Interpersonal problems and impacts: Further evidence for the role of interpersonal functioning in treatment outcome in major depressive disorder

Journal of Affective Disorders 150 (2013) 393–400

Contents lists available at ScienceDirect

Journal of Affective Disorders journal homepage: www.elsevier.com/locate/jad

Research report

Interpersonal problems and impacts: Further evidence for the role of interpersonal functioning in treatment outcome in major depressive disorder$ Lena C. Quilty a,c,n, Brian J. Mainland b, Carolina McBride a,c, R.Michael Bagby c,a a

Centre for Addiction and Mental Health, Toronto, ON, Canada Ryerson University, Toronto, ON, Canada c University of Toronto, Toronto, ON, Canada b

art ic l e i nf o

a b s t r a c t

Article history: Received 5 February 2013 Accepted 19 April 2013 Available online 31 May 2013

Introduction: Empirical research has converged to support the concurrent association between social difficulties and psychiatric symptoms; yet, longitudinal associations between interpersonal problems and treatment outcome require clarification. The current investigation evaluated the influence of interpersonal problems assessed prior to treatment on interpersonal impacts assessed during treatment as well as on treatment outcome in outpatients with major depressive disorder (MDD). Method: 125 participants with a primary diagnosis of MDD were randomized to receive cognitive behavioural therapy or interpersonal therapy. Participants completed the Beck Depression Inventory-II, Hamilton Depression Rating Scale, and Inventory of Interpersonal Problems Circumplex before and after treatment. Therapists completed the Impact Message Inventory during and after treatment. Results: Interpersonal distress improved over the course of treatment; all other interpersonal changes were non-significant when distress was taken into account. Pre-treatment rigidity and agentic problems predicted less reduction in depressive symptoms, whereas agentic and communal impacts upon therapists during treatment predicted greater symptom change. Overall interpersonal distress was only indirectly associated with treatment response later in treatment, through its association with agentic style. Results did not differ across therapy type, and were replicated across self-report and interviewerrated measures of depression severity. Limitations: Limitations include the brief duration of treatment, lack of medication arm, and potentially restricted generalizability of patients in a randomized control trial to those in routine practice. Conclusions: Interpersonal style demonstrated a trait-like stability over treatment, and appears to fluctuate due to co-occurring distress. Yet, specific interpersonal styles were negative prognostic indicators, even within therapy specifically targeting interpersonal functioning. & 2013 Elsevier B.V. All rights reserved.

Keywords: Interpersonal problems Major depressive disorder Treatment outcome Therapeutic alliance

1. Introduction Interpersonal problems are associated with a broad range of psychopathology; yet, empirical research has yielded inconsistent evidence for the link between social difficulties and patient response to treatment (Borkovec et al., 2002; Hardy et al., 2011; Holtforth et al., 2006; Ruiz et al., 2004). Major depressive disorder (MDD) in particular has been associated with a number of interpersonal problems, including social isolation, avoidance, and submissiveness (Barrett and Barber, 2007; Vittengl et al., 2003). Such social behaviours are likely to impact patient relationships ☆ Author Note: This research was supported by the Ontario Mental Health Foundation. n Corresponding author. Tel.: +1 416 535 8501; fax: +1 416 260 4125. E-mail address: [email protected] (L.C. Quilty).

0165-0327/$ - see front matter & 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.jad.2013.04.030

not only in personal and professional domains, but also in a healthcare setting. Recently, Hirsh et al. (2012) and Kushner, et al. (under review) demonstrated that interpersonal traits have direct effects on therapeutic alliance, and through alliance, indirect effects on treatment response. The assessment of interpersonal behaviours has been facilitated by Kiesler’s (1996) interpersonal communications theory, which has yet to be applied to the fulsome investigation of interpersonal problems and treatment process and outcome in patients with MDD. In the current study, we evaluate the clinical relevance of interpersonal difficulties as conceptualized by this theoretical model within this clinical context. According to interpersonal communications theory, the interpersonal behaviours of two interacting people are causally interconnected, such that the social behaviours of one individual pull for specific responses from the other in predictable ways as

