Abstract
Acute stress disorder (ASD) describes acute trauma reactions that purportedly predict subsequent posttraumatic stress disorder (PTSD) [1]. The major difference between ASD and PTSD is the requirement of the ASD diagnosis that one must experience at least three dissociative symptoms. This requirement is based on the premise that dissociation at the time of trauma results in fragmented encoding of the event, which impedes subsequent emotional processing of the experience and purportedly leads to longer-term psychopathology [2]. Prospective studies indicate that approximately 80% of people who initially display ASD still suffer PTSD 6 months later [3–5]. The predictive value of ASD is questioned, however, by findings that significant proportions of people who do not meet the dissociative criterion of ASD also develop PTSD [4, 6]. Further, other studies indicate much lower rates of ASD following trauma and variable predictive relationships with PTSD [7]. Overall, the evidence for the proposed relationship between ASD and chronic PTSD is very mixed.
One factor that may explain the discrepant findings across studies is the role of gender. There is considerable evidence that women are significantly more likely than men to develop PTSD [8]. Moreover, recent evidence indicates that women report peritraumatic dissociation more often than men [9]. The greater prevalence of dissociation in females than males raises the possibility that the predictive power of ASD may be greater for females than males. That is, gender differences in the relationship between acute reactions and PTSD may, in part, explain the mixed evidence for the relationship between ASD and PTSD. Accordingly, this study, funded by the National Health and Medical Research Council, indexed the relationship between ASD diagnosis and PTSD in males and females, respectively.
Method
Patients
One hundred and seventy-one motor vehicle accident (MVA) patients consecutively admitted to hospital were assessed for ASD within one-month posttrauma and 134 (78% of initial sample) were reassessed at 6-months posttrauma. The final sample included 79 males and 55 females. There were no significant differences between the proportions of males and females who were drivers (male = 43%, female = 54%), passengers (10%, 30%), motorbike riders (30%, 5%), pedestrians (11%, 7%), and cyclists (5%, 2%), although there was a trend for more females to be passengers than males, χ2 (1, n = 25) = 3.21, p =.08. Those who participated in the 6-month assessment did not differ from non-participants in terms of age, ASD diagnosis, length of hospital stay, injury severity score, or trauma-ASD assessment interval. Table 1 presents the patient characteristics of the final sample. Males and females did not differ in terms of age, trauma-ASD assessment interval, length of hospital admission, patient's rated severity of trauma, or presence of psychiatric history. More males reported prior trauma than females, and males had higher injury severity scores than females.
Characteristics of patients assessed for acute stress disorder (ASD) and posttraumatic stress disorder
Procedure
Both assessments were carried out by clinical psychologists with more than 5 years experience in assessing traumatized individuals. After describing the study, written informed consent was obtained. Diagnosis of ASD was made using the Acute Stress Disorder Interview (ASDI) [11]. The ASDI is a 19-item structured clinical interview that has sound test–retest reliability (0.88), sensitivity (91%) and specificity (93%) relative to independent clinical diagnosis. Dissociative amnesia was excluded as a possible ASD symptom because of the overlap between dissociative and organic amnesia following brain injury, which is common in MVAs. Severity of injury was assessed using the Abbreviated Injury Scale, which provides a composite score reflecting location and severity of physical injury [12]. The 6-month assessment involved the administration of the PTSD module from the Composite International Diagnostic Interview (CIDI) [13].
Results
Incidence
In terms of the initial sample, full criteria for ASD were met by fewer males (8%, n = 9) than females (23%, n = 14) [χ2 (n = 171, df = 1, with Yates correction) = 6.96, p < 0.01]. Furthermore, females had higher Beck Depression Inventory scores at the initial assessment than males, t (133) = 2.22, p <.05. Posttraumatic Stress Disorder was diagnosed in fewer males (15%, n = 12) than females (38%, n = 21) [χ2 (n = 134, df = 1, with Yates correction) = 9.23, p < 0.002] at 6 months posttrauma.
