Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2017 Feb 1.
Published in final edited form as: J Fam Psychol. 2015 Sep 28;30(1):82–92. doi: 10.1037/fam0000147

Preventing Adolescent Depression with the Family Check-Up: Examining Family Conflict as a Mechanism of Change

Gregory M Fosco 1, Mark J Van Ryzin 2, Arin M Connell 3, Elizabeth A Stormshak 4
PMCID: PMC4742422  NIHMSID: NIHMS713044  PMID: 26414418

Abstract

Family-centered prevention programs are understudied for their effects on adolescent depression, despite considerable evidence that supports their effectiveness for preventing escalation in youth problem behavior and substance use. This study was conducted with 2 overarching goals: (a) replicate previous work that has implicated the Family Check-Up (FCU), a multilevel, gated intervention model embedded in public middle schools, as an effective strategy for preventing growth in adolescent depressive symptoms and (b) test whether changes in family conflict may be an explanatory mechanism for the long-term, protective effects of the FCU with respect to adolescent depression. This trial was conducted with 593 ethnically diverse families who were randomized to intervention (offered the FCU) or middle school as usual. Complier average causal effect (CACE) analysis revealed that engagers in the FCU evidenced less growth in depressive symptoms and family conflict from 6th through 9th grade and post-hoc analyses indicated that the FCU is related to lower rates of Major Depressive Disorder. The second set of analyses examined family conflict as a mechanism of change for families who participated in the FCU. Families who reported short-term intervention benefits had significantly less escalation in family conflict over the middle school years; in turn, growth in family conflict explained risk for adolescent depressive symptoms.

Keywords: Family Check-Up, family-centered prevention, adolescent depression, family conflict, mechanism of family intervention

Adolescent depression, and mood disorders in general, are quite prevalent; approximately 14% of adolescents in the United States qualifying as having a mood disorder (Merikangas et al., 2010). Developmentally, mood disorders are considerably less common in childhood, suggesting that adolescence is a period of risk for problems with depression (Kessler, Chiu, Demler, & Walters, 2005). Recognition of this developmental timing has stimulated the creation of several prevention programs that work with adolescents to target individual factors, such as coping strategies, social skills, and cognitions (e.g., Lewinisohn, Clarke, Hops, & Andrews, 1990; Rhode, Clarke, Mace, Jorgensen, & Seeley, 2004).

More recently, there has been growing recognition that the family system can either elevate risk for depression or be a source of resilience for adolescent stress (Restifo & Bögels, 2009). As such, family-based programs are being evaluated toward the prevention and reduction of adolescent depression. Because family-based prevention models have primarily been used to treat or prevent child and adolescent problem behaviors, such as noncompliance, conduct problems, antisocial behavior, and substance use (Kaslow, Broth, Smith, & Collins, 2012), logic models for these family interventions have focused on the family risk and protective factors for youth problem behaviors. Only recently have efforts been made to articulate a framework that can guide family-centered programs toward addressing risk factors for adolescent depression (Restifo & Bögels, 2009). Thus, the theoretical models guiding family interventions for depression are less-extensively developed and evaluated.

Growing evidence suggests that family involvement in treatment for adolescent depression can be beneficial (Asarnow, Scott, & Mintz, 2002; GS Diamond et al., 2010). Approaches range from a family-centered interventions, such as Attachment-Based Family Therapy (e.g., Diamond et al., 2010) to the addition of a parent adjunct to existing interventions, such as Resourceful Adolescent Parent Program (RAP-P; Pineda & Dadds, 2013). However, other research has documented challenges to family interventions for adolescent depression. In some cases, parents may find family-based treatment of adolescent depression as less credible or even aversive (Stein et al., 2001). Also, there can be difficulties with parental engagement in supplemental parent-training that has been added to individual treatment for depressed adolescents, with engagement rates as low as 10% (Clarke, Rhode, Lewinsohn, Hops, & Seeley, 1999; Lewinsohn et al., 1990; Schochet et al., 2001). Not surprisingly, poorly attended parent-training sessions in these studies did not lead to incremental improvements in treatment outcome over the child-focused treatment components alone. Thus, there is mixed support for family-focused clinical interventions for depression, with parent engagement as a key obstacle.

Prevention programs aimed at reducing family risk factors for adolescent depression also are gaining momentum. One approach has been to target youth in families with a depressed parent for family-based intervention services. Because parental depression poses a risk for children’s development and well-being, interventions have been developed to target processes such as children’s cognitions about their parents depression or targeting children’s depressogenic cognitions within a family context (e.g., Beardslee et al., 2003; Beardslee, Wright, Gladstone, & Forbes, 2007 Compas et al. 2009; Compas et al. 2012). However, interventions that target adolescents of depressed parents seem to have the most promising effects in short-term follow-up, with less reliable effects for long-term outcomes.

We focus on another potentially fruitful approach that draws on strengths from established family-centered interventions that focus on key family management skills and family engagement that may reduce risk for adolescent depression. We hypothesize that interventions designed to prevent substance use or problem behavior target common risk factors that may also reduce risk for adolescent depression. Several family-centered programs now have documented effects for adolescent depression, such as Iowa Strengthening Families Program (Trudeau et al., 2007), Preparing for the Drug Free Years (Mason et al., 2007), the Family Check-Up (Connell & Dishion, 2008), and the Strong African American Families–Teen Program (Brody et al., 2012).

Further evaluation of family programs that may have preventive effects on depression would widen the scope of the public health benefits they already offer. It could also evoke interest in the cost-effective application of existing family-strengthening approaches to problems other than those classified in the externalizing spectrum. However, it is only through exploring the mechanisms that we will be able to understand how and why family-based interventions impact adolescent depression (Liddle, 2004). Such research is both important and lacking (Fosco, Bumbarger, Bamberger, & Van Ryzin, in press; Sandler et al., 2011), and the rare exceptions offer important insights into how family-centered programs designed to address conduct problems or substance use may also impact adolescent depression. Mason and colleagues (2007) found that the Preparing for the Drug Free Years may reduce risk through reductions in substance use. Another study, by Perrino et al (2014) indicates that improved family communication promoted in the Familias Unidas program can reduce risk for depression.

We explore family conflict as another potential avenue by which family-centered interventions may reduce risk for adolescent depression. Family conflict is a widely studied and empirically supported risk factors for adolescent depression (e.g., David, Steele, Forehand, & Armistead, 1996; Formoso, Gonzales, & Aikens, 2000). Families with frequent arguments, escalating hostility, criticism, or anger, creates a stressful family environment that can undermine adolescents’ coping resources and increase their risk for depression (Birmaher et al., 1996). Longitudinal evidence supports the view that family conflict during adolescence is associated with increases in adolescent depression 1 year later (Sheeber et al., 1997).

