|Home | About | Journals | Submit | Contact Us | Français|
To estimate the impact of Hurricane Katrina on mental illness and suicidality by comparing results of a post-Katrina survey with those of an earlier survey.
The earlier survey interviewed 826 adults in the Census Divisions affected by Katrina. The post-Katrina survey interviewed a new sample of 1043 adults who lived in the area prior to the hurricane. Identical questions were asked about mental illness and suicidality. The post-Katrina survey also assessed several dimensions of post-traumatic personal growth (e.g., increased closeness to loved one, increased religiosity). Outcome measures were the K6 screening scale of serious (SMI) and mild-moderate (MMI) mental illness and questions about suicide ideation, plans and attempts.
Post-Katrina respondents had significantly higher estimated prevalence of SMI (13.8% vs. 6.1%, χ21 = 24.6, p < .001) and MMI (21.3% vs. 9.7%, χ21 = 33.8, p < .001) than respondents in the earlier survey. Suicidal ideation and plans given estimated mental illness were significantly lower in the post-Katrina survey (0.6% vs. 8.4%, χ21 = 14.3, p < .001 ideation; 0.2% vs. 3.6%, χ21 = 13.3, p < .001 plans). This lower suicidality was strongly related to two dimensions of post-traumatic personal growth (increased sense of meaning and purpose in life, realization of inner strength), without which between-survey differences in suicidality were insignificant.
Despite the estimated prevalence of mental illness doubling after Hurricane Katrina, suicidality was unexpectedly low. The role of post-traumatic personal growth in ameliorating the effects of trauma-related mental illness on suicidality warrants further investigation.
Hurricane Katrina was the deadliest US hurricane in seven decades and the most expensive natural disaster in US history. Over 500,000 people were evacuated. Nearly 90,000 square miles were declared a disaster area (roughly equal to the land mass of the United Kingdom).1 More than 1,600 confirmed deaths occurred and over 1,000 others remain missing.2 The destruction caused Katrina has lingered much longer than in previous hurricanes.3
An extensive literature documents adverse mental health effects of natural disasters.4, 5 Although these effects vary greatly, the effects of catastrophic disasters are consistently large.6, 7 For example, studies after Hurricane Andrew found 25–50% of respondents afflicted with disaster-related mental disorders.8, 9 Based on these results, and given the extraordinary array of stressors that occurred in conjunction with Hurricane Katrina (e.g., bereavement, exposure to the dead and dying, personal threat to life, massive destruction),10–12 we would expect the mental health effects of Katrina to be at the upper end of the range of previous disasters.
Due to the wide geographic dispersion of the displaced population, comprehensive mental health assessment of Katrina survivors is nonexistent. The Louisiana Department of Public Health documented substantial psychopathology among the 50,000 Katrina survivors cared for in evacuation centers (ECs) shortly after the hurricane,13 but these individuals represented fewer than 1% of survivors. Seven weeks after the hurricane, CDC carried out a household needs assessment survey that found half of adults still living in New Orleans to have clinically significant psychological distress,14 but no information was obtained on the much larger number of pre-hurricane residents who no longer lived in New Orleans. Two public opinion polls, one carried out jointly by Gallup, CNN, and USA Today in a sample of people who sought American Red Cross (ARC) assistance,15 and the other carried out by the New York Times in a sample from the ARC safe list,16 asked a handful of questions about mental health, but without attempting to assess clinical significance. A probability survey of families with children still residing in FEMA-sponsored trailers or hotel rooms in Louisiana as of mid-February, 2006 found 44% of adult caregivers to have clinically significant psychological distress.17 As with the earlier CDC EC survey, though, the sampling frame represented less than 1% of pre-hurricane residents of the affected areas.
Public health decisions cannot be based on such a narrow empirical foundation. The current report presents initial results of an ongoing tracking survey designed to provide broader coverage of the population affected by Hurricane Katrina. The first phase of the study aimed to enroll and carry out a baseline mental health needs assessment survey with a representative sample of 1000 adult (ages 18+) pre-hurricane residents of the FEMA-defined Hurricane Katrina impact areas in Alabama, Louisiana, and Mississippi.18–20 Subsequent phases of the study will monitor evolving needs in follow-up surveys. The focus of this report is on the effects of the hurricane on the prevalence and correlates of mental illness and suicidality. Before-after comparisons are approximated by using baseline data from a 2001–03 national survey that included a probability sub-sample of respondents in the two Census Divisions subsequently affected by Katrina.21 The questions used to assess mental illness and suicidality were identical in the two surveys.
