Download as pdf or txt
Download as pdf or txt
You are on page 1of 14

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/233564565

Psychological Sequelae of Pet Loss Following Hurricane


Katrina

Article  in  Anthrozoos A Multidisciplinary Journal of The Interactions of People & Animals · June 2008
DOI: 10.2752/175303708X305765

CITATIONS READS

97 1,240

3 authors, including:

Melissa G. Hunt
University of Pennsylvania
41 PUBLICATIONS   1,230 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Comparison of Self-Help approaches for IBS: CBT versus Low FODMAP Diet View project

All content following this page was uploaded by Melissa G. Hunt on 05 May 2014.

The user has requested enhancement of the downloaded file.


AZ VOL. 21(2).qxp:Layout 1 4/2/08 7:53 AM Page 109

ANTHROZOÖS VOLUME 21, ISSUE 2 REPRINTS AVAILABLE PHOTOCOPYING © ISAZ 2008


PP 109 – 121 DIRECTLY FROM PERMITTED PRINTED IN THE UK
THE PUBLISHERS BY LICENSE ONLY

Psychological Sequelae
of Pet Loss Following
Hurricane Katrina
Melissa Hunt, Hind Al-Awadi and Megan Johnson
Department of Psychology, University of Pennsylvania, USA
Address for correspondence: ABSTRACT One of the many impacts of natural disasters on the well-being of
Dr Melissa Hunt,
Department of Psychology,
the humans who experience them is enforced abandonment and loss of com-
3720 Walnut Street, panion animals. Hurricane Katrina, which struck the gulf coast of the United
University of Pennsylvania, States in late August, 2005, was such a disaster. This study assessed the
Philadelphia, PA
19104-6241, USA. psychological effects of pet loss on survivors of Hurricane Katrina. Sixty-five pre-
E-mail: dominantly white, female, middle-aged pet owners who lived in affected regions
mhunt@psych.upenn.edu of the country completed online questionnaires, assessing symptoms of de-
pression, acute stress, peri-traumatic dissociation, and posttraumatic stress
disorder (PTSD). Almost all companion animals were cats or dogs. Pet loss was
strongly associated with psychopathology across all measures, even when con-
trolling for displacement from the home (Wilks’ Lambda F(4,57) = 5.22, p = 0.001).
The impact of pet loss on PTSD was mediated by acute stress and dissocia-
tive symptoms during the evacuation (both F(1,61) > 9.3, both p < 0.01). This
suggests that forced abandonment of a companion animal during an evacua-
tion adds considerably to the acute trauma, thereby increasing the risk of long-
term PTSD. The impact of pet loss on depressive symptoms, however, was

Anthrozoös DOI: 10.2752/175303708X305765


independent of acute stress and dissociation (F(1,31) = 15.03, p = 0.001),
suggesting that it is both the acute loss of the pet as well as the continued ab-
sence of the pet itself that contributes to depressive symptom severity.

Keywords: depression, grief, pets, posttraumatic stress disorder, trauma


Hurricane Katrina, a category 5 storm, struck the coasts of
Louisiana, Mississippi, and Alabama in late August, 2005,
wreaking massive destruction, loss of life, and property dam-
age. In the wake of the storm, the popular press was full of images of
desperate evacuees fighting floodwaters. An astounding number of those
images showed people carrying their pets. In the months after the storm,
considerable attention was paid to the plight of abandoned animals and
to the efforts of bereaved owners to find and be reunited with their pets.
While the fate of both animals and owners clearly captured the popular
109
AZ VOL. 21(2).qxp:Layout 1 4/2/08 7:53 AM Page 110

Psychological Sequelae of Pet Loss Following Hurricane Katrina

imagination (Anderson and Anderson 2006), very little is actually known about the psycho-
logical impact of losing a pet during a natural disaster. Research in a number of different areas,
however, suggests that the psychological sequelae of such a loss may be considerable.
Companion animals are an important part of American society. Today, over 60% of house-
holds have pets, an increase from 56% of households in 1988 (Hall et al. 2004). According to
the 2003/2004 Annual National Pet Owners Survey conducted by the Pet Products Manu-
facturing Association, there are more pets than people in the United States (Hall et al. 2004).
The importance of pets in American families is also significant. Close to 100% of pet owners
identify their pets as members of the family, or even as child surrogates (Cohen 2002). Many
owners will celebrate their pets’ birthdays and approximately 50% of adults report confiding
in their pets (Beck and Meyers 1996).
Companion animals confer many social, physical, and psychological benefits. Pets provide
the benefits of love and affection as well as of security and protection (Cusack 1988). Pet own-
ership is associated with better overall physical health (Headey 2003), with benefits like reduced
blood pressure (Katcher et al. 1983), an increased survival rate among recovering cardiac pa-
tients (Friedmann et al. 1980), and a reduced risk of cardiovascular disease (Anderson, Reid and
Jennings 1992). Moreover, the presence of companion animals can mitigate the psychological
and physiological consequences of trauma and stress. Arambasic and colleagues (Arambasic
et al. 2000) found that children with pets living in war zones exhibited lower levels of posttrau-
matic stress disorder (PTSD) compared with children without companion animals.
The flip side of these benefits, however, is that losing a pet can be a significant stressor.
Gerwolls and Labott (1994) found that levels of grief following pet loss were comparable to lev-
els of grief following human loss. In a study of pet loss and mental health, Stallones (1994)
found that in a population that had recently lost a pet, the prevalence of symptoms of de-
pression was three times greater than in a normal population. Another study found that 93%
of owners dealing with the loss or illness of a pet experienced some disruption in their daily rou-
tine, typically erratic sleeping and eating patterns, while 70% reported diminished social ac-
tivities (Quackenbush and Glickman 1984). Hunt and Padilla (2006) found that pet-related
bereavement was strongly correlated with symptoms of depression.
Natural disasters are themselves significant risk factors for negative psychological con-
sequences (e.g., Briere and Elliott 2000). Natural disasters present a significant risk for both
posttraumatic stress disorder (Johnsen et al. 1997; McMillan, Smith and Fisher 1997) and de-
pression (Norris et al. 2004.) One study on Hurricane Andrew, which struck south Florida in
1992, found that six months after the disaster, 20–30% of adults in the area met criteria for
PTSD (Norris et al. 1999). Similarly, it appeared that 33–45% of adults in the area were mean-
ingfully depressed (Norris et al. 1999). Six months after Hurricane Mitch, which struck Hon-
duras and Nicaragua in 1998, one out of 10 primary health care patients from the damaged
areas where the storm hit suffered from PTSD (Caldera et al. 2001). Three to four months after
Hurricane Floyd, which struck North Carolina in 1999, the Medicaid population in affected re-
gions showed a significant increase in the number of visits to psychologists, licensed clinical
Anthrozoös

