The Impact of Partner Support in the Treatment of Postpartum

Depression
ShaUa Misri, MD, FRCPC, Xanthoula Kostaras, BSc^, Don Fox, MSc^, Demetra Kostaras, BSc, MA"

Objective: To determine ihe itnpaci of partner support in ihc treaimcni of mothers suffering from postpartum depression (PPD).
Method: Patients underweni a comprehensive psychiatric assessmcni and were enrolled in the siudy only if ihey mci the DSM-l V criieriafor major depressive disorder with posiparium onsei. Patienis with PPD (n = 29) were assigned randomly to 2 treatment groups:
group I (control group) consisiedofpaiienis only (n = li). while group 2 (support group) consisicdofpaiicnis (n = 16) and iheirpariners. The patienis in boih groups were seenfor 7 psychocdticationai visiis each. In group 2. partners participated in 4 of ihe 7 visiis. Paiienis in boih groups were adminisiered a sei of questionnaires ihai included ihe Edinburgh Posinaial Depression Scale (EPDS). ihe
Kellner Symptom Questionnaire, the Dyadic Adjttstment Scale (DAS), and the Parental Bonding Insirument (PBl). In addiiion. during
visits 1 and 7. all patients underwent assessment using ihe Mini Iniernaiional Neuropsychiairic Insirumeni (MINI), seciion A (major depressive episode). The pariners in boih groups complcicd ihe DAS and ihe General Health Questionnaire (GHQ).
Results: Relative io ihe conirol-grottp paiienis. ihe .nippori-group patienis displayed a significani decrease in depressive sympioms and
other psychiatric conditions. Relative io ihe suppori group, ihe general healih of ihe pariners in the control group deterioraied.
Conclusion: Pariner support has a measurable effect on women experiencing PPD.

(Can J Psychiatry 2000;45:554-558)

Key Words: pregnancy, postpartum depression, partner support
the first 6 postpartum weeks, and, in 60% of the cases, they
represent the woman's first episode of depression.

ostpartum depression (PPD) is a complex and challenging disorder that often takes a woman and her family by
surprise. At a time normally marked by joy and happiness, the
onset of depression can cause devastation and dissension in a
woman's life. The consequences of depression during the
postpartum period are considerably more deleterious than at
another time because a woman faces the added responsibility
of caring for her newborn infant. The demands of infant care
alone are new and require time for adjustment. For a woman
suffering from PPD, the whole experience of motherhood becomes overwhelming (1,2).

After the birth, the husband or partner also goes through an

adjustment period. For some couples, a baby strengthens the
relationship, whieh leads to a supportive milieu for the infant's growth. The partner's role becomes even more crucial
if the mother is afflicted with a major mood disorder. A stable
marital relationship helps new parents adapt to the competing
demands of tnarriage, infant, and family. In contrast, studies
have shown that a poor marital relationship is the most consistent psychosocial predictor of PPD (11,12), and a review of
antenatal psychosocial risk factors and adverse postpartum
outcomes revealed that PPD was most strongly associated
with poor marital adjustment, recent life stressors, and antepartum depression (10).

PPD can manifest itself in a variety of ways. Superimposed

on the typical symptoms of depression are such conditions as
intruding thoughts of hanning the baby (3), psychotic depression that may lead tragically to infanticide (4), and pervasive
thoughts of suicide (5). The prevalence of depression during
the postpartum period is estimated to be between 10% and
28% (6-10). Typically, depressive symptoms appear within

Many women experience a great deal of stress in atteinpting

to handle both matemal and marital roles (13,14). The presence of depression in the mother can cause significant complications in both marital and family relationships (15). In
addition, a troubled environtnent can heighten existing depression during the postpartum period (16). In one study, researchers investigated 71 pregnant women to identify the
influence of marital adjustment on maternal depressive
symptoms (16). The participants were followed at 6,9, and 12
months postpartum. Results support the notion that a disruption in the marital relationship is a predictor of matemal PPD.
Further, it was found that major psychosocial stresses occurred in postpartutn marital adjustment when partners were
not supportive and were not involved in child-rearing (11,17).

Manuscript received October 1999 and accepted Aprii 2000.

