Criticism has mounted in recent years around posttraumatic stress disorder’s (PTSD) diagnostic Criterion A. This criterion in the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV-TR; American Psychiatric Association 2000) explicitly defines what constitutes a “traumatic event”. Some researchers have argued that PTSD’s stressor criterion (Criterion A) is too lenient in defining “trauma”, whereas others have argued that the definition is not broad enough. This paper examines the background and current status of research surrounding PTSD’s Criterion A1 (a subcriterion of Criterion A), as well as discusses the importance and implications of the stressor criterion in legal realms.

History of PTSD’s Criterion A

Since PTSD’s first inclusion in the DSM-III (American Psychiatric Association 1980), each subsequent DSM edition has seen the potential list of Criterion A traumatic events grow because this criterion has become progressively more lenient over time. In fact, currently in DSM-IV and IV-TR (American Psychiatric Association 1994, 2000), the traumatic stressor criterion can be satisfied based on an indirectly experienced trauma (e.g., by witnessing or learning about a trauma occurring to someone other than oneself). This broadening of the stressor definition has resulted in criticism from some experts that it has become too lenient, a state of affairs they refer to as the “conceptual bracket creep” for the PTSD diagnosis (Elhai et al. 2005b; Frueh et al. 2004; McNally 2003; Mikkelsen and Einarsen 2002). However, other researchers have argued that Criterion A should be expanded to include less severe, but still serious, life events such as chronic illness, childbirth complications, sexual harassment, or bullying because they are empirically associated with PTSD symptom endorsement (Avina and O’Donohue 2002; Fontana et al. 2000; Lev-Wiesel et al. 2006; McDermut et al. 2000; Olde et al. 2006; Palmieri and Fitzgerald 2005). We now discuss the stressor criterion since its induction in the DSM and the changes it has undergone over time in order to prepare a more detailed look at the controversy.

PTSD first appeared as a mental health diagnosis in 1980 in the third edition of the DSM (DSM-III; American Psychiatric Association 1980). Its inclusion in the DSM-III was an important step in the mental health field on a number of levels. First, it was in this edition of the DSM that a stressful life event first took on an etiological role in determining a mental health diagnosis. The perception that PTSD’s etiology stemmed from an event outside the individual, rather than from an intrinsic weakness or faulty character trait, was a significant change in the mental health sociopolitical climate (Brunner 2002). It also gave weight to the emerging biopsychosocial notion that the individual and environment in which s/he lives interact and impact on one another (Van der Kolk 1996). Second, the emotional effects of traumatic event exposure were given a name and diagnostic criteria, thus paving the way for empirical evaluation. Finally, the creation of the PTSD diagnosis recognized and substantiated the claims of many individuals who reported suffering the long-term effects of trauma exposure (Andreasen 2004; Lasiuk and Hegadoren 2006a).

PTSD’s Criterion A has been referred to as the “gatekeeper” of the PTSD diagnosis (Davidson and Foa 1991) because a PTSD diagnosis cannot be assigned unless the stressor criterion is met. Criterion A was defined in the DSM-III as “a recognizable stressor that would evoke significant symptoms of distress in almost everyone” (American Psychiatric Association 1980, p. 236). This definition included events such as war, torture, sexual assault, and natural disasters. As such, these events were considered to be distinctly different from more common life stressors such as divorce, bereavement, chronic illness, or job and/or financial loss. Adverse emotional responses to such non-Criterion A events would, in DSM-III terms, be characterized as an adjustment disorder rather than PTSD (American Psychiatric Association 1980).

Due to criticism that the DSM-III stressor criterion lacked specificity and referred primarily to stressors of great magnitude or severity, DSM-III-R (American Psychiatric Association 1987) attempted to better define the quality of the stressor (Weathers and Keane 2007). The DSM-III-R retained the core elements of the previous version’s stressor criterion while adding a list of examples of traumatic events within the stressor criterion itself (American Psychiatric Association 1987). The list provided a more qualitative description of stressors that would meet the definition of a traumatic event than what had been previously described in the previous DSM edition. The type of traumatic events meeting Criterion A was also broadened to include events involving indirect exposure, such as learning about serious threat or harm to a close friend or relative (American Psychiatric Association 1987).

However, with publication of DSM-III-R, critics argued that some traumatic experiences such as sexual assault and motor vehicle accidents were not uncommon or “outside the range of human experience” in the general population. Emerging epidemiological research at that time supported this argument and demonstrated that traumatic event exposure was more prevalent than had been previously believed. For example, in a representative sample of 5,877 individuals from the US general population, lifetime traumatic event exposure rates of 61% in men and 51% in women were found (Kessler et al. 1995). Further, exposure rates of rape/molestation in men and women (4% and 22%, respectively) and life-threatening accidents (25% and 14%, respectively) reinforced the idea that even severe traumatic events that can involve threat to life and/or violence were not uncommon (Kessler et al. 1995).

