Negative life events and depressive symptoms in late
life: Buffering effects of parental and partner
bonding?

VIVIAN KRAAIJ

AND

NADIA GARNEFSKI

Leiden University, The Netherlands

Abstract

The buffering function of parental and partner bonding in the relationship between negative life events and depressive
symptoms at old age was examined. A community sample of 194 people aged 65 years and older was interviewed.
In particular, the control dimension of both parental and partner bonding seems to be of importance in relation to
elderly depression: More psychological control is related to higher depression scores, and low partner control seems
to work as a protective mechanism when negative life events are faced. Developing prevention and intervention
programs aimed at optimizing bonding relationships throughout life seems advisable.

Negative life events have consistently been
found to be related to depression in late life
(Katona, 1993; Kraaij, 2000; Orrell & Davies,
1994). Elderly people who have experienced
events such as death of loved ones, phy-
sical illnesses, abuse, or relational problems
throughout their lives have been found to suffer
from more depressive symptoms. Especially
the accumulation of stress has been found to be
strongly related to depression in the elderly
(Kraaij, Arensman, & Spinhoven, in press;
Kraaij & de Wilde, 2001). Depression in late
life can have a devastating impact; it has been
found to be the leading psychological factor in
suicide among the elderly (Draper, 1994;
Lapierre, Pronovost, Dube, & Delisle, 1992;
McIntosh, Santos, Hubbard, & Overholser,
1994). Therefore, it is important to examine
factors that make people more resilient in the
face of stress. Bonding or attachment has been
considered to be a life span process, and its

buffering role between negative life events and
depressive symptoms was examined in the
present study.

It has been widely hypothesized that secure

attachment is associated with emotional stabil-
ity throughout life, and insecure attachment
may predispose a person to later episodes of
depression, anxiety, and loneliness (Bowlby,
1969, 1973; Parker, 1994). Attachments are
formed at all ages and in different relationships.
During childhood the primary attachment
figures are the parents. Parker distinguished
two key dimensions that underlie the bond
between a parent and child: care and control
(Parker, 1989; Parker, Tupling, & Brown,
1979). The care dimension involves affection
and warmth versus coldness and rejection. The
control dimension involves psychological
autonomy versus psychological control. Dur-
ing adult life the most common attachment
figure is the marital partner or cohabitant.
Levels of affection or control in the adult
intimate relationship are not necessarily dic-
tated by levels of early parental care (Parker,
Barrett, & Hickie, 1992). In line with parental
bonding, Wilhelm and Parker (1988) identified
and defined two key dimensions underlying

205

Correspondence concerning this article should be
addressed to Vivian Kraaij, Division of Clinical and
Health Psychology, Leiden University, P.O. Box 9555,
2300 RB Leiden, The Netherlands; e-mail: Kraaij@
fsw.leidenuniv.nl.

Personal Relationships, 9 (2002), 205–214. Printed in the United States of America.
Copyright

#

2002 ISSPR. 1350-4126/02

partner bonding: care and control. The care
dimension reflects care expressed emotionally
as well as physically. The control dimension
suggests domination, intrusiveness, criticism,
and authoritarian attitudes and behaviors.

In addition to the direct effect that attach-

ment can have on a person’s well-being, a
buffering effect of attachment has also been
suggested. A life-cycle perspective on the
provision of security, specifically during times
of stress, has been suggested by several
authors (Antonucci, 1994; Aro, 1994; Barnas,
Pollina, & Cummings, 1991; West, Livesley,
Reiffer, & Sheldon, 1986). Secure early
attachments through childhood and good
intimate relationships in adulthood make it
more likely that a person will develop feelings
of high self-esteem and self-efficacy. It could
be argued that this might produce a protective
effect against emotional difficulties when
stressful events are faced (Bandura, 1997;
Benight et al., 1999; Rector & Roger, 1997).

The direct effect of bonding on well-being

has been studied extensively. The majority of
studies on perceived parental rearing practices
confirm that anxious and depressed people
recall their parents as having been less affec-
tionate and more controlling and protecting
than do healthy controls (Gerlsma, Emmel-
kamp, & Arrindell, 1990; Parker, 1994). A
parental style of low care and overprotection
has especially been found to be related to
neurotic disorders (Parker, 1989). The impor-
tance of lack of intimacy in adult relationships
as a risk factor of neurotic disorders has also
been suggested in a number of studies (Hickie,
Parker, Wilhelm, & Tennant, 1991; Wilhelm &
Parker, 1988).

Only a few studies have been published on

the direct relationship of parental bonding and
partner bonding with well-being in old age
(Andersson & Stevens, 1993; Murphy, 1982;
Stevens & Andersson, 1996). They have shown
that warm and attentive parental care is related
to lower anxiety scores, and that overprotection
by parents early in life is related to higher
anxiety scores in old age for men. Having an
affectionate partner appears to be related to
lower anxiety and depression scores.