social workers, and nonspeciality providers (Fried, Domino and Shadle 2005).
The literature on natural disasters suggests that the psychological consequences of Hur-
ricane Katrina may be enormous. A few preliminary assessments confirm this assertion, in-
cluding one prepared by the Centers for Disease Control and Prevention (CDC). The CDC
study found that 45% of 166 individuals living in areas hit scored high enough on a PTSD rat-
110

ing scale to receive a referral for mental health services (Voelker 2006). The Substance Abuse
AZ VOL. 21(2).qxp:Layout 1 4/2/08 7:53 AM Page 111

Hunt et al.

and Mental Health Services Administration (SAMSHA) estimates that 25–30% of individuals will
have clinically significant mental health needs and another 10–20% will have important, sub-
clinical mental health needs (Voelker 2006). Altogether, this adds up to 500,000 people who
may require mental health services as a result of this natural disaster (Voelker 2006).
One of the most troubling issues in the field of disaster management is the problem of
evacuation failure. For some people, one of the most upsetting aspects of evacuation in the
face of natural disaster is the enforced abandonment of domestic pets (Anderson and An-
derson 2006). Most human shelters (e.g., those organized by the Red Cross) will not accept
companion animals. Thus, pet owners face the difficult choice of abandoning their pet(s) or re-
maining in a dangerous and potentially life threatening situation. Several studies have found that
animal owners will risk endangering themselves by not evacuating disaster areas unless they
are assured of their animals’ well-being (Lockwood 1997).
Heath (2000) found that, during a natural disaster, the more pets the members of a house-
hold owned, the higher their risk of household evacuation failure. He also found that more than
80% of persons who prematurely returned to an evacuated area after having left did so to res-
cue their pet (Heath 2000). In another study, people living near an army storage site for nerve
gas and blistering agents that was scheduled to be destroyed were asked to prioritize their ac-
tions for a hypothetical evacuation (Linnabary, New and Vogt 1993). Horse owners cited their
greatest concern to be family safety, followed by concern for their horses. Three quarters of
the participants stated that their decision to evacuate would be influenced by the safety of
their horses and almost half gave first priority to their animals.
Despite these findings however, pre-disaster planning by emergency management and
governmental organizations generally places a low priority on pet evacuation (Heath 2000). This
is understandable—when people are dying, it may seem irresponsible to devote search and
rescue, medical and housing resources to cats and dogs. If we are to convince law makers
that their policies and budgets must include animal rescue, it is important to measure the
human costs of pet abandonment, and to find ways to measure the potential benefits of ef-
fective pet evacuation policies in reducing both human suffering and high-risk behaviors such
as refusal to evacuate.
The literature on companion animals, hurricanes as traumatic events, and the role of pets
in evacuation failures, suggests that there might be a relationship between the mental health
of hurricane survivors and the status of their companion animals during and after the disaster.
The current study was designed to examine this relationship by recruiting pet owners who
survived Hurricane Katrina and assessing their levels of psychopathology. Given that pet own-
ership is associated with various benefits and that pet loss is linked with increased depressive
symptoms, we hypothesized that the individuals who were able to keep their pets during Hur-
ricane Katrina would exhibit lower levels of psychopathology than those who were forced to
abandon their pets as a result of the disaster.

Methods
Anthrozoös

Procedure
Participants were comprised of pet owners who lived in the areas of Louisiana, Mississippi, and
Alabama, where Hurricane Katrina hit. They fell into one of two categories: those who kept their
pet(s) and those who lost their pet(s) as a result of the disaster. An invitational message was
posted on a number of Internet message boards, including www.petfinder.org, www.nola.com,
111

www.craigslist.org, and www.wwltv.com. This message informed the viewer that the study
AZ VOL. 21(2).qxp:Layout 1 4/2/08 7:53 AM Page 112

Psychological Sequelae of Pet Loss Following Hurricane Katrina

was about the consequences of pet loss following a natural disaster. The link within the mes-
sage led them to the study website, where they were consented, and invited to complete a
battery of questionnaires.