DSM-IV’s (American Psychiatric Association 1994) Criterion A was expanded in an attempt to provide broader coverage of traumatic events and an explicit definition of the range of events believed to result in PTSD (Kilpatrick et al. 1998). Criterion A now consists of two parts. First, Criterion A1 is defined as events in which the person “experienced, witnessed, or was confronted with an event or events that involved actual or threatened death or serious injury, or the threat to physical integrity of self or others” or “learning about the unexpected or violent death, serious harm, or threat of death or injury experienced by a family member or other close associate” (American Psychiatric Association 2001, pp. 427–428). In the DSM-IV’s text, there is also a longer list of qualifying traumatic events, with some critics indicating that such additions have been responsible for the “bracket creep” of the PTSD diagnosis (Weathers and Keane 2007). Second, Criterion A2 requires that the person also experienced “intense fear, helplessness, or horror” at the time of the event (American Psychiatric Association 1994, p. 428). The text also includes information about differential diagnosis, in particular, that the diagnosis of adjustment disorder should be applied when an individual presents with symptoms of PTSD resulting from a lower magnitude stressor than those qualified in the DSM-IV’s PTSD text (American Psychiatric Association 1994).

PTSD’s Criterion A1 Controversy

Since PTSD first appeared in DSM-III, controversy has surrounded how to operationally define Criterion A (or more appropriately, Criterion A1 in DSM-IV terms) and the appropriateness of linking PTSD with only a distinct class of stressors (Gold et al. 2005). Prior to the formal induction of PTSD in the DSM-III, the prevailing view in the first two editions of the DSM was that a stressful event could have a modifying effect on mental health in that it could worsen or increase its length, but the direct effect was transient or brief (i.e., as found in DSM-I’s gross stress reaction and DSM-II’s transient situational disturbance; O’Brien 1998). This view changed when a stressor took on an etiological role in the DSM-III and resulted in debate of the importance and meaning of the stressor criterion (Lasiuk and Hegadoren 2006b). The debate raised questions regarding what differentiates a traumatic stressor that causes PTSD from a stressor that causes other stress-related emotional problems, as well as the exact nature of traumatic stressors (McNally 2004; O’Brien 1998).

As previously mentioned, an initial focus of debate and research about PTSD involved determining the accuracy of describing the events within the stressor criterion as being outside the range of usual experience (O’Brien 1998). Experts also argued whether PTSD was a normative reaction to severe and uncommon traumatic events or a specific type of reaction to common stressful events (Breslau and Davis 1987b; Yehuda 2004). In a prominent and controversial article, Breslau and Davis (1987b) stated that there was inadequate research to support the claim that PTSD symptoms were associated with a unique set of extreme stressors. A number of prominent experts in the field wrote to the journal’s editor to counter the authors’ statements with their own statements of support for the stressor criterion (Breslau and Davis 1987a; Escobar 1987; Horowitz et al. 1987; Lindy et al. 1987; Ursano 1987).

Complicating the stressor-PTSD association is research evidencing a complex association of findings regarding individual differences and vulnerabilities related to traumatic event exposure and the development of PTSD (McNally 2003; O’Brien 1998). Studies indicating discrepancies between traumatic event exposure and PTSD’s prevalence have raised issues regarding the role of the stressor in the development of the disorder. In a large epidemiological study on PTSD, Kessler et al. (1995) reported lifetime PTSD prevalence rates of 5% and 10% in men and women, respectively, despite 61% of men and 51% of women participants reporting exposure to at least one traumatic event in their lifetime. Thus, a large proportion of individuals who are exposed to a Criterion A1 event do not develop PTSD as a result (Breslau 2002). Even the most devastating stressors do not result in symptoms for all survivors (Kessler et al. 1995; Sutker et al. 1991). There are findings, however, that those who do not develop PTSD after exposure may develop other mental health related symptoms and/or diagnoses (O’Brien 1998).

The importance and controversy of the stressor criterion is only heightened by findings demonstrating high rates of diagnostic comorbidity between PTSD and other mental disorders (Franklin and Zimmerman 2001; Kessler et al. 1995; Norris 1992; O’Brien 1998), even after removing symptoms that overlap between PTSD and these disorders (Elhai et al. 2008). Some factor analyses of PTSD’s symptom structure have indicated a large component of PTSD that may be best conceptualized as general distress and anxiety, which is characteristic of other anxiety and mood disorders (Simms et al. 2002). Thus, the stressor criterion is primarily what makes the otherwise non-unique diagnostic criteria specific to PTSD, given that all but the reexperiencing cluster of symptoms are shared with other diagnoses (Simms et al. 2002). However, other factor analytic work has revealed other models that best represent PTSD’s latent structure (discussed further below; McWilliams et al. 2005).