Relationship between ASD and PTSD
In terms of patients who completed both assessments, 57% (n = 4) of males and 92% (n = 12) of females with a diagnosis of ASD met criteria for PTSD at follow-up [χ2 (n = 20, df = 1, with Yates correction) = 3.51, p < 0.06]. In terms of those who were not diagnosed with ASD, 11% (n = 8) of males and 21% (n = 9) of females met criteria for PTSD [χ2 (n = 114, df = 1, with Yates correction) = 2.22, NS].
Table 2 presents the proportion of males and females who reported each ASD symptom and passed criterion for each symptom cluster. χ2 analyses were conducted with a Bonferonni adjustment of p <.005. More females than males reported fear of the event; reduced awareness; derealization; recurrent images; avoidance of conversations, places, and feelings; and heightened startle response than males. Further, more female than males passed the criterion for Criterion B (dissociation).
Proportion of patients reporting symptoms and predictive power of acute stress disorder (ASD) symptoms
Table 1 also presents the positive and negative predictive power of each ASD symptom for males and females on the basis of diagnostic status at 6-months posttrauma. Positive predictive power was defined as the probability of PTSD developing when an ASD symptom is present. This probability was calculated by dividing the number of participants who reported each ASD symptom and who later developed PTSD by the total number of people who reported each ASD symptom. Negative predictive power was defined as the probability of not developing PTSD when an ASD symptom is absent. This was calculated by dividing the number of participants who did not report each ASD symptom and who later did not develop PTSD by the total number of those who did not report the symptom. Females displayed stronger positive predictive power than males for all symptoms and clusters. With the exception of Criterion C (Re-experiencing), males displayed stronger negative predictive power than females for all symptoms and clusters.
Discussion
The finding that 93% of females who initially met criteria for ASD subsequently had PTSD 6 months posttrauma compared to only 57% of males supports the prediction that the diagnosis of ASD is a more accurate predictor of PTSD for females than males. It appears that the reason for ASD being a stronger predictor in females is that they displayed stronger dissociative reactions, and accordingly more females satisfied the criteria for ASD. The current pattern replicates recent findings of higher prevalence of posttraumatic dissociation [9] and PTSD [8] following MVAs in females than males
The positive predictive power of ASD symptoms and clusters was higher for females. The negative predictive power of ASD symptoms and clusters was higher for males. This indicates that the ASD criteria performed better predicting the presence of PTSD in females and the absence of PTSD in males. The greater prevalence of dissociation and the increased positive predictive power of ASD for females cannot be attributed to prior psychiatric history, trauma severity, age, or prior trauma (because males had more prior trauma and were more severely injured than females).
There are several possible explanations for this pattern. First, it is possible that differential response biases exist in symptom reporting. Cultural or societal factors may result in male patients underreporting symptoms, and this may contribute to relatively higher endorsement of symptoms during clinical interview by women [14]. Second, the elevated risk factor in females may be associated with neurobiological factors [9]. This proposal is supported by findings that females respond more effectively to sertraline than males [15]. Third, it is possible that there is an increased risk factor for PTSD reactions secondary to childhood trauma [16]. This is consistent with evidence of neurological development being affected in female survivors of childhood abuse [17]. Fourth, the greater incidence of depression immediately after the MVA in females may be associated with increased risk for PTSD. Fifth, the greater proportion of females as passengers may have led to greater incidence of PTSD in people who had less responsibility or control in accident. This possibility is consistent with evidence that people who attribute responsibility to another person in a MVA are more likely to develop PTSD [18].
We recognize that predictive analyses are influenced by the prevalence rates of the sample, and it is possible that the lower proportion of females may have influenced the predictive power of symptoms between genders. Further, since we conducted the assessments bias may have influenced the results. Nonetheless, these data may partially explain the mixed findings concerning the ability of ASD and peritraumatic dissociation to predict PTSD [19].