This study documents that the rescue and recovery and cleanup efforts carried out by the workers at the World Trade Center are associated with substantial chronic psychological morbidity and extensive impairment of social functioning. Of 10,132 WTC workers whom we examined, 11.1% had probable PTSD within the month before their mental health examination.
These findings on the high prevalence of PTSD in WTC workers are similar to those encountered in U.S. war veterans. Hoge et al. (2004)
, using the same PTSD diagnostic checklist that we used and the same cutoff score of 50 on this PCL checklist, found a prevalence of 11.5% for probable PTSD among soldiers returning from Afghanistan. Although not directly comparable to our 1-month prevalence rates, the 12-month prevalence estimates for PTSD among the general adult population in the United States range between 3% and 4% (Kessler et al. 2005
Consistent with previous studies, PTSD was not the only type of postdisaster psychopathology observed in this cohort. Almost 9% of the cohort met criteria for probable depression during the month in which they were examined, and 5.0% fulfilled diagnostic criteria for probable panic disorder. Another 17% had probable excess use of alcohol. Personal loss of family and friends appears to have increased these rates. Rates of psychological comorbidity were also high. Among the responders with probable PTSD, 12.7% also met criteria for panic disorder or depression, and 1.7% met criteria for all three disorders: probable PTSD, depression, and panic disorder. Of note, approximately half of the workers with probable PTSD also had a probable comorbid psychiatric condition, and these workers were at far higher risk for social dysfunction and alcohol problems. Further, the point prevalence of PTSD in the comorbid group did not decline over time, suggesting that PTSD in this group may be more chronic.
Our results are consistent with other reports in which PTSD has been strongly associated with functional impairment, including interference with occupation, family, school, and leisure activities, among subjects in community samples (Kessler 2000
; North et al. 2002
), male and female veterans (Kulka et al. 1988
; Stein et al. 1997
), primary care patients (Zatzick et al. 1997
), and female survivors of interpersonal violence (Rapaport et al. 2002
). Level of functional impairment associated with PTSD has been comparable to levels observed in severe chronic depression (Laffaye et al. 2003
). In our population, the odds ratio for social impairment was significantly elevated more than 17-fold among those with PTSD compared to those without probable PTSD.
Many workers who did not meet study criteria for probable PTSD nevertheless reported suffering from PTSD-related symptoms of stress in the month before their evaluation (). For example, approximately one-third of the responders without probable PTSD reported disturbing memories, thoughts, or images; having trouble falling asleep or staying asleep, and being “super-alert.” Nearly half (45%) of all responders without probable PTSD reported suffering from a substantial stress reaction as long as 5 years after the WTC disaster, a rate comparable to the nationally representative sample of U.S. adults surveyed only 3–5 days after the attacks when symptoms typically are at their highest level (Silver et al. 2002
Most research studies and clinical interventions focus on patients who meet full criteria for PTSD. Our data show that such an approach would fail to meet the needs of many of the responders in the present population who were not classified as having probable PTSD but did have a high prevalence of distressing symptoms and functional impairment. In fact, in the current sample, the likelihood of experiencing marked functional impairment in workers with substantial stress symptoms in the absence of probable PTSD was nearly as elevated as this likelihood for workers with panic disorder alone (OR = 3.3; 95% CI, 2.7–4.0). Focusing attention only on probable psychiatric disorders markedly underestimates the full psychological burden and its social ramifications. It is likely that some of these individuals would benefit from appropriate treatment.
Most WTC workers reported one or more psychological/behavioral symptoms in their children during the time that they worked at the disaster site (Duarte et al. 2006
). Stuber et al. (2005)
and Schlenger et al. (2002)
reported similar findings of substantial WTC-related stress among children in New York City at the time of the disaster and for months afterward. Also consistent with previous research, WTC workers with probable PTSD were far more likely to report psychological symptoms and behavioral problems in their children compared with WTC workers without probable PTSD. Similar results were observed in victims of the Chornobyl reactor disaster (Bromet et al. 2002
Unlike most previous findings in civilian trauma survivors of mass disasters or individual traumatic events, the association between post-disaster PTSD and sex was not significant in the present study. Similar findings have been reported in samples of 655 urban police officers (21% female) and 207 exposed disaster workers (11.5% female) as well as in military populations (Sutker et al. 1995
). Pole et al. (2001)
suggested that selection and/or training factors may help to stress-inoculate women involved in police and military work.
