Available online at www.sciencedirect.
com
ScienceDirect
Attachment and aged care: a systematic review of current
research
Gery C Karantzas, Daniel Romano and Juwon Lee
Attachment theory is an important framework from which to
examine familial aged care. Despite this, the role of attachment
in later-life caregiving remains unclear. The current study
presents a systematic review of papers within the last five years
on attachment and various outcomes related to familial aged
care. For the caregiver, attachment anxiety was associated
with poorer mental health, and attachment insecurity with a
more controlling manner of caregiving. The few studies
conducted with care recipients found that attachment
insecurity was associated with greater self-appraisals of
dementia symptoms and a lower sense of security. Research
continues to suffer from the use of inadequate assessments of
individual differences into adult attachment.
Address
School of Psychology, Deakin University, 221 Burwood Highway,
Burwood, Victoria 3125, Australia
Corresponding author: Karantzas, Gery C (gery.karantzas@deakin.edu.
au)
Current Opinion in Psychology 2019, 25:37–46
This review comes from a themed issue on Attachment in adulthood
Edited by Jeffry A Simpson and Gery Karantzas
https://doi.org/10.1016/j.copsyc.2018.02.016
2352-250X/ã 2018 Elsevier Ltd. All rights reserved.
Caring for an ageing family member (usually a parent or
spouse) can be both difficult and taxing [1,2]. This is not
only due to the demands of the caregiving role, but also
because a carer must come to terms with the ailing health
and potential passing of a significant other that has been
an important source of love, comfort, and security [3,4].
We therefore contend that attachment theory [5] is an
important framework for understanding various aspects of
familial aged care, including behaviours and processes
related to caregiving and care receiving, as well as the
psychological and physical wellbeing of both carers and
care recipients.
Research on attachment and families in later life goes as far
back as the mid-1970s [6], and yet, our understanding of
attachment processes during the later stages of family life
and within the context of aged care is somewhat limited.
www.sciencedirect.com
One of the major reasons for this limited understanding
relates to the inherent difficulties in how best to operatio-
nalize and measure adult familial analogues of concepts
central to attachment theory (such as the strength of
attachment bonds, attachment states of mind, and attach-
ment styles) previously studied in infant-parent and roman-
tic attachment bonds. Many studies claiming to have
investigated attachment within the context of caregiving
in later life have often conflated attachment with related
but distinct concepts, such as affection, intergenerational
solidarity, and relationship closeness [7,8], instead of clearly
articulating assessments of attachment styles and attach-
ment states of mind. It is only in the last decade that
researchers studying ageing families have made more con-
certed efforts to directly use or adapt attachment measures
which not only reflect mainstream developmental and
social psychology approaches to the study of individual
differences in attachment, but capture many of Bowlby’s
central ideas regarding aspects of attachment cognitions and
behaviours [4]. To this end, our understanding of attach-
ment in later life familial bonds and aged care is muddied
by these conceptual and measurement issues.
Given these issues, we undertook a systematic review of
the aged care literature using inclusion criteria that reflect
assessments of adult attachment that are widely accepted
to reflect concepts akin with individual differences in
attachment, whether they be differences in: adult attach-
ment styles (i.e. people’s chronic cognitions and beha-
viours regarding close relationships [9,10], attachment
states of mind (i.e. coherence of narrative regarding early
attachment experiences, C. George et al., unpublished),
or attachment strength (i.e. the closeness of the attach-
ment bond [11]). In doing so, we aimed to better under-
stand the associations between individual differences in
adult attachment and processes related to caregiving and
care receiving, as well as the physical and mental health
outcomes for both carers and care recipients. Given that
the focus of COIP is to review research published in the
last few years, we limited our qualitative synthesis to
recently published papers.
Method
The systematic review was conducted in line with the
Preferred Reporting Items for Systematic Reviews and
Meta-Analyses (PRISMA; see Figure 1). Electronic data-
bases were used to conduct literature searches with a
variety of keywords to identify articles (see Figure 1).
Further articles were identified through scanning the
reference lists of publications recovered through the
Current Opinion in Psychology 2019, 25:37–46
38 Attachment in adulthood

Figure 1

Electronic database search included:
PsychINFO, PsycARTICLES, PsycBOOKS,
Psychology and Behavioral Sciences
Collection, AgeLine, CINAL, Medline,
Academic Search Complete, and Scopus.
Identification
Keywords used to search titles, abstracts and full content
included:
attachment, AND care*, AND couple, AND spous*, AND
(son* OR daughter*), AND adult children, AND family NOT
infant*, AND (aging OR aged OR ageing OR elder*), AND
older adult, AND (parent* OR mother OR father OR spouse
OR partner) NOT infan*, AND (dementia OR Alzheimer* OR
chronic illness OR palliative).

