You are on page 1of 27

Personality and Mental Health

(2017)
Published online in Wiley Online Library
(wileyonlinelibrary.com) DOI 10.1002/pmh.1385

Parentchild relationship associated with the


development of borderline personality disorder:
A systematic review

MARIE-VE BOUCHER, JESSICA PUGLIESE, CATHERINE ALLARD-CHAPAIS,


SERGE LECOURS, LOLA AHOUNDOVA, RACHEL CHOUINARD AND
SARAH GAHAM, Department of Psychology, Universit de Montral, C.P. 6128, succursale
Centre-Ville, Montreal, Qubec, Canada, H3C 3J7

ABSTRACT
The parentchild relationship (PCR) is considered as a central factor in most contemporary theories on the
aetiology of borderline personality disorder (BPD). This systematic review aimed to answer the three following
questions: (1) How is the PCR described by BPD participants and their parents in comparison to other
normative and clinical groups? (2) Which aspects of the PCR are specically associated with a BPD diagnosis
in adulthood? (3) How can the facets of the PCR identied in the reviewed studies shed light on the general
aetiological models of BPD? Forty studies were retained and divided into three categories: perspective of BPD
probands, perspective of their parents and perspective of family. Borderline personality disorder participants
consistently reported a much more dysfunctional PCR compared to normal controls. Comparisons with
participants presenting DSM-IV Axis-I and Axis-II disorders were a lot less consistent. BPD probands rated
more negatively the PCR compared to their parents. Low parental care and high parental overprotection may
represent a general risk factor for psychopathology, different from normal controls but common to BPD and
other mental disorders. An interesting candidate for a specic PCR risk factor for BPD appears to be
parental inconsistency, but further studies are necessary to conrm its specicity. Copyright 2017 John
Wiley & Sons, Ltd.

The biosocial interaction theory of the First, physical and sexual abuse, and more
development of borderline personality disorder generally trauma, have been the focus of many
(BPD) has reached a large consensus in the literature reviews 35. However, not every person
scientic literature 1,2. In a nutshell, the theory who has BPD has been abused or traumatized 6.
suggests that biological factors, such as a highly Also, after controlling for family environment, the
reactive temperament, and the social environment relationship between child sexual abuse and later
interact to contribute to the development of BPD. psychological adjustment became non-signicant 7.
Many social factors conducing to BPD have been In other words, family environment was
proposed. consistently confounded with child sexual abuse.

Copyright 2017 John Wiley & Sons, Ltd. (2017)


DOI: 10.1002/pmh
Boucher et al.

Second, insecure attachment has also been (PD) in adult offspring (the only one so far to
strongly related to BPD 8. More specically, the our knowledge), Johnson, Cohen 19 demonstrated
preoccupied, unresolved and fearful types were convincingly that composite indexes of parenting
the most frequent attachment styles found in behaviours were signicantly associated with the
BPD patients 810. In line with these results, a development of various PD in adulthood, over
disorganized attachment between the caregiver and above the inuence of parental psychiatric
and the child has been proposed as a risk factor disorders and childrens emotional and
for the development for BPD 11. According to this behavioural disorders. Johnson and colleagues
author, the caregiver in a disorganized attachment suggested that future research should examine
can be viewed by the child simultaneously as a whether some specic parenting behaviours were
secure base and a source of threat, which can result related to the risk for specic PD.
in an approachavoidance dilemma in later Many literature reviews on the social factors
relationships. In short, studies on attachment in leading to BPD have been published in the last
BPD patients generated non-specic results: many thirty years. They have focused on family factors
2022
types of attachment representing an insecure or , pathological childhood experiences 23 and
disorganized parentchild dynamic were related psychosocial risk factors 24. Yet, none of these
to the aetiology of BPD. Moreover, failure to has put specic emphasis on the PCR. Also, let
develop a secure base and attachment trauma were alone the work of Keinanen and colleagues 24
generally identied as potential factors explaining and Madeddu and colleagues 21, none of these
the aetiology of this personality disorder 12. literature reviews has used a systematic
The parentchild relationship (PCR) is methodology for retrieving relevant references.
probably the variable that best summarizes the Consequently, it is hard to know today what the
social portion of the biosocial interaction theory empirical basis of the PCR in BPD is and how this
because it can be considered as the most proximal knowledge relates to contemporary theories on the
causal environmental factor in the development of aetiology of BPD. Moreover, in line with
BPD. Recent conceptual models of BPD aetiology developmental psychology principles, one can
indeed labelled the PCR in diverse ways: parental wonder whether there are particular aspects of
invalidation of the childs subjective experience the PCR which are specically associated with
13
, biparental failure 14, disorganized attachment a BPD diagnosis in adulthood, but with no
11
or controlled-caregiving/punitive interpersonal other disorder, or if PCR-related variables
strategies 15. represent a non-specic risk for psychopathology
Concurrently with these models specic to in general.
BPD, developmental psychology theorists A systematic review of studies on the PCR in
proposed that geneenvironment interactions BPD should therefore be informative. The review
may explain the emergence of various mental proposed here will synthesize all relevant studies
disorders but also of normality, and that research made since BPD was ofcially introduced in the
efforts should therefore aim at providing a full DSM-III (i.e., 1980). More precisely, this review
understanding of this complex naturenurture intends to gather information on the three
interplay 16,17. More work is also needed to better following questions:
understand the concepts of multinality (i.e., a
single risk factor being associated to multiple 1) How is the PCR described by BPD participants
outcomes) and equinality (i.e., different risk and their parents in comparison to other
factors leading to the same outcome)18. In a normative and clinical groups?
longitudinal study on the effects of parenting 2) Which aspects of the PCR are specically
behaviour on the risk for personality disorders associated with a BPD diagnosis in adulthood?

Copyright 2017 John Wiley & Sons, Ltd. (2017)


DOI: 10.1002/pmh
Parentchild relationship associated with the development of borderline personality disorder: A systematic review

3) How can the facets of the PCR identied in results pertaining specically to an aspect of the
the reviewed studies shed light on the general parentchild interactions which had been
aetiological models of BPD? measured in at least another screened article.
Articles were excluded if they (1) focused
Method on parents diagnosed with BPD and their own
children (without BPD diagnosis); (2)
Search strategy presented results based only on chart reviews;
and (3) were solely theoretical models or case
A computerized search was undertaken on May studies.
2nd 2016 on the three following electronic
databases: PsychInfo, Medline and Web of
Study selection
Science. Search terms were combinations of
borderline personality disorder and keywords Using EndNote, all titles and abstracts were
related to parenthood (mother* or father* or reviewed independently by two researchers (the
caregiver* or caretaker* or parent*). In order to rst author and either LA, CAC, RC, SG or JP).
make the search as efcient as possible, the source When a reference satised all inclusion criteria
type was limited to peer-reviewed journals, and and met no exclusion criteria, it was sorted in
the language was limited to English and French. the to be included group. However, when a
Also, since BPD rst appeared as an ofcial reference met at least one exclusion criterion or
psychiatric diagnosis in the DSM-III, the failed to meet at least one inclusion criterion, it
earliest publication year was set to 1980. The was sorted in the group of the appropriate
complete references retained from all three exclusion criterion/lacking inclusion criterion.
databases were transferred to EndNote. When study selection was completed, the rst
Duplicates were removed. author then compared her article sorting to
each of her colleagues. Any discrepancies
between selections were discussed and resolved.
Inclusion and exclusion criteria See Figure 1 for the study sorting and selection
In line with this reviews rationale, articles were strategy.
included if they (1) focused on the PCR in BPD,
that is the perception of the actual and/or past
Rationale for a narrative review
PCR by either the BPD participants themselves
or by their parents; (2) compared BPD Given the great heterogeneity of concepts and
participants perception of PCR to at least another measures used by the reviewed studies, a
control group AND/OR compared parents of BPD narrative approach was preferred to a meta-
participants perception of PCR to parents of analysis for our systematic review. This
another control group OR predicted BPD qualitative method makes it possible to
diagnosis with the PCR measure; (3) presented a summarize data which would not be included
standardized procedure, that is a valid instrument in a meta-analysis, because there are too few
for conrming BPD diagnosis in late adolescence studies reporting on the same variables and
or adulthood; (4) included participants with BPD measures (for example, the perspective of parents
diagnosis (not only traits); (5) for cross-sectional studies or the abuse and neglect studies). A
studies, included a sample composed of mostly meta-analysis might have been achieved with
adult participants (age 18 years and over); (6) the 17 perspective of BPD probands studies using
had data collected directly from the BPD the Parental Bonding Instrument (PBI) (see
participants and/or their parents; (7) presented section 2.2.1.1). However, we do think that

Copyright 2017 John Wiley & Sons, Ltd. (2017)


DOI: 10.1002/pmh
Boucher et al.

Figure 1: Study sorting and selection strategy

the other 23 studies reviewed allow us to outline Results


a more complex and nuanced picture of the
PCR in BPD. Furthermore, being rst of its Characteristics of the studies
kind, this systematic review on this subject
aimed at describing the diverse variables used Forty studies were retained. Following the type of
to measure the PCR in BPD and not focusing informant used, the studies were divided into
on only two of these variables (i.e., parental care three categories, with one study 25 being presented
and overprotection). in two categories: perspective of BPD probands

Copyright 2017 John Wiley & Sons, Ltd. (2017)


DOI: 10.1002/pmh
Parentchild relationship associated with the development of borderline personality disorder: A systematic review

