Social Support and Relationship Satisfaction in Bipolar Disorder
BRIEF REPORT
Social Support and Relationship Satisfaction in Bipolar Disorder
Grace B. Boyers and Lorelei Simpson Rowe Southern Methodist University
Social support is positively associated with individual well-being, particularly if an intimate partner provides that support. However, despite evidence that individuals with bipolar disorder (BPD) are at high risk for relationship discord and are especially vulnerable to low or inadequate social support, little research has explored the relationship between social support and relationship quality among couples in which a partner has BPD. The current study addresses this gap in the literature by examining the association between social support and relationship satisfaction in a weekly diary study. Thirty-eight opposite-sex couples who were married or living together for at least one year and in which one partner met diagnostic criteria for BPD completed up to 26 weekly diaries measuring social support and relationship satisfaction, as well as psychiatric symptoms. Results revealed that greater social support on average was associated with higher average relationship satisfaction for individuals with BPD and their partners, and that more support than usual in any given week was associated with higher relationship satisfaction that week. The converse was also true, with greater-than-average relationship satisfaction and more satisfaction than usual associated with greater social support. The results emphasize the week-to-week variability of social support and relationship satisfaction and the probable reciprocal relationship between support and satisfaction among couples in which a partner has BPD. Thus, social support may be important for maintaining relationship satisfaction and vice versa, even after controlling for concurrent mood symptoms.
Keywords: bipolar disorder, marriage, social support, longitudinal, relationship satisfaction
Bipolar disorder (BPD) is a severe and chronic illness charac- terized by extreme mood shifts (American Psychiatric Association, 2000) and impairment in occupational and social functioning, even between affective episodes (Fagiolini et al., 2005; Judd & Akiskal, 2003). Individuals with BPD are less likely to marry or live with a romantic partner, and those who do are at higher risk for relationship distress and dissolution compared to individuals with other psychiatric disorders and those without mental illness (Co- ryell et al., 1993; Judd & Akiskal, 2003; Whisman, 2007). Rela- tionship dysfunction has been attributed to a number of factors, including patient mood symptoms (e.g., Lam, Donaldson, Brown, & Malliaris, 2005), caregiver burden (Reinares et al., 2006), and deficits in psychosocial functioning (Coryell et al., 1993). Partners
of individuals with BPD are also at risk for social, occupational, and financial distress, and symptoms of depression and anxiety (Lam et al., 2005). The high risk for individual and couple distress has led to calls to investigate factors that may buffer the negative effects of illness and improve functioning among individuals with BPD and their partners (Reinares et al., 2006).
One potential buffering factor is social support. Multiple studies with nonclinical samples have demonstrated a positive association between social support and individual well-being (for a review, see Cohen & Wills, 1985), particularly when an intimate partner is the support provider (e.g., Beach, Martin, Blum, & Roman, 1993). This effect has been documented with both self-report and ob- served data, concurrently and over time (e.g., Cutrona & Suhr, 1994; Sullivan, Pasch, Johnson, & Bradbury, 2010). Moreover, social support appears to buffer the effects of individual and couple-level stress on individual and relationship functioning (Bodenmann, 1995) and facilitate caring and intimacy (Cohen & Wills, 1985). This research is consistent with the intimacy process model (Reis & Patrick, 1996), which suggests that intimacy de- velops through exchanges that convey validation and understand- ing, especially in response to expressions of vulnerability. In contrast, inadequate or miscarried social support attempts are associated with declines in relationship quality over time (e.g., Brock & Lawrence, 2009).
For individuals with BPD, lack of social support (in either the quality or the number of supportive relationships) is associated with lower medication compliance and greater stress (e.g., Kul- hara, Basu, Mattoo, Sharan, & Chopra, 1999). In contrast, the
Grace B. Boyers and Lorelei Simpson Rowe, Department of Psychology, Southern Methodist University.