Records identified through Additional records identified

database searching through other sources
(n = 321) (n = 6)

Records after duplicates removed

(n = 33)
Screening

Records screened Records excluded

(n = 294) (n = 259)
Inclusion criteria: Reasons for exclusion:
• Written in English • Did not include a valid measure of
• Peer reviewed publication (empirical) attachment individual differences (n = 206)
• Evaluated security priming (as oppose to focusing • Not written in English (n = 21)
on insecurity priming) • Dissertation (n = 25)
• Assessed direct association between individual • Book or chapter review of area (not
reporting new findings [n =7])
Eligibility

differences in attachment and outco mes for

familial carers and/or care recipients
• Sample include carers or care recipients 50+
years of age

Full-text articles excluded,

with reasons
Full-text articles assessed (n = 24)
for eligibility
(n = 35) Reasons for exclusion:

• Did not include a valid measure

Included

of attachment individual
differences (n = 20)
• Attachment style used as an
Studies included in outcome variable (n = 1)
qualitative synthesis • Sample too young (n = 3)
(n = 11)

Current Opinion in Psychology

A flow diagram detailing the application of PRISMA to the qualitative synthesis of published studies into attachment and familial aged care
conducted between January 2013 and Janauary 2018.

databases to ensure relevant studies were not missed. We Results

limited the search to articles published between January
2013 and January 2018. Each study was assessed against a
set of inclusion and exclusion criteria. Excluded studies
were tabulated against reasons for exclusion (see Fig-
ure 1). A full-text review was conducted for eligible
studies, with the finalised set of published studies sub-
jected to qualitative synthesis.
Three hundred and twenty-seven published studies were
identified as part of the systematic search, with a final set of
11 studies included for qualitative synthesis (details for
each study are presented in Tables 1 and 2). The features
and characteristics (research design, samples, and attach-
ment measures and outcome variables used) of each study
along with sample characteristics are presented across

Current Opinion in Psychology 2019, 25:37–46 www.sciencedirect.com

www.sciencedirect.com

Table 1

Sample size and characteristics.

Study Sample size and Recruitment of Data collection Ethnicity Age Relationship type Length of Primary condition for
care group sample method caregiving/care- care
receiving
Ávila, Brandão, N: 124 caregivers Public hospital and Pen and paper Not specified M age = 52.74 100% spousal Clinical diagnosis 100% breast cancer
Coimbra, Lopez non-for-profit (ranging 27–77) caregiver within last 3 years
and Mena Matos, cancer organisation
2016 [12]
Burke, Clark, Ali, N: 57 caregivers Hospital (palliative Pen and paper (in Caregiver: M age 29.8% spousal Not specified Terminal illness (e.g.
Gibson, care unit) person) 61.4% Caucasian caregivers = 56.11 caregiver cancer, renal disease,
Smigelsky and 38.6% African- (ranging 29–87) 42.1% adult child dementia, chronic
Neimeyer, American M age care- caregiver obtrusive pulmonary
2015 [13] Care-recipient: recipients = 71.5 19.3% sibling disease)
59.6% Caucasian (ranging 48–93) caregiver
40.4% African- 8.8% other
American caregiver (e.g.
niece)
Chen, Uzdawinis, N: 91 caregivers Subset of larger Pen and paper 91% Caucasian M age 100% adult child M (months) = 53 100% dementia

Attachment and aged care: a systematic review Karantzas, Romano and Lee 39
Schölmerich and study caregivers = 55 caregiver
Juckel, 2014 [14] M age care-
recipients = 85
(ranging 64–100)
Fasse, Flahault, N: 60 caregivers Hospital (palliative Pen and paper Not specified M age = 62.39 100% adult child M (months since 36.6% breast cancer
Brédart, care unit) caregiver diagnosis) = 68.7 16.% lung cancer
Dolbeault and 10% cervix cancer
Sultan, 2015 [24] 36.6% other cancer
diagnosis
Hasson-Ohayon, N: 60 caregivers Hospital Pen and paper (in Not specified M age = 57 100% spousal At least 3 months 30% stage 3 colorectal
Goldzweig, Sela- person) caregiver cancer
Oren, Pizem, Bar- 70% stage 4 colorectal
Sela and Wolf, cancer
2015 [18]
Hsiao, Jow, Kuo, N: 34 dyads General hospital Pen and paper (in Not specified M age 61.7% spousal Not specified 100% breast cancer
Current Opinion in Psychology 2019, 25:37–46

Huang, Lai, Liu (34 caregivers, (outpatients of person) caregivers = 53.7 caregiver 26.3% stage 1 breast
and Chang, 34 care recipients) breast cancer (ranging 35–70) cancer
2014 [16] surgery) M age care- 42.1% stage 2 breast
recipients = 49.6 cancer
(ranging 21–63)
Lai, Luciani, Galli, N: 60 caregivers Hospital (palliative Pen and paper (in Not specified M age 35% partner Not specified 100% terminal cancer
Morelli, care unit) person) caregivers = 57.4 caregiver
Cappelluti, M age care- 56.6% adult child
Penco, Aceto and recipients = 75.50 caregiver
Lombardo, 8.3% sibling
2015 [26]
Current Opinion in Psychology 2019, 25:37–46

