Double Disaster: Mental Health Of Survivors of Wildfires and Earthquake in a Part of Greece

doi:10.4236/psych.2011.22021

Paper Menu >>

Journal Menu >>

Psychology

2011. Vol.2, No.2, 132-137

Double Disaster: Mental Health of Survivors of Wildfires

and Earthquake in a Part of Greece

Vicky Papanikolaou1, Dimitrios Adamis2,3, Robert C. Mellon4,

Gerasimos Prodromitis4, John Kyriopoulos1

1Department of Health Service Management, National School of Public Health, Athens, Greece;

2Research and Academic Institute of Athens, Athens, Greece;

3Institute of Psychiatry, Kings College, HSPR Department, London, UK;

4Department of Psychology, Panteion University of Social and Political Sciences, Athens, Greece.

Email: [email protected], [email protected], {mellon, gprod}@panteion.gr

Received December 16th, 2010; revised February 8th, 2011; accepted February 22nd, 2011

This paper investigates stress related psychological morbidity in individuals who experienced two disasters 11

months apart (wildfire and earthquake) in a rural area of Greece. A sample of 150 participants has been assessed

after the wildfires and after the earthquake using the Symptom Checklist 90-Revised. Survivors had elevated

levels of psychopathology in all subscales of the SCL-90–R after the earthquake. Significant risk factors for fur-

ther development of psychopathology were damages to property and complete loss of property from both disas-

ters. Double disasters can cause considerable psychological symptoms in victims and there are r easons for policy

makers to create services in order to help and improve the mental health of those affected but also to help them

rebuild their property.

Keywords: Double Disaster, Greece, Psychopathology, Psychological Distress, Adults

Introduction

Natural disasters are frequent events in Greece. It is esti-

mated that from 1980 to 2008, 61 natural disasters happened

and the average number of people killed per year were 54 while

the average number of affected people per year was 10,426.

Earthquakes and wildfires are among the most frequent natural

disasters and they have also caused the most profound eco-

nomic damages (EM-DAT, 2010).

In August of 2007 an intense and destructive wildfire broke

out in the Peloponnesus peninsula in Greece. This was the

worst of the century in Greece. The fires expanded rapidly and

raged out of control for several days. Sixty to eighty people

were reported killed and 5,392 people were affected from the

disaster (EM-DAT, 2010). About 1,500 square kilometres of

forests, olive trees, farmland, and villages were burned in these

fires and the economic damages were estimated around

1,750,000 (X 1,000) US$. A national disaster was declared and

the areas affected by the fires were designated for further sup-

port.

A year later, in July 2008, a deadly earthquake hit the same

area (Peloponnesus) with a moment magnitude of 6.5, accord-

ing to the Athens Geodynamic Institute. The number of people

killed varied in the reports from 2-6, the injured were more than

220 and at least 2,000 people were reported homeless. The

number of affected people was estimated at around 3,708

(EM-DAT, 2010).

After the August wildfires it was perhaps the first time that

the need for an effective public health planning for disasters

was recognised, in order to deliver specific services and to pro-

vide sufficient resources for treating mental disorders, reducing

symptoms, and preventing future problems to those affected by

the disaster. In addition, it was recognised that it was necessary

to estimate the impact of the natural disasters on the mental

health of victims. Parts of this planning were also to carry out

research and to evaluate the affected population mental health

problems.

Although natural disasters differ widely, they usually have

some common characteristics in terms of the risks of survivors’

developing psychopathology and mental distress. Life threat,

injury to oneself or family member, death of loved ones, and

property loss were among the risk factors which have been

indentified (Norris et al., 2002; Norris, Friedman, Watson,

2002). All of these factors exist typically in any natural disaster.

The consequences of these disorders can be long lasting. In

addition, a different kind of disaster may have a different im-

pact on mental health (Norris, Friedman, Watson, 2002) and it

has been suggested (Weiss, Saraceno, Saxena, & van Ommeren,

2003) that it is important to distinguish continuing situations (e.g.,

ongoing war, ongoing drought) fr om acute ones, because chronic

disasters result in simultaneous acute and ongoing disas-

ter-related problems. However, th e impact of double acute disas-

ters in short time on the mental health of survivors is less studied.

Here, the present study investigates the psychological dis-

tress six months after the wildfires and one month after the

earthquake in a selected community population subsample in

the rural region of Peloponnesus, Greece.