The analyses presented in this study were conducted in fulfillment of Grace B. Boyers’s master’s thesis and have not previously been published.
Previous versions of the analyses presented in this study were presented as a poster at the Annual Conference of the Association for Behavioral and Cognitive Therapies in November 2015 and as a paper at the Annual Conference of the Southwestern Psychological Association in April 2016. Other analyses using this data set were presented in Rowe and Miller Morris (2012).
Correspondence concerning this article should be addressed to Lo- relei Simpson Rowe, Department of Psychology, Southern Methodist University, P.O. Box 750442, Dallas, TX 75275-0442. E-mail: lsimpson@smu.edu
T hi
s do
cu m
en t
is co
py ri
gh te
d by
th e
A m
er ic
an P
sy ch
ol og
ic al
A ss
oc ia
ti on
or on
e of
it s
al li
ed pu
bl is
he rs
. T
hi s
ar ti
cl e
is in
te nd
ed so
le ly
fo r
th e
pe rs
on al
us e
of th
e in
di vi
du al
us er
an d
is no
t to
be di
ss em
in at
ed br
oa dl
y.
Journal of Family Psychology © 2018 American Psychological Association 2018, Vol. 32, No. 4, 538 –543 0893-3200/18/$12.00 http://dx.doi.org/10.1037/fam0000400
538
presence of support predicts longer time between recurrence of affective episodes (Cohen, Hammen, Henry, & Daley, 2004; John- son, Lundström, Åberg-Wistedt, & Mathé, 2003) and quicker recovery from mood episodes (Johnson, Winett, Meyer, Green- house, & Miller, 1999). However, no known research has directly studied the association between social support and relationship satisfaction within the context of BPD. This is particularly impor- tant because, although individuals with BPD have a high need for social support, they often do not receive it (Coryell et al., 1993). Likewise, their partners receive less social support than partners of individuals without mental illness; this has been attributed to limited social activities as well as lower support from the partner with BPD (Dore & Romans, 2001). In the current study, we examine the association between social support and relationship satisfaction among individuals with BPD and their intimate part- ners using an intensive longitudinal diary method. This method permits evaluation of fluctuation of variables over time, whereas the existing, predominantly cross-sectional research does not. That is, we can assess the overall association between relationship satisfaction and social support as well as the association between fluctuations in each variable.
Second, we focus on each participant’s report of emotional support they received from their partner (e.g., expressions of care and understanding). We focus on perceived social support because associations between one partner’s report of support provision and the other’s report of support receipt are often weak (Haber, Cohen, Lucas, & Baltes, 2007), reflecting the subjective nature of social support and variability in support provision skill (Howland & Simpson, 2010). That is, one partner may engage in actions in- tended to be supportive that the other partner does not perceive as helpful, which can decrease relationship satisfaction (Bolger & Amarel, 2007). We also focus on emotional support, specifically, because it is more universally acceptable than instrumental support (i.e., active assistance; Cutrona & Suhr, 1992).
We examined weekly reports of partner provision of social support from individuals with BPD and their partners, hypothesiz- ing that (a) individuals with BPD would report receiving more support than would their partners. We also tested the hypotheses that (b) support would be positively associated with relationship satisfaction on average and (c) support in any given week would be positively associated with relationship satisfaction in that week. Finally, because there is reason to believe that social support and relationship satisfaction build upon each other in a reciprocal fashion (Dunkel-Schetter & Skokan, 1990), we tested the converse hypotheses that (d) relationship satisfaction would be associated with support on average and (e) satisfaction in any given week would be positively associated with support in that week. We controlled for patient and partner depressive symptoms and patient manic symptoms because own and partner symptoms correlate with relationship satisfaction and social support (Lam et al., 2005; Lee et al., 2011; Whisman, Uebelacker, & Weinstock, 2004).
