4
As can be seen from the figure above, there are several marked changes in both the definition and the descriptive criteria. Firstly, the
stressor concept itself is amended, from "recognizable" stressors that cause "significant distress in almost everyone" to stressors that
are "outside the range of usual human experience" that are markedly distressing to almost everyone". This marks PTSD off as a
disorder in which the stressor would have to be at least "catastrophic" on Axis IV of DSM-III. It is not clear that this continues or alters
the notion that stress is dimensional, and does not deal with the problem that the stressor is exceptionally difficult to define in
objective terms (Feinsten & Dolan.1991).
Secondly, re-experiencing of the trauma (Criterion B) needs now to be "persistent", with the inclusion of intense distress at exposure to
events that may symbolize the original trauma. Thirdly, numbing of responsiveness, Criterion C in DSM-III, is replaced by persistent
avoidance, Criterion C in DSM-III(R), with a much expanded range of symptoms. Fourthly, there is the replacement of the range of
allowable symptoms in DSM-III, Criterion D, with the notion of persistent increased arousal in DSM-III(R), Criterion D. DSM-III(R) adds a
fifth criterion, Criterion E, which includes a time course of at least one month's duration, as well as allowing for a delayed course. These
changes together indicate a much clearer definition and a disorder that is clearly more severe than had been previously assumed.
Accompanying these classificatory changes has been the validation of the disorder by epidemiological and clinical research. This work
has been recently reviewed by McNally, who argues out the studies indicate that PTSD forms a coherent syndrome (McNally.1993).
There is good reliability in diagnosis, and reliability coefficients range from 0.58 for unstructured interviews to 0.86 for research studies
using trained interviewers. Reliability is thus good, but there can still be questions about validity.
Prevalence is rather variable, and has been assessed in two ways: one approach has been to examine prevalence in the general
population, whilst the other has been to examine prevalence in "at risk" groups. The Epidemiological Catchment Area(ECA) Survey
estimated the lifetime prevalence of PTSD at about 1.3% in the general population, and at 3.5% in persons exposed to civilan or
military violence, whilst a rate of 20% was found for veterans wounded in Vietnam (Helzer,Robins & McEvoy.1987). The clear
suggestion was that the rate was dependent upon actual exposure to violence, and this was bolstered by studies of at risk groups.
The National Vietnam Veterans Readjustment Study reported lifetime prevalence rates of 30.9% for males and 17.5% for females,
whilst, for those exposed to high war zone stress, the current prevalence rates were 38.5% and 17.5% for men and women respectively
(Kulka et al.1988). This again bolstered the notion that the stressor could be objectively determined, and that there was a continuum
along which stressors could be ordered, and according to which the severity and prevalence of PTSD could be predicted.
The degree of exposure to violence seems to affect the course of PTSD, and an American study on children attacked by a sniper at
school showed this clearly (Pynoos et al.1987). One month after the attack, 77% of the children who had been in the playground when
it was attacked showed PTSD, whilst 67% of the children who had been inside the school also showed PTSD. 14 months later, 74% of
the former group still showed PTSD, whilst only 19% of the latter were still suffering from the disorder.
There would therefore appear to be a dose-response effect due to the magnitude of the stressor, and, according to this argument,
torture will represent the most severe of all stressors, and the prevalence of PTSD should be highest in this population. This would
seem to support the dimensional conception of stress, but, as pointed out above, this argument is somewhat modified by a recent
empirical study, which demonstrated that, although PTSD was a probable result of physical injury, it was more a consequence of the
perception of injury than the actual severity of the injury itself (Feinstein & Dolan.1991).
Although there is little doubt that PTSD usually follows trauma, there has been some debate over the notion that there is a delayed
subtype. This argument has been partially resolved by a recent study from Israel, examining veterans who sought psychiatric help
between 6 months and 5 years after the 1982 Lebanon War (Solomon et al.1989). This study suggested that 40% of cases represented
delayed help-seeking, 33% was exacerbation of subclinical PTSD, 13% was reactivation of old PTSD, and 10% was delayed onset PTSD.
The debate is complicated even further by recent studies of Holocaust victims, which demonstrate clinical disturbance in
third-generation survivors (Hardi & Szilagyi.1993), but it is not clear that the disturbance measured in the grandchildren of
concentration camp victims can be described as PTSD. The issues around the course of the disorder and its long-tern effects are yet to
be clearly resolved.
All of this epidemiological work has been substantially supported by the more empirical work. Laboratory studies have shown PTSD
sufferers to have marked and significant response to noxious stimuli, with sufferers showing both psychological distress and
physiological reactivity to battle sounds, war imagery and the like, and marked avoidance of these stimuli: 80% of PTSD sufferers in one
study terminated exposure to audiotaped combat stimuli as opposed to 0% healthy combat veterans (Malloy et al.1983).
Studies of sleep show traumatic dreams occurring in both REM and non-Rem sleep, as well as difficulties in both falling asleep and
maintaining sleep. In general, the sleep studies show a wide range of differences between PTSD sufferers and other populations, both
civilian and military. Other studies, investigating intrusive cognitive activity, have shown marked effects of intrusive cognitions, with