40 Attachment in adulthood
Table 1 (Continued )
Study Sample size and Recruitment of Data collection Ethnicity Age Relationship type Length of Primary condition for
care group sample method caregiving/care- care
receiving
Lee, Sohn, Lee, N: 98 caregivers General hospital Pen and paper (in Not specified M age 100% adult child Not specified 71.4% Alzheimer’s
Seong, Park and person or mail) caregivers = 51.77 caregiver disease
Lee, 2018 [27] M age care- 19.4% vascular
recipients = 80.47 dementia
5.1% Lewy Body
dementia
3.1% fronto-temporal
dementia
1% dementia
otherwise not specified
Milberg, N: 168 care Palliative home- Pen and paper 89.2% Swedish M age = 68.75 Not specified Not specified 100% in palliative care
Friedrichsen, recipients care units (ranging 36–94)
Jakobsson,
Nilsson, Niskala,
Olsson, Wahlberg
and Krevers,
2014 [19]
Monin, Schulz and N: 58 dyads Alzheimer’s disease Pen and paper (in Caregivers: M age 100% spousal At least 3 months of 100% Alzheimer’s
Kershaw, (58 caregivers, research centre person) 94.8% Caucasian caregivers = 71.10 caregiver in-home care disease
2013 [15] 58 care recipients) 3.4% African- M age care-
American recipients = 73.95
1.7% Asian
Care-recipients:
94.8% Caucasian
3.4% African-
American
1.7% Asian
Paulson and N: 165 caregivers Online (MTurk) Online (MTurk) 82.6% Caucasian M age = 52.2 100% adult child Not specified N/A
Bassett, 7.3% African- (ranging 45–65) caregiver
2016 [17] American
2.8% Asian
1.8% Native
American
1.8% Hispanic
3.7% Bi-racial
www.sciencedirect.com
 Attachment and aged care: a systematic review Karantzas, Romano and Lee 41

Table 2

Outcomes of included studies.

Study Study design Caregiver or care- Outcome variable Adult attachment Findings
recipient measure
Ávila, Brandão, Luı́s Cross-sectional; Caregiver Controlling Shortened ECR [28] Controlling
Coimbra, Lopez correlational caregiving – relationship not caregiving
and Mena Matos, Proximity specified Anx +
2016 [12] caregiving Av +
Physical well-being Sec 0
Self-esteem Proximity
caregiving
Anx
Av
Sec 0
Physical well-being
Anx +
Av +
Sec 0
Self-esteem
Anx
Av
Sec +
Burke, Clark, Ali, Cross-sectional; Caregiver Anticipatory grief AAS – measured Anticipatory grief
Gibson, correlational general attachment Anx 0
Smigelsky and Sec
Neimeyer,
2015 [13]
Chen, Uzdawinis, Cross-sectional; Caregiver Caregiver 25-item PBI and Caregiver
Schölmerich and correlational satisfaction Cicirelli’s [11] satisfaction
Juckel, 2014 [14] Psychological attachment Att str +
symptoms of measure – Psychological
dementia Measured symptoms
Stress attachment to Att str
mother Stress
Att str 0
Fasse, Flahault, Cross-sectional; Caregiver Depression ECR – relationship Depression
Brédart, correlational not specified Anx +
Dolbeault and Av 0
Sultan, 2015 [25]
Hasson-Ohayon, Cross-sectional; Caregiver Finding meaning ECR – relationship Finding meaning
Goldzweig, Sela- correlational through caregiving not specified Anx 0
Oren, Pizem, Bar- Av
Sela and Wolf,
2015 [18]
Hsiao, Jow, Kuo, Longitudinal; dyads Caregiver and care- Cortisol levels ECR-Revised [10] – Cortisol levels (CG)
Huang, Lai, Liu recipient relationship not Anx (CG) 0
and Chang, specified Av (CG) 0
2014 [16] Anx (CR) +
Av (CR) 0
Cortisol levels (CR)
Anx (CG) 0
Av (CG) 0
Anx (CR) 0
Av (CR) 0
Lai, Luciani, Galli, Cross-sectional; Caregiver Anxiety Italian ASQ [29] – Anxiety
Morelli, correlational Depression Relationship not Anx +
Cappelluti, Risk of prolonged specified Av 0
Penco, Aceto and grief Sec 0
Lombardo, Depression
2015 [26] Anx 0
Av 0
Sec 0
Risk of prolonged
grief
Anx +
Av 0
Sec 0

www.sciencedirect.com Current Opinion in Psychology 2019, 25:37–46

42 Attachment in adulthood

Table 2 (Continued )
Study Study design Caregiver or care- Outcome variable Adult attachment Findings
recipient measure
Lee, Sohn, Lee, Cross-sectional; Caregiver Activities of daily Korean ECR- ADL (CR)
Seong, Park and correlational living Revised [30] – Anx 0
Lee, 2018 [27] Caregiver burden Relationship not Av 0
Filial obligation specified Caregiver burden
Anx +
Av 0
Filial obligation
Anx +
Av 0

Milberg, Cross-sectional; Care-recipient Sense of security 16-item ECR [31] – Sense of security
Friedrichsen, correlational during palliative relationship not Anx
Jakobsson, care specified Av
Nilsson, Niskala,
Olsson, Wahlberg
and Krevers,
2014 [19]