Method
Participants
Thirty-eight individuals with a lifetime diagnosis of bipolar I (90%) or bipolar II (10%) disorder and their opposite sex partners participated in a 6-month weekly diary study. In 71% of cases, the
individual with bipolar disorder (hereafter referred to as the pa- tient) was female. Participants ranged in age from 25 to 64 years, with a mean age of 44 years (SD � 10) for patients and 46 years (SD � 11) for partners. The sample was predominantly non- Hispanic White (92% of patients, 84% of partners), with the remainder identifying as Hispanic of any race (5% of patients and 8% of partners) or other (3% of patients, 8% of partners). Partic- ipants had 15 years of education on average (SD � 3 years) and 50% of patients and 76% of partners were employed, with a median household income of $4,500 per month. All couples had been living together for at least 1 year, with an average relationship length of 12 years (SD � 10), and 84% were married. In 76% of couples, at least one partner had a biological child (children’s age ranged from 1 to 41 years), with a mean of 2.86 children (SD � 1.66) among couples who had children.
Procedure
The study was conducted in a large southwestern city in the United States. All procedures were approved by the local institu- tional review board. Couples were recruited through Internet and newspaper advertisements and presentations to local mental health consumer organizations. To participate, one partner had to meet Diagnostic and Statistical Manual of Mental Disorders, fourth edition (DSM–IV; American Psychiatric Association, 2000) crite- ria for bipolar I or II disorder, and the other partner could not meet criteria for a bipolar spectrum disorder or a primary psychotic disorder. The couple had to be married and/or cohabiting for at least one year, and partners had to be between the ages of 25 and 64 years, have completed a tenth-grade education or higher, and be able to read and understand English.
After providing informed consent, participants completed a bat- tery of questionnaires and clinical interviews (including those to confirm diagnostic eligibility) at a laboratory assessment. At the end of the assessment, participants completed the first weekly diary, described below, and received instructions for completing and returning weekly diaries for the next 6 months. Participants received $125 each ($250 per couple) in compensation for com- pleting the initial laboratory session and $5 for each completed diary. They were asked to complete the weekly diaries indepen- dently from their partner and return them in self-addressed, stamped envelopes. To encourage timely completion of diaries, participants received payment only if the post date of the diary was within 3 days of the due date; only data from these diaries were included in analyses. Participants completed an average of 20 weekly diaries (range � 2–26, SD � 8), with 74% completing at least 20, 8% completing 10 –20, 9% completing 5–10, and 8% completing 4 or fewer.
Measures
Diagnosis and symptoms. The Structured Clinical Interview for DSM–IV Axis I disorders, research version, patient edition (SCID-I/P; First, Spitzer, Gibbon, & Williams, 2002) was used to confirm diagnostic eligibility. The SCID was administered by clinical psychology doctoral students under the supervision of the primary investigator. Patients and their partners completed the SCID-I/P independently with different interviewers. The SCID-I/P is a reliable and well-validated diagnostic tool (e.g., First, Spitzer,
T hi
s do
cu m
en t
is co
py ri
gh te
d by
th e
A m
er ic
an P
sy ch
ol og
ic al
A ss
oc ia
ti on
or on
e of
it s
al li
ed pu
bl is
he rs
. T
hi s
ar ti
cl e
is in
te nd
ed so
le ly
fo r
th e
pe rs
on al
us e
of th
e in
di vi
du al
us er
an d
is no
t to
be di
ss em
in at
ed br
oa dl
y.
539RELATIONSHIP SATISFACTION IN BPD
Gibbon, & Williams, 2002). Interrater agreement within this study was calculated by rescoring 30% of all interviews (n � 23); current mood episode (� � .89), and mood diagnosis (� � .83) had acceptable agreement.