The baseline survey was the National Comorbidity Survey Replication (NCS-R)21, a face-to-face survey of English-speaking adults (ages 18+) administered between February 2001and February 2003. The NCS-R interviewed 826 people in the two Census Divisions affected by Hurricane Katrina. The response rate was 70.9%. The NCS-R data were weighted to adjust for differential probabilities of selection and for residual discrepancies between the sample and the 2000 Census on a series of social, demographic and geographic variables. The NCS-R design is discussed in more detail elsewhere.22
The post-Katrina survey was administered between January 19 and March 31, 2006 to the telephone probability sample of 1043 people enrolled in the Hurricane Katrina Community Advisory Group (CAG). The survey began with a brief series of screening questions followed by a request for eligible respondents to participate in the CAG. Potential members were informed that CAG participation would require completing quarterly tracking interviews over a period of two years to monitor the ongoing needs of hurricane survivors. The 1043 respondents who agreed to join the CAG were then administered the baseline CAG telephone interview.
The post-Katrina survey screening response rate was 64.9%. This is lower than in typical household surveys due to problems tracing and contacting eligible respondents. The CAG participation rate among screening survey respondents was 41.9%. This relatively low participation rate is presumably due to the high time commitment required for CAG membership. It is noteworthy that both the subjectively rated amount of stress caused by Katrina and the amount of psychological distress experienced in the month before interview were significantly higher among screening survey respondents who declined to join the CAG than those who joined. A number of those who declined to join told interviewers explicitly that their experiences during and after Katrina were so painful to recall that they felt psychologically unable to relive those experiences by participating in the CAG interviews.
The target population for the CAG was English-speaking adults (ages 18+) with pre-hurricane residences in the areas defined by FEMA as affected by Katrina (4,137,000 adult residents in the 2000 Census) in either of two sampling frames: a random-digit dial (RDD) frame that included telephone banks working in the eligible counties-parishes prior to the hurricane; and a frame that included telephone numbers of the roughly 1.4 million families that applied for ARC assistance after Katrina. Pre-hurricane residents of the New Orleans Metropolitan Area were over-sampled in both frames. Many dislocated people were traced in the RDD sample because phone calls were forwarded to new addresses. The ARC sample also included cell phones. The small proportion of evacuees still living in hotels at the time of the survey was represented through a supplemental sample of hotels that housed FEMA-supported evacuees.
Overlap of the two sampling frames was handled in two ways: by confining numbers from the ARC sample to those not in the RDD frame (i.e., cell phones and exchanges outside the hurricane area); and by down-weighting RDD respondents who reported ARC assistance and had additional phone numbers outside the RDD frame. Respondents from the two frames were combined by weighting the participating ARC households to their estimated population proportion based on estimates of the proportion of ARC numbers outside the RDD frame and the proportion of RDD respondents that asked for ARC assistance. Hotel sample respondents were included without a household weight, as they were selected proportionally.
The sample of 1043 CAG members was weighted to adjust for significant differences between screening survey respondents who declined to participate in the CAG and those who participated in measures of perceived stress, psychological distress, and a range of social and demographic variables. A within-household probability of selection weight was also used to adjust for the fact that only one random respondent was invited to join the CAG in each sample household. A post-stratification weight was also used to adjust for residual discrepancies between the sample and data from the 2000 Census for the affected areas on a range social, demographic, and pre-hurricane housing variables. The consolidated weight, finally, was trimmed to increase design efficiency based on evidence that trimming did not significantly affect outcome variable prevalence estimates.
The K6 scale of nonspecific psychological distress23, 24 was used to screen for 30-day DSM-IV anxiety and mood disorders. The K6 is by far the most widely used mental health screening scale in the US.25, 26 Based on previous K6 validation,24 scores on the 0–24 scale in the range 13–24 were classified probable serious mental illness (SMI), those 8–12 were classified probable mild-moderate mental illness (MMI), and those 0–8 were classified probable non-cases. A small clinical reappraisal study was carried out with five respondents selected randomly from each of the three categories (SMI, MMI, non-case). A trained clinical interviewer administered the non-patient version of the Structured Clinical Interview for DSM-IV (SCID)27 blindly to these 15 respondents. The syndromes assessed were DSM-IV Major Depressive Episode, Panic Disorder, Generalized Anxiety Disorder, Post-Traumatic Stress Disorder, Agoraphobia, Social Phobia, and Specific Phobia. SMI was defined as a DSM-IV diagnosis with a Global Assessment of Functioning (GAF)28 of 0–60 and MMI as a DSM-IV diagnosis with a GAF of 61+. K6 classifications were confirmed for 14 of 15 respondents, the exception being a respondent classified SMI by the K6 but MMI by the SCID (based on a GAF of 65). Suicidality was assessed with questions about lifetime occurrence, age of onset, and recency of suicide ideation (ever seriously thought about committing suicide), suicide plans, and suicide attempts. Respondents were classified first-onset cases if they reported 12-month prevalence (the most recent time frame assessed in the NCS-R) and no lifetime history of the suicidal behavior as of the year prior to their current age.