Materials
Demographic Descriptors and Hurricane & Pet-Related Questions: Questions designed to as-
certain demographic characteristics of the respondents included age, city and state of resi-
dence prior to Hurricane Katrina, gender, race, marital status, number of children, education
level, and income level. Additionally, questions concerning human and material loss as a re-
sult of Hurricane Katrina, as well as questions about pet ownership and pet status during and
after the disaster, were posed.

Beck Depression Inventory (BDI-II; Beck et al. 1961; Beck, Steer and Brown 1996): This
21-item, forced-choice questionnaire measures depressive symptoms. Items are scored on a
0–3 scale and summed to produce a total score, with higher scores indicating more depres-
sive symptoms. This measure has moderate to good internal consistency and good test-retest
reliability (Beck et al. 1996). For the BDI-II, Beck at al. (1996) suggest that symptom scores of
0–13 be interpreted as minimal, 14–19 as mild, 20–28 as moderate, and 29–63 as severe.
Thus, 14 is a relatively sensitive cut point for symptomatic levels of depressive symptoms. In
this study, question 9, concerning suicidal ideation, was removed.

PTSD Symptom Scale Self-Report (PSS-SR; Foa et al. 1993): This questionnaire consists of
the 17 symptoms from the DSM-IV (APA 2000) description of PTSD. Respondents are asked
to complete PSSS-R items by rating the frequency of each symptom over the last week on a
scale from 0 (not at all) to 4 (almost always). The highest possible score is 68.

Peri-traumatic Dissociative Experiences Questionnaire–Self-Report (PDEQ-SR; Marmar, Weiss


and Metzler 1997): This self-report, 10-item questionnaire asks respondents to what extent
they experienced depersonalization, derealization, confusion, concentration and memory
deficits during and immediately following the time of the traumatic event. Dissociative symp-
toms include feeling as if events are not real, but more like a movie or dream; feeling as if one
were a spectator observing oneself; experiencing changes in time sense, especially as if things
were happening in slow motion; and experiencing confusion, periods of “blankness” or lack
of awareness, and a feeling of disconnection from one’s body, surroundings, or emotional ex-
perience. Peri-traumatic symptoms are measured on a 5-point scale from 1 (not true at all) to
5 (extremely true). The highest possible score is 50.

Stanford Acute Stress Reaction Questionnaire (SARSQ; Cardena, Classen and Spiegel 1991):
This questionnaire consists of 10 items from the DSM-IV (APA 2000) description of Acute Stress
Disorder, asking respondents if they have experienced numbing, detachment, de-realization, de-
Anthrozoös

personalization, or dissociative amnesia. Respondents are asked to what extent they experienced
these symptoms during and immediately after a disaster. They rate their reaction on a 7-point scale
from 0 (not experienced) to 6 (very often experienced). The highest possible score is 60.

Pet Attachment Questionnaire (PAQ; Stallones et al. 1990): This questionnaire is an 8-item
112

questionnaire devised to assess pet attachment. It has good internal consistency (alpha of
AZ VOL. 21(2).qxp:Layout 1 4/2/08 7:53 AM Page 113

Hunt et al.

0.75) and good factor structure. Stallones et al. (1990) found the questionnaire to be appro-
priate for all companion animals, not only cats and dogs. Pet attachment has been found to
be correlated with degree of grief after pet loss (e.g., Hunt and Padilla 2006).

Pet Bereavement Questionnaire (PBQ; Hunt and Padilla 2006): This questionnaire was devel-
oped and validated by the principal investigator. It is a 16-item questionnaire with good inter-
nal consistency (alpha of 0.87) and good factor structure and construct validity. For the
purposes of this study, it had to be modified slightly. First, two items relating to anger at a vet-
erinarian and at friends or family for not being more helpful were deleted. Second, the word
“disappearance” was added to several items as an alternative to death. In the current study,
the PBQ consisted of 14 items and continued to show good internal consistency, with an
alpha of 0.80. The questionnaire was only given to those participants who lost their animals
due to Hurricane Katrina.

Results
Description of Sample
Sixty-five pet owners (62 women and 3 men) who survived Hurricane Katrina completed the
online questionnaires. The study’s participants were predominantly Caucasian (n = 62), with
only one African American and one Asian American/Pacific Islander participating. The mean
age of the sample was 37 years. Most participants (n = 45) were middle-aged (30–54 years);
17 were young adults (18–29 years); and two were seniors (55+). The majority of participants
were college-educated (n = 47). Most participants (n = 36) made between $15,000–
$45,000/year; 16 made $45,000/year or more; and 12 made $15,000/year or less. Twenty-
one participants were single; 14 were in a relationship; 21 were married; eight were divorced;
and one was widowed. Most participants were either cat (n = 27) or dog owners (n = 23); 14
had both pets; and one was a parrot owner. Forty-two percent of the sample had been forced
to abandon their pet during the natural disaster. See Table 1 for means and standard deviations
of all outcome measures by pet status. Additionally, 45% of participants were displaced from
their homes as a consequence of the hurricane.

Table 1. Mean scores on all measures by pet status.