Patient and partner weekly depressive symptoms were assessed with the Patient Health Questionnaire (PHQ-9; Kroenke, Spitzer, & Williams, 2001), a nine-item measure of DSM–IV depressive symptoms experienced in the past week. Symptoms were rated on a scale ranging from 0, not at all, to 3, nearly every day, with total scores ranging from 0 to 27. The PHQ-9 is well validated and reliable, with good specificity and sensitivity to change (Kroenke et al., 2001). Coefficient alpha for the first diary was .91 for patients and .88 for partners. First-week diary scores were corre- lated with Hamilton Rating Scale for Depression (Hamilton, 1960) scores obtained at the laboratory assessment, r � .67 for patients, .77 for partners, ps � .001.
Patient weekly manic symptoms were measured using the Alt- man Self Rating Scale for Mania (ASRM; Altman, Hedeker, Peterson, & Davis, 1997), a five-item measure of manic symptoms in which participants rate symptoms on a scale ranging from 0 to 4; total scores can range from 0 to 20. The ASRM is correlated with clinician-rated measures of mania and has good reliability and specificity (Altman et al., 1997). Coefficient alpha for ASRM scores in the first diary was .89. First-week diary scores were correlated with the Young Mania Rating Scale (Young, Biggs, Ziegler, & Meyer, 1978) scores obtained at the laboratory assess- ment, r � .77, p � .001.
Relationship satisfaction. Weekly relationship satisfaction was measured by a single item, “All things considered, how happy have you felt in your relationship in the last week?” on a nine-point scale ranging from 0, very unhappy, to 8, perfectly happy. First- week satisfaction scores were positively correlated with self- reported relationship satisfaction at the laboratory assessment us- ing the Dyadic Adjustment Scale (Spanier, 1976), r � .46 for patients and .39 for partners, ps � .05. Previous studies have documented the validity of single-item measures of constructs such as relationship closeness (Aron, Aron, & Smollan, 1992), life satisfaction (Antonucci, Lansford, & Akiyama, 2001), and well- being (Pavot & Diener, 1993).
Social support. Participants reported on weekly support using a single item, “My partner has provided emotional support for me,” on a scale ranging from 0, not at all, to 8, very much. Social support from the first diary week was correlated with reports of overall social support from the partner on the Social Provisions Scale (Cutrona & Russell, 1987), obtained at the laboratory as- sessment, r � .40 for patients and .57 for partners, ps � .05. Although social support and relationship satisfaction are correlated and some older measures of relationship satisfaction have included items about social support (Fincham & Bradbury, 1987), more recent research shows that they are related, but distinct, constructs (e.g., Funk & Rogge, 2007).
Data Analytic Plan
Multilevel models with distinguishable dyads (patient vs. part- ner) across up to 26 weeks of diaries (diary completed at the laboratory assessment plus 25 additional diaries) were used to test the hypotheses. Data that were missing at random, such as skipped individual items in multi-item scales (.01% of the PHQ-9 items
and .002% of the ASRM items), were imputed using EM imputa- tion procedures. Missing single items measuring relationship sat- isfaction and social support were not imputed (8.1% of the rela- tionship satisfaction items, 0.2% of the social support items) because it was impossible to know whether the item was missing at random or on purpose.
Models were estimated in SAS PROC MIXED (SAS Institute, Cary, NC) using restricted maximum likelihood. The intraclass correlation (as calculated for a dual-intercept empty-means model) for relationship satisfaction was .51 for patients and .39 for part- ners, indicating that 51% and 39% of the variance in relationship satisfaction was due to between-person mean differences in pa- tients and partners, respectively, with the remaining variance oc- curring at the within-person level. The intraclass correlation for social support was .56 for patients and .53 for partners, indicating that approximately half of the variance in social support was due to between-person mean differences. Thus, examination of within- person means for both relationship satisfaction and social support was justified.
