Abstract

Although the September 11, 2001, terrorist attacks were the largest human-made disaster in US history, there is little extant research documenting the attacks’ consequences among those most directly affected, that is, persons who were in the World Trade Center towers. Data from a cross-sectional survey conducted 2–3 years after the attacks ascertained the prevalence of long-term, disaster-related posttraumatic stress symptoms and probable posttraumatic stress disorder (PTSD) in 3,271 civilians who evacuated World Trade Center towers 1 and 2. Overall, 95.6% of survivors reported at least 1 current posttraumatic stress symptom. The authors estimated the probable rate of PTSD at 15.0% by using the PTSD Checklist. Women and minorities were at an increased risk of PTSD. A strong inverse relation with annual income was observed. Five characteristics of direct exposure to the terrorist attacks independently predicted PTSD: being on a high floor in the towers, initiating evacuation late, being caught in the dust cloud that resulted from the tower collapses, personally witnessing horror, and sustaining an injury. Working for an employer that sustained fatalities also increased risk. Each addition of an experience of direct exposure resulted in a 2-fold increase in the risk of PTSD (odds ratio = 2.09, 95% confidence interval: 1.84, 2.36). Identification of these risk factors may be useful when screening survivors of large-scale terrorist events for long-term psychological sequelae.

Posttraumatic stress disorder (PTSD) is a psychiatric disorder that occurs after experiencing or witnessing events that threaten death or serious injury and that involves intense feelings of fear, helplessness, or horror (1). National studies have demonstrated that 6.8% of adults develop PTSD in their lifetimes; 3.5% of adults have the condition in any given year, making PTSD the third most common anxiety disorder in the United States (2, 3). Although trauma exposure is a necessary component of PTSD, the presence of a traumatic event by itself does not sufficiently explain why PTSD develops. Substantial variability in the likelihood of developing PTSD after different events exists. For example, the prevalence of PTSD after rape has been estimated to be 55%; in contrast, the prevalence of PTSD after accidents has been estimated to be 7.5% (4).

Intentional disasters, particularly large-scale terrorist attacks, allow for the study of PTSD in unselected communities and circumvent confounding that can occur in community studies of the association of trauma with vulnerability to psychopathology (5, 6). It has been hypothesized that posttraumatic stress symptoms after such disasters may be long-lasting because of the high prevalence of factors known to contribute to the disaster-psychopathology relation (e.g., direct exposure, the witnessing of horror, experience or knowledge of extensive loss of life, relocation, and postdisaster financial strain) (7).

The magnitude and intensity of the attack on the World Trade Center (WTC) on September 11, 2001 (9/11), made it the worst human-made disaster in the history of the United States, with almost 3,000 fatalities, 71,000 jobs lost, and labor and capital losses reaching $36 billion in the months after the attacks (8). Previous research documented a high prevalence of posttraumatic stress symptoms in national and New York City samples within 6 weeks of the attacks and attenuation of symptoms 5 months later (9–13). However, we are unaware of studies devoted exclusively to the psychological consequences among those under direct attack on 9/11: persons who were in the WTC towers. We sought to fill this gap in knowledge by assessing long-term stress reactions among civilians who escaped the WTC towers on the morning of 9/11. Our objectives were to 1) measure the prevalence of posttraumatic stress symptoms and probable PTSD in a large cohort of tower survivors 2–3 years after 9/11; 2) describe the range of direct exposures reported; and 3) assess the independent and collective relations between these exposures and the risk of probable PTSD. Because this study targets those with the greatest exposure, we hypothesized that the burden of posttraumatic stress would be higher than that documented in previous 9/11 studies and that there would be a positive relation between the number of direct exposures reported and the number of cases of posttraumatic stress observed.

MATERIALS AND METHODS

This study was conducted in conjunction with the World Trade Center Health Registry, which was established by the New York City Department of Health and Mental Hygiene and the Agency for Toxic Substances and Disease Registry to monitor the health impact of 9/11 (14). The registry contains interview data from over 71,000 individuals, including rescue/recovery workers and lower Manhattan residents, schoolchildren, building occupants, and passersby. Our study utilized interviews with 3,271 English-speaking, civilian survivors of the WTC attacks. Eligibility criteria included age ≥18 years, physical presence in one of the “twin towers” (1WTC and 2WTC) between the first plane impact and subsequent tower collapses, and no participation in WTC rescue/recovery activities. Psychological sequelae among rescue/recovery personnel, residents, and occupants of other buildings have been reported elsewhere (15–17).

Recruitment and data collection

Between September 2003 and November 2004, tower survivors whose names were gathered from employee lists were contacted as part of an extensive recruitment effort to register all lower Manhattan building occupants (18). Key informants, typically from human resources departments, provided employee names and distributed letters to staff that endorsed enrollment. Recruitment also included a Port Authority of New York and New Jersey list of individuals issued security badges for all 7 WTC buildings during the 5 years before 9/11.

Data collection, conducted concurrent with recruitment, was accomplished through the use of computer-assisted telephone interviews (94.5%) and personal interviews (5.5%). A standardized questionnaire was used, in which participants reported their exact building location on 9/11 after the participants provided informed consent and their eligibility was determined. This precaution, taken because of the sensitive subject matter, precluded our ability to compute outcome rates for tower survivors as opposed to all lower Manhattan building occupants. Contact, cooperation, and response rates for all occupants were 72.5%, 81.3%, and 58.9%, respectively (14). Our cohort's coverage was 21.2%, based on the 15,410 tower survivors estimated by the National Institute of Standards and Technology (NIST) (19).

Posttraumatic stress was assessed with the PTSD Checklist, Stressor-Specific Version (PCL-S) (20). Clinical interviews were not feasible given the participant volume, but the utility of the PCL-S has been verified in numerous studies of civilians exposed to assault, motor vehicle trauma, major illnesses, and terrorism, including the 9/11 attacks (21–25). The PCL-S assesses the full domain of Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, posttraumatic stress symptoms in 3 clusters: 5 intrusive reexperiencing symptoms, 7 numbing and avoidance symptoms, and 5 hyperarousal symptoms. Each symptom's occurrence as a result of 9/11 was queried for the 4 weeks before the interview. Symptoms were considered endorsed as indicating PTSD at moderate disturbance or more (≥3 on the 5-point scale). Frequencies of PCL-S item responses were determined, and the prevalence of probable PTSD (herein referred to as PTSD) was calculated according to a cutoff score of 50 (diagnostic efficiency = 0.94, sensitivity = 0.82, specificity = 0.86) (20).

Potential risk factors included demographic characteristics and 9/11 direct exposures, including physical location (i.e., in which tower the participant was located and the highest floor reached on 9/11), evacuation initiation, exposure to the dust cloud that resulted from the collapse of the towers, witnessing of horrific particulars of the attack, and injuries. Having witnessed a horror was coded as affirmative when a participant personally saw an airplane hitting the WTC, a building collapsing, people running away from a cloud of dust and debris, people who were injured or killed, or people falling or jumping from the WTC towers on 9/11. Having sustained an injury was coded affirmative when a participant reported a burn, broken bone, concussion, cut, sprain, strain, or other injury related to the attacks. We used an additional variable to characterize fatalities at companies for which participants worked, which was accomplished by extracting 9/11 WTC deaths from a publicly available website confirmed in conjunction with federal and local governments (26) and linking this information to the employer of each participant.