The present study has a number of strengths, including use of standardized assessment instruments, comprehensive psychosocial and medical evaluation, inclusion of both males and females, ethnic diversity of subjects, and large cohort size. In an empirical review of the scientific literature from 1981 to 2001 on disaster victims, Norris et al. (2002)
noted problems with small sample sizes as well as demographic and ethnic homogeneity. During that 20-year period, the median sample size for studies related to psychosocial adjustment after disasters was only 159 subjects.
Study limitations include the use of self-administered rather than clinician-administered questionnaires, variability in time to presentation, potential inaccuracy of recall with the passage of time, possible under-reporting of psychological symptoms due to stigma, and lack of assessment within the first few months after 9/11. Because the earliest assessments occurred at least 10 months after the attacks, it is not possible to differentiate delayed onset versus chronic PTSD, and it is not possible to accurately determine rates of acute PTSD. Our self-selected cohort is also a limitation in that we do not know whether workers with psychological symptomatology were more or less likely to enroll. Also, degree of psychological symptomatology may be related to the presence of physical symptoms, and it may be that those with physical illnesses were more likely to seek medical monitoring. Our ability to generalize from this self-selected cohort is enhanced to some extent by its large size, but generalizability is hampered without knowing the true number of WTC rescue and recovery workers nor their sex, race, or ethnicity. Our estimate of 40,000 at-risk workers is considerably lower than the nearly 92,000 estimated by the World Trade Center Registry (Murphy et al. 2007
), largely because of the registry’s less stringent criteria for eligibility into their recovery worker cohort. The registry requires a worker or volunteer to have spent one shift at a WTC site between 11 September 2001 and 30 June 2002. The true number of workers and volunteers undoubtedly falls between the two estimates.
Keeping such limitations in mind, the prevalence of probable PTSD among workers in the present study is comparable to that seen in airline-crash recovery workers 13 months after the event (Fullerton et al. 2004
) and that observed by North et al. (2002)
in firefighters interviewed 34 months after the Oklahoma City bombing.
The present study has a number of implications for public health. Persistent post-disaster mental illness from 10 months to 5 years after the disaster in this cohort underscores the need for long-term mental health screening and treatment programs targeting this population. Chronic mental health disorders constitute a major public health concern. In this cohort, alcohol problems and impairment in occupational, social, and family life were strongly associated with diagnoses of probable PTSD, depression, or panic disorder alone and even with a large number of workers who did not meet criteria for a probable psychiatric disorder but who nevertheless experienced trauma-related psychological symptoms. It is particularly important to screen for comorbid conditions because in our population comorbid conditions were common, and those with these comorbid conditions were at far greater risk for alcohol problems and social dysfunction. The presence of comorbidity may also affect long-term outcome and response to treatment (McFarlane 2002
Psychiatric disability has effects far beyond the personal suffering of the individual and his or her immediate family. For example, PTSD has substantial economic costs from workdays lost and suboptimal performance. It is estimated that PTSD is associated with approximately 3.6 days of work impairment per month, on average (Breslau et al. 1998
). Further, the National Comorbidity Study found that PTSD was associated with marital instability and increased unemployment (Kessler 2000
). Taken together, our findings indicate that a substantial public mental health burden exists in the responder population, which puts them at risk for a variety of adverse health and social consequences.
Unfortunately, once psychopathology, such as PTSD, becomes chronic in nature, it can be difficult to treat. The military, recognizing the high frequency of posttraumatic stress symptoms in response to dangerous and life-threatening situations, as well as the costly effects of these symptoms on psychological well being and performance, has recently instituted periodic behavioral health evaluations on all troops returning from Iraq and Afghanistan. Stigma is reduced by requiring every returning soldier to participate in these evaluations. Similarly, rescue and recovery workers after future environmental disasters would likely benefit from routine behavioral health evaluations that are fully integrated into medical evaluations, as well as early treatment when appropriate. Although such efforts should help reduce chronicity of the mental health sequelae of disaster exposure, long-term provision of accessible mental health services for rescue and recovery workers likely should still constitute part of future disaster planning.
Finally, it will be essential in future environmental disasters to understand that mental health problems will almost certainly accompany effects of toxic exposures on physical health. It is also essential that accurate records be kept of the rescue and recovery cohorts so that postdisaster outreach efforts can be improved and better estimates of injury, illness, and disability can be made. Additional rigorous research is needed to better understand and modify the impacts on health of the physical and psychological risk factors that associated with work after environmental disasters (North 2004