Content uploaded by Marie Jose Enders-Slegers
Author content
All content in this area was uploaded by Marie Jose Enders-Slegers on Jan 30, 2016
Content may be subject to copyright.
Effect of animal-assisted interventions on depression,
agitation and quality of life in nursing home residents
suffering from cognitive impairmen t or dementia: a cluster
randomized controlled trial
Christine Olsen
1
, Ingeborg Pedersen
1
, Astrid Bergland
2
, Marie-José Enders-Slegers
3
, Grete Patil
1
and
Camilla Ihlebæk
1,4
1
Section for Public Health Science, Department of Landscape Architecture and Spatial Planning, Norwegian University of Life Sciences, Ås,
Norway
2
Faculty of Health Sciences, Oslo and Akershus University College, Oslo, Norway
3
Psychology and Educational Sciences, Open University of the Netherlands, Heerlen, The Netherlands
4
Faculty of Health and Social Work Studies, Østfold University College, Fredrikstad, Norway
Correspondence to: C. Olsen, E-mail: christine.olsen@nmbu.no
Objectives:
The prevalence of neuropsychiatric symptoms in cognitively impaired nursing home resi-
dents is known to be very high, with depression and agitation being the most common symptoms.
The possible effects of a 12-week intervention with animal-assisted activities (AAA) in nursing homes
were studied. The primary outcomes related to depression, agitation and quality of life (QoL).
Method: A prospective, cluster randomized multicentre trial with a follow-up measurement 3 months
after end of intervention was used. Inclusion criteria were men and women aged 65 years or older, with
a diagnosis of dementia or having a cognitive deficit. Ten nursing homes were randomized to either
AAA with a dog or a control group with treatment as usual. In total, 58 participants were recruited:
28 in the intervention group and 30 in the control group. The intervention consisted of a 30-min
session with AAA twice weekly for 12 weeks in groups of three to six participants, led by a qualified
dog handler. Norwegian versions of the Cornell Scale for Depression, the Brief Agitation Rating Scale
and the Quality of Life in Late-stage Dementia scale were used.
Results: A significant effect on depression and QoL was found for participants with severe dementia at
follow-up. For QoL, a significant effect of AAA was also found immediately after the intervention. No
effects on agitation were found.
Conclusions: Animal-assisted activities may have a positive effect on symptoms of depression and QoL
in older people with dementia, especially those in a late stage.
Key words: dementia; neuropsychiatric symptoms; depression; agitation; quality of life; non-pharmacological interventions;
animal-assisted interventions
History: Received 06 July 2015; Accepted 23 December 2015; Published online in Wiley Online Library (wileyonlinelibrary.
com)
DOI: 10.1002/gps.4436
Introduction
Dementia is among the leading causes of disability and
death in the elderly (Lobo et al., 2000). Approximately
80% of nursing home residents in Norway suffer from
dementia (Selbæk et al., 2007b), and dementia is the
most common main diagnosis in the nursing home
population in Norway (Nygaard, 2002). In older adults
with a neurodegenerative form of dementia, ongoing de-
generation of brain tissue eventually leads to a loss of
cognitive and physical functions (McKhann et al., 1984;
van Iersel et al., 2004). In addition to impaired cognition,
neuropsychiatric symptoms (NPS) such as apathy,
depressive symptoms, anxiety, agitation, restlessness
Copyright # 2016 John Wiley & Sons, Ltd. Int J Geriatr Psychiatry 2016
RESEARCH ARTICLE
and wandering are common symptoms (Selbæk, 2005;
Selbæk et al., 2007a).
The prevalence of NPS in patients with dementia
has been reported as very high. For example, following
a 2-year longitudinal study, Aalten et al. (2005) found
that 95% of the patients developed one or more NPS.
Lyketsos et al. (2002) found that 75% of the patients
with dementia in their study population had experi-
enced NPS in the preceding month and 55% reported
having two or more symptoms. A recent Norwegian
study found a 31% prevalence of depression among
recently admitted long-term care patients (Iden et al.,
2014). NPS affect patients’ quality of life (QoL)
(Beerens et al., 2013; Mjørud et al., 2014b), and low
QoL is associated with impaired mobility, lack of
social activities and low performance in activities relat-
ing to daily living (Nagatomo et al., 1997; Barca et al.,
2011; Telenius et al., 2013; Mjørud et al., 2014a).
As population ages, health care and social services
face increased demands to provide services for older
people with dementia or cognitive impairment.
Because there is no cure for dementia (Geldmacher
et al., 2006), there is a need for new and innovative
approaches to complement traditional health care.
Medication for NPS is commonly used, but most of
the medicines have major physical and mental side
effects such as abnormal liver function, heart defects,
gastrointestinal problems, apathy, ataxia, restlessness
and insomnia (Tripathi and Vibha, 2010). The finding
of Iden et al. (2014) that antidepressants had been
prescribed for 44% of their study participants indi-
cates extensive use. Little is known about the efficacy
and safety of antidepressant medication when used
to treat symptoms of agitation and psychosis (Seitz
et al., 2011). Therefore, it has been suggested that
non-pharmacological interventions should be imple-
mented on a larger scale in nursing homes (Douglas
et al., 2004; Iden et al., 2014).
Several non-pharmacological alternatives and
complementary treatments have evolved, including
animal-assisted interventions (AAI). The International
Association of Human–Animal Interaction Organiza-
tions (IAHAIO, 2014) defines AAI as ‘a goal oriented
and structured intervention that intentionally includes
or incorporates animals in health, education and hu-
man service for the purpose of therapeutic gains in
humans’. Animal-assisted activities (AAA) are a form
of AAI whereby companion animals are taken by their
human handlers to visit nursing homes for ‘meet and
greet’ activities with residents.