Statistical analyses

We performed 2-tailed chi-square tests to compare our cohort's characteristics with those of 803 telephone survey respondents in the NIST's study of occupant behavior, egress, and emergency communications (19). We also used chi-square tests to identify associations between each demographic or event-exposure variable and PTSD status, and we created bivariate logistic regression models to obtain strength of association. Significant risk factors from bivariate modeling (P < 0.1) were next considered for multivariable logistic regression to determine which demographic and disaster-exposure variables independently predicted PTSD. Forward and backward selection were performed using the chi-square from the difference in –2 log likelihood estimates to determine which variables should be retained in the final model. We only retained explanatory variables that significantly improved model fit to obtain a parsimonious model (P < 0.05). To examine the cumulative effect of 9/11 direct exposures on PTSD, an exposure severity score was constructed as the sum of those exposures found to be predictive of PTSD in the multivariable regression model. We used stepwise logistic regression to assess the relation between cumulative exposure severity and risk of PTSD while accounting for demographic factors.

All analyses were completed with SAS software, version 9.2 (SAS Institute, Inc., Cary, North Carolina). Regression models included tests for interaction among demographic and exposure variables. Respondents with missing data were excluded from regression analyses (n = 40).

RESULTS

Table 1 displays the demographic characteristics of the cohort. Of the survivors, 58.8% were men and 68.2% were white. Most were college graduates (66.5%) and had an annual household income ≥$75,000 (59.8%). The demographic characteristics of the true population of WTC tower survivors are unknown. However, compared with the NIST's study of 803 survivors’ egress patterns, our cohort's age distribution was similar (mean = 41 vs. 45; range, 19–75 vs. 21–74) but contained significantly more women (41% vs. 33%; χ2 = 18.2, P < 0.001).

Table 1.

Demographic Characteristics of Study Participants (n = 3,271), 2003–2004

Characteristic No. of Participantsa 
Gender   
    Male 1,922 58.8 
    Female 1,349 41.2 
Age on September 11, 2001, years   
    Mean 40.8 10.9b 
    ≥65 48 1.5 
    45–64 1,188 36.3 
    25–44 1,837 56.2 
    18–24 198 6.1 
Race/ethnicity   
    White 2,230 68.2 
    African-American 361 11.0 
    Hispanic 339 10.4 
    Asian 260 7.9 
    Multiracial/other 81 2.5 
Educational levelc   
    Postgraduate degree 750 23.1 
    College graduate 1,412 43.4 
    High school graduate or equivalent 1,044 32.1 
    Some high school or less 47 1.4 
Income per yearc   
    ≥$100,000 702 24.2 
    $75,000–$99,999 1,031 35.6 
    $50,000–$74,999 483 16.7 
    $25,000–$49,999 522 18.0 
    <$25,000 161 5.6 
Marital statusc   
    Married 2,026 62.2 
    Divorced/separated 280 8.6 
    Widowed 56 1.7 
    Not married 895 27.5 
Characteristic No. of Participantsa 
Gender   
    Male 1,922 58.8 
    Female 1,349 41.2 
Age on September 11, 2001, years   
    Mean 40.8 10.9b 
    ≥65 48 1.5 
    45–64 1,188 36.3 
    25–44 1,837 56.2 
    18–24 198 6.1 
Race/ethnicity   
    White 2,230 68.2 
    African-American 361 11.0 
    Hispanic 339 10.4 
    Asian 260 7.9 
    Multiracial/other 81 2.5 
Educational levelc   
    Postgraduate degree 750 23.1 
    College graduate 1,412 43.4 
    High school graduate or equivalent 1,044 32.1 
    Some high school or less 47 1.4 
Income per yearc   
    ≥$100,000 702 24.2 
    $75,000–$99,999 1,031 35.6 
    $50,000–$74,999 483 16.7 
    $25,000–$49,999 522 18.0 
    <$25,000 161 5.6 
Marital statusc   
    Married 2,026 62.2 
    Divorced/separated 280 8.6 
    Widowed 56 1.7 
    Not married 895 27.5 
a

Numbers may not add up to 3,271 because not all respondents answered all of the questions.

b

Value is presented as standard deviation, not mean.

c

Reported at the time of interview.

Figure 1 displays the locations of tower survivors on the morning of 9/11 before the planes hit the towers. The distribution across towers was equal (n = 1,666 in 1WTC vs. 1,605 in 2WTC). Although the distribution by floor location of the true population of WTC tower survivors is unknown, there was no difference between our cohort and the NIST's cohort according to the agency's floor categories (52.8% vs 50.0% on floors <42, 29.6% vs. 33.4% on floors 43–75, and 17.7% vs. 16.0% on floors 75–110; χ2 = 5.9, P = 0.05).

Figure 1.

Location on September 11, 2001, reported by adult civilian survivors of the attacks on the World Trade Center. Total n = 3,271; missing n = 11.

Figure 1.

Location on September 11, 2001, reported by adult civilian survivors of the attacks on the World Trade Center. Total n = 3,271; missing n = 11.

Our cohort contained fewer survivors who were located at or above the impact zone of 1WTC (n = 3), the first tower attacked, than at or above the impact zone in 2WTC (n = 370). The vast majority of 2WTC survivors above the impact zone reported that they began evacuation after the first plane hit 1WTC (n = 336; 90.8%). Overall, 60.8% were caught in the dust cloud from the tower collapses. Ninety-four percent witnessed horror (range: 32.3% saw an airplane hit one of the towers to 74.2% saw people run away from a cloud of smoke). Thirty-two percent sustained an injury (range: 2.0% concussion to 17.6% sprain or strain). The overwhelming majority of survivors reported that the WTC was their usual place of work (95.8%). Sixty-seven percent worked for an employer that sustained fatalities on 9/11.

Table 2 shows the prevalence of posttraumatic stress symptoms. Almost all survivors had symptoms; only 145 (4.4%) reported none. The 2 most commonly experienced symptoms were in the group D hyperarousal symptom cluster: hypervigilance (50.4%) and being jumpy or easily startled (37.4%). Fifteen percent of survivors (n = 492) screened positive for PTSD. No significant differences were observed in prevalences of any symptom when results were stratified by tower.

Table 2.