Pet Status BDI PSS-SR SASRQ PDEQ-SR PAQ PBQ
Pet Abandoned 23 (13) 22 (12) 27 (11) 30 (9.5) 17 (3.5) 49 (5)
Pet Never Lost 13 (8) 15 (9) 18 (10) 23 (9) 18 (3) N/A
Note: BDI = Beck Depression Inventory; PSS-SR = PTSD Symptom Scale Self-Report; SASRQ = Stan-
ford Acute Stress Reaction Questionnaire; PDEQ-SR = Peri-traumatic Dissociative Experiences Ques-
tionnaire – Self-Report; PAQ = Pet Attachment Questionnaire; PBQ = Pet Bereavement Questionnaire.
Standard deviations are in parentheses.

Impact of Demographics on Psychopathology, Pet Attachment, and Pet Bereavement


Anthrozoös

There were no main effects of age, income, or type of pet on any of the outcome measures. A
series of one-way ANOVAs revealed significant main effects of marital status on depressive symp-
toms (F(3,60) = 2.75, p = 0.05), acute stress (F(3,60) = 4.4, p < 0.01), peritraumatic dissociation
(F(3,60) = 3.28, p < 0.05), PTSD symptoms (F(3,60) = 2.81, p < 0.05) and pet bereavement (F(3,60) =
4.81, p = 0.01). Planned pairwise comparisons revealed that divorced participants, compared
113

with married, coupled, and single individuals, showed higher levels of depressive symptoms (for
AZ VOL. 21(2).qxp:Layout 1 4/2/08 7:53 AM Page 114

Psychological Sequelae of Pet Loss Following Hurricane Katrina

all comparisons of divorced participants to others, all p < 0.05), acute stress (all p < 0.005), and
PTSD (all p < 0.05). Divorced participants also reported significantly higher levels of pet be-
reavement than single or coupled participants (both p < 0.05), but were not significantly differ-
ent from married individuals. Marital status was not significantly related to pet attachment.
There was a main effect of education on some of the outcome variables. College graduates had
significantly lower levels of depression (t(60) = -2.04, p < 0.05), pet attachment (t(60) = -2.69,
p < 0.01), and pet bereavement (t(60) = -2.18, p < 0.05) than participants with no higher education.
Impact of Pet Loss on Psychopathology
Since our data set included multiple outcome variables that were highly inter-correlated (see
Table 2), we began with a conservative MANOVA predicting all four psychopathology indices
from pet loss, controlling for whether the individuals had been displaced from their homes and
for the interaction between pet loss and displacement. There was a highly significant main
effect of pet loss on psychological outcomes (Wilks’ Lambda F(4,57) = 5.22, p = 0.001). There
was no main effect of displacement (Wilks’ Lambda F(4,57) = 1.01, ns). There was, however, a
non-significant trend suggesting that pet loss might interact with displacement in predicting
psychopathology (Wilks’ Lambda F(4,57) = 1.92, p = 0.12).

Table 2. Correlations among outcome measures.

PSS-SR SASRQ PDEQ-SR PAQ PBQ


BDI 0.72 ** 0.53 ** 0.49 ** 0.17 ns 0.55 **
PSS-SR 0.66 ** 0.69 ** 0.19 ns 0.28 ns
SASRQ 0.67 ** 0.17 ns 0.31 ns
PDEQ-SR 0.27 * 0.23 ns
PAQ 0.26 ns
Note: BDI = Beck Depression Inventory; PSS-SR = PTSD Symptom Scale Self-Report; SASRQ
= Stanford Acute Stress Reaction Questionnaire; PDEQ-SR = Peri-traumatic Dissociative Experi-
ences Questionnaire – Self-Report; PAQ = Pet Attachment Questionnaire; PBQ = Pet Bereavement
Questionnaire.
* Correlation is significant at the 0.05 level (2-tailed). **Correllation is significant at the 0.01 level (2-tailed).

We then conducted separate ANCOVAs for each outcome variable, using pet loss, dis-
placement, and their interaction as predictors. Pet loss was a significant predictor of depres-
sive symptoms (F(1,60) = 16.06, p < 0.001), acute stress (F(1,60)= 8.33, p < 0.01), peri-traumatic
dissociation (F(1,60) = 8.56, p < 0.01) and PTSD symptom severity (F(1,60) = 5.44, p < 0.05).
There was a significant main effect of displacement on acute stress (F(1,60) = 3.93, p = 0.05) but
not on any other outcome variable. There was also a significant interaction between dis-
placement and pet loss on depressive symptoms (F(1,60) = 7.41, p < 0.01) (see Figures 1–4).
We calculated the effect sizes using Cohen’s d (Cohen 1988) for the four measures of psy-
Anthrozoös

chopathology. When comparing participants who were not displaced but differed in terms of
pet status (never lost or lost), the effect sizes were 1.62 for the BDI, 1.00 for the SASRQ, 0.72
for the PDEQ, and 1.02 for the PSS-SR, all large effects according to Cohen’s standards.
When comparing individuals who had been displaced but differed in pet status, the effect sizes
were 0.31 for the BDI, 0.49 for the SASRQ, 0.81 for the PDEQ, and 0.2 for the PSS-SR. These
114

effect sizes were medium, medium, large, and small, respectively, according to Cohen’s
AZ VOL. 21(2).qxp:Layout 1 4/2/08 7:53 AM Page 115

Hunt et al.

Mean Beck Depression Inventory Score


35
30

25

20 Pet Kept
15 Pet Lost

10

0
Not Displaced Displaced

Figure 1. Depressive symptom severity by Pet Status and Displacement.