To our knowledge, even fewer studies have

been performed on the buffering effect that

bonding has on the response to stressful events.
Having an insecure adult attachment style has
been suggested as increasing the chance of
developing difficulties in the presence of
stressful events in adulthood (Carpenter &
Kirkpatrick, 1996). However, a larger body of
literature exists on social support and its
relationship to well-being in adulthood and
old age. Review studies have shown that
intimate confiding relationships might have a
health protective effect when stressful events
are experienced (Bowling, 1994; Pierce, Sar-
ason, & Sarason, 1996; Thoits, 1995).

Clearly, research on the buffering role of

bonding in the setting of stressful life events is
scarce, especially for elderly people, despite the
importance such information could have for
factors of risk and resiliency in elderly depres-
sion. Such information could be used to better
understand depressive symptoms in old age and
could be helpful in the development of preven-
tion and intervention programs. The aim of this
study was therefore to examine the buffering
function of both parental bonding and partner
bonding in the relationship between negative
life events and depressive symptoms at old age.
As far as we know, the present study is the first
study that takes into account the whole life span
of an individual by not only studying parental
bonding or partner bonding separately, but also
by studying parental bonding and partner
bonding simultaneously.

Method

Sample and procedure

The sample consisted of 194 elderly people,
ranging in age from 65 to 94 years (M = 76.5
years, SD = 7.25), and 52% were female.
Almost all respondents (96%) were born in
The Netherlands and all currently held Dutch
nationality. Forty-four percent were single
households, 90% were living independently,
and 76% reported having religious beliefs
(28.4% Catholic, 41.2% Protestant, and 6.2%
another religion).

Subjects were randomly selected from the

community register of the city of Leiden (The
Netherlands). They were informed by letter and
invited by telephone call to participate in the

V. Kraaij and N. Garnefski

206

study. A total of 920 letters were sent. In
response, 487 people chose not to participate
(e.g., the respondents said they received too
many requests to participate in various studies
or they thought research was useless) and
another 239 people could not be reached (e.g.,
the telephone number was incorrect or the
telephone had been disconnected) or could not
participate for practical reasons (e.g., the
respondent was deaf or did not speak the Dutch
language or appeared to be confused). A total of
194 elderly people were interviewed at their
homes after providing informed consent. The
response rate was therefore 28%. All instru-
ments were administered orally.

Measures

Depressive symptoms were measured by the
Geriatric Depression Scale (GDS; Brink,
Yesavage, Heersema, Adey, & Rose, 1982),
consisting of 30 dichotomous questions (e.g.,
Are you basically satisfied with your life?; Do
you often feel downhearted and blue?). Scores
range from 0 to 30, with a high score
indicating more depressive symptoms. The
GDS excludes items that are confounded with
normal aging and the diseases associated with
old age, but assesses primarily psychological
components of depression. Therefore, it is
very suitable for assessing depression in the
elderly. The GDS has been demonstrated as
having a high reliability (Cronbach’s alpha
coefficient: 0.94), good validity, and high
levels of sensitivity and specificity (Kok,
1994; Olin, Schneider, Eaton, Zemansky, &
Pollock, 1992; Yesavage et al., 1983). In the
present sample a Cronbach’s alpha reliability
coefficient of .83 was found.

Parental bonding was measured by the

Parental Bonding Instrument (PBI; Parker,
1983; Parker et al., 1979 ). The PBI is a
25-item self-report questionnaire on recalled
parental rearing practices, using a 4-point
Likert scale (very like to very unlike). The
questionnaire yields scores for the two
principal dimensions of parental behaviors
and attitudes: care and control. The care
dimension consists of 12 items (e.g., ‘‘could
make me feel better when I was upset,’’
‘‘appeared to understand my problems and

worries’’) and scores range from 0 to 36, with
a high score indicating high care. The control
dimension consists of 13 items (e.g., ‘‘tried to
control everything I did,’’ ‘‘was overprotec-
tive of me’’) and scores range from 0 to 39,
with a high score indicating high control
(overprotection). The two dimensions can be
combined to measure four bonding styles:
optimal parenting (high care/low control),
affectionate constraint (high care/high con-
trol), affectionless control (low care/high
control), and neglectful parenting (low care/
low control). In the present study the
combined effect of the care and control
dimension was examined by creating an
interaction term between the two dimensions.

The PBI questions are answered for mother

and father separately. The obtained maternal
and paternal scores can be used separately. In
the present study, both the maternal and
paternal care scales and the maternal and
paternal control scales were highly correlated
(care scales: r = .747, p < .001; control scales:
r = .956, p < .001). Therefore, the maternal and
paternal scales were combined into parental
scales by taking the means of the two scales. If a
respondent reported growing up with one
caregiver, the score of this parent was used.

The PBI has been widely used and has been

found to have good psychometric properties
(Gerlsma, 1994; Parker, 1983, 1989, 1990). In
the present sample the following Cronbach’s
alpha reliability coefficients were found: .87
for maternal care; .84 for paternal care; .77 for
maternal control; and .77 for paternal control.