Posttraumatic Stress Symptoms and Probable Posttraumatic Stress Disorder Among Adult Civilian Survivors of the Attacks on the World Trade Center (n = 3,271), 2003–2004

DSM-IV Posttraumatic Stress Disorder Symptom Symptom Endorsementa
 
No. of Participants 
No symptoms reported 145 4.4 
DSM-IV group B: reexperiencingb 1,620 49.5 
    Intrusive memories 1,094 33.4 
    Dreams or nightmares 442 13.5 
    Flashbacks 592 18.1 
    Upset at reminders 1,217 37.2 
    Physiologic reactivity 768 23.5 
DSM-IV group C: avoidancec 866 26.5 
    Avoidance of thoughts or feelings 1,029 31.5 
    Avoidance of reminders 848 25.9 
    Psychogenic amnesia 499 15.3 
    Loss of interest 691 21.1 
    Detachment 690 21.1 
    Restricted range of affect 484 14.8 
    Sense of shortened future 916 28 
DSM-IV group D: hyperarousald 1,490 45.5 
    Insomnia 1,033 31.6 
    Irritability or anger 829 25.3 
    Difficulty concentrating 830 25.4 
    Hypervigilance 1,648 50.4 
    Being jumpy or easily startled 1,223 37.4 
Probable posttraumatic stress disordere 492 15.0 
DSM-IV Posttraumatic Stress Disorder Symptom Symptom Endorsementa
 
No. of Participants 
No symptoms reported 145 4.4 
DSM-IV group B: reexperiencingb 1,620 49.5 
    Intrusive memories 1,094 33.4 
    Dreams or nightmares 442 13.5 
    Flashbacks 592 18.1 
    Upset at reminders 1,217 37.2 
    Physiologic reactivity 768 23.5 
DSM-IV group C: avoidancec 866 26.5 
    Avoidance of thoughts or feelings 1,029 31.5 
    Avoidance of reminders 848 25.9 
    Psychogenic amnesia 499 15.3 
    Loss of interest 691 21.1 
    Detachment 690 21.1 
    Restricted range of affect 484 14.8 
    Sense of shortened future 916 28 
DSM-IV group D: hyperarousald 1,490 45.5 
    Insomnia 1,033 31.6 
    Irritability or anger 829 25.3 
    Difficulty concentrating 830 25.4 
    Hypervigilance 1,648 50.4 
    Being jumpy or easily startled 1,223 37.4 
Probable posttraumatic stress disordere 492 15.0 

Abbreviation: DSM-IV, Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition.

a

Endorsement for each symptom required answering “moderately” to “quite a bit” (≥3) on the PTSD Checklist.

b

Must have at least 1 of 5 symptoms at moderate or greater level.

c

Must have at least 3 of 7 symptoms at moderate or greater level.

d

Must have least 2 of 5 symptoms at moderate or greater level.

e

Must have a PTSD Checklist, Stressor-Specific Version Score ≥50.

Table 3 shows that gender, race/ethnicity, educational level, income, and marital status were significantly associated with PTSD in chi-square and bivariate logistic regression analyses. Significant event exposures included having been in the towers above the point of airplane impact, having initiated evacuation late, having been exposed to the dust cloud, having witnessed horror, and having sustained an injury. Survivors who worked for an employer that experienced fatalities on 9/11 had a higher PTSD prevalence than did those who worked for an employer with no fatalities.

Table 3.

Associations of Demographic Characteristics, Event Exposures, and Probable Posttraumatic Stress Disorder Among Adult Civilian Survivors of the Attacks on the World Trade Center (n = 3,271), 2003–2004

 Posttraumatic Stress Disorder
 
Chi-Square Likelihood of Posttraumatic Stress Disorder 
 No.a Test of Association COR 95% CI AORb 95% CI 
Demographic characteristics        
    Gender        
        Male 190 9.9 96.9*** 1.00  1.00  
        Female 302 22.4 2.63 2.16, 3.20*** 1.67 1.31, 2.14*** 
    Age on September 11, 2001, years        
        ≥65 6.3 7.12 1.00    
        45–64 176 14.8 2.61 0.80, 8.48   
        25–44 273 14.9 2.62 0.81, 8.48   
        18–24 40 20.2 3.8 1.12, 12.84*   
    Race/ethnicity        
        White 234 10.5 137.7*** 1.00  1.00  
        African-American 103 28.5 3.41 2.61, 4.44*** 1.54 1.11, 2.14* 
        Hispanic 96 28.3 3.37 2.57, 4.43*** 1.80 1.29, 2.52** 
        Asian 41 15.8 1.60 1.11, 2.29* 1.56 1.02, 2.37* 
        Multiracial/other 18 22.2 2.44 1.42, 4.19** 1.39 0.69, 2.80 
    Educational levelc        
        Postgraduate degree 60 8.0 122.5*** 1.00    
        College graduate 163 11.5 1.50 1.10, 2.05*   
         High school graduate or equivalent 249 23.9 3.60 3.67, 4.86***   
        Some high school or less 17 36.2 6.52 3.40, 12.49***   
    Income per yearc        
        ≥$100,000 42 6.0 253.3*** 1.00  1.00  
        $75,000–$99,999 110 10.7 1.88 1.30, 2.72** 1.58 1.06, 2.36* 
        $50,000–$74,999 88 18.2 3.50 2.38, 5.16*** 2.32 1.51, 3.57*** 
        $25,000–$49,999 139 26.6 5.70 3.95, 8.24*** 3.57 2.34, 5.46*** 
        <$25,000 79 49.1 15.14 9.76, 23.48*** 8.45 5.04, 14.14*** 
    Marital statusb        
        Currently married 239 11.8 51.0*** 1.00    
        Divorced or separated 71 25.4 2.54 1.88, 3.43***   
        Widowed 12 21.4 2.04 1.06, 3.92*   
        Not married 167 18.7 1.72 1.38, 2.13***   
Event exposures        
    Residence on September 11, 2001        
        Manhattan 59 14.8 49.2*** 1.00    
        Bronx 27 24.1 1.83 1.09, 3.05*   
        Brooklyn 86 19.9 1.42 0.99, 2.05   
        Queens 84 20.2 1.45 1.01, 2.10*   
        Staten Island 32 22.1 1.63 1.01, 2.63*   
        Elsewhere in New York 50 9.1 0.57 0.38, 0.86   
        New Jersey 131 12.9 0.85 0.61, 1.19**   
        Other 14 11.5 0.75 0.40, 1.39   
    Tower        
        1 World Trade Center (North Tower) 263 15.8 1.5 1.00    
         2 World Trade Center (South Tower) 229 14.3 0.89 0.73, 1.08   
    Location on September 11, 2001, by airplane impact zoned        
        Below 421 14.6 7.3* 1.00  1.00  
        In impact zone 21 14.3 0.98 0.61, 1.57 1.43 0.83, 2.45 
        Above impact zone 48 21.2 1.58 1.13, 2.21** 1.77 1.17, 2.67** 
    Evacuation initiationd        
        During or after plane's impact into 1 World Trade Center 372 14.0 18.7*** 1.00  1.00  
         During or after plane's impact into 2 World Trade Center 91 18.3 1.38 1.07, 1.77* 1.39 1.02, 1.88* 
         During or after collapse of 2 World Trade Center 28 27.5 2.33 1.49, 3.64** 1.67 0.96, 2.92 
    Exposure to dust and debris cloud that resulted from the collapse of the towers        
        No 102 8.1 78.7*** 1.00  1.00  
        Yes 387 19.5 2.76 2.12, 3.47*** 1.78 1.36, 2.35*** 
    Witnessed horrore        
        No 4.2 18.8*** 1.00  1.00  
        Yes 484 15.7 3.96 3.24, 4.83*** 3.01 2.39, 3.81*** 
    Sustained an injuryf        
        No 197 8.9 201.8*** 1.00  1.00  
        Yes 295 27.9 4.29 2.10, 8.76*** 2.72 1.26, 5.86** 
    Usual place of work was the World Trade Center        
        No 18 13.0 0.4 1.00    
        Yes 470 15.1 1.18 0.71, 1.96   
    Worked for an employer that sustained fatalities        
        No 129 12.5 6.5* 1.00  1.00  
        Yes 335 16.0 1.33 1.07, 1.64* 1.34 1.04, 1.74* 
 Posttraumatic Stress Disorder
 