30
Mean PTSD Symptom Scale Score

25

20
Pet Kept
15
Pet Lost
10

0
Not Displaced Displaced

Figure 2. PTSD Symptom Severity by Pet Status and Displacement.


Mean Peri-Traumatic Dissociation Score

40
35
30
25
Pet Kept
20
Pet Lost
15
10
5
0
Not Displaced Displaced
Anthrozoös

Figure 3. Peri-Traumatic Dissociation by Pet Status and Displacement.

standards. These effect sizes suggest that pet loss had an impact on psychopathology over
and above being displaced from one’s home. Comparing individuals who lost their pets (but
kept their homes) and those who lost their homes (but kept their pets) yielded effect sizes of
115

0.8 for depression and 0.58 for peri-traumatic dissociation. These are large enough to suggest
AZ VOL. 21(2).qxp:Layout 1 4/2/08 7:53 AM Page 116

Psychological Sequelae of Pet Loss Following Hurricane Katrina

35

Mean Stanford Acute Stress


30

25

Inventory Score
20
Pet Kept
15 Pet Lost
10

0
Not Displaced Displaced

Figure 4. Acute Stress by Pet Status and Displacement.

that for some individuals, losing a pet is a more important predictor of psychological outcome
than losing one’s home.
Mediation of Pet Loss on PTSD by Acute Stress and Peri-Traumatic Dissociation
Both acute stress and peri-traumatic dissociation are well known risk factors for PTSD, so we
tested whether the impact of pet loss on PTSD might be mediated by increased rates of acute
stress and peri-traumatic dissociation during the evacuation. Mediation requires that pet loss
predict PTSD, that pet loss predict acute stress and dissociation, that acute stress and peri-
traumatic dissociation predict PTSD, and, finally, that there is no variance in PTSD that can be
accounted for by pet loss once acute stress and peri-traumatic dissociation are controlled for.
We had already established that pet loss was a significant predictor of PTSD, acute stress
and peri-traumatic dissociation. Not surprisingly, both acute stress and peri-traumatic disso-
ciation contributed uniquely to predicting PTSD symptom severity (both F(1,62) > 9.5, both p <
0.01). Finally, when pet loss, acute stress and peri-traumatic dissociation were all included in
the ANCOVA, acute stress and dissociation remained significant predictors (both F(1,61) > 9.3,
both p < 0.01) but pet loss dropped out completely (F(1,61) = 0.02, ns). Thus, all of the condi-
tions for mediation were met (see Figure 5).

Acute Stress & Peri-Traumatic Dissociation

Both p < 0.05 Both p < 0.001

p = 0.88, ns
Pet Loss PTSD Severity

p = 0.05
Anthrozoös

Figure 5. Acute Stress and Peri-Traumatic Dissociation Mediate the Impact of Pet
Loss on PTSD.
116
AZ VOL. 21(2).qxp:Layout 1 4/2/08 7:53 AM Page 117

Hunt et al.

Mediation of Pet Loss on Depression by Acute Stress and Peri-Traumatic


Dissociation
We also tested whether acute stress and peri-traumatic dissociation might mediate the
relationship between pet loss and depressive symptoms. Since we had previously found that
displacement status interacted with pet loss in predicting depressive symptoms, we con-
ducted the mediation analyses separately by displacement status. For the non-displaced
individuals, the preconditions of mediation were all met. As previously established, pet loss
was a significant predictor of depressive symptoms (F(1,33) = 27.57, p < 0.001), acute stress,
and dissociation (see above). We also established that both acute stress and dissociation pre-
dicted depressive symptoms (both F(1,33) > 11.5, both p < 0.01). In the final ANCOVA control-
ling for pet loss, acute stress and dissociation, we found that both dissociation and pet loss
remained significant predictors of depressive symptoms (F(1,31) = 4.34, p < 0.05 and F(1,31) =
15.03, p = 0.001, respectively) while acute stress dropped out (F(1,31) = 0.45, ns). For displaced
individuals, the preconditions of mediation were not met because pet loss was not a signifi-
cant predictor of depressive symptoms, so no further analyses were undertaken.
Pet Bereavement Questionnaire
PBQ total scores were significantly correlated with depressive symptoms (r = 0.55, p < 0.01).
The Anger/Trauma factor of the PBQ was significantly related to depressive symptoms (r =
0.52, p < 0.05) and to symptoms of PTSD (r = 0.42, p = 0.05). This is not surprising, since it
includes items that reflect re-experiencing symptoms such as “memories of my pet’s last
moments haunt me” and “I have had nightmares about my pet’s death/disappearance.”
We also conducted four independent t-tests on the PBQ total and factor scores by
displacement. (Note that only individuals who lost their pets completed the PBQ, which
presupposes that a pet has died.) There was a significant effect of displacement on PBQ total
scores, such that those participants who lost their homes reported higher overall scores on the
PBQ (t(19) = 2.86, p = 0.01). However, this entire effect appears to be due to a significant
difference on the trauma factor score which was highly significant (t(19) = 3.69, p < 0.01), while
the grief and guilt factor scores showed no differences as a function of displacement (both
t(19) ≤ 1.0, both ns).
Pet Attachment
Pet attachment was significantly correlated with peri-traumatic dissociation (r = 0.27, p < 0.05),
but not with depressive symptoms, acute stress, PTSD, or pet bereavement.