We used modified Actor-Partner Interdependence Models (Kenny, 1996), including separate fixed and random intercepts for patients and partners (Atkins, 2005), as shown in the equation for relationship satisfaction below. Independent variables were disag- gregated into Level 2 person-mean (PM) and Level 1 within- person (WP) components (Singer & Willett, 2003). Person-mean variables were grand-mean centered by partner, and WP variables were centered at each individual’s mean score. We included both actor and partner effects for weekly depressive symptoms but only the actor effect of manic symptoms for patients and the partner effect for partners because partners, by definition, had very low levels of manic symptoms. The autoregressive coefficient for the dependent variable (i.e., the individual’s score from the previous week) was included in all models to control for the possibility that the association between present week satisfaction and support was due to the effect of past week values. Patient sex was not included in the analyses reported below because it did not moderate effects (analyses including sex as a moderator are available from the authors upon request).
Relationship satisfactionti � (patient)��00 � �01(PM social supporti)
� �02(PM actor PHQ-9i) � �03(PM partner PHQ-9i)
� �04(PM actor ASRMi) � �10(WP social supportti)
� �20(WP actor PHQ-9ti) � �30(WP partner PHQ-9ti)
� �40(WP actor ASRMti) � �50(previous week satisfactionti)
� ε0i� � (partner)��100 � �101(PM social supporti)
� �102(PM actor PHQ-9i) � �103(PM partner PHQ-9i)
��104(PM partner ASRMi) � �110(WP social supportti)
� �120(WP actor PHQ-9ti) � �130(WP partner PHQ-9ti)
� �140(WP partner ASRMti) � �150(previous week satisfactionti)
� ε10i�
Results
Participants reported moderate levels of relationship satisfaction (patients: M � 4.71, SD � 2.34; partners: M � 4.76, SD � 2.17) and emotional support (patients: M � 5.04, SD � 2.17; partners: M � 4.26, SD � 2.26) on average over the course of the study.
T hi
s do
cu m
en t
is co
py ri
gh te
d by
th e
A m
er ic
an P
sy ch
ol og
ic al
A ss
oc ia
ti on
or on
e of
it s
al li
ed pu
bl is
he rs
. T
hi s
ar ti
cl e
is in
te nd
ed so
le ly
fo r
th e
pe rs
on al
us e
of th
e in
di vi
du al
us er
an d
is no
t to
be di
ss em
in at
ed br
oa dl
y.
540 BOYERS AND SIMPSON ROWE
Patients had moderate symptoms of depression (M � 7.10, SD � 6.78) and mild symptoms of mania (M � 2.23, SD � 3.55) on average, whereas partners had mild symptoms of depression (M � 1.85, SD � 3.12).
As expected, a test of the difference of the intercept coefficients using an empty-means model revealed that partners reported less emotional support than patients, t(37.7) � �2.96, p � .005 (Hy- pothesis 1). Next, we tested the hypotheses that support on average would be positively associated with relationship satisfaction on average (Hypothesis 2) and that support in any given week would be associated with concurrent relationship satisfaction in that week (Hypothesis 3). We regressed weekly satisfaction onto person- mean and within-person support, controlling for past week satis- faction and person-mean and within-person psychiatric symptoms. As hypothesized, person-mean support was positively associated with average relationship satisfaction for patients, b � .51, SE � .09, p � .001, and partners, b � .35, SE � .10, p � .002, and within-person support was positively associated with within- person relationship satisfaction for patients, b � .32, SE � .04, p � .001, and partners, b � .33, SE � .04, p � .001 (see Table 1).
Finally, we tested the converse hypotheses that satisfaction would be positively associated with support, on average (Hypoth- esis 4), and that satisfaction in any given week would be associated with concurrent support (Hypothesis 5), controlling for past week support and person-mean and within-person psychiatric symp- toms. As expected, person-mean relationship satisfaction was pos- itively associated with average support for patients, b � .58, SE � .11, p � .001, and partners, b � .34, SE � .14, p � .02, and within-person relationship satisfaction was positively associated with within-person support for patients, b � .24, SE � .04, p � .001, and partners, b � .23, SE � .03, p � .001 (see Table 2).
