Twelve months after being directly exposed to the terrorist attack perpetrated in London on March 22^nd, 2017, 25.6% of the teenagers in our sample met DSM-5 criteria for PTSD. Concurrent to PTSD, elevations in psychiatric comorbidities, changes in consumption of toxic substances, and impaired functioning in all areas of their life were also evident. Other than PTSD, 64.1% of the participants also expressed a current psychiatric disorder. Apart from an increased prevalence of impaired concentration, days of reduced effectiveness and personal history of anxiety disorders, in the sample which had received mental health care, there were no statistically significant differences did, vs. did not, consult beyond 1 week after the attack.

Notably, few studies have investigated a similar target population (adolescents aged 16 years and directly exposed to a terrorist attack) or with the same experimental protocol (intentional and man-made event; clinical assessment at 1 year). The PTSD prevalence of 25.6% in our study (with 25.6% satisfying criteria for intrusive, avoidance and hyperarousal symptoms and 23.1% satisfying criteria for negative alterations in cognition and mood) is consistent with the overall literature in this area, including in children, adolescents (4) and adults (6, 7). A large sample of Norwegian high school (27) students were assessed following the Oslo terror attack on July 22^nd, 2011 (n = 10,220; mean age 16.9 years; 53% of girls), although only 1.8% of this sample were on site. Consequently, PTSD prevalence varied according to exposure level, with only 0.8% of all respondents reporting substantial distress on the reexperiencing item (DSM IV), 4.9% on the avoidance item, 1.1% on the hyperarousal item and 0.4% on all three areas, approximately 7 months after the event. Following the World Trade Center Attack, a sample of 817 adolescents, aged 13–18 years, was drawn from a Jewish high school, although these young people were not directly exposed to the attack. Consequently, at 1 year after the attack, only 2.3% seemed to meet PTSD criteria (28). Among adults involved in the terror attacks in January 2015 in Paris (civilian, n = 190, median age 41), 31% of directly threatened participants met PTSD criteria at 6 months (29).

Other psychological disorders were also evident at 1 year in our target population, which is also in accordance with the literature (2, 3, 8, 9). A systematic review (30) suggested the risk of major depressive episode, in adults directly exposed to a terrorist attack, ranged from 20 to 30% in the first months after its occurrence. Chemtob et al. (31) also found an association between exposure to terrorism and PTSD, suicidal ideation and functional impairment in a sample of 2,094 Israeli adolescents, aged 12 to 18 years. At 1 year following the World Trade Center attack, adolescents with PTSD showed an increased risk for suicidal ideation (28) and after 18 months, increased substance use was associated with impaired school performance (11). In the Paris, 2015 attacks, within the 58 directly threatened civilians (29), 19% had depression, 38% anxiety disorder and 19% increased suicide risk, at 6 months after the attacks. Given the high prevalence of 38.5% of participants with PTSD having a depressive episode and 20.5% having suicidal ideation in our target population, it may be important to note that data collection occurred around the first anniversary date on the event. Consequently, data collection occurred at an emotionally heightened time-point, although the attacks in London were more discreetly commemorated in France, in contrast to the widely commemorated Paris, 2015 attacks.

Epidemiologically, the risk factors for PTSD in our population were generally in agreement with the wider literature. In a meta-analysis, Brewin et al. (32) classified risk factors according to the strength of their predictive effects in three categories. (1) Some factors, such as gender, did not appear in all studies, although a consistent pattern for women to be at higher risk for PTSD than men was evident. In our study, both groups (with and without PTSD), included more girls but the groups did not statistically differ on gender. (2) Other factors, such as previous trauma, predicted PTSD more consistently, with variations at least partly being explained by the methodology used. In our work, having witnessed a physical assault was statistically significant in predicting the development of PTSD. (3) Factors such as personal or family psychiatric history, childhood abuse, had a more uniform predictive effect, and the significant effects of these factors were also present in our results. In our study, depressive disorders were again noted here, as a risk factor for PTSD such as personal or familial history of mental disorders.