The Family Check-Up Model: Implications for Preventing Adolescent Depression

The Family Check-Up (FCU; Dishion & Kavanagh, 2003; Dishion & Stormshak, 2007) is a family-centered, school-based preventive intervention designed to reduce coercive family interactions by increasing parents’ positive reinforcement, effective limit setting, monitoring, and effective family problem-solving skills. Studies evaluating the FCU have resulted in considerable evidence for its efficacy in preventing substance use and antisocial behavior outcomes (e.g., Stormshak et al., 2011; Van Ryzin & Dishion, 2012; Van Ryzin, Stormshak, & Dishion, 2012; for a review, see Fosco et al., 2014). The FCU also is related to other outcomes, including high-risk sexual behavior (Caruthers, Van Ryzin, & Dishion, 2014), and, the focus of our study, adolescent depression (Connell & Dishion, 2008; Stormshak et al., 2010). We hypothesized that the FCU achieves these broad effects because of its impact on common risk factors across domains of functioning (Connell & Dishion, 2008). Many family risk processes are common across depression and conduct problems, including low levels of parental warmth and support and high levels of coercive family dynamics and family conflict (Dishion & Patterson, 2006; Sheeber, Hops, & Davis, 2001). Thus, by virtue of the fact that the FCU is designed to reduce coercive family interactions, it may also have collateral benefits in reducing adolescents’ risk for developing depression.

The Family Check-Up Intervention Model

The FCU is a tiered intervention model embedded in middle schools that uses a multiple gating approach for the identification of students at risk (Dishion & Kavanagh, 2003). At the universal level, a family resource center (FRC) is established in schools to provide infrastructure for family services, to offer general parenting information, and to send home information to parents about their students. Selected-level family services are intended for families with youth that are exhibiting some risk. Families are invited to seek out services if they believe they will be helpful. In addition, FRC staff seek out families for participation if their student has exhibited risk factors in school (e.g., behavior problems, attendance problems, academic decline). The family is invited to participate in a three-meeting assessment that includes an interview, a family assessment, and a feedback session that focuses on motivation to improve parenting. Then, a menu of family services is offered to support parents’ efforts to improve their family management practices, communication, and relationships. Interventionists work collaboratively with parents to identify the most appropriate intervention services on the basis of systematic assessments of family interactions, available resources, and parent motivation (Dishion & Stormshak, 2007). As opposed to traditional intervention strategies where all participants receive identical services, the FCU is guided by an adaptive intervention framework and tailors intervention services to the individual needs of each family in an effort to optimize efficiency by allocating only the needed intervention resources to each family (Collins et al., 2004).

Evaluating Tiered Interventions

The tiered intervention design used by the FCU model in schools calls for thoughtful attention to the implications of different methods for evaluating its effectiveness. An intent-to-treat (ITT) approach compares the outcomes for all participants in the control group with the outcomes for all participants assigned to the intervention group, regardless of their actual level of engagement with treatment services. A few studies have found ITT effects for the Family Check-Up model, when evaluating substance use (Connell, Dishion, & Deater-Deckard, 2006) and antisocial behavior (Smith, Noble, Zerr, Dishion, & Stormshak, 2014), and intervention targets, such as family relationship quality (Caruthers, Van Ryzin, & Dishion, 2014; Smith et al., 2014) and adolescent self-regulation (Fosco, Frank, Stormshak, & Dishion, 2013). These studies suggest that FCU can have net effects at the population level, regardless of the design intended to only engage families with indications of risk (i.e., selected, indicated levels).

Other studies have focused on a closer match between the tiered intervention design and the analytic framework. ITT analyses implicitly assume that all participants allocated to the intervention group will receive the intervention. However, the FCU’s tiered approach is designed around the expectation that most families will not need a family intervention, and thus offers a cost-effective allocation of resources to families who need them. In this approach, engagement in the intervention, particularly at selected and indicated levels of intervention, is what drives intervention effects (e.g., Connell & Dishion, 2008). To evaluate this approach, many of the studies testing the FCU’s efficacy have utilized complier average causal effect (CACE) modeling, which uses indications of engagement and non-engagement among the intervention group (Jo, 2002). Using mixture modeling techniques, observed compliance behavior in the intervention group is used to estimate latent classes (compliers and non-compliers) in the control condition that match those observed in the intervention group. Subsequently, intervention effects can be calculated as the difference between the outcomes in the intervention compliers group and those of the matched latent class of “complier” families in the control group. As one might expect given the adaptive nature of the FCU, the non-compliers are typically families whose children are not demonstrating significant problems (Van Ryzin et al., 2012); likewise, compliers may be more motivated to make changes in their family than non-compliers. CACE models offer insight into the degree to which the intervention is effective with those families that engage in support services as compared to a relevant group of families in the control condition.

This Study

To understand how and why the FCU might reduce risk for adolescent depression, analyses were conducted in two steps. The first set of analyses tested for FCU intervention effects for adolescent depressive symptoms and family conflict in the same model. Using ITT and CACE analytic approaches, we evaluated the degree to which the FCU might impact risk for depression for youth in middle school. These findings would replicate previous studies indicating effects for depression (Connell & Dishion, 2008) and family conflict (Van Ryzin & Dishion, 2012). Assuming significant effects of the FCU on depression and family conflict, the second set of analyses were conducted to test family conflict as a mechanism for long-term protective effects of the FCU with respect to adolescent depression. Given the complexity of CACE models, it is understandable that tests of mediation within an intervention framework are still evolving (Jo, 2008). Therefore, we conducted within-group analyses of intervention engagers, using previously validated methods (Fosco, Van Ryzin, Stormshak, & Dishion, 2014). A latent growth modeling (LGM) approach to examine how caregivers’ reported improvements in FCU-targeted adolescent behaviors (e.g., self-management, behavior concerns), labeled FCU response, were related to trajectories of family conflict from sixth through eighth grades, which in turn would predict adolescent depression in ninth grade.

Method

Participants

Participants were 593 adolescents and their families from three public middle schools in an urban area of the country. All three middle schools in this study served an at-risk population, with 35%, 89%, and 39% of families across the three schools receiving free/reduced-price lunch. All three schools were Title 1 schools, and approximately 20% of the school populations combined qualified for special education services. Youths and families were recruited in sixth grade across two cohorts. Parents of all sixth grade students were invited to participate in the study, and 80% of all parents agreed to do so. Consent forms were mailed to families or sent home with youths. Youth sex was evenly split in the sample (49% female). The ethnicity of the sample was as follows: European American (36%), Latino/Hispanic (18%), African American (16%), Asian (8%), American Indian (3%), and biracial/mixed ethnicity (19%).

A total of 386 families (65%) were randomly assigned to the intervention condition, and 207 families (35%) were randomly assigned to the control condition in which families experienced “school as usual,” (see Figure 1) including regular services offered by the schools but no access to any of the intervention services available to families in the intervention condition. No differences were found between intervention and control groups on any variables.

Figure 1.

Figure 1

Open in a new tab

Participant enrollment, randomization, and retention.

Student surveys were collected annually about all youths enrolled in the study. More than three quarters of the youths were retained across the 4 years of the study (Wave 2, n = 525, 89% of sample; Wave 3, n = 511, 86% of sample; Wave 4, n = 493, 83% of sample), 84% of the intervention group and 83% of the control group were retained by Wave 4.

