Family processes and behavioral outcomes in Functional Family Therapy: Results from a treatment study

Technical Summary: Family Processes and Behavioral Outcomes in Functional Family Therapy

Problem Statement and Theoretical Context
Functional Family Therapy (FFT) is a widely recommended intervention for adolescent disruptive behavior, operating on a phase-based model that posits changes in family functioning (e.g., reduced negativity, improved communication, increased cohesion) as the primary mechanism driving reductions in youth problem behaviors. While theoretical frameworks and prior research suggest that family functioning acts as a mediator in this process, empirical evidence directly testing this mediation pathway within FFT is limited. Specifically, there is a lack of studies examining whether changes in family functioning statistically account for (mediate) the observed reductions in aggressive and rule-breaking behaviors over time. This study addresses this gap by analyzing whether changes in family functioning mediate behavioral outcomes in a sample of Norwegian families receiving FFT in regular service settings.

Methodology

• Sample: The study utilized a subsample of 69 youths (mean age = 14.8 years; 43.5% female) and their families from a larger randomized controlled trial conducted in Norway between 2013 and 2017. The analytic sample was restricted to families who received a minimum of eight FFT sessions (the developer-defined threshold for treatment completion).
• Design: The researchers employed a within-participant analytic framework using data collected at three time points: pre-treatment (baseline), 6 months post-treatment start, and 18 months post-treatment start.
• Measures:
  • Outcomes: Aggressive and rule-breaking behaviors were assessed via parent reports on the Child Behavior Checklist (CBCL).
  • Mediators: Eight family functioning variables were measured using parent and youth reports: Family Environment Scale (cohesion, conflict), a social support scale, and the Inventory of Parent and Peer Attachment (trust, communication, alienation).
• Statistical Analysis: The study utilized two-condition within-participant mediation models (Montoya & Hayes, 2017) to estimate indirect effects. This approach separates within-participant change from between-participant differences. The analysis examined:
  1. Single Mediator Models: Testing indirect effects for each family functioning variable separately on short-term (6-month) and long-term (18-month) behavioral changes.
  2. Parallel Mediator Models: Testing combinations of mediators to detect potential additive effects.
  3. Exploratory Reversed Causality Models: Testing whether changes in behavior mediated subsequent changes in family functioning.
  • Significance was assessed using percentile bootstrapped confidence intervals.

Key Results

• Behavioral Outcomes: Significant improvements were observed in both aggressive and rule-breaking behaviors from pre-test to follow-up.
• Mediation Analysis (Primary Hypothesis): The study found no significant indirect effects through any of the eight family functioning mediators. Neither short-term nor long-term reductions in disruptive behavior were statistically mediated by changes in family cohesion, conflict, social support, trust, communication, or alienation.
• Mediator Changes: No significant average changes (a-paths) were detected across the eight family functioning measures during the initial 6 months.
• Direct Effects: Significant direct effects (c'-paths) remained, indicating that behavioral improvements occurred even when controlling for changes in family functioning and average mediator levels.
• Exploratory Reversed Causality: Analyses suggesting a reversed causal order yielded significant indirect effects. Specifically, reductions in aggressive and rule-breaking behaviors appeared to mediate subsequent improvements in family functioning (cohesion, conflict, and social support) in both the short and long term.
• Parallel Models: No significant total indirect effects were found in the parallel mediator models, suggesting that combinations of mediators did not collectively explain the behavioral changes.

Significance and Claims
The paper claims that its findings challenge the hypothesized "cascading" mechanism of FFT, where therapeutic changes in family interaction are expected to precede and cause reductions in youth problem behavior.

• Contradiction of Expectations: Contrary to the FFT theoretical model and previous literature suggesting family functioning is a central mediator, this study found no evidence that changes in family functioning mediated behavioral outcomes in this Norwegian sample.
• Temporal Ordering: The exploratory results suggest that, in this specific context, improvements in youth behavior may precede improvements in family functioning, rather than the reverse.
• Contextual Factors: The authors suggest that the lack of mediation may be related to the specific intake criteria in Norwegian Child Welfare Services, where families were referred primarily for youth behavior rather than assessed for family dysfunction. Consequently, families may have entered treatment with sufficient baseline family functioning to leave little room for measurable improvement, or the treatment may have targeted behavior directly without waiting for family system changes.
• Methodological Contribution: The study extends knowledge of FFT by explicitly modeling within-participant change processes in a real-world effectiveness setting, highlighting the complexity of testing mediation in routine care where control groups may receive similar family-based interventions.

The authors conclude that while FFT remains effective in reducing disruptive behavior in this sample, the specific theoretical pathway—that family functioning changes drive these behavioral reductions—was not supported. They call for further research to examine the conditions under which family functioning is altered by FFT and how these alterations relate to behavioral outcomes.