(33 studies, see Table 1), perspective of parents of discriminated well BPD participants from normal
BPD probands (3 studies, see Table 2) and controls (NC). Indeed, whether using the PBI
perspective of family (5 studies, see Table 3). A two-factor or three-factor structure and whatever
summary of results specic to each reviewed study the sample composition (inpatients only,
can be found in the supplemental material available outpatients only or mixed sample), BPD
on the journals website (Tables S1, S2 and S3). participants tended to report signicantly lower
parental care and higher parental overprotection
Perspective of borderline personality disorder compared to NC.
probands When PBI scores of BPD participants were
compared to those of other DSM-IV Axis-I
Parental care and overprotection studies and Axis-II groups, results were less consistent.
However, compared to parental overprotection,
These two variables are presented in the same parental care seemed to more robustly
category, because they were always measured discriminate BPD participants from Axis-I
together in the various studies included. participants (i.e., psychiatric controls, schi-
Subsections were named after the type of zophrenics, anorectics and participants without
instrument used to measure parental care and PDs), with BPD participants reporting sig-
overprotection. Because almost half of all reviewed nicantly lower parental care compared to
studies used the PBI, a separate subsection was Axis-I participants.
formed for studies using this questionnaire (see As for comparisons to participants with other
subsection below). The two following subsections personality disorders (OPD), results were even less
present studies using other standardized measures consistent, some studies stating no signicant
of parental care and overprotection and non- differences for all four PBI subscales between
standardized instruments respectively. BPD and OPD participants, while other studies
reported that either parental care or parental
Studies using the Parental Bonding Instrument. overprotection scores were signicantly different
Seventeen studies compared BPD participants between both groups. These inconsistent results
scores on the PBI to those of other clinical groups. might be explained by the type of patients
Seven had an all-female sample 2531, while two included in the samples. Indeed, only studies with
had an all-male sample 32,33. As for the eight an inpatient-only or a mixed (i.e., inpatient and
remaining studies, seven used a predominantly outpatient) sample reported no signicant
female sample 3440, and one study did not specify differences between BPD and OPD participants
the sex proportion of its sample 41. for all four PBI subscales scores. Conversely,
All studies except for one 28 used the two-factor studies using outpatient-only samples reported
structure of the PBI, measuring care and signicant score differences between BPD and
overprotection for both parents. As for Laporte OPD participants on at least one PBI subscale.
and Guttman 28, they used the new three-factor Parental care and overprotection could therefore
structure, consisting of care, denial of be associated to psychopathology severity,
psychological autonomy, and denial of instead of being specic relationship factors of
behavioural freedom (each measured for both BPD. Moreover, prediction of BPD diagnosis
parents). Most studies compared BPD groups to with PBI subscales seemed inconsistent: in most
other clinical and normative groups. studies including other predictors in logistic
The most consistent result concerning parental regressions, PBI subscales became non-signicant
care and overprotection measured with the PBI and other variables, such as childhood sexual
was the fact that all four PBI subscales abuse, remained signicant predictors. Finally,

Copyright 2017 John Wiley & Sons, Ltd. (2017)


DOI: 10.1002/pmh
Table 1: Characteristics of perspective of BPD proband studies

Measures of parentchild
Authors (year) BPD evaluation Age range/age mean relationship, related
Boucher et al.

Country Study design procedure N (SD), % of women section of results

Allen et al. (2005) Cross-sectional SCID-II 40 BPD OUT Total sample: Family Interaction Measure
USA 10 ST-BPD OUT 2348 years,
24 PC OUT 69.5%
19 NC
Arens, Grabe, Spitzer Longitudinal SCID-II 17 BPD C-D Total sample: EMBU (T0)
& Barnow (2011) (5-year 17 DEP C-D TO: 15.3 years (2.2)

Copyright 2017 John Wiley & Sons, Ltd.


Germany follow-up) 34 NC T1: 19.6 years (2.4)
Baker, Silk, Westen, Cross-sectional DIB 31 BPD IN BPD: 30.4 years Adjective Check List
Nigg & Lohr (1992) 15 MDD IN (8.0), 87.1%
USA 14 NC MDD: 40.7 years
(9.4), 60.0%
NC: 37.8 years
(11.1), 64.3%
Bandelow, Krause, Cross-sectional SCID-II 66 BPD OUT -BPD: 30.2 years Standardized questionnaire
Wedekind, Broocks, 109 NC (9.4), 71.2% of 203 items evaluating
Hajak & Ruther (2005) -NC: 32.3 years 4 domains, one of them
Germany (6.6), 60.6% being parental attitudes
Byrne, Velamoor, Cernovsky, Cross-sectional DSM-III criteria-based 15 BPD IN and OUT BPD: 24.6 years -PBI
Cortese & Losztyn (1990) diagnosis made by 14 SZ IN and OUT (N/A), 86.7% -Childhood Life Events
Canada two psychiatrists 873 NC SZ: 25.9 years and Family Characteristics
and one resident (3.3), 71.4% Questionnaire
Dubo, Zanarini, Lewis & Cross-sectional DIB-R 42 BPD IN BPD: 1860 years, Childhood Experiences
Williams (1997) DIPD-R 17 OPD IN 71.4% Questionnaire
USA OPD: 2057 years,
58.8%
Fletcher, Parker, Bayes, Cross-sectional DIPD-IV 22 BPD OUT BPD: 36.7 years Measure of Parental Style
Paterson & McClure (2014) and C-D (N/A), 87,5%
Australia 22 BiD-II OUT BiD-II: 32.8 years
and C-D (N/A), 50%
Fossati, Donati, Donini, Cross-sectional SCID-II 44 BPD IN and OUT BPD: 27.6 years PBI
Novella, Bagnato & 98 CB IN and OUT (7.2), 56.8%
Maffei (2001) CB: 30.4 years
Italy (8.7), 54.1%
(Continues)

DOI: 10.1002/pmh
(2017)
Table 1: (continued)

Measures of parentchild
Authors (year) BPD evaluation Age range/age mean relationship, related
Country Study design procedure N (SD), % of women section of results

39 CAC IN and OUT CAC: 31.0 years


70 NoPD IN (8.6), 64.1%
and OUT NoPD: 30.8 years
206 NC (8.7), 75.7%
NC: 24.4 years
(8.8), 82.0%

Copyright 2017 John Wiley & Sons, Ltd.


Frank & Hoffman (1986) Cross-sectional DIB-2 10 BPD OUT BPD: 2333 years, Parent Activity Inventory
Canada 14 NEU OUT 100%
NEU: 2128 years,
100%
Ghiassi, Dimaggio & Cross-sectional SCID-II 50 BPD IN BPD: 26.2 years German translation of EMBU
Brune (2010) 20 NC (6.6), 92%
Germany NC: 26.0 years
(4.2), 65%
Goldberg, Mann, Wise & Cross-sectional DSM-III diagnosis 24 BPD IN BPD: 24.9 years PBI
Segall (1985) 22 PC IN (6.0), N/A
USA 10 NC PC: 34.4 years
(12.7), N/A
NC: 28.4 years
(8.4), N/A
Guttman & Laporte (2002) Cross-sectional DIB-R 21 BPD OUT Total sample: PBI
Canada 23 AN OUT 1640 years, 100%
25 NC
Hayashi, Suzuki & Cross-sectional DSM-III R criteria- 13 BPD IN BPD: 27 years Semi-structured video-recorded
Yamamoto (1995) based evaluation 13 PC IN (6.4), 77% interview about the
Japan by psychiatrist PC: 30.0 years experience of the parent
(8.4), 85% child relationship
Heffernan & Cloitre (2000) Cross-sectional SCID-II 26 BPD + PTSD C-D Total sample: Child Maltreatment
USA 45 PTSD C-D 37.3 years Interview Schedule
(11 years)
100%
Helgeland & Torgersen (1997) Cross-sectional SCID-II 14 BPD IN BPD: 30.9 years PBI
Norway 19 SZ IN (N/A), 79%
(Continues)
Parentchild relationship associated with the development of borderline personality disorder: A systematic review

DOI: 10.1002/pmh
(2017)
Table 1: (continued)

Measures of parentchild
Authors (year) BPD evaluation Age range/age mean relationship, related
Boucher et al.

Country Study design procedure N (SD), % of women section of results

15 NC SZ: 27.9 years


(N/A), 27%
NC: 26 years
(N/A), 47%
Hernandez, Arntz, Gaviria, Cross-sectional SCID-II 32 BPD IN and OUT Total sample: PBI
Labad & Gutierrez- DIB-R 33 OPD IN and OUT 38.8 years (10.7),

Copyright 2017 John Wiley & Sons, Ltd.


Zotes (2012) 34 NoPD IN and OUT 100%
Spain
Huang, Napolitano, Wu, Cross-sectional MSI-BPD 152 BPD OUT Total sample: 28.4 years EMBU
Yang, Xi, Li & Li (2014) SCID-II 79 OPD OUT (9.1), 59.1%
China 55 NoPD OUT
Joyce et al. (2003) Cross-sectional SCID-II 30 BPD OUT Total sample: PBI (mother and father
USA 43 AvPD OUT 1835 years, combined score)
58.9%
Katerndahl, Burge & Cross-sectional SCID-II 27 BPD OUT Total sample: PBI
Kellogg (2005) 43 PC OUT 28.7 years (7.0),
USA 22 NC 100%
Laporte & Guttman (2007) Cross-sectional DIB-R 35 BPD OUT Total sample: PBI (3-factor structure)
Canada 34 AN OUT 1640 years, 100%
33 NC
Machizawa-Summers (2007) Cross-sectional -DSM-IV-TR criteria- 45 BPD OUT BPD: 28.5 years PBI
Japan based diagnosis by 45 NoPD OUT (6.5), 100%
psychiatrist PC: 32.9 years
-BSI (7.6), 100%
Merza, Papp & Szab (2015) Cross-sectional SCID-II 80 BPD IN BPD: 30.5 years Traumatic Antecedents
Hungary 73 MDD IN (10.9), 85% Questionnaire
51 NC MDD: 44.3 years
(5.9), 82.2%
NC: 33.6 years
(8.7), 86,3%
Paris, Zweig-Frank & Cross-sectional DIB-R 13 BPD-Life OUT BPD-Life: PBI
Guzder (1993) 26 BPD OUT 34.9 years, 100%
Canada BPD: 37.7 years, 100%
(Continues)

DOI: 10.1002/pmh
(2017)
Table 1: (continued)

Measures of parentchild
Authors (year) BPD evaluation Age range/age mean relationship, related
Country Study design procedure N (SD), % of women section of results

Paris, Zweig-Frank & Cross-sectional DIB-R 78 BPD OUT BPD: 28.3 years -PBI
Guzder (1994a) 72 OPD OUT (6.3), 100% -Developmental interview,
Canada OPD: 29.7 years assessing physical
(7.2), 100% abuse by a caretaker
during the 16 rst years
Paris, Zweig-Frank & Cross-sectional DIB-R 61 BPD OUT BPD: 30.6 years -PBI

Copyright 2017 John Wiley & Sons, Ltd.


