Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 Mar 1.
Published in final edited form as: J Clin Child Adolesc Psychol. 2019 Aug 20;50(2):202–214. doi: 10.1080/15374416.2019.1644648

The Depression Prevention Initiative: Mediators of Interpersonal Psychotherapy – Adolescent Skills Training

Jason D Jones a, Robert Gallop b, Jane E Gillham c, Laura Mufson d, Alyssa McCarthy e, Rebecca Kanine a, Jami F Young a,f
PMCID: PMC7031021  NIHMSID: NIHMS1534778  PMID: 31429601

Abstract

Objective:

Several adolescent depression prevention programs have demonstrated effects on depressive symptoms and overall functioning. Yet, despite an increasing emphasis on elucidating mechanisms of change in interventions, few studies have identified mediators of these preventive interventions. In this study, we examined interpersonal mediators of Interpersonal Psychotherapy–Adolescent Skills Training (IPT-AST), an evidence-based depression prevention program.

Method:

The Depression Prevention Initiative (DPI) is a school-based randomized controlled trial in which 186 adolescents (mean age = 14.01, SD = 1.22; 66.7% female; 32.2% racial minority) were assigned to receive either IPT-AST (n = 95) or Group Counseling (GC; n = 91). We examined whether change in interpersonal conflict, social support, or social functioning from baseline to mid-intervention mediated the effects of IPT-AST on depressive symptoms and overall functioning at post-intervention.

Results:

At post-intervention, youth in IPT-AST had lower depressive symptoms (d = −.31) and higher overall functioning scores (d = .32) than youth in GC. Improvements in adolescent romantic functioning, reductions in peer conflict, and improvements in a factor score reflecting mother-adolescent conflict and difficulties in family functioning emerged as significant mediators. However, the effects of the intervention on change in the mediators were not statistically significant.

Conclusions:

These findings add to the sparse literature on mediators of psychosocial interventions, provide partial support for the theoretical mechanisms underlying change in IPT-AST, and highlight important directions for future prevention and intervention research.

Keywords: prevention, depression, mediators, randomized controlled trial, adolescence

Depression is a major public health concern. It is highly prevalent and the leading cause of disability worldwide (World Health Organization, 2017). Depression symptoms and diagnoses spike during adolescence (Hankin et al., 2015), and the majority of adult depression cases have their onset in adolescence (Kim-Cohen et al., 2003). This has led to increased interest in developing depression prevention programs that target adolescents to reduce the public health burden of depression. Several adolescent depression prevention programs have been developed, stemming from various theoretical orientations (e.g., cognitive-behavioral, interpersonal). Encouragingly, recent comprehensive reviews and meta-analyses of randomized controlled trials (RCT) of depression prevention programs support the promise of these interventions for reducing depressive symptoms and improving overall functioning, at least through short-term follow-up (Hetrick, Cox, Witt, Birr, & Merry, 2016; Werner-Seidler, Perry, Calear, Newby, & Christensen, 2017). However, a major gap in the literature is a lack of understanding of how these programs work. That is, what are the mediators or mechanisms by which these interventions exert their effects on outcomes?

The importance of testing mediators of intervention effects has been described at length (e.g., Kazdin, 2007, 2009). Identifying mechanisms of change not only provides support for the theoretical model underlying a particular intervention, but also highlights the components of an intervention that are most effective, facilitating the refinement or tailoring of a program to achieve optimal outcomes. Such adjustments may improve an intervention’s efficiency and enhance its longer-term effects. Yet, to date, there has been limited investigation of mechanisms of change in depression prevention programs and few mediators have been identified (e.g., Brunwasser, Freres, & Gillham, 2018; Compas et al., 2010; Duong et al., 2016; Stice, Rohde, Seeley, & Gau, 2010). In the current study, we examined potential mediators of an interpersonal depression prevention program for adolescents, Interpersonal Psychotherapy – Adolescent Skills Training (IPT-AST; Young, Mufson, & Schueler, 2016).

IPT-AST

IPT-AST is an adaptation of interpersonal psychotherapy (IPT; Mufson, Dorta, Moreau, & Weissman, 2004; Weissman, Markowitz, & Klerman, 2000) and is derived from interpersonal theories of depression which posit a dynamic and transactional interplay between relationships and mood. According to these theories, interpersonal risk factors increase individuals’ vulnerability to depressive symptoms. The depressive symptoms, in turn, further compromise interpersonal functioning, resulting in the perpetuation of depressive symptoms (Rudolph, Flynn, & Abaied, 2008; Hames, Hagan, & Joiner, 2013). For instance, maladaptive communication strategies may lead to problems in interpersonal relationships, which lead to an increase in depressive symptoms. These symptoms then further negatively impact interpersonal relationships and functioning, maintaining and exacerbating the depression.

In the IPT-AST pre-group sessions, the program leader conducts an interpersonal inventory of the adolescent’s current relationships. Based on the inventory, each adolescent identifies two or three interpersonal goals that he or she would like to address during the intervention. Examples of goals include decreasing conflict with a parent about chores, spending more time with a specific friend or family member, and communicating feelings to others to gain support. During the IPT-AST group sessions, adolescents learn specific interpersonal communication strategies to help address their interpersonal goals. For instance, in the case of conflict, teens might be encouraged to “strike while the iron is cold” to prevent the situation from escalating or to put themselves in the other person’s shoes to communicate that they understand how the other person is feeling, diffuse the situation, and achieve a more positive outcome. If the goal is to communicate better with a family member to gain support, a teen would be encouraged to use “I statements” to express his or her feelings and to work with the family member to identify possible solutions to the situation that is causing distress. Given IPT-AST’s focus on the application of interpersonal communication strategies to reduce conflict, increase support, and improve social functioning more broadly, we believe that these may be potential mechanisms of change through which IPT-AST impacts depressive symptoms and overall functioning. Further, in line with interpersonal theories of depression, on which IPT-AST is based, deficits in social functioning and social support and high levels of interpersonal conflict have been associated with adolescent depressive symptoms (La Greca & Harrison, 2005; Sheeber, Davis, Leve, Hops, & Tildesley, 2007).

IPT-AST targets multiple relational domains rather than one specific relationship. Adolescence is a time of rapid social development during which teens spend less time with parents, more time with peers, and form their first romantic relationships (Collins & Laursen, 2004). Compelling evidence suggests that functioning in each of these relational domains is important to adolescents’ psychosocial well-being (Cassidy & Shaver, 2016) and that each of the domains contributes uniquely to internalizing symptoms (e.g., La Greca & Harrison, 2005; Sheeber et al., 2007). As such, it is important to examine the proposed mediators of IPT-AST across relationships.