Monin, Schulz and Correlational; Caregiver and care- Physical symptoms 12-item ECR – Physical dementia
Kershaw, Dyads recipient of dementia Relationship not symptoms (CG)
2013 [15] Psychological specified Anx (CG) 0
symptoms of Av (CG) 0
dementia Anx (CR) 0
Av (CR) 0
Physical dementia
symptoms (CR)
Anx (CG) 0
Av (CG) 0
Anx (CR) +
Av (CR) +
Psychological
dementia
symptoms (CG)
Anx (CG) 0
Av (CG) +
Anx (CR) 0
Av (CR) 0
Psychological
dementia
symptoms (CR)
Anx (CG) +
Av (CG) 0
Anx (CR) +
Av (CR) 0
Paulson and Cross-sectional; Caregiver Preparation for Revised AAS [32] – Preparation
Bassett, Correlational caregiving Measured general activities
2016 [17] Filial obligation attachment Anx 0
Sec +
Filial obligation
Sec +

AAS = Adult Attachment Scale; ASQ = Attachment Style Questionnaire; ECR = Experiences in Close Relationships Questionnaire; PBI = Parental
Bonding Instrument; Anx = Attachment anxiety; Att Str = Attachment strength; Av = Attachment avoidance; Sec = Attachment security; CG = Care-
giver; CR = Care-recipient.

Tables 1 and 2. In Table 3, a narrative summary is provided
to orient the reader to the trends across studies regarding
study features, designs, and sample characteristics.
Review findings
Carers
Attachment anxiety was associated with mental health
issues (i.e. increased depression, anxiety, burden, and
grief symptoms) and poor self-esteem in 66.6% of mental
health outcomes assessed across studies. In contrast,
attachment avoidance was not associated with mental
health outcomes in any study, but was negatively associ-
ated with self-esteem in the only study to assess this
outcome [12]. Attachment security was not significantly
associated with mental health problems or self-esteem,
other than in one study of anticipatory grief in which it
was negatively associated [13]. Attachment strength
(measured in one study [14]) was found to be positively