Previous studies have shown mixed results regarding
the effectiveness of AAI on depression, agitation and
QoL for dementia patients (Richeson, 2003; Mossello
et al., 2011; Majic et al., 2013; Nordgren and Engstrom,
2014a, 2014b; Friedmann et al., 2015; Thodberg et al.,
2015). Further, much of the research on AAI and
dementia to date has lacked adequate study designs for
investigating the effects of interventions, and because
of the limited use of control groups and follow-up mea-
sures, the conclusions are disputable. For this reason,
the aim of this study was to examine the possible effects
on depression, agitation and QoL in nursing home res-
idents with dementia or cognitive impairment, through
an intervention with AAA and a follow-up study.
Methods
Design
The study was conducted in Norway as a prospective
and cluster randomized multicentre 12-week trial with
a 3-month follow-up. Computer-generated random
numbers were used to randomize nursing home units
to either an AAA group with a dog or to a control
group with treatment as usual. The study was regis-
tered by ClinicalTrials.gov (identifier: NCT02008630).
Data collection was carried out at baseline before
the intervention started (T
0
), when finishing the inter-
vention after 12 weeks (T
1
), and at follow-up 3 months
after the intervention had ended (T
2
).
Participants and recruitment
Of 90 eligible nursing homes in three Norwegian
counties, 10 adapted units for residents with dementia
agreed to participate in the project (Figure 1). The
nursing homes included in the study had to provide
the facilities required to carry out the interventions.
They also had to abstain from any dog-visiting activi-
ties for 3 months prior to the intervention, as well as
during the whole intervention period from T
0
to T
2
.
The health personnel in the nursing homes were
asked to recruit between five and eight participants
each. The inclusion criteria were as follows: aged
65 years or older and having dementia or a cognitive
deficit score of less than 25 on the mini-mental state
examination test (Folstein et al., 1975; Strobel and
Engedal, 2009). The exclusion criteria were nursing
home residents with fear of dogs or with a dog allergy.
Of 130 eligible patients in the 10 units, 58 patients
(45%) agreed to participate; seven patients (12%) died
during the study period and were subsequently ex-
cluded from the study. Thus, the study population
consisted of 51 participants. Three participants
dropped out of the study after baseline data were
2 C. Olsen et al.
Copyright # 2016 John Wiley & Sons, Ltd. Int J Geriatr Psychiatry 2016
collected but were included in the study population
(Figure 1).
The study was conducted during winter –spring
2013 (n = 12), autumn–winter 2013 (n = 22) and
spring–summer 2014 (n = 24).
Intervention and intervention content
A protocol was developed by the project group to stan-
dardize the AAA intervention across different units
and dog handlers. The intervention consisted of a
30-min session with AAA twice weekly for 12weeks
in groups of three to six participants. The AAA
sessions were led by a qualified dog handler.
For each session, the participants were randomly
seated in a half-circle. Each session started with a
greeting round, when each participant had the opportu-
nity to pet the dog and feed it treats. Thereafter, the han-
dler started the different activities, which included any
of the following: petting the dog, feeding the dog a treat
and throwing a toy for the dog to fetch. All activities
were supposed to follow the protocol but should be in-
dividually tailored to each participant based on the
health personnel’s knowledge of the participant. How-
ever, no activities were mandatory, and the sessions
therefore included activities that occurred between the
participants and between each participant and the dog.
The control groups were not offered any new
activities, and their treatment continued as usual, in-
cluding diverse group activities such as reminiscence,
music therapy, sensory garden, singing, exercise,
cooking and handicrafts.
Figure 1 Consort flow diagram of participants.
3Animal-assisted interventions for dementia patients
Copyright # 2016 John Wiley & Sons, Ltd. Int J Geriatr Psychiatry 2016
Dogs and their handlers
Both dogs and their handlers were carefully selected
for their suitability to work with AAIs. The dogs had
to take and pass a mentality test containing different
elements with respect to, for example, aggressiveness,
sociability, anxiety and handling. Similarly, their han-
dlers completed at least one course in AAIs for visiting
dogs. To enhance the similarity between the 10 units,
all handlers were informed about the protocol for
the sessions both verbally and in writing.
All handlers, except one, had either a theoretical or
practical background in health care or biological
science.
Assessments and procedures for data collection
The instruments used in the study have all been tested
for their validity and reliability and have been designed
and/or are commonly used for older people with
dementia. Prior to the start of the project, two health
professionals from each nursing home unit attended
lectures with instructions on how to use the instru-
ments. They later scored all assessments at all three
time points (T
0
,T
1
and T
2
).
Depression was measured using the Cornell Scale
for Depression in Dementia (CSDD) (Alexopoulos
et al., 1988; Barca et al., 2010); a validated Norwegian
version was used (Korner et al., 2006). The scale
contains 19 symptoms of depression in five domains
(mood-related signs, behavioural disturbance, physical
signs, cyclic functions and ideational disturbance). Each
item is rated on a scale from absent, mild/intermittent
to severe, with a sum score ranging from 0 to 38
(Cronbach’s alpha=0.74). A sum score below 6 indi-
cates the absence of depressive symptoms, scores above
10 indicate probable major depression and scores above
18 indicate definite major depression (Alexopoulus
et al., 1988).
Agitation and restlessness were measured using the
Brief Agitation Rating Scale (BARS) (Finkel et al.,
1993), derived from the 29-item Cohen-Mansfield
Agitation Inventory (Cohen-Mansfield et al., 1989).
The BARS is used to assess the presence and severity
of physically aggressive, physically non-aggressive and
verbally agitated behaviours in older nursing home
residents. It is a seven-level scale of frequency from 1
(never)to7(a few times per hour or continuously for
half an hour or more). The validated Norwegian
version of the instrument (Swift et al., 2002; Sommer
and Engedal, 2011) is a nine-item inventory with a
sum score ranging from 9 to 63 (Cronbach’s
alpha = 0.76), where a high score indicates higher fre-
quency of agitated behaviour.
