was constructed using three items: (i) a high subjective exposure
score(a score >5 indicated high exposure), (ii) the type of exposure(which was constructed
using 20 questions that defined whether people had been directly threatened, indirectly
threatened or had indirectly witnessed) and (iii) physical proximity with the perpetrators(where
10
the person was during the event: within 10 meters of the terrorists/very close to the event site,
in an adjacent room/in a nearby building or on an adjacent street/elsewhere).We have therefore
considered three types of physical proximity: immediate proximity, distant from the scene of
the event and having lost a loved one.
The variable for severe peri-traumatic reactions around the event was constructed using two
items : (i) acute peri-traumatic stress measured by the STRS score (Shortness of breath,
Tremulousness, Racing heart, and Sweating) (Bracha et al., 2004) which provided information
on the adrenergic reaction (high STRS score>24, we chose the median as the cut-off in the
absence of indications in the literature) and (ii) marked experience of peritraumatic dissociation,
i.e. a high level of exposure felt during the attack (Peritraumatic Dissociative Experience
Questionnaire) (Birmes et al., 2005)(high PDEQ score, >15, we chose a cut-off of 15 on the
basis of the literature (Marmar et al., 1996; Zambaldi et al., 2011)).
The presence of psychological and psychiatric histories in the medium term was constructed
using four items: (i) the presence of psychological support in the past (yes/no), (ii) the presence
of life difficulties experienced in 2014 (yes/no), (iii) the presence of a history of trauma before
the first round of data collection (yes/no) and (iv) taking sleeping pills in the past (yes/no).
The perception of lack of social support at T1 and T2 was constructed from the three perceived
social support items (Melrose et al., 2015; Robert et al., 2017) collected in the medium and long
term after the events: (i) not feeling supported financially (being able to count on someone to
help in case of financial or material need: yes/no), (ii) not feeling supported in daily life (being
able to count on someone to help in daily life : yes/no) and (iii) not feeling supported (being
able to count on someone to provide moral or emotional support: yes/no).
Other indicators. We introduced four other variables into the model that were directly measured
in the survey questionnaire and maintained as solely observed variables because they did not
11
enable the construction of latent variables: (i) gender (male/female), (ii) presence of a history
of trauma since the first round of data collection (yes/no), (iii) presence of exposure to the
November 2015 attacks (yes/no) and (iv) having responded to the questionnaire in the second
round of the survey, that is to say after the attacks in July 2016 (yes/no).
2.4.Data analyses
The characteristics of the population were described and the proportions and medians were
calculated with 95% confidence intervals.
First of all, we tested the measurement model. We estimated the correlations between the
different observed variables for the same latent variable using Spearman's correlation
coefficient (non-parametric). We then considered there was a correlation if ρ> 0.3 (Falissard,
2011). Secondly, we checked the uni-dimensionality of each latent variable using scree-plots.
Thirdly, we estimated the measurement model using CFA (Confirmatory Factor Analysis) to
check that the measured variables contributed significantly to the constructed latent variables.
We used the WLSMV (Weighted Least Squares Means and Variance) estimator adapted to the
use of ordinal categorical variables (Newsom, 2018). Finally, we estimated the structural
equation model with standardised regression coefficients (-1 to 1). Using a descending stepwise
procedure, the least significant relationships were removed one after the other. The goodness-
of-fit of the CFA and structural model was checked using goodness-of-fit indexes for the
categorical variables (Hutchinson & Olmos, 1998) : the Root Mean Square Error of
Approximation (RMSEA) (>0.08), the Comparative Fit Index (CFI) (<0.95) and the Tucker-
Lewis Index (TLI) (<0.95). In order to estimate the statistical power of the model, we used
Preacher’s method (Preacher & Coffman, 2006), which requires the number of degrees of
freedom to be known (Rigdon, 1994) :
, were m is the number of
observed variables and e the number of latent variables.
12
All analyses were conducted using R software (version 4.0.2), with the "lavaan" package for
estimation of SEM.
3. RESULTS
3.1.Population
A total of 190 people responded in T1. The participation rate in T2 was 64.7% with 123
participants. Among the participants in T2, the majority were women (61%) and born in France
(95.9%) (Table 1). The median age was 42 years at T1. Concerning the type of exposure, 32
(26%) people were indirect witnesses, 55 (44.7%) people were indirectly threatened and 36
(29.3%) were directly threatened. Concerning their physical proximity with the perpetrators, 13
(10.6%) people were had lost a loved one, but not at the scene, 55 (44.7%) people were in the
immediate vicinity and 55 (44.7%) people were some distance from the street where the attack
took place. Finally, concerning subjective exposure, 29 (24.2%) people reported low exposure
and 91 (75.8%) high exposure.
