Many people are exposed to traumatic events; however, the proportion of those who develop post-traumatic stress disorder (PTSD) is lower than the proportion of those who have been exposed (Breslau et al., 1998). Some distress after an abnormal event is considered normal, but often this is followed by a rapid return to predisaster functioning (Bonanno et al., 2004).
Interest in resilience and recovery has increased among researchers. Resilience seems to be better characterized as adaptability rather than stability (Norris et al., 2009) and more about “bouncing back” from harm rather than immunity from harm (Layne et al., 2007). Psychological resilience is suggested to be differentiated from recovery (Bonanno et al., 2004), with recovery involving longer periods of distress, followed by gradual return to pre-event functioning.
Development of post-traumatic stress reactions is related to severity of trauma exposure (Bonanno et al., 2006; Holen, 1990; Johannesson et al., 2009; North et al., 2004). Over the time, both exposure factors and additional trauma exposure need to be considered (Boscarino and Adams, 2009; McFarlane and Van Hooff, 2009). Predictors of symptoms can include ongoing stressors, other traumatic events, lower income (Cerda et al., 2008), poor mental health, female gender, and low education (Brewin et al., 2000; Galea et al., 2007). Older age and state of being married seem to be protective factors (Bonanno et al., 2007).
Post-traumatic stress reactions are prominent during the first weeks of postdisaster, but decreasing during the subsequent year (Ford et al., 2007; North et al., 2002; Shalev, 2002), with continued attenuation during the second year (Norris and Sloane, 2007). It has been stated that one-third of those who develop PTSD do not remit, even after many years (Norris and Sloane, 2007), although this opinion has been challenged (Kessler et al., 1995).
There are few studies of long-term effects after disasters and the results are conflicting. Fifty years after a Greek earthquake (Lazaratou et al., 2008), three-quarters of the victims reported a continued impact of the earthquake on their lives. Male survivors of an oilrig disaster (Holen, 1990) had signs of psychological distress 5 years later, compared with controls. In a 10-year disaster follow-up, 21% of oil platform respondents had PTSD compared with 73% in the first 3 months after the disaster (Hull et al., 2002). Twenty-five years after a hotel fire few survivors had signs of distress, even though 21% indicated that the fire still had an impact on daily life (Lundin and Jansson, 2007). Heightened autonomic reactivity to trauma reminders may persist for many years in highly exposed survivors (Tucker et al., 2007).
Many long-term follow-up studies focus on survivors residing in the affected areas (Green et al., 1990; Norris et al., 2004; Ohta et al., 2003; van den Berg et al., 2008; Tang, 2007; Tsai et al., 2007). Long-term follow-up of large samples of disaster-exposed nonresident survivors, who were able to return to an intact domestic society, is lacking.
The 2004 tsunami badly affected Southeast Asia, where over 227,000 people perished (Telford et al., 2006). According to Swedish authorities, about 19,000 Swedish tourists traveled in Southeast Asia at that time, and approximately 7000 were in the most affected areas. In terms of number of victims, Sweden was one of the most affected countries in Europe (Telford et al., 2006). The aim of this study was to examine how a well-defined tourist population was affected by the 2004 tsunami by longitudinally measuring symptoms and rates of recovery. We hypothesized that post-traumatic stress reactions and affected mental health would be reduced after 3 years, compared with 14 months, but are still prominent in those exposed to life threat and in the bereaved.
Returning Swedish citizens from Southeast Asia were registered at Swedish airports during the first weeks after the disaster. Individuals 16 years of age or older (n = 10,501), accounting for 77% of those registered, were invited by mail to participate in the study. Youngsters, 16 to 18 years, needed written consent from caregivers. A comparison between participants and nonparticipants revealed no difference in distribution of age or gender. The sample represented inhabitants of urban and rural areas, including major metropolitan areas.
A questionnaire was mailed 14-months postdisaster (T1), followed by one reminder to nonresponders. The 4910 (49%) individuals who responded to the T1 questionnaire were invited for a 3-years follow-up (T2). The T2 questionnaire included similar items as in the T1 study, with addition of questions concerning prior health conditions. The study was approved by the Regional Ethical Vetting Board.