In particular, the study aims to (a) assess the differences in

the prevalence of psychopathology 6 months after the wildfires

and one month after the earthquake, and (b) examine socio-

demographic, and disaster related losses variables predictive of

psychiatric casenes s .

V. PAPANIKOLAOU ET AL. 133

Method

Design of the Stu dy

The first phase of the study (6 months after the wildfires)

was a cross sectional case control study. The two samples

(cases and controls) were closely matched for gender, age,

educational, marital and regional distributions. Participants

were residents aged from 18 years to 65 years old who lived in

the five prefectures designated by the Hellenic Republic Minis-

try of Interior to be disaster areas. The number of respondents

surveyed in each prefecture was proportional to its adult popu-

lation. The sample was composed of 409 residents of desig-

nated disaster areas and 391 residents of directly adjoining

areas in which there was no fire damage in the immediate vi-

cinity. For more details about the methods for this phase see

(Mellon, Papanikolau, & Prodromitis, 2009).

Participants

For the second phase of study which is examined here, the

recruited population were a subsample (n = 150) of those af-

fected from wildfires and who lived mainly in the affected from

the earthquake area (see also Figure 1). They were reassessed in

one month’s time after the earthquake (five months after first

assessment and 11 months after the wildfire disaster). The par-

ticipants were residents aged from 18 years to 65 years old.

Measurements

1) Demographic characteristics (age, gender, educational

background, marital status, occupation). All the demographic

characteristics were self-reported. Age was categorised in five

groups, education in three, marital status in four and occupation

in three.

2) Symptom Checklist 90-Revised (SCL-90-R) (Derogatis &

Psychometric, 1992). A Greek validated version of SCL-90-R

was used (Donias, Karastergiou, & Manos, 1991). A sensitivity

of 0.98 and a specificity of 0.74 in indicating active psychiatric

patients of SCL-90-R were reported (Donias, Karastergiou, &

Manos, 1991). The SCL-90-R has 90 items, which measure the

degree of distress experienced by the individual during the last

7 days, using a 5-point scale (0 to 4) that ranges from “not at

all” to “extremely.” The SCL-90R can be scored for nine symp-

tom dimensions (Somatization, Obsessive-Compulsive, Inter-

pesonal Sensitivity, Depression, Anxiety, Hostility, Phobic

Anxiety, Paranoid, and Psychoticism). In addition to the nine

scales, there are three global indices that are computed. The

Global Severity Index (GSI), is the sum of all the nonzero re-

sponses, divided by 90, (if there are no missing responses) and

reflects both the number of symptoms endorsed and the inten-

sity of perceived distress. The Positive Symptom Total (PST) is

defined as the number of symptoms to which the patient indi-

cates a nonzero response. This is a measure of the number of

symptoms endorsed. Thus, it can be interpreted as a measure-

ment of symptoms span. The Positive Symptom Distress Index

(PSDI) is calculated by dividing the sum of all item values by

the PST; this is a measure of “intensity” corrected for the num-

ber of symptoms. The internal consistency reliability of

SCL-90-R was found quite satisfactory ranging from a 0.77 for

the dimension of Psychoticism to 0.90 for Depression. Simi-

larly test retest reliability was found ranging from 0.78 for the

dimension of Hostility to 0.90 for Phobic Anxiety. Regarding

validity, the sensitivity of SCL-90-R to detect psychiatric cases

was reported equal to 0.90 and specificity to 0.87 (Derogatis &

Psychometric, 1992).

3) Number and type of losses as a result of the fire or the

earthquake including: a) damage to property (Yes vs. No), b)

Missi ng n = 38.

Missing answers on

SCL-90-R or “fake

good or bad”

Missing n = 12.

Missing answers on

SCL-90-R or “fake

good or bad”

“wave”

Included n = 112 2

“wave”

Included n = 138

Finally analysed n = 112

“wave”

Eligible and Approached

N = 150

“wave”

Eligible and Approached

N = 150

Figure 1.

Chart flow of participants.

V. PAPANIKOLAOU ET AL.

134

complete damage and loss of property (Yes vs. No), c) personal

injury or injury of a close family member (Yes vs. No), and d)

deaths of close family members (Yes vs. No). The responses to

questions a and b were mutually exclusive. If more than one

loss had happened all of them counted (number of losses). Also

the types of losses were self-reported and the number of losses

was calculated from the different types.