In the medium term, 16.3% presented PTSD, 32% had at least one anxiety disorder and 42.3%
had current depression. In the long term, 14.6% had PTSD, 43.1% had anxiety disorders and
30.9% had depressive disorders.
3.2.Validation of latent constructs
The correlations across the observed variables of the latent variables ranged from -0.03 to 0.59.
Some observed variables of the latent variable "Psychological and psychiatric history at T1"
were not correlated with each other, so the latent variable was not retained. For this latent
construct, in the model we retained the observed variable: having experienced at least one
trauma in the medium term after the events. The observed variable "subjective exposure” was
removed from the latent variable "High level of exposure" because it was not correlated with
the other observed variables. For latent variables with more than 3 observed variables, the uni-
dimensionality was validated.
13
3.3.Measurement models
A factor analysis was run with five latent variables in the hypothetical model. The estimated
standardised loadings were statistically significant (p-value<0.001) and their values were
greater than 0.5, which meant a robust relationship between the observed variables and their
latent variable. This model enabled close agreement with the data and the proposed dimensional
structure was therefore not rejected: TLI = 0.936, RMSEA = 0.031[0.000-0.072], CFI = 0.958.
The measurement model thus contained 19 measured variables and 5 latent variables. On the
basis of the Preacher’s method presented in "statistical analyses", we estimated the number of
degrees of freedom:
= 125. The estimated statistical power for
a sample of 123 individuals, an RMSEA between 0.05 and 0.08, a first species risk at 0.05 and
a number of degrees of freedom at 125, was 84% (Sideridis et al., 2014).
3.4.Final model
The final model is shown in Fig 2.PTSD in the medium term (6-10 months) was linked to the
presence of a history of trauma in the medium term (βSD= 0.13 [95% CI: 0.04-0.40]) as well as
to severe peri-traumatic reactions experienced during the attacks (βSD= 0.40, [95% CI: 0.30-
0.74]) and to the absence of perceived social support in the medium term (βSD= 0.42, [95% CI:
0.02-0.56]).
Long-term (18-22 months) PTSD was associated with the presence of comorbidities (βSD=
0.64, [95% CI: 0.31-0.70]). The presence of comorbidities was linked to female gender (βSD=
0.22, [95% CI: 0.17-0.63]), PTSD at 6-10 months (βSD= 0.52, [95% CI: 0.32-0.86]) and lack of
social support in the medium term (βSD= 0.49, [95% CI: 0.16-0.61]). Severe peri-traumatic
reactions were linked to a high level of exposure (βSD= 0.26, [95% CI: 0.07-0.62]). The
adjustment indices of the structural model were satisfactory, the RMSEA was 0.00[0.00-
0.03], the CFI was 1.00, and the TLI was 1.05.
14
Fig 2: Final model of relationships between PTSD and the various latent and observed
variables, in the medium and long term after the attacks, IMPACTS 2015-2016 survey.
Circles: latent variables; Boxes: observed variables
4. DISCUSSION
This study shows that the prevalence of PTSD was 16.3% at 6-10 months and 14.6% at 18-22
months after the attacks. An online survey of a nationally representative panel 1 month after
these events showed a prevalence of PTSD of 7.6% (Ben-Ezra et al., 2015). Among people
exposed to the 1996 tube train attacks in France, 41% had PTSD at 6 months and 34% at 18
15
months (Jehel et al., 1999).Among civilians exposed to the November 2015 Paris terror attacks,
48% presented PTSD at 8-11 months of the events (Pirard et al., 2018).
Following the 11 September 2001 attacks in the USA, 17% of residents and workers enrolled
in the World Trade Center Health Registry presented PTSD 2 to 3 years after the events (Welch
et al., 2016). The prevalence of PTSD in our study was lower than the values found in the
literature, which could be explained by the difference in the diagnostic scales used, since the
prevalence of PTSD is more restrictive when measured using the MINI than when measured
with the DTS (Davidson Trauma Scale
1
) (Vázquez et al., 2006).
We found a relationship between medium-term PTSD and severe peri-traumatic reactions to
the attacks, which is consistent with the literature (Brewin et al., 2000; Ozer et al., 2003).
However, SEM showed that peri-traumatic reactions were linked to a high level of exposure.
Our results also highlight a relationship between a history of trauma and PTSD in the medium
term, as others studies (Delahanty & Nugent, 2006; Gilbar et al., 2020; Scharff et al., 2021).
Finally, our results suggest that the lower was the person's perceived social support, the more
likely were these individuals to develop post-traumatic stress disorder. In the literature, social
support, when perceived as positive by the individual, has a positive influence on his or her
mental health (Dyb et al., 2014; Harandi et al., 2017), as well as helping the person to adapt to
a stressful event (Folkman, 2013).