Quality of life was measured using the validated
Norwegian version of Quality of Life in Late-stage
Dementia (QUALID) (Weiner et al., 2000; Røen
et al., 2015). The scale consists of 11 items with a pos-
sible score of 1–5 on each item. The items are rated by
frequency of occurrence, comprising both positive and
negative dimensions of concrete and observable mood
and performance. Scores are summed to range from
11 to 55 (Cronbach’s alpha = 0.79). A low score indi-
cates a high QoL.
The Clinical Dementia Rating Scale (CDR) is a 5-
point scale used to assess six domains of cognitive and
functional performance-applicable dementia (Hughes
et al., 1982; Engedal and Haugen, 1993; Nygaard and
Ruths, 2003). CDR staging is a valid substitute for a de-
mentia assessment among nursing home residents to de-
termine the severity of dementia (Engedal and Haugen,
1993; Nygaard and Ruths, 2003). A CDR of 0 implies
no cognitive impairment,0.5very mild dementia,1mild
dementia,2moderate dementia and 3 severe dementia.
The study participants’ sociodemographic charac-
teristics on age, gender, education, use of walking aids,
social contact, hobbies and animal contact were col-
lected at baseline (Table 1).
Ethics
The project was performed in accordance with the
Helsinki Declaration and the Regional Committee
for Medical and Health Research Ethics approved the
project. Nursing staff at each participating nursing
home allocated eligible participants, provided infor-
mation about the study and obtained written consent.
Written and verbal information about the study was
given to the patients and their relatives by the primary
caregiver. A procedure was developed for health per-
sonnel to evaluate the participants’ cognitive capacity
to give informed written consent. Those with suffi-
cient cognitive capacity were informed about the pro-
ject and gave written consent to participate. For those
with reduced capacity, health personnel and/or the
next of kin took this decision on their behalf and gave
written consent. All participants were informed that
they could withdraw from the study at any stage.
Statistical analyses
Prior to commencing the study, a power calculation
was made using statistical software
JMP version 12
(SAS Institute, Cary, SC, USA) with BARS as the
4 C. Olsen et al.
Copyright # 2016 John Wiley & Sons, Ltd. Int J Geriatr Psychiatry 2016
primary outcome measure. A power calculation for
change of means in BARS with 80% probability of
detecting differences between groups, alpha 0.05, and
a least significant difference of 7.0 points (SD = 8.4)
between the intervention group and the control group
indicated a necessary total of 30 participants in each
group at the respective units. The power calculation
took into account a 20% dropout rate.
Intraclass correlation coefficient
To test the level of agreement between the different
raters, health personnel from five units with the same
training in BARS scored the same participants (n = 28),
intraclass correlation (ICC) = 0.84 (single measures).
Values between 0.75 and 1.0 are considered to indicate
excellent interrater reliability (Hallgren, 2012). ICC
was also used to test for cluster effect of facilities
(ICC BARS = 0.02; ICC CSDD = -0.04; ICC
QUALID = 0.28).
Missing data
The person mean substitution method was used to im-
pute missing data on item level for CSDD, BARS and
QUALID if three or fewer items were missing.
Analyses
All analyses were computed using statistical software
IBM SPSS Statistics for Windows, Version 22.0.
Armonk, NY: IBM Corp. To assess the internal consis-
tency of CSDD, BARS and QUALID, Cronbach’s al-
pha was calculated for the sum scores, all of which
showed acceptable consistency. One-way ANOVA for
continuous data and chi-square for categorical data
were used to test the differences in means between
the intervention and control groups at T
0
.
A mixed model was used to investigate changes over
time and differences between the intervention group
and the control group (West, 2009). The dependent var-
iables were the three main types of assessment: CSDD,
BARS and QUALID. Time was modelled as a repeated
variable, and an autoregressive covariance structure
(AR1) was used to accommodate dependencies between
the three points in time. The type of intervention was
included as fixed effect; nursing home within group
was included as random effect. T
0
wasusedasreference
point for time. The control group was set as the refer-
ence group. To accommodate different time trends be-
tween the groups, an interaction term was included
between the intervention group and control group and
points of time—the effect of interest in the study.
As severity of dementia is known to affect main as-
sessments (Beerens et al., 2013; Mjørud et al., 2014a),
also stratified analyses of cognitive and functional per-
formance (CDR) were conducted. Before the analyses,
CDR was dichotomized into either mild/moderate or
severe dementia.
To test the clinically significant change in depres-
sion, a modified method developed by Teri et al.
(1997) was used. The participants’ sum scores for T
0
,
T
1
and T
2
were categorized into four levels according
the administration and scoring guidelines for the
CSDD by George S. Alexopoulos (2002). Subjects with
Table 1 Demographic data for control and animal-assisted activity (AAA)
Control
(n = 26)
AAA
(n = 25)
p-
value
Gender, women (%) 17 (65.4) 15 (60.0) 0.69
Missing 0 0
Age, mean (SD) 84.1 (6.7) 82.9
(8.5)
0.60
Missing 1 1
Enjoy animal contact (%) 24 (92.3) 18 (72.0) 0.78
Missing 0 5 (20.0)
Clinical Dementia Rating
Scale (%)
0.72
000
0.5 1 (3.9) 0
1 1 (3.9) 2 (8.0)
2 12 (46.2) 11 (44.0)
3 12 (46.2) 12 (48.0)
Missing 0 0
Education (%) 0.20
Primary school 17 (65.4) 9 (36.0)
Secondary school 4 (15.4) 3 (12.0)
Higher education 3 (11.5) 2 (8.0)
Other 2 (7.7) 3 (12.0)
Missing 0 8 (32.0)
Walking aids (%) 0.16
None 8 (30.8) 10 (40.0)
Walking sticks 0 0
Cane 3 (11.5) 1 (4.0)
Crutches 0 0
Rollator 8 (30.8) 12 (48.0)
High walker 4 (15.4) 0
Wheelchair 3 (11.5) 1 (4.0)
Supported walking 0 1 (4.0)
Missing 0 0
Social contact (%) 0.10
Daily 0 2 (8.0)
Several times per week 9 (34.6) 7 (28.0)
Once per week 10 (38.5) 14 (56.0)
Every other week 4 (15.4) 0
Rare 3 (11.5) 1 (4.0)
Missing 0 1 (4.0)
Hobbies (%) 0.30
Cognitive activities 7 (26.9) 3 (12.0)
Physical activities 11 (42.3) 8 (32.0)
Other 1 (3.85) 2 (8.0)
Combination 4 (15.4) 8 (32.0)
Missing 3 (11.5) 4 (16.0)
SD, standard deviation.