Guzder (1994b) and C-D (7.9), 0% -Developmental interview
Canada 60 OPD OUT OPD: 32.3 years
and C-D (6.5), 0%
Schwarze, Hellhammer, Cross-sectional SCID-II 100 BPD IN, OUT BPD: 31.6 years PBI
Stroehle, Lieb & and C-D (9.7), 90%
Mobascher (2015) 100 NC NC: 32.0 years
Germany (10.3), 90%
Timmerman & Cross-sectional PDQ-R (for Pr 8 BPD FPI FPI: 35 years PBI
Emmelkamp (2005) and NC) 39 FPI (11.3), 20
Netherlands IPDE (for FPI) 43 BPD Pr 71 years, 0%
192 Pr Pr: 33.8 years
22 BPD C-D (8.9), 2166
195 NC years, 0%
NC: 43.6 years
(11.7), 2265
years, 0%.
Torgersen & Cross-sectional SIPD 19 SZT OUT Total sample: PBI
Alnaes (1992) 36 BPD OUT 1859 years, 70%
Norway 165 OPD OUT
52 NoPD OUT
Zanarini et al. (1997) and Cross-sectional SCID-II 358 BPD OUT BPD: 27.6 years Revised Childhood
Zanarini et al. (2000a) DIB-R 109 OPD OUT (6.8), 77.1% Experiences
USA DIPB-R OPD: 29.3 years Questionnaire
(9.1), 56.0%
Zanarini, Gunderson, Cross-sectional DIPB 50 BPD OUT BPD: 29.2 years -Retrospective Family
Marino, Schwartz & DIB 29 AnPD OUT (6.4), 66% Pathology Questionnaire
Frankenburg (1989) 26 D-OPD OUT
(Continues)
Parentchild relationship associated with the development of borderline personality disorder: A systematic review

DOI: 10.1002/pmh
(2017)
Table 1: (continued)

Measures of parentchild
Authors (year) BPD evaluation Age range/age mean relationship, related
Boucher et al.

Country Study design procedure N (SD), % of women section of results

USA AnPD: 25.2 years


(6.1), 27.6%
D-OPD: 31.3 years
(7.2), 84.6%
Zweig-Frank & Paris (1991) Cross-sectional DIB 62 BPD OUT BPD: 29.6 years PBI
Canada 99 PC OUT (9.6), 54.8%

Copyright 2017 John Wiley & Sons, Ltd.


PC: 29.3 years
(8.6), 54.5%
Zweig-Frank & Paris (2002) Longitudinal DIB, retrospective 59 BPD IN BPD: 50.9 years PBI
Canada (but PBI-related version (based on and OUT (7.6), 83%
results are cross- chart review)
sectional)

Note. AN, anorexia nervosa diagnosis; AvPD, avoidant personality disorder; BiD-dep, bipolar disorderdepressed condition; BiD-II, bipolar II disorder; BPD,
borderline personality disorder; BPD-Life, participants who met the criteria for lifetime BPD but who no longer had the diagnosis; BPD-NA, borderline personality
disorder with no reported childhood abuse; BPD-PA, borderline personality disorder with reported parental physical and/or sexual abuse; BSI, Borderline Syndrome
Index; CAC, cluster A or C personality disorder patients; CB, cluster B personality disorder patients; C-D, community-drawn; D-OPD, dysthymic participants with
other personality disorder; DEP, depressive disorder; DIB, Diagnostic Interview for Borderlines; DIB-2, Diagnostic Interview for Borderlines (second edition);
DIB-R, Revised Diagnostic Interview for Borderlines; DIPD, Diagnostic Interview for Personality Disorders; DIPD-R, Revised Diagnostic Interview for Personality
Disorders; DIS, Diagnostic Interview Schedule; CSA, childhood sexual abuse; FPI, forensic psychiatric inpatients; IN, inpatients; IPDE, International Personality
Disorder Examination; LBSI, Lifetime Borderline Symptom Inventory; LIDR, Linking Interview with Diagnostic Rules Manual; MDD, major depressive disorder;
MMPI, Minnesota Multiphasic Personality Inventory; N/A, non-available information; NC, normal controls; NEU, neurotic controls; NEU-PD, neurotic and
other personality disorder controls; NPD, narcissistic personality disorder; NoPD1, participants with no personality disorder diagnosis; OPD, participants with other
personality disorders controls; OUT, outpatients; PBI, Parental Bonding Inventory; PC1, psychiatric controls; PDQ-R, Personality Disorder Questionnaire
Revised; PIP, Physician Interview Program; Pr, prisoners; PTSD, post-traumatic stress disorder; SCID-II, Structured Clinical Interview for DSM-IV Axis II
Disorders; SIPD, Structured Interview for DSM-III, Axis II Personality Disorders; ST-BPD, subthreshold for borderline personality disorder; SZ, schizophrenia;
SZT, schizotypal personality disorder; UniD, major depression (unipolar).
1
As opposed to psychiatric controls, participants with no personality disorders were patients formally evaluated for Axis II diagnoses and who did not receive any
personality disorder (PD) diagnosis. Hence, the presence of PD other than BPD was not ofcially ruled out for psychiatric controls in Goldberg and colleagues
(1995) and Zweig-Frank & Paris (1991) studies.

DOI: 10.1002/pmh
(2017)
Table 2: Characteristics of perspective of parents of BPD proband studies

Authors (year) BPD evaluation Selection of Age range/age mean Measures of parentchild
Country Study design procedure clinical groups N (SD), % of women relationship

Bezirganian, Cohen & Longitudinal T3: Modied questions C-D T3: Total sample Parentchild interactions

Copyright 2017 John Wiley & Sons, Ltd.


Brook (1993) (2.5 years of the (then proposed) 57 BPD -T2 (N = 778): scales (T2)
USA between T2 SCID-II for DSM-III-R 66 NPD 13.7 years
and T3) 41 HPD (2.8), 50%
70 CD -T3 (N = 776):
113 CB 16.4 years
118 CAC (2.8 years), 48%
123 Axis-I
Goodman et al. (2010) Cross-sectional -MSI-BPD C-D 321 BPD BPD: 24.0 years Anonymous Internet survey,
USA -Formal BPD diagnosis 87 Non-BPD (8.4), 100% consisting of 109 items
by a professional at some sisters Non-BPD sisters: and covering aspects of
point in the probands life 21.6 years the probands life from
(9.2), 100% pregnancy through young
adulthood.
Goodman, Patel, Cross-sectional -MSI-BPD C-D 97 BPD BPD: 25.8 years Anonymous Internet survey,
Oakes, Matho & 166 Non-BPD (7.3), 0% same as Goodman
Triebwasser (2013) -Formal BPD diagnosis by a siblings Non-BPD siblings: et al. (2010)
USA professional at some point 25.7 years
in the probands life (8.2), 0%

Note. Axis-I, axis-I disorder; BPD, borderline personality disorder; CAC, cluster A or C personality disorder patients; CB, cluster B personality disorder patients;
CD, conduct disorder; C-D, community-drawn; HPD, histrionic personality disorder; MSI-BPD, McClean Screening Instrument for Borderline Personality
Disorder; NPD, narcissistic personality disorder; SCID-II, Structured Clinical Interview for DSM-III-R Axis II Disorders.
Parentchild relationship associated with the development of borderline personality disorder: A systematic review

DOI: 10.1002/pmh
(2017)
Boucher et al.

Table 3: Characteristics of perspective of family studies

Authors (year) BPD evaluation Age range/age mean Measures of parentchild


Country Study design procedure N (SD), % of women relationship

Gunderson and Cross-sectional DIPD 21 families (BPD IN BPD daughters: Dyadic Relationship Scale

Copyright 2017 John Wiley & Sons, Ltd.


Lyoo (1997) daughter, mother 2035 years
USA and father) 100%
Guttman & Laporte (2000) Cross-sectional DIB-R Number unclear for Total sample: Family Interview
Canada BPD OUT (26 (in 1640 years for Protectiveness
method) or 27 BPD: 32 years and Empathy
(in abstract)) (N/A), 100%
28
AN OUT and C-D AN: 22 years
27
NC (N/A), 100%
NC: 21 years
(N/A), 100%
Guttman & Laporte (2002) Cross-sectional DIB-R 21 BPD OUT and C-D Total daughter PBI
Canada 23 AN OUT and C-D sample:
*also presented in Table 1 25 NC OUT and C-D 1640 years, 100%
Laporte, Paris, Guttman & Cross-sectional DIB-R 56 BPD OUT-sister dyads 1845 years, 100% -PBI
Russell (2011) DIPD-IV -The 2012 paper reported -Sibling Inventory of
and only on the 53 dyads Differential Experience
Laporte, Paris, Guttman, with one sister free of
Russell & Correa (2012) psychopathology
Canada

Note. AN, anorexia nervosa diagnosis; BPD, borderline personality disorder; C-D, community-drawn; DIB-R, Revised Diagnostic Interview for Borderlines; DIPD,
Diagnostic Interview for Personality Disorders (1987 version); DIPD-IV, Diagnostic Assessment for Personality Disorders (1996 version); IN, inpatients; N/A, non
available information; NC, normal controls; OUT, outpatients; PBI, Parental Bonding Inventory.

DOI: 10.1002/pmh
(2017)
Parentchild relationship associated with the development of borderline personality disorder: A systematic review

no clear patterns for PBI scores could be that only prospectively reported parental
described using proportion of sex in samples of overprotection discriminated adolescents who
reviewed studies. would later develop BPD from those who would
not. Conversely, in adult samples with
Studies using other standardized care and retrospective scores, parental lack of emotional
overprotection measures. Five studies measured warmth and parental rejection and punishment
perceived parental care and overprotection with were more frequently reported by BPD
standardized questionnaires other than the PBI. participants than by other clinical and non-
Three studies used the EMBU (Swedish acronym clinical groups. Results pertaining to parental
for own memories concerning upbringing) with overprotection in adult samples were inconsistent.
predominantly female samples 4244. One study
used the Adjective Checklist to measure, in a Studies using non-standardized instruments for
predominantly female sample, seven parental measuring parental care and overprotection. Three
behaviours, one of them being parental studies used a non-standardized instrument to
nurturance 45. In a predominantly female sample, measure perceived parental care and
one study used the Measure of Parental Style, an overprotection 4749
. Frank and Hoffman (1986)
abbreviated version of the PBI with an additional used an all-female sample, while the other studies
abuse subscale for both parents 46. used samples mostly composed of females.
In short, studies using other standardized In summary, studies using non-standardized
measures of parental care and overprotection measures of parental care and overprotection
reported results in line for the most part with those reported ndings similar to those of studies using
of studies using the PBI. More precisely, standardized measures. Compared to other clinical
retrospectively reported lack of emotional warmth, and non-clinical groups, BPD participants
especially from the mother, was found to be a consistently reported signicantly lower parental
signicant predictor of BPD diagnosis: adult BPD care. However, parental overprotection
participants reported signicantly lower levels of discriminated BPD participants from NC, but not
maternal emotional warmth compared to NC from other clinical groups.
and participants with bipolar II disorder. However,
these other standardized measures of parental care
Parental inconsistency
and overprotection consistently failed to
discriminate BPD participants from depressive Perceived parental inconsistency in BPD
participants, a clinical group which was not participants was evaluated and compared to other
specically evaluated in studies using the PBI. clinical groups in ve studies 14,5053.
According to the only longitudinal study reviewed Predominantly female samples were used in all of
in this subsection 42, an elevated degree of those studies.
internalized disorders was the only signicant In short, compared to normal and psychiatric
predictor that distinguished between BPD and controls, BPD participants reported signicantly
[participants with depressive disorders] (p.37). more conicting responses and signicantly less
However, the interaction between temperamental appropriate responses from their rst parental
harm avoidance and perceived maternal gure 50. The four other studies compared BPD
overprotection was a signicant predictor of the participants scores of caretaker inconsistency to
risk for BPD when compared to NC ve years those of participants with OPD. Studies with low
later. Finally, there seemed to be an age as well statistical power (i.e. N = 50 BPD participants
as a measure effect on the EMBU scores: results and less than 30 control participants) failed to
of three studies using this instrument suggested detect a signicant difference for rates of