Current Opinion in Psychology 2019, 25:37–46 www.sciencedirect.com

www.sciencedirect.com
Table 3
A summary of the design and characteristics of the 11 studies included in the systematic review.
Recruitment and study
characteristics
 3 studies were conducted in the
United States and Europe (Italy,
France, Portugal). The remaining
studies were conducted in Israel,
Korea, Taiwan, and Sweden.
 7 studies recruited participants
from hospitals (palliative care units,
cancer outpatient departments or
community palliative care), 1 study
also recruited cancer patients
through a not-for-profit cancer
organization [12], and 1study
recruited participants online via
MTurk [17]. 1 study drew on
existing data as part of the German
national study [14].
 10 studies administered paper-
and-pen surveys, with 1 study [17]
administering questionnaires online.
 10 studies were cross-sectional/
correlational in design and 1 of these
studies was a dyadic study of
spouses [15].
 1 study was a dyadic longitudinal
design focused on romantic couples
(however the sample was not
Current Opinion in Psychology 2019, 25:37–46
subjected to dyadic modelling).
a
Given that the differing measures of adult attachment style have been shown to converge along the broad dimensions of attachment anxiety and avoidance [9] (with some researchers extracting a
broad dimension of attachment security [8,12]), we grouped associations between attachment style and outcome variables along these three dimensions where possible.
b
The assessment of the PBI as a measure of attachment was not included in the qualitative synthesis as it is care measure of bonding rather than an explicit measure of adult attachment.
Sample characteristics
 The sample size across all
11 studies totalled 1235 participants
comprising 975 carers and 260 care
recipients.
 Sample size varied from 60 to
160 participants across studies, with
the two dyadic studies comprising
92 couples overall.
 Eight studies included samples
exclusively comprising carers, while
only one study exclusively
comprising care recipients
 Of the studies including carers
exclusively, 50% recruited adult
children while 25% recruited
spousal carers. One study recruited
a variety of informal carers.
 The time carers spent providing
care ranged from 3 to 69 months.
 Regarding the context of
caregiving/care receiving: 5 studies
investigated attachment and aged
care in people with cancer, 3 studies
exclusively focused on people with
various forms of dementia, 2 studies
included samples dealing with a
range of terminal illnesses, and
1 study involved prospective carers
of older parents.
 The mean age of carers across
studies ranged from 51 to 71 years
 The mean age of care recipients
ranged from 50 to 80 years.
Outcomes
 10 studies utilized self-report
assessments of the outcome variables for
carers and care recipients.
 1 dyadic study assessed carer and care
recipient cortisol levels [16].
 Numerous studies included multiple
outcome variables.
 Regarding carer outcomes:
 6 studies assessed various mental
health issues or strain (i.e. depression,
anxiety, stress, carer burden, grief
symptoms, and self-esteem)
 2 studies assessed carers’ reports of
psychological and/or physical symptoms of
dementia in care recipients.
 1 study investigated a carer’s style of
caregiving while another study examined
carer satisfaction.
 1 study examined carers’ preparedness
to provide care [17].
 2 studies focused on carers’ sense of
filial obligation [17,27].
1 study assessed finding meaning in the
caregiving role [18].
 1 study measured stress levels indexed
through cortisol readings [16].
 1 study assessed physical wellbeing
[12].
 Regarding care recipient outcomes:
 1 study investigated cortisol levels [16].
 1 study assessed sense of security after
being placed in palliative care [19].
 1 study assessed self-reported
symptoms of dementia [15].
Attachment measures
 All studies included dimensional assessments
of attachment.a
 1 study of adult daughter carers used the
Parental Bonding Instrumentb (PBI;
2 dimensions – care and overprotection) and
Cicirelli’s Adult Attachment Scale (a
unidimensional measure of the strength of an
adult child’s attachment to their older parent)
[14].
 Most studies used the Experiences in Close
Relationship (ECR) Scale and its derivatives
Attachment and aged care: a systematic review Karantzas, Romano and Lee 43
(the ECR-R [10]; and short-forms [28,31]; two
dimensions – attachment anxiety and
attachment avoidance; n = 5 carer studies,
n = 2 dyadic studies [carers and care-
recipients]; n = 1 care-recipient study). The
suite of ECR measures assessed attachment
style in to close others in general.
 Two studies of carers [13,17] used the Adult
Attachment Scale [32,33]; three factors —
close, depend [two facets of attachment
security], and anxiety — a measure of
attachment style towards one’s romantic
partner.
 The two studies that implemented Collin’s
AAS entailed samples that included non-
romantic carers. One study used the
Attachment Style Questionnaire (ASQ [29]; five
factors – confidence [a facet of security],
discomfort with closeness and relationships as
secondary [facets of attachment avoidance],
need for approval and preoccupation with
relationships [facets of attachment anxiety]).
The ASQ is also an assessment of attachment
style to close others in general.
44 Attachment in adulthood
associated with carer satisfaction, and negatively associ-
ated with the severity of care recipients’ psychological
symptoms of dementia as perceived by the carer. Attach-
ment anxiety and was not associated with carers’ percep-
tions of care recipients’ physical and psychological symp-
toms of dementia, however, attachment avoidance was
positively associated (n = 1, [15]). In the one study that
investigated carers’ physical wellbeing, attachment anxi-
ety and avoidance were positively associated with this
outcome [12]. In a dyadic study of stress indexed by
levels of cortisol, care recipient attachment anxiety (but
not care recipient attachment avoidance) was associated
with elevation of this stress hormone in carers [16].
Regarding aspects of caregiving (i.e. caregiving style,
preparedness for caregiving, filial obligation towards car-
ing), carer attachment security was not associated with
