In early September 2004, approximately 32 terrorists attacked a school in Beslan, North Ossetia, in the Russian Federation. At the time of the attack, the school was holding festivities to mark the first day of classes after summer holiday. Over 1300 adults and children were taken hostage in the siege, and during a three-day period of detention, 344 hostages died (including 186 children) and more than 700 were injured. This event had an extensive negative impact on the health of children, families, and the entire Beslan community [1].

For three years following the Beslan school hostage tragedy, the Swiss Department of Development and Cooperation (SDC) fully financed a number of psychosocial programs under the auspices of the North Ossetian Ministry of Education. These programs were offered by staff of the local psychosocial education centre "Doverie". From the perspective of resilience [2–5], a number of recreation, sport, and psychosocial rehabilitation activities were offered to support the recovery of youth in the Beslan area. One of these programs was a "resilience-enhancing program", which was set in the nearby mountain resort of Tsey in the summer of 2006. In this program, "resilience enhancement" was attempted by offering a variety of outdoor experiences under the guidance of caring, non-family adults who instructed participants in useful problem-solving strategies, as well as life and coping skills, in a community experience with supportive peers.

Resilience enhancement and development were selected as the focus of the program for two primary reasons. First, it was hypothesized that the youth involved in these programs would benefit from enhancement of their emotional, mental and social capacities to overcome the adversities they faced. Research has shown that building resilience among survivors of high-risk environments can help develop and/or maintain: social competence, empathy, caring, problem-solving skills, critical and creative thinking, task mastery, and a sense of purpose and social connectedness [6]. Problem-solving skills are a particularly strong predictor of improved resilience in children and youth in the long-term, as improved problem-solving skills can enhance the possibility that future life challenges will be resolved successfully [5, 7–10]. And second, it was determined by the government that to intervene directly on trauma issues might produce alternative outcomes that would have regressive and destructive psychological affects on the children. Thus, the program needed to focus on strengths and capacity building.

This research project is based on the developmental psychology perspective of resilience, first conceptualized by Luthar, and later extended and applied by others [2, 4, 11–13]. Luthar's basic definition of resilience is: "A dynamic process encompassing positive adaptation within the context of significant adversity. Implicit within this notion are two critical conditions: 1.) An exposure to significant threat or severe adversity, and 2.) the achievement of positive adaptation despite major assaults on the developmental process" [2, 3, 14]. As defined by Luthar, resilience is neither a "trait" or "state," but rather a multi-dimensional dynamic process, which implies that it can occur in a person in different ways, at different times, in response to different situations [2, 14]. Resilience thus exists as a potential that may be fostered and supported within individuals, without regard to age or developmental phase.

The importance of the Beslan program is primarily in its relevance to resilience-enhancing interventions in community-based settings - interventions that may be especially influential when undertaken during significant developmental transitions, such as entry into school, advancing to higher levels of school, moving from childhood into adolescence, from school into the workforce, and from youth into adulthood [2, 3, 14]. Thus, the results of effective resilience-fostering interventions can offer practitioners in the field unique opportunities for promoting positive adaptation in a variety of adverse situations, and assisting transitions through various developmental stages [3, 5, 14].

Past research on psychological resilience has identified four key protective factors [10, 15–19] that can contribute to the development, support and sustaining of resilience processes in youth. They are: (a) healthy attachments to related and unrelated older adults, who provide them with support, encouragement, and guidance; (b) healthy and connected peer relationships; (c) effective problem-solving skills and coping strategies; and (d) community involvement, in support of the common good. These protective factors are interactive with resilience factors, as these both help develop and can later help sustain resilience processes and trajectories [5, 16].