Chi-Square Likelihood of Posttraumatic Stress Disorder 
 No.a Test of Association COR 95% CI AORb 95% CI 
Demographic characteristics        
    Gender        
        Male 190 9.9 96.9*** 1.00  1.00  
        Female 302 22.4 2.63 2.16, 3.20*** 1.67 1.31, 2.14*** 
    Age on September 11, 2001, years        
        ≥65 6.3 7.12 1.00    
        45–64 176 14.8 2.61 0.80, 8.48   
        25–44 273 14.9 2.62 0.81, 8.48   
        18–24 40 20.2 3.8 1.12, 12.84*   
    Race/ethnicity        
        White 234 10.5 137.7*** 1.00  1.00  
        African-American 103 28.5 3.41 2.61, 4.44*** 1.54 1.11, 2.14* 
        Hispanic 96 28.3 3.37 2.57, 4.43*** 1.80 1.29, 2.52** 
        Asian 41 15.8 1.60 1.11, 2.29* 1.56 1.02, 2.37* 
        Multiracial/other 18 22.2 2.44 1.42, 4.19** 1.39 0.69, 2.80 
    Educational levelc        
        Postgraduate degree 60 8.0 122.5*** 1.00    
        College graduate 163 11.5 1.50 1.10, 2.05*   
         High school graduate or equivalent 249 23.9 3.60 3.67, 4.86***   
        Some high school or less 17 36.2 6.52 3.40, 12.49***   
    Income per yearc        
        ≥$100,000 42 6.0 253.3*** 1.00  1.00  
        $75,000–$99,999 110 10.7 1.88 1.30, 2.72** 1.58 1.06, 2.36* 
        $50,000–$74,999 88 18.2 3.50 2.38, 5.16*** 2.32 1.51, 3.57*** 
        $25,000–$49,999 139 26.6 5.70 3.95, 8.24*** 3.57 2.34, 5.46*** 
        <$25,000 79 49.1 15.14 9.76, 23.48*** 8.45 5.04, 14.14*** 
    Marital statusb        
        Currently married 239 11.8 51.0*** 1.00    
        Divorced or separated 71 25.4 2.54 1.88, 3.43***   
        Widowed 12 21.4 2.04 1.06, 3.92*   
        Not married 167 18.7 1.72 1.38, 2.13***   
Event exposures        
    Residence on September 11, 2001        
        Manhattan 59 14.8 49.2*** 1.00    
        Bronx 27 24.1 1.83 1.09, 3.05*   
        Brooklyn 86 19.9 1.42 0.99, 2.05   
        Queens 84 20.2 1.45 1.01, 2.10*   
        Staten Island 32 22.1 1.63 1.01, 2.63*   
        Elsewhere in New York 50 9.1 0.57 0.38, 0.86   
        New Jersey 131 12.9 0.85 0.61, 1.19**   
        Other 14 11.5 0.75 0.40, 1.39   
    Tower        
        1 World Trade Center (North Tower) 263 15.8 1.5 1.00    
         2 World Trade Center (South Tower) 229 14.3 0.89 0.73, 1.08   
    Location on September 11, 2001, by airplane impact zoned        
        Below 421 14.6 7.3* 1.00  1.00  
        In impact zone 21 14.3 0.98 0.61, 1.57 1.43 0.83, 2.45 
        Above impact zone 48 21.2 1.58 1.13, 2.21** 1.77 1.17, 2.67** 
    Evacuation initiationd        
        During or after plane's impact into 1 World Trade Center 372 14.0 18.7*** 1.00  1.00  
         During or after plane's impact into 2 World Trade Center 91 18.3 1.38 1.07, 1.77* 1.39 1.02, 1.88* 
         During or after collapse of 2 World Trade Center 28 27.5 2.33 1.49, 3.64** 1.67 0.96, 2.92 
    Exposure to dust and debris cloud that resulted from the collapse of the towers        
        No 102 8.1 78.7*** 1.00  1.00  
        Yes 387 19.5 2.76 2.12, 3.47*** 1.78 1.36, 2.35*** 
    Witnessed horrore        
        No 4.2 18.8*** 1.00  1.00  
        Yes 484 15.7 3.96 3.24, 4.83*** 3.01 2.39, 3.81*** 
    Sustained an injuryf        
        No 197 8.9 201.8*** 1.00  1.00  
        Yes 295 27.9 4.29 2.10, 8.76*** 2.72 1.26, 5.86** 
    Usual place of work was the World Trade Center        
        No 18 13.0 0.4 1.00    
        Yes 470 15.1 1.18 0.71, 1.96   
    Worked for an employer that sustained fatalities        
        No 129 12.5 6.5* 1.00  1.00  
        Yes 335 16.0 1.33 1.07, 1.64* 1.34 1.04, 1.74* 

Abbreviations: AOR, adjusted odds ratio; CI, confidence interval; COR, crude odds ratio.

* P < 0.05; **P < 0.01; ***P < 0.0001.

a

Numbers may not add up to 3,271 because not all respondents answered all the questions.

b

Variables were considered for inclusion in the multivariable logistic regression model if their bivariate regression model results were significant at P < 0.1. Only variables that significantly contributed to final model fit were retained to present a parsimonious model.

c

Reported at the time of interview.

d

The plane impacts occurred at 8:46 AM between floors 93 and 99 for 1 World Trade Center and 9:02 AM between floors 77 and 85 for 2 World Trade Center. Two World Trade Center collapsed at 9:59 am, and 1 World Trade Center collapsed at 10:28 AM.

e

Includes personally seeing at least 1 of the following: an airplane hitting one of the World Trade Center towers, people falling or jumping from the World Trade Center, buildings collapsing, people running away from a cloud of smoke, or injured or dead individuals.

f

Includes sustaining at least 1 of the following: broken bone, burn, concussion, cut, sprain or strain, or other injury.

In the multivariable logistic regression model, gender, race/ethnicity, and income were associated with PTSD, but educational level and marital status were not significantly associated. Tests for interaction among gender, race/ethnicity, and income variables in the model resulted in no significance. Compared with whites, the risk of PTSD was highest among Hispanics (adjusted odds ratio = 1.80, 95% confidence interval (CI): 1.29, 2.52) and African Americans (adjusted odds ratio = 1.54, 95% CI: 1.11, 2.14). There was a strong inverse relation between income and PTSD. Specifically, survivors with incomes of <$25,000 per year were 8 times more likely to have PTSD than those earning ≥$100,000 per year (49.1% vs. 6.0%; adjusted odds ratio = 8.45, 95% CI: 5.04, 14.14).