Partner bonding was measured by the

24-item Intimate Bond Measure (IBM; Wil-
helm & Parker, 1988). The IBM, which is
similar in development to the PBI, is a self-
report measure of the care and control exerted
by the intimate partner. The care dimension
and the control dimension each consist of
12 items (using a 4-point Likert scale ranging
from very true to not at all), and scores range
from 0 to 36, with a high score indicating
respectively high care and high control.
Examples of items from the care scale are
‘‘understands my problems and worries,’’ ‘‘is
gentle and kind to me,’’ ‘‘is very considerate of
me.’’ Examples of items from the control scale
are ‘‘wants to know exactly what I’m doing

Negative life events and bonding

207

and where I am,’’ ‘‘tends to try and change
me,’’ ‘‘seeks to dominate me.’’ Similar to the
PBI, the two scales can be combined to
measure four styles of intimate relationships:
optimal intimacy (high care/low control),
affectionate constraint (high care/high control),
affectionless control (low care/high control),
and absence of intimacy (low care/low con-
trol). In the present study the combined effect
of the care and control dimension was
examined by creating an interaction term
between the two dimensions. The subjects
were asked to answer the questions for their
(current or past) partner. (Respondents who
had had more than one partner answered the
questions for the partner with whom they had
been together the longest.) The IBM has been
found to have good psychometric properties. In
the present sample Cronbach’s alpha reliability
coefficients of .94 for partner care and .79 for
partner control were found. In the present
study, there was no significant relationship
between the presence or absence of the partner
referred to in the IBM and the scores on the care
and control dimensions. Moreover, there ap-
peared to be no significant interaction between
presence or absence of the partner and the care
and control dimensions on depressive symp-
toms. Therefore, the IBM could be used for the
whole group in the same manner.

Negative life events experienced before the

interview were measured by the Negative Life
Events Questionnaire. This questionnaire is an
adaptation of the Life Events Questionnaire,
an instrument used in the WHO multicenter
study on parasuicide (Kerkhof, Schmidtke,
Bille-Brahe, de Leo, & Lo¨nnqvist, 1994).
Among others, the adapted version is extended
with negative life events specific to elderly
people (e.g., dementia of a partner and
wartime-related events). The Negative Life
Events Questionnaire contains 107 dichoto-
mous items on negative life events concerning
self or significant others, such as death of
significant others, sexual abuse, and imprison-
ment during the war. The questionnaire is a
lifetime instrument questioning the occurrence
of all events for different developmental
periods, ranging from childhood to the year
prior to the interview. Questions are formu-
lated in a detailed way to minimize judgmental

and subjective estimates (e.g., with regard to
physical abuse one of the questions is ‘‘were
you ever severely beaten, kicked, or deliber-
ately wounded by [one of] your parents’’ rather
than ‘‘were you ever physically abused?’’). In
the present study a sum score was used
by adding all negative events experienced
throughout life, up to the initial interview.

Data analysis

Before studying the buffering effect of parental
and partner bonding, we examined whether
there was an interaction effect between the care
and control dimensions of, respectively, par-
ental bonding and partner bonding. This was
studied by performing moderated regression
analyses (entering all variables simultaneously)
for parental bonding and partner bonding
separately. Depressive symptoms were re-
gressed on care, control, and its interaction
term. In the case of a nonsignificant interaction
effect, it is sufficient to use the two dimensions.
However, in the case of a significant interaction
effect, the interaction term should be included
in the following analysis.

To study the moderating impact of parental

bonding and partner bonding on the relation-
ship between negative life events and depres-
sive symptoms, we used moderated regression
analysis. Depressive symptoms were regressed
on parental and partner bonding, negative life
events, and the interaction terms of the
bonding dimensions with negative life events.
All variables were entered simultaneously.

To examine the meaning of the significant

interaction term in more detail, the group was
divided into four equal-sized subgroups, based
on the quartiles of the moderating variable.
For each group the Pearson correlation
between negative life events and depressive
symptoms was computed.

Results

Preliminary analyses

Prior to performing the main analyses, de-
scriptive statistics and Pearson correlations
among the variables were computed (Table 1).
In the present sample, the elderly people

V. Kraaij and N. Garnefski

208

recalled on average both their parents and
partner as exerting more care than control. The
respondents had a mean score of 6 depressive
symptoms and had experienced on average 19
negative life events. There was no evidence of
multicollinearity (see Tabachnick & Fidell,
1996) among the measures of parental bond-
ing, partner bonding, and negative life events,
which means that they could be used in the
same analysis. As expected, however, there
were a number of significant correlations.

Interaction between the care and control
dimensions

The interaction effect between the care and
control dimensions of parental bonding and
partner bonding was studied by moderated
regression analyses (Table 2). With regard to
parental bonding, only parental control had a
significant (positive) relationship with depres-
sive symptoms. There appeared to be no
interaction effect between the parental care
and control dimension. With regard to partner
bonding, again only the control dimension had
a significant (positive) relationship with de-
pressive symptoms. There appeared to be no
interaction effect between the partner care and
control dimension. As there were no significant
interaction effects between the care and control
dimensions, these interaction terms were not
included in the following analyses.