In Beslan, the goal was to examine a program that incorporated these four vital protective factors for enhancing the development of resilience processes. The approach was to combine sport activities, safety training, and various therapeutic rehabilitation activities in a group setting, where participants camped together away from their homes. Specifically, the program provided: (1) first aid, cardio-pulmonary reanimation (CPR), and life-saving rescue techniques; (2) mountaineering and survival skills training; (3) alpine sport activities, such as safe skiing, climbing, and alpine walking; and (4) informal arts, play, and supportive talk therapy sessions. It was anticipated that each of the components would foster resilience via multiple mechanisms. For example, the physical activities in the mountain wilderness were expected to build internal resilience though boosting confidence, supporting the experience of positive emotions, improving somatic health, and encourage a climate of solidarity, both with peer participants and adult leaders.

Further explication of the therapeutic component of the program may be useful. This element was primarily designed to offer participants assistance in sorting out emotions provoked by the activities involved in the other parts of the program. In the evenings, children played parlor games, sang songs, told stories, created art, and engaged in other similar leisure activities, all of which were facilitated by psychosocial therapists. These sessions were informal and primarily supportive in focus, with no special training offered to the therapists and no particular therapeutic approach or intervention emphasized. Therapy was administered as needed. That is, in cases where a child's behavior suggested that the daytime or evening activities provoked exacerbation of psychopathology, psychologists addressed the children's experiences and emotions within the group. In the rare case where children were extremely disruptive, a psychologist would remove them, and address the behavior or emotions in a one-to-one setting.

It is important to note that the actual outdoor activities and training were conducted by rescuers (emergency medical technicians and professional mountain guides) from the North Ossetian Search and Rescue Services, a department of the Russian Federal Ministry for Emergencies. All participating rescuers were professionals who had been involved in the liberation mission of the Beslan hostages, and were thus considered quite positively by the camp's participants. Additionally, all these programs were supported by staff of the local psychosocial education centre "Doverie". All psychologists were university qualified in the Russian Federation, and all attended specific courses for psycho-traumatology in Israel prior to participating in the camp.

The primary research objective of this project was to assess whether a youth's involvement in the overall program would result in enhanced resilience processes and towards more resilient life trajectories. In the absence of a comparison group, within-subject resilience was compared at three time points: before the program, at the completion of the program, and at a six-month follow-up assessment. A secondary objective was to investigate variations in the effect of the interventions across groups, which differed in their experiences of the 2004 Beslan school tragedy. For this study, the Connor-Davidson Resilience Scale (CD-RISC) [6], was utilized to measure resilience. At the time of the study, the CD-RISC was one of the few instruments in existence that expressly assessed resilience processes. The CD-RISC has been validated by a number of studies, and is also one of the few resilience instruments that have been used cross-culturally [20–24].

Since September 2004, when members of the Beslan community were held hostage by terrorists, merely one study [37] and one clinical report [38] have addressed the community's posttraumatic stress reaction. Two further articles have chronicled narratives of caregivers [39] and observations of the affected community [40]. The current study takes a new and different perspective, evaluating a resilience intervention program for Beslan youth offered during summer holidays of 2006. Adopting a resilience perspective, rather than a trauma and psychopathology perspective, the primary goal of this program was to promote physical, mental, and social well-being in youth of the Beslan community. Specifically, the program was designed to enhance the resilience of participants via acquisition of new skills and strategies, and social engagement with peers and older non-family adults. The program was undertaken with two major scientific objectives in mind: to evaluate intra-subject change in resilience at the end of the program, and then 6 months later; and to determine whether this self-reported resilience differed, cross-sectionally and in terms of change, according to the number of trauma experiences resulting from the Beslan school tragedy. In support of the second objective, we also sought to document the extent of injuries and death associated with program participants.

The first suggestion of our findings is that the program intervention had a measurable effect on subsequent self-reported resilience levels within individuals. The average sample member reported statistically significant increases in resilience from the baseline value, at both the end of the one-week intervention and at 6-month follow-up. The second relevant finding is that participants who had been taken hostage during the school tragedy differed from non-hostages in two ways: one, hostages had lower resilience scores than non-hostages at baseline, and immediately after the program; and two, hostages experienced greater gains in resilience than non-hostages six months after the program's completion. A similar finding was implied by the analysis of differential effects according to loss and injury experience. That is, participants in the highest losses and injuries categories had significantly lower resilience than those in the lowest categories at the baseline measure, but reported greater increases in resilience from the baseline to the second follow-up.