Intervention Protocol

The FCU is a tiered intervention model. At the universal level, a family resource center (FRC) was established in each of the participating public middle schools; the FRC was available for services to the entire intervention group (see Stormshak et al., 2005 for more description). A parent consultant in each FRC provided information to parents about their student’s behavior, attendance, and homework completion and offered brief consultations to parents to support homework routines and home-to-school planning. Also, special seminars were provided about topics of interest to families in the school (e.g., supervising your teen during the summer).

The selected intervention was the FCU, a three-session intervention modeled on the Drinker’s Check-Up (Miller & Rollnick, 2002). Although all families were invited to participate in the FCU, concerted efforts were placed on engaging youth who were identified by school staff for behavior, attendance, or academic concerns. Families that participated in the selected intervention (42% of those in the intervention condition) completed three FCU sessions consisted of an initial interview, an assessment session, and a feedback session. The initial interview focused on exploring parent concerns and encouraging the parents to engage in a family assessment. The assessment session included survey assessments and videorecorded family interactions in families’ homes or at school to help therapists evaluate parent–child interactions. The feedback session was delivered using motivational interviewing strategies (Miller & Rollnick, 2002) to summarize the results of the assessment for parents and explore the need for indicated intervention services to target changes in family management practices at home. Indicated interventions were drawn from empirically validated sessions described in the Everyday Parenting curriculum (Dishion, Stormshak, & Kavanagh, 2011), which focuses on intensive support around four key domains of family management: positive behavior support, monitoring, limit setting, and family problem-solving.

Of the 386 families in the intervention condition, 52% (n = 199) received consultation from a parent consultant, the majority of which (n = 163) received the full selected-level FCU intervention (42% of the 386 families). Only families that completed the FCU and received feedback were considered “intervention compliers” for CACE analyses. Of the families receiving the FCU, 80% (n = 131) received indicated-level support after the feedback, such as parent skills training or development of a home-to-school plan. Specifically, 36% received positive behavior support, 71% received support in limit setting and monitoring skills, and 73% received communication and problem solving support. School-related support was received by 79%. The average intervention family received 365 minutes (or 6.1 hours) of intervention time. The majority of the FCU interventions were completed during seventh and eighth grades (73%). Interventionists received direct supervision from intervention developers, including direct consultation and feedback on video-recorded sessions FCU sessions to ensure high levels of fidelity to the intervention model. Observer ratings of fidelity indicated consistently high fidelity to the FCU model that were invariant across family racial groups (Bustos, 2011).

A basic consideration is the level of engagement of families of all demographics, such as gender and ethnic status. With respect to gender, 47% (n = 76) of families who engaged in the FCU had a girl, and 53% (n = 87) of families had a boy as a target of the intervention. The percentages of families engaging in the intervention by family ethnicity were as follows: 47% (n = 33) Latino families, 42% (n = 59) European American families, 36% (n = 22) African American families, 51% (n = 37) youths from multiethnic backgrounds, and 7% (n = 12) youths from other ethnic backgrounds engaged in the FCU.

Measurement Procedures

Measurement was conducted in two ways. Data for the full sample (intervention and control groups) were collected by means of adolescent self-report on in-school surveys during the spring semester of sixth through ninth grades. This assessment was an instrument developed and reported on by colleagues at the Oregon Research Institute (Metzler, Biglan, Rusby, & Sprague, 2001). If students moved out of their original schools, we followed them up at their new location. Each youth who participated received $20 for each year he or she completed the assessment. These data were the basis for our first aim. Additional information was collected as part of the FCU assessment session, thus, caregiver data were available from families who engaged in the FCU intervention. For the second aim, we were therefore able to include caregiver reports of family resources as a covariate, which has been found to be an important predictor of family conflict (Fosco et al., 2014). These caregiver assessments also were used to create the change in FCU-targeted adolescent behavior scores, as described later.

Measures

Family resources

Caregivers who completed FCU assessment materials provided information on families’ ecological context with respect to access to resources or stress involved in meeting basic needs, consistent with the FCU model (Dishion & Stormshak, 2007). Similar to measures of SES, this assessment surveyed families about circumstances such as income, employment, education, and use of financial aid programs and helped guide feedback about community resources that may benefit families. When data were available for both parents, the highest level of each variable among the two parents was chosen. These variables were standardized and averaged (α = .74) to create a single measure of family resources.

Parent Depression

Depression was assessed as a part of the FCU assessment, using the Center for Epidemiological Studies-Depression scale (CES-D; Radloff, 1977). The CES-D is a widely used measure of depression consisting of 20 items used to assess levels of symptoms during the past week. Sample items include “I felt depressed,” “I could not get going,” and “I felt that everything I did was an effort.” This scale had good internal consistency (α = .88).

Change in FCU-targeted adolescent behavior concerns

This index was created to capture caregivers’ concerns about their child’s behavior that they would like to see change. During the intake and 1-year follow-up assessments, parents reported about their concerns in four general domains: youth behavior problems (e.g., stealing, defiance), youth self-management (e.g., self-control with peers, managing feelings and emotions), youth positive behavior concerns (e.g., wishing youth would help out around the house, engage in prosocial activities), and internalizing behavior concerns (e.g., being sad or depressed, fearful or anxious). Each scale had adequate internal consistency (αs ranged from .61 to .70 on the three, four-item scales and from .73 to .78 for the eight-item Behavior Problem Concerns scale). Comparisons of those who provided data at T2 to those who did not indicated no differences in the groups. Confirmatory factor analysis indicated that these four domains converged on one factor. Following guidelines by McArdle and Nesselroade (1994), a latent change score was calculated for the latent variables over time, so that higher values reflected decreases in FCU-targeted adolescent behavior concerns from baseline to 1-year follow-up. The latent change model yielded good fit with the data and had statistically significant variance in the degree of change across families. These latent change scores were saved and used as a manifest variable for analysis in this study. Previous findings with this approach can be found elsewhere (Fosco et al., 2014).

Family conflict

Youth reports of conflict with parents were measured by averaging across four items. Youth were asked how many times in the past month did the following happen between you and someone in your family: “We got angry at each other”, “we argued”, “one of us got so mad, we hit the other person”, and “I got my way by being angry.” Responses were scored 0 (never), 1 (once), 2 (twice), 3 (3 times), 4 (4 or 5 times), 5 (6 or 7 times), or 6 (more than 7 times). This scale had good reliability across waves (α = .80–.82).

Depressive symptoms

Depressive symptoms were assessed annually using a 14-item measure adapted from the Child Depression Inventory (Kovacs, 1992) to more directly correspond to the Diagnostic and Statistical Manual for Mental Disorders IV (DSM-IV-TR; APA, 2000). Adolescents were asked about the frequency of symptoms such as feeling sad, moody, or hopeless and having trouble sleeping, with all DSM-IV-TR diagnostic criteria for depression addressed by at least one item. Participants rated each symptom on a 5-point scale ranging from never or almost never to always or almost always for the past month. This adapted measure has been shown to have high internal reliability (Stormshak et al., 2010). In this study, Cronbach’s alpha was between .92 and .95 for this measure across the four waves of data.