394

L.C. Quilty et al. / Journal of Affective Disorders 150 (2013) 393–400

specified by “the principle of complementarity” (Kiesler, 1996). This principle can be operationalized using interpersonal behaviours as captured by the interpersonal circumplex (Kiesler, 1983, 1992). The interpersonal circumplex is comprised of two independent dimensions known as communion and agency. Communal interpersonal behaviours vary between the poles of warmth vs. coldness, and tend to pull for complimentary behaviours from others (e.g., warmth tends to pull for kindness and affection, whereas coldness tends to pull for hostility and interpersonal distance). Agentic interpersonal behaviours vary between the poles of dominance vs. submissiveness, and tend to pull for reciprocal responses from others (e.g., dominant behaviour tends to elicit submissiveness, and vice versa) (Kiesler, 1983,, 1992). Maladaptive interpersonal behaviours associated with a wide range of psychopathology can be usefully represented according to the interpersonal circumplex. Problems related to communion range from being cold and inhibited to being intrusively involved. In contrast, problems associated with agency range from being non-assertive and exploitable to being domineering and authoritarian (Horowitz et al., 1988). Ravitz et al. (2008) explored the interpersonal dynamics that contribute to maladaptive relational patterns in those with depressive difficulties. Specifically, adults with MDD have a tendency to withdraw from social supports and to exhibit timid, acquiescent behaviours. This interpersonal style is not likely to recruit social proximity or support, and instead creates even greater interpersonal distance during times of need (Ravitz et al., 2008). Thus, social impairment in adults with depressive difficulties is exacerbated by “a cycle of maladaptive interpersonal transactions that act to amplify depressogenic processes” (p. 13). Empirical evidence supports the presence of interpersonal problems in depressed samples. Barrett and Barber (2007) found that patients with MDD reported moderate levels of interpersonal distress as compared to a normative sample. More specifically, depressed patients endorsed more problems associated with social coldness and submissiveness, including social avoidance, lack of assertiveness and interpersonal distance. Patients further endorsed fewer problems in being overly nurturing, which the authors attributed to the decreased opportunity for such behaviours inherent in their social isolation. However, it has also been suggested that patients with MDD endorse problems associated with social warmth. Vittengl et al. (2003) found that depressed patients were generally non-assertive, socially avoidant and exploitable. Thus, whereas patients in both of the studies cited above were characterized by submissive (or non-assertive) interpersonal styles, the patients in the latter study were more variable in terms of their degree of interpersonal warmth, ranging from socially avoidant (or cold submissive) to exploitable (or warm submissive). Interpersonal problems as assessed by interpersonal circumplex instruments appear to improve over the course of both pharmacotherapy and psychotherapy for MDD (Huber et al., 2007; Markowitz et al., 1996). Evidence suggests that interpersonal circumplex instruments may assess both state-like interpersonal distress and trait-like interpersonal style, and that interpersonal style remains stable over the course of treatment when interpersonal distress is taken into account (Renner et al., 2012; Vittengl et al., 2003). Investigations incorporating more heterogeneous patient samples have revealed similar results (Ruiz et al., 2004) or have remarked upon the stability of interpersonal difficulties over the course of treatment (Berghout et al., 2012; Schauenburg et al., 2000). At present, the prognostic utility of pre-treatment interpersonal problems for treatment outcome in MDD remains unclear. Research has demonstrated an inverse relation between pretreatment overall interpersonal problems or distress and

treatment outcome for MDD across treatment modalities (Markowitz et al., 1996; Vittengl et al., 2003). Most recently, Renner et al. (2012) replicated this effect, and further demonstrated a marginal association between agency and symptom severity after treatment. Investigations incorporating more heterogeneous patient samples as well as additional therapeutic modalities have generally supported a positive link between communal interpersonal difficulties and treatment response (Dinger et al., 2007; Filak et al., 1986; Gurtman, 1996; Schauenburg et al., 2000), although some exceptions exist (Puschner et al., 2004; Ruiz et al., 2004). Agentic interpersonal difficulties have been more inconsistently associated with treatment response (Borkovec et al., 2002; Filak et al., 1986; Gurtman, 1996; Ruiz et al., 2004; Schauenburg et al., 2000). Interpersonal problems may influence treatment response via the therapeutic alliance. Renner et al. (2012) reported a positive relation between pre-treatment communal problems and alliance, and a negative relation between both agentic problems and alliance and interpersonal distress and alliance, over and above depressive severity. Further, in an earlier investigation of depressed patients with multiple sclerosis, results provided support for the mediating role of early working alliance in the association between overall interpersonal problems and treatment outcome (Howard et al., 2006). In a sample of patients with affective and anxious diagnoses, Muran et al. (1994) reported that cold dominant problems negatively predicted whereas warm submissive problems positively predicted alliance after three weeks of cognitive therapy. In another heterogeneous patient sample, Dinger et al. (2007) reported that communal problems were associated with better retrospectively assessed therapeutic alliance, and that therapeutic alliance was associated with better treatment outcome, in separate analyses. These investigations provide preliminary support for the influence of interpersonal problems on therapeutic alliance, and through alliance, on treatment response. Therapeutic alliance is partially defined by the interpersonal impact of the patient on the therapist. Constantino and colleagues have led a research effort focused upon the “impact messages” of patients with depressive difficulties, wherein the principle of complementarity permits the assessment of patient interpersonal functioning through others’ social responses to the patient. In an initial investigation, Constantino et al. (2008) reported that the interpersonal “impacts” of patients with chronic depression were characterized by therapists as cold and submissive, and that these impacts improved over the course of treatment. Subsequently, Constantino et al. (2010) reported that communal interpersonal impacts were associated with improved early therapeutic alliance in patients with acute depression (i.e., after three weeks of treatment). Most recently, Constantino et al. (2012) reported that decreased hostile and submissive impacts during treatment were associated with treatment outcome in patients with chronic depression. Taken together, this body of work strongly supports the prognostic utility of therapist-rated, patient interpersonal impact in therapy process and outcome. 1.1. The current investigation Empirical research has thus converged to support the interpersonal impairment in those with depressive difficulties, which tends to manifest as submissive and cold social behaviours. Interpersonal function improves over treatment for depression; yet, the degree to which changes in interpersonal style occur over and above changes in interpersonal distress has been questioned. To date, investigations have evaluated the association between interpersonal problems and therapeutic alliance or response, and interpersonal impacts and therapeutic alliance or response; no

