Effect of Problem Area on Outcomes Among Adolescents Treated With Interpersonal Psychotherapy for Depression

Interpersonal psychotherapy for adolescents (IPT-A) with depression is an evidence-based developmental adaptation of IPT for adults (1–5). IPT’s premise is that interpersonal difficulties are key to the development and maintenance of depression (2). Those difficulties have been categorized into four problem areas: grief (difficulty coping with the death of a loved one), role disputes (disagreements between an adolescent and others about expectations of each other in a relationship), role transitions (difficulty adjusting to a life change that requires a new role), or interpersonal deficits (social isolation and feelings of loneliness) (2). IPT-A’s goal is to help adolescents learn the communication and interpersonal problem-solving skills needed to resolve the problem area most closely related to their depression (1, 6).

It is not yet understood how patients with different problem areas may differ from one another or how problem areas may differentially affect treatment outcomes with IPT-A. This lack of knowledge has implications for understanding how adolescents’ characteristics may make them more vulnerable to specific interpersonal difficulties. Such knowledge might also inform prognosis and personalization of treatment. Some researchers have posited that the interpersonal deficits problem area may be the most difficult to treat, because individuals with these deficits may have more pervasive and long-standing impairment in their relationship skills (7). However, very little empirical examination of this hypothesis has occurred. Most trials have not reported the frequencies of participants’ primary problem areas. To our knowledge, no IPT-A trials have reported problem areas by frequency of occurrence; only 11 IPT studies have done so, and only one study (8) has examined IPT outcomes among patients with different problem areas. That study, which targeted adults experiencing depression, found no significant differences in time to depression remission among patients with different interpersonal problem areas. However, the researchers (8) reported that the study’s therapists tended to select interpersonal deficits as the secondary treatment focus, even when such deficits appeared to be their patients’ primary difficulty, which may have affected the results. The current study aimed to examine whether characteristics of adolescents differed by primary interpersonal problem area and whether these problem areas differentially affected outcomes after treatment with IPT-A.

Methods

Participants were 40 adolescents (ages 12–17) from a prior study (9) of treatment strategies for depression that included IPT-A. Demographic characteristics of the sample were as follows: mean±SD age was 14.8±1.8 years; 78% (N=31) were female and 22% (N=9) were male; 10% (N=4) were Latino; and 80% (N=32) were White, 8% (N=3) were Asian American, 8% (N=3) were American Indian or Alaska Native, and 5.0% (N=2) were biracial.

The study was a 16-week, sequential, multiple-assignment randomized trial. The parent study received institutional review board approval from the University of Minnesota, and the design and methods have been published (9). The initial treatment plan consisted of 12 sessions of IPT-A delivered over 16 weeks. At week 1, adolescents’ depressive symptoms were assessed via the Hamilton Rating Scale for Depression (HRSD), a 17 item-scale with scores ranging from 0 to 52 (higher scores indicate more severe depression). The participants were then randomly assigned for symptom assessment again at week 4 or week 8 of therapy. Adolescents whose reduction in HRSD score suggested that they would be treatment responders after 12 sessions of IPT-A (≥20% reduction in HRSD score at week 4 or ≥40% reduction in HRSD score at week 8) (10) continued the initial treatment plan of 12 IPT-A sessions. Adolescents with an insufficient reduction in HRSD score were randomly assigned to receive either fluoxetine or four additional IPT-A sessions, scheduled twice a week.

IPT-A was delivered in accordance with the IPT-A manual (1). Therapists were trained and supervised by a certified IPT-A trainer and supervisor. Selection of the primary problem area for each adolescent was made in consultation with the supervisor and was then discussed and agreed on collaboratively with the adolescent. None of the participants disagreed with the proposed problem area. The distribution of the primary problem areas was as follows: role disputes, N=15; role transitions, N=12; interpersonal deficits, N=12; and grief, N=1. Because only one adolescent had a primary problem area of grief, that adolescent was not included in the analyses.

Measures were administered at baseline (week 1) and at week 16 by independent evaluators who were blind to participants’ treatment condition (augmentation vs. no augmentation), the timing of potential augmentation, and participants’ interpersonal problem area. The following measures were included in the current study: Beck Depression Inventory, Beck Hopelessness Scale, Children’s Depression Rating Scale–Revised (CDRS-R), Children’s Global Assessment Scale (CGAS), Columbia–Suicide Severity Rating Scale, Conflict Behavior Questionnaire, Dysfunctional Attitudes Scale, Experiences in Close Relationships–Revised (ECR-R), Multidimensional Scale of Perceived Social Support, Revised Helping Alliance Questionnaire (HAQ-II), Social Adjustment Scale–Self-Report (SAS-SR), and Social Problem-Solving Inventory. Study data were collected and managed with REDCap tools hosted at the University of Minnesota.

Group comparisons of adolescents randomly assigned to a week 4 versus a week 8 decision point (i.e., whether to augment the planned 12 IPT-A treatment sessions) showed no significant differences in demographic characteristics, baseline CDRS-R score, or interpersonal problem area. Because no significant differences in CDRS-R scores at week 4 versus week 8 were evident, we collapsed the two groups for our analyses (N=39).

Analyses were conducted on the intent-to-treat sample by using SPSS Statistics, version 26. Analysis of variance was used to compare the three interpersonal problem area groups on continuous baseline characteristics. A chi-square test was used to compare the presence of a comorbid anxiety diagnosis (9) by problem area. To evaluate the impact of problem area on outcomes (CDRS-R, CGAS, and SAS-SR), we used two-level linear mixed models, in which multiple waves of assessments were nested within each individual. The impact of interpersonal problem area on change in CDRS-R score was examined by testing an interaction term of problem area × time. The analyses were controlled for number of therapy sessions attended and medication status.