Socio-demographic correlates included age, sex, race/ethnicity, family income, education, marital status, and employment status. Income was coded into low (below the federal poverty line for the size/composition of the respondent’s family), low-average (no higher than the population median), high-average (up to three times the median), and high (above three times the median). The final measures assessed dimensions of post-traumatic personal growth found in previous research to occur after trauma exposure and to facilitate psychological adjustment by making sense of trauma or finding some positive aspect of the trauma.29, 30 We focus on five such dimensions based on their presence in both of the two most commonly used inventories of post-traumatic personal growth:31, 32 post-traumatic increases in (i) emotional closeness to loved ones, (ii) faith and trust in people, (iii) spirituality or religiosity, (iv) meaning or purpose in life, and (v) recognition of inner strength or competence.
Differences in estimated prevalence of mental illness and suicidality were compared in the two samples. Socio-demographic variation in between-survey differences was assessed in pooled logistic regression equations predicting outcomes from a 0–1 variable for survey (0 = NCS-R, 1 = post-Katrina survey), the socio-demographic variables, and interactions between survey and socio-demographic variables. The role of post-traumatic growth was examined in subgroup analysis. Because the surveys featured weighting (both surveys) and geographic clustering (NCS-R), analyses used the Taylor series linearization method.33 Multivariate significance was calculated using Wald χ2 tests based on design-corrected coefficient variance-covariance matrices. Statistical significance was evaluated using two-sided .05 level tests.
The proportion of respondents estimated to have SMI is significantly higher in the Katrina sample than the NCS-R (13.8% vs. 6.1%, χ21 = 24.6, p = .040). The same is true for estimated MMI (21.3% vs. 9.7%, χ21 = 33.8, p < .001) and any estimated mental illness (35.1% vs. 15.7%, χ21 = 64.4, p < .001), with odds-ratios (ORs) in the range 2.5–2.9. (Table 1) The between-survey difference in suicidality is not significant either for ideation (2.5% vs. 2.8%, χ21 = 0.1, p = .75), plans (0.5% vs. 1.1%, χ21 = 0.9, p = .35), or attempts (1.0% vs. 0.6%, χ21 = 0.4, p = .52).
As 12-month suicidality was reported almost entirely by people estimated to have mental illness in both samples (results available on request), the higher estimated prevalence of mental illness but not suicidality implies that the conditional prevalence of suicidality given estimated mental illness is lower in the Katrina sample than the NCS-R. More detailed analysis found this especially true for first onset of suicidality in the past year among respondents estimated to have mental illness. (Table 2) These differences are significant for ideation (0.6% vs. 8.4%, χ21 = 14.3, p < .001) and plans (0.2% vs. 3.6%, χ21 = 13.3, p < .001), but not attempts (1.8% vs. 2.3%, χ21 = 0.1, p = .77).
Significant socio-demographic correlates of estimated SMI in the Katrina sample include female gender, low education, pre-hurricane not married, and pre-hurricane “other” employment status (mostly unemployed or disabled). (Table 3) None of these associations differs significantly in the post-Katrina sample compared to the NCS-R. No significant between-survey differences were found in predictors of estimated MMI or any mental illness. (Results available on request.)
Suicide ideation given estimated mental illness was the focus of subsequent analysis due to suicide plans and attempts being too uncommon to study with adequate statistical power. The only statistically significant socio-demographic correlate of ideation was age less than 40. (Table 3) None of the socio-demographic correlates differs significantly in the two samples.