Another point of contention raised has been that equal weighting of a large variety of traumatic event types in a single diagnosis, from witnessing the death of another person to long-term internment in a concentration camp, results in a muddled measurement of traumatic event exposure (Ehrenreich 2003). Some traumatic stress researchers indicate that using a single diagnostic label for responses to such a diversity of stressful events results in a “watered down” stressor criterion (Yehuda 2004). Others argue, however, that because PTSD reactions occur in response to a wide range of events, the stressor criterion should be abolished or widened to include less severe events (Amir et al. 1996; Avina and O’Donohue 2002; Fontana et al. 2000; Lev-Wiesel et al. 2006; McDermut et al. 2000; Mol et al. 2005; Olde et al. 2006; Palmieri and Fitzgerald 2005). Experts siding with widening the stressor criterion first cited Horowitz’s early work as evidence that PTSD symptoms could develop from less stressful events (Horowitz 1975; Horowitz and Wilner 1976; Sutker et al. 1991). In that research, the lines between PTSD and other mental disorders were shown to be blurred, and the PTSD diagnosis was questioned when PTSD’s only distinguishing feature, reexperiencing symptoms, was suggested by the authors to also occur after less stressful events such as exposure to sad or gruesome films (Horowitz 1975; Horowitz and Wilner 1976; Sutker et al. 1991).

With continued controversy surrounding Criterion A1’s definition, proponents of broadening the stressor criterion argue that doing so increases the political/social acknowledgement of the relationship between different types of traumatic event exposure and PTSD symptomatology (Avina and O’Donohue 2002). It is argued that such acknowledgement will result in increased resources for victims, including financial remuneration for their loss and suffering and treatment (Avina and O’Donohue 2002; O’Brien 1998). Other experts in the traumatic stress field argue that the broadening of Criterion A1 results in a more heterogeneous group of people diagnosed with PTSD, many of who probably should not qualify for a PTSD diagnosis. In fact, Breslau and Kessler (2001) found that DSM-IV’s broader stressor definition resulted in reported trauma exposure rates increasing by 59% since the DSM-III-R. The added events in the stressor criterion were reportedly responsible for 38% of PTSD diagnoses (Breslau and Kessler 2001). Experts against such a broadening argue that such diversity in traumatic events complicates epidemiological research and studies attempting to identify underlying mechanisms of PTSD, as well as trivializes the impact of severe trauma exposure, and limits resources for those who experience truly and severely traumatic events (Elhai et al. 2005b; McNally 2004). There is also the question that it has led to an explosion of legal claims, including for frivolous reasons or complaints of suspect validity (see below).

Current Status of the Research on Criterion A1

Despite the ongoing controversy regarding PTSD’s Criterion A1, there is little empirical research exploring whether events meeting its current definition are associated with a different PTSD symptom pattern or diagnostic prevalence than non-Criterion A1 events. There is a paucity of studies examining a wide range of events that contribute to PTSD, yet studying a range of events may shed light on a universal application of PTSD (Davidson and Foa 1991). The range of events studied in the past has been limited because most studies investigated only events already defined by DSM’s Criterion A (McFarlane 1991). Several experts suggested that the lack of stressor research is due to the focus on either single catastrophes, life stressors, or on individual factors (Amir et al. 1996; Breslau 1990). A few recent studies have found that, in contrast to Criterion A1 events, non-Criterion A1 events result in similar or higher rates of PTSD diagnoses and severity. Given their import to the relevance of the A1 criterion, these studies will be discussed next.

A DSM field trial consisting of treatment-seeking and community sample participants was conducted prior to the publication of the DSM-IV (American Psychiatric Association 1994). As part of testing the most empirically valid definition of PTSD, one goal of the field trial was to examine the adequacy of the stressor criterion (Kilpatrick et al. 1998). The authors’ proposed expanded definition for the stressor criterion was broadened further to include the sudden death or illness of someone close to the person, witnessing of death, threat to personal integrity, and the subjective component (i.e., that which is now Criterion A2; Kilpatrick et al. 1998). The authors assessed a total of five definitions of Criterion A, including a liberal definition that incorporated non-Criterion A1 events. They reported that there was only a slight variability in the prevalence of PTSD based on how liberal the definitions were and that few people appeared to develop PTSD unless they had been exposed to traumatic events (Kilpatrick et al. 1998). Proponents of the new, expanded definition of Criterion A argued that these findings supported the broadening of the definition, but that there was not a basis for expanding to a stressor criterion which included non-Criterion A1 events (Avina and O’Donohue 2002). The authors pointed out weaknesses in the study, however, including the very small portion of the sample that reported only having non-Criterion A1 events and the retrospective nature of the self-reports (Kilpatrick et al. 1998).

Gold et al. (2005) examined 430 college students for differences in PTSD between those endorsing Criterion A1 events and stressful, but non-Criterion A1 events. Those with non-Criterion A1 events reported, on average, less exposure to traumatic events, higher rates of PTSD according to the Posttraumatic Stress Diagnostic Scale (Foa et al. 1997), higher severity of total PTSD symptoms, higher scores on the reexperiencing subscale, and similar rates of global distress using the Brief Symptom Inventory (Derogatis and Melisaratos 1983).