Discussion
Consistent with our predictions, we found that the status of their companion animals was
strongly related to the mental health of Hurricane Katrina survivors. Pet loss was found to be
associated with significantly higher levels of psychopathology, including acute stress and peri-
traumatic dissociation, as well as symptoms of depression and PTSD, even when controlling
for displacement from the home.
Anthrozoös

Reactions to traumas can be both short term (acute stress, peri-traumatic dissociation) and
long term (depression, PTSD). Being forced to abandon one’s pet during the flooding was
clearly a significant stressor for evacuees. Pet loss was strongly associated with both acute
stress and peri-traumatic dissociation, both of which are known risk factors for long-term
PTSD (Harvey and Bryant 1998; Candel and Merckelback 2004). While there was a main effect
117

of pet loss on PTSD symptoms, we found that it was mediated by acute stress and
AZ VOL. 21(2).qxp:Layout 1 4/2/08 7:53 AM Page 118

Psychological Sequelae of Pet Loss Following Hurricane Katrina

peri-traumatic dissociation. This suggests that the impact of pet loss on PTSD can be ac-
counted for by the acute trauma of being forced to abandon one’s pet during an evacuation.
Displacement from the home in and of itself also had a significant effect on acute stress,
but not, interestingly, on PTSD or depression. This is somewhat surprising, given that the de-
gree of home damage a hurricane survivor sustains as well as their continued displacement
from their home have previously been found to be strongly related to subsequent levels of
PTSD (Lonigan et al. 1994).
Pet loss and displacement interacted in predicting depressive symptoms. There was a sig-
nificant main effect of pet loss such that individuals who lost their pets reported, on average,
higher levels of depressive symptoms (BDI M = 23, SD = 13) than those who did not
(BDI M = 12, SD = 7). However, this effect was most pronounced for people who were not
displaced from their home. Indeed, individuals who were forced to abandon their pet(s) but
kept their homes had the highest BDI scores of any group in the sample (M = 26, SD = 13.6),
suggesting moderate to severe levels of depressive symptoms that are highly likely to reflect
clinically significant depression (Beck et al. 1996). In comparison, individuals who were able
to keep both their pets and their homes had the lowest levels of depressive symptoms
(BDI M = 9, SD = 5), which is well below the standard cut point for even mild depressive
symptoms (Beck et al. 1996). Moreover, the relationship between pet loss and depressive
symptoms was not mediated by either acute stress or peri-traumatic dissociation. Thus, the
impact of pet loss on depressive symptoms cannot be accounted for by the acute trauma of
being forced to abandon one’s pet. Rather, it appears to be the actual loss that contributes
to depressive symptom severity.
For individuals who were displaced from their homes, the impact of pet loss on depres-
sive symptoms was much less striking. In our sample, people who lost both their home and
their pet(s) had a mean BDI score of 20 (SD = 12), while people who lost their homes but kept
their pets had a mean BDI score of 17 (SD = 8), both of which are in the clinically significant,
but mild range of symptoms. That is, it is possible that losing one’s home is so depressing
and traumatizing, that losing one’s pet(s) on top of that does not seem to be associated with
much worsening of depressive symptoms. Another possibility is that “survivor guilt” was
stronger for those individuals who kept their homes but lost their pets, perhaps because the
inevitability of the loss of the pets is reduced. However, the data from the PBQ appear to
support the “trauma saturation” hypothesis over the “survivor guilt” hypothesis. Individuals
who were displaced from their homes reported significantly higher scores on the trauma fac-
tor of the PBQ, but not on the guilt factor. Perhaps the most important aspect of these data
overall are that the loss of a pet alone led to levels of depression that were not significantly
different from the loss of a home.
Some of our demographic variables were related to outcome. Married individuals exhib-
ited the lowest levels of psychopathology and divorced individuals exhibited the highest. This
might be due to the fact that marriage is associated with significant social support. Married and
divorced individuals showed the highest scores on pet bereavement. Additionally, education
Anthrozoös

level was related to lower scores on all the outcome measures, suggesting that education
may be a buffer against developing psychopathology in the wake of a disaster.
One of the more interesting findings concerning demographics was that income level was
not significantly related to any of the measures of psychopathology. This suggests that finan-
cial resources did not necessarily buffer people from the trauma of the hurricane. This is com-
118

pelling in light of the media’s focus on how the poorer areas of Louisiana, Mississippi, and
AZ VOL. 21(2).qxp:Layout 1 4/2/08 7:53 AM Page 119

Hunt et al.