Prior studies of the efficacy of IPT-AST for adolescents with elevated depressive symptoms provide preliminary support for the proposed mediators. In the first RCT, which compared IPT-AST to usual school counseling (SC) in parochial middle and high schools, IPT-AST was superior to SC in reducing depressive symptoms and diagnoses and improving overall functioning through 6-month follow-up (Young, Mufson, & Davies, 2006). In this study, IPT-AST youth also experienced a significant decrease in mother-adolescent conflict through 12-month follow-up, whereas SC youth reported an increase in conflict, providing initial evidence that IPT-AST effectively targets interpersonal conflict (Young, Gallop, & Mufson, 2009). In the second RCT, IPT-AST again was superior to SC in reducing depressive symptoms and diagnoses and improving overall functioning through 6-month follow-up (Young, Mufson, & Gallop, 2010). In addition, from baseline to post-intervention, IPT-AST youth showed significantly greater improvements in overall social functioning and functioning in the friend domain compared to SC youth (Young, Kranzler, Gallop, & Mufson, 2012), indicating that IPT-AST effectively targets social functioning more broadly. The youth in IPT-AST also showed significant improvements in social functioning in the family and dating domains, but this change was not significantly greater than change in SC youth.

We recently published the short- and long-term outcomes of the third and largest RCT comparing IPT-AST to enhanced group counseling (GC) for adolescents with elevated depressive symptoms, called the Depression Prevention Initiative (DPI). Through 6-month follow-up, youth in IPT-AST showed significantly greater improvements in depressive symptoms and overall functioning than youth in GC (Young et al., 2016). Through 2-year follow-up, youth in both conditions showed significant improvements in depressive symptoms and overall functioning, but there were no significant between-group differences (Young et al., 2019).

The aim of the present study is to examine whether interpersonal conflict, social support, or social functioning more broadly mediated intervention effects in DPI. Given the lack of between-group differences in outcomes over long-term follow-up, and recommendations to test for mediation during intervention delivery when intervention effects are likely to be most pronounced (Stice et al., 2010), we focused on mediators of change in depressive symptoms and overall functioning during the IPT-AST prevention program. Specifically, we examined whether change in interpersonal conflict, support, or social functioning from baseline to mid-intervention mediated the effects of IPT-AST on depressive symptoms and overall functioning at post-intervention.

Method

Participants

The sample included 186 students in 7th through 10th grades enrolled in participating middle and high schools in central New Jersey (mean age = 14.01 years, SD = 1.22; 66.7% female). The racial distribution of the sample was 68% White, 20% African American, 4% Asian, and 8% other or mixed race. In addition, 38% of the adolescents identified as Hispanic/Latino ethnicity (see Young et al., 2016, for further details about the sample and study procedures).

Procedures

We identified adolescents with elevated depressive symptoms using a two-stage screening process (see Figure 1). First, adolescents who returned signed parental consent forms and who assented to take part in the screening completed the Center for Epidemiologic Studies–Depression Scale (CES-D; Radloff, 1977). Adolescents with a CES-D score of 16 or higher were eligible to participate in the second stage of the screening process. Of the 593 youth with a CES-D score ≥ 16, 271 (47%) agreed to participate in the second screening evaluation and the larger prevention study, if eligible. The adolescents who agreed to participate in the second screening did not differ significantly from those who declined to participate in terms of age, gender, or screening CES-D scores (all ps > .05).

Figure 1.

Figure 1.

Open in a new tab

Participant Flow Chart through Post-Intervention Assessment. Note. CES-D = Center for Epidemiologic Studies–Depression Scale; IPT-AST = Interpersonal Psychotherapy–Adolescent Skills Training; GC = Group Counseling.

In the second stage, adolescents completed the Kiddie Schedule for Affective Disorders and Schizophrenia (K-SADS-PL; Kaufman et al., 1997). Adolescents were eligible to participate in the study if they had at least two subthreshold or threshold depression symptoms on the K-SADS-PL, one of which was depressed mood, anhedonia, or irritability. Exclusionary criteria included: (a) not meeting the depression symptom criterion (n = 24); (b) current major depression or dysthymia (n = 36), bipolar disorder, psychosis (n = 1), substance abuse, or conduct disorder (n = 3) per Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, Text Revision criteria (DSM-IV-TR; American Psychiatric Association, 2000); (c) significant suicidal ideation or non-suicidal self-injury (n = 11); and/or (d) severe cognitive or language impairment (n = 1).

After stratifying by gender within each school, we used a computer-generated random number sequence to randomly assign the adolescents to receive either IPT-AST (n = 95) or GC (n = 91). Adolescents in the two conditions did not differ by age, gender, race, ethnicity, family income, baseline depressive symptoms, or baseline overall functioning. Adolescents were re-assessed at mid-intervention, post-intervention, and then at 6-month intervals through 24-month follow-up. This paper reports on mediation of intervention effects on depressive symptoms and overall functioning at post-intervention. All study procedures were approved by the Institutional Review Board and the school boards of participating school districts.

Interventions

IPT-AST.

All IPT-AST sessions were held in participating schools and co-led by group leaders who were clinical psychology graduate students and licensed clinical psychologists. There were 18 IPT-AST groups. In DPI, IPT-AST involved two individual pre-group sessions, eight group sessions, one individual mid-group session, and four individual post-intervention booster sessions in the 6 months after group. Parents were invited to attend part of the individual sessions. In the pre-group sessions of IPT-AST, adolescents receive an orientation to the IPT-AST framework and complete an interpersonal inventory with a group leader to establish interpersonal goals for the group. In the first two group sessions, adolescents learn about depressive symptoms and about the links between relationships and mood. In the third group session, adolescents learn interpersonal communication strategies to help address their interpersonal goals for group. In the remainder of the group (sessions 4 through 8), teens are given the opportunity to apply these strategies to their own relationships through role plays and work-at-home assignments. The individual mid-group session, which occurs after session 4, provides an opportunity for intensive work on the adolescents’ specific interpersonal goals. As described in Young et al. (2016), IPT-AST was delivered with high fidelity.

GC.

There were 16 GC groups, which were led by school counselors, most of whom had a Master’s degree in Education or a related field. GC matched IPT-AST in terms of frequency and duration of sessions, including the four booster sessions. However, GC only included one individual pre-group session. GC leaders were not given explicit instructions or limitations on the content or focus of the sessions. Some counselors ran manual-based cognitive-behavioral groups while others utilized predominately supportive and psychodynamic strategies.

Measures

Descriptive statistics for all measures are presented in Table 1.

Table 1.