Variables used were as follows: gender; age; family situation; education; previous mental problems; previous trauma experiences in childhood, adulthood, and post-tsunami; experience of social support; and reactions afterward. Age was coded into 4 categories as follows: 16–24 years, 25–40 years, 41–60 years, and 61+ years. Gender and family situation (married/partner or no partner) and education (≤12, >12 years) were coded as binary variables. Traumatic events were classed as follows: experienced or not during childhood (0–16 years), adulthood (>16 years of age and until the tsunami), and until T2. In all, 13 types of following traumatic events were included: accidents; disasters; war/terror; violence/abuse; own serious illness/injury; relative's serious injuries/illnesses; serious family conflicts/divorce; parents' divorce; death of parents, siblings, own children, partners, and/or other family members. Previously experienced traumatic events were coded into 3 categories (0, 1–2, and ≥3 trauma experiences). Prior psychiatric problems were defined as present, if the disturbance had negatively affected general functioning or had required treatment and if either of the following alternative questions were endorsed: (a) Have you ever in life, before the tsunami disaster, felt depressed or in a low mood, or had feelings of hopelessness, for more than 2 weeks or longer? (b) Have you ever in life, before the tsunami disaster, had problems with panic reactions or persistent anxiety or anguish for more than 4 weeks?
Social support was measured by following 1 question: “Are you, on the whole, pleased with the social support you have received after the tsunami?” The respondents rated contentment on a 7-point Likert scale, ranging from 1 (never) to 7 (always).
The Impact of Event Scale-Revised (IES-R) (Weiss, 2004), a well-validated scale for estimation of post-traumatic stress reactions, was used. The respondents were instructed to evaluate how disturbing the symptoms had been during the past 7 days, on a 5-point Likert scale ranging from 0 (not at all) to 4 (extremely disturbing).
The General Health Questionnaire 12 (GHQ-12) (Goldberg et al., 1997), similarly well-validated, was used to evaluate the respondents' mental health during the weeks before participating. The GHQ-12 scale has an internationally accepted cut-off score of ≥3, indicating impaired mental health (Connor et al., 2006).
Suicidal ideation was measured by respondents' reports of active or passive suicidal thoughts during the last 12 months (yes/no).
Exposure was coded into following 3 categories: high exposure, medium exposure, and low exposure. High exposure included those who ticked one of the alternatives in the question: “Did you get caught by the wave?” The 3 response alternatives were as follows: (a) Yes, and I lost all physical control; (b) Yes, but I could keep some physical control; (c) No, but I was very close to being caught by the wave. Medium exposure included those who reported: “No, I was not in the neighborhood of being caught by the tsunami wave,” but experienced one or more of the following consequences of the disaster: loss of relatives, subjectively felt their life was threatened, physical injury to themselves or others, feelings of anxiety regarding the fate of relatives, perceived dead bodies, perceived others suffering, and/or perceived young children on their own, or had helped other victims. The low-exposure group included those who could have been indirectly exposed, e.g., by being in close vicinity or by talking to affected persons.
Loss included those who ticked loss of any relative in the tsunami (yes/no). Relatives included partner, children, parents, siblings, grandparents, or parents-in-law.
Demographic variables and outcome measures were checked for anomalies such as outliers, points of high influence, and nonrandom missing data. Among low exposed, only one individual reported losing a relative. This observation was removed from the analysis. Conclusions on loss are only applicable to high- and medium- exposure groups.
Response to at least 70% of the items in the IES-R was minimum for being included in the analyses. When measuring prevalence of severe post-traumatic reactions, the IES-R mean item score was set to ≤1.89 (Weiss, 2004, pp 168–189), an equivalent to a sum score of ≤41.6. Thus, at T2 in relation to T1, following 4 trajectories of traumatic stress reactions were defined: resilient = below cut off at both time points; remitted = above cut off at T1 and below cut off at T2; delayed = below cut off at T1 and above cut off at T2; persistent = above cut off at both T1 and T2.
