Improving Cardiometabolic Health in Adolescent Girls With Elevated Anxiety: Web-Based Multisite, Pilot, and Feasibility Randomized Controlled Trial.

Abstract

Anxiety is common in adolescents with elevated BMI (kg/m²), affecting 33%-43% of adolescents with overweight or obesity (age- and sex-specific BMI ≥85th percentile ). Anxiety has been associated with poorer cardiometabolic health, and disinhibited eating in response to anxiety may be a key explanatory mechanism. Therefore, reducing anxiety in adolescents may improve disinhibited eating patterns and prevent worsening cardiometabolic health. This pilot study primarily aimed to assess the feasibility and acceptability of 2 virtual group psychological interventions, interpersonal psychotherapy (IPT) and cognitive behavioral therapy (CBT), in adolescent girls with elevated anxiety and above-average weight. Secondary aims were to describe changes in anxiety, eating behavior, and cardiometabolic health indices. This 2-site, 2-year, pilot and feasibility randomized controlled trial enrolled 12- to 17-year-old girls with elevated anxiety symptoms (State-Trait Anxiety Inventory for Children total score ≥32) and above-average weight (BMI ≥75th percentile for age and sex), who were in good general health and not concurrently receiving therapy or medication treatment. Girls were randomly assigned to 1 of 2 web-based arms: 12-week, psychologist-led, group IPT or CBT. In both arms, groups of 5-7 adolescents met for 90 minutes on secure Zoom (Zoom Communications, Inc) once per week. Anxiety, disinhibited eating, BMI indices, blood pressure, lipids, hemoglobin A_1c (HbA_1c), and fasting glucose were measured at baseline, 12-week follow-up, and 1-year follow-up. Feasibility of recruitment, retention, and study protocol, and acceptability and fidelity of IPT and CBT were evaluated. Of 47 eligible adolescents, 40 (85%) enrolled. Enrollment of the total sample (n=40) was completed in 7 months. Retention was 85% (n=17) in IPT and 100% (n=20) in CBT at 12-week follow-up. Protocol adherence exceeded 96% for all intervals. Missing data varied, with higher rates for phlebotomy (up to 30%). Acceptability ratings were above-average for likability (mean 3.03, SD 1.14) and credibility (mean 3.26, SD 0.75; 0-4 scale with 4=Highest) across arms. Most adolescents received ≥80% intervention dosage (IPT: 90%, CBT: 95%). Expert fidelity ratings of IPT were 78% (SD 11%) and CBT were 94% (SD 9%). From baseline to 1-year, IPT demonstrated improvements in general anxiety (95% CI -8.95 to -1.40), social anxiety (95% CI -12.37 to -1.88), and HbA_1c (95% CI -0.48 to -0.23); CBT demonstrated improvement in emotional eating (95% CI -1.18 to -0.18), and both arms demonstrated improvements in BMI percentile (effect sizes=-0.82 and -0.63, IPT and CBT, respectively). This pilot and feasibility trial provided initial support for the feasibility of recruitment and retention and acceptability of web-based group IPT and CBT, as well as yielded areas for optimizing protocol feasibility and intervention fidelity. These preliminary results inform the planning of larger-scale efficacy trials with powered samples to assess effects on BMI and cardiometabolic health indicators among adolescents with elevated anxiety and above-average weight.