Descriptive Statistics for Potential Interpersonal Mediators and Clinical Outcomes

GC (N = 91) M (SD) IPT-AST (N = 95) M (SD) p value
Potential Mediators (Baseline)
 Interpersonal Support – Mother (NRI) 26.27 (5.64) 26.48 (6.43) .81
 Interpersonal Support – Father (NRI) 23.76 (6.87) 23.67 (7.47) .94
 Interpersonal Support – Peer (NRI) 27.09 (6.05) 27.32 (6.22) .80
 Conflict – Mother (NRI) 13.49 (6.08) 13.83 (5.92) .70
 Conflict – Father (NRI) 12.93 (6.08) 11.97 (5.46) .27
 Conflict – Peer (NRI) 9.39 (4.03) 9.13 (3.20) .63
 Conflict – Mother (CBQ) 4.90 (4.88) 4.75 (5.36) .84
 Conflict – Father (CBQ) 4.76 (5.02) 4.49 (4.69) .71
 Family Functioning (SAS-SR) 1.97 (.70) 1.98 (.68) .95
 Friend Functioning (SAS-SR) 2.18 (.46) 2.19 (.52) .85
 Romantic Functioning (SAS-SR) 3.83 (.97) 3.81 (1.05) .90
Potential Mediators (Mid-Intervention)
 Interpersonal Support – Mother (NRI) 27.00 (5.56) 26.07 (6.42) .30
 Interpersonal Support – Father (NRI) 24.18 (6.92) 23.98 (7.02) .85
 Interpersonal Support – Peer (NRI) 27.60 (5.74) 26.88 (5.94) .41
 Conflict – Mother (NRI) 13.31 (6.27) 13.41 (5.98) .91
 Conflict – Father (NRI) 12.55 (5.56) 11.73 (6.28) .36
 Conflict – Peer (NRI) 9.50 (4.20) 9.40 (4.06) .87
 Conflict – Mother (CBQ) 4.48 (4.72) 4.69 (5.26) .77
 Conflict – Father (CBQ) 4.63 (5.01) 4.30 (5.17) .67
 Family Functioning (SAS-SR) 1.98 (.64) 1.92 (.61) .48
 Friend Functioning (SAS-SR) 2.09 (.51) 2.05 (.55) .62
 Romantic Functioning (SAS-SR) 3.69 (.93) 3.70 (1.02) .96
Clinical Outcomes (Baseline)
 Depressive Symptoms 24.41 (6.88) 23.14 (6.37) .19
 Overall Functioning 67.55 (5.24) 67.44 (4.98) .89
Clinical Outcomes (Post-Intervention)
 Depressive Symptoms 13.12 (7.62) 11.09 (7.12) .04
 Overall Functioning 72.27 (5.33) 73.88 (4.76) .04
Open in a new tab

Note. Values for variables at baseline and mid-intervention are observed means and standard deviations. Values for clinical outcomes at post-intervention are model-based estimated means and standard deviations. NRI = Network of Relationships Inventory. CBQ = Conflict Behavior Questionnaire. SAS-SR = Social Adjustment Scale – Self-Report.

Potential interpersonal mediators.

Network of Relationships Inventory-Short Form (NRI-SF; Furman & Buhrmester, 1985).

The NRI-SF is a 13-item questionnaire designed to assess interpersonal support (7 items) and conflict (6 items) across multiple relationships (e.g., with parents, peers). Adolescents indicated the degree to which different statements apply to each of the relationships from 1 (little or none) to 5 (the most). Sample items include, “How much does this person help you figure out or fix things?” and “How much do you and this person disagree or quarrel?” In the present study, we focused on support and conflict with mother, father, and peers (α ranged from .86 to .93 for interpersonal support and from .92 to .95 for conflict). The peer variable is an average rating across same-sex peers, opposite-sex peers, and romantic partners.

Conflict Behavior Questionnaire (CBQ; Robin & Weiss, 1980).

The CBQ is a 20-item assessment of parent-adolescent conflict and negative communication. Adolescents completed separate forms for their mother and father. Adolescents indicated whether each of the 20 statements (e.g., “We almost never seem to agree”) is true or false (α ranged from .90 to .91).

Social Adjustment Scale – Self-Report (SAS-SR; Weissman & Bothwell, 1976).

The SAS-SR is a 23-item questionnaire that assesses social functioning in four domains: family, friends, romantic, and school. In the present analyses, we did not examine the school subscale of the SAS-SR given our focus on interpersonal mediators. The family subscale includes items about being able to talk about problems with parents and feeling unfairly treated by family members (e.g., Have you been able to talk about your feelings and problems with your parents in the last 2 weeks?”). The friend subscale includes items about amount of time spent with friends and feelings of loneliness (e.g., “Have you felt lonely and wished for more friends during the last 2 weeks?”). The romantic subscale includes two items about interest in dating and frequency of dating (e.g., “Have you been interested in dating during the last 2 weeks?”). Items are rated on a 5-point scale with higher values reflecting poorer social functioning. Cronbach’s alphas for the friend and family subscales ranged from .52 to .72. Reliability of the two-item romantic subscale was assessed with the Spearman-Brown coefficient (.54 at both time points).

Clinical outcomes.

Center for Epidemiologic Studies–Depression Scale (CES-D; Radloff, 1977).

The CES-D is a widely-used 20-item assessment of depressive symptoms. Adolescents rated how often from 0 (Rarely or none of the time [less than one day]) to 3 (Most or all of the time [5–7 days]) they experienced each symptom over the past week (α = .85 at baseline and .89 at post-intervention assessment).

Children’s Global Assessment Scale (CGAS; Shaffer et al., 1983).

The CGAS is a clinician-rated scale of overall functioning across various domains (e.g., home, school, peers) that was completed by evaluators who were naïve to intervention condition. Evaluators received extensive training and completed a 10-case reliability assessment (ICC = .89). Ten percent of completed assessments were randomly selected and re-rated by a senior investigator (ICC = .96).

Results

Analytic Approach

Data reduction.

Our interpersonal measures yielded 11 potential mediator variables that reflect change in the interpersonal constructs from baseline to mid-intervention. To reduce the number of potential mediators, and therefore the number of tests performed, we used principal axis factor analysis (PAF) with varimax rotation to obtain orthogonal solutions. Results of the PAF yielded three latent factors that provided the best fit to the data according to traditional estimation criteria: scree test, eigenvalues > 1, and interpretability of the solutions. These three factors accounted for 93.4% of the total variance. Variables with factor loadings of > .40 were retained as indicators of a respective factor (see Table 2). The factors were labeled mother-adolescent conflict/family functioning (Factor 1), social support (Factor 2), and father-adolescent conflict as measured by the NRI-SF (Factor 3). In addition, four variables did not load onto any of the three factors (friend functioning, romantic functioning, peer conflict, and father-adolescent conflict on the CBQ); we examined these variables individually as potential mediators.