Relationship of care, control, and negative life
events with depressive symptoms

To study the moderating impact of parental
bonding and partner bonding on the relation-

ship between negative life events and depres-
sive symptoms, we performed moderated
regression analysis (see Table 3). The findings
showed that the strongest predictor of depres-
sive symptoms was the number of negative
life events experienced. Further, both parental
and partner control had a direct significant
relationship with depressive symptoms. If a
respondent reported more parental control and
more partner control, more depressive symp-
toms were reported. A significant moderating
effect of partner control was also found. In
total, 19.5% of the variance was explained
(F[9, 158] = 4.25; p < .001).

1

Examination of buffering function

To examine the meaning of the significant
interaction term into more detail, the group was
divided into four equal-sized subgroups based
on the quartiles of partner control. This resulted
in the following groups: (1) those with the
lowest scores on partner control (first quartile,
n = 40), (2) those with a low to median score on
partner control (second quartile, n = 44), (3)
those with a median to high score on partner
control (third quartile, n = 47), and (4) those
with the highest scores on partner control
(fourth quartile, n = 44). For each group the
Pearson correlation between negative life
events and depressive symptoms was com-
puted. The groups that reported having the

Table 1. Descriptive statistics and Pearson correlations between all variables

Variables

M

(SD)

1

2

3

4

5

1 Depressive symptoms

5.89 (4.86)

—

2 Parental care

26.59 (7.13)

.206**

—

3 Parental control

12.26 (6.81)

.239**

.439***

—

4 Partner care

29.37 (7.95)

.115

.184*

.080

—

5 Partner control

8.12 (6.25)

.237**

.058

.007

.457***

—

6 No. of negative life events 19.11 (8.33)

.405***

.287***

.128

.350*** .310***

Note. Due to missing values, n ranged from 168 to 188.
*p < .05. **p < .01. ***p < .001.

1. When gender and age were included in the analysis,

they did not have a significant relationship with
depressive symptoms.

Negative life events and bonding

209

highest partner control (groups 3 and 4) had the
strongest significant (positive) relationship
between the number of negative life events
and depressive symptoms (respectively,
r = .333; p < .05, and r = .540; p < .001),
with group 4 showing the highest correlation.
The groups who reported having the lowest
partner control (groups 1 and 2) had no
(significant) relationship between the number
of negative life events and depressive symp-
toms (respectively, r = .031 and r = .248), with
group 1 showing the lowest correlation. These
findings showed a much stronger relationship
between the number of negative life events and
depressive symptoms for those who reported
having a partner characterized by high control
than for those who did not.

Discussion

A major finding of the present study is that in
particular the control dimension of both
parental bonding and partner bonding seems
to be of importance in relation to elderly
depression. More psychological control by
parents and partner appears to be a risk factor
for depressive symptoms. Further, partner
control may function as a buffer between
the experience of negative life events and
dep-ressive symptoms, in that low partner
control seems to work as a protective
mechanism when many negative life events
are faced, whereas high partner control seems

to work as a vulnerability mechanism when
many stressful events are experienced. These
findings are in line with the results of other
studies (Andersson & Stevens, 1993; Brown
& Harris, 1978; Carpenter & Kirkpatrick,
1996; Gerlsma et al., 1990; Murphy, 1982;
Parker, 1994; Stevens & Andersson, 1996).
Control suggests domination, intrusiveness,
criticism, and authoritarian attitudes and behav-
ior. These attitudes and behaviors can co-occur
with marital dissatisfaction, but are not neces-
sarily the same (Rook & Ituarte, 1999). In
general, it has been shown that marital dis-
satisfaction or not getting along with one’s
spouse is related to psychological disorders
(Whisman, 1999; Whisman & Bruce, 1999;
Whisman, Sheldon, & Goering, 2000). This
study adds that especially controlling attitudes
and behaviors of the partner seem to be related
to psychological well-being. Future studies
should examine whether partner control and
marital dissatisfaction are related. Finally, the
relative impact of parental and partner care on
depression was tested in a regression analysis
including control and negative life events.
Parental care and partner care appeared to
have no significant unique contribution to
depression.

Some methodological considerations of the

present study do need to be taken into account.
The first concerns the representativeness of the

Table 3. Moderated regression analysis of
parental bonding, partner bonding, and
negative life events on depressive symptoms

Independent variables

Beta

Parental care

.027

Parental control

.181*

Partner care

.164

Partner control

.178*

No. of negative life events

.285**

Life events

parental care

.058

Life events

parental control

.007

Life events

partner care

.015

Life events

partner control

.183*

Note. Variables were entered simultaneously. R

2

= .195,

F(9,158) = 4.25, p < .001.
*p < .05. **p <.01. ***p < .001.

Table 2. Examination of interaction effect
between care and control dimensions on
depressive symptoms by moderated regression
analyses

Independent
variables

Parental
bonding

(Beta)

a

Partner

bonding

(Beta)

b

Care

.131

.067

Control

1.77*

.223**

Care

control

.020

.142

Note. Variables were entered simultaneously.
*p < .05, **p < .01.

a

R

2

= .072, F(3,181) = 4.72, p < .01.

b

R

2

= .070, F(3,171) = 4.28, p < .01.