L.C. Quilty et al. / Journal of Affective Disorders 150 (2013) 393–400

investigation has provided a concurrent evaluation of interpersonal problems and impacts as well as treatment outcome. The current investigation undertakes to do just that in a sample of outpatients with MDD receiving short-term psychotherapy. The current investigation has several design strengths including (1) statistical control for interpersonal distress; (2) assessment of interpersonal impacts in both early and later stages of treatment; and (3) replication across both self-report and therapist-rated measures of depressive severity. We utilized repeated measures analyses of variance to evaluate change in interpersonal problems and impact over the course of treatment. We then utilized structural equation modeling to explore: (1) the effect of pre-treatment interpersonal problems on interpersonal impacts on therapists, at both early and later stages of treatment; (2) the effect of pre-treatment interpersonal problems on treatment response, both direct and indirect via interpersonal impacts on therapists; and (3) the effect of interpersonal impacts on treatment response. In line with previous research, we hypothesized that interpersonal distress would decrease over treatment, and that other measures of interpersonal problems would be stable after controlling for distress. We further hypothesized that agentic and communal impacts upon therapists would increase over the course of treatment. We finally hypothesized that difficulties with communion would be positively associated with treatment response, whereas difficulties with agency would be negatively associated.

2. Method 2.1. Participants A total of 125 outpatients (43 men, 82 women) participated in a randomized treatment trial. All participants met diagnostic criteria for DSM-IV MDD as determined by the Structured Clinical Interview for DSM-IV, Axis I Disorders—Patient version (First et al., 1995), were between the ages of 18 and 60 years, free of antidepressant medication, had received no electroconvulsive therapy in the past six months, did not have a concurrent medical illness, had minimum 8 years education, were fluent in reading English, and had the capacity to give written informed consent. Exclusion criteria included the presence of bipolar disorder, psychotic disorder, substance use disorders, organic brain syndrome, or either borderline or antisocial personality disorder, as assessed by the Structured Clinical Interview for DSM-IV Axis II Personality Disorders (First et al., 1997). A total of 377 individuals were assessed; 164 were excluded because they did not meet study criteria; 39 declined to participate; and 49 were assigned to a medication condition not included in the current investigation as interpersonal impacts measures were not completed during treatment. A total of 63 participants were randomly assigned to CBT and 62 to IPT. In the CBT condition, 16 participants dropped out early, leaving a sample of 47 (13 men and 34 women); for IPT, 12 participants dropped out early, leaving a sample of 50 (18 men and 32 women). There was no difference among participants who dropped out or completed treatment for age (t¼1.82, p ¼.07) or sex (χ2 ¼.99, p ¼.32). There was no difference among participants who dropped out and completed treatment on any pre-treatment clinical characteristics (all ts o1.36, all ps4 .18) or interpersonal problems (all ts o1.71, all ps4.09). 2.2. Measures Beck Depression Inventory-II (BDI-II) (Beck et al., 1996). The BDIII is a 21-item measure of depressive symptoms with good internal

395

consistency, retest reliability, and validity. It is the most widely used self-report measure in depression studies (Santor et al., 2006). A score of 0–13 indicates normal mood, 14–20 mild depression, 21–30 moderate depression, and 30–63 severe depression. Hamilton Depression Rating Scale (Ham-D17) (Hamilton, 1960). The Ham-D17 is a semi-structured, clinician-rated interview designed to assess severity of depression in clinical trials. The Ham-D is the most widely used interview measure of depression severity in clinical trials (Bagby et al., 2004). A score of 0–7 is considered to be normal, 8–13 mild, 14–18 moderate, 19–22 severe and 423 very severe. Inventory of Interpersonal Problems Circumplex (IIP-32) (Horowitz et al., 2000). Patients completed the IIP-32 at pretreatment (week 0) and post-treatment (week 16). The IIP-32 is a 32-item derivative of the original 64-item IIP-C, including items with the highest factor loadings from the original IIP-C (Alden et al., 1990; Horowitz et al., 2000). The measure yields a score that indicates global interpersonal distress (i.e., IIP-32 Elevation), two dimension scores of agentic and communal distress (i.e., IIP-32 Agency, IIP-32 Communion), as well as scores for octant subscales including: PA (domineering/controlling), NO (intrusive/needy), LM (self-sacrificing), JK (overly accommodating), HI (non-assertive), FG (socially inhibited), DE (distant/cold), BC (vindictive/ selfcentred). Internal consistency and test-retest reliability of the IIP-32 has been found to be comparable with those of the IIP-C, suggesting that the abbreviated form loses little from the 64-item form (Hardy et al., 2011). The following formulas were previously reported by Hardy et al. (2011) and were used to calculate IIP-32 Communion and IIP-32 Agency scores. IIP−32 Communion ¼ LM þ :71ðNO þ JKÞ−:71ðBC þ FGÞ−DE IIP−32 Agency ¼ PA þ :71ðBC þ NOÞ−:71ðFG þ JKÞ−HI IIP-32 Elevation (interpersonal distress) and IIP-32 Amplitude (interpersonal rigidity) scores were calculated with the following formulas (Hardy et al., 2011). IIP−32 Elevation ¼ ðΣIIP−32 scoresÞ=8 IIP−32 Amplitude ¼

qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 2 Affiliation þDominance2

Impact Message Inventory (IMI) (Kiesler and Schmidt, 1993). Therapists completed the octant scale version of the IMI at weeks 3, 13 and 16 (post-treatment). This version of the IMI consists of 56 items of the original 90 items and possesses good internal consistency and structure (Schmidt et al., 1999). Subscales include: dominant (D), friendly-dominant (FD), friendly (F), friendlysubmissive (FS), submissive (S), hostile-submissive (HS), hostile (H), and hostile-dominant (HD). Using formulas reported by Constantino et al. (2010), we calculated weighted vector scores based on the geometry of the circle and taking into account information from each adjacent vector. For example, the S weighted formula is S+.707(FS+HS) and the FD weighted formula is FD+.707(D+F). The theoretical minimum and maximum for the weighted vectors are 16.90 and 67.59, respectively (Constantino et al., 2010). Weighted Communion and weighted Agency factors from vector scores and their weighted proximity to the pure communion and agency axes were also calculated using the following formulas: IMI Communion ¼ F−H þ :707ðFD þ FSÞ−:707ðHD þ HSÞ IMI Agency ¼ D−S þ :707ðHD þ FDÞ−:707ðHS þ FSÞ The theoretical minimum and maximum for IMI Communion and Agency are −50.69 and 50.59, respectively (Constantino et al., 2010).

396

L.C. Quilty et al. / Journal of Affective Disorders 150 (2013) 393–400

2.3. Treatment protocol Participants in treatment conditions received 16 to 20 weeks of CBT or IPT. CBT was delivered with the use of Greenberger and Padesky (1995) manual and IPT with the Weissman et al., (2000) manual. All psychotherapists were master's or doctoral level psychologists, formally trained in and sponsored by experts in the delivery of either IPT or CBT. Further details of concerning the sample and procedures are reported by McBride et al., (2006). 2.3.1. Statistical analyses We evaluated interpersonal problems and impacts during treatment with a series of repeated measures analyses of variance (ANOVAs). Separate analyses were conducted for IIP-32 Agency, IIP-32 Communion, IIP-32 Amplitude, IIP-32 Elevation, IMI Agency and IMI Communion, which served as dependent variables. Treatment condition served as the independent variable for all analyses. Analyses including IIP-32 derived scales were replicated with pretreatment IIP-32 Elevation as a covariate. We evaluated a series of stacked path models using AMOS 16.0, applying maximum likelihood method of estimation. Due to missing data, multivariate indices of normality could not be computed. Some forms of multivariate nonnormality can be indicated by univariate nonnormality, however, with acceptable levels indicated by univariate skewo 3 and kurtosis o 8 (Kline, 2005). In each model, pre-treatment IIP-32 Agency, Communion, Amplitude, and Elevation scales served as antecedent variables. IMI Agency and Communion at Weeks 3 and 13 served as mediator variables in separate models. Ham-D17 and BDI-II change scores served as criterion variables in separate models. In each model, parameter estimates were estimated for patients receiving CBT and IPT simultaneously. Initial fit for all models was equal, due to the fact that models were just-identified (i.e., the number of parameters to be estimated was equal to the degree of freedom; χ2 undefined; CFI¼ 1.00; RMSEA ¼0.00). Parameter estimates were restricted to be equal across treatment groups to evaluate any moderating effect of treatment modality. Model fit following this restriction was evaluated via change in χ2 and CFI, as proposed by Cheung and Rensvold (2002).

3. Results 3.1. Interpersonal characteristics over treatment Patient interpersonal and clinical characteristics are displayed in Table 1. For IMI Agency, there was no effect for Time, F¼.17, p ¼.80, partial η2 o .01 or Time X Treatment Group interaction, F¼1.37, p ¼.26, partial η2 ¼.02. In contrast, for IMI Communion, there was an effect for Time, F¼13.74, p o.01, partial η2 ¼.20. This finding indicates that patient affiliative impact on therapists increased over the course of treatment, to a similar degree across treatment conditions. For IIP-32 Agency, there was an effect for Time, F¼6.46, p ¼ .01, partial η2 ¼ .07 but not Time X Treatment Condition, F¼.58, p¼ .45, partial η2 o.01. For IIP-32 Communion, there was no effect for Time, F¼1.04, p¼ .31, partial η2 ¼.01 or Time X Treatment Group, F¼.01, p ¼.94, partial η2 o.01. For both IIP-32 Amplitude and Elevation, there was a significant effect for Time (F¼ 4.74, p¼ .03, partial η2 ¼.05 and F¼22.22, p o.01, partial η2 ¼ .20, respectively), but not for Time X Treatment Condition (F¼.17, p¼ .68, partial η2 o .01 and F¼3.13, p ¼.08, partial η2 ¼.03, respectively). When pre-treatment IIP-32 Elevation was included as a covariate, there was no effect for Time or Time X Treatment Group for any of the three IIP-32 scales: Agency Time F¼ .09, p ¼.77, partial η2 ¼.00, Time X Treatment Condition F¼.19, p ¼.67, partial η2 ¼.00;

Table 1 Interpersonal and clinical characteristics of sample across time points. Variable