Table 2.

Principal Axis Factor Analysis of Potential Interpersonal Mediators

Factor Loadings
Variable Factor 1 Factor 2 Factor 3
Conflict – Mother (CBQ) .73 −.06 .10
Conflict – Mother (NRI) .66 .33 .39
Family Functioning (SAS-SR) .65 −.01 .16
Interpersonal Support – Mother (NRI) −.30 .74 .06
Interpersonal Support – Father (NRI) −.26 .67 .20
Interpersonal Support – Peer (NRI) .002 .48 −.05
Conflict – Father (NRI) .02 .08 .996
Friend Functioning (SAS-SR) .21 −.06 −.03
Romantic Functioning (SAS-SR) −.10 −.35 .04
Conflict – Peer (NRI) .06 .25 .26
Conflict – Father (CBQ) .13 −.11 .25
Open in a new tab

Note. Variables included in the factor analysis reflect change in the potential interpersonal mediators from baseline to mid-intervention. Estimates are varimax rotated factor loadings. Estimates in bold reflect variables retained for each factor. NRI = Network of Relationships Inventory. CBQ = Conflict Behavior Questionnaire. SAS-SR = Social Adjustment Scale – Self-Report.

Mediation analyses.

For the mediation analyses, we used mixed effects analysis of covariance modeling to test the effects of early change in the proposed mediators (from baseline to mid-intervention) on depression and overall functioning scores at post-intervention, controlling for school, income, gender, age, baseline depression and overall functioning, respectively, and the clustering attributable to the group randomization. Traditional mediation models assume there are no unmeasured confounders affecting the relationship between the mediator and outcome, defined as sequential ignorability (Preacher, 2015; Ten Have & Joffe, 2012). Because we defined the mediators as change in the relevant variables from baseline to mid-intervention, patients were not randomized to levels of the mediator, thus raising the possibility that any observed relations between mediators and subsequent outcomes might be spurious and depend on unknown confounding variables. Whether this is the case is an untestable assumption (Zheng et al., 2015). Accordingly, we implemented a specific structural nested mean model called the rank preserving model (RPM; Ten Have et al., 2007), which does not require the assumption of sequential ignorability, and which connects the common mediation parameters to causal parameters (Imai, Keele, & Tingley, 2010). Following current approaches to mediation analysis (Preacher & Hayes, 2004), we tested the significance of the multiplicative paths from intervention to the outcome (path c), intervention to the posited mediator (path a), and the mediator to outcome (path b) controlling for intervention under the RPM. Path c′ corresponds to the direct effect of the intervention on the outcome after the mediator is introduced into the model. As described by Ten Have et al. (2007), weighted G-estimation was used to obtain consistent RPM estimators of the structural parameters θMS and θRS for the RPM specified under the discussed assumptions. These estimates are obtained by solving weighted G-estimation equations dependent on the selection of baseline covariates, called instruments, which have variation in the effect of the mediator across the strata formed by the instruments and have a strong main effect on the mediator in the group randomized to the intervention. In the mediation analyses, the romantic functioning variable, Factor 1 scores, and CES-D scores were normalized with a square root transformation.

Main Effects of Intervention on the Outcomes

At post-intervention, youth in IPT-AST had significantly lower depressive symptoms than youth in GC, t(173) = 2.06, p < .05, d = −.31. In addition, youth in IPT-AST had significantly higher overall functioning scores than youth in GC at post-intervention, t(173) = −2.12, p < .05, d = .32.

Mediation Models

We identified one mediator of intervention effects on depressive symptoms (romantic functioning) and three mediators of intervention effects on overall functioning (mother-adolescent conflict/family functioning factor scores [Factor 1], romantic functioning, and peer conflict). None of the other interpersonal variables emerged as mediators. See Table 3. Figure 2 provides an illustration of the mediation models tested in this paper and the corresponding effect sizes for each path in the model.

Table 3.

Model-Based Estimated Change in Potential Interpersonal Mediators and Rank Preserving Model Results

TX → Mediator (path a) Rank Preserving Model (path b)
Potential Mediator Estimated Change IPT-AST (SE) Estimated Change GC (SE) Treatment Contrast  t(df)   p Mediator→CES-D | INT t(df)  p Mediator→CGAS |INT t(df)  p
FACTOR 1 −.07 (.09) −.06 (.08) t(156) = .10 .92 t(153) =−.58 .56 t(154) =− 2.18 .03
FACTOR 2 −.06 (.11) .17 (.11) t(156) = 1.71 .09 t(153) = −1.40 .16 t(153) = 1.36 .17
FACTOR 3 −.001 (.13) −.06 (.12) t(156) = − .39 .70 t(153) = −1.25 .21 t(153) = 1.70 .09
NON-LOADERS
Friend Functioning (SAS-SR) −.13 (.06) −.06 (.05) t(173) = .97 .33 t(170) = 1.26 .21 t(170) = −1.57 .12
Romantic Functioning (SAS-SR) −.18 (.08) −.03 (.07) t(174) = 1.53 .13 t(171) = 1.99 .047 t(171) = −1.98 .047
Conflict – Peer (NRI) −.17 (.26) .42 (.27) t(171) = 1.66 .10 t(168) = 1.69 .09 t(168) = −2.00 .046
Conflict – Father (CBQ) −.13 (.39) −.15 (.35) t(162) = −.05 .96 t(159) = −1.10 .27 t(159) = .81 .42
Open in a new tab

Notes. INT= intervention

Figure 2.

Figure 2.

Open in a new tab

Example Mediation Model: Romantic Functioning Mediates Intervention Effects on Depressive Symptoms

Depressive symptoms.

The RPM casual mediation model including romantic functioning revealed a significant mediation effect, t(171) = 1.99, p < .05, d = .30. Including romantic functioning in the model reduced the intervention effect on post-intervention depressive symptoms by 48% (d for path c′ = −.16). The corresponding coefficient for romantic functioning is 6.88, SE = 3.46. As higher scores on the SAS-SR indicate poorer functioning, this positive coefficient indicates that improvements in romantic functioning from baseline to mid-intervention were associated with lower depressive symptoms at post-intervention, on average. The effect of the intervention on change in romantic functioning (path a) was not significant, t(174) = 1.53, p = .13, d = −.23. However, IPT-AST youth showed a significant improvement in romantic functioning, on average (β = −.18, SE = .08, p < .05), whereas GC youth showed a non-significant improvement in romantic functioning, on average (β = −.03, SE = .07, p = .66). Although traditional approaches to mediation require each of the individual paths in the mediation model to be significant, contemporary approaches to mediation do not have this requirement. Mediation is possible even if one (or more) of the individual paths in the model is not statistically significant (Hayes, 2009; Hayes & Rockwood, 2017). The significant mediation effect in the RPM indicates that adding romantic functioning to the model reduced the main effect of the intervention on depressive symptoms (path c) to non-significance.

