Purpose The purpose of this study was to determine behavioral and clinical outcomes of the DECIDE (Decision-Making Education for Choices in Diabetes Everyday) diabetes support program trial participants with and without a mental health (MH) history by treatment arm. Methods A secondary analysis was conducted of data from the DECIDE trial sample of urban African American adults with type 2 diabetes (T2DM; N = 137) who received the DECIDE diabetes support program in 1 of 3 delivery formats: self-study (n = 46), individual (n = 45), and group (n = 46). Positive screen on the Patient Health Questionnaire-2 and/or reported MH diagnosis were coded as MH history. Self-management, knowledge, problem-solving, and A1C data at baseline and 1 week and 6 months postintervention were analyzed for participants with and without MH history. Results Prevalence of MH history was 37% in the sample. Among those with no MH history, knowledge and problem-solving improved at 6 months postintervention in all intervention arms. For those with MH history, knowledge and problem-solving improved in the self-study and individual arms but not in the group arm. Clinically but not statistically significant changes in A1C were observed at 6 months. Conclusions In an urban minority T2DM sample, those with an MH history benefited from the intervention, but delivery format mattered, with robust improvements when participants with an MH history received self-directed or one-on-one formats rather than group.

Objective: To evaluate changes in behavioral and clinical outcomes from the DECIDE (Decision-making Education for Choices in Diabetes Everyday) self-management support program trial among participants with and without a mental health (MH) history. Methods: Participants were urban African American adults with type 2 diabetes (T2DM) (N=182). DECIDE is a problem solving diabetes self-management support program. Participants were randomized to 1 of 4 intervention arms: Self Study (n=46), Individual (n=45), Group (n=46), or Enhanced Print Education (EPE; n=45). Those who screened positive on the Patient Health Questionnaire 2 or reported a mental health diagnosis were coded as having MH history. Outcomes were self-management, knowledge, problem solving, and A1C assessed at baseline and 6 months post intervention, stratified by MH history status. Results: 72 participants (40%) reported a MH history, including depression, bipolar, and schizophrenia. Among those with no MH history, knowledge and problem solving improved at 6 months post intervention in all intervention arms (knowledge: Self Study p<.001, Individual p=.011, Group p=.004, EPE p=.026; problem solving: Self Study p=.023, Individual p=.015, Group p=.010, EPE p=.015). For those with MH history, knowledge and problem solving improved in the Self Study and Individual arms (diabetes knowledge: Self Study p=.018, Individual p=.037; health problem solving: Self Study p=.006, Individual p=.011) but not in the Group or EPE arms (ps≥.10). Although changes in A1C were not significant in any group (ps≥.20), A1C decreased by 1% among those with MH history in the Self Study arm (baseline A1C=9.1%; 6 month A1C=8.1%; p=.20). Conclusion: In an urban minority sample with T2DM, those with a MH history benefited from a diabetes self-management intervention, but delivery modality mattered. Robust improvements in skills and A1C were observed when those with a MH history had self-paced or one-on-one formats rather than group. Disclosure E.A. Vrany: None. A. Jennings: None. F. Hill-Briggs: None. Funding National Heart, Lung, and Blood Institute (R01HL089751, 5T32HL007180)