Yifat Fundoiano-Hershcovitz ^1* Inbar Breuer Asher ¹ Halit Kantor ² Sandy Rahmon ² Marilyn D. Ritholz ³ David L. Horwitz ⁴ Omar Manejwala ¹ Pavel Goldstein ⁵

• ¹ DarioHealth, Caesarea, Israel
• ² School of Public Health, University of Haifa, Haifa, Haifa, Israel
• ³ Joslin Diabetes Center, Harvard Medical School, Boston, Massachusetts, United States
• ⁴ DLH Biomedical Consulting, Las Vegas, NV, United States
• ⁵ Integrative Pain Laboratory, School of Public Health, University of Haifa, Haifa, Israel

Depression can exacerbate diabetes by impairing self-care behaviors and increasing the risk of complications, however, the underlying mechanism is still unclear. Given the suggested associations between walking activity, depression status, and blood glucose levels, this study explores the intricate relationship between depression and blood glucose (BG) control, with a focus on walking activity as a behavioral mediator.The purpose of this study is to examine walking activity mediating role in depression's impact on BG levels, investigating and validating non-linear association between BG levels and walking activity.This retrospective real-world study demonstrates the potential of regular walking activity as a simple and accessible intervention to mitigate the negative effects of depression on BG levels in T2D and prediabetes. A cohort of 989 users with T2D and prediabetes, who regularly tracked their steps levels and BG levels for 12 months using the Dario digital health platform was evaluated. The mediating role of the monthly average number of steps on the relationship between the self-reported depression status and lagged monthly average BG was assessed.Additionally, the association between monthly walking activity and monthly average BG was tested using a piecewise linear mixed effects model.Users with self-reported depression demonstrated increased BG levels compared to users without depression (B=8.00, P=.01). The association between depression and monthly average number of steps was significant (B=-.27, P<.005) and monthly average number of steps significantly predicted the following months' average BG (B=-.81, P=.001), adjusting for depression. The monthly average number of steps significantly mediated the effect of selfreported depression on the following month's average BG (M=.22, P<.005). Further sensitivity analysis demonstrated model robustness over various periods. Finally, non-linear dynamics of walking activity over time was validated using unseen data showing a decrease in monthly average BG for users with over an average of 400 steps per day (B=-1.87, P<.01).This study shows how regular walking may reduce the negative impact of depression on BG levels in people with T2D. Our findings advocate for the integration of walking activity into treatment protocols as a cost-effective, accessible intervention strategy to improve glycemic management and depressive symptoms in this population.