Overall functioning.

The model including Factor 1 (mother-adolescent conflict/family functioning) scores revealed a significant mediation effect, t(154) = −2.18, p < .05, d = −.33. Including Factor 1 in the mediation model reduced the intervention effect on post-intervention overall functioning by 53% (d for path c′ = .15). The corresponding coefficient for Factor 1 is −10.14, SE = 4.66. As higher scores on Factor 1 indicate more conflict and poorer family functioning, the negative coefficient indicates that improvements in mother-adolescent conflict/family functioning from baseline to mid-intervention were associated with better overall functioning at post-intervention, on average. The effect of the intervention on change in Factor 1 was not significant, t(156) = .10, p = .92, d = −.02. On average, youth in both IPT-AST (β = −.07, SE = .09, p = .45) and GC (β = −.06, SE = .08, p = .47) showed non-significant improvements on this factor. As described above, the significant mediation effect in the RPM indicates that adding Factor 1 to the model reduced the main effect of the intervention on overall functioning (path c) to non-significance.

The model including romantic functioning also revealed a significant mediation effect, t(171) = - 1.98, p < .05, d = −.30. Including romantic functioning in the model reduced the intervention effect on post-intervention overall functioning by 28% (d for path c′ = .23). The corresponding coefficient for romantic functioning is −6.60, SE = 3.33. This indicates that improvements in romantic functioning from baseline to mid-intervention were associated with better overall functioning at post-intervention, on average. As noted above, the effect of the intervention on change in romantic functioning was not significant.

The model including peer conflict revealed a significant mediation effect, t(168) = −2.00, p < .05, d = −.30. Including peer conflict in the mediation model reduced the intervention effect on post-intervention overall functioning by 28% (d for path c′ = .23). The corresponding coefficient for peer conflict is −1.84, SE = .92. Reductions in peer conflict from baseline to mid-intervention were associated with better overall functioning at post-intervention, on average. The effect of the intervention on change in peer conflict was marginally significant, t(171) = 1.66, p = .10, d = −.25. On average, IPT-AST youth showed a non-significant reduction in peer conflict from baseline to mid-intervention (β = −.17, SE = .26, p = .53), whereas GC youth showed a non-significant increase in peer conflict over that period (β = .42, SE = .27, p = .12).

Discussion

Understanding mechanisms of change in psychosocial interventions is a priority of the National Institute of Mental Health’s Strategic Plan for Research (NIMH, 2015), yet few studies have investigated mediators of youth depression prevention programs. Several studies demonstrate that IPT-AST is a promising depression prevention program (Young et al., 2006, 2010, 2016); however, there has be no formal examination of interpersonal mediators of IPT-AST’s effects on clinical outcomes. In the present study, we examined whether three theoretically-based and empirically-supported interpersonal mechanisms (social functioning, interpersonal conflict, and social support) mediated intervention effects on depressive symptoms and overall functioning.

The results revealed four significant mediation effects that spanned parent, peer, and romantic relationships. Specifically, we found that functioning in romantic relationships was a significant mediator of intervention effects on both depressive symptoms and overall functioning. In addition, mother-adolescent conflict/family functioning (Factor 1) and peer conflict significantly mediated intervention effects on overall functioning. However, it is important to note that the effects of the intervention on change in the mediators (path a) were marginally significant (peer conflict) or non-significant (mother-adolescent conflict/family functioning, romantic functioning). Although mediation is possible when the effect of the independent variable on the mediator is non-significant (Hayes & Lockwood, 2017), the lack of statistically significant effects of IPT-AST on these mediators tempers the conclusions we can draw from this study about mechanisms of IPT-AST. The significant mediation effects in the rank preserving models indicate that adding the interpersonal mediators to the model reduces the main effects of the intervention on the clinical outcomes (path c) to non-significance. Thus, we have evidence that improving these aspects of interpersonal functioning mediates the effects of IPT-AST on the outcomes, but we have insufficient evidence to conclude that IPT-AST is causing these improvements. Although the intervention contrasts were not statistically significant, the observed effects of IPT-AST on the mediators were in the expected directions. Youth in IPT-AST showed significant improvements in romantic functioning from baseline to mid-intervention, whereas GC youth did not. IPT-AST youth also showed non-significant reductions in peer conflict, whereas GC youth showed a non-significant increase in peer conflict. Both groups showed non-significant improvements in mother-adolescent/family functioning (Factor 1).

There are several possible explanations for the lack of significant effects of the intervention on the mediators (path a). One possibility is that participating in the early sessions of IPT-AST makes adolescents’ previously underestimated interpersonal difficulties more salient. Our clinical experience in IPT-AST suggests that completing the interpersonal inventory as part of the individual pre-group session makes adolescents think hard about their relationships and often results in the insight that their relationships are not as good as they previously thought. As a result, it may be difficult to detect changes in interpersonal functioning from the baseline evaluation to the mid-intervention evaluation if the baseline scores were artificially positive. This is one of the limitations of relying solely on self-report measures of relationships and highlights the need to include objective measures, such as observed interactions.

A second consideration is that IPT-AST is a multi-targeted and individualized intervention. At the beginning of the prevention program, each adolescent identifies specific relational difficulties and interpersonal goals for the intervention. For some adolescents, the main difficulties may be in the peer domain, while for others, the biggest struggles are with their parents. Further, some adolescents may be specifically interested in increasing social support and may not have problems with interpersonal conflict, whereas others are focused on reducing conflict and do not have concerns about social support. Despite the individualized nature of the intervention, we are testing for common mediators across the whole group, which may obscure intervention effects. Effects might emerge or be stronger if we took a more individualized analytic approach. For example, does peer conflict mediate intervention effects for adolescents who specified peer conflict as their main area of difficulty? This is an important question for future intervention research.