'Ciinicai Professor, Departments of Psychiatry and Obstetrics/Gynaecology, University of British Columbia; Codirector, Reproductive Mental
Heahh Programs, St. Paul's Hospital and BC Women's Hospital and Health
Centre, Vancouver, British Columbia.
^Research Assistant, Reproductive Mental Health Program, St. Paul's Hospital, Vancouver, British Columbia.
^Statistician, Vancouver, British Columbia.
''Research Assistant, Reproductive Mental Health Program, St. Paul's Hospital, Vancouver, British Columbia.
Address for Correspondence: Dr S Misri, Reproductive Mental Health
Program, St. Paul's Hospital, 2B-250 1081 Burrard Street, Vancouver, BC
V6Z 1Y6
e-mail: xkostaras@provideneehealth.bc.ca

Can J Psycbiatry, Vol 45, August 2000

Women who have histories of depressive disorders are more

likely to have a relapse ofthe illness after childbirth if they are
554

August 2000

Partner Support in the Treatment of Postpartum Depression

dissatisfied with their partners. A lack of communication is

the most common complaint among these women. Conversely, there is evidence that if a psychologically vulnerable
woman is in a relationship within which she is appreciated by
her partner this appreciation may actually protect her from
PPD(18).
Several studies have indicated that supportive partners play a
significant role in reduction of stress levels and improvements of mood in new mothers (18,19). This study investigated the impaet of partner support in the treatment of
mothers suffering from PPD.
Method
Partieipants
The sample of participants was obtained from those women
referred to the Reproductive Mental Health Programs at 2
major university hospitals in Vancouver, British Columbia,
for treatment of postpartum illness. The 29 women included
in the study all met the DSM-IV criteria for major depression
with postpartum onset. All women were married or cohabiting, and all had an Edinburgh Postnatal Depression Scale
(EPDS) score of 12 or more at the start ofthe study. Informed
consent was obtained for all participants and their partners.
Table 1 characterizes the demographic characteristics ofthe
participants in this study.
Measures
The following measures were employed in this study:
1. Mini Intemational Neuropsychiatric Interview (MINI),
section A (major depressive episode). The MINI is a semistructured, clinician-rated interview tool based on the DSMIV. It is used to help establish the diagnosis of depression
(20).
2. Edinburgh Postnatal Depression Scale (EPDS). This instrument consists of 10 short statements, each with 4 possible
responses. The mother identifies the response that is closest to
how she has been feeling during the past week. Validation
studies show that mothers who score above 12 or 13 ofthe
maximum score of 30 are likely to be suffering from a postpartum depressive illness (21).
3. Kellner Symptom Questionnaire. This is a self-rated scale
that measures distress and well-being. The patient is instructed to read quickly through a list of 92 psychiatric and
somatic conditions and choose the response (yes or no, true or
false) that best describes how she has been feeling during the
past week and on the day of the interview. Respondents are
given a rating of 1 for each symptom that is checked "yes" or
"true" and for each statement of well-being that is checked
"no" or "false." A higher score indicates more distress than a
lower score (22).
4. Dyadic Adjustment Scale (DAS). This is a 37-item selfreport questionnaire that uses a 5-point scale ranging from
"always disagree" = 0 to "always agree" = 5. High total and
subscale scores indicate positive appraisal of the marriage
(23). This scale is designed to detect changes in the marital