Copyright 2017 John Wiley & Sons, Ltd. (2017)


DOI: 10.1002/pmh
Boucher et al.

inconsistent treatment by a caretaker between more BPD than OPD participants reported
BPD participants and OPD. Conversely, the two abusive and/or neglectful behaviours from both
studies with high statistical power revealed that male and female caretakers.
compared to OPD inpatients, a signicantly
greater proportion of BPD inpatients reported Parental attitudes
inconsistent treatment by at least one full-time
caretaker, and that parental inconsistency, Using predominantly female samples, two studies
especially from the mother, was a signicant compared rates of parental attitudes (i.e.,
predictor of BPD diagnosis. The group difference aggressiveness, short temper, dominance) of BPD
of this parental characteristic however was of a participants to those of non-clinical and other
small magnitude, perhaps because of the stringent clinical groups 45,48. In sum, BPD participants
comparison of BPD and OPD groups. consistently reported negative parental attitudes.
However, it is not possible to specify which
parental attitude especially discriminated BPD
Parental abuse and neglect: participants from NC because results from the rst
Eleven studies compared rates of perceived study were the complete opposite of the second.
parental abuse and neglect in BPD participants Moreover, using a Bonferroni correction and after
to those of different clinical groups adjusting for age and sexual abuse, signicant
14,31,32,34,46,47,5155
. Three studies used an all- differences in ratings of parental attitudes were
female sample 31,47,54, one used an all-male sample not found between BPD participants and
32
while the other seven used a predominantly participants with major depressive disorder.
female sample.
All studies but one reviewed in this section Perspective of parents of borderline personality
measured parental abuse, while only four studies disorder probands
also evaluated parental neglect. Compared to
participants with Axis-I disorders (bipolar Only one study 56 evaluated parental reports of
disorder, post-traumatic stress disorder and major parental behaviours toward their adolescent. A
depressive disorder), BPD participants consistently maternal style made of both overprotection and
reported more frequently parental abuse and inconsistency predicted BPD diagnosis two and a
neglect. As for comparisons of rates of parental half years later, but did not predict any other
abuse with participants with Axis-II disorders, Axis-I or II diagnosis. Bezirganian, Cohen 56 also
only parental verbal abuse seemed to consistently reported that maternal overprotection alone
discriminate BPD participants from participants specically predicted histrionic PD diagnosis and
with OPD, while results pertaining to parental that maternal inconsistency alone specically
physical and sexual abuse were not as consistent. predicted BPD diagnosis two and a half years later.
Types of parental neglect which discriminated Moreover, two studies 57,58 measured parental
BPD from participants with OPD all referred to perspective on BPD childs behaviour compared
the parental inability in meeting their childs to their non-BPD siblings. The 2010 study asked
emotional needs, but not physical needs. Only parents to compare their BPD daughter to their
two low-powered studies failed to nd signicant other non-BPD daughters, while the 2013 study
differences for parental abuse and neglect between compared BPD sons to their non-BPD brothers.
BPD participants and other clinical groups. In short, participating parents retrospectively
Finally, the concept of biparental failure was described their BPD child as being unusually
suggested as an aetiological factor of BPD by sensitive and having what is commonly called a
Zanarini, Frankenburg 14 because signicantly difcult temperament as early as their rst year

Copyright 2017 John Wiley & Sons, Ltd. (2017)


DOI: 10.1002/pmh
Parentchild relationship associated with the development of borderline personality disorder: A systematic review

of life. Parents also reported a troubled daughter, while AN and NC daughters


relationship with their BPD child in childhood perception of maternal empathy only correlated
and adolescence. Verbal abuse discriminated with their fathers perception. Similarly, in BPD
BPD adolescent daughters from their non-BPD and AN families, daughters perception of paternal
sisters, while various violent and antisocial empathy correlated only with their mothers
behaviours discriminated BPD adolescent sons perception.
from their non-BPD brothers. For women,
variables related to temperamental traits Borderline personality disorder daughters and their
(moodiness, unusual sensitivity) were the only sisters
predictors of BPD diagnosis to remain signicant
when controlling for age and household income. Two papers compared BPD female participants
Excessive separation anxiety as infant was a perception of PCR to their sisters drawn from
signicant predictor of BPD diagnosis across all the same sample 61,62 and reported rather
developmental periods for men only. contradictory results. The 2011 paper 61 reported
results on the complete sample, while the 2012
paper 62 reported results only on the BPD-
Perspective of family discordant dyads, that is dyads where one sister
was free of psychopathology. Surprisingly, in both
Borderline personality disorder daughters and their studies, BPD probands reported signicantly
parents. higher maternal care1 on the Parental Bonding
Three studies compared BPD female participants Inventory (PBI), compared to their non-BPD
perception of the PCR to their parents perception sisters. However, in the 2011 paper 61, maternal
25,59,60
and both studies by Guttman and Laporte care was not a signicant predictor of BPD in
25,60
compared BPD families perceptions to those multivariate analysis predicting Revised
of families with anorectic (AN) and NC Diagnostic Interview for Borderlines (DIB-R)
daughters. Borderline personality disorder scores. Personality traits of affective instability and
daughters generally reported a more negative impulsivity were instead the signicant predictors,
perception of their parents behaviour compared over and above trauma. Laporte, Paris 61 therefore
to their parents perception of themselves. More suggested that personality traits mediated the
precisely, BPD daughters reported less parental relationship between childhood adversities and
care, more maternal overprotection and development of BPD or resilience in adulthood.
inconsistent parental values and norms, while In the 2012 paper 62, on a different
their mothers and fathers both described questionnaire (Sibling Inventory of Differential
themselves in a more normative fashion. This ExperienceSIDEquestionnaire), both BPD
marked daughterparent disagreement over probands and their non-BPD sisters generally
parental care seems to be a characteristic unique reported low scores of parental affection and high
to BPD families. Indeed, AN and NC daughters scores of parental overprotection. Similar rates
reported ratings of parental care similar to at least
one of their parents. Moreover, compared to 1
Results pertaining to maternal overprotection were not clear,
parents of AN and NC daughters, mothers and because an inconsistency is found between Laporte and
fathers of BPD daughters reported similar maternal colleagues 62 text, which reports lower maternal
overprotection in BPD participants compared to their non-
care, but signicantly lower paternal care.
BPD sisters (p.322), and Table 3 (p.323), which reports a
However, all three family members in BPD higher, but non-signicant maternal overprotection score for
families agreed on one variable, the lack of BPD participants. Explanations given by authors did not
maternal empathy directed toward the BPD completely clarify this confusion.

Copyright 2017 John Wiley & Sons, Ltd. (2017)


DOI: 10.1002/pmh
Boucher et al.

for type of perpetrator (both parents, father only, families. However, as reported in one of the
mother only) of various types of abuse and reviewed studies, all three family members in BPD
physical neglect were reported as well by BPD families agreed on one variable, the lack of maternal
probands and their sisters. Laporte and her empathy directed toward the BPD daughter, while
colleagues therefore concluded that lack of NC daughters perception of maternal empathy
parental care and marked parental abuse and only correlated with their fathers perception.
neglect were not directed specically toward the Borderline personality disorder participants
BPD daughter, but rather reected generally perception of parental care and overprotection was
dysfunctional PCRs in those families. also compared to their non-BPD sisters perception.
As opposed to comparisons with both parents
perception, the comparisons with siblings
Discussion
perception were generally convergent, BPD
probands and their sisters both reporting lack of
Summary of results in relation to the aims of the
parental care and high parental overprotection.
review
Surprisingly, compared to their non-BPD sisters,
How is the parentchild relationship described by borderline BPD participants reported more maternal care.
personality disorder participants and their parents in Comparisons with participants presenting
comparison to other normative and clinical groups?. DSM-IV Axis-I and Axis-II disorders were a lot
Borderline personality disorder participants and less consistent, but some trends can still be
their parents consistently reported a much more delineated. Compared to parental overprotection,
dysfunctional PCR compared to NC. First, which yielded inconsistent results, lack of parental
compared to community-drawn participants who care seemed to more robustly discriminate BPD
had no mental disorder, BPD participants tended participants from Axis-I participants, with the
to report signicantly lower parental care, higher exception of depressive participants. Parental
parental overprotection and higher parental inconsistency, abuse and neglect also consistently
inconsistency. Second, compared to their other discriminated BPD participants from Axis-I
children who did not develop BPD, parents of participants.
BPD probands retrospectively described their BPD As for comparisons with other Axis-II
child as being unusually sensitive and having what participants (OPD), parental abuse and neglect
is commonly called a difcult temperament as early were globally the most consistent discriminating
as their rst year of life. Parents also reported a variables, BPD participants reporting signicantly
conictual relationship with their BPD child in more frequently different forms of parental abuse
childhood and adolescence: verbal abuse and neglect than OPD. When considering specic
discriminated BPD adolescent daughters from their types of abuse (i.e., physical, sexual, and verbal abuse),
non-BPD sisters, while various violent and only parental verbal abuse seemed to consistently
antisocial behaviours discriminated BPD adolescent discriminate BPD from OPD participants, while
sons from their non-BPD brothers. Third, family- results pertaining to parental physical and sexual
perspective studies revealed that compared to NC abuse were not as consistent. Furthermore, less
daughters, who had a perception similar to their consistent results concerning parental care and
parents, BPD daughters reported less parental care, overprotection can be explained by the type of
more maternal overprotection and inconsistent patients included in the samples (see the next
parental values and norms, while their parents both subsection for further discussion). Overall, these
described themselves in a more normative fashion. results suggest that inpatients with BPD are better
This divergence of perception between daughter discriminated from inpatients with OPD based on
and parents seems to be characteristic of BPD rates of parental abuse and neglect, but not on rates