V. Kraaij and N. Garnefski

210

group that was studied. The response rate was
moderately low, which makes it possible that
selection occurred. One could hypothesize that
people with depressive symptoms were more
inclined to nonparticipation in order to avoid
stimuli reminding them of negative life events.
Actually, the invitation letter stated that people
who were in treatment for emotional problems
should discuss participation with their treat-
ment provider (this in accordance with rules of
the Ethical Committee). There is, in fact, no
indication that people who were depressed
were less willing to participate: 16% of the
present sample scored above the cutoff score
for depression of the GDS, which is in
accordance with rates identified in other
studies (Beekman et al., 1997; Cappeliez,
1988; Gurland & Cross, 1982; van Marwijk,
1995; O’Hara, Kohout, & Wallace, 1985;
Ruegg, Zisook, & Swerdlow, 1988). Further-
more, the sample had a gender distribution
comparable to the Dutch elderly population,
and the subjects had the same assessment of
their general health as the Dutch elderly
population (CBS, 1997). Finally, the majority
of the participants were religious, which is
representative for the Dutch elderly age-group
(Braam, Beekman, Deeg, & van Tilburg, 1994).

A second point of concern is that because

the data were gathered retrospectively we
cannot make causal inferences about the
relationships between the variables. Previous
research has shown the bidirectional interplay
of marital discord and depressive symptoms
(Fincham, Beach, Harold, & Osborne, 1997).
For the present study this means, for example,
that partner control may not only be a possible
cause of depressive symptoms but also a
consequence of depressive symptoms. De-
pressed people may suffer from sleeplessness
and irritability, which may take its toll on the
spouse, potentially resulting in more criticism
and authoritarian behaviors of the partner. In
order to solve these cause and effect issues,
longitudinal research is needed. Further, retro-
spective designs have been criticized because
they can result in recall errors (Henry, Moffitt,
Caspi, Langley, & Silva, 1994; Norris &
Kaniasty, 1992; Yarrow, Campbell, & Burton,
1970). With regard to the recall of life events, it
has been suggested that depressed persons, in

order to account for their current emotional
state, may report more negative life events than
nondepressed persons (Brown, 1972; Teasdale,
1983). There has also been research suggesting
that the central characteristics of events are
generally remembered well, but the temporal
details of events, such as frequencies and dates,
are less likely to be reliable (Brewin, Andrews,
& Gotlib, 1993). To minimize the possible
effects of recall bias, a highly structured
interview schedule was used and the questions
were limited to whether or not the respondent
had experienced the negative life event without
asking for details. With regard to the recall of
parental bonding and partner bonding, it has
been often suggested that depressed mood
influences the perception or recall of these
phenomena. However, numerous studies have
failed to find evidence of such a response bias
(Gerlsma, Das, & Emmelkamp, 1993; Gerls-
ma, Kramer, Scholing, & Emmelkamp, 1994;
Parker, 1981, 1990; Wilhelm & Parker, 1988).
Future studies should try to limit these recall
biases, for example by using methods to cue
memories, or by using secondary instruments
to assess the reliability of the self-reports.

Finally, although the PBI and the IBM were

designed to measure perceived parental and
partner characteristics, it is often questioned
whether they are also measures of actual
parental and partner characteristics. Various
studies have offered support for the PBI and
IBM as valid measures of actual characteristics
(Parker, 1990, 1994; Wilhelm & Parker,
1988). Whether these constructs reflect actual
relationships versus the personality of respon-
dents remains to be studied.

The present study has some important

strengths. First, well-designed and validated
instruments were used. Further, as far as we
know, it is the first study examining the
buffering role of bonding in the presence of
negative life events and at the same time
taking into account the whole life span of an
individual by studying parental bonding and
partner bonding simultaneously. It is the first
study suggesting (a) that low partner control
seems to protect an individual against emo-
tional difficulties when faced with stressful
events, and (b) that depression in old age
seems to be related to lifelong bonding

Negative life events and bonding

211

processes. Optimal bonding, defined espe-
cially by low control, with attachment figures
throughout life, seems to be associated with
emotional stability at old age.

As psychological control seems to be the

most important dimension of relationships
throughout life with regard to well-being, it
could be argued that prevention and interven-
tion programs should pay attention to this
aspect as early in life as possible. In infant
welfare centers and child health centers, health
care workers could focus on the establishment
of optimal bonding between child and parent
by educating parents or by skills training
programs.

In adult life, if a patient has a current

intimate relationship, relational therapy could
focus on the bonding aspects of the relation-
ship. In particular, control aspects in the
relationship, such as domination, intrusive-
ness, criticism, and authoritarian attitudes and
behaviors of a partner, could be the focus of
counseling. Social skills training programs for
couples could also be developed, focusing on
the development of attitudes and behaviors in
favor of psychological autonomy.