Among 9/11 exposures, the following experiences indicating severe exposure were associated with PTSD in both bivariate and multivariable logistic regression models: having witnessed horror (adjusted odds ratio = 3.01, 95% CI: 2.39, 3.81), having sustained an injury (adjusted odds ratio = 2.72, 95% CI: 1.26, 5.86), having been exposed to the dust cloud (adjusted odds ratio = 1.78, 95% CI: 1.36, 2.35), having been above the impact zones rather than below (adjusted odds ratio = 1.77, 95% CI: 1.17, 2.67), and having evacuated later rather than early (adjusted odds ratio = 1.39, 95% CI: 1.02, 1.88 during or after the plane's impact into 2WTC). Tests of interaction between significant demographic (i.e., gender, race/ethnicity, and income) and exposure variables resulted in no significance.

An exposure severity score was created to test the relation between cumulative exposure and PTSD. It included the following dichotomous variables: highest floor reached relative to the impact zone (below vs. at or above), initiation of evacuation (during or after the plane's impact into 1WTC vs. during or after the plane's impact into 2WTC), exposure to the dust cloud (no vs. yes), witnessing of any horror event (no vs. yes), and sustaining any injury (no vs. yes). The score, ranging from 0 to 5 (low to high), was normally distributed, and there was a positive, monotonic relation between the exposure severity score and the PCL-S score (Figure 2; mean PCL-S scores ranged from 25 at 0 exposures to 44 at 4 exposures).

Figure 2.

Relation of direct exposure and posttraumatic stress symptoms among adult civilian survivors of the attacks on the World Trade Center. Total n = 3,271; n = 102 and mean PTSD Checklist, Stressor-Specific Version (PCL-S) score = 25 at exposure severity score 0; n = 691 and mean PCL-S score = 28 at exposure severity score 1; n = 1272 and mean PCL-S score = 28 at exposure severity score 2; n = 901 and mean PCL-s score 32 at exposure severity score 3; n = 248 and mean PCL-S score = 44 at exposure severity score 4; and n = 17 and mean PCL score = 42 at exposure severity score 5.

Figure 2.

Relation of direct exposure and posttraumatic stress symptoms among adult civilian survivors of the attacks on the World Trade Center. Total n = 3,271; n = 102 and mean PTSD Checklist, Stressor-Specific Version (PCL-S) score = 25 at exposure severity score 0; n = 691 and mean PCL-S score = 28 at exposure severity score 1; n = 1272 and mean PCL-S score = 28 at exposure severity score 2; n = 901 and mean PCL-s score 32 at exposure severity score 3; n = 248 and mean PCL-S score = 44 at exposure severity score 4; and n = 17 and mean PCL score = 42 at exposure severity score 5.

Table 4 displays the stepwise logistic regression models that examined the relation between the exposure severity score and the risk of PTSD. In the unadjusted model, each increase in severity score yielded a 117% greater likelihood of a survivor's screening positive for symptoms (crude odds ratio = 2.17; 95% CI: 1.94, 2.42). When the score was controlled for demographic confounders, the relation between the score and PTSD was slightly attenuated. In the final model, each additional exposure was associated with a 109% greater likelihood of PTSD (adjusted odds ratio = 2.09, 95% CI: 1.84, 2.36).

Table 4.

Stepwise Regression Models Demonstrating the Association Between Severity of Direct Exposure During the 9/11 Attacks and Probable Posttraumatic Stress Disorder, 2003–2004

 Step 1
 
Step 2
 
Step 3
 
Step 4
 
 COR 95% CI AOR 95% CI AOR 95% CI AOR 95% CI 
Exposure severity scorea 2.17 1.94, 2.42*** 2.12 1.90, 2.37*** 2.06 1.84, 2.30*** 2.09 1.84, 2.36*** 
Gender         
    Male   1.00  1.00  1.00  
    Female   2.47 2.01, 3.03*** 2.12 1.72, 2.63*** 1.83 1.45, 2.32*** 
Race/ethnicity         
    White     1.00  1.00  
    African-American     2.43 1.82, 3.23*** 1.53 1.12, 2.10** 
    Hispanic     2.78 2.08, 3.72*** 1.86 1.35, 2.57*** 
    Asian     1.62 1.11, 2.36* 1.49 0.99, 2.23 
    Other     2.05 1.16, 3.62* 1.19 0.60, 2.40 
Income per year         
    ≥$100,000       1.00  
    $75,000–$99,999       1.47 1.00, 2.17* 
    $50,000–$74,999       2.28 1.50, 3.44*** 
    $25,000–$49,999       3.27 2.18, 4.92*** 
    <$25,000       8.24 5.06, 13.43*** 
   Adjusted R2 0.11  0.15  0.18  0.25  
 Step 1
 
Step 2
 
Step 3
 
Step 4
 
 COR 95% CI AOR 95% CI AOR 95% CI AOR 95% CI 
Exposure severity scorea 2.17 1.94, 2.42*** 2.12 1.90, 2.37*** 2.06 1.84, 2.30*** 2.09 1.84, 2.36*** 
Gender         
    Male   1.00  1.00  1.00  
    Female   2.47 2.01, 3.03*** 2.12 1.72, 2.63*** 1.83 1.45, 2.32*** 
Race/ethnicity         
    White     1.00  1.00  
    African-American     2.43 1.82, 3.23*** 1.53 1.12, 2.10** 
    Hispanic     2.78 2.08, 3.72*** 1.86 1.35, 2.57*** 
    Asian     1.62 1.11, 2.36* 1.49 0.99, 2.23 
    Other     2.05 1.16, 3.62* 1.19 0.60, 2.40 
Income per year         
    ≥$100,000       1.00  
    $75,000–$99,999       1.47 1.00, 2.17* 
    $50,000–$74,999       2.28 1.50, 3.44*** 
    $25,000–$49,999       3.27 2.18, 4.92*** 
    <$25,000       8.24 5.06, 13.43*** 
   Adjusted R2 0.11  0.15  0.18  0.25  

Abbreviations: AOR, adjusted odds ratio; CI, confidence interval; COR, crude odds ratio.

* P < 0.05; **P < 0.01; ***P < 0.0001.

a

The direct exposure severity score included the following dichotomous variables: highest floor reached according to impact zone (below vs. at or above impact zone), time of initiation of evacuation (during or after the plane's impact into 1 World Trade Center vs. during or after the plane impact into 2 World Trade Center), exposure to the dust and debris cloud (no or yes), having witnessed any horror event (no or yes), and having sustained any injury (no or yes).

DISCUSSION

In the largest cohort of civilian survivors of the WTC attacks assembled to date, we found that the majority of survivors experienced multiple 9/11-related posttraumatic stress symptoms 2–3 years after the attacks. Almost 1 in 6 participants screened positive for probable PTSD. This finding underscores the long-lasting mental health effects of mass violence but is lower than the rate reported in similar research on other terrorist attack survivors (6). Despite differences in levels of exposure, study design, screening instruments, and length of follow-up, assessments of bombings in France between 1982 and 1987 found PTSD rates of 31% and 18% among survivors 2–3 and 8 years later, respectively (27, 28). Six months after the 1987 bombing in Enniskillen, Northern Ireland, 50% of survivors were determined to have PTSD (28). In the United States, 34% of survivors of the 1995 bombing of the Alfred P. Murrah Federal Building in Oklahoma City had PTSD 6 months after the attack, and PTSD was chronic in 100% of cases, lasting ≥3 months (30). In a follow-up study 17 months after the bombing, 89% of cases were unremitted (31).