Copyright 2017 John Wiley & Sons, Ltd. (2017)


DOI: 10.1002/pmh
Parentchild relationship associated with the development of borderline personality disorder: A systematic review

of parental care and overprotection, while The two longitudinal studies reviewed 42,56
outpatients of both groups can be discriminated provide insights as to which aspects of the PCR
well using both types of variables. Alternately, specically predict the risk for BPD diagnosis.
parental inconsistency was reported signicantly Their prospective results suggested that maternal
more frequently by BPD than OPD participants, overprotection might be a risk factor for various
but only in studies with high statistical power. mental disorders, but not for BPD specically.
According to Arens, Grabe 42, the interaction
Which aspects of the parentchild relationship are between temperamental harm avoidance and
specically associated with a borderline personality perceived maternal overprotection was a
disorder diagnosis in adulthood?. The majority of signicant predictor of the risk for BPD when
studies included in the present review evaluated compared to NC ve years later: however, this
parental care and overprotection, mainly from interaction could not discriminate between BPD
the perspective of BPD probands and sometimes and depressive individuals ve years later.
of their parents, or of both dyad members Bezirganian, Cohen 56 reported that it was a
simultaneously. Results on parental care and maternal style made of both overprotection and
overprotection generally conrmed the concept inconsistency that specically predicted BPD
of multinality 18, that is the repeated observation diagnosis two and a half years later, but did not
that a single risk factor (i.e., low parental care or predict any other Axis-I or II diagnosis. Likewise,
high parental overprotection) can be associated parental inconsistency discriminated BPD from
with multiple negative outcomes in the childs OPD in the studies on perspective of BPD
later development. While parental care and probands with high statistical power. This possible
overprotection consistently discriminated BPD specicity of parental inconsistency in the
from NC participants, these two variables appear development of BPD was also suggested by Stepp,
to be related to the severity of psychopathology, Whalen 64 following a literature review on
instead of being specic risk factors of BPD. children of mothers with BPD. According to
Indeed, only studies on the perspective of BPD Stepp and her colleagues, a pattern of
probands with an inpatient-only or a mixed (i.e., oscillations between maternal hostile control
inpatient and outpatient) sample reported no and passivity would be unique to interactions
signicant differences between BPD and OPD between mothers with BPD and their children.
participants for all four PBI subscales scores. Moreover, parental inconsistency could explain
Additionally, studies using other standardized the marked divergence of perception between
measures of parental care and overprotection BPD probands and their parents in family
failed to discriminate between BPD and depressive perspective studies, especially when maternal care
participants. These non-signicant results are in was concerned. Mothers of BPD daughters could
line with recent ndings suggesting that BPD possibly have focused on their overinvolvement
diagnostic criteria may represent a general factor with their child and perceived this behaviour as
of personality pathology and/or core features of an indicator of normative parental care, while
PD severity 63. This new conceptualization of BPD daughters might have focused on their
BPD could explain the less consistent results of mothers tendency to sometimes emotionally
studies comparing BPD participants ratings of disengage and reported low maternal care.
parental care and overprotection to Axis-I or
Axis-II patients: the more severe the level of How can the facets of the parentchild relationship
psychopathology reported by participants, the identied in the reviewed studies shed light on the general
more extreme their ratings of parental care and aetiological models of borderline personality disorder?.
overprotection were. The oldest reviewed studies, conducted in the

Copyright 2017 John Wiley & Sons, Ltd. (2017)


DOI: 10.1002/pmh
Boucher et al.

1980s, tried to verify two opposite hypotheses results from many other studies reviewed in this
regarding the aetiology of BPD: (1) the article, such as Bezirganian, Cohen 56 and
overprotection hypothesis, which posited that the Gunderson and Lyoo 59, are in line with this
mother of the preborderline child could not tolerate general idea of parental contradictory beha-
separation from her child and therefore maintained viours associated with BPD.
an attitude promoting dependency 65,66; (2) the Kernbergs theory on the aetiology of BPD 75
neglect and deprivation hypothesis, which stated was empirically tested in one of the studies
that adults with BPD had received insufcient included 45. Following negative ratings of both
parental love and caring as children 67,68. Empirical parents by BPD participants, Baker and colleagues
support was given to both hypotheses, with slightly concluded that the results imply that borderlines
more support in favour of the neglect and have a greater tendency to view the world in
deprivation hypothesis, which more robustly negative, malevolent ways than to split their
discriminated BPD from clinical and non-clinical object representations (p.258). The concept of
groups compared to overprotection 38,41,47,69. biparental failure2 76 can be easily related to
Torgersen and Alnaes 70 were the only authors Kernbergs concept of malevolence. This similar
conrming the overprotection hypothesis and concept has received indirect empirical support
contradicting the neglect and deprivation by many of the reviewed studies which reported
alternative, but later studies using the PBI and negative ratings of both mother and father by
DSM-based standardized evaluations of PBD again BPD probands, and was explicitly tested and
supported both theories29. As mentioned in the conrmed by one of the studies included 14.
previous subsection, in two prospective studies Linehans biosocial theory 13, which posits that
42,56
, maternal overprotection as rated at present the interaction between a childs highly reactive
time by adolescents or mothers themselves temperament and an invalidating environment
signicantly predicted BPD diagnosis years later. explains the development of BPD, was empirically
Conversely, retrospectively assessed parental tested by only one3 of the studies reviewed 42:
overprotection did not consistently discriminate reporting an anxious temperament and perceived
BPD participants from participants with Axis-I maternal overprotection during adolescence was
and other Axis-II disorders. associated with a 1.7 increased risk of having a
It became more and more evident that a single BPD diagnosis in young adulthood. However, this
psychosocial factor could not account alone for interaction could not discriminate between BPD
the development of BPD. As proposed by Zweig- and depressive individuals ve years later. Indirect
Frank and Paris 71, Masterson and Rinsleys theory support was also given by other studies included in
on the aetiology of BPD 66 could be this review, such as Fletcher, Parker 46 and Joyce,
conceptualized as an alternating pattern between McKenzie 77, which suggested an interaction
parental overprotection toward the childs between specic parental behaviours and specic
dependent behaviours and parental neglect
toward the childs independent behaviours.
Similarly, building on the propositions of Walsh 2
Zanarini, Frankenburg 96 provided the following denition
72
and Melges and Swartz 73, Allen and Farmer of biparental failure: both parents suffering from a psychiatric
74
proposed that parental behavior that alternates disorder that is associated with aspects of poor parental
between the polarities of overinvolvement and functioning or both parents being remembered as having a
disturbed relationship with their preborderline child (p.265).
underinvolvement may serve as a direct trigger 3
It is important to note that two studies 97,98 which tested and
to self-destructive behavior in patients with validated Linehans theory were not included in the present
BPD (p.46). While Allen and Farmers model review, since they recruited participants with BPD symptoms
was explicitly tested in only one study 50, instead of BPD diagnosis.

Copyright 2017 John Wiley & Sons, Ltd. (2017)


DOI: 10.1002/pmh
Parentchild relationship associated with the development of borderline personality disorder: A systematic review

strategies of emotion regulation, and between occur once the BPD child becomes an adult and
specic parental behaviours and specic would trigger self-destructive behaviour. Taking
temperamental traits, respectively, in BPD. A into account the various facets of PCR associated
recent study by Laporte and colleagues 61,62 also with BPD in the studies reviewed, it appears that
underlined the important role of temperamental each polarity of parental inconsistency could take
traits in the development of BPD, those traits several forms: the overinvolvement could
being the main difference between BPD probands comprise either parental overprotection, hostile
and their sisters free of psychopathology, who both control or even marked abusive behaviours, while
reported similar relationships with their parents. the underinvolvement could include parental
Likewise, studies focusing on the perspective of passivity, emotional withdrawal, overt rejection
parents of BPD probands reported that both a or various other marked neglectful behaviours.
childs difcult temperament (i.e., unusual An expanded concept of parental inconsistency
sensitivity, moodiness) and a conictual PCR could thus successfully encompass both
were signicant predictors of BPD diagnosis 57,58. overprotection and neglect/deprivation hypo-
Finally, the mentalization decit theory 12,78, theses and represent a parsimonious proposal for
which states that BPD patients would present describing the relational familial adversity
marked impairments in the ability to think about experienced by BPD patients.
mental states in themselves and others in
emotionally intense relationships contexts, was
Theoretical and clinical implications
also explicitly tested in only one of the studies
reviewed 43. Although BPD participants did not The ndings reviewed previously have various
have poorer mentalizing abilities compared to theoretical implications. First, there seems to be
NC, signicant negative correlations were found specic and non-specic factors associated with
in BPD between mentalizing abilities and maternal the risk for BPD. On the one hand, ratings of
overprotection, lack of emotional availability and parental care and overprotection might be
rejection. Ghiassi, Dimaggio 43 therefore concluded associated with psychopathology severity and
that the quality of parental care during early represent a general risk factor for psychopathology,
childhood plays a role in the development of instead of being specic relationship factors
mentalizing skills in BPD (p.657). leading to BPD. The many reviewed studies which
In sum, ndings of the present review explicitly measured these two variables reported ndings in
highlight the various ways the PCR can be line with the concept of multinality 18:
adversely experienced by BPD probands and their participants with various mental disorders
parents. Indeed, the aetiological models reviewed frequently reported similar ratings of low parental
in this section refer to diverse manifestations of care and high parental overprotection. On the
relational adversity experienced in the PCR, as other hand, parental inconsistency appears to be
indicated by the signicant associations between a specic risk factor for BPD. As discussed
BPD and the PCR variables reviewed. However, previously, this particular aspect of the PCR was
relational adversity per se appears to be previously identied as being characteristic of
insufciently specic for the understanding of mothers with BPD 64 and as a trigger of self-
BPD development. Aetiological models of BPD destructive behaviour in BPD. 74. Further studies
should consider parental inconsistency as a are however necessary to conrm this possible
potential specic risk factor for BPD. This concept specic factor.
was dened by Allen and Farmer 74 as an Second, ndings of this review are at least
alternation between parental overinvolvement partially in line with contemporary theories which
and underinvolvement, which would continue to posit a geneenvironment interaction in the

Copyright 2017 John Wiley & Sons, Ltd. (2017)


DOI: 10.1002/pmh
Boucher et al.

