This study adds to the scientific literature examining stress response syndromes within the context of community disasters. Consistent with previous research (Bromet et al., 1998
; Kessler et al., 1995b
), we show that being 18–29, being female, experiencing more WTCD-related events, reporting low social support, and having low self-esteem were risk factors for the onset of Y1 PTSD, even after taking other stressful events into account. Two years after the attacks, however, younger age, gender, and social support were no longer related to PTSD. Instead, Latinos and respondents between 30 and 64 were at risk for PTSD, as well as those with low self-esteem, again controlling for other stressful events. Finally, our study documents a small but significant number of respondents who had increases in PTSD symptoms between 1 and 2 years post-WTCD, which also is consistent with previous research (Creamer et al., 2001
; Gray et al., 2004
; Orcutt et al., 2004
However, our study went beyond prior studies and explored some of the reasons for delayed and remitted PTSD. Although each of these two symptom trajectory groups represent only 3% of the sample, they show that changes both in social circumstances and psychological resources are potential explanations as to why some persons had their PTSD symptoms remit, while others experienced delayed PTSD. More specifically, individuals with delayed PTSD reported experiencing more negative life events postdisaster and had a marked decline in self-esteem, whereas remitters reported fewer negative events and showed an increase in self-esteem during this same period. PTSD symptoms and symptom severity show a similar pattern, with delayed PTSD respondents becoming more symptomatic and experiencing the symptoms more severely, whereas remitted PTSD cases show significant improvement in these areas. In multivariate comparisons with other groups, those most at risk for delayed PTSD were Latinos, those who experienced more negative life events, and those with low self-esteem. It is noteworthy that the delayed-onset group was more like the resilient group at Y1 and more like the acute group at Y2, while, again, the opposite was observed in the remission group. Thus, as has been previously noted (Gray et al., 2004
), changes in these diagnostic categories may not be due to minor fluctuations in PTSD symptomatology per se.
Even though DSM-IV included delayed PTSD, theoretical and conceptual work on it has lagged in the larger discussion of causes and consequences of PTSD. From a measurement error perspective, changes in diagnosis may be due to an underreporting of symptoms at the initial assessment or an overreporting of symptoms at later assessments for delayed PTSD and the opposite for remitted PTSD cases. It has also been speculated that delayed PTSD may result from classical conditioning for fear and anxiety responses to trauma cues, reinforcing avoidance and re-experiencing symptoms (Gray et al., 2004
). Finally, and in line with our findings, changes in diagnosis may reflect changes in trauma survivors’ psychosocial circumstances, whereby exposure to negative life events and/or events leading to lower self-esteem may result in increased PTSD symptom reporting.
Additional research on individuals who do not meet full PTSD criteria but nevertheless have many PTSD symptoms may also provide insight into the course of postdisaster stress response disorders. Referred to as partial (Breslau et al., 2004a
) or subsyndromal PTSD (Galea et al., 2003
), these classifications typically require individuals to have a certain number of symptoms from criteria B, C, and D (re-experiencing, avoidance, and arousal). These designations are not without controversy in that they do not adhere to DSM-IV criteria. On the other hand, Breslau et al. (2004a)
note that individuals meeting criteria for partial PTSD have some impairment in work and social interaction domains, although not nearly the level of impairment exhibited by persons meeting the full criteria. Thus, examining how changes in social circumstances relate to changes in the number and severity of symptoms for individuals not meeting full criteria may further illuminate factors underlying delayed PTSD. The main point, though, is that more detailed research on the postdisaster environment is strongly supported by our study.
The question that remains unanswered is why some individuals experience more negative life events and diminished self-esteem in the postdisaster period. Some investigators contend that the postdisaster period can be characterized by an adverse social environment, defined as “a consistent pattern of chronic [negative] impacts to individuals and communities” (Picou et al., 2004
, p. 1496). In other words, if the social environment is more conflict prone, as indicated by transportation problems, litigation, social conflicts, disruption of living arrangements, and the breakdown of public service agencies, then mental disorders will likely increase. In their study of the economic and social consequences of the Exxon Valdez oil spill into the Prince William Sound, Palinkas et al. (1993a
) noted that this environmental disaster was not particularly life-threatening. Nevertheless, the oil spill and subsequent clean-up disrupted subsistence food production (e.g., fishing), strained family and community relations, and increased social inequality. Individuals living in communities most affected by these social and interpersonal changes also reported more depressive, PTSD, and anxiety symptoms compared with those living in less affected communities. Clearly, research is needed to understand better how disasters disrupt familial and community support systems and may alienate victims from local institutions and increase PTSD symptoms.
As with any study, ours has limitations and strengths. First, we omitted individuals without telephones, those who did not speak English or Spanish, and those too disabled to undertake a survey or institutionalized. Given that the sample matched the 2000 census for NYC, elimination of these persons did not appear to introduce overall demographic bias. We are limited, though, in generalizing these findings beyond white, African American, and Latino groups. To date, little research has focused on how the World Trade Center attacks affected the physical or mental health of immigrant communities or the wide variety of ethnic groups living in NYC. It is possible that individuals in these communities may have suffered greater psychological problems due to fewer economic resources and greater job instability (Thiel de Bocanegra and Brickman, 2004
). There may also be cultural differences in the desirability of reporting psychiatric symptoms that could affect the ethnicity-PTSD relationship (Norris et al., 2001
), which we did not explore. We also did not have predisaster data, with much of our data being based on retrospective measures. Although we included lifetime depression to partially account for predisaster psychological problems, we did not have data on other difficulties such as anxiety or personality disorders. Additionally, our mental health measures were based on self-report. Although there has been significant progress in assessing individual mental health with standardized instruments administered by interviewers (Adams et al., 2002
; Kessler et al., 1995b
), there continue to be discrepancies between lay and clinician-based assessments within community population samples. The strengths of our study include the use of a large random sample representative of NYC, the assessment of physical and mental well-being using standard scales and measurements, the focus on a specific time-bound event that meets criteria for community-wide disaster, and a focus on common postdisaster problems faced by survivors.
Community disasters can have a significant impact on the well-being of survivors (Bromet and Dew 1995
). As with all environmental challenges faced by individuals, the degree of exposure to adverse events and the amount of change in personal circumstances can have long-lasting impact on psychological well-being. With regard to the WTCD, further research is needed to determine whether these are transient changes affecting psychological well-being or reflect larger changes in the psychosocial environment of disaster survivors.