A Randomized, Controlled Multisite Study of Behavioral Interventions for Veterans with Mental Illness and Antipsychotic Medication-Associated Obesity

Zachary D. Erickson, Crystal L. Kwan, Hollie A. Gelberg, Irina Y. Arnold, Valery Chamberlin, Jennifer A. Rosen, Chandresh Shah, Charles T. Nguyen, Gerhard Hellemann, Dixie R. Aragaki, Charles F. Kunkel, Melissa M. Lewis, Neena Sachinvala, Patrick A. Sonza, Joseph M. Pierre, Donna Ames

Weight gain and other metabolic sequelae of antipsychotic medications can lead to medication non-adherence, reduced quality of life, increased costs, and premature mortality. Of the approaches to address this, behavioral interventions are less invasive, cost less, and can result in sustained long-term benefits. We investigated behavioral weight management interventions for veterans with mental illness across four medical centers within the Veterans Affairs (VA) Healthcare System. We conducted a 12-month, multi-site extension of our previous randomized, controlled study, comparing treatment and control groups. Veterans (and some non-veteran women) diagnosed with mental illness, overweight (defined as having a BMI over 25), and required ongoing antipsychotic therapy. One group received "Lifestyle Balance" (LB; modified from the Diabetes Prevention Program) consisting of classes and individual nutritional counseling with a dietitian. A second group received less intensive "Usual Care" (UC) consisting of weight monitoring and provision of self-help. Participants completed anthropometric and nutrition assessments weekly for 8 weeks, then monthly. Psychiatric, behavioral, and physical assessments were conducted at baseline and months 2, 6, and 12. Metabolic and lipid laboratory tests were performed quarterly. Participants in both groups lost weight. LB participants had a greater decrease in average waist circumference [F(1,1244) = 11.9, p < 0.001] and percent body fat [F(1,1121) = 4.3, p = 0.038]. Controlling for gender yielded statistically significant changes between groups in BMI [F(1,1246) = 13.9, p < 0.001]. Waist circumference and percent body fat decreased for LB women [F(1,1243) = 22.5, p < 0.001 and F(1,1221) = 4.8, p = 0.029, respectively]. The majority of LB participants kept food and activity journals (92%), and average daily calorie intake decreased from 2055 to 1650 during the study (p < 0.001). Behavioral interventions specifically designed for individuals with mental illness can be effective for weight loss and improve dietary behaviors. "Lifestyle Balance" integrates well with VA healthcare's patient-centered "Whole Health" approach. ClinicalTrials.gov identifier NCT01052714.