One explanation for the lower prevalence of PTSD documented here compared with the few other extant postterrorism studies is that many survivors of the attacks on the WTC did not realize that their lives were in extreme danger during the event. Reports from evacuees included descriptions of calm descent (32–34). Many survivors did not imagine that the towers would collapse, and it was not until after a safe evacuation that the magnitude of the situation was apparent. It is possible that awareness of life threat is a driver of high risk for PTSD (35), and the absence of such knowledge accounts for the lower prevalence documented here compared with other terrorism studies. Still, longitudinal study of WTC survivors to determine whether symptoms increase or decrease over time is warranted; research on the Oklahoma City bombing showed that symptoms and functional impairment of bombing survivors might not be diagnosable for years after an attack (31).

As expected, we estimated a higher prevalence of PTSD among tower survivors than among the general population after 9/11 (9–13). Although population studies demonstrated that rates of 9/11-related PTSD declined over time, they also found that those with the highest exposure burden experienced the worst mental health. For example, in a small group of individuals who were physically in any part of the WTC complex at the time of the attacks and part of a representative sample of New York City residents, 36.7% had PTSD within 6 months of 9/11 (13), and among primary care patients who reported being in the WTC or lower Manhattan on 9/11, 17.1% had PTSD 1 year after the attack (25).

We examined demographic risk factors and found that their relation supported the general association of social disadvantage with adverse mental health indicators (6). Our observations about female gender were consistent with findings from other post-9/11 studies (36), and could have been explained by women's higher prevalence of pretrauma psychological disorders, ancillary stressors, caretaking responsibilities, and gender role expectations (37). African Americans and Hispanics were more likely to have PTSD than were whites, similar to what was seen in other studies (38). Explanations for this observation include differential vulnerability (39), pre-9/11 exposure (39), immigrant status (25), pre-9/11 mental health problems (39), lack of social supports, and perievent emotional reactions (39). However, by far the strongest demographic risk factor in our study was low income. We observed a pronounced dose-response relation between income and PTSD that was independent of other covariates. Potential explanations for this relation include findings that poor populations tend to avoid seeking (40) or receiving (41, 42) treatment from mental health specialists, as well as broader sociologic theories on the effect of marginalization, lack of resources, and powerlessness on coping with negative life events (25, 43–45).

It has been well-documented that the extent of exposure to a disaster is one of the most important factors in the development of disaster-related PTSD. However, most studies used relatively crude measures of exposure, often combining the loss of a loved one or property and participation in rescue efforts into one broad categorization of direct exposure. Our large cohort allowed us to focus on experiences by civilians. PTSD risk was greater among survivors whose experiences indicated severe life threat, as defined by location in the towers, time of evacuation initiation, or dust cloud exposure. We also found strong relations between having witnessed horror or having sustained injuries and PTSD, consistent with other studies on human-made disasters (12, 13, 46–48). One potential explanation as to why those who witness horror are more susceptible to PTSD might lie in a biologic understanding of PTSD etiology, as images of grotesque and unimaginable scenes are encoded into memory and may be relived upon stimuli.

We found an interesting relation between coworker casualties and PTSD. In a meta-analysis, Rubonis and Bickman (7) found that the number of human casualties clearly affects the relation between disasters and consequent psychological problems. It is possible that WTC fatalities within one's company indicated increased life threat for survivors, but bereavement for colleagues who perished may have also contributed to this finding (7, 46).

The cumulative effect of severe exposure was found to be significantly associated with PTSD in stepwise regression, demonstrating that although tower survivors may have shared a collective experience, individuals who were exposed to several of the most troubling and life-threatening events during the disaster were at the greatest risk of PTSD. This cumulative effect is an important observation that has implications for future disaster response and that warrants further attention. We are aware of only 1 prior study related to the WTC disaster that used a comparable measure of exposure severity (49). Brackbill et al. (50). found that residents with a very high exposure level (i.e., ≥6 of 14 events, including having been afraid of being killed, having a friend or relative killed, having been forced to move, having financial difficulties, etc.) were significantly more likely to screen positive for PTSD at a 1-year assessment.

Strengths of the present study include the procurement of a large cohort of civilians who survived the WTC attacks that had characteristics comparable to the NIST's smaller egress study, the use of a validated measure of PTSD, and detailed characterizations of 9/11 exposures that were not explored in previous studies. A central limitation to this study was the recruitment strategy, which may have produced selection bias and limited generalizability to all those who evacuated the towers on 9/11. The true effect of this bias is unknown because of the lack of information in the literature on all tower survivors, but if those most affected were more likely to participate in our study, our estimates would have been inflated. Alternatively, if those most affected avoided participation, true symptom prevalence would have been underestimated. Our findings were also based on a cross-sectional survey, which typically makes the temporal relation of exposures and outcomes unclear. However, exposures assessed here were ones that occurred on 9/11, and all stress symptoms were reported as occurring in the last 4 weeks before participation. This study relied on self-reporting, which could have affected the precision of true symptom prevalence. Although the PCL-S is one of the most widely used screening instruments today, it is inevitably less precise than a clinical interv iew. Confounding and effect modification could have led to nonrandom error in our risk factor analyses: Lack of information on preevent psychological functioning and trauma, subjective appraisal of life threat, bereavement, social support, and mental health services may account for some of the observed relations between the risk factors examined here and PTSD (6). Further, our data collection instrument did not include other psychiatric disorders, such as depression, generalized anxiety disorder, or substance abuse, that are commonly observed after trauma (12, 30).

As the long-term effects of the WTC disaster emerge (50, 51), the results from this study suggest that some survivors of the WTC disaster will continue to report psychological symptoms years after their exposure to the events of 9/11. The implication of this finding is that the impact of terrorism on survivors, particularly those in low socioeconomic positions, could be substantial, as PTSD is known to be comorbid with other disorders and harmful behaviors that affect daily functioning, wellness, and relationships (52).

As disaster literature moves toward understanding long-term risks in the general population, this study reminds us that the relation between direct exposures and PTSD is clear and suggests potential avenues for planning policy to reduce the burden of terrorism-related psychopathology. Disaster preparedness training for the civilian workforce should incorporate disclosure of potential experiences that contribute to PTSD risk such as evacuation, injury, and personally witnessing the horrors of death and destruction. The additive effect of such direct exposures should also be taken into consideration during postdisaster treatment.

Abbreviations

    Abbreviations
  • CI

    confidence interval

  • NIST

    National Institute of Standards and Technology

  • PCL-S

    PTSD Checklist, Stressor-Specific Version

  • PTSD

    posttraumatic stress disorder

  • WTC

    World Trade Center

Author affiliations: Division of Environmental Health, New York City Department of Health and Mental Hygiene, New York, New York (Laura DiGrande); Department of Psychiatry, Columbia University Medical Center, New York, New York (Yuval Neria); New York State Psychiatric Institute, New York, New York (Yuval Neria); Department of Epidemiology, Mailman School of Public Health, Columbia University, New York, New York (Yuval Neria, Sandro Galea); Agency for Toxic Substances and Disease Registry, Atlanta, Georgia (Robert M. Brackbill); and Research Triangle Institute, International, Chicago, Illinois (Paul Pulliam).