Negative binomial regression with log link was used to estimate the association between independent variables and IES-R sum score at T2. This was done because the IES-R sum had a count distribution with no negative values and was heavily right skewed with a variance greater than the mean. To estimate the association between independent variables and the odds of affected mental health and suicidal ideation at T2, logistic regression was used. For all analyses the variable “social support” was treated as a continuous covariate; all other independent variables were entered as factors.
Generalized estimating equations (GEE) analysis with an unstructured covariance matrix was used to estimate association between independent variables and change in outcome measures over time, while considering correlation between time points. When IES-R sum was the dependant variable, negative binomial distribution with log link was observed. For the dichotomized GHQ-12 measure and for suicidal ideation, binomial distribution with logit link was used. All standard errors were estimated empirically unless noted otherwise. The independent variables were the same as in the analysis of outcomes in T2, but with following 2 additions: the time point for the measurement (T1 and T2, were coded as 0 and 1) was included as an independent variable. For each independent variable, an interaction term of time and the independent variable were included. The interaction terms were removed stepwise by a modified manual backwards elimination procedure: the interaction term with the highest type III p value was removed in each step, until only interaction terms with a Type III p value of <0.05 remained. However, the interaction terms exposure*time and loss*time were not removed even if they had a type III test p > 0.05.
Proc Genmod and Proc Logistic in SAS software 9.2 (SAS, 2002) were used for the analyses.
Responders Versus Nonresponders
The response rate at T2 was 70% (n = 3457) of original sample. Women responded to a somewhat higher degree (74% vs. 66%). Being educated more than 12 years resulted in a higher response rate (74% vs. 68%) than lower education. Respondents were older than nonresponders (median age 43 vs. 39 years). High exposure resulted in higher response rate (74%) than medium (69%) and low exposures (66%). No differences between responders and nonresponders for post-traumatic stress reactions, affected mental health, or suicidal ideation, as measured at T1, were discerned, nor was there a difference in response rate between bereaved and nonbereaved.
Mean age of the sample was 42.3 years, 44% were educated more than 12 years, 75% were married/cohabiting, and 87% were employed. Of the respondents, 58% were women. Previous psychiatric problems during lifetime were reported by 23%. The majority of the respondents (96%) had been on vacation in Southeast Asia. The general satisfaction with social support after the tsunami was high (median = 6 on a 7-point scale).
Trajectories of Post-traumatic Stress Reactions
Following 4 trajectories were identified: resilient, remitted, delayed, and persistent. Of the survivors identified as resilient (Fig. 1), 76.6% were high-exposed, 91% medium-exposed, 98.2% low-exposed, and 51.9% belonged to the bereaved group. Of the survivors identified as remitted, 11.9% were high-exposed, 6% medium-exposed, 1.5% low-exposed, and 26% belonged to the bereaved group. Of the survivors identified as delayed, 3% were high-exposed, 1% medium-exposed, 0.1% low-exposed, and 3% belonged to the bereaved group. Among those identified with persistent reactions, 8.5% were high-exposed, 1.9% medium-exposed, 0.2% low-exposed, and 19.1% belonged to the bereaved group.
Post-traumatic Stress Reactions at T2
Exposure and loss were associated with higher levels of post-traumatic stress reactions at T2 (Table 1). In the high-exposure group, estimated log count difference was 1.15 (p = <0.0001), corresponding to approximately 3 times the score of the low-exposure group. In the medium-exposure group, the estimated log count difference was 0.71 (p = <0.0001) corresponding to approximately twice the score of the low-exposure group. In the bereaved group, the estimated log count difference was 0.62 (p = <0.0001), corresponding to 86% higher score than those with no loss.
Change in Severity of Post-traumatic Stress Reactions
Post-traumatic stress reactions decreased from T1 to T2 (Fig. 2) in the 3 exposure groups (p = <0.0001). However, high-exposed generally recovered from higher levels of post-traumatic stress reactions.
The GEE analysis revealed that the predicted level of post-traumatic stress reactions was lower for all 3 exposure groups at T2 than at T1. The proportional recovery, expressed as regression lines, was not as large among high-exposed (estimate of log count difference between time points = −0.36, SE = 0.018, corresponding to 30% lower IES-R at T2) and medium-exposed group (estimate = −0.44, SE = 0.027, corresponding to 35% lower IES-R at T2) as among low-exposed (estimate = −0.55, SE = 0.042, corresponding to 42% lower IES-R at T2). Loss of relatives predicted a further slower recovery (Fig. 3). High- and medium-exposed bereaved were predicted on average to have 0.09 log IES points more (9%) at T2 than their nonbereaved counterparts.