5Animal-assisted interventions for dementia patients
Copyright # 2016 John Wiley & Sons, Ltd. Int J Geriatr Psychiatry 2016
a score that showed improvement on at least two levels
from T
0
to T
1
or from T
0
to T
2
were considered as
having a clinically significant improvement in their
depression symptoms. A subanalysis using mixed
models was used to test for the effect of attendance
at the AAA sessions. Attendance was grouped into
high (>90%) and low (<90%).
Results
No significant differences were found between the
intervention group and the control group at baseline
(Table 1). All of the participants in the control group
had a dementia diagnosis, but five did not in the AAA
group. For the latter participants, the mean mini-
mental state examination was 13.80 (SD =6.61, range:
7–23). There were 26 complete cases in the control
group (65.4% women) and 25 in the intervention
group (60% women). The mean age was 84.1 years in
the control group and 82.9 years in the intervention
group. Regarding CDR, 92% of the participants in each
of the two groups scored moderate or severe on the rat-
ing scale. The majority of the participants reported that
they enjoyed contact with animals.
The main effects of intervention and time are listed
in Table 2. No significant effects of the intervention
were found from T
0
to T
1
for depression in the total
sample (Table 3). However, the intervention group
had a continual decrease in the CSDD score, while
the control group had a continual increase in the
CSDD score, and a significant effect of the interven-
tion was found from T
0
to T
2
(Table 3). When strati-
fied on CDR, there was a close to and significant effect
on depression from T
0
to T
1
(p = 0.054) and T
0
to T
2
(p = 0.001) among participants with severe dementia
(Table 4). For participants with mild to moderate de-
mentia, the intervention showed no significant effects.
Also the significant difference between the groups
with regard to depression from T
0
to T
2
showed
clinical significance. More participants in the AAA
group improved than in the control group (p = 0.03)
(Table 5). A total of eight (17%) participants in the in-
tervention group improved by two levels on the CSDD
score, from T
0
to T
2
, but none in the control group.
Three participants (6.4%) from both the AAA group
and the control improved one level (Table 5).
There were no significant effects of the intervention
on change in agitation from either T
0
to T
1
or T
0
to T
2
(Table 3) or when stratified on cognitive level
(Table 4).
Significant effects of the intervention were found on
QoL for persons with severe dementia from both T
0
to
T
1
and T
0
to T
2
(Table 4). The control group showed
an increase in the QUALID score over the study
period, indicating a decline in QoL, whereas the AAA
group showed a decrease in the QUALID score. There
were no significant effects on QoL in the total sample
(Table 3) or in persons with mild to moderate demen-
tia (Table 4).
The number of sessions attended did not affect the
outcome of the CSDD, BARS or QUALID scores
(data not shown). The participation rate was high:
16 (64%) of the participants attended 90% or more
of the group sessions.
Discussion
The main finding in the study was significant statistical
and clinical improvement in symptoms of depression
from baseline (T
0
) to follow-up 12 weeks after end of
the intervention (T
2
) in the AAA group compared
with the control group. The intervention effect on
depression was found to be associated with severe
dementia. For patients with severe dementia, the inter-
vention also showed significant effects on QoL in the
change from T
0
to T
1
and T
2
. In the control group,
the symptoms gradually worsened during the study
period. The intervention showed no significant effects
on agitation.
Although there have been inconsistent findings
regarding the effect of AAI on depression in patients
with dementia (Moretti et al., 2011; Mossello et al.,
2011), the decline in symptoms found in the AAA
group is in line with findings from earlier studies
(Majic et al., 2013; Friedmann et al., 2015). In a simi-
lar study with AAI group intervention, Friedmann
et al. (2015) found that depression decreased during
the intervention period, while the reminiscing group,
used for comparison, did not experience a decrease
in depression. However, in contrast to the study
reported in the present article, no significant effect
was found between groups (Friedmann et al., 2015).
Majic et al. (2013) studied the effect of individual-
based AAI on depression in nursing home residents.
When using the Dementia Mood Assessment Scale,
they found that while the control group worsened
during the intervention period, the intervention group
showed constant frequency and severity in symptoms
of depression (Majic et al., 2013).
The level of agitation observed at baseline was in
line with a reliability study of the Norwegian version
of BARS (mean 24.2, SD 12.6) (Sommer et al., 2009)
and indicate observed agitated behaviour once or
twice per week. Agitation is one of the most difficult
6 C. Olsen et al.
Copyright # 2016 John Wiley & Sons, Ltd. Int J Geriatr Psychiatry 2016
NPS to manage in dementia patients. The lack of a
significant effect on agitation is in line with findings
from other AAI studies (Nordgren and Engstrom,
2014a; Friedmann et al., 2015; Thodberg et al.,
2015), although some early research have reported
positive effects (McCabe et al., 2002; Richeson, 2003;
Sellers, 2006).