Mol et al. (2005) randomly surveyed 2,997 patients from a list of 67,500 patients of 12 general medical practices in the Netherlands. Of the 1,498 patients who completed the surveys, 832 were used in the subsequent analyses. The authors did not find a significant difference between Criterion A1 and non-Criterion A1 groups on PTSD scores until time since the event was controlled for in post hoc analyses. When time since the event was controlled for, the non-Criterion A group evidenced higher average total PTSD Symptom Scale (Foa et al. 1993) scores for events that occurred in the past 30 years.

In an article by Shapinsky et al. (2005), college students were instructed to rate PTSD symptoms based on a college exam as the index event. A large proportion of participants yielded PTSD scores that were above conservative cut-off scores on PTSD measures, that is, obtained for 10% of participants on each of the Impact of Event Scale-Revised (Weiss and Marmar 1996), PTSD Checklist (Weathers et al. 1994), and Revised Civilian Mississippi PTSD Scale (Vreven et al. 1995). However, the analyses that were undertaken did not control for history of exposure to trauma or other recent stressful life events. Moreover, a formal assessment of traumatic event exposure with a valid and reliable measure is important when attempting to empirically analyze the relationship between events and psychological variables (Netland 2005). Without controlling for trauma history, it is impossible to conclude that PTSD resulting from non-Criterion A1 events was not a result of previous trauma exposure.

Finally, Long et al. (2008) reported that in a sample of 119 college students, after statistically controlling for traumatic event history, non-Criterion A1 events were associated with a greater proportion of “probable” PTSD diagnoses (based on PTSD Symptom Scale cut-off scores; Foa et al. 1993) and symptom frequency than Criterion A1 events. The authors found, however, that the PTSD symptom frequency differences were moderated by the order in which the measures were administered. Specifically, using a mixed (within-subjects and between groups) design, PTSD scores from non-Criterion A1 events were higher only when participants were asked about PTSD from non-Criterion A1 events before queried about PTSD from Criterion A1 events. Similar patterns of differences in PTSD scores between the stressor types were also found across the three PTSD symptom criteria (Long et al. 2008).

Recent Findings for Criterion A1: Possible Explanations and Solutions

Recent evidence finds that non-Criterion A1 events can result in levels of PTSD severity and diagnostic prevalence on par with or even higher than Criterion A1 events. Some of this research possesses methodological flaws—not controlling for prior traumatic event exposure, failing to inquire about Criterion A1 and non-Criterion A1 events with behaviorally specific measures, failing to inquire about PTSD from both Criterion A1 and non-Criterion A1 events, failing to explicitly link PTSD symptom ratings to an index Criterion A1 and/or non-Criterion A1 event, and neglecting to counterbalance the order of measure administration. Despite these limitations, the studies’ findings challenge whether DSM-IV PTSD’s list of qualifying traumatic stressors is an accurate list of stressors most relevant for the PTSD diagnosis. We now discuss possible explanations for these studies’ findings and possible solutions for lower magnitude stressors being just as (or more) associated with PTSD than high magnitude stressors. These issues are discussed in greater detail in Long et al. (2008).

Altering PTSD’s Criterion A1

The first possible explanation for the greater PTSD prevalence from lower magnitude stressors is that PTSD’s Criterion A1 is flawed. One possible solution, therefore, is to modify Criterion A1. It has been argued that because PTSD reactions have occurred in response to a wide range of events, the stressor criterion should be widened to include these other events (Avina and O’Donohue 2002; Dansky and Kilpatrick 1997; Fitzgerald et al. 1997; Fontana et al. 2000; Fontana and Rosenheck 1998; Lev-Wiesel et al. 2006; McDermut et al. 2000; Mol et al. 2005; Olde et al. 2006; Palmieri and Fitzgerald 2005; Wolfe and Keane 1990). Proponents of broadening the stressor criterion argue that doing so increases the political/social acknowledgement of the relationship between different types of traumatic event exposure and PTSD symptomatology (Avina and O’Donohue 2002). Further, such acknowledgement could result in better resources for trauma victims (e.g., compensation, treatment; Avina and O’Donohue 2002). Carried to the extreme, one could argue that Criterion A1 is unnecessary and should be removed entirely if researchers continue to find that a large number of non-traumatic events are associated with PTSD severity and diagnoses (Gold et al. 2005; Mol et al. 2005; Shapinsky et al. 2005).

Other experts in the traumatic stress field, however, argue that there is already heterogeneity among the PTSD symptoms that result from trauma exposure. They contend that if Criterion A1 is broadened or removed, a greater diversity of posttraumatic experiences will result, and an accurate PTSD symptom pattern will be even more difficult to uncover (McNally 2003; Simms et al. 2002; Watson 2005). These experts have pointed to Breslau and Kessler’s (2001) study that, unlike Kilpatrick et al. (1998), clearly evidenced that broadening Criterion A1 resulted in a higher PTSD prevalence rate. As stated earlier, experts against broadening Criterion A1 also argue that excessive diversity in traumatic events may result in difficulties in identifying the underlying mechanisms of PTSD’s development from such diverse stressors; however, extensive factor analytic work has yet to be conducted on whether different trauma-exposed groups yield non-invariant (i.e., diverse) PTSD factor solutions. Additionally, such broadening could result in limitations in resources for those who experience truly severe traumatic events, medicalization of normal distress to stressful life events, and increases in PTSD-related litigation (Elhai et al. 2005b; McNally 2004). Finally, the circular contention that if an individual has the symptoms of PTSD, s/he must meet Criterion A1 has been argued as logically unsound (Elhai et al. 2005b; McNally 2003).