General Mental Health at T2
Descriptive data of affected mental health and suicidal ideation at T1 and T2 are presented in Table 2.
High and medium exposure increased the odds for affected mental health at T2 (p < 0.0001; Fig. 4). In addition, loss increased the odds similarly.
Suicidal ideation at T2 was increased through high exposure and loss.
Satisfaction with social support was associated with decreased risk for affected mental health and suicidal ideations.
Change in General Mental Health
The GEE analysis revealed no change in odds of affected mental health at T2 compared with T1 for low-exposed respondents. However, for high- (odds ratio [OR] = 0.60, confidence interval [CI] = 0.44–0.82) and medium-exposed (OR = 0.61, CI = 0.44–0.86), these odds had decreased at T2. A similar result was found for bereaved: a further decreased risk was found at T2, compared with T1 (OR = 0.39, CI = 0.23–0.64).
Among demographic variables, younger age (16–24, 24–40) predicted increased risk of mental health impairment from T1 to T2 (16–24 years, OR = 1.56, CI = 1.13–2.14: 25–40 years, OR = 1.30, CI = 1.00–1.69) as did exposure for 1 to 2 post-tsunami traumas (OR = 1.34, CI = 1.06–1.70).
Three years after the tsunami, compared with 14 months, we found a pattern of resilience and recovery among survivors. However, those severely exposed still had increased post-traumatic stress reactions and affected general mental health. There was a decrease of symptoms in all survivors, although the rate of recovery was slower among respondents exposed to life threat and among bereaved.
The studied sample was healthy and socioeconomically stable. The majority was employed and almost a half had a college/university education. Three quarters of them were married or cohabiting.
Resilience is common in highly exposed individuals (Bonanno et al., 2004), which our findings also suggest. Of those exposed to life threat during the tsunami, 12% had aggravated post-traumatic stress reactions 3 years later. This can be compared with 3% of those exposed to other consequences of the tsunami, as opposed to bereaved survivors, all high- or medium-exposed, which was approximately 21%, a lower proportion than reported by Neria et al. (2007).
Trauma Exposure, Bereavement, and Post-traumatic Stress Reactions
Exposure was associated with increased levels of post-traumatic stress reactions even 3 years after the disaster. The high-exposed had approximately 3 times the IES-R scores of the low-exposed, and medium-exposed had 2 times the IES-R scores of the low-exposed. This indicates a dose-response relationship, which is in agreement with previous studies (Bonanno et al., 2006; Goenjian et al., 2005; North et al., 2004). Bereavement had a similar association where bereaved had 86% higher score than those with no loss, confirming previous findings (Chou et al., 2007).
Higher levels of symptoms in exposed subjects at T1 rendered difficulties to compare the change between T1 and T2 in different exposure groups. Therefore, the proportional change of scores from T1 to T2 was chosen for analyses. The comparison of symptoms between T1- and T2-indicated signs of recovery from post-traumatic stress reactions in all 3 exposure groups which is in line with other studies (Green et al., 1990). High-exposed (30% estimated recovery) and medium-exposed (35% estimated recovery) groups did not recover as much as low-exposed (42% estimated recovery) group. Hence, the effect of exposure seemed to slow the process of recuperation. These data might be consistent with a model of heightened amygdala reactivity after high-intensity trauma exposure, with relatively slow recovery (Ganzel et al., 2007; Tucker et al., 2007). Similarly, bereaved survivors, starting from higher levels, recovered proportionally slower than nonbereaved. The additional burden of bereavement could complicate recovery, consistent with other findings (Chou et al., 2007).
High exposure was associated with psychological distress, with 28% presenting signs of affected general mental health at T2, compared with 20% of medium-exposed. For the bereaved, corresponding figure was 43%, indicating the profound effect of bereavement after 3 years. Suicidal ideation was modest in all groups: 12% in the high-exposed, 8% in the medium-exposed, and 6% in the low-exposed. However, 23% of the bereaved, reported suicidal ideation the year prior to T2.