Older persons with dementia often have a dimin-
ished QoL (Bárrios et al., 2012). This was confirmed
in the results of the study as there was a substantial de-
crease in QoL over time in participants with severe
dementia in the control group. AAA was found to
have an effect on both QoL and depression in the
group of patients with severe dementia. It is possible
that the AAA intervention might have been of particu-
lar value for this group, as patients with severe demen-
tia have been found to have a high prevalence of
unmet needs regarding meaningful activities and so-
cial contact (Cohen-Mansfield et al., 2015). Not only
might being part of a group intervention where a
dog is the centre of attention reduce the pressure in
social interaction, but also the dog might serve as a
Table 2 Estimates of main effects of intervention and time for CSDD, BARS and QUALID
Estimates of main effects
1
Dependent
variables
Control–intervention T
1
T
0
T
2
T
0
Estimate 95% CI Estimate 95% CI Estimate 95% CI
CSDD 1.78 2.88, 6.44 1.16 1.38, 3.70 0.89 1.29, 3.08
BARS 0.67 9.65, 10.99 1.25 5.35, 2.86 0.03 3.24, 3.17
QUALID 1.00 5.05, 7.06 0.33 3.74, 3.08 0.63 3.27, 2.00
CSDD, Cornell Scale for Depression in Dementia; BARS, Brief Agitation Rating Scale; QUALID, Quality of Life in Late-stage Dementia; T
0
,pre-test;
T
1
, post-test; T
2
, follow-up; CI, confidence interval..
1
A mixed model was used to estimate main effects.
Table 3 CSDD, BARS and QUALID for control and AAA (mean ± SD) and estimates of fixed effects
Dependent
variables
Pre-test
(T
0
)
Post-
test (T
1
)
Follow-
up (T
2
)
Estimates of fixed effects
1
T
1
T
0
T
2
T
0
Estimate tp
2
95% CI Estimate tp 95% CI
CSDD
Control 6.88
± 4.70
(n = 26)
8.28
± 5.62
(n = 25)
9.58
± 6.61
(n = 24)
2.09 1.38 0.171 5.09, 0.92 3.73 2.11 0.037 7.23, 0.23
AAA 8.35
± 4.65
(n = 23)
7.86
± 4.42
(n = 22)
7.41
± 5.01
(n = 22)
BARS
Control 23.19
± 11.39
(n = 26)
24.65
± 13.95
(n = 26)
24.00
± 13.20
(n = 25)
1.43 0.64 0.525 5.88, 3.02 0.50 0.17 0.864 6.20, 5.21
AAA 23.44
± 7.64
(n = 25)
23.75
± 7.13
(n = 24)
24.87
± 8.34
(n = 23)
QUALID
Control 22.92
± 8.50
(n = 26)
25.31
± 10.26
(n = 26)
26.48
± 10.05
(n = 25)
1.75 0.95 0.344 5.41, 1.92 3.60 1.50 0.136 8.34, 1.15
AAA 23.92
± 6.99
(n = 25)
24.80
± 5.79
(n = 24)
24.57
± 6.58
(n = 23)
CSDD, Cornell Scale for Depression in Dementia; BARS, Brief Agitation Rating Scale; QUALID, Quality of Life in Late-stage Dementia; AAS,
animal-assisted activity; CI, confidence interval.
1
A mixed model was used to estimate time trends between the groups.
2
Significance level 0.05.
7Animal-assisted interventions for dementia patients
Copyright # 2016 John Wiley & Sons, Ltd. Int J Geriatr Psychiatry 2016
mediator for conversation and lead to social cohesion
within the group (Beetz et al., 2012). The effect found
at T
2
for both depression and QoL may indicate that
the intervention initiated a process that continued be-
yond the end of intervention period. The intervention
may have contributed to an increase in social
interaction in general between the participants and
staff. Earlier research has shown that AAI might im-
prove social behaviour (Filan and Llewellyn-Jones,
2006), increase social interactions and conversations
(Bernstein et al., 2000; Kramer et al., 2009) and reduce
loneliness (Banks and Banks, 2002).
The study had several weaknesses that should be
considered when interpreting the results. Generaliza-
tion of the results should be done with caution be-
cause both the recruitment of the nursing homes and
participants might have been biased towards those
who regarded AAA as a positive activity.
Table 4 CSDD, BARS, QUALID stratified on CDR for control and AAA (mean ± SD) and estimates of fixed effects
Dependent
variables
Pre-test
(T
0
)
Post-
test (T
1
)
Follow-
up (T
2
)
Estimates of fixed effects
1
T
1
T
0
T
2
T
0
Estimate tp
2
95% CI Estimate tp95% CI
CSDD mild/moderate dementia
Control 6.36
± 5.56
(n = 14)
8.15
± 6.09
(n = 13)
10.50
± 8.18
(n = 14)
1.81 0.66 0.513 7.35, 3.73 4.46 1.45 0.151 10.58, 1.67
AAA 8.77
± 6.39
(n = 13)
9.36
± 6.02
(n = 11)
8.55
± 6.64
(n = 11)
CSDD severe dementia
Control 11.25
± 6.74
(n = 12)
12.92
± 8.08
(n = 12)
16.70
± 11.72
(n = 10)
5.04 1.99 0.054 10.17, 0.09 11.00 3.67 0.001 17.01,5.00
AAA 13.50
± 5.28
(n = 10)
11.00
± 6.91
(n = 11)
7.91
± 5.43
(n = 11)
BARS mild/moderate dementia
Control 21.43
± 10.09
(n = 14)
21.71
± 12.63
(n = 14)
21.79
± 11.40
(n = 14)
0.48 .017 0.866 5.23, 6.20 0.09 0.03 0.980 7.40, 7.21
AAA 21.92
± 6.13
(n = 13)
22.69
± 5.92
(n = 13)
21.92
± 8.80
(n = 12)
BARS severe dementia
Control 25.25
± 12.88
(n = 12)
28.08
± 15.17
(n = 12)
26.82
± 15.27
(n = 11)
3.68 1.02 0.317 11.04, 3.67 0.95 0.24 0.811 8.89, 6.99
AAA 25.08
± 8.99
(n = 12)
25.00
± 8.47
(n = 11)
28.09
± 6.77
(n = 11)
QUALID mild/moderate dementia
Control 20.36
± 5.96
(n = 14)
23.07
± 9.50
(n = 14)
23.00
± 6.56
(n = 14)
1.05 0.40 0.692 4.27, 6.38 1.47 0.47 0.643 4.85, 7.79
AAA 21.46
± 7.00
(n = 13)
25.23
± 5.10
(n = 13)
25.83
± 8.08
(n = 12)
QUALID severe dementia
Control 25.91
± 10.21
(n = 12)
27.92
± 10.90
(n = 12)
30.91
± 12.15
(n = 11)
5.08 2.33 0.035 9.79,
0.37 9.79 3.15 0.003 16.03,3.54
AAA 26.58
± 6.17
(n = 12)
24.27
± 6.72
(n = 11)
23.18
± 4.40
(n = 11)
Cornell Scale for Depression in Dementia (CSDD), Brief Agitation Rating Scale (BARS), and Quality of Life in Late-stage Dementia (QUALID);
AAS, animal-assisted activity; CI, confidence interval.