Our results suggest an indirect relationship between PTSD in the medium term and in the long
term, depending on the presence of comorbidities, a finding that is consistent with the results
of other studies. On the one hand, we know PTSD in the medium term is linked to the presence
of comorbidities (Bugge et al., 2017; Vlahov et al., 2002) - symptoms that usually develop over
the medium to long term. On the other hand, the presence of co-morbidities has been reported
to contribute to PTSD (Adams et al., 2019; Vázquez et al., 2006). Furthermore, our results
1
The DTS is a 17-item self-report measure that assesses the 17 DSM-IV symptoms of PTSD
16
indicate the influence of gender on the presence of comorbidities, as reported in the literature
(Hales et al., 2014; Salguero et al., 2011).
We found an association between social support and medium-term PTSD and comorbidities.
However, we did not find any association between social support and PTSD in the long term.
We therefore hypothesise that the absence of social support indirectly influences PTSD in the
long term, but that this relationship is mediated by PTSD in the medium term and the presence
of comorbidities in the long term.
Nor did we find a relationship between a history of trauma and PTSD in the long term. We
therefore hypothesise that a history of trauma is more likely to influence medium-term PTSD,
as suggested by our results.
We also initially hypothesised that secondary exposure to the attacks in November 2015 attacks
would be linked to PTSD in the long term as well as to the presence of comorbidities. This
relationship was not however found, so we suppose that the small number of people exposed
secondarily (n = 23) did not enable us to highlight this relationship. Similarly, we do not find
an association between PTSD in the long term and having been interviewed after July 2016 at
T2 probably in relation to the geographical distance of this event, which did not cause
reactivation of PTSD among these people.
Finally, we did not find a relationship between gender and peri-traumatic reactions, or a history
of psychological and psychiatric disturbances. Some authors consider that it is the over-
activation of neural networks in the processing of fear that could explain gender differences in
the prevalence of PTSD (Olff et al., 2007). It can be assumed from our results that gender is
linked to trauma in the medium term (with men experiencing fewer traumatic events than
women) and to peritraumatic reactions (with men having a lesser need to process fear than
women) and thus indirectly PTSD.
17
5. LIMITATIONS
This study has several limitations. Firstly, recruitment was carried out on a voluntary basis, it
is therefore possible that there was a selection bias in the initial participation (at T1): the least
exposed, those who were doing well, may have felt that they did not qualify to participate or,
conversely, the most exposed, those who were doing worst, were not in a state to participate.
Secondly, the small sample size probably did not enable us to find certain relationships, for
instance between secondary exposure to the November 2015 attacks and PTSD. Thirdly, we
could not infer causal relationships because we were not able to measure the state of mental
health before the attack, but we do have proxy variables measuring the existence of treatment
or psychological follow-up before the attack and some characteristics could have changed
between measurement times (social support, etc.). Finally, our study certainly presents biases.
We admit a potential memory bias concerning issues relating to peritraumatic reactions and the
level of perceived exposure, which were collected 6-10 months after the event.
Nevertheless, our study has several strengths. Firstly, the study was carried out very soon after
the event, and recruitment aiming to reach as many exposed people as possible was a demanding
process. Secondly, mental health outcomes and peri-traumatic reactions were measured using
reliable, standardised diagnostic tools. Thirdly, a study conducted as part of the IMPACTS
survey (Vuillermoz et al., 2020) indicated that attrition in the long term was not associated with
exposure or mental health disorders in the medium term. Thus, it is unlikely that there was an
attrition bias between the two rounds of surveys (for instance those who were the most exposed
or the least disturbed). Fourthly, the statistical power of the model (84%) and the fit indices
were very satisfactory (Asparouhov & Muthen, 2010). However, we found a TLI adjustment
index value slightly higher than 1, which indicates that the indices were probably slightly
overestimated. This can be the case when categorical variables are used (Savalei, 2020; Xia &
Yang, 2019). Fifthly, this survey enabled the study of the phenomenon of secondary exposure
18
of some participants to other attacks that occurred in France (the November 2015 attacks in
Paris and the Nice attack of July 13, 2016 took place during the second round of the survey).
6. CONCLUSION
Our study confirms the role of social support on mental health in the context of mass traumatic
events, or of gender on the presence of comorbidities. It suggests that it is not medium-term
PTSD directly that influences the presence of long-term PTSD but that medium-term PTSD
leads to comorbidities which in turn influence the presence of long-term PTSD. Also, that it is
not having a high level of exposure that influences PTSD but the presence of peri-traumatic
reactions (itself influenced by a high level of exposure).