Child gender

Gender was coded as 0 = male and 1 = female in the analyses.

Intervention status

Random assignment was coded as 0 = control and 1 = intervention.

Engagement status

Consistent with CACE modeling guidelines (Jo, 2002), engagement status was coded with two dichotomous variables, “comply” and “noncomply” to reflect family participation in the FCU. Families in the intervention condition who elected to receive the FCU were coded as comply = 1 and noncomply = 0. Families in the intervention condition who did not receive the FCU were coded as comply = 0 and noncomply = 1. Families in the control group were coded identically so that the software could make its determination of latent class membership (Jo, 2002).

Analysis Plan

The first aim of this study was to examine FCU intervention effects on adolescent depression and the proposed mechanism (family conflict). For this test of the FCU intervention effects, we conducted two types of analysis, drawing on ITT and CACE modeling approaches. All analyses were conducted with Mplus 6.1, using full information maximum likelihood estimation to account for missing data (Muthén & Muthén, 2008). ITT analyses were conducted to evaluate the effects of the FCU as a package within schools, in a manner akin to a population-level effect. Significant findings would reflect an overall effect on the whole school population, rather than those families that were engaged by intervention services. To evaluate ITT effects, we fit growth models for adolescent depression and family conflict and regressed the slopes on intervention condition, controlling for ethnic minority status and adolescent gender.

A CACE analysis was also conducted to evaluate the specific effects of the FCU intervention model. The CACE model was computed to evaluate intervention effects for families who engaged with the FCU using latent growth model (LGM) embedded in CACE analyses (Jo, 2002). In CACE models, researchers can examine outcomes specifically for participants who engaged in (or “complied with”) the intervention and compare effects with the matched portion of the control group. Model fit is evaluated using entropy, which is a summary measure of the probability of membership in the most-likely class for each individual (i.e., complier or noncomplier). Possible values range from 0 to 1.0, with higher values representing better fit.

The LGM intercepts in the CACE model were parameterized to reflect the initial level of family conflict and depressive symptoms in sixth grade (prior to any intervention), and the slope was parameterized to reflect the linear rate of change from sixth through ninth grades. The intervention condition was used to predict the change over time (i.e., slope) in the complier class but not in the noncomplier class, since by definition this class did not receive the intervention. The intervention condition was not set to be a predictor of the initial levels (i.e., intercepts) in either class, and was not set as a predictor of membership in the latent compliance classes. Also, to ensure that the latent compliance classes were the same for each outcome variable under consideration, we ran both family conflict and depression in a single CACE model.

The second aim of this study was to examine family conflict as a mechanism of the FCU intervention effects on long-term adolescent depression outcomes using a subsample comprised only of the intervention engagers. First, an unconditional LGM was estimated for family conflict; then, accounting for gender and family resources as covariates, response to the FCU intervention (measured by short-term change in FCU-targeted adolescent behavior concerns) was tested as a predictor of the rate of long-term growth (slope) in family conflict over time. In turn, family conflict slope was tested as a predictor of adolescent depressive symptoms in ninth grade, accounting for previous levels at baseline (sixth grade).

Results

As a preliminary step, we compared non-complier (n = 223) and complier (n = 163) families in the intervention condition with regard to proportion of the adolescents who were female (non-compliers = 49.8%, compliers = 46.0%; χ2(1) = .534, n.s.) and of ethnic minority status (non-compliers = 65.4%, compliers = 62.6%; χ2(1) = .343, n.s.). We also compared the groups in terms of baseline family conflict (non-complier M = 2.67, SD = 1.57, complier M = 2.51, SD = 1.31; F(1, 384) = .51, n.s.) and depressive symptoms (non-complier M = 1.81, SD = .74, complier M = 1.87, SD = .82; F(1, 382) = 1.16, n.s.). Means and standard deviations for all variables are reported in Table 1. Both family conflict and depressive symptoms tended to increase across time, and girls reported more family conflict and depression than boys.

Table 1.

Correlations and Sample Descriptive Statistics

Family conflict (by grade)
 Depressive symptoms (by grade)
Variable 6 7 8 9 6 7 8 9
1. Intervention status .03 .00 .01 .01 .01 −.01 .01 .01
2. Gender .08 .13** .24*** .25*** .06 .19*** .30*** .26***
3. Ethnicity −.07 −.06 −.05 −.09 .02 −.04 −.02 −.09*
n 588 524 508 459 582 523 508 459
M 2.57 2.97 2.94 2.90 1.84 1.91 1.96 1.96
SD 1.45 1.46 1.42 1.33 .77 .80 .86 .85
Open in a new tab
*

p < .05,

**

p < .01,

***

p < .001.

Ethnicity was coded as a dichotomous variable such that 0 = European American/White, 1 = all other racial/ethnic groups.

FCU Intervention Effects on Family Conflict and Adolescent Depressive Symptoms

ITT Analyses

First, unconditional growth models were computed for adolescent depression and family conflict separately. The adolescent depression unconditional model provided acceptable fit with the data, in which χ2(5) = 10.185, p = .07, CFI = .99, TLI = .99, RMSEA = .04 (90% C.I.: .00–.08). On average, the slope for depression was increasing over time (M = .04, SE = .01, p < .01), and there was statistically significant variance in the slope (Var = .05, SE = .01, p < .01). The family conflict unconditional model yielded marginal fit, where χ2(5) = 44.22, p = .00, CFI = .94, TLI = .92, RMSEA = .12 (90% C.I.: .08–. 15). However, there were meaningful patterns of change and variance in the slope parameter. On average, family conflict increased over time (M = .09, SE = .02, p < .05), and there was statistically significant variance in the slope across families (Var = .13, SE = .02, p < .01). We then regressed intervention group (intervention or control group), youth gender, and ethnic minority status on the slope coefficients for adolescent depression and family conflict. No ITT effects were detected predicting depression (β = .00, p = 99) or family conflict (β = −.04, p = .57).

CACE Models

Second, we conducted CACE analyses. In this model, entropy was acceptable (0.79). Results for predictors of engagement status and the results for within-class variation in trajectories of both family conflict and depressive symptoms are shown in Table 2.

Table 2.