In order to advance the understanding of PTSD’s etiology and conceptualization, an empirical method of defining the specific events linked to PTSD is warranted (Weathers and Keane 2007). There is a dearth of studies that include a wide range of events that contribute to PTSD, despite experts’ indications that studying a range of events may improve application of the disorder (Davidson and Foa 1991; Weathers and Keane 2007). A difficulty with broadening the stressor criterion would be assembling an empirically valid list of life stressor events that result in PTSD symptoms. Based on the wide variety of non-Criterion A1 events already reported as resulting in PTSD, it would be difficult to know which events to include.

Altering PTSD’s Diagnosis or Nosology Within DSM

A competing explanation for the finding that non-Criterion A1 events are associated with greater PTSD estimates than Criterion A1 events is that there is a problem with PTSD’s symptom criteria in tapping emotional reactions from such events. One possible solution would be to recognize and utilize the diagnosis of adjustment disorder that DSM-IV-TR indicates is to be used in instances such as these (American Psychiatric Association 2000). Weathers and Keane (2007) noted that DSM-IV-TR already specifies in its text that, for those individuals meeting all criteria for PTSD except Criterion A1, the most appropriate diagnosis would be adjustment disorder. A positive aspect of this solution is its parsimony; no diagnostic changes would be necessary. This retention of the status quo would be advantageous because diagnostic changes are often complicated by multiple other implications, such as the need for epidemiological re-assessment of the disorder, changes in psychological measures and treatments, dissemination of information regarding the changes, and general shifts in mindset regarding the disorder. A disadvantage of using the current adjustment disorder diagnosis is that it is only appropriate for individuals whose symptoms have an onset within 3 months of the stressor’s onset, and symptoms must resolve within 6 months of the stressor’s termination (American Psychiatric Association 2000). Alternatively, another diagnostic category could be created for individuals who have PTSD symptoms related to a stressful but not traumatic stressor. A separate category for PTSD symptoms related to bereavement has already been proposed (Gold et al. 2005; Prigerson et al. 1999). However, the disadvantages of implementing diagnostic changes would apply to this solution.

Another solution to the problem of an imperfect PTSD diagnosis could be to redefine the symptom criteria such that they would be better able to distinguish between events of lower and higher magnitude. PTSD’s symptom structure has been the subject of debate in recent years (McWilliams et al. 2005; Simms et al. 2002; Watson 2005). One reason for the debate is that PTSD’s three-factor structure as defined by DSM-IV-TR has not received much empirical support (McWilliams et al. 2005; Simms et al. 2002; Watson 2005). There has been argument for a four-factor structure, similar to the DSM-IV-TR’s three-factor structure, by separating PTSD’s avoidance and emotional numbing into separate factors (for a review, see Asmundson et al. 2004). However, newer studies argue and find empirical support for a four-factor structure that includes a strong dysphoria component (McWilliams et al. 2005; Simms et al. 2002). Thus, there is currently no clear consensus as to the optimal PTSD factor structure, though one of the two competing four-factor solutions appears best. Other problems with specific PTSD symptoms not loading well in factor solutions (emotional numbing symptoms, traumatic amnesia) have also been reported.

In addition to conflicting factor structure results for PTSD, the disorder is highly comorbid with other mental disorders (Franklin and Zimmerman 2001). PTSD is particularly comorbid with a number of anxiety and mood disorders, with rates of co-occurring PTSD and major depressive disorder, for example, ranging from 26% to 53% across studies (Franklin and Zimmerman 2001). One of the possible causes of such comorbidity is symptom overlap with other anxiety and mood disorders (Kessler et al. 1995; Simms et al. 2002; Watson 2005), although symptom overlap has not been found to be a primary reason for comorbidity (Elhai et al. 2008; Franklin and Zimmerman 2001). These shared symptoms have been referred to as a general distress component, the presence of which has been supported in a number of empirical studies (McWilliams et al. 2005; Shapinsky et al. 2005; Simms et al. 2002; Watson 2005).

Another potential solution for redefining PTSD’s symptom pattern and/or structure and distinguishing traumatic events from those that are less severe is placing increased importance or weight on the reexperiencing criteria. Watson (2005) proposed an integrated hierarchical model in which each of the anxiety and mood disorders that has the shared dysphoria component also appears to evidence a specific characteristic that makes it unique. At this time, there is some empirical support for PTSD’s reexperiencing symptoms being unique to PTSD (Simms et al. 2002). If the reexperiencing symptoms were weighted more heavily in the PTSD diagnosis, especially because there are fewer of them than dysphoria symptoms, individuals with PTSD may present differently than those with general depression and anxiety symptoms stemming from other causes, thus helping with differential diagnosis and subsequent provision of resources. The general distress symptoms would still play a role in the overall symptom presentation and would be useful for prevention and treatment. The main advantage of this solution of differentially weighting symptoms is that it could improve PTSD’s symptom structure and thus diagnostic clarity. An argument against the removal or lower weighting of the general distress symptoms is that, as previously noted, several studies have found that removal of PTSD’s overlapping symptoms does not appreciably reduce PTSD diagnostic rates or comorbidity (Elhai et al. 2008; Franklin and Zimmerman 2001). Arguments such as this support the need for increased study of more of models that redefine PTSD’s symptom structure.