555

relationship and has been used often in studies related to postpartum adjustment (1).
5. General Health Questionnaire (GHQ). This scale includes
4 subscales: somatic symptoms, anxiety and insomnia, social
dysfunction, and severe depression. A higher total sum on the
combined subscales is indicative of poorer general health
(24,25). This 12-item screening questionnaire has been
widely used for detecting psychiatric disorders (25). The 12item GHQ has been validated with more detailed assessments, such as the GHQ-60 (60 items), the GHQ-30 (30
items), and the GHQ-28 (28 items), and is used often in general practice studies of physical illness and distress.
6. Parental Bonding Instrument (PBI). This 25-item selfreport measure is used to measure recollections of affection
and control from each parent over the first 16 years ofthe patient's life. Subjects score their perceptions of parental behaviour on a 4-point Likert scale (0-3) applied to 12 items
conceming affection (range 0-36, where 36 = "very affectionate") and 13 items conceming control (range 0-39, where
39 = "very controlling") (26).
Proeedure
Each new postpartum patient who met the study criteria and
consented to participate was randomly assigned either to the
control group, in which the patient attended all therapy sessions alone, or to the support group, in which the patient's
partner attended selected therapy sessions. All patients attended 6 clinical visits, I week apart. Visit 7, which was the
follow-up, took place 1 month later.
Study Visit I (Baseline). All patients underwent a psychiatric
assessment using the MINI. Patients also completed a baseline assessment package consisting of a demographic questionnaire, the EFDS, the Symptom Questionnaire, the DAS,
and the PBI. In addition, patients in both groups took home a
questionnaire package for their partners to complete. This
package included the DAS and the GHQ.
Study Visits 2 to 5. Control-group patients attended visits 2
through 5 alone. At each visit, the patient's mood was assessed, and treatment with medications was reviewed. In the
support group, partners attended visits 2 and 4, and the first
author encouraged positive interaction between the couple by
focusing specifically on such postpartum issues as helping
with the baby and participating in housework and other related tasks.
Study Visit 6. On study visit 6, patients in both groups were
again assessed by the first author, using the MINI. Once
again, control-group patients attended visit 6 alone, whereas
support-group patients attended visit 6 with their partners.
Patients in both groups completed a questionnaire package
consisting ofthe EPDS, the Symptom Questionnaire, and the
DAS. Support-group partners were asked to complete the
DAS and the GHQ during their visit. Control-group partners
were sent a package consisting ofthe same 2 questionnaires
and were instmcted to mail it back when completed.
Study Visit 7 (Follow-up). Study visit 7 took place 1 month after study visit 6. Patients in the control group attended this

The Canadian Journal of Psychiatry

556

Vol 45, No 6

Table 1. Demographic characteristics

Control group (n = 13)

Characteristic
Number
Age of patients (years)
Marital status

Support group (n = 16)

Mean

Range

SD

33.5

26-42

4.4

Number

Common-law

Married

12

14

Mean

Range

SD

32.9

23-46

7.2

Comparison of ControlGroup and Support-Group

Partieipants
Descriptive statistics of the
sample's demographies are
listed in Table 1. Both groups
displayed similar demographic
characteristics, except for a
greater range of ages among
mothers in the support group
compared with the control
group ( 2 3 - 4 6 years and
26-42 years, respectively).

Ethnicity
Asian

Caucasian

II

16

Other

Education
Some high school

High school

Post-secondary

University

Employment area
None

Homemaker

Technical

Professional

Age of child (months)

5.4

0.7-11.0

3.2

4.9

1-11

3.11

1.5

1-3

0.66

1.6

1-3

0.72

Sex of child
Female

Male

First child
Yes

No

Unknown

Total number of children

Medical history
Major depression

Eating disorders

Postpartum depression

Premenstrual syndrome

Anxiety or Panic disorder

0
1

Obsessive-compulsive
disorder
None

session alone, and patients in the support group attended this

visit with their partners. As in visits 1 and 6, the first author
again completed a psyehiatrie assessment using the MINI.
Patients in both groups completed the EPDS and the Symptom Questionnaire. Partners in the support group completed
the GHQ during visit 7, whereas those in the control group
were sent a package consisting of the GHQ and were instructed to complete and retum it.
Analysis
The significance of differenees between the control group
and the support group on baseline measures, postintervention
measures, and follow-up assessments was determined using
independent sample /-tests.
Results
The results are summarized in Tables 2 through 4.

Attrition
All 29 patients attended all 7
visits and completed all
assessments.

Baseline Charaeteristies
All statistical analyses were
performed at an alpha level of
0.05. Tables 2a and 2b show
that there were no significant
differences between the control- and support-group patients in terms of baseline
assessments. For the partners,
significant differences were
found on the DAS scores: the
support-group partners had a
higher level of dyadic adjustment, indicating that they had a
more positive appraisal oftheir
marriage than did their
eontrol-group counterparts.