Week 0

BDI-II 30.31 (8.00) Ham-D 18.00 (3.77) IIP-32 Agency −2.77 (2.17) Communion 0.94 (2.54) Amplitude 4.04 (1.85) Elevation 6.29 (2.04) IMI Therapist rated Agency – Communion –

3

13

16

22.11 (9.46) –

15.38 (10.46) –

– – – –

– – – –

−7.39 (15.46) 19.85 (27.09)

28.85 (26.17) −7.79 (14.27)

12.80 (10.80) 8.65 (7.45) −2.44 1.10 3.61 5.45

(1.95) (2.23) (1.69) (2.23)

−10.18 (12.85) 33.61 (24.70)

Note. Means are presented with standard deviations in parentheses. BDI-II ¼Beck depression inventory-II; Ham-D ¼Hamilton depression rating scale; IIP32¼Inventory of interpersonal problems—circumplex; IMI ¼Impact message inventory, Weighted scores.

Communion Time F¼ 2.25, p¼ .14, partial η2 ¼.02, Time X Treatment Condition F¼ 1.59, p¼ .21, partial η2 ¼.02; Amplitude Time F¼.03, p¼ .86, partial η2 ¼.00, Time X Treatment Condition F¼.68, p¼ .41, partial η2 ¼ .01. 3.2. Path analyses Path models evaluating the prediction of change in depression as assessed by the Ham-D17 by interpersonal problems and impacts are displayed in Fig. 1A (including Week 3 IMI scores) and 1B (including Week 13 IMI scores). Both models did not demonstrate a significant decrease in fit with the restriction of parameter estimates to be equal across treatment groups, Week 3 Δχ2 ¼15.72, p4 .05, ΔCFIo.01; Week 13 Δχ2 ¼11.87, p 4.05, ΔCFIo.01. Pre-treatment IIP-32 Agency and Amplitude were consistently associated with poor treatment outcome, such that increased difficulties with dominance and increased rigidity of interpersonal problems were associated with decreased change in depression over the course of treatment. Both Week 3 and Week 13 IMI Communion were associated with positive treatment outcome, indicating that higher levels of affiliative impacts at both early and later stages of treatment are associated with greater change in depression. Week 13 IMI Agency was associated with positive treatment outcome, indicating that higher levels of dominant impact was associated with increased depression change only in later phases of treatment. IIP-32 Elevation was indirectly associated with poor treatment outcome at Week 13, through its negative association with IMI Agency. Path models evaluating the prediction of change in depression as assessed by the BDI-II by interpersonal problems and impacts are displayed in Fig. 2A and B, and reveal the same pattern of results. There was mixed evidence for the equivalence of parameter estimates across treatment groups at Week 3 Δχ2 ¼16.55, p4 .05, ΔCFI¼.016; however, there was no significant decrease in fit with the restriction of parameter estimates at Week 13 Δχ2 ¼11.70, p4 .05, ΔCFIo.01. As an unrestricted analysis of Week 3 revealed a similar pattern of results, the restricted parameter estimates are presented here. Again, pre-treatment IIP-32 Agency and Amplitude were consistently associated with poor treatment outcome. IMI Communion was associated with positive treatment outcome at both Weeks 3 and 13, whereas IMI Agency was associated with positive treatment outcome at only Week 13. IIP-32 Elevation was again indirectly associated with poor treatment outcome at Week 13, through its negative association with IMI Agency.

L.C. Quilty et al. / Journal of Affective Disorders 150 (2013) 393–400

397

-.42**

Pre-treament IIP-32 Amplitude

.03

-.10 .45 .01 -.04 -.73

.26

Pre-treatment IIP-32 Communion

Week 3 IMI Communion

.01

.31**

-.07

.25

Ham-D Change Score

.04 .07

.06

Pre-treatment IIP-32 Agency

-.01

.01

Week 3 IMI Agency

-.51** .11

-.08 -.11

Pre-treatment IIP-32 Elevation

-.11 -.12 -.29*

Pre-treament IIP-32 Amplitude

-.18

-.22 .45 .01 .09 -.73

.26

Pre-treatment IIP-32 Communion

Week 13 IMI Communion

.01

.30**

-.16

.26

Ham-D Change Score

.04

-.01

.16

.04

Pre-treatment IIP-32 Agency

-.45** .07

.24**

Week 13 IMI Agency

-.09 .07

Pre-treatment IIP-32 Elevation

-.24* -.10

Fig. 1. The prediction of depression change as assessed with the Ham-D17 with pre-treatment interpersonal problems and during treatment interpersonal impacts assessed at week 3 (a) and week 13 (b). Error variables have been omitted for clarity. n signifies p o .05, nn signifies po .01 for individual parameter estimates.