Identifying PTSD Measurement Issues

Comparable PTSD rates between Criterion A1 and non-Criterion A1 events may not be solely due to problems with Criterion A1 or with PTSD’s symptom criteria, but rather with the measurement of PTSD. That is, it is possible that respondents have difficulty sequestering their ratings of PTSD to a single traumatic event. Or, perhaps respondents have difficulty disaggregating their PTSD ratings to a Criterion A1 event from PTSD ratings to a non-Criterion A1 event. Empirical investigation is needed to increase general knowledge of the influence of PTSD measurement issues. One obvious benefit of examining such measurement issues would be that empirical investigations might uncover influences on participant responses that lead to unstable or unreliable results. Investigations may also be helpful in uncovering issues related to repeated measures designs, as well as to the ability of participants to separate the influence of having PTSD symptoms to multiple stressors (i.e., multiple types of discrete stressors or multiple incidents of a specific type of stressor) when responding to a single event. A disadvantage of focusing on measurement issues to the exclusion of diagnostic changes is that there is a growing body of research indicating that there are problems with the PTSD diagnosis that cannot be ignored and cannot be solely attributed to measurement issues (McHugh and Treisman 2007; Watson 2005).

Subjective Appraisal

A fourth explanation for the findings that non-Criterion A1 events are associated with comparable or greater PTSD severity and diagnoses than Criterion A1 events is that the perception of an experience as stressful or traumatic may have a stronger association with the development of PTSD symptoms than the objective nature of the experience. This perspective of including subjective appraisal in the assessment of what is considered a stressor is not new to the stress field; in fact, it represents Criterion A2 of PTSD. According to Lazarus and Opton (1966), the stressful event cannot be taken out of the context of subjective appraisal, and stress is a result of the perceived imbalance between demands and resources (Lazarus and Opton 1966; Li and Olsen 2005; Sutker et al. 1991). Likewise, Spielberger (1972) emphasized that the emotional state of anxiety develops from the perception of threat of physical or psychological danger along with associated activations of the autonomic nervous system. Individuals higher in trait anxiety have a propensity to perceive a wider variety of situations as threatening and tend to have more severe reactions to anxiety-producing situations (Spielberger et al. 1995). It is possible that some respondents perceive the most distressing non-Criterion A1 events as more threatening currently than the Criterion A1 events, perhaps because the non-Criterion A1 events may be more salient in the lives of such respondents. Further empirical investigation of the role of trait anxiety when rating responses to an objective event is warranted, including examination of the differential effects of trait anxiety on both PTSD’s Criterion A2 and PTSD’s symptoms.

Retaining the Original Criterion A1 Definition’s Intent

Clearly, the original intent of DSM-III’s introduction of the PTSD diagnosis was to offer a disorder emanating from a severe traumatic stressor (Spitzer et al. 2007). Yet, given the discussion above, it is possible that Criterion A1 stressors are not, in fact, more associated with PTSD diagnoses and severity. Nonetheless, another option to the Criterion A1 conundrum is to keep with the original intent of the stressor criterion, despite empirical findings to the contrary. That is, perhaps Criterion A1 should be defined as it currently stands, despite evidence that some lower magnitude stressors may be just as likely to result in PTSD.

Resilience Factors

Finally, it should be emphasized that even in response to high magnitude events clearly meeting PTSD’s Criterion A1, not all exposed individuals develop PTSD. Individual differences in resilience to aversive events clearly play a prominent role in determining psychopathology from the events (for a review, see Bonanno 2004). Such individual differences could play a role in explaining why some non-Criterion A events result in comparable PTSD estimates to that yielded from Criterion A events.

Legal Implications of PTSD’s Criterion A1

The subtleties of PTSD’s Criterion A may have important implications in criminal and civil legal settings. These issues are discussed next.

PTSD claims may present in legal settings in a number of ways. For example, in criminal cases PTSD may be raised as the basis for an insanity defense, to acquit the defendant, or mitigate his/her sentence. In civil cases, PTSD claims may be filed subsequent to alleged exposure to a traumatic event in the workplace (worker’s compensation), alleged inability to work/function because of PTSD symptoms (disability determinations), or in a lawsuit in which the claimant proposes to have PTSD due to the actions of another person, persons, or organization. We next discuss the importance of corroborating Criterion A1 and the use of standardized, structured methods of assessment in legal settings.

