Therapeutic change may begin before the therapist and patient ever meet: Pre-treatment alliance changes predict clinical outcomes

Mechanisms of change are widely assumed to be activated by the onset of treatment. This foundational premise underlies decades of psychotherapy research and guides the timing of measurement and intervention in clinical science. The present study challenges this assumption by demonstrating that one of the most well-established predictors of treatment outcome—the therapeutic alliance—not only changes before therapy begins, but that these early, pre-treatment changes predict both alliance development and symptom reduction throughout treatment. In a randomized controlled trial for depression (N=100), the therapeutic alliance was assessed twice before patients met their therapist, and weekly thereafter. The pre-treatment change in alliance significantly predicted subsequent alliance dynamics and clinical improvement, above and beyond baseline levels. The findings remained robust even after controlling for pre-treatment outcome expectations and previous treatment experience. These findings suggest that mechanisms traditionally viewed as the result of treatment targeting may, in fact, be antecedents. By revealing pre-treatment shifts in a mechanism once thought to be non-existent prior to patient–therapist contact, this work may challenge fundamental assumptions in clinical psychology and highlights the need to reconceptualize therapeutic change as a dynamic, temporally extended process that may begin before any intervention has occurred. Such a shift has implications for how we define, measure, and intervene on core mechanisms of human change.