4. Discussion In the current investigation, we evaluated the impact of interpersonal problems prior to treatment upon interpersonal impacts during treatment as well as depressive symptoms following one of two forms of treatment. Interpersonal behaviour has long been viewed as a fundamental component of personality (Benjamin, 2003; Kiesler, 1992; Ruiz et al., 2004), and researchers have promoted the use of personality assessment in psychotherapy (Ben-Porath, 1997; Costa and McCrae, 1992). Similar to other diagnostic groups (Bjerke et al., 2011), MDD is associated with

significant interpersonal distress, including submissive social behaviours in particular. Consistent with our hypotheses and previous research, interpersonal distress improved over the course of treatment (Renner et al., 2012; Vittengl et al., 2003). Changes in interpersonal difficulties with dominance and rigidity were not significant when overall interpersonal distress was taken into account. Both previous investigations utilized a factor analytic method to compute interpersonal distress and style scores, different from the scoring used in this study. These results thus support the robust nature of this finding across such differences in assessment strategy. More

398

L.C. Quilty et al. / Journal of Affective Disorders 150 (2013) 393–400

-.48**

Pre-treament IIP-32 Amplitude

.04

-.10 .45 .01 -.04 -.73

.26

Pre-treatment IIP-32 Communion

Week 3 IMI Communion

.00

.32**

-.07

.26

BDI-II Change Score

.04

-.01

.07

.07

Pre-treatment IIP-32 Agency

-.01

Week 3 IMI Agency

-.49** .11

-.09 -.10

Pre-treatment IIP-32 Elevation

-.11 -.09 -.38**

Pre-treament IIP-32 Amplitude

-.15

-.17 .45 .02 .08 -.73

.26

Pre-treatment IIP-32 Communion

Week 13 IMI Communion

-.02

.31**

-.17

.26

BDI-II Change Score

.04

-.01

.15

.00

Pre-treatment IIP-32 Agency

-.44** .03

.18*

Week 13 IMI Agency

-.09 .10

Pre-treatment IIP-32 Elevation

-.24* -.10

Fig. 2. The prediction of depression change as assessed with the BDI-II with pre-treatment interpersonal problems and during treatment interpersonal impacts assessed at week 3 (a) and week 13 (b). Error variables have been omitted for clarity. n signifies po .05, nn signifies p o.01 for individual parameter estimates.

substantively, these results support the trait-like character of interpersonal styles, which may appear to fluctuate in stability due to co-occurring distress. Partially consistent with our hypotheses and previous research, interpersonal impacts upon therapists increased over the course of treatment; however, this was restricted to communal behaviours (Constantino et al., 2010). Significant others also observe similar changes in interpersonal impact over treatment in adults with depressive difficulties (Holtforth et al., 2012). The limitation of this effect to communion in this investigation may be a function of the brief, protocolized nature of treatment within this study, which may have precluded opportunity to exhibit a full range of agency.

Path analyses revealed that pre-treatment interpersonal problems were of prognostic value; however, many hypotheses were disconfirmed. More specifically, difficulties associated with dominant social behaviour prior to treatment were robustly associated with poor treatment outcome. In addition, a more narrow range or rigid interpersonal style was consistently associated with less change in depression severity over the course of treatment. Results further indicated that interpersonal impacts during treatment were predictive of treatment response. An affiliative impact upon therapists was associated with good treatment outcome, at both early and later phases of treatment; however, a more dominant impact upon therapists was associated with good treatment

L.C. Quilty et al. / Journal of Affective Disorders 150 (2013) 393–400

response only later in treatment. The former result is sensible: given that friendly behaviour pulls for the same in others, a working alliance appears likely to benefit from this form of interpersonal impact. A dominant impact early in therapy may not result in an optimal dynamic, due to the more directive role often assumed by clinicians at this stage of treatment (e.g., socializing patients to the form of care, providing psychoeducation and skills training). In contrast, increased assertiveness is appropriate at later stages of treatment during which patients are preparing to transition from regular follow-up. Thus, although analyses evaluating interpersonal features during treatment appear to support the contention that interpersonal style is not inherently adaptive or maladaptive, and that extreme social engagement may be the most reliable indicator of impairment, the path analyses (which controlled for Elevation) suggest otherwise. Certain interpersonal styles – namely, interpersonal problems with agency and rigidity – appear to negatively impact treatment outcome in this context. It is notable that these results were consistent across therapy type, despite the explicit attention of IPT on impacting interpersonal functioning. The current research boasts several strengths, including the statistical control of overall interpersonal distress in analyses, the consideration of both early and later stage impacts upon therapists, and the replication across self-report and interviewer-rated measures of depressive severity. Neither treatment response nor relations among interpersonal problems, impacts and depressive symptom severity differed across treatment groups, suggesting that this pattern of results is indeed robust. Yet, the current research has several limitations as well. First, patients in randomized controlled trials may differ in qualitative ways from those in routine clinical practice (Kushner et al., 2009). Second, patients in the medication arm of this study were not administered the IMI, precluding the investigation of interpersonal dynamics within this treatment modality. Finally, the treatment protocol was relatively brief, and as noted by Renner et al. (2012), treatment duration is closely tied change in interpersonal functioning (Barkham et al., 2002).

Role of funding source Funding for this study was provided by the Ontario Mental Health Foundation. This organization had no further role in study design; in the collection, analysis and interpretation of data; in the writing of the report; and in the decision to submit the paper for publication.

Conflict of interest All authors declare that they have no conflicts of interest.

Acknowledgements We are indebted to all participants for their time and effort to complete all assessments. None of the authors has a financial interest in the results of the study.