Subjectivity of PTSD’s Criterion A Event Determination

Clearly, some events would be universally accepted by mental health professionals, including traumatic stress specialists, as meeting PTSD’s Criterion A1. For example, such events as forcible rape, combat or war zone exposure, and serious physical assault requiring medical treatment would be clearly agreed upon Criterion A1 events. However, courts are often faced with “gray area” stressors such as those discussed earlier in this paper (e.g., sexual harassment, bullying, childbirth complications) that may not have universal support in the mental health field for meeting PTSD’s Criterion A1. Thus, a certain amount of subjectivity is required in judging whether a stressor meets Criterion A1.

Given such subjectivity, it will often rest on mental health clinicians to use their own subjective clinical judgment to determine whether a stressor meets Criterion A1. And, given this subjectivity, there may be little convergence in judgment across clinicians regarding whether a stressor meets Criterion A1. The result is that for such gray area stressors, two (or more) well-intentioned and opposing experts, respectively testifying for the prosecution and defense, will take different positions regarding whether the alleged stressor meets Criterion A1. These different positions may be so at odds with each other that juries and judges may simply be confused about determining a Criterion A1 stressor for the PTSD diagnosis and may even doubt the mental health profession’s credibility for not having stricter standards.

Ideally, mental health professionals working in legal settings should be sensitive to the issues surrounding PTSD’s Criterion A controversy and should appropriately consider and discuss such issues in their testimony. Mental health experts should not take their position too seriously on whether a gray area event meets Criterion A1. Rather, they should simply acknowledge to themselves and the court that there is substantial subjectivity that can be involved in making such a determination. And, perhaps rather than assigning a diagnosis of PTSD when there is a gray area event, the examinee should be diagnosed as having “PTSD features” or “subclinical” or “subsyndromal” PTSD; in fact, subclinical PTSD is found to be quite debilitating in its own right (e.g., Grubaugh et al. 2005).

Assessment of PTSD’s Criterion A1 Events

To some extent, the subjectivity of Criterion A1 can be reduced with the aid of good practices in assessment. Research indicates that querying an individual about his/her history of traumatic events should not be open-ended and unstructured, but rather should be conducted using a specific list of potentially traumatic events. Without using such a list, some potentially important traumatic events may not be considered for inquiry. Such a list of traumas should use a standardized wording of its items (so that deviation in inquiries across different respondents is minimized), with as little ambiguity in terms as possible. That is, rather than inquiring about such vague terms as “sexual abuse” or “rape”, of which people often have different definitions, terms that are more behaviorally specific should be used instead. These issues are discussed in detail elsewhere (Frueh et al. 2004; Weathers and Keane 1999). Instruments assessing trauma history are reviewed in sources such as Elhai et al. (2005a) and Norris and Hamblen (2004).

Verifying PTSD’s Criterion A Events

There is reason to be concerned about individuals exaggerating or fabricating a traumatic experience in a high-stakes setting, such as in a court of law or examination for disability. To ensure that such an event actually occurred, PTSD’s Criterion A should be corroborated objectively. Such corroboration is not usually required in clinical treatment settings, where the patient is typically trusted that a stressful event in fact happened to him/her. A higher standard is required in legal settings, thus calling for objective corroboration of traumatic stressors; however, corroboration is not always possible. Traumatic events involving law enforcement typically, but not always, result in an arrest report, complaint, order of protection, or some related form of reporting. Traumatic events are not always reported to police for a variety of reasons (e.g., it may have been too stigmatizing for the victim to disclose; the perpetrator lives with the victim and does not feel safe to disclose). The absence of corroboration, therefore, does not necessarily indicate that the traumatic event did not happen.

Recently, Frueh et al. (2005) investigated the prevalence of exaggerated/fabricated combat exposure reports in a VA’s mental health clinic that treats PTSD. Through Freedom of Information Act requests, the authors requested the official military records of 100 consecutively presenting individuals (from 1997 to 1999) registered with VA, claiming to have served in combat during the Vietnam War. Some surprising findings emerged, especially that 32% of the sample had no evidence of combat exposure, 3% served in the military but not in Vietnam, and 2% had never served in the military (despite being registered with VA). Because 62% of the sample reported applying for VA disability compensation, financial incentives for appearing to suffer from PTSD could have played a role in combat exposure exaggeration or fabrication. The authors also discussed issues regarding limitations of records keeping in the military and thus difficulties in completing the corroboration of combat exposure. Nonetheless, these findings pose problems for PTSD claimants in legal cases who lack objective substantiation of Criterion A1 events they alleged to have experienced.

Linking PTSD Symptom Queries to a Single Criterion A Event

If PTSD’s Criterion A1 is satisfied and corroborated, in legal settings, it still must be supported as the cause of current PTSD symptoms. It is important to note that it is essential to the PTSD diagnosis that the inquiry of symptoms is specifically linked to a single, particular index traumatic event, as required by the DSM-IV. Without linking PTSD symptom queries to a specific index trauma, false positive endorsement of PTSD symptoms often results because some of PTSD’s symptoms can easily be attributed to other causes than a traumatic event (e.g., difficulty remembering the trauma can be attributed to age-related memory problems; sleep difficulty can be caused by other stressors or substance use; difficulty concentrating can be attributed to depression or restlessness; Elhai et al. 2005b; Long et al. 2008). The linking of PTSD symptoms to an index trauma is also essential given PTSD’s symptom overlap with other mood and anxiety mental disorders. Thus, a primary means of distinguishing these other disorders from PTSD is based on whether the symptoms are linked to a specific trauma (Elhai et al. 2008).