A majority of post-Katrina survey respondents reported post-traumatic growth: becoming closer to their loved ones (81.2%), developing more faith and trust in others (68.4%), becoming more spiritual or religious (64.7%), finding deeper meaning and purpose in life (78.3%), or discovering inner strength (82.3%). (Table 4) None of these is strongly related to estimated mental illness
Two of the five dimensions are significantly related to low suicide ideation among people estimated to have mental illness: purpose in life and inner strength. These two specify the comparatively low prevalence of suicide ideation among post-Katrina survey respondents. (Table 5) Specifically, the lower prevalence of suicide ideation in the post-Katrina sample than the NCS-R is limited to those who reported either increased purpose in life or increased inner strength. The odds of suicidal ideation among post-Katrina respondents with neither of these cognitions do not differ significantly from those of NCS-R respondents (1.0).
The two-survey comparison method is an inexact way to estimate the effects of Hurricane Katrina because the surveys differed in sample frame (all households in two Census Divisions vs. contactable telephone households in areas within these Divisions affected by Katrina), data collection mode (face-to-face vs. telephone interviews), and response rates. An additional limitation concerns the K6. Although good concordance with clinical interviews has been consistently documented in published reports23, 24, the K6 is merely a screen and not a clinical interview.
Notwithstanding these limitations, the fact that estimated prevalence of SMI and MMI more than doubled after Hurricane Katrina is consistent with other evidence of adverse mental health effects of major disasters.34, 35 The socio-demographic correlates are also largely consistent with previous research.36, 37 That the associations of socio-demographic predictors did not differ across samples suggests that the adverse mental health effects of Katrina were equally distributed across broad segments of the population. Although analysis of treatment patterns goes well beyond the scope of this report, these results document high and widely dispersed need for mental health treatment.
Our most striking finding is the lower conditional likelihood of suicidality among persons estimated to have mental illness after Katrina. This finding is not unprecedented. A cross-national epidemiological survey of suicide ideation found the lowest prevalence in war-torn Beirut despite a higher prevalence of depression than in virtually any other study site.38 While post hoc methodological interpretations can be constructed (e.g., that the mental illness associated with exposure to trauma might have a lower intensity not detected by standard measures), they seem implausible in light of independent evidence that severity and impairment of mental illness are similar after disasters compared to other times.39, 40
A more plausible explanation is that the effects of increased mental illness were offset by protective factors activated by the hurricane. Although this possibility has not previously been studied, post-traumatic personal growth in such areas as self-efficacy,41 optimism,30 hope,42 and perceived social support43 have been documented after disasters and these changes have been linked to low post-disaster distress.44 Our findings go beyond these earlier results to suggest that some dimensions of post-traumatic personal growth might be protective against suicidality among people with clinically significant psychological distress. It is noteworthy that the indicators of post-traumatic growth were not strongly related to SMI or MMI, which means that a great many Katrina survivors are, understandably, depressed by their loses and anxious about their uncertain future. However, the suicidality often associated with these syndromes in the general population is much lower among people in the CAG sample who were able to develop a sense of new purpose and meaning and inner strength in the wake of the hurricane. The causal processes underlying this pattern presumably involve the creation of positive future orientations that provide psychological scaffolding protecting against the suicidality often associated with extreme distress. Although processes of this sort have long been discussed in the psychoanalytic literature,45, 46 the current study is, to our knowledge, the first to provide quantitative evidence regarding such a pattern in an epidemiological sample of a disaster population.
This finding suggests that more systematic investigation of post-traumatic personal growth might be useful in guiding public health mass media efforts in the aftermath of future disasters. Research has suggested that public health messages play an important part in affecting psychological reactions to disasters.47–49 Promotion of positive cognitions might be an important pathway for these effects. Systematic research to explore this possibility is needed. In a more immediate way, the finding documents a psychological strength in the population affected by Katrina that is at least temporarily linked to an unexpectedly low prevalence of suicidality. It is important for public health officials to recognize, though, that this low suicidality might be temporary. For example, if the feelings of purpose, meaning, and inner strength reported by so many respondents are linked to an expectation that the practical problems of living created by the hurricane will soon be solved, and if these expectations are not met as time goes on, one could imagine the positive cognitions eroding and being replaced with a sense of hopelessness that, in the presence of the high levels of estimated mental illness found here, could lead to a substantial increase in suicidality. The finding of low suicidality, then, should be considered evidence of a short-term postponement rather than of a permanent absence of suicidality in this population.
Funding/Support: The Hurricane Katrina Community Advisory Group is supported by the National Institute of Mental Health (R01 MH070884-01A2), with supplemental support from the Federal Emergency Management Agency. The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; and preparation, review, or approval of the manuscript.