Alabama were affected by the disaster. Similarly, pet type was not related to any of the out-
come measures. This finding supports previous research that has shown that levels of pet at-
tachment do not differ between dog and cat owners (Stallones et al. 1990).
There were some significant limitations to this study. In particular, the sample consisted
mostly of Caucasian, middle-aged women. This clearly limits the generalizability of the results.
It is possible that this limited sample resulted from the fact that this was an Internet-based
study and those who were hit hardest by the natural disaster might not have been able to ob-
tain Internet access. However, prior research has also strongly suggested that the prototypic
bereaved pet owner is a middle-aged white woman, so our sample is at least consistent with
previous work in the area (e.g., McCutcheon and Fleming 2001–2002; Brown 2002.) More-
over, we recruited participants through Katrina-related websites, including pet-specific sites that
were attempting to reunite lost animals with their owners. While we also posted study links on
a number of other Katrina-related sites, in addition to general sites such as Craig’s List, it is
probable that our sample was highly self-selected. That is, only those individuals who were par-
ticularly invested in companion animals were likely to have seen the study link and spent the
time to participate.
A related limitation is that both the advertisements for the study and consent form stated
that the study was focusing on the consequences of pet loss, assuring that the participants
were not blind to the objective of the study. Thus, the recruitment and informed consent
process may have resulted in demand effects that led to a systematic bias toward over-
reporting negative experiences and symptoms regarding pet loss. In an ideal world, we
would have framed the study more neutrally, perhaps as an investigation of how pet own-
ers were “impacted” by the hurricane. However, even had we done so, it is highly likely that
the nature of the questionnaires used would have made clear to people the sorts of impact
we were investigating.
Second, we relied entirely on self-report symptom inventories, which cannot be used to
diagnose clinical disorders. While instruments like the BDI have been used in hundreds of
studies, and have well-established and well-validated cut points, it is still not appropriate to use
them diagnostically. In an ideal world, researchers would be equipped to travel to disaster
areas and conduct clinical interviews with the relevant participants. However, the consider-
able advantages of Internet research, which include anonymity, convenience, and access,
preclude options like structured clinical interviews.
Finally, the study started running three months after Hurricane Katrina hit, meaning that
we had to rely on retrospective reports of acute stress and peri-traumatic dissociation. Because
people who have experienced a traumatic event like a hurricane find it particularly difficult to
give accurate descriptions of past emotional states (Candel and Merckelback 2004), it is pos-
sible that our results were compromised by retrospective bias.
Despite the limitations, we believe that our results shed some light on the importance of
companion animals in the lives of their owners during a natural disaster. The fact that we were
able to demonstrate statistically and clinically significant effects of pet loss on psychopathol-
Anthrozoös

ogy suggests that there are many potential benefits to including pets in evacuation policies.
While rescuing pets may be logistically and economically challenging, the long-term benefits
to their human owners in terms of psychological well-being are probably substantial. As
clinicians and human beings, we must consider the heartbreak of an individual owner being
forced to abandon a beloved companion. As a society, we must consider the cost to us all of
119

the PTSD and depressive symptoms such enforced abandonment is associated with.
AZ VOL. 21(2).qxp:Layout 1 4/2/08 7:53 AM Page 120

Psychological Sequelae of Pet Loss Following Hurricane Katrina

One positive consequence to emerge in the aftermath of Hurricane Katrina was the
passage of the Pets Evacuation and Transportation Standards Act in September of 2006. This
act was an amendment to the Stafford Disaster Relief and Emergency Assistance Act. It
provides the Federal Emergency Management Agency (FEMA) Director with the authority to
make financial contributions to the States to help with state and local preparedness to address
the needs of individuals with household pets and service animals following a major disaster or
emergency. We have yet to see how this legislation will actually impact disaster preparedness
in the future, but it is to be hoped that enforced abandonment, and the negative psychological
sequelae that come with it, may be reduced in the future.

References
American Psychiatric Association. 2000. Diagnostic and Statistical Manual of Mental Disorders. Text Revision.
DSM-IV-TR. 4th edn. Washington, DC: Author.
Anderson, A. and Anderson, L. 2006. Rescued: Saving Animals from Disaster. Novato, CA: New World Library.
Anderson W. P., Reid, C. M. and Jennings G. L. 1992. Is pet ownership good for your heart? The results of a
survey of risk factors for cardiovascular disease in Melbourne, Australia. In Urban Animal Management:
Proceedings of the First National Conference on Urban Animal Management in Australia, 150–153, ed. R. W.
Murray. Mackay, Australia: Chiron Media.
Arambasic, L., Kerestes̆, G., Kuterovac-Jagodić, G. and Vizek-Vidović, V. 2000. The role of pet ownership as
a possible buffer variable in traumatic experiences. Studia Psychologica 42: 135–146.
Beck, A. M. and Meyers N. M. 1996. Health enhancement and companion animal ownership. Annual Review
of Public Health 17: 247–257.
Beck, A. T., Steer, R. A. and Brown, G. K. 1996. Beck Depression Inventory – II Manual. The Psychological
Corporation. San Antonio, TX: Harcourt Brace and Company.
Beck, A.T., Ward, C.H., Mendelson, M., Mock, J. and Erbaugh, J. 1961. An inventory for measuring depres-
sion. Archives of General Psychiatry 4: 561–571.
Briere, J. and Elliot, D. 2000. Prevalence, characteristics, and long-term sequelae of natural disaster exposure
in the general population. Journal of Traumatic Stress 13: 661–679.
Brown, S. 2002. Ethnic variations in pet attachment among students at an American school of veterinary
medicine. Society & Animals 10(3): 249–266.
Caldera, T., Palma L., Penayo, U. and Kullgren, G. 2001. Psychological impact of the hurricane Mitch in
Nicaragua in a one-year perspective. Social Psychiatry and Psychiatric Epidemiology 36: 108–114.
Candel, I. and Merckelback, H. 2004. Peri-traumatic Dissociation as a predictor of Post-traumatic Stress Dis-
order: A critical review. Comprehensive Psychiatry 45: 44–50.
Cardena, E., Classen, C. and Spiegel, D. 1991. Stanford Acute Stress Reaction Questionnaire. Stanford, CA:
Stanford University Medical School.
Cohen, J. 1988. Statistical Power Analysis for the Behavioral Sciences. 2nd edn. Hillsdale, NJ: Lawrence
Earlbaum Associates.
Cohen, S. P. 2002. Can pets function as family members? Western Journal of Nursing Research. Special Issue:
Human–Animal Interaction 24: 621–638.
Cusack, O. 1988. Pets and Mental Health. New York: Haworth Press.
Foa, E. B., Riggs, D. S., Dancu, C. V. and Rothbaum B. O. 1993. Reliability and validity of a brief instrument for
assessing Post Traumatic Stress Disorder. Journal of Traumatic Stress 6: 459–473.
Fried, B. J., Domino, M. E. and Shadle, J. 2005. Use of mental health services after Hurricane Floyd in North
Carolina. Psychiatric Services 11: 1367–1373.
Anthrozoös