1
A mixed model was used to estimate time trends between the groups.
2
Significance level 0.05.
8 C. Olsen et al.
Copyright # 2016 John Wiley & Sons, Ltd. Int J Geriatr Psychiatry 2016
The instruments used to measure the outcomes
were standardized, validated and reliable (Swift et al.,
2002; Korner et al., 2006; Barca et al., 2010; Sommer
and Engedal, 2011); moreover, an excellent interrater
reliability was found. However, the raters were not
blind to whether the participants were part of an
AAA group or a control group. Although this might
have influenced the positive change seen for depres-
sion and QoL, the trend towards increased agitation
indicates that raters were not biased.
When using treatment as usual as a control condi-
tion, there is always a possibility that any observed
effect of the intervention is merely a novelty effect.
However, all participants in the study were offered a
range of regular activities, and the AAA were additional
to these. Using another activity as control condition
would therefore b e both difficult in practice and imply
a wish to compare different intervent ions’ effective-
ness, which was not within the scope of the study.
Furthermore, it could be argued that the dog handler,
not the dog, is the decisive factor in AAIs. By defini-
tion, AAA implies a human and animal team, a nd
using a control condition without a dog was theref ore
not considered.
A strength of the study lies in its design, as random-
ized controlled trials are the most robust evaluative
method (Puffer et al., 2005). Methodological issues
in cluster randomized trials are straightforward and
manageable (Murphy et al., 2006), and we considered
these issues carefully. The assessment of the long-term
effects is a further strength of our study. The moderate
dropout rate (17%) was as expected, because of the
population’s age and progressive decease.
There is a need for high-quality research in non-
pharmacological interventions for older people with
dementia (Iden et al., 2014), and the present results
contribute to a better understanding of the feasibility
and effect of AAA programmes for older people with
dementia. The fact that the statistical difference in
the CSDD also showed significant clinical relevance
renders the results valuable for clinical practice.
Conclusion
The significant improvements in depression and QoL
show that complementary treatment such as AAA
may be useful in dementia care. The effects were
found for persons with severe dementia, which sup-
ports the importance of individually tailored interven-
tions where participants’ cognitive and functional
levels are taken into account.
Conflict of interest
The first-named author owns a share in the Norwegian
Centre of Anthrozoology, which was a partner in the
study project.
Key points
• The prevalence of neuropsychiatric symptoms
in cognitively impaired nursing home residents
is high.
• Non-pharmacological treatment is recommended.
• Significant improvements to both the severity
of depression and quality of life were found in
persons with severe dementia in the animal-
assisted intervention group compared with the
control group.
• Animal-assiste d activity may be effective in de-
mentia care.
Acknowledgements
The project was funded by grant no. 217516 from
Oslofjordfondet and RFF Hovedstaden, NMBU, and
cooperating partners (the Norwegian Centre of An-
throzoology, Buskerud, and Vestfold University
College, Centre for Development of Institutional and
Home Care Services, Vestfold). Cooperating partners
supported the project through internal funding.
The authors also thank the participants, the nursing
homes and health workers, the dogs and their handlers
and the cooperating partners.
Table 5 Clinically significant change on subject level in Cornell Scale
for Depression in Dementia (chi-square and p-value)
T
1
T
0
T
2
T
0
Control
group
(n = 26)
N (%)
AAA
group
(n = 23)
N (%)
Control
group
(n = 25)
N (%)
AAA
group
(n = 22)
N (%)
Improved 3.00 0 (0) 1 (2.0) 0 (0) 0 (0)
2.00 2 (4.1) 2 (4.1) 0 (0) 8 (17.0)
1.00 4 (8.2) 4 (8.2) 3 (6.4) 3 (6.4)
No
change
0.00 11
(22.4)
11
(22.4)
13
(68.4)
6 (31.6)
Worse 1.00 7 (14.3) 5 (10.2) 5 (10.6) 3 (6.4)
2.00 1 (2.0) 0 (0) 3 (12) 2 (4.3)
3.00 1 (2.0) 0 (0) 1 (2.1) 0 (0)
χ
2
= 3.16, p = 0.79 χ
2
= 12.14, p = 0.03
T
0
, pre-test; T
1
, post-test; T
2
, follow-up; AAS, animal-assisted
activity.
9Animal-assisted interventions for dementia patients
Copyright # 2016 John Wiley & Sons, Ltd. Int J Geriatr Psychiatry 2016
References
Aalten P, de Vugt ME, Jaspers N, Jolles J, Verhe y FR. 2005. The course of neuropsy-
chiatric symptoms in dementia. Part I: findings from the two-year longitudinal
Maasbed study. Int J Geriatr Psychiatry 20: 523–530.
Alexopoulos GS. 2002. The Cornell Scale for Depression in Dementia: Administra-
tion and Scoring Guidelines, WMCoCU Cornell Institute of Geriatric
Psychiatry (ed.)(ed.).. Weill Medical College of Cornell University: White
Plains, NY; 10.
