Use of a voice-activated virtual assistant was associated with reduced mental distress and modest improvements in glycemic control among older adults with type 2 diabetes, according to the IVAM-ED randomized clinical trial published in JAMA Network Open.
The open-label study, conducted at a single academic center in Brazil between June 2023 and February 2024, enrolled 112 adults aged 65 years or older with type 2 diabetes. Participants were randomly assigned to receive a smart speaker programmed with a 12-week behavioral health and diabetes support intervention or to continue usual care.
The intervention included medication and glucose monitoring reminders, daily health messages, and weekly educational audio content focused on diabetes self-management and mental health. The control group received standard outpatient care and written educational materials.
The primary outcome was mental distress at 12 weeks, measured using the Self-Reporting Questionnaire 20. Among 103 participants who completed follow-up (mean age, 73 years; 63% female), the adjusted mean questionnaire score was 6.29 in the intervention group compared with 7.75 in the usual care group (mean difference, −1.46).
Secondary outcomes included quality of life, diabetes self-care behaviors, perceived stress, and hemoglobin A1c (HbA1c). Quality-of-life scores improved in the intervention group (mean difference, 9.46 points), and adherence to self-care behaviors was higher (mean difference, 3.40).
At 12 weeks, HbA1c decreased by about half a percentage point in the intervention group compared with usual care. No significant differences were observed in blood pressure, lipid levels, or perceived stress.
No participants withdrew due to adverse events. Three deaths occurred during follow-up and were not considered related to the intervention.
The authors note that the open-label design, short follow-up, and requirement for home Wi-Fi access may limit generalizability. The findings suggest that voice-based digital tools may support diabetes self-management and may be associated with measurable changes in patient-reported outcomes and HbA1c over a 12-week period.
Source: JAMA Network Open