Despite these interpretative challenges, the present findings advance the sparse literature on mediators of depression prevention programs broadly and of IPT-AST specifically. Our results are partially consistent with the theoretical mechanisms of IPT-AST. Conflict and difficulties in social functioning were significant mediators, but social support was not. This pattern of findings is interesting in light of substantial evidence suggesting that negative aspects of relationships have a much stronger effect on both short- and long-term relationship outcomes than positive aspects (Baumeister, Bratslavsky, Finkenauer, & Vohs, 2001). Therefore, it is possible that reducing the negative (e.g., conflict), rather than increasing the positive (e.g., social support), has a larger impact on overall interpersonal functioning, which may, in turn, lead to greater improvements in depressive symptoms and overall functioning. This possibility is also consistent with the findings of the limited research examining interpersonal mediators of depression interventions, which has found that reducing difficulties in social functioning (Reyes-Portillo, McGlinchey, Yanes-Lukin, Turner, & Mufson, 2017), but not increasing parent-adolescent communication (Duong et al., 2016), mediated intervention effects on depressive symptoms. Alternatively, it is possible that it takes longer to see meaningful changes in social support than it does to see reductions in conflict and difficulties in social functioning. If this is the case, mediation effects of social support may not become apparent until after the intervention is completed. Further research is warranted to explore these possibilities.

It is noteworthy that our mediation results spanned multiple relationships: peers, romantic partners, and mothers/family. This highlights the need for prevention programs to focus on multiple relational domains, rather than just focusing on parents/family or peers. Notably, functioning in romantic relationships was the only construct to emerge as a mediator for both depressive symptoms and overall functioning. The two-item romantic functioning scale of the SAS-SR assesses frequency of and interest in dating during the past two weeks. Although this is a coarse measure of romantic functioning, it appears to be capturing something meaningful that is sensitive to intervention efforts. In a prior RCT, adolescents in IPT-AST showed significant improvements in this romantic functioning variable during the intervention, but this improvement was not significantly greater than that seen in the comparison group (Young et al., 2012). Similarly, in the present study, although the between-group comparison was not statistically significant, adolescents in IPT-AST showed significant improvements in romantic functioning, whereas youth in GC did not demonstrate significant change. Further, in a treatment study of depressed adolescents, those receiving interpersonal psychotherapy showed significantly greater improvements in this romantic functioning variable post-treatment compared to adolescents in the treatment as usual control group (Mufson et al., 2004). Notably, other studies using similarly brief measures of romantic functioning have found important associations with psychological distress. For example, a simple dichotomous indicator of dating or not dating has been associated with depression and anxiety symptoms (Davila et al., 2004; Joyner & Udry, 2001; La Greca & Harrison, 2005). Taken together, the evidence suggests that even coarse measures of romantic functioning have meaningful associations with clinical outcomes.

Involvement in romantic relationships is a major developmental milestone of adolescence and a salient part of adolescents’ social worlds. There is compelling evidence that adolescents’ romantic relationships have important implications for development and adjustment. Adolescents’ romantic relationships are both a source of interpersonal support and positive emotions and a source of stress and negative emotions (Collins, Welsh, Furman, 2009). In fact, involvement in a romantic relationship has consistently been associated with elevated depressive symptoms in adolescence, and adolescents are more likely to attribute strong emotions to romantic relationships compared to other sources, such as school or family (Davila et al., 2004; Joyner & Udry, 2001; La Greca & Harrison, 2005; Larson, Clore, & Wood, 1999). In addition, experiences with a romantic partner have a significant impact on adolescents’ daily negative mood (Rogers, Ha, Updegraff, & Iida, 2018). Given the highly emotional and volatile nature of adolescents’ romantic relationships and their effects on daily functioning and mood, it is important for depression prevention programs, particularly interpersonal prevention programs, to address problems in these relationships, as well as difficulties in other relational domains. Across IPT-AST groups, only a small subset of adolescents identify an interpersonal goal related to a romantic partner. However the findings across multiple studies suggest that the communication and conflict reduction strategies that adolescents learn in IPT-AST may generalize to their romantic relationships, increase their desire to date, and reduce the effects of romantic conflict on daily negative mood. Further research with more detailed assessments of romantic relationships is warranted to see which aspects of romantic functioning, if any, are changed by IPT-AST and associated with depressive symptoms.

It is also interesting that our mediation results were more consistent for overall functioning than for depressive symptoms, particularly since many depression prevention programs do not include measures of general adjustment as an outcome (Merry et al., 2011). Further, to our knowledge, no other study has examined how interpersonal variables may mediate the effects of a depression prevention program on overall functioning. Our results suggest that reducing conflict with peers and mothers and improving functioning in romantic relationships may have a more generalized effect on overall functioning than a specific effect on depressive symptoms. Of course, depressive symptoms and overall functioning are not independent constructs, as improving depressive symptoms is likely to also improve overall functioning and vice versa. Additional research on the effects of depression prevention programs on diverse outcomes, as well as mediators of these effects, is needed to advance the field.

Limitations and Additional Future Directions

The present findings should be interpreted in light of several study limitations. First, all of our potential interpersonal mediators were assessed with self-report measures, which are susceptible to various reporting biases, including the initial underestimation of interpersonal difficulties noted above. Future research on mechanisms of IPT-AST (and other interventions) should include multi-method, multi-informant assessments of potential mediators. Second, consistent with what others have found (e.g., Reyes-Portillo et al., 2017), the internal consistency of the SAS-SR subscales was low. Further, the romantic functioning subscale of the SAS-SR is based only on two items that assess the frequency of and interest in dating, rather than the quality of romantic relationships. Future research should include more thorough measures of social functioning with better psychometric properties. Third, our effort to reduce the number of mediator variables by using factor analysis was only partially successful. Several of the variables did not load onto any of the factors, Factor 1 was a combination of both mother-adolescent conflict and difficulties in family functioning which muddies the interpretation of this factor, Factor 3 consisted of only a single item, and the two measures of father-adolescent conflict did not load onto the same factor. Fourth, there are other theoretically-based potential mediators of IPT-AST, such as perspective-taking and open communication, that were not assessed in this study but should be considered in future research. Fifth, we did not assess overall functioning (CGAS) at the mid-intervention evaluation, which precluded us from testing alternative direction of effects. It is possible that improvements in overall functioning led to improvements in interpersonal relationships rather than vice versa. Future research should explore this possibility.