In contrast to post-traumatic stress reactions, the level of change in general mental health appeared modest, although there was a positive progress toward recovery among high- and medium-exposed respondents. Distress was lower than previous findings among earthquake survivors (Ohta et al., 2003). Possibly, the examined sample had higher level of functioning and was less vulnerable than an average population. Low-exposed demonstrated a small increase in reactions (from 10% to 13.5%), a change that could be explained by the effect of age and post-tsunami stressful events. Similarly, medium-exposed had decreased odds although the unadjusted proportions were the same between the 2 time points, also explained by age and post-tsunami events. Bereaved respondents had a decreased risk at T2, although the effect on this group was still substantially higher (43%) than for nonbereaved (21%). In contrast to Chou et al. (2007), no change was found in prevalence of suicidal ideation for different exposure conditions: bereaved respondents had decreased prevalence between the 2 time points, indicating a process toward recovery.
Factors Associated With Long-term Trauma Reactions
Among demographic characteristics, female gender, lower education, previous psychiatric illness, and further traumatic experiences after the tsunami were associated with increased post-traumatic stress reactions at T2, which concurred with findings after the 2001 World Trade Centre attack (Boscarino and Adams, 2009). Demographic variables did not significantly affect the rate of recovery from T1 to T2, suggesting other factors being more important in the trajectories toward recovery; alternatively, the analyses were not sufficiently powerful to detect the impact. North et al. (2002) suggest that female gender and previous trauma do not affect recovery over time. In our study, young age (16–24, 25–40 years), and traumatic experiences post-tsunami increased the odds of adverse outcome and affected general mental health over time.
Previous psychiatric problems and previously experienced traumatic life events were associated with higher levels of long-term trauma reactions and higher risk for psychological distress, in line with previous studies (Brewin et al., 2000; Boscarino and Adams, 2009). However, it is difficult to make any statement about the causality between effects of additional traumas and higher levels of reactions without knowledge of relationship between earlier life patterns, personality, and previous disorders.
Factors Associated With Recovery
A factor positively associated with recovery at both the time points was satisfaction with social support, confirming social support as an important factor for remission (Norris and Kaniasty, 1996). Other factors were middle age, being married/cohabiting, male gender, and higher level of education, confirmed also in other studies (Bonanno et al., 2007). Personality factors were not considered in the present study. However, it is reasonable to assume that the above-mentioned factors reflect underlying capacity such as satisfying social network and ability to use societal resources of support which might be helpful in the recovery process. Another positive factor could be that one-third of respondents had revisited the area, of which 90% reported this as important. This is in line with other findings (Heir and Weisaeth, 2006).
Limitations and Strengths
A limitation in this study is rather low response rate in the first (T1) survey. Low response rates are common in studies of disasters (Bonanno et al., 2006; McFarlane et al., 1997). However, nonresponse is suggested to be related more to lack of interest and lack of relevant experience than to fear of retraumatization (Hussain et al., 2009). Another limitation is the lack of systematic assessment of disaster survivors upon their return to Sweden, making an estimation of development of reactions from the acute phase to T1 not possible. A further limitation concerns the limited use of self-report questionnaires, as these are not clinical measures of mental health and PTSD. Length of time between the disaster and T1 should also be considered.
Notwithstanding these limitations, a major strength of this study was the large cohort of wide age range and the possibility to assess a homogenous, nonclinical population with different levels of trauma exposure. We could identify a large group traveling in the area, unexposed to the tsunami and consequently an appropriate comparison group. We could also analyze the long-term effects of extreme exposure to trauma in a presumably previously healthy and normal population, who could return to an intact domestic society.
This study highlights the long-term negative effects of severe exposure and traumatic loss, which appear to slow down recovery. Identification of symptoms and subsequent support or trauma-focused psychotherapy might facilitate optimal recovery.
The authors thank Jon Bisson, Cardiff University School of Medicine, United Kingdom, for valuable comments and constructive criticism.