CACE Model Results

Within-class variation Within-class variation
Class
Membership		 Noncomplier class Complier class Noncomplier class Complier class
Complier vs.
noncomplier
logit (SE)		 Intercept
est. (SE)		 Slope
est. (SE)		 Intercept
est. (SE)		 Slope
est. (SE)		 Intercept
est. (SE)		 Slope
est. (SE)		 Intercept
est. (SE)		 Slope
est. (SE)
Variable/parameter
 Family conflict
 Depressive symptoms
Intervention status Fixed at 0 Fixed at 0 Fixed at 0 Fixed at 0 −.21* (.09) Fixed at 0 Fixed at 0 Fixed at 0 −.15* (.07)
Gender .02 (.03) .05* (.02) .02* (.01) .01 (.03) .02* (.01) .02* (.01) .02* (.01) .02 (.02) .02* (.01)
Ethnicity .06 (.21) −.24 (.15) .09 (.06) −.08 (.21) −.09 (.08) .06 (.08) −.07* (.04) −.02 (.13) .01 (.05)
Parameter intercept 2.49* (.16) −.05 (.07) 2.76* (.23) .19 (.13) 1.69* (.07) .00 (.03) 1.86* (.09) .05 (.06)
Residual variance 1.30* (.17) .12* (.03) 1.12* (.16) .11* (.03) .26* (.04) .05* (.01) .37* (.06) .05* (.01)
Open in a new tab
*

p < .05.

Compliance class membership

Results for predictors of engagement followed a logistic regression framework, in that predictors attempted to discriminate between the complier and noncomplier classes. Our results indicated that none of the predictors was significantly related to family compliance with the FCU (see Table 2), suggesting equal participation among ethnicities and families of boys and of girls.

Predictors of within-class variation

Of primary interest is the finding that assignment to the intervention condition predicted rate of change in both outcomes (see Table 2). The negative coefficients indicate that the intervention compliers possessed a significantly lower rate of change across time compared with compliers in the control group. As can be seen in Figure 2, the intervention compliers maintained a relatively flat level of family conflict, whereas the control compliers demonstrated an increase over time (2a). Similar effects appeared for depressive symptoms (2b). Control variables also predicted significant within-group variance in the expected directions; for example, girls had higher levels of depressive symptoms than did boys, both in terms of initial levels (i.e., intercept) and in terms of change over time (i.e., slope).

Figure 2.

Figure 2

Open in a new tab

Complier average causal effect results of the intervention effect on the development of family conflict and adolescent depressive symptoms from Grade 6 through 9.

Effect sizes

To calculate effect sizes, complier class estimates were saved and outcomes for individuals assigned to the complier class in the control and intervention groups were compared in ninth grade. Based upon Cohen’s criteria (i.e., moderate, d = .50; and large, d = .80), effect sizes were small for family conflict (.14) and depressive symptoms (.24). To assess clinical significance, we conducted a post-hoc analysis of symptom counts and approximated diagnoses of major depressive disorders (MDD). To do so, we re-scored depression items rated as “often” or “almost always” as symptoms, and summed the total number of depressive symptoms for each participant. In addition, diagnosis of MDD was assigned when youth reported having at least 5 symptoms of depression, with at least one being feelings of depressed mood or diminished interest in activities (APA, 2000). Symptoms counts and proportion of youths with MDD are reported in Table 3. Complier youth in the control condition had more symptoms of depression (t(172) = 2.34, p < .05) and nearly double the rate of MDD than the intervention group compliers by wave 4. However, categorical comparisons for MDD diagnosis failed to reach statistical significance (χ2(1) = 2.12, p = 15).

Table 3.

Symptom counts and proportion of major depressive disorder diagnoses by complier group and grade

Grade 6 Grade 7 Grade 8 Grade 9
M (SD) % MDD M (SD) % MDD M (SD) % MDD M (SD) % MDD
Intervention: compliers 1.47 (2.85) 13 1.48 (2.78) 13 1.74 (3.08) 13 1.34 (2.68) 10
Control: compliers 2.60 (3.91) 20 2.77 (3.90) 21 3.17 (3.89) 28 3.25 (4.42) 18
Intervention: non-compliers 1.32 (2.47) 09 1.66 (2.67) 11 1.97 (3.41) 13 2.11 (3.54) 16
Control: non-compliers 1.05 (2.15) 05 1.27 (2.68) 09 1.24 (2.61) 10 1.11 (2.23) 08
Open in a new tab

Testing Intervening Effects of Family Conflict on Adolescent Depression

Building on the intervention effects on family conflict and depressive symptoms when comparing compliers in the intervention and control groups, we then examined family conflict as an intervening process by which the FCU intervention may impact adolescent depressive symptoms. This was accomplished by evaluating the intervention complier group, including all families who initiated FCU services in schools. By using a measure of change in caregivers’ reports of adolescent behavior concerns that was validated in a previous study (Fosco, Van Ryzin, Stormshak, & Dishion, 2014), it was possible to test associations among short-term linear change in adolescent behavior concerns with trajectories of family conflict from sixth grade through eighth grade (the proposed mechanism) and ultimately adolescents’ levels of depression in ninth grade, after accounting for sixth grade depression. Table 4 reports correlations and descriptive statistics.

Table 4.

Correlations for the Structural Equation Model

1 2 3 4 5 6 7 8 9
1. Youth gender —
2. Family resources −.14 —
3. Parent Depression −.04 −.46* —
4. Depression Gr. 6 .01 −.08 .17* —
5. FCU-targeted behav. .00 .17 −.16 −.15 —
6. Family conflict Gr. 6 .03 .07 −.06 .50* −.04 —
7. Family conflict Gr. 7 .13 −.13 .02 .27* .04 .47* —
8. Family conflict Gr. 8 .15 −.08 .03 .37* .15 .36* .50* —
9. Depression Gr. 9 .19* −.06 .01 .42* −.05 .12 .17 .36* —
M — — .43 1.86 .01 2.50 2.82 2.88 1.82
SD — — .41 .81 .69 1.31 1.35 1.29 .77
n 163 163 160 158 115 160 155 146 123
Open in a new tab

The unconditional model provided marginal fit with the data, in which χ2(1) = 3.33, p = .07, CFI = .97, TLI = .92, RMSEA = .11 (90% C.I.: .00–.26). On average, the slope for family conflict was increasing over time (M = .22, SE = .06, p < .01), and there was statistically significant variance in the slope across families (Var = .26, SE = .12, p < .05). The full model, as shown in Figure 3, demonstrated good fit with the data, χ2(8) = 11.79, p = . 17; CFI = .98; TLI = .92; RMSEA = .05 (90% C.I.: .00–.11). Adolescent gender and family resources were included as covariates because of their association with family conflict or adolescent depression in previous studies (e.g., Fosco et al., 2014). None of these covariates was associated with change in FCU-targeted adolescent behavior concerns or with the slope in family conflict over time. Adolescent depressive symptoms were moderately stable over time (β = .46), and girls’ levels of depressive symptoms at ninth grade were higher than those of boys (β = .17). Consistent with our hypothesis, improvement in the FCU-targeted adolescent behavior concerns over the course of 1 year was associated with less growth in family conflict over time (β = − 20). In turn, increasing levels of family conflict during middle school were associated with higher levels of adolescents’ depressive symptoms by ninth grade (β = .47), after accounting for previous levels of depression in sixth grade. We then examined the indirect effect by which improvements in FCU-targeted adolescent behavior concerns were associated with depressive symptoms as a function of a reduction in the growth in family conflict. The standardized indirect effect was marginally significant (β = −.11, p = .08, 95% C.I.: −.02 to .24). Due to the relatively small sample for these analyses, we suspected that limited power may affect the test of indirect effects. Thus, we tested a model in which all nonsignificant predictive paths to depression were constrained to zero; this did not change model fit (χ2(4) = 4.79, ns), but the indirect effect achieved trend-level statistical significance, (β = −.08, p = .051, 95% C.I.: .00–.15).