Friedmann, E., Katcher A. H., Lynch J. J. and Thomas, S. A. 1980. Animal companions and one year survival
of patients after discharge from a coronary care unit. Public Health Reports 95: 307–312.
Gerwolls, M. K. and Labott, S. M. 1994. Adjustment to the death of a companion animal. Anthrozoös 7: 172–176.
Hall, M. J., Ng, A., Ursano, R. J., Holloway, H., Fullerton, C. and Casper, J. 2004. Psychological impact of the
animal–human bond in disaster preparedness and response. Journal of Psychiatric Practice 10: 368–374.
Harvey, A. G. and Bryant, R. A. 1998. The relationship between Acute Stress Disorder and Posttraumatic Stress
120

Disorder: A prospective evaluation of motor vehicle accident survivors. Journal of Consulting and Clinical
Psychology 66: 507–512.
AZ VOL. 21(2).qxp:Layout 1 4/2/08 7:53 AM Page 121

Hunt et al.

Headey, B. 2003. Pet ownership: good for health? Medical Journal of Australia 179: 460–461.
Heath, S. E. 2000. An epidemiological study of public and animal health consequences of pet ownership in a
disaster: The January 1997 flood of Yuba County, California. Dissertation Abstracts International: Section B:
The Sciences and Engineering 60: 5384.
Hunt, M. G. and Padilla, Y. 2006. Development and validation of the Pet Bereavement Questionnaire. Anthrozoös
19(4): 308–324.
Johnsen, B. H., Eid, J., Lovstad, T. and Michelsen, L. T. 1997. Posttraumatic stress symptoms in non-exposed,
victims, and spontaneous rescuers after an avalanche. Journal of Traumatic Stress 10: 133–140.
Katcher, A. H., Beck, A., Friedmann, E. and Lynch, J. 1983. Looking, talking and blood pressure: The physio-
logical consequences of interaction with the living environment. In New Perspectives on Our Lives with
Companion Animals, 351–359, ed. A. Katcher and A. Beck. Philadelphia: University of Pennsylvania Press.
Linnabary, R. D., New, J. C. and Vogt, B. M. 1993. Emergency evacuation of horses: A Madison County,
Kentucky survey. Journal of Equine Veterinary Science 13: 153–158.
Lockwood, R. 1997. Through Hell and High Water: Disasters and the Human–Animal Bond. Washington, DC:
The Humane Society of the United States.
Lonigan, C .J., Shannon M. P., Taylor C. M., Finch A. J. and Sallee, F. R. 1994. Children exposed to disaster: II.
Risk factors for the development of post-traumatic symptomatology. Journal of the American Academy of
Child and Adolescent Psychiatry 33: 94–105.
Marmar, C. R., Weiss, D. S. and Metzler, T. J. 1997. The Peri-traumatic Dissociative Experiences Questionnaire.
In Assessing Psychological Trauma and PTSD: A Handbook for Practitioners, 412–428, ed. J. P. K. Wilson
and T. M. Keene. New York: Guilford.
McCutcheon, K. and Fleming, S. 2001–2002. Grief resulting from euthanasia and natural death of companion
animals. OMEGA: The Journal of Death and Dying 44(2): 169–188.
McMillen, J. C., Smith, E. M. and Fisher, R. H. 1997. Perceived benefit and mental health after three types of
disaster. Journal of Consulting and Clinical Psychology 65: 733–739.
Norris, F. H., Murphy, A. D., Baker, C. K. and Perilla, J. L. 2004. Postdisaster PTSD over four waves of a panel
study of Mexico’s 1999 flood. Journal of Traumatic Stress 17: 283–292.
Norris, F. H., Perilla, J. L., Riad, J. K., Kaniasty, K. and Lavizzo, E. A. 1999. Stability and change in stress,
resources, and psychological distress following natural disaster: findings from Hurricane Andrew. Anxiety,
Stress, and Coping 12: 363–396.
Quackenbush, J. E. and Glickman, L. 1984. Helping people adjust to the death of a pet. Health Social Work 9:
42–48.
Stallones, L. 1994. Pet loss and mental health. Anthrozoös 7: 43–54.
Stallones, L., Garrity, T. F., Johnson, T. P. and Marx, M. B. 1990. Quality of attachment to companion animals
among U.S. adults 21–64 years of age. Anthrozoös 3: 171–175.
Voelker, R. 2006. Post-Katrina mental health needs prompt group to compile disaster medical guide. Journal
of the American Medical Association 295: 259–260.

Anthrozoös
121

View publication stats

You might also like