Psychiatry · Medical Literature
Symptom Trajectories in Patients With Anxiety Disorders During Transdiagnostic Cognitive Behavioral Therapy.
Abstract
Growing research confirms the efficacy of transdiagnostic cognitive behavioral psychotherapy, that is treatments applicable across multiple diagnoses within a disorder class or spectrum, by examining group-level changes from pre- to posttreatment. However, less is known about interindividual differences and dynamic changes during therapy. Understanding different symptom trajectories and their associations with treatment outcome would allow early detection of nonresponders and treatment adjustments. Using data from the treatment group of a randomized controlled trial, we aimed to investigate trajectories in transdiagnostic cognitive behavioral therapy (CBT) with the Unified Protocol (UP) in patients with anxiety disorders. Patients reported on demographics, clinical history, coping strategies, treatment motivation, and expectations before starting transdiagnostic CBT. Participants provided symptom ratings across 16 sessions of the UP as well as at posttreatment, 6-, and 12-month follow-up (N = 64, 71.9% female, age: M = 32.67, SD = 11.91). Using latent class growth analysis (LCGA), we identified symptom trajectory classes and tested whether these could be predicted from baseline data. We also examined associations between classes and treatment outcome, using multilevel modeling and reliable change indices (RCIs). LCGA revealed four distinct trajectory classes: low severity-improved (n = 21, 32.81%), high severity-improved (n = 19, 29.69%), high severity-stagnant (n = 12, 18.75%), and low severity-worsened (n = 12, 18.75%). Only the number of comorbidities emerged as a significant predictor of class, with patients in the high severity-stagnant group having more comorbidities. They also showed higher symptom severity at baseline, posttreatment, and follow-ups compared to other classes, despite clear improvement. Latent classes of symptom changes during transdiagnostic CBT can be distinguished and important pretreatment factors identified, informing treatment selection and personalization. Particularly patients with high initial burden and a higher number of comorbidities may require monitoring and adaptive treatment strategies, which should be explored in future studies. Trial Registration: ClinicalTrials.gov identifier: NCT03945617.
Clinical relevance
This study addresses a critical gap in transdiagnostic CBT research by moving beyond group-level pre-to-post analyses to characterize interindividual symptom trajectories during active treatment with the Unified Protocol (UP) for anxiety disorders. The findings are directly relevant to the large and heterogeneous population of patients with anxiety disorders and psychiatric comorbidities, for whom one-size-fits-all treatment approaches frequently yield suboptimal outcomes. Clinicians need this granular trajectory data now to enable early identification of nonresponders and to justify adaptive, personalized treatment modifications before treatment failure is fully realized.
Key findings
1. Latent class growth analysis (LCGA) of 64 patients (71.9% female; mean age 32.67 ± 11.91 years) undergoing 16 sessions of the Unified Protocol identified four distinct symptom trajectory classes: low severity-improved (n=21, 32.81%), high severity-improved (n=19, 29.69%), high severity-stagnant (n=12, 18.75%), and low severity-worsened (n=12, 18.75%). 2. Approximately 37.5% of participants demonstrated an unfavorable treatment course, belonging to either the high severity-stagnant or low severity-worsened classes — underscoring a clinically significant nonresponse rate not captured by group-level analyses. 3. Number of psychiatric comorbidities was the only statistically significant baseline predictor of trajectory class membership, with patients in the high severity-stagnant group carrying a greater comorbidity burden than those in other classes. 4. The high severity-stagnant class exhibited greater symptom severity at baseline and maintained higher symptom burden at posttreatment and at both 6- and 12-month follow-up assessments compared to all other trajectory classes, despite demonstrating measurable improvement. 5. Treatment outcome was evaluated using both multilevel modeling and reliable change indices (RCIs), providing a methodologically rigorous, clinically meaningful threshold for assessing individual-level response beyond statistical significance. 6. Baseline variables including demographics, coping strategies, treatment motivation, and expectations did not significantly predict trajectory class, suggesting comorbidity load is a more robust early stratification marker than patient-reported pretreatment characteristics.
Practice implications
Clinicians delivering transdiagnostic CBT with the Unified Protocol should systematically assess psychiatric comorbidity burden at intake, as higher comorbidity load is the most robust predictor of a stagnant high-severity trajectory and warrants proactive treatment intensification or augmentation — such as more frequent monitoring, adjunctive pharmacotherapy, or step-up to higher-intensity care — rather than a wait-and-see approach. Given that nearly 1 in 5 patients followed a high severity-stagnant course despite active treatment, routine within-treatment symptom monitoring (e.g., session-by-session self-report) should be implemented to enable early detection of nonresponse, ideally using reliable change indices to distinguish true clinical change from measurement error. Future clinical protocols and RCTs should incorporate adaptive treatment designs that allow modification of the UP based on early trajectory signals, particularly for patients presenting with multiple comorbid anxiety or related disorders.