And in fact, for a diagnosis of PTSD, all of PTSD’s symptom criteria must be satisfied in reference to the same traumatic event; it is not permissible to satisfy one symptom criterion from one trauma and another criterion from another trauma for example. The ideal method of assessment (in the interest of time constraints) is therefore typically conducted by querying the subject about PTSD symptoms in relation to his/her most distressing trauma (if more than one trauma was endorsed; Briere 2004; Frueh et al. 2004). In legal cases, the index trauma most likely will be the event that the individual is alleging to have caused his/her PTSD (e.g., the trauma involved in a worker’s compensation case, VA disability).

Ruling out the Overreporting of PTSD

Finally, although this paper focuses on assessment of PTSD’s traumatic stressor criterion, it is also important to comment on the overreporting of PTSD symptoms. Overreporting of PTSD is a substantial problem in forensic practice (see review by Guriel and Fremouw 2003), whether it is intentionally done (i.e., malingered) or done unconsciously by the patient as a cry for help to play the sick role, etc. Moreover, it is possible that overreporting PTSD can also include fabrication or exaggeration of the traumatic event as part of the overreporting presentation. Therefore, it is important to rule out PTSD overreporting in the context of assessing PTSD’s Criterion A.

Conclusion

The ongoing controversy surrounding the PTSD’s stressor criterion (Criterion A1) definition has been explored throughout this paper. This issue of Criterion A1’s definition is very important and has many clinical and legal implications. Given the way in which the PTSD diagnosis is worded at this time, Criterion A1 is necessary to create the link between a stressor and symptom responses, and this link must be empirically supported for the validity of the diagnosis (Breslau 1990). The stressor criterion also adds uniqueness to the PTSD diagnosis, given that several of PTSD’s symptoms are found in other anxiety and mood disorders (Simms et al. 2002).

One of the main elements of the controversy is whether the stressor criterion should be expanded to include more life stressor events or, if in fact, it has already become too inclusive (O’Brien 1998). Some experts in the field of traumatic stress argue that Criterion A1 has progressively become too lenient for the PTSD diagnosis (Elhai et al. 2005b; Frueh et al. 2004; McNally 2003; Mikkelsen and Einarsen 2002). Other researchers argue, however, that Criterion A1 is not too lenient and should be expanded to include less severe, but still serious, life events (Avina and O’Donohue 2002; Lev-Wiesel et al. 2006; Olde et al. 2006).

Experts on each side of the controversy argue that the definition of Criterion A1 and its application impact on issues such as the identification of trauma victims, allocation of resources for victims, and clarification of trauma-related research (McNally 2004; O’Brien 1998). Because of its importance, we argue that until further research can clarify the objective definition of the events that should be included as meeting Criterion A1, assessors in the medico/legal field should strictly adhere to the current Criterion A1 definition and use behaviorally specific assessment measures (as described in this paper) to assess traumatic event exposure. Following the current definition closely and using appropriate assessment measures will improve chances that there will be a more homogenous population from which to conduct further research to clarify the current controversy, as well as better identification of trauma victims.

This issue’s controversial nature and subjectivity in opinion can been demonstrated by the fact that the present study’s authors disagree on it. One of the authors argues for narrowing Criterion A1 to exclude indirect exposure; the other author argues for broadening the criterion to include any stressor that the individual perceives as traumatic, provided that other PTSD symptom criteria are met in relation to that stressor. Both authors agree that the issue of Criterion A needs further empirical study, and both have been open to modifying their stance based on emerging empirical results. Future research should more closely examine PTSD symptom and diagnostic patterns across lenient and conservative Criterion A1 definitions of trauma exposure. Research should also examine differences in PTSD’s factor structure based on lenient and conservative Criterion A1 definitions.

Ultimately, given the problems inherent with PTSD’s Criterion A discussed above, it should be considered for restructuring in DSM-V. Perhaps Criterion A should not merely be considered one of PTSD’s several criteria because doing so can inadvertently and mistakenly send a message to clinicians that (1) Criterion A is as equally important as the symptom criteria (when in fact Criterion A is a crucial gatekeeper), and (2) the presence of a gray area stressor can be waived in as a definite traumatic event eligible for the PTSD diagnosis as long as other symptom criteria are met (when in fact an event that is not a definite trauma should not be waived in, despite meeting other PTSD criteria). Instead, Criterion A could be set up as an initial, required entry point for assessing the remaining diagnostic criteria, with more detail on assessing this entry point, as required (e.g., using objective and subjective assessment of the trauma, and also commenting on the extent of external corroboration of trauma exposure).