This implies for research that these variables/relationships should be taken into account in the
analyses, for public health, this article suggests the importance of providing "early" social
support in preventing the presence of PTSD, anxiety and depressive disorders. In the case of
future mass traumatic events there is a need (i) to screen more for peri-traumatic reactions
during the attack, (ii) to take into account the "feeling" of exposure rather than the categories
of exposure, (iii) to raise awareness in the entourage of those exposed about the importance of
their support and (iv) to screen for anxiety and depressive disorders and PTSD at an early stage.
It would be interesting to test this model on another population exposed to mass trauma in order
to clarify, and validate or invalidate our hypotheses.
Funding Statement
The IMPACTS survey was financially funded by the “Fondationd’Aide aux Victimes” and was
coordinated by Santé publique France and the Greater Paris regional health agency (ARS-IdF).
The funding bodies had no role in the design of the study, data collection, analysis,
interpretation of data, and writing the manuscript.
19
Acknowledgements
The authors are most grateful to all the study participants for their involvement especially given
the very difficult context for them, and to all the interviewers. The authors would like to
acknowledge the support from the national French agency for public health as well as the
regional agency for health of Paris area: Anne Gallay, Thierry Cardoso, Anne Laporte, Jean
Claude Desenclos, ClothildeHachin, NaineIsserlis, Michel Gentile, Laurent Kosorotoff, Martial
Mettendorf, Claude Evin, and François Bourdillon for generously contribution their time and
energy to the conduct of IMPACTS study. A special thanks to Laurent Bernard-Brunel, Thierry
Baubet, Alexandra Botero, Jean-Michel Coq, Nathalie Cholin, Nicolas Dantchev, Elise Neff,
Marc Grohens, Aurelia Rochedreux, Toufik Selma, Laure Zeltner and Julien Sonnesi for their
field investigation. We wish to thank the members of the scientific committee of the study. We
wish to thank Sarah Verdier Leyshon and Angela Swaine Verdier for their careful reading of
the final manuscript.
Availability of data and materials
The data that support the findings of this study are available from The French Public Health
Agency (Santé Publique France), but restrictions apply to the availability of these data, which
were used under license for the current study, and so are not publicly available. Data are
however available from the authors upon reasonable request and with permission of The French
Public Health Agency.
Ethics approval and consent to participate
The IMPACTS survey received approval from the Committee of Ethics and Deontology (CED)
of Santé Publique France in 2015, and from CNIL (the French National Commission on
Informatics and Liberties, notice No. 915262), CPP (the French ethical research committee,
notice No. 3283) and CCTIRS (the French Advisory Committee on Information Processing in
20
Material Research in the Field of Health, notice No. 150522B-31). Written informed consent
was obtained from all participants.
Consent for publication
Not applicable.
Declaration of interest: none
Category of proposed article: Article with an empirical dimension
References
Adams, S. W., Allwood, M. A., & Bowler, R. M. (2019). Posttraumatic Stress Trajectories in
World Trade Center Tower Survivors: Hyperarousal and Emotional Numbing Predict
Symptom Change. Journal of Traumatic Stress, 32(1), 67–77.
https://doi.org/10.1002/jts.22357
Amoretti, N. (2018). Terrorist Attack in Nice, France, in July 2016: Massive Influx of Patients
to a Radiology Department. Radiology, 288(1), 2–3.
https://doi.org/10.1148/radiol.2018172915
Asparouhov, T., & Muthen, B. (2010). Weighted Least Squares Estimation with Missing Data
(Mplus technical appendix).
https://www.statmodel.com/download/GstrucMissingRevision.pdf
Bachet, F. (1999). L’utilisation des modèles dans l’analyse des structures de covariance.
L’année psychologique, 99(1), 99–122. https://doi.org/10.3406/psy.1999.28549
Ben-Ezra, M., Leshem, E., & Goodwin, R. (2015). In the Wake of National Trauma:
Psychological Reactions Following the Charlie Hebdo Terror Attack. American Journal
of Psychiatry, 172(8), 795–796. https://doi.org/10.1176/appi.ajp.2015.15020239
21
Birmes, P., Brunet, A., Benoit, M., Defer, S., Hatton, L., Sztulman, H., & Schmitt, L. (2005).
Validation of the Peritraumatic Dissociative Experiences Questionnaire self-report
version in two samples of French-speaking individuals exposed to trauma. European
Psychiatry: The Journal of the Association of European Psychiatrists, 20(2), 145–151.
https://doi.org/10.1016/j.eurpsy.2004.06.033
Bonanno, G. A., Brewin, C. R., Kaniasty, K., & Greca, A. M. L. (2010). Weighing the Costs of
Disaster: Consequences, Risks, and Resilience in Individuals, Families, and
Communities. Psychological Science in the Public Interest, 11(1), 1–49.
https://doi.org/10.1177/1529100610387086
Bracha, H. S., Williams, A. E., Haynes, S. N., Kubany, E. S., Ralston, T. C., & Yamashita, J.