Procedure

Data were collected in face-to-face interviews. In the first

wave of assessments interviewers asked if there was an adult in

the random sampled household (at least 18 years of age) who

would be willing to participate anonymously. In each house-

hold only one interview was conducted. In the second wave of

assessments the same individual from the same household was

asked again to participate in the survey.

Ethics

The study has been approved by the Ministry of Health and

informed consent was obtained from each participant.

Analysis

The Q Local v 2.1.11, (NCS Pearson Inc, MN, USA), was

used for the estimation of the standardized T scores from the

raw data for the SCL-90-R scale. Data were analysed with

PASW (SPSS) v18, using appropriate paired test statistics. For

the non-normally distributed data, non-parametric tests were

used. To estimate the effects of losses and individual character-

istics on the caseness status, as it was recorded with the

SCL-90-R scale, a multinomial logistic regression model was

constructed.

Results

Demographic Characteristics of the Sample

A total number of 150 subjects took part in both “waves” of

the survey, of whom 75 (50%) were males. The demographic

characteristics are presented in Table 1.

Psychological Distress after the Wildfires and the

Earthquake

To investigate the differences of psychological symptoms as

they were measured with the SCL-90-R scale between the two

assessments (after wildfires and after the earthquake) a paired

test was used. As the variables did not conform to a normal

distribution, a non-parametric test was used (Wilcoxon Signed

Ranks Test). Because of missing answers in SCL-90-R, or same

answer in all 90 questions or fake “good or bad”, the total

number of participants analysed were N = 112 (Figure 1). Table

2 shows some descriptive statistics ( N, mean, SD) of each vari-

able and the statistical significance of the comparisons. As it

can be seen from Table 2 all the participants had significantly

increased psychological distress after the earthquake in all the

measured symptoms and the three indices of SCL-90-R com-

pared to the previous 5 months ago.

Number and Ty p e o f Los ses

Table 3 shows the numbers and the percentages of the indi-

viduals who had each type of loss. The number of losses was

estimated by adding each category of loss: the minimum is 0

and the maximum 3.

Caseness

A caseness, according to SCL-90-R, is when a respondent

has a GSI score greater or equal to a T score of 63 or if any of

two dimensions scores are greater than or equal to a T score of

63. Thus, in the first wave of assessments (n = 112) 23 cases

were identified (20.5%) while in the second (n = 138) 93 cases

(67.4%). In the total analysed sample (N = 103) 23 participants

(22.3%) were not cases in both assessments, 58 (56.3%) were

new cases, 15 (14.6%) remained the same and 7 (6.8%) have

recovered. There was a significant increase of cases after the

earthquake (McNemar Test, x2 = 38.462, df: 1, p < .001) com-

pared to cases after the wildfires.

Effects of Socio-Demographic Factors and Losses

from the Disasters on the Caseness

To further investigate the factors that significantly influence

the caseness status (new cases, constantly cases, recovered

cases, and never cases) a multinomial regression model was

conducted. The caseness status (four levels) was the dependent

variable and demographic characteristics (gender, age, educa-

tion, occupation, and marital status), type of losses and the

number of losses were the independent variables. Table 4

shows only the significant effects of the independent variables

on the caseness status.

As it can be seen from Table 4 new cases were more likely to

be those who had old and new damages to their property and

those who have completely lost their property from both natural

disasters compared to those who did not develop significant

psychopathology (no cases in both assessments). However, they

were less likely to have deaths of loved ones. In addition, those

who were constantly cases in both waves of the survey were

Table 1.

Demographic characteristics of the sample.

Count

(N = 150)N%

18 - 25 27 18.0

26 - 35 31 20.7

36 - 45 32 21.3

46 - 55 32 21.3

Age group

56 - 65 28 18.7

Primary school 39 26.0

Secondary s chool 93 62.0

Education

College/university 18 12.0

married 103 68.7

single 41 27.3

divorced 1 .7

Marital status

widowed 5 3.3

professional occupation 22 14.7

sales and customer se rv ice occupation 19 12.7

Occupation

elementary occupation 109 72.7

V. PAPANIKOLAOU ET AL. 135

Table 2.

Comparison between the assessments.