If an intimate partner is unavailable or does

not provide the needed support, developing
other relationships that are characterized by
psychological autonomy could possibly help
elderly depressed people to regain emotional
stability. As stated above, optimal bonds can
be formed with caretakers and intimate
partners, but secure attachment relationships
can also be formed with (adult) children, other
family members, and friends (Ainsworth,
1989, 1991). Future research should be
performed to study the influence of these
other secure attachment relationships. Provid-
ing elderly people with social activities, such
as going on day trips or participating in art
clubs or game clubs, with people from this
social network could aid in the formation of
optimal bonding relationships.

Giving attention in intervention and pre-

vention programs to the formation of optimal
bonds in relationships could not only make
people more resilient against emotional pro-
blems in general, but could also protect a
person against emotional difficulties in the
presence of negative life events.

To further unravel the relationship between

negative life events and well-being at old age,
other outcome measures such as anxiety and
health should also be studied. The possible
buffering role of other aspects, such as
coping, personality characteristics, and other
sources of social support, should also be the
focus of future research. As it remains
unclear to what extent the findings of this
study can be generalized to other populations,
studies should also be performed with clinical
samples and other age groups. Finally, the
buffering role of bonding when specific
negative life events are experienced (i.e.,
death of spouse or natural disaster) should
also be studied. Previous studies found a
similar relationship between various negative
life events and depression in old age (Kraaij
et al., in press). Whether bonding has a
similar or different buffering role with
different kinds of negative life events remains
to be studied.

As elderly people face multiple losses,

such as death of friends, loss of functional
abilities, and loss of autonomy with the move
to an institutional home, they are at risk of
developing emotional problems. Considering
the strong increase in the aging population,
the pressure placed on the health care system
by the elderly will increase. The present
study shows the importance of optimal
bonding relationships throughout life to make
elderly people more resilient in the presence
of life stress. Developing prevention and
intervention programs aimed at developing
optimal bonding relationships seems of para-
mount importance.

References

Ainsworth, M. D. S. (1989). Attachments beyond infancy.

American Psychologist, 44, 709–716.

Ainsworth, M. D. S. (1991). Attachments and other

affectional bonds across the life cycle. In C. M.

Parkes, J. Stevenson-Hinde, & P. Marris (Eds.),
Attachment across the life cycle (pp. 33-51). London:
Routledge.

Andersson, L., & Stevens, N. (1993). Associations

V. Kraaij and N. Garnefski

212

between early experiences with parents and well-being
in old age. Journal of Gerontology, 48, 109–116.

Antonucci, T. C. (1994). Attachment in adulthood and

aging. In M. B. Sperling & W. H. Berman (Eds.),
Attachment in adults: Clinical and developmental
perspectives (pp. 256–272). New York: Guilford.

Aro, H. (1994). Risk and protective factors in depression:

A developmental perspective. Acta Psychiatrica Scan-
dinavica, 89, 59–64.

Bandura, A. (1997). Self-efficacy: The exercise of control.

New York: W.H. Freeman.

Barnas, M. V., Pollina, L., & Cummings, E. M. (1991).

Life-span attachment: Relations between attachment
and socioemotional functioning in adult women.
Genetic, Social, and General Psychology Monographs,
117, 175–202.

Beekman, A. T. F., Deeg, D. J. H. Tilburg, T. G. van,

Schoevers, R. A., Smit, J. H. Hooijer, C., & van
Tilburg, W. (1997). Depressie bij ouderen in de
Nederlandse bevolking: Een onderzoek naar de pre-
valentie en risicofactoren. [Depression in the elderly in
the Dutch community: A study of prevalence and risk
factors]. Tijdschrift voor Psychiatrie, 39, 294–307.

Benight, C. C., Ironson, G. Klebe, K., Carver, C. S.,

Wynings, C., Burnett, K., Greenwood, D., Baum, A.,
& Schneiderman, N. (1999). Conservation of resources
and coping self-efficacy predicting distress following a
natural disaster: A causal model analysis where the
environment meets the mind. Anxiety, Stress, and
Coping, 12, 107–126.

Bowlby, J. (1969). Attachment and loss. Volume 1.

Attachment. New York: Basic Books.

Bowlby, J. (1973). Attachment and loss. Volume 2.

Separation: Anxiety and anger. New York: Basic
Books.

Bowling, A. (1994). Social networks and social support

among older people and implications for emotional
well-being and psychiatric morbidity. International
Review of Psychiatry, 6, 41–58.

Braam, A. W., Beekman, A. T. F., Deeg, D. J. H., & van

Tilburg, W. (1994). Religiositeit en depressieve
symptomen bij ouderen: Een studie bij ouderen in de
bevolking van Sassenheim. [Religion and depressive
symptoms in the elderly: A study of the elderly in the
community of Sassenheim]. Tijdschrift voor Psychia-
trie, 36, 509–519.

Brewin, C. R., Andrews, B., & Gotlib, I. H. (1993).