References Alden, L.E., Wiggins, J.S., Pincus, A.L., 1990. Construction of circumplex scales for the inventory of interpersonal problems. Journal of Personality Assessment 55, 521–536. Bagby, R.M., Ryder, A.G., Schuller, D.R., Marshall, M.B., 2004. The Hamilton depression rating scale: has the gold standard become a lead weight? The American Journal of Psychiatry 16, 2163–2177. Barkham, M., Rees, A., Stiles, W.B., Hardy, G.E., Shapiro, D.A., 2002. Dose-effect relations for psychotherapy of mild depression: a quasi-experimental comparison of effects of 2, 8, and 16 sessions. Psychotherapy Research 12, 463–474. Barrett, M.S., Barber, J.P., 2007. Interpersonal profiles in major depressive disorder. Journal of Clinical Psychology 63, 247–266. Beck, A.T., Steer, R.A., Brown, G.K., 1996. Manual for Beck Depression Inventory-II. Psychological Corporation, San Antonio, TX. Ben-Porath, Y.S., 1997. Use of personality assessment instruments in empirically guided treatment planning. Psychological Assessment 9, 361–367. Benjamin, L.S., 2003. Interpersonal Reconstructive Therapy: Promoting Change in Nonresponders. Guilford Press, New York, NY.

399

Berghout, C.C., Zevalkink, J., Katzko, M.W., De Jong, J.T., 2012. Changes in symptoms and interpersonal problems during the first 2 years of long-term psychoanalytic psychotherapy and psychoanalysis. Psychology and Psychotherapy 85, 203–219. Bjerke, E., Hansen, R.S., Solbakken, O.A., Monson, J.T., 2011. Interpersonal problems among 988 Norwegian psychiatric outpatients. a study of pretreatment selfreports. Comprehensive Psychiatry 52, 273–279. Borkovec, T.D., Newman, M.G., Pincus, A.L., Lytle, R., 2002. A component analysis of cognitive-behavioral therapy for generalized anxiety disorder and the role of interpersonal problems. Journal of Consulting and Clinical Psychology 70, 288–289. Cheung, G.W., Rensvold, R.B., 2002. Evaluating goodness-of-fit indexes for testing measurement invariance. Structural Equation Modeling 9, 233–255. Constantino, M.J., Schwaiger, E.M., Smith, J.Z., DeGeorge, J., McBridge, C., Ravitz, P., Zuroff, D.C., 2010. Patient interpersonal impacts and the early therapeutic alliance in interpersonal therapy for depression. Psychotherapy: Theory, Research, Practice, Training 47, 418–424. Constantino, M.J., Manber, R., DeGeorge, J., McBride, C., Ravitz, P., Zuroff, D.C., Klein, D.N., Markowitz, J.C., Rothbaum, B.O., Thase, M.E., Arnow, B.A., 2008. Interpersonal styles of chronically depressed outpatients: profiles and therapeutic change. Psychotherapy 45, 491–506. Constantino, M.J., Laws, H.B., Arnow, B.A., Klein, D.N., Rothbaum, B.O., Manber, R., 2012. The relation between changes in patients' interpersonal impact messages and outcome in treatment for chronic depression. Journal of Consulting and Clinical Psychology 80, 354–364. Costa, P.T., McCrae, R.R., 1992. Normal personality assessment in clinical practice: the NEO Personality Inventory. Psychological Assessment 4, 5–13. Dinger, U., Strack, M., Leichsenring, F., Henning, S., 2007. Influences of patient’s and therapists’ interpersonal problems and therapeutic alliance on outcome in psychotherapy. Psychotherapy Research 17, 148–159. Filak, J., Abeles, N., Norquist, S., 1986. Client’s pretherapy interpersonal attitudes and psychotherapy outcome. Professional Psychology: Research and Practice 17, 217–222. First, M.B., Spitzer, R.L., Williams, J.B.W., Gibbon, M.1995. Structured Clinical Interview for DSM-IV Axis I Disorders—Patient Edition (SCID-I/P), Version 2. , Biometric Research Department, New York State Psychiatry Institute, New York, NY. First, M.B., Gibbon, M., Spitzer, R.L., 1997. Structured Clinical Interview for DSM-IV Axis II Personality Disorders (SCID-II): Interview and Personality Questionnaire. American Psychiatric Press, Washington, DC. Greenberger, D., Padesky, C.A., 1995. Mind Over Mood: A Cognitive Therapy Treatment Manual for Clients. Guilford Press, New York, NY. Gurtman, M.B., 1996. Interpersonal problems and the psychotherapy context: the construct validity of the inventory of interpersonal problems. Psychological Assessment 8, 241–255. Hamilton, M., 1960. A rating scale for depression. Journal of Neurology, Neurosurgery, & Psychiatry 23, 56–61. Hardy, A.O., Tracey, T.J.G., Glidden-Tracy, C., Hess, T.R., Rohling, J.E., 2011. Interpersonal contribution to outcome: the relation of interpersonal distress and symptomatic improvement as a result of psychotherapy. Clinical Psychology & Psychotherapy 18, 225–233. Hirsh, J.B., Quilty, L.C., Bagby, R.M., McMain, S.F., 2012. The relationship between agreeableness and the development of the working alliance in patients with borderline personality disorder. Journal of Personality Disorders 26, 616–627. Holtforth, M.G., Altenstein, D., Ansell, E., Schneider, C., Caspar, F., 2012. Impact messages of depressed outpatients as perceived by their significant others: profiles, therapeutic change, and relationship to outcome. Journal of Clinical Psychology 68, 319–333. Holtforth, M.G., Lutz, W., Grawe, K., 2006. Structure and change of the IIP-D pre-and postpsychotherapy: a replication using a Swiss clinical sample. European Journal of Psychological Assessment 22, 98–103. Horowitz, L.M., Alden, L.E., Wiggins, J.S., Pincus, A.L., 2000. Inventory of Interpersonal Problems Manual. Psychological Corporation, San Antonio, TX. Horowitz, L.M., Rosenberg, S.E., Baer, B.A., Ureno, G., Villasenor, V.S., 1988. Inventory of Interpersonal Problems: psychometric properties and clinical applications. Journal of Consulting and Clinical Psychology 56, 885–892. Howard, I., Turner, R., Olkin, R., Mohr, D.C., 2006. Therapeutic alliance mediates the relationship between interpersonal problems and depression outcome in a cohort of multiple sclerosis patients. Journal of Clinical Psychology 62, 1197–1204. Huber, D., Henrich, G., Klug, G., 2007. The inventory of interpersonal problems (IIP): sensitivity to change. Psychiatry Research 17, 474–481. Kiesler, D.J., 1996. Contemporary Interpersonal Theory and Research: Personality, Psychopathology, and Psychotherapy. John Wiley & Sons, New York, NY. Kiesler, D.J., 1992. Interpersonal circle inventories: pantheoretical applications to psychotherapy research and practice. Journal of Psychotherapy Integration 2, 77–99. Kiesler, D.J., 1983. The 1982 interpersonal circle: a taxonomy for complementarity in human transactions. Psychological Review 90, 185–214. Kiesler, D.J., Schmidt, J.A., 1993. The Impact Message Inventory: Form IIA Octant Scale Version. Mind Garden, Redwood City, CA. Kushner, S.C., Uliaszek, A.A., Quilty, L.C., McBride, C., Bagby, R.M. Therapeutic alliance mediates the association between personality and treatment outcomes in patients with major depressive disorder. Under review. Kline, R.B., 2005. Principles and Practice of Structural Equation Modeling, second ed Guildford Press, New York, NY.