development of BPD2,12,13. More specically, this self-regulate should also be provided (see 64, for
systematic review mainly highlights, by choice, a detailed discussion of such intervention
the relational adversity experienced by BPD targets).
probands and their parents in the PCR. Second, ndings from many studies reviewed
Additionally, studies which measured the earlier suggest that verbal abuse characterizes
perspective of parents and family showed the interactions between preborderline adolescents
intervention of some temperamental traits, such and their parents, as both dyad members have
as emotion dysregulation and impulsivity, in the described each other as verbally abusive. This
aetiology of BPD. Laporte and colleagues papers is to us a robust nding which clinicians might
61,62
showed very clearly that both biological nd useful to keep in mind when treating
(i.e., temperamental traits) and social (i.e., familial either member of the parentchild dyad.
environment perceived as abusive and uncaring) Professional intervention could be aimed at
factors are necessary for a young woman to develop improving verbal communication between
BPD and that these factors interact with each other. adolescents with BPD symptoms and their
These results are also in line with familial and twin parents and could therefore contribute to
studies, which taken together, suggest that the minimize the detrimental effects of verbal abuse
heritability for BPD is approximately 40% (for a on self-esteem. The gradual deterioration of the
systematic review, see 79). Amad, Ramoz 79 parentadolescent relationship could also be
proposed a paradigmatic shift, where genes would slowed down and even stopped given pro-
be understood as inuencing personal reactivity to fessional help is offered.
ones environment, good or bad, (plasticity genes) A third clinical implication of this systematic
instead of making one vulnerable only to negative review is the importance of subjective reality.
outcomes (vulnerability genes). This interesting Simply put, one of the most frequently reproduced
hypothesis gives hope regarding help that might results across all reviewed studies was that BPD
be provided to persons with plasticity genes, who daughters reported having lacked parental positive
may then react signicantly more positively to attention and love, while their parents did not
adequate parental care. This hypothesis is also in describe their behaviours in such extreme terms.
line with ideas from research on resilience and As mentioned earlier, this marked disagreement
differential susceptibility theory 80. remains to be claried. Studies on parental abuse
As for clinical implications, the main ndings and neglect indicated similarly that the types of
of this review highlight the relational adversity parental neglect which discriminated BPD from
experienced by BPD participants and their parents participants with OPD all referred to the parental
in the PCR. This relational adversity took several inability in meeting their childs emotional
forms (low care, high overprotection, inco- needs, but not physical needs. Clinicians should
nsistency, abuse, neglect). A rst clinical always keep in mind this perceived lack of
implication resides in prevention, such as the parental love in BPD patients and reect on its
fostering of preventive parental coaching. Parents impact in the therapeutic relation. Frequent
of high-risk children (i.e., with a marked reactions of clinicians working with BPD
emotional vulnerability) should be provided patients, such as an alternating desire to both
professional help fostering positive PCR. More save and reject the patient, could easily be
specically, these parents should be actively explained by this perceived lack of parental love,
coached and reinforced in enacting caring, which could be reenacted in the therapeutic
autonomy-supportive and coherent behaviours relationship.
toward their child; psychoeducation about A fourth clinical implication of this review is
emotion regulation and ways to help their child the suggestion that a corrective emotional

Copyright 2017 John Wiley & Sons, Ltd. (2017)


DOI: 10.1002/pmh
Parentchild relationship associated with the development of borderline personality disorder: A systematic review

relational experience should be necessary in order numerous industrialized countries. However,


to improve these patients global functioning and perspective of parents of BPD participants and
symptomatology by modifying the internalized perspective of family studies were a lot less
past relational adversity which is reenacted with numerous and took place only in Canada and
actual signicant others. Because relational the United States of America.
adversity experienced by BPD patients can take Furthermore, the fact that most of the results
several forms, such corrective emotional reviewed were based on retrospective self-reports
relational experience implies non-specic factors of BPD patients should not be seen as a major
and can be provided through different techniques shortcoming of this systematic review. Even if
and theoretical principles. The non-specicity of some authors have criticized the existence of a
this corrective experience is also reected by the retrospective bias of current symptoms, several
fact that many therapeutic models (transference- studies have demonstrated that retrospective self-
focused therapy, mentalization-based therapy, reports of PCRs are relatively stable over time
dialectical behavioural therapy, etc.) with and independent of mood variations (for a review,
different relational positions have been validated see 87), as much in depressive adult inpatients 88 as
for effectively treating BPD (see 81, for a recent in adolescents from the community 89. Brewin,
meta-analysis). Moreover, ndings of this Andrews 87 also concluded that patients
systematic review are in line with actual treatment memories are in as much agreement with external
recommendations for BPD, which point to criteria as are controls, whether the criteria be
psychotherapy as primary treatment (a relational siblings memories or independent records
mode of intervention that is) and (p.91). These results are in line with the ones from
pharmacotherapy as an adjunctive time-limited Laporte and colleagues study 61,62, where BPD
treatment for specic symptoms 82,83. Many patients and their sisters free of psychopathology
recent papers on the treatment of BPD 8486 reported similar ratings of parental care and
proposed that such treatment be based on control. However, other studies demonstrated a
common factors instead of specic theories and small but signicant impact of depressive mood
techniques, one of those common factors being on retrospective self-reports of PCR (i.e., 90). In
a strong alliance between BPD patients and their brief, retrospective self-reports of PCR can be
therapist. considered as carrying some validity: small
variations over time can be observed on PCR
questionnaires such as the PBI and the EMBU,
Strengths and limitations of the present review
but scores still remain globally stable and do not
A major strength of this review on PCR in BPD is change as much as the patients symptoms do.
that a systematic methodology was applied. All Parentchild relationship retrospective reports
titles and abstracts were reviewed and sorted therefore appear to have to a stable core. For
independently by two researchers, and any example, patients who reported an abusive or
discrepancies in the sorting process were discussed overprotective PCR at one point in time might
and resolved. Moreover, this reviews focus on the reinterpret and modify slightly their report over
PCR in BPD is innovative and makes it possible to time, but not to the extent of describing their
shed light on contemporary aetiological theories parents as non-abusive or supporting their
of BPD. Additionally, the 33 perspective of BPD autonomy as children.
probands studies were realized in 11 different Another reason for which one could question
countries and yet frequently reached similar the validity of PCR retrospective self-reports is
conclusions. Results pertaining to BPD probands the fact that Zweig-Frank and Paris 39 reported
perception can therefore be generalized to PBI scores of older BPD patients (i.e., mean age

Copyright 2017 John Wiley & Sons, Ltd. (2017)


DOI: 10.1002/pmh
Boucher et al.

of 50.9 years, SD = 7.6) signicantly higher for ratings of maternal care compared to parents of
parental care and lower for parental overprotection AN and NC daughters, while BPD daughters
than those of all other reviewed studies, where reported levels of maternal care signicantly lower
participants mean ages spanned between 20 and than those of daughters of various other clinical
35 years. Those contrasted results could be groups and than their own parents. This parent
explained by the possible idealization of deceased daughter disagreement over parental care in BPD
parents, demonstrated by Richter, Eisemann 91. It families seems to be only about maternal care,
is quite probable that many participants in Zweig- because parents of BPD daughters reported lower
Frank and Pariss study 39 had a least one parent paternal care compared to parents of other types
deceased when they lled out the PBI. Moreover, of families. Another contradictory result regarding
Zweig-Frank and Paris 39 did not measure PBI maternal care was found in studies comparing BPD
ratings at an earlier point and therefore one probands perception to their non-BPD sisters.
cannot know what the perception of those Higher maternal care was reported on the PBI by
participants was of their PCR when they were BPD probands compared to their sisters: this is
20 or 30 years old. Other biases could have quite surprising considering BPD probands
affected Zweig-Frank and Paris longitudinal consistently reported lower maternal care
study 39: the remaining participants who lled compared to NC in perspective of BPD probands
out the PBI could have had initially better studies. On another questionnaire, similar
relationships with their parents compared to relationships with parents were reported by both
participants who did not complete the 27-year groups of sisters. In summary, taking the reviewed
follow-up. Similarly, these remaining participants ndings as a whole, it is difcult to describe the
could have concretely improved their PCR globally and coherently. The exact parent
relationships with their parents over the long child dynamic involving parental care remains
course of the study. In brief, very little evidence unclear, which could be explained by the fact that
suggests that age could affect systematically PCR very few studies measured simultaneously the
retrospective self-reports. perspective of both BPD probands and their
Several limitations can be identied due to the parents.
many exclusion criteria dened earlier. Thus,
many relevant studies were excluded due to their
Quality assessment of included studies
dimensional evaluation of BPD. Likewise, cross-
sectional studies on children or adolescents with More than half of the reviewed studies used
BPD and all studies on mothers with BPD and validated instruments, such as the PBI and the
their own children were excluded. Results of the EMBU, which is certainly a methodological
present review can therefore only be generalized strength. However, only two of the studies
to the relationship between adults/late adolescents included had a longitudinal design. Consequently,
with BPD (evaluated categorically) and their very little is known about the predictive power of
parents. Additionally, some results were reported PCR in the later development of BPD. Being
by a single study only: these results should based on almost exclusively cross-sectional studies
therefore be interpreted with great caution. using adult samples, the present review can only
A nal limitation is the heterogeneity of the condently report on a signicant contemporary
variables measured, of the instruments used and relationship between those two variables. Finally,
of the perspectives taken by the studies reviewed. ve family-perspective studies were reviewed, and
Some contradictory ndings also make it difcult it is quite possible that only family members who
to integrate the results into a general model. More still had regular contact or at least who had not
specically, mothers and fathers reported similar completely broken off all ties with one another

Copyright 2017 John Wiley & Sons, Ltd. (2017)