Psychopathology and early experience: A reappraisal
of retrospective reports. Psychological Bulletin, 113,
82–98.

Brink, T. L., Yesavage, J. A., Heersema, P. H., Adey, M.,

& Rose, T. L. (1982). Screening tests for geriatric
depression. Clinical Gerontologist, 1, 37–43.

Brown, G. W. (1972). Life-events and psychiatric illness:

Some thoughts on methodology and causality. Journal
of Psychosomatic Research, 16, 311–320.

Brown, G. W., & Harris, T. (1978). Social origins of

depression: A study of psychiatric disorder in women.
London: Tavistock.

Cappeliez, P. (1988). Some thoughts on the prevalence and

etiology of depressive conditions in the elderly. Special
issue: Francophone research in gerontology in Canada.
Canadian Journal on Aging, 7, 431–440.

Carpenter, E. M., & Kirkpatrick, L. A. (1996).

Attachment style and presence of a romantic partner
as moderators of psychophysiological responses to a
stressful laboratory situation. Personal Relationships,
3, 351–367.

CBS. (1997). Vademecum gezondheidsstatistiek Neder-

land [On-line] [Handbook of health statistics, The
Netherlands]. Available at: http://statline.cbs.nl.

Draper, B. (1994). Suicidal behaviour in the elderly.

International Journal of Geriatric Psychiatry, 9,
655–661.

Fincham, F. D., Beach, S. R. H., Harold, G. T., & Osborne,

L. N. (1997). Marital satisfaction and depression:
Different causal relationships for men and women?
Psychological Sciences, 8, 351–357.

Gerlsma, C. (1994). Parental rearing styles and

psychopathology: Notes on the validity of ques-
tionnaires for recalled parental behavior. In C.
Perris, W. A. Arrindell, & M. Eisemann (Eds.),
Parenting and psychopathology (pp. 75–105). Chi-
chester: Wiley.

Gerlsma, C., Das, J., & Emmelkamp, P. M. G. (1993).

Depressed patients’ parental representations: Stability
across changes in depressed mood and specificity
across diagnoses. Journal of Affective Disorders, 27,
173–181.

Gerlsma, C., Emmelkamp, P. M. G., & Arrindell, W. A.

(1990). Anxiety, depression, and perception of early
parenting: A meta-analysis. Clinical Psychology
Review, 11, 251–277.

Gerlsma, C., Kramer, J. J. A. M., Scholing, A., &

Emmelkamp, P. M. G. (1994). The influence of mood
on memories of parental rearing practices. British
Journal of Clinical Psychology, 33, 159–172.

Gurland, B. J., & Cross, P. S. (1982). Epidemiology of

psychopathology in old age: Some implications for
clinical services. Psychiatric Clinics of North America,
5, 11–26.

Henry, B., Moffitt, T. E., Caspi, A., Langley, J., & Silva, P.

A. (1994). On the ‘‘Remembrance of things past’’: A
longitudinal evaluation of the retrospective method.
Psychological Assessment, 6, 92–101.

Hickie, I., Parker, G., Wilhelm, K., & Tennant, C. (1991).

Perceived interpersonal risk factors of non-endogenous
depression. Psychological Medicine, 21, 399–412.

Katona, C. (1993). The aetiology of depression in old age.

International Review of Psychiatry, 5, 407–416.

Kerkhof, A. J. F. M., Schmidtke, A., Bille-Brahe, U., de

Leo, D., & Lo¨nnqvist, J. (1994). Attempted suicide in
Europe. Findings from the WHO multicentre study on
parasuicide by the WHO regional office for Europe.
Leiden, The Netherlands: DSWO.

Kok, R. M. (1994). Zelfbeoordelingsschalen voor depres-

sie bij ouderen. [Self report instruments for depression
in the elderly]. Tijdschrift voor Gerontologie en
Geriatrie, 25, 150–156.

Kraaij, V. (2000). Depressive symptoms in the elderly:

Negative life events and buffering factors. Dissertation,
Leiden University, Leiden, The Netherlands.

Kraaij, V., Arensman, E., & Spinhoven, Ph. (in press).

Negative life events and depression in the elderly: A
meta-analysis. Journal of Gerontology: Psychological
Sciences.

Kraaij, V., & de Wilde, E. J. (2001). Negative life events

and depressive symptoms in the elderly: A life span
perspective. Aging & Mental Health, 5, 84–91.

Lapierre, S., Pronovost, J., Dube, M., & Delisle, I. (1992).

Risk factors associated with suicide in elderly persons
living in the community. Canada’s Mental Health, 40,
8–12.

McIntosh, J. L., Santos, J. F., Hubbard, R. W., &

Overholser, J. C. (1994). Special high-risk factors in
suicide among older adults. In J. L. McIntosh, J. F.

Negative life events and bonding

213

Santos, R. W. Hubbard, & J. C. Overholser (Eds.),
Elder suicide: Research, theory, and treatment
(pp. 101–131). Washington DC: American Psycho-
logical Association.

Murphy, E. (1982). Social origins of depression in old age.