This work was supported by cooperative agreement U50/ATU272750 from the Agency for Toxic Substances and Disease Registry.

The authors thank Sharon Campolucci, Jay Sapp, Lisa Thalji, Joseph Murphy, Melissa Dolan, Pauline Thomas, Deborah Walker, David Wu, Mark Farfel, and Lorna Thorpe, all of whom helped establish and/or maintain the World Trade Center Health Registry, and Robyn Gershon and Steven Stellman, who provided invaluable input on the design and description of this particular study.

The contents of this article are solely the responsibility of the authors and do not necessarily represent the official views of the Agency for Toxic Substances and Disease Registry.

Conflict of interest: none declared.

References

1.
American Psychiatric Association
Diagnostic and Statistical Manual of Mental Disorders
 , 
1994
Fourth Edition
Washington, DC
American Psychiatric Publishing, Inc
2.
Kessler
RC
Berglund
P
Demler
O
, et al.  . 
Lifetime prevalence and age-of-onset distributions of DSM-IV disorders in the National Comorbidity Survey Replication
Arch Gen Psychiatry
 , 
2005
, vol. 
62
 
6
(pg. 
593
-
602
)
3.
Kessler
RC
Chiu
WT
Demler
O
, et al.  . 
Prevalence, severity, and comorbidity of 12-month DSM-IV disorders in the National Comorbidity Survey Replication
Arch Gen Psychiatry
 , 
2005
, vol. 
62
 
6
(pg. 
617
-
627
)
4.
Kessler
RC
Sonnega
A
Bromet
E
, et al.  . 
Posttraumatic stress disorder in the National Comorbidity Survey
Arch Gen Psychiatry
 , 
1995
, vol. 
52
 
12
(pg. 
1048
-
1060
)
5.
North
CS
Pfefferbaum
B
Research on the mental health effects of terrorism
JAMA
 , 
2002
, vol. 
288
 
5
(pg. 
633
-
636
)
6.
Neria
Y
Nandi
A
Galea
S
Post-traumatic stress disorder following disasters: a systematic review
Psychol Med
 , 
2008
, vol. 
38
 
4
(pg. 
467
-
480
)
7.
Rubonis
AV
Bickman
L
Psychological impairment in the wake of disaster: the disaster-psychopathology relationship
Psychol Bull.
 , 
1991
, vol. 
109
 
3
(pg. 
384
-
399
)
8.
Bram
J
Orr
J
Rappaport
C
Measuring the effects of the September 11 attack on New York City
FRBNY Econ Pol Rev
 , 
2002
, vol. 
8
 
2
(pg. 
5
-
20
)
9.
Schlenger
WE
Caddell
JM
Ebert
L
, et al.  . 
Psychological reactions to terrorist attacks: findings from the National Study of Americans’ Reactions to September 11
JAMA
 , 
2002
, vol. 
288
 
5
(pg. 
581
-
588
)
10.
Schuster
MA
Stein
BD
Jaycox
L
, et al.  . 
A national survey of stress reactions after the September 11, 2001, terrorist attacks
N Engl J Med
 , 
2001
, vol. 
345
 
20
(pg. 
1507
-
1512
)
11.
Silver
RC
Holman
EA
McIntosh
DN
, et al.  . 
Nationwide longitudinal study of psychological responses to September 11
JAMA
 , 
2002
, vol. 
288
 
10
(pg. 
1235
-
1244
)
12.
Galea
S
Ahern
J
Resnick
H
, et al.  . 
Psychological sequelae of the September 11 terrorist attacks in New York City
N Engl J Med
 , 
2002
, vol. 
346
 
13
(pg. 
982
-
987
)
13.
Galea
S
Vlahov
D
Resnick
H
, et al.  . 
Trends of probable post-traumatic stress disorder in New York City after the September 11 terrorist attacks
Am J Epidemiol
 , 
2003
, vol. 
158
 
6
(pg. 
514
-
524
)
14.
World Trade Center Health Registry
New York City Department of Health and Mental Hygiene
Data File User's Manual
 , 
2006
New York, NY
New York City Department of Health and Mental Hygiene
 
15.
Perrin
MA
DiGrande
L
Wheeler
K
, et al.  . 
Differences in PTSD prevalence and associated risk factors among World Trade Center disaster rescue and recovery workers
Am J Psychiatry
 , 
2007
, vol. 
164
 
9
(pg. 
1385
-
1394
)
16.
DiGrande
L
Perrin
MA
Thorpe
LE
, et al.  . 
Posttraumatic stress symptoms, PTSD, and risk factors among lower Manhattan residents 2–3 years after the September 11, 2001 terrorist attacks
J Trauma Stress
 , 
2008
, vol. 
21
 
3
(pg. 
264
-
273
)
17.
Brackbill
RM
Thorpe
LE
DiGrande
L
, et al.  . 
Surveillance for World Trade Center disaster health effects among survivors of collapsed and damaged buildings
MMWR Surveill Summ
 , 
2006
, vol. 
55
 
2
(pg. 
1
-
18
)
18.
World Trade Center Health Registry
New York City Department of Health and Mental Hygiene
Sample Building and Denominator Estimation
 , 
2006
New York, NY
New York City Department of Health and Mental Hygiene
 
19.
Averill
JD
Mileti
DS
Peacock
RD
, et al.  . 
Federal Building and Fire Safety Investigation of the World Trade Center Disaster. Occupant Behavior, Egress, and Emergency Communications
 , 
2005
Washington, DC
National Institute of Standards and Technology
 
http://wtc.nist.gov/pubs/NISTNCSTAR1-7.pdf (Accessed September 22, 2010)
20.
Weathers
FW
Litz
BT
Herman
DS
, et al.  . 
 
The PTSD Checklist (PCL): reliability, validity, and diagnostic utility. Presented at the Sixth Annual Meeting of the International Society for Traumatic Stress Studies, San Antonio, TX, 1993
21.
Smith
MY
Redd
W
DuHamel
K
, et al.  . 
Validation of the PTSD Checklist—Civilian version in survivors of bone marrow transplantation
J Trauma Stress
 , 
1999
, vol. 
12
 
3
(pg. 
485
-
499
)
22.
Ruggiero
KJ
Del Ben
K
Scotti
JR
, et al.  . 
Psychometric properties of the PTSD Checklist—Civilian version
J Trauma Stress
 , 
2003
, vol. 
16
 
5
(pg. 
495
-
502
)
23.
Blanchard
EB
Hickling
EJ
Taylor
AE
, et al.  . 
Who develops PTSD from motor vehicle accidents?
Behav Res Ther
 , 
1996
, vol. 
34
 
1
(pg. 
1
-
10
)
24.
Neria
Y
Olfson
M
Gameroff
MJ
, et al.  . 
The mental health consequences of disaster-related loss: findings from primary care one year after the 9/11 terrorist attacks
Psychiatry
 , 
2008
, vol. 
71
 