400

L.C. Quilty et al. / Journal of Affective Disorders 150 (2013) 393–400

Kushner, S.C., Quilty, L.C., McBride, C., Bagby, R.M., 2009. A comparison of depressed patients in randomized versus nonrandomized trials of antidepressant medication and psychotherapy. Depression and Anxiety 26, 666–673. Markowitz, J.C., Friedman, R.A., Miller, N., Spielman, L.A., Moran, M.E., Kocsis, J.H., 1996. Interpersonal improvement in chronically depressed patients treated with desipramine. Journal of Affective Disorders 41, 59–62. McBride, C., Atkinson, L., Quilty, L.C., Bagby, R.M., 2006. Attachment as moderator of treatment outcome in major depression: a randomized control trial of interpersonal psychotherapy versus cognitive behavior therapy. Journal of Consulting and Clinical Psychology 74, 1041. Muran, J.C., Segal, Z.V., Samstag, L.W., Crawford, C.E., 1994. Patient pretreatment interpersonal problems and therapeutic alliance in short-term cognitive therapy. Journal of Consulting and Clinical Psychology 62, 185–190. Puschner, B., Kraft, S., Bauer, S., 2004. Interpersonal problems and outcome in outpatient psychotherapy: findings from a long-term longitudinal study in Germany. Journal of Personality Assessment 83, 223–234. Ravitz, P., Maunder, R., McBride, C., 2008. Attachment, contemporary interpersonal theory and IPT: an integration of theoretical, clinical, and empirical perspectives. Journal of Contemporary Psychotherapy 38, 11–21. Renner, F., Jarrett, R.B., Vittengl, J.R., Barrett, M.S., Clark, L.A., Thase, M.E., 2012. Interpersonal problems as predictors of therapeutic alliance and symptom

improvement in cognitive therapy for depression. Journal of Affective Disorders 138, 458–467. Ruiz, M.A., Pincus, A.L., Borkovec, T.D., Echemendia, R.J., Castonguay, L.G., Raguasea, S.A., 2004. Validity of the inventory of interpersonal problems for predicting treatment outcome: an investigation with the Pennsylvania Practice Research Network. Journal of Personality Assessment 83, 213–222. Santor, D.A., Gregus, M., Welch, A., 2006. Eight decades of measurement in depression. Measurement 4, 135–155. Schmidt, J.A., Wagner, C.C., Kiesler, D.J., 1999. Psychometric and circumplex properties of the octant scale Impact Message Inventory (IMI-C): A structural evaluation. Journal of Counseling Psychology 46, 325–334. Schauenburg, H., Kuda, M., Sammet, I., Strack, M., 2000. The influence of interpersonal problems and symptom severity on the duration and outcome of short-term psychodynamic psychotherapy. Psychiatry Research 10, 133–146. Vittengl, J.R., Clark, L.A., Jarrett, R.B., 2003. Interpersonal problems, personality pathology, and social adjustment after cognitive therapy for depression. Psychological Assessment 15, 29–40. Weissman, M.M., Markowitz, J.C., Klerman, G.L., 2000. Comprehensive Guide to Interpersonal Psychotherapy. Basic Books, New York, NY.