caregiving style, but was positively associated with pre-
paredness to care and filial obligation. In addition, attach-
ment anxiety was positively associated with filial obliga-
tion. Attachment anxiety and avoidance were also
positively associated with controlling caregiving, but neg-
atively associated with proximal caregiving (n = 1, [12]).
No association was found between attachment anxiety and
preparedness to provide care (n = 1, [17]) (the association
with attachment avoidance was not measured). In terms of
finding meaning in the caregiving role, attachment avoid-
ance was negatively associated with this outcome, whereas
attachment anxiety was not (n = 1, [18]).
Care recipients
In the three studies that included care recipients, attach-
ment anxiety and avoidance were not associated with
cortisol levels (n = 1, [16]). However, attachment anxiety
and avoidance were positively associated with percep-
tions of one’s own physical symptoms of dementia, while
attachment anxiety was also positively associated with
perceptions of psychological dementia symptoms (n = 1,
[15]). Attachment anxiety and avoidance were also nega-
tively associated with a sense of security once placed into
palliative care (n = 1, [19]).
Discussion
Our qualitative synthesis of in the area of attachment and
aged care provides a contemporary review of the research
over the last five years. Before outlining the key findings
and issues, it important to highlight our qualitative syn-
thesis only comprised of published studies. Therefore, we
cannot remark on the number of unpublished studies that
report non-significant or largely inconsistent results.
Therefore, future reviews in the area could widen the
scope to include published and unpublished studies.
Despite this, our review highlights a number of important
findings and methodological issues that need to be con-
sidered when appraising the ‘state of the field’ and in
undertaking research going forward.
Current Opinion in Psychology 2019, 25:37–46
Firstly, the last five years has continued to see little by
way of research investigating individual differences in
adult attachment (whether it be attachment style, attach-
ment strength, or states of mind) and aged care within
families, especially with regard to the care recipient. One
important reason for this reflects an existing and pervasive
issue in the area, namely, numerous studies continue to
apply inadequate or inappropriate measures of adult
attachment, with several studies continuing to assess
related, but distinct, constructs such as bonding and
affection. Due to these measurement issues, over 85%
of studies published in the last five years were excluded
from this systematic review.
However, the assessment of attachment within the stud-
ies included as part of this synthesis is still not without
issue. The use of Cicirelli’s AAS as a measure of the
strength of the attachment bond between adult child and
older parent is questionable. Higher scores on the AAS
reflect greater strength of attachment and are implied to
reflect attachment security [11,20]. However, various
items on the AAS tap into emotional and physical depen-
dence – characteristics that are incongruent with a strong
and secure attachment bond between parent and child in
later life. Rather, these items reflect characteristics more
akin to clinginess and relationship preoccupation — hall-
marks of attachment anxiety. Therefore, it seems more
appropriate to infer that moderate scores on the AAS reflect
secure familial attachment rather than high scores [4].
The application of widely used self-report measures of
attachment styles within caregiving research is also some-
what problematic. The majority of these measures were
designed to assess romantic attachment (in general or
towards a specific partner). The use of such measures in
studies of spousal caregiving may seem appropriate; how-
ever, in some studies, measures of romantic attachment
are used despite samples (at least in part) encompassing
non-romantic relationships [13,17]. Merely changing the
questionnaire instructions or item wording to reference
familial bonds is unlikely to render some of the measure-
ment issues that can ensue when using existing (espe-
cially romantic) attachment measures. For instance, items
reflecting themes of merging emotionally, or feeling
nervous around a romantic partner or close other, may
not apply particularly well in familial relationships in later
life. Therefore, it seems important that future research
into familial attachment and aged care work towards the
development of appropriate attachment style measures
from the ground up to ensure items reflect context-
specific and age-appropriate assessments of familial
attachment. To this end, the findings of the systematic
review need to be interpreted with some caution.
A particularly important finding is that attachment anxi-
ety is associated with aspects of carer mental health and
psychological wellbeing rather than attachment
www.sciencedirect.com
 Attachment and aged care: a systematic review Karantzas, Romano and Lee 45
avoidance and security. The hyperactivating strategies
that characterise individuals high in attachment anxiety,
such as the chronic intensification of distress and negative
affect, may increase the likelihood that mental health
issues manifest. Moreover, this likelihood of mental
health issues may be further exacerbated as a function
of the stressors and strains that accompany the caregiving
role [4]. In contrast, secure individuals’ constructive
strategies to deal with the challenge of caregiving, and
avoidant individuals’ excessive self-reliance and use of
cognitive and behavioural deactivating strategies, may
inoculate these carers from experiencing mental health
problems [21]. However, for carers high on attachment
avoidance, their strategies may not be sustainable for long
periods. Indeed research has demonstrated that the use of
this strategy may compromise the mental health of these
carers over time [22]. To this end, assessments of indi-
vidual differences in carers’ attachment may be an impor-
tant first line of intervention in identifying carers at risk of
these outcomes (especially those who are anxiously
attached), and to then tailor services and interventions
to enhance their wellbeing [3,4].