N Mean Std. Deviati on (1-2) Z Sig. (2-tailed)

Somatisati on 2 138 65.17 9.891

Somatisati on 1 112 50.62 11.296 –7.231a p < .001

Obsessive -compulsive (OC) 2 138 60.23 9.337

Obsessive -compulsive (OC) 1 112 48.59 10.010 –6.794a p < .001

Interpersonal sensitivity (IS) 2 138 60.25 9.849

Interpersonal sensitivity (IS) 1 112 49.21 8.680 –6.554a p < .001

Depression 2 138 63.06 9.099

Depression 1 112 52.04 9.862 –6.750a p < .001

Anxiety 2 138 57.11 11.333

Anxiety 1 112 47.55 10.558 –5.935a p < .001

Hostility 2 138 55.51 10.005

Hostility 1 112 49.33 9.666 –4.587a p < .001

Phobic anxie t y 2 138 58.90 10.295

Phobic anxie t y 1 112 51.86 7.933 –5.171a p < .001

Paranoid 2 138 62.22 10.792

Paranoid 1 112 51.46 10.459 –6.269a p < .001

Psychotism 2 138 56.30 10.786

Psychotism 1 112 47.47 7.428 –5.743a p < .001

GSI 2 138 63.09 9.226

GSI 1 112 48.95 11.053 –7.454a p < .001

PSDI 2 138 58.72 7.137

PSDI 1 112 50.53 9.634 –6.377a p < .001

PST 2 138 61.91 8.995

PST 1 112 48.59 11.464 –7.098a p < .001

a. Based on positive ranks.

Table 3.

Number and type of losses in t h e t w o w a v e s o f as s e s s m e n t s.

First wave n

(%) Second wave n

(%)

No 58 (38.7) 30 (20)

damages to pr operty Yes 92 (61.3) 120 (80)

No 117 (78) 12 5 (83.3)

complete loss of property Yes 33 (22) 25 (16.7)

No 144 (96) 141 (94)

injuries of individual or

close members of the family Yes 6 (4) 9 (6)

No 141 (94) 138 (92)

deaths of close members Yes 9 (6) 12 (8)

0 21 (14) 8 (5.3)

1 118 (78.7) 12 0 (80)

2 11 (7.3) 20 (13.3)

the numbe r of losses

3 - 2 (1.3)

more likely to be males (borderline significance) and those who

had completely lost their property in both disasters.

Discussion

The results show that the one disaster after another increases

the psychopathology of the survivors. Only few victims had

recovered (in terms of psychopathology) from the first disaster

during the five months time and none of the examined factors

contributed significantly to their recovery. Previous research in

survivors of disasters has showed that the psychological dis-

tress and psychopathology after disasters are long lasting (Nor-

ris et al., 2002; Norris, Friedman, Watson, 2002; Weiss et al.

2003). Those who were constantly cases during the five

months’ time were more likely to be males and to have com-

pletely lost their property by the two disasters. It seems that

during the time between the disasters, they had tried to rebuild

their property but the new disaster destroyed everything again.

This perhaps explains why the male gender was a risk factor.

Culturally in Greece and especally in rural areas males are i

V. PAPANIKOLAOU ET AL.

136

Table 4.

Parameter Estimates (only significant results) fro m m ul t i no m i al an a l ys i s .

Casenessa B Std. ErrorWald df Sig.

Intercept 17.9632.803 41.070 1 **

Not at all damages –0.8291.955 0.180 1 NS

New damages –0.2840.787 0.131 1 NS

Old damages –3.3321.489 5.009 1 *

Old and new damages 0b . . 0

Complete loss of property (no) –20.1221.971 104.224 1 **

Complete loss of property (New) –16.8362.230 57.025 1 **

Complete loss of p roperty (Old) –18.9191.811 109.176 1 **

Complete loss of property (old and new) 0b . . 0

Deaths (no) 2.806 1.218 5.308 1 .*

Deaths (New) 0.268 1.583 0.029 1 NS

Deaths (Old) 0b . . 0

Old caseness

(recovered n = 7) No significant effects of examined variables

Intercept 19.7642.759 51.315 1 **

Male 1.789 0.911 3.858 1 *

Female 0b . . 0 .

Complete loss of property (no) –20.1011.217 272.957 1 **

Complete loss of property (New) –18.3571.692 117.674 1 **

old and new

(constantly caseness = 15)

Complete loss of property (old and new) 0b . . 0

a. The reference category is: no caseness in both assessments; b. This parameter is set to zero because it is redundant; Statistical significant p < .05

marked *, p < .001 marked **, NS = no significance

responsible for the building and the maintenance of property

thus conceivably they are under more stress. Although the fe-

male gender has been identified in literature as a risk factor to

develop psychopathology after disasters (Norris et al., 2002)

not all the studies agree with it (e.g. Den Ouden, 2007) while

others reported that females are more likely to develop depres-

sion or phobia (Heir & Weisaeth, 2008; Hussain, Weisaeth &

Heir, 2010). In addition, when we analysed the data separately,

ignoring the first wave of assessments we did not find any sig-

nificant effect of gender on the caseness (analysis was not re-

ported in the results). Similarly, analysis of the total sample of

the wildfires victims (first wave of assessments), gender was

found not to influence the caseness but females were more

likely to develop somatisation symptoms (Unpublished data).