Alexopoulos GA, Abram s RC, Young RC, Shamoian CA. 1988. Cornell scale for
depression in dementia. Biol Psychiatry 23: 271–284. Til norsk ved Dag Årsland.
Alexopoulus GS, Abrams RC, Young RC. 1988. Cornell scale for depression in
dementia. Biol Psychiatry 23: 271–284.
Banks MR, Banks WA. 2002. The effects of animal-assisted therapy on loneliness in
an elderly population in long-term care facilities. J Gerontol A Biol Sci Med Sci
57: 428–432.
Barca ML, Engedal K, Selbæk G. 2010. A reliability and validity study of the Cornell
scale among elderly inpatients, using various clinical criteria. Dement Geriatr Cogn
Disord 29: 438–447.
Barca ML, Engedal K, Laks J, Selbæk G. 2011. Quality of life among elderly patients
with dementia in institutions. Dement Geriatr Cogn Disord 31: 435–442.
Bárrios H, Narciso S, Guerreiro M, et al. 2012. Quality of life in patients with mild
cognitive impairment. Aging Ment Health 17: 287–292.
Beerens HC, Zwakhalen SM, Verbeek H, Ruwaard D, Hamers JP. 2013. Factors asso-
ciated with quality of life of people with dementia in long-term care facilities: a
systematic review. Int J Nurs Stud 50: 1259–1270.
Beetz A, Uvnas-Moberg K, Julius H, Kotrschal K. 2012. Psychosocial and psychophys-
iological effects of human-animal interactions: the possible role of oxytocin. Front
Psychol 3: 234.
Bernstein PL, Friedmann E, Malaspina A. 2000. Animal-assisted therapy enhances
resident social interaction and initiation in long-term care facilities. Anthrozoös
13: 213–224.
Cohen-Mansfield J, Marx MS, Rosenthal AS. 1989. A description of agitation in a
nursing home. J Gerontol Biol Med Sci 44: M77–M84.
Cohen-Mansfield J, Dakheel-Ali M, Marx MS, Thein K, Regier NG. 2015. Which
unmet needs contribute to behavior problems in persons with advanced dementia?
Psychiatry Res 228:59–64.
Douglas S, James I, Ballard C. 2004. Non-pharmacological interventions in dementia.
Adv Psychiatr Treat 10: 171–177.
Engedal K, Haugen PK. 1993. The prevalence of dementia in a sample of elderly
Norwegians. Int J Geriatr Psychiatry 8: 565–570.
Filan SL, Llewellyn-Jones RH. 2006. Animal-assisted therapy for dementia: a review of
the literature. Int Psychogeriatr/IPA 18: 597–611.
Finkel SI, Lyons JS, Anderso n RL. 1993. A brief agitation rating scale (BARS) for
nursing home elderly. J Am Geriatr Soc 41:50–52.
Folstein MF, Folstein SE, McHugh PR. 1975. ‘Mini-mental state’. A practical method for
grading the cognitive state of patients for the clinician. J Psychiatr Res 12:189–198.
Friedmann E, Galik E, Thomas SA, et al. 2015. Evaluation of a pet-assisted living
intervention for improving functional status in assisted living residents with mild
to moderate cognitive impairment: a pilot study. Am J Alzheimers Dis Other Demen
30: 276–289.
Geldmacher DS, Frolich L, Doody RS, et al. 2006. Realistic expectations for treatment
success in Alzheimer’s disease. J Nutr Health Aging 10: 417–429.
Hallgren KA. 2012. Computing inter-rater reliability for observational data: an
overview and tutorial. Tutor Quant Methods Psychol 8:23–34.
Hughes CP, Berg L, Danziger WL, Coben LA, Martin RL. 1982. A new clinical scale
for the staging of dementia. Br J Psychiatry 140: 566–572.
IAHAIO. 2014. IAHAIO White Paper. In The IAHAIO Definitions for Animal
Assisted Intervention and Guidelines for Wellness of Animals Involved. Amster-
dam: International Association of Human-Animal Interaction Organizations.
Iden KR, Engedal K, Hjorleifsson S, Ruths S. 2014. Prevalence of depression among
recently admitted long-term care patients in Norwegian nursing homes: associa-
tions with diagnostic workup and use of antidepressants. Dement Geriatr Cogn
Disord 37: 154–162.
van Iersel MB, Hoefsloot W, Munneke M, Bloem BR, Olde Rikkert MG. 2004.
Systematic review of quantitative clinical gait analysis in patients with dementia.
Z Gerontol Geriatr 37:27–32.
Korner A, Lauritzen L, Abelskov K, et al. 2006. The Geriatric Depression Scale and the Cor-
nell Scale for Depression in Dementia. A validity study. Nord J Psychiatry 60:360–364.
Kramer SC, Friedmann E, Bernstein PL. 2009. Comparison of the effect of human
interaction, animal-assisted therapy, and AIBO-assisted therapy on long-term care
residents with dementia. Anthrozoös 22:43–57.
Lobo A, Launer LJ, Fratiglioni L,
et al. 2000. Prevalence of dementia and major
subtypes in Europe: a collaborative study of population-based cohorts. Neurologic
diseases in the elderly research group. Neurology 54:S4–S9.
Lyketsos CG, Lopez O, Jones B, et al. 2002. Prevalence of neuropsychiatric symptoms
in dementia and mild cognitive impairment: results from the cardiovascular health
study. JAMA 288: 1475–1483.
Majic T, Gutzmann H, Heinz A, Lang UE, Rapp MA. 2013. Animal-assisted therapy
and agitation and depression in nursing home residents with dementia: a matched
case–control trial. Am J Geriatr Psychiatr 21: 1052–1059.
McCabe BW, Baun MM, Speich D, Agrawal S. 2002. Resident dog in the Alzheimer’s
special care unit. West J Nurs Res 24: 684–696.