Results

Baseline characteristics of the sample, stratified by interpersonal problem area, are reported in Table 1. The presence of baseline comorbid anxiety diagnosis was noted as follows: role disputes, 47% (N=7 of 15); role transitions, 33% (N=4 of 12); interpersonal deficits, 58% (N=7 of 12); all groups, 46% (N=18 of 39). All variables were normally distributed. Levene’s test for equality of variances indicated that all variables had homogeneity of variance.

Significant differences were found among adolescents in the three problem area groups on the following three baseline measures: age (F=5.01, df=2 and 36, p=0.012), expectations about the quality of the therapeutic alliance (HAQ-II) (F=4.44, df=2 and 34, p=0.019), and attachment avoidance (ECR-R) (F=4.08, df=2 and 35, p=0.026). Tukey post hoc tests showed that adolescents in the role disputes group were significantly younger than adolescents in the interpersonal deficits (mean±SD difference=1.77±0.63, p=0.021) and role transitions (mean difference=1.60±0.63, p=0.039) groups. Adolescents in the role disputes group also had significantly worse expectations for therapeutic alliance than did participants in the role transitions group (mean difference=9.90±3.49, p=0.020). Participants in the interpersonal deficits group had higher attachment avoidance than adolescents in the role transitions group (mean difference=1.13±0.45, p=0.042).

Groups did not differ in number of therapy sessions attended or in whether treatment was augmented with additional IPT-A sessions or medication. A significant interaction effect of problem area × time was found for CDRS-R (F=8.11, df=14 and 99.63, p<0.001) and SAS-SR (F=3.70, df=14 and 95.55, p<0.001) scores. At week 16, adolescents in the role transitions group had more severe depression (beta estimate=11.15±5.37, t=2.08, df=73.97, p=0.041) and social adjustment problems (beta estimate=0.79±0.25, t=3.17, df=86.90, p=0.002) than those in the interpersonal deficits group and more severe social adjustment problems than those in the role disputes group (beta estimate=0.69±0.24, t=2.92, df=86.90, p=0.004).

Discussion

The results of this study suggested some differences in baseline characteristics, which varied by interpersonal problem area, among our sample of adolescents with depression. Adolescents in the role disputes group were significantly younger than those in the interpersonal deficits and role transitions groups. This difference may be due to the normative increase in parent-child conflict that begins during early adolescence (11). Younger adolescents may not have yet developed effective strategies for navigating this new experience of conflict, which may have increased their risk for depression. If this was the case, then it may be helpful for therapists to inform families of this normative increase in conflict to help them understand why their family member may be experiencing depression symptoms during early adolescence and why an intervention that equips an adolescent to negotiate conflicts effectively may be valuable.

At baseline, the role disputes group also had worse expectations for therapeutic alliance than the role transitions group. Adolescents who have experienced significant interpersonal disputes (often with a parent or adult caregiver) may start therapy with expectations of similar difficulties in their relationship with their therapist. If replicated in future studies, this result would suggest that therapists need to make extra efforts to form a good therapeutic alliance with adolescents experiencing role disputes.

Finally, adolescents in the interpersonal deficits group had higher baseline attachment avoidance than adolescents in the role transitions group. That is, they reported more discomfort with interpersonal closeness and intimacy and greater hesitancy to rely on others for emotional support. These interpersonal patterns are consistent with the interpersonal deficits problem area, in which social isolation is experienced.

Our results also suggested that the type of interpersonal problem area affected treatment outcomes. Adolescents in the role transitions group had more severe depression posttreatment than adolescents in the interpersonal deficits group. They also had more posttreatment social adjustment difficulties than those in the other two groups. These results suggested that role transitions may be more difficult to address with IPT-A than other problem areas. It is not clear whether this finding is specific to IPT-A or whether role transitions are more difficult to address with any type of treatment. Role transitions may simply be more complicated to navigate compared with the other problem areas. To resolve a role transition, adolescents need to not only adjust their expectations of the people in the new situation, negotiate any disagreements about expectations, and learn interpersonal skills needed to be successful in the new role but also come to terms with the loss of how things used to be (2). In addition, role transitions are sometimes not chosen by adolescents, which may decrease their sense of control over the situation. If these results are replicated in subsequent studies, future researchers and therapists might consider how IPT-A may need to be modified for individuals with the role transitions problem area to improve outcomes. For example, these adolescents may need additional IPT-A sessions or other types of coping strategies to augment IPT-A.

Whereas the interpersonal deficits problem area has generally been thought to carry a worse prognosis than other problem areas, we did not find this to be the case. Adolescents in this group did report higher levels of attachment avoidance, which has been found to predict a poor treatment response in some IPT studies with adults (12). However, in separate analyses with the current study sample, higher baseline attachment avoidance instead predicted a greater decrease in depressive symptoms with IPT-A (13). Attachment may be more malleable during adolescence than during adulthood, thus IPT may have better potential for addressing interpersonal deficits during adolescence, before problematic relationship patterns become entrenched. However, these results require replication and further study.

Our study had some limitations, the main one being its small sample size. Larger studies are needed to determine whether the findings are replicable. A larger sample would also likely increase the frequency of the grief problem area, enabling assessment of that problem area’s impact on outcomes. In addition, not all participants received the same number of IPT-A sessions, and some had their treatment augmented with medication. However, by controlling the analyses for medication status and number of therapy sessions, we were able to evaluate the impact of IPT-A problem areas on treatment outcomes.

Conclusions

This preliminary study provided information about characteristics of adolescents with depression receiving IPT-A for different problem areas and about the impact of problem areas on outcomes after 16 weeks of IPT-A. If these results are replicated, close attention to the therapeutic relationship for adolescents experiencing role disputes, and to the potential need to augment treatment for adolescents experiencing a role transition, may be helpful.