Bonanno GA, Galea S, Bucciarelli A, Vlahov D (2006) Psychological resilience after disaster: New York City in the aftermath of the September 11th terrorist attack. Psychol Sci
Bonanno GA, Galea S, Bucciarelli A, Vlahov D (2007) What predicts psychological resilience after disaster? The role of demographics, resources and life stress. J Consult Clin Psychol
Bonanno GA, Wortman CB, Nesse RM (2004) Prospective patterns of resilience and maladjustment during widowhood. Psychol Aging
Boscarino JA, Adams RE (2009) PTSD onset and course following the World Trade Center disaster: findings and implications for future research. Soc Psychiatry Psychiatr Epidemiol
Breslau N, Kessler RC, Chilcoat HD, Schultz LR, Davis GC, Andreski P (1998) Trauma and posttraumatic stress disorder in the community: the 1996 Detroit Area Survey of Trauma. Arch Gen Psychiatry
Brewin C, Andrews B, Valentine J (2000) Meta-analysis of risk factors for posttraumatic stress disorder in trauma-exposed adults. J Consult Clin Psychol
Cerda M, Vlahov D, Tracy M, Galea S (2008) Alcohol use trajectories among adults in an urban area after a disaster: evidence from a population-based cohort study. Addiction
Chou FH, Wu HC, Chou P, Su CY, Tsai KY, Chao SS, Chen MC, Su TTP, Sun WJ, Ou-Yang WC (2007) Epidemiologic psychiatric studies on post-disaster impact among Chi-Chi earthquake survivors in Yu-Chi, Taiwan. Psychiatry Clin Neurosci
Connor KM, Foa EB, Davidson JR (2006) Practical assessment and evaluation of mental health problems following a mass disaster. J Clin Psychiatry
. 67(suppl 2):26–33.
Ford JD, Adams ML, Dailey WF (2007) Psychological and health problems in a geographically proximate population time-sampled continuously for three months after the September 11th, 2001 terrorist incidents. Anxiety Stress Coping
Galea S, Brewin CR, Gruber M, Jones RT, King DW, King LA, McNally RJ, Ursano RJ, Petukhova M, Kessler RC (2007) Exposure to hurricane-related stressors and mental illness after Hurricane Katrina. Arch Gen Psychiatry
Ganzel B, Casey BJ, Glover G, Voss HU, Temple E (2007) The aftermath of 9/11: Effect of intensity and recency of trauma on outcome. Emotion
Goenjian AK, Walling D, Steinberg AM, Karayan I, Najarian LM, Pynoos R (2005) A prospective study of posttraumatic stress and depressive reactions among treated and untreated adolescents 5 years after a catastrophic disaster. Am J Psychiatry
Goldberg DP, Gater R, Sartorius N, Ustun TB, Piccinelli M, Gureje O, Rutter C (1997) The validity of two versions of the GHQ in the WHO study of mental illness in general health care. Psychol Med
Green B, Lindy J, Grace M, Gleser G, Leonard AC, Korol M, Winget C (1990) Buffalo Creek survivors in the second decade: Stability of stress symptoms. Am J Orthopsychiatry
Heir T, Weisaeth L (2006) Back to where it happened: Self-reported symptom improvement of tsunami survivors who returned to the disaster area. Prehospl Disaster Med
. 21(2 suppl 2):59–63.
Holen A (1990) A Long-term Outcome Study of Survivors from a Disaster: The Alexander Kielland Disaster in Perspective
. Oslo (Norway): University of Oslo.
Johannesson KB, Michel PO, Hultman CM, Lindam A, Arnberg F, Lundin T (2009) Impact of exposure to trauma on posttraumatic stress disorder symptomatology in Swedish tourist tsunami survivors. J Nerv Ment Dis
Hull AM, Alexander DA, Klein S (2002) Survivors of the Piper Alpha oil platform disaster: Long-term follow-up study. Br J Psychiatry
Hussain A, Weisaeth L, Heir T (2009) Nonresponse to a population-based postdisaster postal questionnaire study. J Trauma Stress
Kessler R, Sonnega A, Bromet E, Hughes M, Nelson C (1995) Posttraumatic stress disorder in the National Comorbidity Study. Arch Gen Psychiatry
Layne C, Warren J, Watson P, Shalev A (2007) Risk, vulnerability, resistance and resilience. In M Friedman, T Keane, P Resick (Eds), Handbook of PTSD Science and Practice
(pp 497–520). New York (NY): Guilford Press.