M. (2004). The STRS (shortness of breath, tremulousness, racing heart, and sweating):
A brief checklist for acute distress with panic-like autonomic indicators; development
and factor structure. Annals of General Hospital Psychiatry, 3(1), 8.
https://doi.org/10.1186/1475-2832-3-8
Brewin, C. R., Andrews, B., & Valentine, J. D. (2000). Meta-analysis of risk factors for
posttraumatic stress disorder in trauma-exposed adults. Journal of Consulting and
Clinical Psychology, 68(5), 748–766. https://doi.org/10.1037//0022-006x.68.5.748
Bugge, I., Dyb, G., Stensland, S. Ø., Ekeberg, Ø., Wentzel-Larsen, T., & Diseth, T. H. (2017).
Physical Injury and Somatic Complaints: The Mediating Role of Posttraumatic Stress
Symptoms in Young Survivors of a Terror Attack. Journal of Traumatic Stress, 30(3),
229–236. https://doi.org/10.1002/jts.22191
Carragher, N., Sunderland, M., Batterham, P. J., Calear, A. L., Elhai, J. D., Chapman, C., &
Mills, K. (2016). Discriminant validity and gender differences in DSM-5 posttraumatic
stress disorder symptoms. Journal of Affective Disorders, 190, 56–67.
https://doi.org/10.1016/j.jad.2015.09.071
22
Delahanty, D. L., & Nugent, N. R. (2006). Predicting PTSD prospectively based on prior trauma
history and immediate biological responses. Annals of the New York Academy of
Sciences, 1071, 27–40. https://doi.org/10.1196/annals.1364.003
Dyb, G., Jensen, T., Glad, K. A., Nygaard, E., & Thoresen, S. (2014). Early outreach to
survivors of the shootings in Norway on the 22nd of July 2011. European Journal of
Psychotraumatology, 5. https://doi.org/10.3402/ejpt.v5.23523
Elhai, J. D., Patrick, S. L., Anderson, S., Simons, J. S., & Frueh, B. C. (2006). Gender- and
trauma-related predictors of use of mental health treatment services among primary care
patients. Psychiatric Services (Washington, D.C.), 57(10), 1505–1509.
https://doi.org/10.1176/ps.2006.57.10.1505
Falissard, B. (2011). Analysis of Questionnaire Data with R. Chapman and Hall/CRC.
https://doi.org/10.1201/b11190
Fekih-Romdhane, F., Chennoufi, L., & Cheour, M. (2017). PTSD and Depression Among
Museum Workers After the March 18 Bardo Museum Terrorist Attack. Community
Mental Health Journal, 53(7), 852–858. https://doi.org/10.1007/s10597-017-0085-z
Folkman, S. (2013). Stress: Appraisal and Coping. In M. D. Gellman & J. R. Turner (Eds.),
Encyclopedia of Behavioral Medicine (pp. 1913–1915). Springer New York.
https://doi.org/10.1007/978-1-4419-1005-9_215
Gilbar, O., Taft, C., & Dekel, R. (2020). Male intimate partner violence: Examining the roles
of childhood trauma, PTSD symptoms, and dominance. Journal of Family Psychology:
JFP: Journal of the Division of Family Psychology of the American Psychological
Association (Division 43), 34(8), 1004–1013. https://doi.org/10.1037/fam0000669
Glad, K. A., Jensen, T. K., Hafstad, G. S., & Dyb, G. (2016). Posttraumatic stress disorder and
exposure to trauma reminders after a terrorist attack. Journal of Trauma & Dissociation:
23
The Official Journal of the International Society for the Study of Dissociation (ISSD),
17(4), 435–447. https://doi.org/10.1080/15299732.2015.1126777
Hales, R. E., Yudofsky, S. C., MA, L. W. R. M., & Kupfer, D. J. (2014). The American
Psychiatric Publishing Textbook of Psychiatry (6th Revised edition). American
Psychiatric Association Publishing.
https://doi.org/10.1176/appi.books.9781615372980
Hamama-Raz, Y., Palgi, Y., Shrira, A., Goodwin, R., Kaniasty, K., & Ben-Ezra, M. (2015).