However, a previous study after a series of disasters (storm,

tornado, floods, exposure to dioxin) in St. Louis area (USA) in

1982 (Solomon, Smith, Robins, & Fischbach, 1987) reported

that males and females differ in how they display negative ef-

fects of disaster exposure. Males showed increased symptoms

of alcohol abuse and depression as a result of either personal, or

both personal and indirect, exposure to disaster, while the psy-

chopathology of females was not affected by the personal ex-

posure to disaster. Thus males are more prone to develop

symptomatology in directed exposure to disaster compared to

females (Solomon, Smith, Robins, & Fischbach, 1987).

Furthermore, those who became new cases in terms of psy-

chopathology after the earthquake were those who had property

losses in both disasters. It seems that the new losses had an

additive effect on the development of psychopathology. Al-

though material losses seem to be a risk factor for caseness, it is

surprising that injuries to oneself or to a family member had not

any effect. In addition, deaths of loved ones seem to be a pro-

tective factor for psychological caseness. However, it is impor-

tant to note here that the number of deaths was small, so the

power to detect true differences is low. Also, the official num-

ber of deaths reported for the second disaster (earthquake) was

two, and those two (if not more) have accounted for twelve

(affected) survivors. This could happen as the disaster area was

rural and the family ties are perhaps tighter. On the contrary,

the number of injured was high but injuries did not have any

effect on psychopathology. However, similar findings have

been reported by others (Heir & Weisaeth, 2008; Clayer, Book-

less-Pratz, & Harris, 1985). In fact, Heir and Weisaeth (Heir &

Weisaeth, 2008) pointed out that having a near relative or close

friend injured could be a protective factor because of distraction,

and because a caretaking role for a close relative may increase

resilience and self efficacy.

This study also has limitations. First of all, the number of

participants in the second wave of assessments and inevitably

the comparison with the first wave of assessments is relatively

small. Because, as we discussed above, the number of deaths

was small, maybe the effect of deaths on psychological distress

is wrongly estimated. Since the number of participants was

fixed from the first wave of assessments and given that the

study was purely observational and pragmatic we could not

increase the number of participants at the second wave of as-

V. PAPANIKOLAOU ET AL. 137

sessments. Similarly, we have found some risk factors from

variables which we have measured and maybe other unob-

served factors have also influenced the psychopathology in both

disasters. A second more important limitation of this study is

that the sample may be biased. In the second wave of assess-

ments the sample was not random as in the first. Simply, we

followed-up those who participated in the first part of the study

and we have included those who lived in the most affected from

the earthquake area (prefecture of Ilia). Thus, it is very likely

that we have compared a less affected from wildfires popula-

tion with the most affected from the earthquake and thus the

differences in psychopathology perhaps were more pronounced.

In addition, the time of assessments after the disaster was un-

equal. Our data can only be interpreted in one way; that is, a

disaster after another increases psychological distress. On the

other hand, despite these limitations this study is maybe the

first which examined the impact of two “acute” natural disas-

ters on the mental health of victims.

Thus, it seems from the above that two natural disasters in

relatively short time increase the psychological distress of the

affected population. Risk factors for further development of

psychopathology are damages to property and complete loss of

property. It can also be concluded that it takes time for the

psychological distress to cease after a disaster.

These findings have implications for policy makers, disaster

response agencies, and communities affected by disaster. If the

losses are risk factors for producing psychological distress then

efforts to protect positions or to replace them to some degree

may reduce distress and thus reduce the impact of disaster on a

community.

Our findings are similar with general disaster literature de-

spite cultural differences. Loss of property and income together

with life threat were significantly related to distress in previous

studies e.g. (Gibbs, 1989; O’Neill et al., 1999; Freedy et al.,

1994; Smith, & Freedy, 2000 ) .

Thus there are reasons for policy makers not only to create

services in order to help and improve the mental health of those

affected but also to support victims in restoring their properties.