Figure 3.

Figure 3

Open in a new tab

Family conflict as a mediator of the FCU and adolescent depression.

Note. Path coefficients reflect standardized betas

Nonsignificant paths are presented as dashed lines. *p < .05; Bold lines depict the hypothesized mediational pathway.

Model fit: χ2(8, N = 163) = 11.79, p = .16; CFI = .98; TLI = .92; RMSEA = .05 (90% CI: .00–.11)

Discussion

Family-centered interventions have been underutilized in efforts to prevent adolescent depressive symptoms (Restifo & Bögels, 2009). Effective interventions for prevention of problem behaviors in adolescents have articulated key aspects of family functioning that should be targeted for optimal benefits, such as the family environment (e.g., conflict, cohesion, bonding) and effective parenting skills (Kumpfer & Alvarado, 2003). Interestingly, many of these components also are implicated in the development of adolescent depression (Restifo & Bögels, 2009). As such, family-centered programs have potential to help forestall adolescent depression. Although the scant existing literature about family interventions for depression has indicated mixed results (and has highlighted the challenge of engaging families into treatments for depression), several family-focused prevention programs designed for adolescent conduct problems and substance use have been found to yield improvements in depressive symptoms (Brody et al., 2012; Mason et al., 2007; Trudeau et al., 2007).

Our study had two goals: (a) evaluate FCU intervention effects for adolescent depression using an ITT and CACE approach, and (b) examine family conflict as a possible intervening process in these intervention effects. Regarding the first goal, our findings did not find support for an overall ITT effect of the FCU on adolescent depression. However, within a CACE model approach, the FCU does appear to have enduring intervention effects over 4 years, from sixth through ninth grades, for those families who engage in the FCU. These findings are consistent with those of an earlier reported RCT study of the FCU for adolescent depression (Connell & Dishion, 2008) and replicated findings that revealed the FCU’s effectiveness for adolescent depression. Such findings bolster the evidence for the value of embedding a family-centered prevention program in public middle schools, in that the scope of benefits can be expanded from problem behaviors to include depressive symptoms.

Based on family models of depression (e.g., Sheeber et al., 1997), logic models of family interventions for depression (Restifo & Bögels, 2009), and findings that the FCU can reduce escalation in family conflict (Van Ryzin & Dishion, 2012), the second goal of this study was to examine family conflict as an intervening process linking engagement in the FCU with adolescent depression. We examined the subsample of families who participated in the FCU intervention and found that caregiver-reported improvements in FCU-targeted adolescent behavior concerns were related to long-term rates of growth in family conflict from sixth through eighth grades. Specifically, the degree to which the FCU helped parents address problems with adolescent behavior was associated with less escalation in family conflict over time. In turn, the rate of escalation in family conflict predicted change in adolescents’ depressive symptoms from sixth to ninth grade. These findings are the first test of an FCU intervention mechanism for adolescent depression and lend insight into the ways in which family-centered services targeting substance use and problem behavior can also diminish growth in adolescent depression.

These findings are a promising step toward a comprehensive prevention approach that can be embedded in public middle schools to maximize public health impact. They suggest the possibility of using a multilevel, tiered prevention approach to identify youths at risk for developing depression and directing them and their families to receive appropriate services before problems escalate and become entrenched. Through the use of an adaptive and tailored approach to intervention, the FCU offers a cost-effective and efficient approach that capitalizes on the school environment in which youths are observed for considerable periods of time, enabling early detection of problems (Dishion & Kavanagh, 2003).

This family-centered approach to reducing risk for adolescent depression may have advantages to other programs that focus primarily on the adolescent. By engaging parents at the start of the intervention, targeting their motivation to change in important family domains, and providing a strengths-based approach, parents may be more involved in implementing intervention strategies at home, and less likely to become defensive or resistant to intervention services. The current findings support a future direction for the FCU that broadens screening approaches to specifically identify risk for adolescent depression, and to develop depression-specific intervention modules for parents and adolescents.

Limitations

This study is not without limitations. It was limited primarily to adolescent self-report data. Although self-report data are preferred for measures of depression (e.g., Lagattuta, Sayfan, & Bamford, 2012), integrating a multimethod assessment of family conflict would strengthen our confidence in the findings. Our test of indirect effects is limited to a within-group analysis of the intervention group engagers, with a relatively small sample that had some retention issues. Although we found direct effects of the FCU intervention for family conflict over time, in our test of indirect effects, it was not possible to further compare this mechanism across intervention and control groups, thereby precluding causal interpretations. Further, interpretation of the indirect effects in this study warrant caution until replicated, given the marginally significant results.

Our post-hoc analyses should also be interpreted with caution. The measure of depression was not based on a diagnostic interview, it did not assess for thoughts of death or suicide (because the data collection occurred in schools), and focused on past month levels, rather than focusing on a two week period. Our intention was to provide a descriptive approximation of MDD diagnoses.

Finally, this study is limited to only one family mechanism that may account for adolescent depression. Other mechanisms warrant testing, such as positive family relations (Caruthers et al., 2014) and youth individual factors, such as self-regulation (e.g., Fosco et al., 2013), that have been found to explain linkages with other outcomes. In addition, other studies suggest that engaging in substance use or conduct problem behavior may be risk factors for later depression (e.g., Miettunen et al., 2014; O’Neil, Conner, & Kendall, 2011). Given the established effects of the FCU on antisocial behavior and substance use (e.g., Stormshak et al., 2011; Van Ryzin et al., 2012), future research might examine these processes as potential explanatory mechanisms for preventing risk for depression.

Conclusion

This study responds to the call for increased attention to mechanisms of family interventions (Sandler et al., 2011) and provides compelling support, through replication and extension of previous work, for the role of family-centered prevention in addressing adolescent depression symptoms. By capitalizing on the early-adolescence years, a noted developmental period of risk for onset and development of depression (Kessler, Chiu et al., 2005), family-centered prevention programs can have enduring effects with lasting implications for youth well-being. Although the current findings offer promise for preventing adolescent depression, the small effect sizes suggest that it would be valuable to develop depression-specific family and youth components that can be integrated into the menu of treatment options in the FCU.

Acknowledgments

This project was supported by grant DA018374 from the National Institute on Drug Abuse to Elizabeth Stormshak. We gratefully acknowledge the contributions of the Project Alliance staff, Portland public schools, and the participating youth and families, to the success of this project.