4
(pg. 
339
-
348
)
25.
Neria
Y
Gross
R
Olfson
M
, et al.  . 
Posttraumatic stress disorder in primary care one year after the 9/11 attacks
Gen Hosp Psychiatry
 , 
2006
, vol. 
28
 
3
(pg. 
213
-
222
)
26.
CNN
September 11: A Memorial
 , 
2001
Atlanta, GA
Cable News Network
 
(http://www.cnn.com/SPECIALS/2001/memorial). (Accessed December 1, 2004)
27.
Abenhaim
L
Dab
W
Salmi
LR
Study of civilian victims of terrorist attacks (France 1982–1987)
J Clin Epidemiol
 , 
1992
, vol. 
45
 
2
(pg. 
103
-
109
)
28.
Verger
P
Dab
W
Lamping
DL
, et al.  . 
The psychological impact of terrorism: an epidemiologic study of posttraumatic stress disorder and associated factors in victims of the 1995–1996 bombings in France
Am J Psychiatry
 , 
2004
, vol. 
161
 
8
(pg. 
1384
-
1389
)
29.
Curran
PS
Bell
P
Murray
A
, et al.  . 
Psychological consequences of the Enniskillen bombing
Br J Psychiatry
 , 
1990
, vol. 
156
 (pg. 
479
-
482
)
30.
North
CS
Nixon
SJ
Shariat
S
, et al.  . 
Psychiatric disorders among survivors of the Oklahoma City bombing
JAMA
 , 
1999
, vol. 
282
 
8
(pg. 
755
-
762
)
31.
North
CS
Pfefferbaum
B
Tivis
L
, et al.  . 
The course of posttraumatic stress disorder in a follow-up study of survivors of the Oklahoma City bombing
Ann Clin Psychiatry
 , 
2004
, vol. 
16
 
4
(pg. 
209
-
215
)
32.
Dwyer
J
Flynn
K
102 Minutes: The Untold Story of the Fight to Survive Inside the Twin Towers
 , 
2005
New York, NY
Times Books
33.
Gershon
RR
Qureshi
KA
Rubin
MS
, et al.  . 
Factors associated with high-rise evacuation: qualitative results from the World Trade Center Evacuation Study
Prehosp Disaster Med
 , 
2007
, vol. 
22
 
3
(pg. 
165
-
173
)
34.
Proulx
G
Fahy
RF
Walker
A
Analysis of First-Person Accounts From Survivors of the World Trade Center Evacuation on September 11, 2001
  
(Research report no. 178). Ottawa, Canada: Institute for Research in Construction; 2004. (http://nparc.cisti-icist.nrc-cnrc.gc.ca/npsi/ctrl?action=rtdoc&an=5752510&article=0&lang=en). (Accessed September 22, 2010)
35.
Marshall
RD
Bryant
RA
Amsel
L
, et al.  . 
The psychology of ongoing threat: relative risk appraisal, the September 11 attacks, and terrorism-related fears
Am Psychol
 , 
2007
, vol. 
62
 
4
(pg. 
304
-
316
)
36.
Pulcino
T
Galea
S
Ahern
J
, et al.  . 
Posttraumatic stress in women after the September 11 terrorist attacks in New York City
J Womens Health (Larchmt)
 , 
2003
, vol. 
12
 
8
(pg. 
809
-
820
)
37.
Kimmerling
R
Ouimette
P
Wofe
J
Gender and PTSD
 , 
2002
New York, NY
The Guilford Press
38.
Galea
S
Vlahov
D
Tracy
M
, et al.  . 
Hispanic ethnicity and post-traumatic stress disorder after a disaster: evidence from a general population survey after September 11, 2001
Ann Epidemiol
 , 
2004
, vol. 
14
 
8
(pg. 
520
-
531
)
39.
Norris
FH
Alegria
M
Mental health care for ethnic minority individuals and communities in the aftermath of disasters and mass violence
CNS Spectr
 , 
2005
, vol. 
10
 
2
(pg. 
132
-
140
)
40.
Howard
KI
Cornille
TA
Lyons
JS
, et al.  . 
Patterns of mental health service utilization
Arch Gen Psychiatry
 , 
1996
, vol. 
53
 
8
(pg. 
696
-
703
)
41.
Gallo
JJ
Marino
S
Ford
D
, et al.  . 
Filters on the pathway to mental health care, II. Sociodemographic factors
Psychol Med
 , 
1995
, vol. 
25
 
6
(pg. 
1149
-
1160
)
42.
Leaf
PJ
Bruce
ML
Tischler
GL
, et al.  . 
Factors affecting the utilization of specialty and general medical mental health services
Med Care
 , 
1988
, vol. 
26
 
1
(pg. 
9
-
26
)
43.
Dohrenwend
BS
Social status and stressful life events
J Pers Soc Psychol
 , 
1973
, vol. 
28
 
2
(pg. 
225
-
235
)
44.
McLeod
JD
Kessler
RC
Socioeconomic status differences in vulnerability to undesirable life events
J Health Soc Behav
 , 
1990
, vol. 
31
 
2
(pg. 
162
-
172
)
45.
Ross
CE
Mirowsky
J
Pribesh
S
Powerlessness and the amplification of threat: neighborhood disadvantage, disorder, and mistrust
Am Soc Rev.
 , 
2001
, vol. 
66
 (pg. 
568
-
591
)
46.
Neria
Y
Gross
R
Litz
B
, et al.  . 
Prevalence and psychological correlates of complicated grief among bereaved adults 2.5–3.5 years after September 11th attacks
J Trauma Stress
 , 
2007
, vol. 
20
 
3
(pg. 
251
-
262
)
47.
Gregg
W
Medley
I
Fowler-Dixon
R
, et al.  . 
Psychological consequences of the Kegworth air disaster
Br J Psychiatry
 , 
1995
, vol. 
67
 
6
(pg. 
812
-
817
)
48.
Hull
AM
Alexander
DA
Klein
S
Survivors of the Piper Alpha oil platform disaster: long-term follow-up study
Br J Psychiatry
 , 
2002
, vol. 
181
 (pg. 
433
-
438
)
49.
Adams
RE
Boscarino
JA
Predictors of PTSD and delayed PTSD after disaster: the impact of exposure and psychosocial resources
J Nerv Ment Dis
 , 
2006
, vol. 
194
 
7
(pg. 
485
-
493
)
50.
Brackbill
RM
Hadler
JL
DiGrande
L
, et al.  . 
Asthma and posttraumatic stress symptoms 5 to 6 years following exposure to the World Trade Center terrorist attack
JAMA
 , 
2009
, vol. 
302
 
5
(pg. 
502
-
516
)
51.
Neria
Y
Olfson
M
Gameroff
MJ
, et al.  . 
Long-term course of probable PTSD after the 9/11 attacks: a study in urban primary care
J Trauma Stress
 , 
2010
, vol. 
23
 
4
(pg. 
474
-
482
)
52.
Vlahov
D
Galea
S
Resnick
H
, et al.  . 
Increased use of cigarettes, alcohol, and marijuana among Manhattan, New York, residents after the September 11th terrorist attacks
Am J Epidemiol
 , 
2002
, vol. 
55
 
11
(pg. 
988
-
996
)