DOI: 10.1002/pmh
Parentchild relationship associated with the development of borderline personality disorder: A systematic review

accepted to participate. Consequently, results of develop BPD and which participants do not, or
those studies cannot in all likelihood be develop other mental disorders and which
generalized to all BPD families. specic factors are related to those diverse
trajectories. Also, in line with the concept of
subjective reality, being able to compare
Recommendations for future research
measures of PCR collected in childhood to the
First, as suggested by Brewin, Andrews 87, we same measures collected retrospectively in
recommend that future studies on PCR in BPD adulthood would be very interesting and
include multiple informants as much as possible. informative. Finally, longitudinal studies could
The richness of this methodology has been provide insight regarding the specic predictive
demonstrated in earlier sections of this review, role of parental inconsistency in the risk for BPD.
and recruiting only one member of the parent
child dyad can be justied only in dyads where
one member has broken all ties with the other. Acknowledgements
New studies including the perspective of family
members are necessary to obtain a more global This study, which was part of the rst authors
picture of the PCR. Qualitative research could doctoral research, was supported by a doctoral
be especially useful to gather a detailed description scholarship from the Fonds de recherche du
of the subjective experience of each member of Qubec-Socit et Culture.
the parentchild dyad using the participants
own words instead of predened concepts. Such
studies would allow comparisons between the References
childs perception and the parents perception 1. Cole PM, Llera SJ, & Pemberton CK. Emotional
and help determine which parts of their subjective instability, poor emotional awareness, and the
experience are convergent and divergent. development of borderline personality. Dev Psychopathol
2009; 21(4): 1293310.
Second, more studies on PCRs in men with 2. Hughes AE, Crowell SE, Uyeji L, & Coan JA. A
BPD are needed, all the more so as prevalence developmental neuroscience of borderline pathology:
studies in large community-drawn samples Emotion dysregulation and social baseline theory. J
revealed that BPD was diagnosed as frequently in Abnorm Child Psychol 2012; 40(1): 2133.
males as in females 92,93. The few included studies 3. Paris J. Traumatisme et trouble de la personnalit limite.
Sante Ment Que 1996; 21(1): 17787.
with an all-male sample suggested that there could
4. Sabo AN. Etiological signicance of associations
be specic aspects of PCR in men with BPD, such between childhood trauma and borderline personality
as an abusive fatherchild relationship, and disorder: Conceptual and clinical implications. J Pers
developmental factors different from women with Disord 1997; 11(1): 5070.
BPD, such as excessive anxiety separation in 5. Goodman M, & Yehuda R. The relationship between
infancy and toddlerhood. psychological trauma and borderline personality disorder.
Psychiatr Ann 2002; 32(6): 33745.
Third, there is an important need for more 6. Graybar SR, & Boutilier LR. Nontraumatic pathways to
longitudinal studies in order to dene more clearly borderline personality disorder. Psychother Theory Res
the predictive role of specic aspects of PCR in the Pract Train 2002; 39(2): 15262.
later development of BPD in adulthood. 7. Rind B, Tromovitch P, & Bauserman R. A meta-
Longitudinal studies on high-risk samples (i.e., 94) analytic examination of assumed properties of child
sexual abuse using college samples. Psychol Bull 1998;
and on a birth cohort of twins 95 followed from
124(1): 2253.
childhood to adolescence have already been 8. Agrawal HR, Gunderson J, Holmes BM, & Lyons-Ruth
published, but these samples need to be followed K. Attachment studies with borderline patients: A
past adolescence to verify which participants review. Harv Rev Psychiatry 2004; 12(2): 94104.

Copyright 2017 John Wiley & Sons, Ltd. (2017)


DOI: 10.1002/pmh
Boucher et al.

9. Barone L. Developmental protective and risk factors in psychosocial risk factors for understanding of borderline
borderline personality disorder: A study using the Adult personality disorder. Psychoanal Psychother 2012; 26(1):
Attachment Interview. Attach Hum Dev 2003; 5(1): 6591.
6477. 25. Guttman HA, & Laporte L. Family members retrospective
10. Choi-Kain LW, Fitzmaurice GM, Zanarini MC, perceptions of intrafamilial relationships. Contemporary
Laverdiere O, & Gunderson JG. The relationship Family Therapy: An Int J 2002; 24(3): 50521.
between self-reported attachment styles, interpersonal 26. Hernandez A, Arntz A, Gaviria AM, Labad A, &
dysfunction, and borderline personality disorder. J Nerv Gutierrez-Zotes JA. Relationships between childhood
Ment Dis 2009; 197(11): 81621. maltreatment, parenting style, and borderline personality
11. Holmes J. Disorganized attachment and borderline disorder criteria. J Pers Disord 2012; 26(5): 72736.
personality disorder: A clinical perspective. Attach Hum 27. Katerndahl D, Burge S, & Kellogg N. Predictors of
Dev 2004; 6(2): 18190. development of adult psychopathology in female victims
12. Fonagy P, Target M, Gergely G, Allen JG, & of childhood sexual abuse. J Nerv Ment Dis 2005; 193(4):
Bateman AW. The developmental roots of borderline 25864.
personality disorder in early attachment relationships: 28. Laporte L, & Guttman H. Recollections of parental
A theory and some evidence. Psychoanal Inq 2003; bonding among women with borderline personality
23(3): 41259. disorder as compared with women with anorexia nervosa
13. Linehan MM. Cognitive-behavioral treatment of borderline and a control group. Aust J Psychol 2007; 59(3): 1329.
personality disorder. New York, NY: Guilford Press, 1993. 29. Machizawa-Summers S. Childhood trauma and parental
14. Zanarini MC, Frankenburg FR, Reich D, Marino MF, bonding among Japanese female patients with borderline
Lewis RE, Williams AA et al. Biparental failure in the personality disorder. Int J Psychol 2007; 42(4): 26573.
childhood experiences of borderline patients. J Pers 30. Paris J, Zweig-Frank H, & Guzder H. The role of
Disord 2000; 14(3): 26473. psychological risk factors in recovery from borderline
15. Gunderson JG, & Lyons-Ruth K. BPDs interpersonal personality disorder. Compr Psychiatry 1993; 34(6):
hypersensitivity phenotype: A geneenvironment 41013.
developmental model. J Pers Disord 2008; 22(1): 2241. 31. Paris J, Zweig-Frank H, & Guzder J. Psychological risk
16. Masten AS. Developmental psychopathology: Pathways factors for borderline personality disorder in female
to the future. Int J Behav Dev 2006; 30(1): 4754. patients. Compr Psychiatry 1994; 35(4): 3015.
17. Rutter M, & Sroufe L. Developmental psychopathology: 32. Paris J, Zweig-Frank H, & Guzder J. Risk factors for
Concepts and challenges. Dev Psychopathol 2000; 12(3): borderline personality in male outpatients. J Nerv Ment
26596. Dis 1994; 182(7): 37580.
18. Cicchetti D, & Rogosch FA. Equinality and 33. Timmerman IG, & Emmelkamp PM. Parental rearing
multinality in developmental psychopathology. Dev styles and personality disorders in prisoners and forensic
Psychopathol 1996; 8(4): 597600. patients. Clin Psychol Psychother 2005; 12(3): 191200.
19. Johnson JG, Cohen P, Chen H, Kasen S, & Brook JS. 34. Byrne C, Velamoor V, Cernovsky Z, Cortese L et al. A
Parenting behaviors associated with risk for offspring comparison of borderline and schizophrenic patients for
personality disorder during adulthood. Arch Gen childhood life events and parentchild relationships.
Psychiatry 2006; 63(5): 57987. Can. J. Psychiatry/La Rev. Can. Psychiatr. 1990; 35(7):
20. Gagnon P. The role of the family in the development of 5905.
borderline personality disorder (BPD). Can. J. 35. Fossati A, Donati D, Donini M, Novella L, Bagnato M,
Psychiatry/La Rev. Can. Psychiatr. 1993; 38(9): 61116. & Maffei C. Temperament, character, and attachment
21. Madeddu F, Visintini R, Fossati A, & Maffei C. patterns in borderline personality disorder. J Pers Disord
Borderline personality and family factors: An overview 2001; 15(5): 390402.
of empirical literature. New Trends Exp Clin Psychiat 36. Helgeland MI, & Torgersen S. Maternal representations
1991; 7(4): 15168. of patients with schizophrenia as measured by the
22. Gunderson JG, & Englund DW. Characterizing the Parental Bonding Instrument. Scand J Psychol 1997;
families of borderlines. A review of the literature. 38(1): 3943.
Psychiatr. Clin. N. Am. 1981; 4(1): 15968. 37. Torgersen S, & Alnaes R. Differential perception of
23. Zanarini MC. Childhood experiences associated with the parental bonding in schizotypal and borderline
development of borderline personality disorder. Psychiatr personality disorder patients. Compr Psychiatry 1992;
Clin N Am 2000; 23(1): 89101. 33(1): 3438.
24. Keinanen MT, Johnson JG, Richards ES, & Courtney 38. Zweig-Frank H, & Paris J. Parents emotional neglect and
EA. A systematic review of the evidence-based overprotection according to the recollections of patients

Copyright 2017 John Wiley & Sons, Ltd. (2017)


DOI: 10.1002/pmh
Parentchild relationship associated with the development of borderline personality disorder: A systematic review