Visit 6 Assessments
Table 3a indicates that, relative
to control-group women,
support-group women experienced a decrease in depressive
symptoms by visit 6, as measured by the MINI, Symptom
Questionnaire, and EPDS.
With respect to dyadic adjustment, by visit 6 there was a more
profound margin and a statistically significant difference in
DAS scores between the 2 groups of women. The controlgroup subjects displayed a marked decrease in DAS scores
from the baseline visit to visit 6, indicating that the lack of
partner support may have resulted in a more negative appraisal ofthe relationship. Conversely, when compared to the
baseline visit, the support-group subjects demonstrated a
slight increase in DAS scores, indicating a slightly more positive appraisal oftheir relationships.
It is interesting that the DAS scores for both groups of partners decreased from visit 1 to visit 6. However, the general
trend in the data still shows that the support-group partners
had significantly higher overall DAS scores than did their
counterparts in the control group at visits 1 and 6. The health

August 2000

Partner Support in the Treatment of Postpartum Depression

the GHQ) and the control group displaying

a decrease in general health (as indicated
by a higher score on the GHQ).

Table 2. Basetine characteristics

Control
Measure

Number

Support

Mean
seore

SD

Number

18.4

5.1

15

Mean
score

SD

Discussion

Part a: Patients
MINI: major depression

13

EPDS

13

16

Kellner

13

9.0

6.0

15

DAS

13

99.8

16.0

16

PBI

13

35.9

15.0

15

17.0

6.9

0.55

9.3

6.6

0.92

125.6

28.9

0.071

38.5

4.7

0.23

Part b: Partners
DAS

II

96.3

47.1

16

132.6

21.6

0.012

OHQ

13

25.9

17.1

15

20.0

13.3

0.31

PBI

10

38.3

5.1

16

37.9

4.7

0.86

Table 3. Post-intervention assessments: visit 6

Control
Measure

Number

Mean
seore

Support
SD

Number

Mean
seore

SD

Part a: Patients
MINI: major depression

EPDS

13

14.6

7.2

15

11.4

6.2

0.20

Kellner

13

8.1

6.0

15

4.5

4.7

0.083

DAS

12

81.3

42.0

16

127.1

22.6

0.0009

DAS

13

89.6

47.0

16

126.0

29.0

0.016

GHO

13

38.0

28.7

15

20.7

14.3

0.049

Part b: Partners

Table 4. Follow-up assessments: visit 7

Control

557

Support

That partner support is an important factor

in the treatment of PPD was supported by
the data from both visit 6 and visit 7. This
finding concurs with the findings of past
studies showing that difficulties within the
marital relationship are good predictors of
maternal symptoms of PPD (11,17). Further, these studies show that partners who
are unsupportive in child-rearing are a
source of psychosocial stress in the normal
postpartum marital adjustment that occurs
with the arrival of a new infant. In addition,
a woman's perception of her partner's support helps to increase her sense of wellbeing as a woman, wife, and mother (27).
It is clear from this study that partner support has an effect on women experiencing
PPD that is measurable and statistically
significant. The results indicate that
women experience a more rapid recovery
and are also more appreciative of their
partner's contribution to the relationship
when the partner is supportive. There is
also evidence to indicate a connection between the woman's illness and the partner's general health.