In terms of variables pertaining to caregiving specifically,
attachment security and anxiety were positively associ-
ated with filial obligation, while attachment anxiety and
avoidance were positively associated with delivering care
in a manner that was controlling and lacked physical and
emotional closeness. Attachment security was positively
associated with a sense of preparedness around the care-
giving role. The findings may suggest that securely
attached carers are not afraid to think of the future in
which an ageing spouse or older parent may experience
failing health, and despite the challenges this may cause,
they maintain a strong filial responsibility around their
caregiving duties. Insecurely attached individuals appear
to engage in care, but do so in a way that may undermine
the care recipient, with care rendered in a way that is
controlling and lacks emotional closeness.
Of the limited studies that examined carers’ appraisals of
the care recipient’s dementia symptoms, attachment
strength (but not attachment anxiety or security) was
associated with the reporting of fewer psychological
dementia symptoms, while attachment avoidance was
associated with reporting more symptoms. It may be that
individuals with higher attachment strength may perceive
care recipients as having less symptoms in order to mini-
mize distress and the subsequent activation of the attach-
ment system that may ensue if one were to acknowledge
the actual dementia severity of an attachment figure. For
carers high in attachment avoidance, their chronic nega-
tive views of others may result in perceiving care recip-
ients as demonstrating greater signs of dementia.
Care recipients’ self-appraisals of dementia symptoms
and their sense of security were predicted by their own
www.sciencedirect.com
attachment style. Specifically, individuals high on attach-
ment anxiety or avoidance perceived greater physical
symptoms of dementia and a lower sense of security once
placed into palliative care than securely attached individ-
uals [15,19]. Thus, attachment insecurity within care
recipients seems to heighten concerns around one’s vul-
nerability when faced with dementia and the transition to
palliative care. These findings may speak to the labile
nature of insecurely attached individuals’ perceptions of
the self, especially during periods of great strain and
challenge such as chronic health problems and difficult
life transitions [21,23,24]. In particular, insecure individ-
uals perceive themselves to be less competent and more
vulnerable when exposed to threat or insecurity.
Conclusion
The findings of this systematic review suggest that
research into attachment and familial care in later life
continues to be a highly under-researched area. Further-
more, the field needs to revisit the assessment of familial
attachment in later life as the use of existing measures may
not adequately account for the role of individual differ-
ences in attachment within the context of caregiving. It is
also important that a concerted effort is made to conduct
studies that focus on the care recipient. Despite this lack
of research, it appears that carers high on attachment
anxiety are the most likely to experience mental health
issues, yet attachment anxiety and security are associated
with filial duty for carers. Furthermore, attachment inse-
curity (i.e. anxiety and avoidance) is associated with
controlling ways of rendering care to older adults. The
attachment insecurities of care recipients seem to exacer-
bate self-perceptions of vulnerability. The findings point
to the importance of considering attachment theory as an
important framework for understanding familial caregiv-
ing processes in the later stages of life.
Conflict of interest statement
Nothing declared.
Acknowledgements
Funding: This work was supported by the Australian Research Council
[DP160102874].
References and recommended reading
Papers of particular interest, published within the period of review,
have been highlighted as:
 of special interest
 of outstanding interest
1. Pinquart M, Sörensen S: Associations of caregiver stressors
and uplifts with subjective well-being and depressive mood: a
meta-analytic comparison. Aging Ment Health 2004, 8:438-449
http://dx.doi.org/10.1080/13607860410001725036.
2. Pinquart M, Sörensen S: Correlates of physical health of
informal caregivers: a meta-analysis. J Gerontol B 2007,
62B:126-137 http://dx.doi.org/10.1093/geronb/62.2.P126.
3. Karantzas GC, Evans L, Foddy M: The role of attachment in
current and future parent caregiving. J Gerontol B 2010,
65B:573-580 http://dx.doi.org/10.1093/geronb/gbq047.
Current Opinion in Psychology 2019, 25:37–46
46 Attachment in adulthood
4. Karantzas GC, Simpson JA: Attachment and aged care. In
 Attachment Theory and Research: New Directions and Emerging
Themes. Edited by Simpson JA, Rholes WS. Guilford Press;
2015:319-345.
One of the most comprehensive theoretical and empirical reviews of adult
attachment within the context of aged care and families in later life.
5. Bowlby J: Attachment and Loss: Vol. 1. Attachment. 2nd ed.. Basic
Books; 1982.
6. Troll E, Smith J: Attachment through the life-span: some
questions about dyadic bonds among adults. Hum Dev 1976,
19:156-170 http://dx.doi.org/10.1159/000271526.
7. Bradley JM, Cafferty TP: Attachment among older adults:
current issues and directions for future research. Attach Hum
Dev 2001, 3:200-221 http://dx.doi.org/10.1080/14616730126485.
8. Wilkinson RB: Measuring attachment dimensions in
adolescents: development and validation of the Experiences
in Close Relationships-Revised-General Short Form. J Soc
Relatsh Res 2011, 2:53-62 http://dx.doi.org/10.1375/jrr.2.1.53.
9. Brennan KA, Clark CL, Shaver PR: Self-report measurement of
adult attachment: an integrative overview. In Attachment
Theory and Close Relationships. Edited by Simpson JA, Rholes
WS. Guilford Press; 1998:46-76.
10. Fraley RC, Waller NG, Brennan KA: An item response theory
analysis of self-report measures of adult attachment. J Pers
Soc Psychol 2000, 78:350-365 http://dx.doi.org/10.1037//0022-
3514.78.2.350.
11. Cicirelli VG: Attachment and obligation as daughters’ motives
for caregiving behavior and subsequent effect on subjective
burden. Psychol Aging 1993, 8:144-155 http://dx.doi.org/
10.1037//0882-7974.8.2.144.