There are numerous limitations of note in this study. First and foremost, our intervention was performed without a control group, which weakens any argument for the efficacy of the resilience enhancing program. We recommend that future studies build upon our pilot findings by incorporating quasi-experimental elements. Previous literature on school-based interventions for non-environmental traumatic events suggests that the use of a wait-list control group provides a feasible method for analysis [41–43]. Nevertheless, as wait-list controls are generally not blinded to treatment status, this approach can results in biased results. An alternative to the wait-list control approach is the use of a generic support group as a potential control against the effect of time and attention on recovery from PTSD [42]. Earlier research has, however, suggested that the implementation of such alternative methods may be deemed insensitive and unfair by family members of children in the control group [44].

Second is the lack of a specific PTSD screen in the study, a shortcoming common to other research performed with the CD-RISC, which makes it difficult to place our findings in the context of previous research. Two earlier studies of the CD-RISC, which showed that treatment of PTSD significantly improves resilience parallel to symptom reduction [22, 23], partially support our findings. However, because our study was limited to assessing resilience by request of local health authorities, who were concerned about negative impacts of trauma assessment, its results cannot help clarify to what extent low resilience scores might mirror PTSD pathology [45–47].

Third, although we have uncovered that traumatic exposure appears to modify the intra-subject change in resilience attributable to the intervention, we could not explore whether hostages and non-hostages at the same levels of trauma severity differed significantly in their resilience changes. With 94 study participants the analyses were not sufficiently powered to answer this question. Small and missing cell sizes of non-hostages at higher levels of trauma experiences precluded the type of doubly-stratified analysis--or 3-way interaction of time, hostage status, and losses or injuries--which would be necessary to draw such inferences.

Three additional points should be highlighted. One, the results should be read with some caution, as our analyses are sensitive to a type of regression-to-the-mean effect. That is, the baseline resilience reports of hostages may be an underestimate of the true value, owing to the extent of trauma experiences associated with the hostage taking. In this case, it would be natural that the reported resilience of this group would improve substantially over time, conceivably even in the absence of an intervention such as ours. Two, at present, there are no CD-RISC norms. We are thus left to consider the pre-/post-intervention changes in CD-RISC scores in a purely statistical manner, without a clear context for the clinical meaning of a given CD-RISC score or change. Efforts should be directed to independently establishing validity and reliability of the instrument for this age group, as well as developing age-related norms, so that future research of this kind may have greater clinical meaning. Finally, our study was not designed to identify mechanisms for resilience enhancement. In fact, we did not provide specific resilience training to the counselors and therapists; rather, we selected activities that we hypothesized would naturally lead to resilience enhancement, which renders impossible any isolation of mediating influences. Thus, the observed changes in resilience cannot be directly attributed to any single element of the intervention, and it remains possible that the occasional therapeutic intervention, provoked by daytime and evening activities (e.g., first aid training, game playing, music, art) explains our findings. Research has suggested that there are multiple pathways to resilience change, and individuals are remarkably resilient to extremely traumatic exposures [48, 49]. Future controlled studies should therefore be systematized so that the relative influences of various intervention components can be ascertained.

Despite the significant limitations of this study, we do know that mass violence not only manifests in posttraumatic mental health problems, but also in disrupted human attachment systems [50]. Therefore newly empowered social systems on the peer level, in addition to those of family and extended non-family associations [51], should be considered as a major element of future psychosocial response for afflicted communities. The results of this study do support this notion, and further suggests that a mixed trauma intervention group experience (i.e., groups comprising those who experienced traumatic events and those who have not) may improve the likelihood of an effective outcome.