Although dyadic adjustment did show a

marked ehange for patients between visits
1 and 6, the same was not true oftheir partPart a: Patients
ners. This may indicate that, in the course
MINI: major depression
8
3
of treatment, the women in the control
EPDS
13
14.7
7.2
8.6
5.2
16
0.013
group (compared with those in the support
Kellner
6.6
6.3
13
16
2.1
3.7
0.021
group) became more sensitive both to the
Part b: Partners
lack of partner support and to the breaking
43.0
13
30.5
18.4
17.1
0.013
GHO
15
down oftheir relationships. There was also
DAS =Dyadie Adjustment Scale ; EPDS = Edinburgh Postnatal Depression Scale; GHQ = General Health
noticeable deterioration in the general
Questionnaire; MINI= Mini Intemational Neuropsychiatric Instrument; PBI = Parental Bonding Instrument.
health of the control-group partners relative to those in the support group. As always in such studies, it is difficult to
of the control-group partners was also significantly worse
distinguish
between
cause and effeet, but it is possible that the
than that ofthe support-group partners, as indicated by higher
escalating
ill
health
ofthe
partners contributed to the deteriototal scores on the GHQ.
ration ofthe relationships, that the lack of improvement in the
women's mental health resulted in a consequent deterioration
Visit 7 Follow-up Assessments
in the partners' health, or that these conditions acted
synergistically.
Tables 4a and 4b indicate that, relative to control-group patients, support-group patients continued to show marked improvements from baseline scores in depressive
symptomatology and other psychiatric conditions. With reCareful studies to unravel which elements in this investigaspect to the partners, the margin of difference between the
tion constitute cause and which constitute effect would be
general health of the control group and that of the support
worthwhile, but design and implementation would be diffigroup continued to widen, with the support group displaying
cult because of the considerable effort required to eliminate
an increase in general health (as indicated by a lower score on
confounding factors.
Measure

Number

Mean
seore

SD

Number

Mean
seore

SD

558

The Canadian Journal of Psychiatry

Vol 45, No 6

Acknowledgements
The authors thank Dr Michael Myers for his support and advice in
the review of an earlier version of this manuscript.

References
1. Wolman WL, Chalmers B, Hofmeyer GJ. Nikodem VC. Postpartum depression
and companionship in the clinieal birth environment: a randomized, controlled
study. Am J Obstet Gynecol 1993; 168:1388-93.
2. Cooper PJ, Campbell EA, Day A, Kennerly H, Botid A. Nonpsychotic psychiatric
disorder after ehildbirth. Br J Psychiatry 1988;I52:799-8O6.
3. Sichel DA, Cohen L, Rosenbaum JF, Driscoll J. Postpartum onset of obsessivecompulsive di.sorder. Psychosomatics l993;34:277-9.
4. Wisner KL, Peindl K, Hanusa BH. Symptomatology of affective and psychotic illnesses related to childbearing. J Affect Disord 1994;30:77-87.
5. Appleby L. Suicide during pregnancy and in the first postnatal year. Br Med J
199i;302:137^0.
6. Wisner KL, Wheeler SB. Prevention of recurrent postpartum major depression.
Hosp Community Psyehiatry 1994;45:1191-6.
7. Zuckerman B, Amaro H, Baucher H, Cabral H. Depressive symptoms during pregnancy: relationship to poor health behaviour. Am J Obstet Gynecol
1989; 160:1170-81.
8.0'Hara M W, Swain AM. Rates and risk of postpartum depression - a meta analysis.
Int Rev Psychiatry 1996;8:37-54.
9. Beek CT. A meta-analysis of predietors of postpartum depression. Nurs Res
I996;45:297-3O3.
10. Wilson LM, Reid AJ, Midmer DK, Biringer A, Carroll JC, Stewart DE. Antenatal
psychosocial risk factors associated with adverse postpartum family outcomes.
Can Med Assoe J 1996; 154:785-99.
11. Alfonso D, Lovett S, Paul S, Arizmendi T, Nussbaum R, Newman L, and others.
Predictors of depression symptoms during pregnancy and postpartum. J Psychosom Obstet Gynaeeol 1991; 12:255-71.
12.0'Hara MW, Neunaber DJ, Zekoski EM. Prospective study of postpartum depression: prevalence, course and predictive factors. J Abnorm Psyehol
1984;93:I58-71.
13. Misri S, Duke M. Depression during pregnancy and post-partum. Joumal ofthe Soeiety of Obstetrieians and Gynaecologists of Canada 1995; 17:657-63.
14. Fleming AS, Flett Gl, Ruble DN, Shaul DL. Post-partum adjustment in first-time
mothers: relations between mood, maternal attitudes, and mother-infant interactions. Dev Psyehol 1988;24:7I-81.
15. Pridham KF. The meaning for mothers of a new infant: relationship to maternal experienee. Ameriean Joumal of Maternal Child Nursing: MCN 1987; 16:103-22.

Clinical Implications
Partner support has a significant positive effect on women experiencing postpartum depression.
Husbands or partners should be routinely included in women's visits
with both primary eare physicians and psychiatrists.