Gender differences in psychological reactions to Hurricane Sandy among New York
Metropolitan Area residents. The Psychiatric Quarterly, 86(2), 285–296.
https://doi.org/10.1007/s11126-014-9333-3
Harandi, T. F., Taghinasab, M. M., & Nayeri, T. D. (2017). The correlation of social support
with mental health: A meta-analysis. Electronic Physician, 9(9), 5212–5222.
https://doi.org/10.19082/5212
Henriksen, C. A., Bolton, J. M., & Sareen, J. (2010). The psychological impact of terrorist
attacks: Examining a dose-response relationship between exposure to 9/11 and Axis I
mental disorders. Depression and Anxiety, 27(11), 993–1000.
https://doi.org/10.1002/da.20742
Hutchinson, S., & Olmos, A. (1998). Behavior of Descriptive Fit Indexes in Confirmatory
Factor Analysis Using Ordered Categorical Data. Structural Equation Modeling-a
Multidisciplinary Journal - STRUCT EQU MODELING, 5, 344–364.
https://doi.org/10.1080/10705519809540111
Jehel, L., Duchet, C., Paterniti, S., Louville, P., & Carli, P. (1999). Évaluation des troubles
psychotraumatiques des victimes d’un attentat terroriste. Une étude prospective.
Médecine de Catastrophe - Urgences Collectives, 2, 132–138.
24
Jolly, A. (2000). Evénements traumatiques et état de stress post-traumatique. Une revue de la
littérature épidémiologique. Annales Médico-Psychologiques, Revue Psychiatrique,
158(5), 370–378.
Maguen, S., Luxton, D. D., Skopp, N. A., & Madden, E. (2012). Gender differences in traumatic
experiences and mental health in active duty soldiers redeployed from Iraq and
Afghanistan. Journal of Psychiatric Research, 46(3), 311–316.
https://doi.org/10.1016/j.jpsychires.2011.11.007
Mansiaux, Y. (2014). Analyse d’un grand jeu de données en épidémiologie: Problématiques et
perspectives méthodologiques [Phdthesis, Université Pierre et Marie Curie - Paris VI].
theses.fr. https://tel.archives-ouvertes.fr/tel-01089321
Marmar, C. R., Weiss, D. S., Metzler, T. J., Ronfeldt, H. M., & Foreman, C. (1996). Stress
responses of emergency services personnel to the loma prieta earthquake interstate 880
freeway collapse and control traumatic incidents. Journal of Traumatic Stress, 9(1), 63–
85. https://doi.org/10.1002/jts.2490090107
Melrose, K., Brown, G., & Wood, A. (2015). When is received social support related to
perceived support and well-being? When it is needed. Personality and Individual
Differences, 77, 97–105. https://doi.org/10.1016/j.paid.2014.12.047
Monfort, E., & Afzali, M. H. (2017). Traumatic stress symptoms after the November 13th 2015
Terrorist Attacks among Young Adults: The relation to media and emotion regulation.
Comprehensive Psychiatry, 75, 68–74.
https://doi.org/10.1016/j.comppsych.2017.02.015
Morin, T., Jaluzot, L., & Picard, S. (2013, November). Femmes et hommes face à la violence—
Insee Première. INSEE PREMIÈRE, 1–4.
Motreff, Y., Baubet, T., Pirard, P., Rabet, G., Petitclerc, M., Stene, L. E., Vuillermoz, C.,
Chauvin, P., & Vandentorren, S. (2020). Factors associated with PTSD and partial
25
PTSD among first responders following the Paris terror attacks in November 2015.
Journal of Psychiatric Research, 121, 143–150.
https://doi.org/10.1016/j.jpsychires.2019.11.018
Newsom, J. (2018). Practical Approaches to Dealing with Nonnormal and Categorical
Variables. Psychiatry. http://web.pdx.edu/~newsomj/semclass/ho_estimate2.pdf
Olff, M., Langeland, W., Draijer, N., & Gersons, B. P. R. (2007). Gender differences in
posttraumatic stress disorder. Psychological Bulletin, 133(2), 183–204.
https://doi.org/10.1037/0033-2909.133.2.183
Ozer, E. J., Best, S. R., Lipsey, T. L., & Weiss, D. S. (2003). Predictors of posttraumatic stress
disorder and symptoms in adults: A meta-analysis. Psychological Bulletin, 129(1), 52–
73.
Perlman, S. E., Friedman, S., Galea, S., Nair, H. P., Erős-Sarnyai, M., Stellman, S. D., Hon, J.,
& Greene, C. M. (2011). Short-term and medium-term health effects of 9/11. The
Lancet, 378(9794), 925–934. https://doi.org/10.1016/S0140-6736(11)60967-7
Pettersson, A., Modin, S., Wahlström, R., af Winklerfelt Hammarberg, S., & Krakau, I. (2018).
The Mini-International Neuropsychiatric Interview is useful and well accepted as part
of the clinical assessment for depression and anxiety in primary care: A mixed-methods
study. BMC Family Practice, 19(1), 19. https://doi.org/10.1186/s12875-017-0674-5
Pirard, P., Motreff, Y., Lavalette, C., Vandentorren, S., Baubet, T., & Messiah, A. (2018).
Mental health impact and health care use of the bereaved persons 8-11 months after the
November 2015 Paris terror attacks. European Journal of Public Health, 28(suppl_4),
cky213.635. https://doi.org/10.1093/eurpub/cky213.635
Preacher, K. J., & Coffman, D. L. (2006, May). Power and Sample Size using RMSEA.