12. Ávila M, Brandão T, Coimbra JL, Lopez F, Matos PM:
 Experiencing an intimate partner’s breast cancer: attachment,
caregiving, and burden in men. Psychiatry 2016, 79:236-248
http://dx.doi.org/10.1080/00332747.2016.1158003.
This study collected data from men who were caring for their wives who
had breast cancer. The results showed that carers with more insecure
attachment styles were more likely to provide care that was more con-
trolling or intrusive, and less sensitive, and that proximal caregiving
mediated the attachment style to carer burden association.
13. Burke LA, Clark KA, Ali KS, Gibson BW, Smigelsky MA,
Neimeyer RA: Risk factors for anticipatory grief in family
members of terminally ill veterans receiving palliative care
services. J Soc Work End Life Palliat Care 2015, 11:244-266 http://
dx.doi.org/10.1080/15524256.2015.1110071.
14. Chen CK, Uzdawinis D, Schölmerich A, Juckel G: Effects of
attachment quality on caregiving of a parent with dementia.
Am J Geriatr Psychiatry 2014, 22:623-631 http://dx.doi.org/
10.1016/j.jagp.2012.12.003.
15. Monin JK, Schulz R, Kershaw TS: Caregiving spouses’
attachment orientations and the physical and psychological
health of individuals with Alzheimer’s disease. Aging Ment
Health 2013, 17:508-516 http://dx.doi.org/10.1080/
13607863.2012.747080.
16. Hsiao F, Jow G, Kuo W, Huang C, Lai Y, Liu Y, Chang K: The
partner’s insecure attachment, depression and psychological
well-being as predictors of diurnal cortisol patterns for breast
cancer survivors and their spouses. Stress 2014, 17:169-175
http://dx.doi.org/10.3109/10253890.2014.880833.
17. Paulson D, Bassett R: Prepared to care: adult attachment and
 filial obligation. Aging Ment Health 2016, 20:1221-1228 http://dx.
doi.org/10.1080/13607863.2015.1072800.
This study provides one of the only investigations into the role of attach-
ment theory in the preparation activities of prospective carers of older
parents. This study found that daughters with a secure attachment style
are more likely to engage in preparation activities in regard to future
caregiving for their parents. Moreover, this association was fully mediated
by the daughter’s sense of filial obligation.
18. Hasson-Ohayon I, Goldzweig G, Sela-Oren T, Pizem N, Bar-
Sela G, Wolf I: Attachment style, social support and finding
meaning among spouses of colorectal cancer patients:
Current Opinion in Psychology 2019, 25:37–46
gender differences. Palliat Support Care 2015, 13:527-535 http://
dx.doi.org/10.1017/S1478951513000242.
19. Milberg A, Friedrichsen M, Jakobsson M, Nilsson E, Niskala B,
Olsson M, Wahlberg R, Krevers B: Patients’ sense of security
during palliative care-What are the influencing factors? J Pain
Symptom Manage 2014, 48:45-55 http://dx.doi.org/10.1016/j.
jpainsymman.2013.08.021.
20. Thompson L, Walker AJ: Mothers and daughters: aid patterns
and attachment. J Marriage Fam 1984, 46:313-322 http://dx.doi.
org/10.2307/352463.
21. Gillath O, Karantzas GC, Fraley RC (Eds): Adult Attachment: A
 Concise Introduction to Theory and Research. Academic Press;
2016.
An important introductory text regarding theory and research into adult
attachment. The text provides a comprehensive discussion of many of the
central questions in the study of adult attachment.
22. Berant E, Mikulincer M, Shaver PR: Mothers’ attachment style,
their mental health, and their children’s emotional
vulnerabilities: a 7-year study of children with congenital heart
disease. J Pers 2008, 76:31-66 http://dx.doi.org/10.1111/j.1467-
6494.2007.00479.x.
23. Klohnen EC, Weller JA, Luo S, Choe M: Organization and
predictive power of general and relationship-specific
attachment models: one for all, and all for one? Pers Soc
Psychol Bull 2005, 31:1665-1682 http://dx.doi.org/10.1177/
0146167205278307.
24. Strodl E, Noller P: The relationship of adult attachment
dimensions to depression and agoraphobia. Pers Relatsh 2003,
10:171-186 http://dx.doi.org/10.1111/1475-6811.00044.
25. Fasse L, Flahault C, Brédart A, Dolbeault S, Sultan S: Describing
and understanding depression in spouses of cancer patients
in palliative phase. Psychooncology 2015, 24:1131-1137 http://
dx.doi.org/10.1002/pon.3777.
26. Lai C, Luciani M, Galli F, Morelli E, Cappelluti R, Penco I, Aceto P,
Lombardo L: Attachment style dimensions can affect
prolonged grief risk in caregivers of terminally ill patients with
cancer. Am J Hosp Palliat Med 2015, 32:855-860 http://dx.doi.
org/10.1177/1049909114547945.
27. Lee J, Sohn BK, Lee H, Seong SJ, Park S, Lee JY: Attachment
 style and filial obligation in the burden of caregivers of
dementia patients. Arch Gerontol Geriatr 2018, 75:104-111
http://dx.doi.org/10.1016/j.archger.2017.12.002.
Using a Korean filial obligation measure developed for the study, the
authors found that although greater attachment anxiety predicted higher
burden, the association of attachment avoidance with burden was con-
tingent on filial obligation.
28. Wei M, Russell DW, Mallinckrodt B, Vogel DL: The Experiences in
Close Relationship Scale (ECR)-short form: reliability, validity,
and factor structure. J Pers Assess 2007, 88:187-204 http://dx.
doi.org/10.1080/00223890701268041.
29. Fossati A, Feeney JA, Donati D, Donini M, Novella L, Bagnato M,
Acquarini E, Maffei C: On the dimensionality of the Attachment
Style Questionnaire in Italian clinical and nonclinical
participants. J Soc Pers Relatsh 2003, 20:55-79 http://dx.doi.org/
10.1177/02654075030201003.
30. Kim S-H, Sherry AR, Lee Y-S, Kim C-D: Psychometric properties
of a translated Korean adult attachment measure. Mea Eval
Couns Dev 2011, 44:135-150 http://dx.doi.org/10.1177/
0748175611409842.
31. Lo C, Walsh A, Mikulincer M, Gagliese L, Zimmermann C, Rodin G:
Measuring attachment security in patients with advanced
cancer: psychometric properties of a modified and brief
experiences in close relationships scale. Psychooncology
2009, 18:490-499 http://dx.doi.org/10.1002/pon.1417.
32. Collins NL: Working models of attachment: implications for
explanation, emotion, and behavior. Pers Soc Psychol 1996,
71:810-832 http://dx.doi.org/10.1037//0022-3514.71.4.810.
33. Collins NL, Read SJ: Adult attachment, working models, and
relationship quality in dating couples. J Pers Soc Psychol 1990,
58:644-663 http://dx.doi.org/10.1037/0022-3514.58.4.644.
www.sciencedirect.com