British Journal of Psychiatry, 141, 135–142.

Norris, F. H., & Kaniasty, K. (1992). Reliability of delayed

self-reports in disaster research. Journal of Traumatic
Stress, 5, 575–588.

O’Hara, M. W., Kohout, F. J., & Wallace, R. B. (1985).

Depression among the rural elderly: A study of
prevalence and correlates. Journal of Nervous and
Mental Disease, 173, 582–589.

Olin, J. T., Schneider, L. S., Eaton, E. M., Zemansky, M.

F., & Pollock, V. E. (1992). The Geriatric Depression
Scale and the Beck Depression Inventory as screening
instruments in an older adult outpatient population.
Psychological Assessment, 4, 190–192.

Orrell, M. W., & Davies, A. D. M. (1994). Life events

in the elderly. International Review of Psychiatry, 6,
59–71.

Parker, G. (1981). Parental reports of depressives: An

investigation of several explanations. Journal of
Affective Disorders, 3, 131–140.

Parker, G. (1983). Parental overprotection: A risk factor in

psychosocial development. New York: Grune &
Stratton.

Parker, G. (1989). The Parental Bonding Instrument:

Psychometric properties reviewed. Psychiatric Devel-
opments, 4, 317–335.

Parker, G. (1990). The Parental Bonding Instrument: A

decade of research. Social Psychiatry and Psychiatric
Epidemiology, 25, 281–282.

Parker, G. (1994). Parental bonding and depressive

disorders. In M. B. Sperling & W. H. Berman (Eds.),
Attachment in adults: Clinical and developmental
perspectives (pp. 299–312). New York: Guilford.

Parker, G., Tupling, H., & Brown, L. B. (1979). A Parental

Bonding Instrument. British Journal of Medical
Psychology, 52, 1–10.

Parker, G. B., Barrett, E. A., & Hickie, I. B. (1992). From

nurture to network: Examining links between percep-
tions of parenting received in childhood and social
bonds in adulthood. American Journal of Psychiatry,
149, 877–885.

Pierce, G. R., Sarason, I. G., & Sarason, B. R. (1996).

Coping and social support. In M. Zeidner & N. S.
Endler (Eds.), Handbook of coping: Theory, research,
applications (pp. 434–451). New York: Wiley.

Rector, N. A., & Roger, D. (1997). The stress buffering

effects of self-esteem. Personality and Individual
Differences, 23, 799–808.

Rook, K. S., & Ituarte, P. H. G. (1999). Social control,

social support, and companionship in older adults’
family relationships and friendships. Personal Rela-
tionships, 6, 199–211.

Ruegg, R. G., Zisook, S., & Swerdlow, N. R. (1988).

Depression in the aged: An overview. Psychiatric
Clinics of North America, 11, 83–99.

Stevens, N. L., & Andersson, L. (1996). Overprotection by

parents and anxiety in later life. Journal of Mental
Health and Aging, 2, 39–50.

Tabachnick, B. G., & Fidell, L. S. (1996). Using

multivariate statistics, third edition. New York:
HarperCollins.

Teasdale, J. D. (1983). Negative thinking in depression:

Cause, effect or reciprocal relationship? Advances in
Behavior Research and Therapy, 5, 3–25.

Thoits, P. A. (1995). Stress, coping, and social support

processes: Where are we? What next? Journal of
Health and Social Behavior (Extra Issue), 53–79.

van Marwijk, H. W. J. (1995). Depression in the elderly as

seen in general practice: Prevalence, diagnosis, and
course. Dissertation, Leiden University, Leiden, The
Netherlands.

West, M., Livesley, W. J., Reiffer, L., & Sheldon, A.

(1986). The place of attachment in the life events
model of stress and illness. Canadian Journal of
Psychiatry, 31, 202–207.

Whisman, M. A. (1999). Marital dissatisfaction and

psychiatric disorders: Results from the National
Comorbidity Survey. Journal of Abnormal Psychology,
108, 701–706.

Whisman, M. A., & Bruce, M. L. (1999). Marital

dissatisfaction and incidence of major depressive
episode in a community sample. Journal of Abnormal
Psychology, 108, 674–678.

Whisman, M. A., Sheldon, C. T., & Goering, P.

(2000). Psychiatric disorders and dissatisfaction with
social relationships: Does type of relationship
matter? Journal of Abnormal Psychology, 109,
803–808.

Wilhelm, K., & Parker, G. (1988). The development of a

measure of intimate bonds. Psychological Medicine,
18, 225–234.

Yarrow, M. R., Campbell, J. D., & Burton, R. V. (1970).

Recollection of childhood: A study of the retrospective
method. Monographs of the Society for Research in
Child Development, 35, (5, serial no. 138).

Yesavage, J. A., Brink, T. L., Rose, T. L., Lum, O., Huang,

V., Adey, M., & Leirer, O. von (1983). Development
and validation of a geriatric depression screening scale:
A preliminary report. Journal of Psychiatric Research,
17, 37–49.

V. Kraaij and N. Garnefski

214