Lazaratou H, Paparrigopoulos T, Galanos G, Psarros C, Dikeos D, Soldatos C (2008) The psychological impact of a catastrophic earthquake: A retrospective study 50 years after the event. J Nerv Ment Dis
Lundin T, Jansson L (2007) Traumatic impact of a fire disaster on survivors—a 25-year follow-up of the 1978 hotel fire in Boras, Sweden. Nord J Psychiatry
McFarlane AC, Clayer JR, Bookless CL (1997) Psychiatric morbidity following a natural disaster: An Australian bushfire. Soc Psychiatry Psychiatr Epidemiol
McFarlane AC, Van Hooff M (2009) Impact of childhood exposure to a natural disaster on adult mental health: 20-year longitudinal follow-up study. Br J Psychiatry
Neria Y, Gross R, Litz B, Maguen S, Insel B, Seirmarco G, Rosenfeld H, Suh EJ, Kishon R, Cook J, Marshall RD (2007) Prevalence and psychological correlates of complicated grief among bereaved adults 2.5–3.5 years after September 11th attacks. J Trauma Stress
Norris FH, Kaniasty K (1996) Received and perceived social support in times of stress: A test of the social support deterioration deterrence model. J Pers Soc Psychol
Norris FH, Murphy AD, Baker CK, Perilla JL (2004) Postdisaster PTSD over four waves of a panel study of Mexico's 1999 flood. J Trauma Stress
Norris F, Sloane L (2007) The Epidemiology of trauma and PTSD. In M Friedman, T Keane, P Resick (Eds), Handbook of PTSD Science and Practice
(pp 78–98). New York (NY): Guilford Press.
Norris FH, Tracy M, Galea S (2009) Looking for resilience: Understanding the longitudinal trajectories of responses to stress. Soc Sci Med
North CS, Kawasaki A, Spitznagel EL, Hong BA (2004) The course of PTSD, major depression, substance abuse and somatization after a natural disaster. J Nerv Ment Dis
North CS, McCutcheon V, Spitznagel EL, Smith EM (2002) Three-year follow-up of survivors of a mass shooting episode. J Urban Health
Ohta Y, Araki K, Kawasaki N, Nakane Y, Honda S, Mine M (2003) Psychological distress among evacuees of a volcanic eruption in Japan: A follow-up study. Psychiatry Clin Neurosci
SAS (2002) SAS 9.2. ed. Cary (NC): SAS Institute Inc.
Shalev AY (2002) Acute stress reactions in adults. Biol Psychiatry
Tang CS (2007) Trajectory of traumatic stress symptoms in the aftermath of extreme natural disaster: A study of adult Thai survivors of the 2004 Southeast Asian earthquake and tsunami. J Nerv Ment Dis
Tsai KY, Chou P, Chou FH, Su TT, Lin SC, Lu MK, Ou-Yang WC, Su CY, Chao SS, Huang MW, Wu HC, Sun WJ, Su SF, Chen MC (2007) Three-year follow-up study of the relationship between posttraumatic stress symptoms and quality of life among earthquake survivors in Yu-Chi, Taiwan. J Psychiatr Res
Tucker PM, Pfefferbaum B, North CS, Kent A, Burgin CE, Parker DE, Hossain A, Jeon-Slaughter H, Trautman RP (2007) Physiologic reactivity despite emotional resilience several years after direct exposure to terrorism. Am J Psychiatry
van den Berg B, Grievink L, van der Velden PG, Yzermans CJ, Stellato RK, Lebret E, Brunekreef B (2008) Risk factors for physical symptoms after a disaster: A longitudinal study. Psychol Med
Weiss DS (2004) The Impact of Event Scale-Revised. In J Wilson, T Keane (Eds), Assessing Psychological Trauma and PTSD
(pp 168–189). New York (NY): Guilford Press.