Conclusion

Despite study limitations and the interpretative challenges noted above, the present findings have important implications and advance the prevention literature in a number of ways. This study is one of only a few to examine mediators of depression prevention programs, and it identified several potentially important interpersonal targets for depression prevention. Although the comparative effects of the two interventions on the interpersonal mediators were marginally significant or non-significant, the mediation results indicate that reducing interpersonal conflicts and difficulties in social functioning led to improvements in depressive symptoms and overall functioning. Thus, although the findings do not concretely demonstrate that IPT-AST causes improvements in these facets of interpersonal functioning, they highlight the benefits of targeting interpersonal vulnerabilities in depression prevention programs. These findings also raise questions about the potential utility of taking a personalized analytic approach to identifying mediators of intervention effects. IPT-AST (as well as other interventions) is a multi-faceted intervention that targets multiple aspects of interpersonal functioning, and the primary targets of the intervention may vary across adolescents. As a result, testing for common mediators across the whole sample may mask effects that would be revealed using a more personalized analytic approach. This could explain why significant mediation effects have been elusive in the intervention literature. As the field continues to focus on mechanisms of psychosocial interventions, we may need to consider alternative approaches to examining mediators of these multi-component interventions.

Footnotes

Dear Dr. White,

Thank you for the opportunity to revise and resubmit our manuscript, “The Depression Prevention Initiative: Mediators of Interpersonal Psychotherapy – Adolescent Skills Training,” (MS ID: JCCAP-2019–0056.R2) to the Journal of Clinical Child & Adolescent Psychology.

We apologize for not properly formatting the abstract. We have now revised the abstract so that it is in compliance with the journal’s structured format.

Thank you again for the opportunity to revise and resubmit our manuscript to the Journal of Clinical Child & Adolescent Psychology. Please feel free to contact me with any questions.