with borderline personality disorder. Am J Psychiatry borderline patients. Compr Psychiatry 1989; 30(1):
1991; 148(5): 64851. 1825.
39. Zweig-Frank H, & Paris J. Predictors of outcome in a 27- 52. Zanarini MC, Williams AA, Lewis RE, Reich R, Vera
year follow-up of patients with borderline personality SC, Marino MF et al. Reported pathological childhood
disorder. Compr Psychiatry 2002; 43(2): 1037. experiences associated with the development of
40. Schwarze CE, Hellhammer DH, Stroehle V, Lieb K, & borderline personality disorder. Am J Psychiatry 1997;
Mobascher A. Lack of breastfeeding: A potential risk 154(8): 11016.
factor in the multifactorial genesis of borderline 53. Dubo ED, Zanarini MC, Lewis RE, & Williams AA.
personality disorder and impaired maternal bonding. J Childhood antecedents of self-destructivenes in
Pers Disord 2015; 29(5): 61026. borderline personality disorder. Can J Psychiatry/La Rev
41. Goldberg RL, Mann LS, Wise TN, & Segall EA. Can Psychiatr 1997; 42(1): 6369.
Parental qualities as perceived by borderline 54. Heffernan K, & Cloitre M. A comparison of
personality disorders. Hillside J Clin Psychiatry 1985; posttraumatic stress disorder with and without borderline
7(2): 13440. personality disorder among women with a history of
42. Arens EA, Grabe HJ, Spitzer C, & Barnow S. Testing the childhood sexual abuse: Etiological and clinical
biosocial model of borderline personality disorder: characteristics. J Nerv Ment Dis 2000; 188(9): 58995.
Results of a prospective 5-year longitudinal study. 55. Merza K, Papp G, & Szab IK. The role of childhood
Personal Ment Health 2011; 5(1): 2942. traumatization in the development of borderline
43. Ghiassi V, Dimaggio G, & Brune M. Dysfunctions in personality disorder in Hungary. Eur J Psychiatr 2015;
understanding other minds in borderline personality 29(2): 10518.
disorder: A study using cartoon picture stories. Psychother 56. Bezirganian S, Cohen P, & Brook JS. The impact of
Res 2010; 20(6): 65767. motherchild interaction on the development of
44. Huang J, Napolitano LA, Wu J, Yang Y, Xi Y, Li Y et al. borderline personality disorder. Am J Psychiatry 1993;
Childhood experiences of parental rearing patterns 150(12): 183642.
reported by Chinese patients with borderline personality 57. Goodman M, Patel U, Oakes A, Matho A, &
disorder. Int J Psychol 2014; 49(1): 3845. Triebwasser J. Developmental trajectories to male
45. Baker L, Silk KR, Westen D, Nigg JT, & Lohr NE. borderline personality disorder. J Pers Disord 2013;
Malevolence, splitting, and parental ratings by 27(6): 76482.
borderlines. J Nerv Ment Dis 1992; 180(4): 25864. 58. Goodman M, Patil U, Triebwasser J, Diamond E, Hiller
46. Fletcher K, Parker G, Bayes A, Paterson A, & McClure A, Hoffman P et al. Parental viewpoints of trajectories
G. Emotion regulation strategies in bipolar II disorder to borderline personality disorder in female offspring. J
and borderline personality disorder: Differences and Pers Disord 2010; 24(2): 20416.
relationships with perceived parental style. J Affect Disord 59. Gunderson JG, & Lyoo IK. Family problems and
2014; 157: 5259. relationships for adults with borderline personality
47. Frank H, & Hoffman N. Borderline empathy: An disorder. Harv Rev Psychiatry 1997; 4(5): 2728.
empirical investigation. Compr Psychiatry 1986; 27(4): 60. Guttman HA, & Laporte L. Empathy in families of
38795. women with borderline personality disorder, anorexia
48. Bandelow B, Krause J, Wedekind D, Broocks A, Hajak nervosa, and a control group. Fam Process 2000; 39(3):
G, & Ruther E. Early traumatic life events, parental 34558.
attitudes, family history, and birth risk factors in patients 61. Laporte L, Paris J, Guttman H, & Russell J.
with borderline personality disorder and healthy Psychopathology, childhood trauma, and personality
controls. Psychiatry Res 2005; 134(2): 16979. traits in patients with borderline personality disorder
49. Hayashi N, Suzuki R, & Yamamoto N. Parental and their sisters. J Pers Disord 2011; 25(4): 44862.
perceptions of borderline personality disorders in video- 62. Laporte L, Paris J, Guttman H, Russell J, & Correa JA.
recorded interviews. Psychiatry Clin Neurosci 1995; Using a sibling design to compare childhood adversities
49(1): 3537. in female patients with BPD and their sisters. Child
50. Allen DM, Abramson H, Whitson S, Al-Taher M, Maltreat 2012; 17(4): 31829.
Morgan S, Veneracion-Yumul A et al. Perceptions of 63. Sharp C, Wright AGC, Fowler JC, Frueh BC, Allen JG,
contradictory communication from parental gures by Oldham J et al. The structure of personality pathology:
adults with borderline personality disorder: A preliminary Both general (g) and specic (s) factors? J Abnorm
study. Compr Psychiatry 2005; 46(5): 34052. Psychol 2015; 124(2): 38798.
51. Zanarini MC, Gunderson JG, Marino MF, Schwartz 64. Stepp SD, Whalen DJ, Pilkonis PA, Hipwell AE, &
EO, & Frankenburg FR. Childhood experiences of Levine MD. Children of mothers with borderline

Copyright 2017 John Wiley & Sons, Ltd. (2017)


DOI: 10.1002/pmh
Boucher et al.

personality disorder: Identifying parenting behaviors as 80. Rutter M. Resilience as a dynamic concept. Dev
potential targets for intervention. Personality Disorders: Psychopathol 2012; 24(2): 33544.
Theory, Research, and Treatment 2012; 3(1): 7691. 81. Cristea IA, Gentili C, Cotet CD, Palomba D, Barbui C,
65. Levy D. Maternal overprotection, p. 417. New York: & Cuijpers P. Efcacy of psychotherapies for borderline
Columbia University Press, 1943. personality disorder: A systematic review and meta-
66. Masterson JF, & Rinsley DB. The borderline syndrome: analysis. JAMA Psychiat 2017; 74(4): 31928.
The role of the mother in the genesis and psychic 82. Paris J. The treatment of borderline personality
structure of the borderline personality. Int J Psychoanal disorder: Implications of research on diagnosis,
1975; 56(2): 16377. etiology, and outcome. Annu Rev Clin Psychol 2009;
67. Guntrip H. Schizoid phenomena, object relations and the 5: 27790.
self, p. 437. Oxford, England: International Universities 83. Bateman AW, Gunderson J, & Mulder R. Treatment
Press, 1968. of personality disorder. The Lancet 2015; 385(9969):
68. Adler G. Borderline psychopathology and its treatment. New 73543.
York: Jason Aronson, 1985. 84. Livesley W. Moving beyond specialized therapies for
69. Frank H, & Paris J. Recollections of family experience in borderline personality disorder: The importance of
borderline patients. Arch Gen Psychiatry 1981; 38(9): integrated domain-focused treatment. Psychodynamic
10314. Psychiatry 2012; 40(1): 4774.
70. Torgersen S, & Alnaes R. Differential perception of 85. Paris J. Applying the principles of psychotherapy
parental bonding in schizotypal and borderline integration to the treatment of borderline personality
personality disorder patients. Compr Psychiatry 1992; disorder. J Psychother Integr 2015; 25(1): 139.
33(1): 3438. 86. McMain SF, Boritz TZ, & Leybman MJ. Common
71. Zweig-Frank H, & Paris J. Parents emotional neglect and strategies for cultivating a positive therapy relationship
overprotection according to the recollections of patients in the treatment of borderline personality disorder.
with borderline personality disorder.[Erratum appears in J. Psychother. Integr. 2015; 25(1): 2029.
Am J Psychiatry 1991 Sep;148(9):1282]. Am J Psychiatry 87. Brewin CR, Andrews B, & Gotlib IH. Psychopathology
1991; 148(5): 64851. and early experience: A reappraisal of retrospective
72. Walsh F. Family study 1976: 14 borderline cases. In: reports. Psychol Bull 1993; 113(1): 8298.
Grinker RR, Sr, & Werble B (eds). The borderline patient, 88. Richter J, & Eisemann M. Stability of memories of
pp. 15877. New York: Aronson, 1977. parental rearing among psychiatric inpatients. Clin
73. Melges FT, & Swartz MS. Oscillations of attachment in Psychol Psychother 2000; 7(4): 25661.
borderline personality disorder. Am J Psychiatry 1989; 89. Asselmann E, Knappe S, Wittchen HU, Lieb R, &
146(9): 111520. Beesdo-Baum K. Stability of recalled parental
74. Allen DM, & Farmer RG. Family relationships of adults rearing behavior in a community sample of
with borderline personality disorder. Compr Psychiatry adolescents and young adults. J Res Adolesc 2015;
1996; 37(1): 4351. 25(4): 61421.
75. Kernberg O. Borderline personality organization. J Am 90. Gillham JE, Putter P, & Kash VM. The effects of sad
Psychoanal Assoc 1967; 15(3): 64185. mood on reports of parents caregiving behaviors. J
76. Frank H, & Paris J. Recollections of family experience in Psychopathol Behav Assess 2007; 29(2): 12533.
borderline patients. Arch Gen Psychiatry 1981; 38(9): 91. Richter J, Eisemann M, Richter G, & Perris C.
10314. Determinants of recall of parental rearing behavior:
77. Joyce PR, McKenzie JM, Luty SE, Mulder RT, Carter JD, The inuence of age or loss of parents by separation or
Sullivan PF et al. Temperament, childhood environment death. Psychopathology 1992; 25(3): 1207.
and psychopathology as risk factors for avoidant and 92. Grant BF, Chou S, Goldstein RB, Huang B, Stinson FS,
borderline personality disorders. Aust New Zeal J Psychiatr Saha TD et al. Prevalence, correlates, disability, and
2003; 37(6): 75664. comorbidity of DSM-IV borderline personality disorder:
78. Fonagy P, & Luyten P. A developmental, mentalization- Results from the Wave 2 National Epidemiologic Survey
based approach to the understanding and treatment of on Alcohol and Related Conditions. J Clin Psychiatry
borderline personality disorder. Dev Psychopathol 2009; 2008; 69(4): 53345.
21(4): 135581. 93. Zanarini MC, Horwood J, Wolke D, Waylen A,
79. Amad A, Ramoz N, Thomas P, Jardri R, & Gorwood P. Fitzmaurice G, & Grant BF. Prevalence of DSM-IV
Genetics of borderline personality disorder: Systematic borderline personality disorder in two community
review and proposal of an integrative model. Neurosci samples: 6,300 English 11-year-olds and 34,653
Biobehav Rev 2014; 40: 619. American adults. J Pers Disord 2011; 25(5): 60719.

Copyright 2017 John Wiley & Sons, Ltd. (2017)


DOI: 10.1002/pmh
Parentchild relationship associated with the development of borderline personality disorder: A systematic review

94. Stepp SD, Scott LN, Jones NP, Whalen DJ, & Hipwell 98. Sauer SE, & Baer RA. Validation of measures of
AE. Negative emotional reactivity as a marker of biosocial precursors to borderline personality disorder:
vulnerability in the development of borderline personality Childhood emotional vulnerability and environmental
disorder symptoms. Dev Psychopathol 2016; 28(1): 21324. invalidation. Assessment 2010; 17(4): 45466.
95. Belsky DW, Caspi A, Arseneault L, Bleidorn W, Fonagy P,
Goodman M et al. Etiological features of borderline Supporting information
personality related characteristics in a birth cohort of 12-
year-old children. Dev Psychopathol 2012; 24(1): 25165. Additional supporting information may be found in the
96. Zanarini MC, Frankenburg FR, Reich DB, Marino MF, online version of this article at the publishers web site.
Lewis RE, Williams AA et al. Biparental failure in the
childhood experiences of borderline patients. J Pers
Disord 2000; 14(3): 26473. Address correspondence to: Marie-ve Boucher,
97. Robertson CD, Kimbrel NA, & Nelson-Gray RO. The Dpartement de psychologie, Universit de
Invalidating Childhood Environment Scale (ICES):
Psychometric properties and relationship to borderline
Montral, C.P. 6128, succursale Centre-Ville,
personality symptomatology. J Pers Disord 2013; 27(3): Montral, Qubec, Canada, H3C 3J7. Email:
40210. marie-eve.do.boucher@umontreal.ca

Copyright 2017 John Wiley & Sons, Ltd. (2017)


DOI: 10.1002/pmh