A study of the mental health effects of the 1988 Armenian earthquake: The search for post-traumatic stress disorder

Background. The nosologic structure of post-trauma mental distress has been based on the post-traumatic stress disorder (PTSD) diagnosis since its introduction into the psychiatric nomenclature in the DSM-III. At its inception, PTSD was used to explain the psychological symptoms expressed by Vietnam War veterans, but over the course of time, and with the multiple revisions of the DSM, the disorder has been applied more broadly to a wide variety of traumatic exposures. Unfortunately, the utility of the diagnosis for other trauma exposures (e.g. natural disasters) has generally not been questioned. Given the evidence of a general increase in prevalence for several other mental disorders following natural disasters, we examined whether PTSD denotes a cohesive syndrome that can be qualitatively differentiated from major depressive disorder and generalized anxiety disorder among survivors of the 1988 Armenian earthquake.; Methods. As part of a larger epidemiologic study, data were collected on survivors 2 years following the 1988 Armenian earthquake. Study participants (N = 1781) completed a standardized interview measuring earthquake exposures, a set of psychological symptoms for several disorders, and general information about health and economic outcomes since the earthquake. The analytic sample (N = 818) was comprised of participants who affirmed that they had been traumatized by the earthquake and had experienced symptoms of depression and anxiety since the earthquake. Multiple statistical methods were used to examine symptom data in the search for a discrete PTSD syndrome. Factor and latent class analyses were used to identify latent constructs representing dimensional and categorical syndromes, respectively. Multivariate statistics were used to test whether the prevalence of any symptom or set of symptoms were specifically associated with a high degree of exposure to the acute effects of the earthquake.; Results. All analyses were conducted stratified by gender. An examination of the factors from multiple factor analysis models revealed several made up of different types of symptoms providing evidence for dimensional syndromes within the data. However, none of the factors contained the constellation of symptoms currently used to diagnose PTSD. Evidence from the latent class models indicated that adequate fit to the data was accomplished assuming an underlying categorical structure. An evaluation of the different models did not reveal a single class that appeared to contain the set of PTSD symptoms as it is currently diagnosed in the DSM. The multivariate analysis using all the symptoms did not result in finding a set of symptoms more associated with high acute exposure to the earthquake. Instead, the prevalence rates for all the symptoms were consistently higher in the high exposure compared with low exposure group.; Conclusion. The results suggest a relatively complex typology of post-trauma syndromes associated with earthquake exposure. They indicate that the structure of PTSD currently proposed by the DSM was not the best way to represent the data. The findings shed light on the mental health effects of disasters and indicate that new strategies are warranted for classification of post-trauma distress. For intervention purposes, it may be more useful to classify people by the presence or absence of symptoms or by the number of symptoms rather than by whether or not they meet criteria for any specific diagnosis.