Quantpsy.Org. http://quantpsy.org/rmsea/rmsea.htm
26
Rigdon, E. E. (1994). Calculating degrees of freedom for a structural equation model. Structural
Equation Modeling: A Multidisciplinary Journal, 1(3), 274–278.
https://doi.org/10.1080/10705519409539979
Robert, S., Lesieur, S., Chastang, J., Kergoat, V., Dutertre, J., & Chauvin, P. (2017). [Health
and access to health care of 18-25 years old people not in employment, education or
training (NEETs) attending the French ’missions locales’]. Revue D’epidemiologie Et
De Sante Publique, 65(4), 265–276. https://doi.org/10.1016/j.respe.2017.01.120
Rouillon, F., Barrier, G., Baruffol, E., Benayoun, S., Chambon, E., Evrard, M., Groult, J.,
Lamping, D., Lebigot, F., Lert, F., Rondet, P., & Verger, P. (2001, January 31).
Présentation de l’étude épidémiologique des conséquences sanitaires des attentats
(ECSAT). S.O.S. Attentats, France.
Salguero, J. M., Fernández-Berrocal, P., Iruarrizaga, I., Cano-Vindel, A., & Galea, S. (2011).
Major depressive disorder following terrorist attacks: A systematic review of
prevalence, course and correlates. BMC Psychiatry, 11(1), 96.
https://doi.org/10.1186/1471-244X-11-96
Santiago, P. N., Ursano, R. J., Gray, C. L., Pynoos, R. S., Spiegel, D., Lewis-Fernandez, R.,
Friedman, M. J., & Fullerton, C. S. (2013). A systematic review of PTSD prevalence
and trajectories in DSM-5 defined trauma exposed populations: Intentional and non-
intentional traumatic events. PloS One, 8(4), e59236–e59236. PubMed.
https://doi.org/10.1371/journal.pone.0059236
Savalei, V. (2020). Improving Fit Indices in Structural Equation Modeling with Categorical
Data. Multivariate Behavioral Research, 0(0), 1–18.
https://doi.org/10.1080/00273171.2020.1717922
Scharff, A., Ortiz, S. N., Forrest, L. N., & Smith, A. R. (2021). Comparing the clinical
presentation of eating disorder patients with and without trauma history and/or
27
comorbid PTSD. Eating Disorders, 29(1), 88–102.
https://doi.org/10.1080/10640266.2019.1642035
Sheehan, D. V., Lecrubier, Y., Sheehan, K. H., Amorim, P., Janavs, J., Weiller, E., Hergueta,
T., Baker, R., & Dunbar, G. C. (1998). The Mini-International Neuropsychiatric
Interview (M.I.N.I.): The development and validation of a structured diagnostic
psychiatric interview for DSM-IV and ICD-10. The Journal of Clinical Psychiatry, 59
Suppl 20, 22-33;quiz 34-57.
Sideridis, G., Simos, P., Papanicolaou, A., & Fletcher, J. (2014). Using Structural Equation
Modeling to Assess Functional Connectivity in the Brain: Power and Sample Size
Considerations. Educational and Psychological Measurement, 74(5), 733–758.
https://doi.org/10.1177/0013164414525397
Silver, R. C., Holman, E. A., McIntosh, D. N., Poulin, M., & Gil-Rivas, V. (2002). Nationwide
longitudinal study of psychological responses to September 11. JAMA, 288(10), 1235–
1244.
Simard, C. (2018). Conception et validation d’un modèle théorique de la résilience auprès des
étudiants universitaires (http://hdl.handle.net/20.500.11794/28371) [Doctorat en
Mesure et évaluation Philosophiæ Doctor (Ph. D.), Québec].
https://corpus.ulaval.ca/entities/publication/89d1728e-af59-45c1-b44b-6fb1d724172f
Solla, F., Carboni, J., Bréaud, J., Babe, P., Brézac, G., Chivoret, N., Dupont, A., & Fernandez,
A. (2018). July 14, 2016, Terror Attack in Nice, France. Academic Pediatrics, 18(4),
361–363. https://doi.org/10.1016/j.acap.2018.01.004
Spoont, M., Arbisi, P., Fu, S., Greer, N., Kehle-Forbes, S., Meis, L., Rutks, I., & Wilt, T. (2013).
Screening for Post-Traumatic Stress Disorder (PTSD) in Primary Care: A Systematic
Review (Washington (DC): Department of Veterans Affairs (US)).
https://www.ncbi.nlm.nih.gov/books/NBK126691/
28
Vandentorren, Pirard, Sanna, Aubert, Motreff, Dantchev, N., Lesieur, S., Chauvin, P., &
Baubet, T. (2018). Healthcare provision and the psychological, somatic and social
impact on people involved in the terror attacks in January 2015 in Paris: Cohort study.