Discussion
As expected, partners received less social support than patients, suggesting that partners of individuals with BPD may be at risk for inadequate social support in their relationships. Also as hypothe-
sized, average social support was positively associated with aver- age relationship satisfaction, and greater-than-average support within any given week was associated with greater-than-average relationship satisfaction that week, controlling for patient and partner mood symptoms and previous week relationship satisfac- tion. The converse hypotheses, with support as the dependent variable and person-mean and within-person relationship satisfac- tion as the independent variables, were also supported. These results are consistent with the literature (e.g., Cutrona & Suhr, 1994; Sullivan et al., 2010) and expand the existing body of knowledge by demonstrating a reciprocal association between support and satisfaction. This pattern is consistent with the inti- macy process model (Reis & Patrick, 1996), in which support in times of vulnerability enhances intimacy, increasing the likelihood of future expressions of vulnerability.
Our results also highlight the important relationship between social support and relationship satisfaction among couples in which a partner has BPD, over and above the well-documented effects of patient and partner mood symptoms on relationship functioning (e.g., Lam et al., 2005). Indeed, our results emphasize the need to go beyond the focus on patient symptoms and func- tioning alone in understanding BPD and to include broader rela- tionship outcomes. Specifically, although individuals with BPD and their partners are at high risk for relationship distress and dissolution (Coryell et al., 1993; Whisman, 2007), the current study shows that at least some couples coping with BPD are able to sustain high levels of satisfaction. However, the association between social support and relationship satisfaction may also indicate that low levels of either variable may have reciprocal effects, leading to declines in the other. In addition, the lower levels of support reported by partners may reflect an imbalance in support provision that could contribute to eventual relationship distress and caregiver burden (Brock & Lawrence, 2009; Lam et al., 2005). Alternatively, it may be that individuals with BPD simply need more support than their partners and the results reflect the differential need.
Table 1 Predicting Relationship Satisfaction by Patient and Partner Social Support
Variable Patient B (SE) Partner B (SE)
Intercept 3.29��� (.21) 3.42��� (.22) Weekly emotional support,
person-mean .51��� (.09) .35�� (.10) Weekly emotional support,
within-person .32��� (.04) .33��� (.04) Control variables
Previous week satisfaction .29��� (.03) .29��� (.03) Own PHQ-9, person-mean �.09�� (.03) .04 (.06) Own PHQ-9, within-person �.07��� (.02) �.15��� (.03) Own ASRM, person-mean �.03 (.07) — Own ASRM, within-person .03 (.02) — Partner PHQ-9, person-mean �.003 (.06) �.05 (.03) Partner PHQ-9, within-person �.02 (.04) .001 (.02) Partner ASRM, person-mean — .004 (.08) Partner ASRM, within-person — .05� (.02)
Note. PHQ-9 � Patient Health Questionnaire; ASRM � Altman Self- Rating Scale for Depression. � p � .05. �� p � .01. ��� p �.001.
Table 2 Predicting Social Support by Patient and Partner Relationship Satisfaction
Variable Patient B (SE) Partner B (SE)
Intercept 3.88��� (.25) 3.23��� (.23) Weekly relationship satisfaction,
person-mean .58��� (.11) .34� (.14) Weekly relationship satisfaction,
within-person .24��� (.04) .23��� (.03) Control variables
Social support the previous week .23��� (.04) .26��� (.04) Own PHQ-9, person-mean .03 (.04) �.01 (.07) Own PHQ-9, within-person �.002 (.02) �.12��� (.03) Own ASRM, person-mean .02 (.09) — Own ASRM, within-person .03 (.09) — Partner PHQ-9, person-mean .02 (.07) �.06 (.04) Partner PHQ-9, within-person �.07� (.03) �.04�� (.02) Partner ASRM, person-mean — .15 (.09) Partner ASRM, within-person — .03 (.02)
Note. PHQ-9 � Patient Health Questionnaire; ASRM � Altman Self- Rating Scale for Depression. � p � .05. �� p � .01. ��� p � .001.
