Effects of the multidomain non-pharmacological interventions on anxiety symptoms and depressive symptoms in older adults with suspected cognitive impairment: a propensity score matching and difference-in-differences analysis

Abstract

Background Mental health problems among older adults with cognitive impairment have attracted increasing attention. Although non-pharmacological interventions (NPIs) are considered safe and feasible, evidence regarding their effects on older adults with suspected cognitive impairment in Southwest China remains limited. Methods This intervention study was conducted in an urban district of Chongqing, China, including older adults with suspected cognitive impairment. A total of 159 individuals were assigned to the intervention group and 161 to the control group. The intervention group received routine health education in addition to a 2-month multidomain NPIs, while the control group received routine health education only. The multidomain NPIs included multiple components, such as music, physical exercise, and sensory stimulation. Data were collected through face-to-face interviews using validated assessment tools, including the Generalized Anxiety Disorder Scale and the 15-item Geriatric Depression Scale. Propensity score matching (PSM) was used to balance baseline characteristics, and difference-in-differences (DID) models estimated the net intervention effect. Difference-in-Difference-in-Differences (DDD) models further examined whether baseline cognitive status modified the effects on anxiety and depressive symptoms. Results In this study, the prevalence of anxiety and depressive symptoms was 24.38% and 10.94%, respectively. After PSM, 202 participants were included in the analysis. DID results showed that the multidomain NPIs significantly reduced anxiety symptoms (β = −1.44, 95% CI: −2.22, −0.65, P < 0.001) and depressive symptoms (β = −2.04, 95% CI: −2.83, −1.25, P < 0.001) among older adults with suspected cognitive impairment. Model 3 of the DDD analysis indicated that baseline cognitive status significantly moderated the effect of the multidomain NPIs on depressive symptoms (Treatment × Time × Cognitive status: β = 2.14, 95% CI: 0.25–4.04, P = 0.027). These findings suggest that, compared with older adults with baseline cognitive impairment, the intervention showed a greater effect on reducing depressive symptoms among older adults without objective cognitive impairment at baseline. Conclusion The multidomain NPIs may reduce anxiety and depressive symptoms in older adults with suspected cognitive impairment, with effects on depressive symptoms potentially moderated by baseline cognitive status. Integrating the multidomain NPIs into early cognitive impairment management may improve emotional well-being.