Contributor Information

Gregory M. Fosco, Pennsylvania State University

Mark J. Van Ryzin, University of Oregon, Child and Family Center

Arin M. Connell, Case Western Reserve University

Elizabeth A. Stormshak, University of Oregon, Child and Family Center

References

  1. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders DSM-IV-TR Fourth Edition (Text Revision) Author: American Psychiatr. 2000 [Google Scholar]
  2. Asarnow JR, Scott CV, Mintz J. A combined cognitive-behavioral family education intervention for depression in children: A treatment development study. Cognitive Therapy and Research. 2002;26(2):221–229. [Google Scholar]
  3. Beardslee WR, Gladstone TR, Wright EJ, Cooper AB. A family-based approach to the prevention of depressive symptoms in children at risk: Evidence of parental and child change. Pediatrics. 2003;112(2):e119–e131. doi: 10.1542/peds.112.2.e119. [DOI] [PubMed] [Google Scholar]
  4. Beardslee WR, Wright EJ, Gladstone TR, Forbes P. Long-term effects from a randomized trial of two public health preventive interventions for parental depression. Journal of Family Psychology. 2007;21(4):703. doi: 10.1037/0893-3200.21.4.703. [DOI] [PubMed] [Google Scholar]
  5. Birmaher B, Ryan ND, Williamson DE, Brent DA, Kaufman J. Childhood and adolescent depression: A review of the past ten years. Part I. Journal of the American Academy of Child and Adolescent Psychiatry. 1996;35:1427–1439. doi: 10.1097/00004583-199611000-00011. [DOI] [PubMed] [Google Scholar]
  6. Brody GH, Chen YF, Kogan SM, Yu T, Molgaard VK, DiClemente RJ, Wingood GM. Family-centered program deters substance use, conduct problems, and depressive symptoms in black adolescents. Pediatrics. 2012;129(1):108–115. doi: 10.1542/peds.2011-0623. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Bustos CE. Doctoral dissertation. University of Oregon; 2011. Parent experiences of a family-centered intervention: Examining ethnocultural group differences. [Google Scholar]
  8. Caruthers AS, Van Ryzin MJ, Dishion TJ. Preventing high-risk sexual behavior in early adulthood with family interventions in adolescence: Outcomes and developmental processes. Prevention Science. 2014;15(1):59–69. doi: 10.1007/s11121-013-0383-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Clarke G, Rhode P, Lewinsohn P, Hops H, Seeley J. Cognitive-behavioral treatment of adolescent depression: Efficacy of acute group treatment and booster sessions. Journal of the American Academy of Child and Adolescent Psychiatry. 1999;38:272–279. doi: 10.1097/00004583-199903000-00014. [DOI] [PubMed] [Google Scholar]
  10. Collins L, Murphy S, Bierman K. A conceptual framework for adaptive preventive interventions. Prevention Science. 2004;5:185–196. doi: 10.1023/b:prev.0000037641.26017.00. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Compas BE, Forehand R, Keller G, Champion JE, Rakow A, Reeslund KL, Cole DA. Randomized controlled trial of a family cognitive-behavioral preventive intervention for children of depressed parents. Journal of consulting and clinical psychology. 2009;77(6):1007. doi: 10.1037/a0016930. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Connell A, Dishion TJ. Reducing depression among at-risk early adolescents: Three-year effects of a family-centered intervention embedded within schools. Journal of Family Psychology. 2008;22(4):574–585. doi: 10.1037/0893-3200.22.3.574. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. David C, Steele R, Forehand R, Armistead L. The role of family conflict and marital conflict in family functioning. Journal of Family Violence. 1996;11:81–91. [Google Scholar]
  14. Diamond GS, Wintersteen MB, Brown GK, Diamond GM, Gallop R, Shelef K, Levy S. Attachment-based family therapy for adolescents with suicidal ideation: a randomized controlled trial. Journal of the American Academy of Child & Adolescent Psychiatry. 2010;49(2):122–131. doi: 10.1097/00004583-201002000-00006. [DOI] [PubMed] [Google Scholar]
  15. Dishion TJ, Kavanagh K. Intervening in adolescent problem behavior: A family-centered approach. New York, NY: Guilford; 2003. [Google Scholar]
  16. Dishion TJ, Nelson SE, Kavanagh K. The Family Check-Up with high-risk young adolescents: Preventing early-onset substance use by parent monitoring [Special issue] Behavior Therapy. 2003;34:553–571. [Google Scholar]
  17. Dishion TJ, Patterson S. Parenting young children with love, encouragement, and limits. New York, NY: Research Press; 2006. [Google Scholar]
  18. Dishion TJ, Stormshak EA. Intervening in children’s lives: An ecological, family-centered approach to mental health care. Washington, DC: American Psychological Association; 2007. [Google Scholar]
  19. Dishion TJ, Stormshak EA, Kavanagh K. Everyday parenting: A professional’s guide to building family management skills. Champaign, IL: Research Press; 2011. [Google Scholar]
  20. Formoso D, Gonzales NA, Aiken LS. Family conflict and children’s internalizing and externalizing behavior: Protective factors. American Journal of Community Psychology. 2000;28:175–199. doi: 10.1023/A:1005135217449. [DOI] [PubMed] [Google Scholar]
  21. Fosco GM, Bumbarger B, Bamberger KT, Van Ryzin MJ. Thinking systemically for enduring family change. In: Van Ryzin K, Kumpfer GM, Fosco, Greenberg M, editors. Family-Based Prevention Programs for Children and Adolescents: Theory, Research, and Large-Scale Dissemination. New York: Routledge; in press. [Google Scholar]
  22. Fosco GM, Frank JL, Stormshak EA, Dishion TJ. Opening the “black box”: Family Check-Up intervention effects on self-regulation that prevents growth in problem behavior and substance use. Journal of School Psychology. 2013;51(4):455–468. doi: 10.1016/j.jsp.2013.02.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Horowitz JL, Garber J. The prevention of depressive symptoms in children and adolescents: A meta-analytic review. Journal of Consulting and Clinical Psychology. 2006;74:401–415. doi: 10.1037/0022-006X.74.3.401. [DOI] [PubMed] [Google Scholar]
  24. Jo B. Estimation of intervention effects with noncompliance: Alternative model specifications. Journal of Educational and Behavioral Statistics. 2002;27:385–409. [Google Scholar]
  25. Jo B. Causal inference in randomized experiments with mediational processes. Psychological Methods. 2008;13:314–336. doi: 10.1037/a0014207. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Kaslow NJ, Broth MR, Smith CO, Collins MH. Family-based interventions for child and adolescent disorders. Journal of Marital and Family Therapy. 2012;38:82–100. doi: 10.1111/j.1752-0606.2011.00257.x. [DOI] [PubMed] [Google Scholar]