T hi
s do
cu m
en t
is co
py ri
gh te
d by
th e
A m
er ic
an P
sy ch
ol og
ic al
A ss
oc ia
ti on
or on
e of
it s
al li
ed pu
bl is
he rs
. T
hi s
ar ti
cl e
is in
te nd
ed so
le ly
fo r
th e
pe rs
on al
us e
of th
e in
di vi
du al
us er
an d
is no
t to
be di
ss em
in at
ed br
oa dl
y.
541RELATIONSHIP SATISFACTION IN BPD
Limitations
The primary limitation in the current study is the use of single- item measures of relationship satisfaction and social support. Single-item measures limit the information that can be obtained about multifaceted constructs; future studies of social support in couples with BPD using more comprehensive measures of both variables are important to replicate our findings. In addition, the sample was relatively small and is not likely to be representative of all couples in which a partner has BPD. Indeed, the low levels of mood symptoms, on average, suggest that this may be a relatively high-functioning sample, although many patients in the study experienced weeks in which depressive and/or manic symptoms were quite high. Finally, the majority of participants were White, so the results may not generalize to a more diverse sample.
Implications and Future Directions
Social support from an intimate partner is highly beneficial (e.g., Cutrona & Suhr, 1994) as long as support is provided with some degree of skill and balanced, with neither partner experiencing too much burden of support provision or feeling inadequate as a result of needing support (Bolger & Amarel, 2007; Brock & Lawrence, 2009). Our findings extend the literature on social support to individuals with BPD and their partners. Unfortunately, couples in this population may be less skilled in support provision and ac- ceptance than couples without severe mental illness, given the high rates of relationship dysfunction in BPD (Coryell et al., 1993; Judd & Akiskal, 2003; Whisman, 2007). Future research will need to explore the skill with which patients with BPD and their partners provide support to each other and factors that may interfere with support provision (e.g., severe mood episodes, substance abuse, and stress). Experimental manipulation of support provision through psychoeducation or instructions may also enhance our understanding of the association between support and relationship satisfaction within this population. Such research has the potential to inform relationship and family-based interventions that may benefit individuals with BPD and their loved ones.
References
Altman, E. G., Hedeker, D., Peterson, J. L., & Davis, J. M. (1997). The Altman self-rating mania scale. Biological Psychiatry, 42, 948 –955. http://dx.doi.org/10.1016/S0006-3223(96)00548-3
American Psychiatric Association. (2000). Diagnostic and statistical man- ual of mental disorders (4th ed., text revision). Washington, DC: Author.
Antonucci, T. C., Lansford, J. E., & Akiyama, H. (2001). Impact of positive and negative aspects of marital relationships and friendships on well-being of older adults. Applied Developmental Science, 5, 68 –75. http://dx.doi.org/10.1207/S1532480XADS0502_2
Aron, A., Aron, E. N., & Smollan, D. (1992). Inclusion of other in the self scale and the structure of interpersonal closeness. Journal of Personality and Social Psychology, 63, 596 – 612. http://dx.doi.org/10.1037/0022- 3514.63.4.596
Atkins, D. C. (2005). Using multilevel models to analyze couple and family treatment data: Basic and advanced issues. Journal of Family Psychology, 19, 98 –110. http://dx.doi.org/10.1037/0893-3200.19.1.98
Beach, S. R. H., Martin, J. K., Blum, T. C., & Roman, P. M. (1993). Effects of marital and co-worker relationships on negative affect: Testing the central role of marriage. American Journal of Family Therapy, 21, 313–323. http://dx.doi.org/10.1080/01926189308251002
Bodenmann, G. (1995). A systematic-transactional conceptualization of stress and coping in couples. Swiss Journal of Psychology, 54, 34 – 49.