Acknowledgements

The study has been funded by the Ministry of Health & So-

cial Solidarity within the framework of Health Disaster Man-

agement in cooperation with World Association for Disaster

Emergency Medicine.

References

Clayer, J. R., Bookless-Pratz, C., & Harris, R. L. (1985). Some health

consequences of a natural disaster. Medical Journal of Australia, 143,

182-184.

Den Ouden, D. J., Van Der Velden, P. G., Grievink, L., Morren, M.,

Dirkzwager, A. J. E., & Yzermans, C. J. (2007). Use of mental health

services among disaster survivors: Predisposing factors. BMC Public

Health, 7, 173. doi:10.1186/1471-2458-7-173

Derogatis, L. R. & Psychometric, R. C. (1992). SCL-90-R: Administra-

tion, scoring & procedures manual-II, for the revised version and

other instruments of the psychopathology rating scale series. Clinical

Psychometric Research, Towson.

Donias, S., Karastergiou, A., & Manos, N. (1991). Standardization of

the symptom checklist 90 rating scale in a Greek population. Psy-

chiatriki, 2, 42-48.

EM-DAT. (2008). The OFDA/CRED International Disaster Database

PreventionWeb, Louvain Centre for Research on the Epidemiology

of Disasters, Ecole de Santé Publique, Université Catholique de Lou-

vain.

http://www.preventionweb.net/english/countries/statistics/index.php?

cid=68.” Retrieved 15/9/2010. .

Freedy, J. R., Saladin, M. E., Kilpatrick, D. G., Resnick, H. S., &

Saunders, B. E. (1994) Understanding acute psychological distress

following natural disaster. Journal of Trauma Str ess , 7, 257-273.

doi:10.1002/jts.2490070207

Gibbs, M. S. (1989). Factors in the victim that mediate between disaster

and psychopathology: A review. Journal of Traumatic Stress, 2, 489-

514. doi:10.1002/jts.2490020411

Heir, T. & Weisaeth, L. (2008). Acute disaster exposure and mental

health complaints of Norwegian tsunami survivors six months post

disaster. Psychiatry, 71, 266-276. doi:10.1521/psyc.2008.71.3.266

Hussain, A., Weisaeth, L., & Heir, T. (2010, In Press) Psychiatric dis-

orders and functional impairment among disaster victims after expo-

sure to a natural disaster: A population based study. Journal of Affec-

tive Disorders.

Mellon, R. C., Papanikolau, V., & Prodromitis, G. (2009). Locus of

control and psychopathology in relation to levels of trauma and loss:

Self-reports of Peloponnesian wildfire survivors. Journal of Trauma

Stress, 22, 189-196. doi:10.1002/jts.20411

Norris, F. H., Frie dman, M. J., Wats on, P. J., Byrn e, C. M., Diaz, E., &

Kaniasty, K. (2002). 60,000 disaster victims speak: Part I. An em-

pirical review of the empirical literature, 1981-2001. Psychiatry, 65,

207-239. doi:10.1521/psyc.65.3.207.20173

Norris, F. H., Friedman, M. J., & Watson, P. J. (2002). 60,000 disaster

victims speak: Part II. Summary and implications of the disaster

mental health research. Psychiatry, 65, 240-260.

doi:10.1521/psyc.65.3.240.20169

O'Neill, H. K., Evans, B. A., Bussman, M. D., & Strandberg, D. K.

(1999). Psychological distress during the Red River flood: Predictive

utility of the Conservation of Resources model. Applied Behavioral

Science Review, 7, 159-169. doi:10.1016/S1068-8595(00)80016-1

Smith, B. W. and Freedy, J. R. (2000) Psychosocial resource loss as a

mediator of the effects of flood exposure on psychological distress

and physical sym ptoms. Journal of Trauma Stress, 13, 349.

doi:10.1023/A:1007745920466

Solomon, S., Smith, E. M., Robins, L., & Fischbach, R. L. (1987) So-

cial involvement as a mediator of disaster-induced stress. Journal of

Applied Social Psychology, 17, 1092-1112.

doi:10.1111/j.1559-1816.1987.tb02349.x

Weiss, M. G., Saraceno, B., Saxena, S., & van Ommeren, M. (2003).

Mental health in the aftermath of disasters: Consensus and contro-

versy. Journal of Nervous and Mental Disease, 191, 611- 615.

doi:10.1097/01.nmd.0000087188.96516.a3