British Journal of Psychiatry, 212(4), 207–214. https://doi.org/10.1192/bjp.2017.63
Vázquez, C., Hervás, G., & Pérez-Sales, P. (2008). Chronic thought suppression and
posttraumatic symptoms: Data from the Madrid March 11, 2004 terrorist attack. Journal
of Anxiety Disorders, 22(8), 1326–1336. https://doi.org/10.1016/j.janxdis.2008.01.014
Vázquez, C., Pérez-Sales, P., & Matt, G. (2006). Post-traumatic stress reactions following the
March 11, 2004 terrorist attacks in a Madrid community sample: A cautionary note
about the measurement of psychological trauma. The Spanish Journal of Psychology,
9(1), 61–74. https://doi.org/10.1017/s1138741600005989
Vlahov, D., Galea, S., Resnick, H., Ahern, J., Boscarino, J. A., Bucuvalas, M., Gold, J., &
Kilpatrick, D. (2002). Increased use of cigarettes, alcohol, and marijuana among
Manhattan, New York, residents after the September 11th terrorist attacks. American
Journal of Epidemiology, 155(11), 988–996.
Vuillermoz, C., Stene, L. E., Aubert, L., Motreff, Y., Pirard, P., Baubet, T., Lesieur, S.,
Chauvin, P., & Vandentorren, S. (2020). Non-participation and attrition in a longitudinal
study of civilians exposed to the January 2015 terrorist attacks in Paris, France. BMC
Medical Research Methodology, 20(1), 63. https://doi.org/10.1186/s12874-020-00943-
x
Welch, A. E., Caramanica, K., Maslow, C. B., Brackbill, R. M., Stellman, S. D., & Farfel, M.
R. (2016). Trajectories of PTSD Among Lower Manhattan Residents and Area Workers
Following the 2001 World Trade Center Disaster, 2003–2012. Journal of Traumatic
Stress, 29(2), 158–166. https://doi.org/10.1002/jts.22090
29
Wood, C. M., Salguero, J. M., Cano-Vindel, A., & Galea, S. (2013). Perievent panic attacks
and panic disorder after mass trauma: A 12-month longitudinal study. Journal of
Traumatic Stress, 26(3), 338–344. https://doi.org/10.1002/jts.21810
Xia, Y., & Yang, Y. (2019). RMSEA, CFI, and TLI in structural equation modeling with
ordered categorical data: The story they tell depends on the estimation methods.
Behavior Research Methods, 51(1), 409–428. https://doi.org/10.3758/s13428-018-
1055-2
Zambaldi, C. F., Cantilino, A., Farias, J. A., Moraes, G. P., & Sougey, E. B. (2011). Dissociative
experience during childbirth. Journal of Psychosomatic Obstetrics and Gynaecology,
32(4), 204–209. https://doi.org/10.3109/0167482X.2011.626092
30
Table
Table 1: Descriptive Table of Characteristics of IMPACTS study participants
31
Participants at 6-10 months
n = 123
Participants at 18-22 months
n = 123
N (%)
N (%)
Female
75 (61.0)
Median age [min - max]
42 [23 - 79]
43 [24 - 80]
Marital status
Single
43 (35)
47 (38.2)
Divorced
11 (8.9)
9 (7.3)
Married
69 (56.1)
67 (54.5)
Educational level
Third-level education
89 (73)
High-school diploma or less
33 (27)
Born in France
118 (95.9)
Type of exposure
Indirect witnesses
32 (26)
Indirectly threatened
55 (44.7)
Directly threatened
36 (29.3)
Physical proximity with the perpetrators
Having lost a loved one
13 (10.6)
Immediate proximity
55 (44.7)
Distant from the scene of the event
55 (44.7)
Median STRS score [min - max]
24 [0 - 52]
High PDEQ score (>15)
66 (54.5)
High level of perceived terror exposure
91 (75.8)
No everyday-support in life
14 (11.4)
14 (11.4)
No financial support
26 (21.1)
24 (19.5)
Social isolation
20 (16.3)
27 (22)
Pre-event psychological follow-up
37 (30.1)
Difficult life event the previous year
46 (38)
32
History of trauma
68 (55.3)
51 (41.5)
Psychological follow-up in the past
26 (21.1)
Secondary exposure in the November 2015 attacks
23 (18.7)
Having been interviewed at T2 before the terrorist
attack in Nice in 2016
49 (39.8)
PTSD in the last month
20 (16.3)
18 (14.6)
At least one anxiety disorder
39 (32)
53 (43.1)
Current depression
52 (42.3)
38 (30.9)