However, several studies have enhanced the stronger influence of peri- and post-traumatic factors, than pre-traumatic factors (13, 32–34), on PTSD risk. Trickey et al. (13) added the impact of superior subjective experience to the objective elements, even if as in the situation studied, strong objective elements were present such as the severity of the events and the level of exposure (the car just stopped a few meters away from the participants). As to the importance of post-trauma factors, it is notable that in our study, low social support was significantly associated with PTSD risk at 1 year. This is similar to the meta-analysis by Trickey et al. (13), which focussed on children and adolescents, as well as with data on the first responders at 12 months after Paris terror attacks (34) in November 2015, or in a systematic review (35) about consequences of 9/11 on children's and young adult's mental health. About adolescents, in this last review of Rousseau et al. (35), low family support as parental unavailability to discuss the events or parent-adolescent conflict was highlighted. By way of qualifying social support as a predictor, in addition, with a comparison of two meta-analysis (32, 33), social support appeared a stronger predictor when the event had happened at least 3 years before the moment of the study.

According to our results, it is also worth mentioning these PTSD risk factors did not overlap with the effects of predictive factors as to whether follow up did, vs. did not, occur beyond 1 week.

As a consequence of France's experience of wars, disasters, and terrorist attacks, medico-psychological units have been deployed since July 1995, providing initial and longer-term support (36, 37). Such immediate and longer-term support was provided for the attacks in Paris 2015 (29, 36) and in Nice 2016, where more than 30,000 people were present, including babies, children and adolescents (38, 39). The Nice attack, with a young population impacted (39) (in the early care phase, 365 children and adolescents were registered at the pediatric medico-psychological emergency unit with 25.5% <6 years old, 44.7% 6–11, and 29.8% 12–17 years old) and experience of Lenval University Children's Hospital, is relevant to our work. It was indeed observed a great number of families (39) spontaneously converged to Children's Hospital in search of psychological or medical help, as in our work where all teenagers with PTSD 1 year after the attack received post-immediate care and 70% were still being followed up at the time of our study. The Nice study (39), like ours, highlighted and encouraged the follow-up of adolescents, especially if they are vulnerable, in a standardized manner, to ensure the screening and treatment of all young people.

Cohort studies conducted following other terrorist attacks, such as 9/11, the Madrid, 2004 car bombings, the 2011 Oslo/Utoya massacre and Paris, 2015, showed participation rates ranging from 40 to 70% in the first wave, and 50 to 75% in the second wave (40, 41). Although these attacks vary considerably in their details, the participation rates are similar, and may be improved in younger people via the utilization of social media, peers, and educational environments, as noted by Vuillermoz et al. following the Paris, 2015 attacks (41). The higher level (73.6 %) of participation in the current study could be explained by various details: (1) the small size of the total target population which allowed us to identify each teenager present in the event; (2) links with the network set up at the local level (high-school, academy and diocese, partners of justice, medico-psychological emergency unit, care, and political bodies); (3) awareness of all the partners in order to deploy a prevention and care message, when the research was launched, in an information meeting in the high school; (4) personalized phone call if necessary; and (5) creation of a specific email address, allowing participants to engage in discussions, including with family and high-school staff. Such a simple provision may be especially important for an adolescent population. Moreover, beyond the research dimension, raising awareness in the network about adolescents after the attack, involving parents, professors, but also political actors, seemed essential to us to build this social link, which is recognized as positive in the prognosis (13, 34, 35). Our idea was already to open the discussion so that young people in difficulty do not remain isolated. On several occasions, contact details of professionals to support these teenagers were sent and some were then able to benefit from an adapted and specialized follow up. Regarding adolescent psychopathology, in fact, taking the first step is sometimes necessary to initiate care.

Furthermore, about care and therapeutics, a whole new literature is appearing: for example, current research focuses on the treatment with ketamine of major depressive episodes, suicidal ideation (42–44) and chronic PTSD (45, 46) in adult populations. Common pathophysiological factors could indeed be involved in the complexity of PTSD, but also in the depressive episode, suicidality (42). In France in addition, suicide is the second leading cause of death for young people aged 15 to 24 (47). Nevertheless, ketamine is beginning to be explored in pediatrics in acute pain (48), but studies will be continued and questioned for PTSD and suicidality in children, teenagers. The therapeutic work also remains non negligible and Bianchini et al. (49) showed a significantly decreased in PTSD symptomatology and psychological distress severity with Cognitive Behavior Therapy in 39 young subjects, followed and evaluated 1 year after l'Aquila Earthquake.