Limitations
The sample size for both groups was small.
Most of the subjects were white.
There was a short follow-up period.
16. Merchant DC, Affonso DD, Mayberry LJ. Infiuence of marital relationship and
child-care stress on maternal depression symptoms in the postpartum. J Psychosom
Obstet Gynaeeol 1995; 16:193-200.
17. Braverman J, Roux JF. Sereening for the patient at risk for postpartum depression.
Obstet Gyneeol 1978;52:731-6.
18. Marks M, Wieek A, Cheekley S, Kumar C. How does marriage protect women with
histories of affective disorder from post-partum relapse? Br J Med Psyehol
1996;69:329-42.
19. Logsdon MC, Birkimer JC, Barbee AP. Social support providers for postpartum
women. Joumal of Social Behaviour and Personality 1997;12:89-102.
20. Sheehan DV, Lecrubier Y, Sheehan KH, Amorim P, Janavs J, Weiller E, and others. The mini-intemational neuropsychiatric interview (M.I.N.I.): the development
and validation of a structured diagnostic psychiatric interview for DSM-IV and
ICD-IO. J Clin Psyehiatry I998;59 (Suppl 20):22-3.
21. Cox JL, Holden JM, Sagovsky R. Detection of postnatal depression: development
of the 10-item Edinburgh postnatal detection scale. Brit J Psyehiatry
1987; 150:782-6.
22. Kellner R. A symptom questionnaire. J Clin Psyehiatry l987;48:268-74.
23. Spanier GB. Measuring dyadic adjustment: new scales for assessing the quality of
marriage and similar dyads. Journal of Marriage and the Family 1976;38:l5-28.
24. Goldberg DP, Hillier VF. A scaled version ofthe General Health Que.stionnaire.
Psyehol Med 1979;9:l39-45.
25. Goldberg D, Williams P. The user's guide to the General Health Questionnaire.
Windsor (UK): NFER-Nelson Publishing; 1988.
26. Parker G, Tupling H, Brown LB. A parental bonding instrument. Br J Med Psyehol
1979;52(1):I-IO.
27. Hopkins J, Campbell S, Marcus M. Role of infant-related stressors in postpartum
depression. J Abnorm Psyehol 1987:96:237-41.

Resume L'effet du soutien du partenaire dans le traitement de la depression du

post-partum
Objeetif: Determiner l'effet du soutien du partenaire dans le traitement des meres souffrant de depression du postpartum (DPP).
Metttode: Lespalientes ont regu une evaluation psychiatrique complete et n 'ont ete inserites a I 'etude que si elles satisfaisaient aux eriteres du DSM-IV de trouble depressifmajeur apparu durant le post-partum. Les patientes souffrant de
DPP (n = 29) ont ete affeetees au hasarda deux groupes de Iraitement: le groupe 1 (groupe temoin) etaitforme de patientes seulement (n = 13), tandis que le groupe 2 (groupe de soutien) se eomposait de patientes (n = 16) et de leurs partenaires. Les palientes des deux groupes devaient se presenter a sept visites psyehopedagogiques ehaeune. Dans le
groupe 2, les partenaires participaient a quatre des sept visites. Les patientes des deux groupes ont repondu a une serie
de questionnaires incluant I'eehelle de depression postnatale d'Edinburgh (EPDS), le questionnaire de symptomes
Kellner, I'echelle d'adaptation dyadique (DAS) et I'instrument de lien parental (PBI). En outre, au eours de la premiere et de la septieme visite, toutes lespatienles ont ete evaluees a I 'aide du mini-instrument neuropsychiatrique international (MINI), section A (episode depressifmajeur). Les partenaires des deux groupes ont repondu a la DAS et au
questionnaire sur I 'etat de sante general (GHQ).
Resuttats: Relativement aux patientes du groupe temoin, eelles du groupe de soutien afftchaient une reduction significative des symptomes depressifs et d 'autres etats psychiatriques. Relativement au groupe de soutien, I 'etat de sante general des partenaires du groupe lemoin se deteriorait.
Conctusion : Le soutien du partenaire a un effet mesurable sur les femmes qui souffrent de DPP.