This study aimed to evaluate the gender-specific effect of a couple-based intervention on the management behaviors and mental well-being of community-dwelling older adults with type 2 diabetes mellitus during the COVID-19 partial lockdown in Guangzhou. Out of 207 participants involved in a prior randomized controlled trial (Trial no. ChiCTR1900027137), 156 (75%) completed the COVID-19 survey. Gendered differences in management behaviors and depressive symptoms between the couple-based intervention group and the patient-only control group were compared by distance to the high-risk areas cross-sectionally and longitudinally using random intercept models. Cross-sectionally, female patients of the intervention group had more positive behavior change scores (β = 1.53, p = 0.002) and fewer depressive symptoms (β = −1.34, p = 0.02) than the control group. Over time, female patients lived closer to the high-risk areas (<5 km) and showed decreasing depressive symptoms (β = −4.48, p = 0.008) in the intervention group vs. the control group. No statistically significant between-group difference was found for males. Females tended to benefit more from the coupled-based intervention than males did, particularly among these closer to the high-risk areas. Chronic disease management can be better sustained with active spousal engagement.