Sincerely,

The Authors

References

  1. American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorders (4th ed., text rev.). Washington, DC: Author. [Google Scholar]
  2. Baumeister RF, Bratslavsky E, Finkenauer C, & Vohs KD (2001). Bad is stronger than good. Review of General Psychology, 5, 323–370. [Google Scholar]
  3. Brunwasser SM, Freres DR, & Gillham JE (2018). Youth cognitive-behavioral depression prevention: Testing theory in a randomized controlled trial. Cognitive Therapy and Research, 42, 468–482. [PMC free article] [PubMed] [Google Scholar]
  4. Cassidy J, & Shaver PR (Eds.). (2016). Handbook of attachment: Theory, research, and clinical applications (3rd ed.). New York, NY: Guilford. [Google Scholar]
  5. Collins WA, & Laursen B (2004). Changing relationships, changing youth: Interpersonal contexts of adolescent development. The Journal of Early Adolescence, 24, 55–62. [Google Scholar]
  6. Collins WA, Welsh DP, & Furman W (2009). Adolescent romantic relationships. Annual Review of Psychology, 60, 631–652. [DOI] [PubMed] [Google Scholar]
  7. Compas BE, Champion JE, Forehand R, Cole DA, Reeslund KL, Fear J, & Merchant MJ (2010). Coping and parenting: Mediators of 12-month outcomes of a family group cognitive–behavioral preventive intervention with families of depressed parents. Journal of Consulting and Clinical Psychology, 78, 623–634. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Davila J, Steinberg SJ, Kachadourian L, Cobb R, & Fincham F (2004). Romantic involvement and depressive symptoms in early and late adolescence: The role of a preoccupied relational style. Personal Relationships, 11, 161–178. [Google Scholar]
  9. Duong MT, Kelly BM, Haaland WL, Matsumiya B, Huey SJ, & McCarty CA (2016). Mediators and moderators of a school-based cognitive-behavioral depression prevention program. Cognitive Therapy and Research, 40, 705–716. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Furman W & Buhrmester D (1985). Children’s perceptions of the personal relationships in their social networks. Developmental Psychology, 21, 1016–1022. [Google Scholar]
  11. Hames JL, Hagan CR, & Joiner TE (2013). Interpersonal processes in depression. Annual Review of Clinical Psychology, 9, 355–377. [DOI] [PubMed] [Google Scholar]
  12. Hankin BL, Young JF, Abela JR, Smolen A, Jenness JL, Gulley LD, & Oppenheimer CW (2015). Depression from childhood into late adolescence: Influence of gender, development, genetic susceptibility, and peer stress. Journal of Abnormal Psychology, 124, 803–816. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Hayes AF (2009). Beyond Baron and Kenny: Statistical mediation analysis in the new millennium. Communication monographs, 76, 408–420. [Google Scholar]
  14. Hayes AF, & Rockwood NJ (2017). Regression-based statistical mediation and moderation analysis in clinical research: Observations, recommendations, and implementation. Behaviour Research and Therapy, 98, 39–57. [DOI] [PubMed] [Google Scholar]
  15. Hetrick SE, Cox GR, Witt KG, Bir JJ, & Merry SN (2016). Cognitive behavioural therapy (CBT), third‐wave CBT and interpersonal therapy (IPT) based interventions for preventing depression in children and adolescents. Cochrane Database of Systematic Reviews, Issue 8, Art. No.: CD003380. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Imai K, Keele L, & Tingley D (2010). A general approach to causal mediation analysis. Psychological Methods, 15, 309. [DOI] [PubMed] [Google Scholar]
  17. Joyner K, & Udry JR (2000). You don’t bring me anything but down: Adolescent romance and depression. Journal of Health and Social Behavior, 41, 369–391. [PubMed] [Google Scholar]
  18. Kaufman J, Birmaher B, Brent D, & Rao U (1997). Schedule for affective disorders and schizophrenia for school-age children-present and lifetime version (K-SADS-PL): Initial reliability and validity data. Journal of the American Academy of Child & Adolescent Psychiatry, 36, 980–988. [DOI] [PubMed] [Google Scholar]
  19. Kazdin AE (2007). Mediators and mechanisms of change in psychotherapy research. Annual Review of Clinical Psychology, 3, 1–27. [DOI] [PubMed] [Google Scholar]
  20. Kazdin AE (2009). Understanding how and why psychotherapy leads to change. Psychotherapy Research, 19, 418–428. [DOI] [PubMed] [Google Scholar]
  21. Kim-Cohen J, Caspi A, Moffitt TE, Harrington H, Milne BJ, & Poulton R (2003). Prior juvenile diagnoses in adults with mental disorder: Developmental follow-back of a prospective-longitudinal cohort. Archives of General Psychiatry, 60, 709–717. [DOI] [PubMed] [Google Scholar]
  22. La Greca AM, & Harrison HM (2005). Adolescent peer relations, friendships, and romantic relationships: Do they predict social anxiety and depression? Journal of Clinical Child and Adolescent Psychology, 34, 49–61. [DOI] [PubMed] [Google Scholar]
  23. Larson RW, Clore GL, & Wood GA (1999). The emotions of romantic relationships: Do they wreak havoc on adolescents? In Furman W, Brown BB & Feiring C (Eds.), The development of romantic relationships in adolescence; the development of romantic relationships in adolescence (pp. 19–49). New York, NY: Cambridge University Press. [Google Scholar]
  24. Merry SN, Hetrick SE, Cox GR, Brudevold-Iverson T, Bir JJ, & McDowell H (2011). Psychological and educational interventions for preventing depression in children and adolescents. Evidence-Based Child Health: A Cochrane Review Journal, 7, 1409–1685. [DOI] [PubMed] [Google Scholar]
  25. Mufson L, Dorta KP, Moreau D, & Weissman MM (2004). Interpersonal psychotherapy for depressed adolescents (2nd ed.). New York, NY: Guilford Press. [Google Scholar]
  26. National Institute of Mental Health. (2015). NIMH Strategic Plan for Research (NIH Publication No. 02–2650). Retrieved from http://www.nimh.nih.gov/about/strategic-planning-reports/index.shtml
  27. Preacher KJ (2015). Advances in mediation analysis: A survey and synthesis of new developments. Annual Review of Psychology, 66, 825–852. [DOI] [PubMed] [Google Scholar]
  28. Preacher KJ, & Hayes AF (2004). SPSS and SAS procedures for estimating indirect effects in simple mediation models. Behavior Research Methods, Instruments, & Computers, 36, 717–731. [DOI] [PubMed] [Google Scholar]
  29. Radloff LS (1977). The CES-D scale: A self-report depression scale for research in the general population. Applied Psychological Measurement, 1, 385–401. [Google Scholar]
  30. Reyes-Portillo JA, McGlinchey EL, Yanes-Lukin PK, Turner JB, & Mufson L (2017). Mediators of interpersonal psychotherapy for depressed adolescents on outcomes in Latinos: The role of peer and family interpersonal functioning. Journal of Latina/o Psychology, 5, 248–260. [Google Scholar]
  31. Robin AL, & Weiss JG (1980). Criterion-related validity of behavioral and self-report measures of problem-solving communication skills in distressed and non-distressed parent–adolescent dyads. Behavioral Assessment, 2, 339–352. [Google Scholar]
  32. Rogers AA, Ha T, Updegraff KA, & Iida M (2018). Adolescents’ daily romantic experiences and negative mood: A dyadic, intensive longitudinal study. Journal of Youth and Adolescence, 47, 1517–1530. [DOI] [PubMed] [Google Scholar]
  33. Rudolph KD, Flynn M, & Abaied JL (2008). A developmental perspective on interpersonal theories of youth depression. In Abela JRZ & Hankin BL (Eds.), Child and Adolescent Depression: Causes, Treatment and Prevention (pp. 79–102). New York: Guilford. [Google Scholar]
  34. Shaffer D, Gould MS, Brasic J, Ambrosini P, Fisher P, Bird H, & Aluwahlia S (1983). A Children’s Global Assessment Scale (CGAS). Archives of General Psychiatry, 40, 1228–1231. [DOI] [PubMed] [Google Scholar]
  35. Sheeber LB, Davis B, Leve C, Hops H, & Tildesley E (2007). Adolescents’ relationships with their mothers and fathers: Associations with depressive disorder and subdiagnostic symptomatology. Journal of Abnormal Psychology, 116, 144–154. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Stice E, Rohde P, Seeley JR, & Gau JM (2010). Testing mediators of intervention effects in randomized controlled trials: An evaluation of three depression prevention programs. Journal of Consulting and Clinical Psychology, 78, 273–280. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Ten Have TR, & Joffe MM (2012). A review of causal estimation of effects in mediation analyses. Statistical Methods in Medical Research, 21, 77–107. [DOI] [PubMed] [Google Scholar]
  38. Ten Have TR, Joffe MM, Lynch KG, Brown GK, Maisto SA, & Beck AT (2007). Causal mediation analyses with rank preserving models. Biometrics, 63, 926–934. [DOI] [PubMed] [Google Scholar]
  39. Weissman MM, & Bothwell S (1976). Assessment of social adjustment by patient self-report. Archives of General Psychiatry, 33, 1111–1115. [DOI] [PubMed] [Google Scholar]
  40. Weissman MM, Markowitz JC, & Klerman GL (2000). Comprehensive guide to interpersonal psychotherapy. New York, NY: Basic Books. [Google Scholar]
  41. Werner-Seidler A, Perry Y, Calear AL, Newby JM, & Christensen H (2017). School-based depression and anxiety prevention programs for young people: A systematic review and meta-analysis. Clinical Psychology Review, 51, 30–47. [DOI] [PubMed] [Google Scholar]
  42. World Health Organization. (2017). Depression and other common mental disorders: Global health estimates. Geneva: World Health Organization. [Google Scholar]
  43. Young JF, Benas JS, Schueler CM, Gallop R, Gillham JE, & Mufson L (2016). A randomized depression prevention trial comparing Interpersonal Psychotherapy-Adolescent Skills Training to group counseling in schools. Prevention Science, 17, 314–324. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Young JF, Gallop R, & Mufson L (2009). Mother-child conflict and its moderating effects on depression outcomes in a preventive intervention for adolescent depression. Journal of Clinical Child and Adolescent Psychology, 38, 696–704. [DOI] [PubMed] [Google Scholar]
  45. Young JF, Jones JD, Sbrilli M, Benas JS, Spiro C, Haimm C, Gallop R, Mufson L, & Gillham JE (2019). Long-term effects from a school-based trial comparing Interpersonal Psychotherapy-Adolescent Skills Training to group counseling. Journal of Clinical Child and Adolescent Psychology, 48(sup1), S362–S370. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Young JF, Kranzler A, Gallop R, & Mufson L (2012). Interpersonal Psychotherapy-Adolescent Skills Training: Effects on school and social functioning. School Mental Health, 4, 254–264. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Young JF, Mufson L, & Davies M (2006). Efficacy of Interpersonal Psychotherapy-Adolescent Skills Training: An indicated preventive intervention for depression. Journal of Child Psychology and Psychiatry, 47, 1254–1262. [DOI] [PubMed] [Google Scholar]
  48. Young JF, Mufson L, & Gallop R (2010). Preventing depression: A randomized trial of Interpersonal Psychotherapy-Adolescent Skills Training. Depression and Anxiety, 27, 426–433. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Young JF, Mufson L, & Schueler CM (2016). Preventing Adolescent Depression: Interpersonal Psychotherapy-Adolescent Skills Training. Oxford University Press. [Google Scholar]
  50. Zheng C, Atkins DC, Zhou XH, & Rhew IC (2015). Causal models for mediation analysis: An introduction to structural mean models. Multivariate Behavioral Research, 50, 614–631. [DOI] [PubMed] [Google Scholar]

RESOURCES