McKhannG,DrachmanD,FolsteinM,et al. 1984. Clinical diagnosis of Alzheimer’sdisease:
report of the NINCDS-ADRDA work group under the auspices of Department of Health
and Human Services Task Force on Alzheimer’sDisease.Neurology 34: 939–944 .
Mjørud M, Kirkevold M, Rosvik J, Selbæk G, Engedal K. 2014a. Variables associated
to quality of life among nursing home patients with dementia. Aging Ment Health
18: 1013–1021.
Mjørud M, Røsvik J, Rokstad AMM, Kirkevold M, Engedal K. 2014b. Variables
associated with change in quality of life among persons with dementia in nursing
homes: a 10 months follow-up study. PLoS One 9: e115248: .
Moretti F, De Ronchi D, Bernabei V, et al. 2011. Pet therapy in elderly patients with
mental illness. Psychogeriatrics 11: 125–129.
Mossello E, Ridolfi A, Mello AM, et al. 2011. Animal-assisted activity and emotional sta-
tusofpatientswithAlzheimer’s disease in day care. Int Psychogeriatr/IPA 23:899–905.
Murphy AW, Es terman A, Pilotto LS. 2006. Cluster randomized controlled trials in
primary care: an introduction. Eur J Gen Pract 12:70–73.
Nagatomo I, Kita K, Takigawa M, Nomaguchi M, Sameshima K. 1997. A study of the
quality of life in elderly people using psychological testing. Int J Geriatr Psychiatry
12: 599–608.
Nordgren L, Engstrom G. 2014a. Animal-assisted intervention in dementia: effects on
quality of life. Clin Nurs Res
23:7–19.
Nordgren L, Engstrom G. 2014b. Effects of dog-assisted intervention on behavioural
and psychological symptoms of dementia. Nurs Older People 26:31–38.
Nygaard HA. 2002. Sykehjemmet som medisinsk institusjon—et fatamorgana?
Tidsskr Nor Laegeforen 122: 823– 825.
Nygaard HA, Ruths S. 2003. Missing the diagnosis: senile dementia in patients admit-
ted to nursing homes. Scand J Prim Health Care 21: 148–152.
Puffer S, Torgerson DJ, Watson J. 2005. Cluster randomized controlled trials. J Eval
Clin Pract 11: 479–483.
Richeson NE. 2003. Effects of animal-assisted therapy on agitated behaviors and
social interactions of older adults with dementia. Am J Alzheimers Dis Other Demen
18: 353–358.
Røen I, Selbæk G, Kirkevold O, et al. 2015. The reliability and validity of the
Norwegian version of the quality of life in late-stage dementia scale. Dement
Geriatr Cogn Disord 40: 233–242.
Seitz DP, Adunuri N, Gill SS, et al. 2011. Antidepressants for agitation and psychosis
in dementia. Cochrane Database Syst Rev. DOI: 10.1002/14651858.CD008191.pub2
Alternate journal: The Cochrane database of systematic reviews, ISSN: 1361-6137.
Selbæk G. 2005. Atferdsforstyrrelser og psykologiske symptomer ved demens. Tidsskr
Norske Lægeforening 125: 1500– 1502.
Selbæk G, Kirkevold Ø, Engedal K. 2007a. The prevalence of psychiatric symptoms
and behavioural disturba nces and the use of psychotropic drugs in Norwegian
nursing homes. Int J Geriatr Psychiatry 22: 843–849.
Selbæk G, Kirkevold Ø, Engedal K. 2007b. The prevalence of psychiatric symptoms
and behavioural disturba nces and the use of psychotropic drugs in Norwegian
nursing homes. Int J Geriatr Psychiatry 22: 843–849.
Sellers DM. 2006. The evaluation of an animal assisted therapy intervention for elders
with dementia in long-term care. Activities, Adaptation & Aging 30:61–77.
Sommer OH, Engedal K. 2011. Reliability and validity of the Norwegian version of
the Brief Agitation Rating Scale (BARS) in dementia. Aging Ment Health 15:
252–258.
Sommer OH, Aga O, Cvancarova M, et al. 2009. Effect of oxcarbazepine in the treat-
ment of agitation and aggression in severe dementia. Dement Geriatr Cogn Disord
27: 155–
163.
Strobel C, Engedal K. 2009. Norsk Revidert Mini Mental Status Evaluering
(MMSE-NR).
Swift RH, Harrigan EP, Cappelleri JC, Kramer D, Chandler LP. 2002. Validation of
the behavioural activity rating scale (BARS)
™
: a no vel measure of activity in agi-
tated patients. J Psychiatr Res 36:87–95.
Telenius EW, Engedal K, Bergland A. 2013. Physical performance and quality of life
of nursing-home residents with mild and moderate dementia. Int J Environ Res
Public Health 10: 6672–6686.
Teri L, Logsdon RG, Uomoto J, McCurry SM. 1997. Behavioral treatment of depres-
sion in dementia patients: a controlled clinical trial. J Gerontol B Psycho l Sci Soc Sci
52B: P159–P166.
Thodberg K, Sorensen LU, Christensen JW, et al. 2015. Therapeutic effects of dog
visits in nursing homes for the elderly. Psychogeriatrics. DOI: 10.1111/psyg.12159
ISSN: 1346-3500.
Tripathi M, Vibha D. 2010. An approach to and the rationale for the pharmacological
management of behavioral and psychological symptoms of dementia. Ann Indian
Acad Neurol 13: S94–S98.
Weiner MF, Martin-Cook K, Svetlik DA, et al. 2000. The quality of life in late-stage
dementia (QUALID) scale. J Am Med Dir Assoc 1: 114–116.
West BT. 2009. Analyzing longitudinal data with the linear mixed models procedure
in SPSS. Eval Health Prof 32: 207–228.
10 C. Olsen et al.
Copyright # 2016 John Wiley & Sons, Ltd. Int J Geriatr Psychiatry 2016