  27. Kessler RC, Chiu WT, Demler O, Walters EE. Prevalence, severity, and comorbidity of 12-month DSM-IV disorders in the National Comorbidity Survey Replication. Archives of General Psychiatry. 2005;62:617–627. doi: 10.1001/archpsyc.62.6.617. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Kovacs M. The Children’s Depression Inventory (CDI) Psychopharmacology Bulletin. 1985;21(4):995–999. [PubMed] [Google Scholar]
  29. Kumpfer KL, Alvarado R. Family strengthening approaches for the prevention of youth problem behaviors. American Psychologist. 2003;58:457–465. doi: 10.1037/0003-066X.58.6-7.457. [DOI] [PubMed] [Google Scholar]
  30. Lagattuta KH, Sayfan L, Bamford C. Do you know how I feel? Parents underestimate worry and overestimate optimism compared to child self-report. Journal of Experimental Child Psychology. 2012;113:211–232. doi: 10.1016/j.jecp.2012.04.001. [DOI] [PubMed] [Google Scholar]
  31. Lewinsohn PM, Clarke GN, Hops H, Andrews J. Cognitive-behavioral treatment for depressed adolescents. Behavior Therapy. 1990;21:385–401. [Google Scholar]
  32. Liddle HA. Family-based therapies for adolescent alcohol and drug use: Research contributions and future research needs. Addiction. 2004;99:76–92. doi: 10.1111/j.1360-0443.2004.00856.x. [DOI] [PubMed] [Google Scholar]
  33. Mason WA, Kosterman R, Hawkins JD, Haggerty KP, Spoth RL, Redmond C. Influence of a Family-Focused Substance Use Preventive Intervention on Growth in Adolescent Depressive Symptoms. Journal of Research on Adolescence. 2007;17(3):541–564. [Google Scholar]
  34. Merikangas KR, He JP, Burstein M, Swanson SA, Avenevoli S, Cui L, Swendsen J. Lifetime prevalence of mental disorders in US adolescents: results from the National Comorbidity Survey Replication-Adolescent Supplement (NCS-A) Journal of the American Academy of Child & Adolescent Psychiatry. 2010;49(10):980–989. doi: 10.1016/j.jaac.2010.05.017. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Metzler CW, Biglan A, Rusby JC, Sprague JR. Evaluation of a comprehensive behavior management program to improve school-wide positive behavior support. Education & Treatment of Children. 2001;24:448–479. [Google Scholar]
  36. Miettunen J, Murray GK, Jones PB, Mäki P, Ebeling H, Taanila A, Moilanen I. Longitudinal associations between childhood and adulthood externalizing and internalizing psychopathology and adolescent substance use. Psychological medicine. 2014;44(08):1727–1738. doi: 10.1017/S0033291713002328. [DOI] [PubMed] [Google Scholar]
  37. Miller WR, Rollnick S. Motivational interviewing: Preparing people for change. New York, NY: Guilford; 2002. [Google Scholar]
  38. Muthén LK, Muthén BO. Mplus user’s guide. 4th ed. Los Angeles, CA: 2008. [Google Scholar]
  39. O’Neil KA, Conner BT, Kendall PC. Internalizing disorders and substance use disorders in youth: Comorbidity, risk, temporal order, and implications for intervention. Clinical Psychology Review. 2011;31:104–112. doi: 10.1016/j.cpr.2010.08.002. [DOI] [PubMed] [Google Scholar]
  40. Perrino T, Pantin H, Prado G, Huang S, Brincks A, Howe G, Brown CH. Preventing internalizing symptoms among Hispanic adolescents: a synthesis across familias unidas trials. Prevention Science. 2014;15(6):917–928. doi: 10.1007/s11121-013-0448-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Pineda J, Dadds MR. Family intervention for adolescents with suicidal behavior: a randomized controlled trial and mediation analysis. Journal of the American Academy of Child & Adolescent Psychiatry. 2013;52(8):851–862. doi: 10.1016/j.jaac.2013.05.015. [DOI] [PubMed] [Google Scholar]
  42. Radloff LS. The CES-D scale a self-report depression scale for research in the general population. Applied psychological measurement. 1977;1(3):385–401. [Google Scholar]
  43. Restifo K, Bögels S. Family processes in the development of youth depression: Translating the evidence to treatment. Clinical Psychology Review. 2009;29(4):294–316. doi: 10.1016/j.cpr.2009.02.005. [DOI] [PubMed] [Google Scholar]
  44. Rhode P, Clarke GN, Mace DE, Jorgensen JS, Seeley JR. An efficacy/effectiveness study of cognitive-behavioral treatment for adolescents with comorbid major depression and conduct disorder. Journal of the American Academy of Child and Adolescent Psychiatry. 2004;43:660–668. doi: 10.1097/01.chi.0000121067.29744.41. [DOI] [PubMed] [Google Scholar]
  45. Sandler IN, Schoenfelder EN, Wolchik SA, MacKinnon DP. Long-term impact of prevention programs to promote effective parenting: Lasting effects but uncertain processes. Annual Review of Psychology. 2011;62:299–329. doi: 10.1146/annurev.psych.121208.131619. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Schochet I, Dadds M, Holland D, Whitefield K, Harnett P, Osgarby S. The efficacy of a universal school-based program to prevent adolescent depression. Journal of Clinical Child Psychology. 2001;30:303–315. doi: 10.1207/S15374424JCCP3003_3. [DOI] [PubMed] [Google Scholar]
  47. Sheeber L, Hops H, Alpert A, Davis B, Andrews J. Family support and conflict: Prospective relations to adolescent depression. Journal of Abnormal Child Psychology. 1997;25:333–344. doi: 10.1023/a:1025768504415. [DOI] [PubMed] [Google Scholar]
  48. Sheeber L, Hops H, Davis B. Family processes in adolescent depression. Clinical Child & Family Psychology Review. 2001;4(1):19–35. doi: 10.1023/a:1009524626436. [DOI] [PubMed] [Google Scholar]
  49. Stein D, Brent DA, Bridge J, Kolko D, Birmaher B, Baugher M. Predictors of parent-rated credibility in a clinical psychotherapy trial for adolescent depression. Journal of Psychotherapy Practice & Research. 2001;10(1):1–7. [PMC free article] [PubMed] [Google Scholar]
  50. Stormshak EA, Dishion TJ, Light J, Yasui M. Implementing family-centered interventions within the public middle school: Linking service delivery to change in student problem behavior. Journal of Abnormal Child Psychology. 2005;33:723–733. doi: 10.1007/s10802-005-7650-6. [DOI] [PubMed] [Google Scholar]
  51. Trudeau L, Spoth R, Randall G, Azevedo K. Longitudinal effects of a universal family-focused intervention on growth patterns of adolescent internalizing symptoms and poly-substance use: Gender comparisons. Journal of Youth and Adolescence. 2007;36:725–740. [Google Scholar]
  52. Van Ryzin MJ, Dishion TJ. The impact of a family-centered intervention on the ecology of adolescent antisocial behavior: Modeling developmental sequelae and trajectories during adolescence. Development and Psychopathology. 2012;24:1139–1155. doi: 10.1017/S0954579412000582. [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Van Ryzin MJ, Stormshak EA, Dishion TJ. Engaging parents in the Family Check-Up in middle school: Longitudinal effects on family conflict and problem behavior through the transition to high school. Journal of Adolescent Health. 2012;50(6):627–633. doi: 10.1016/j.jadohealth.2011.10.255. [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES