Non-pharmacological interventions for older adults with depressive symptoms: a network meta-analysis of 35 randomized controlled trials

Objective: To assess the effectiveness of non-pharmacological interventions for seniors with depressive symptoms. Methods: A comprehensive literature search was performed. We conducted network meta-analysis in two ways, intervention classes (psychosocial, psychotherapy, physical activity, combined, treatment as usual) and individual intervention (11 categories). Whenever included studies used different scales, the different instruments were converted to the units of the scale most frequently used (the Geriatric Depression Scale), such that the effect size was reported as a mean difference (MD) with 95% confidence interval (CI). The risk of bias of RCTs included in this review was assessed according to the Cochrane Handbook. Bayesian NMA was conducted using R-3.4.0 software. Results: A total of 35 RCTs with 3,797 enrolled patients were included. Compared to conventional treatment, physical activity and psychotherapy resulted in significant improvements in depressive symptoms (MD: 2.25, 95%CrI: 0.99-3.56; SUCRA = 86.07%; MD: 1.75, 95% CrI: 0.90-2.64; SUCRA = 66.44%, respectively). Similar results were obtained for music (MD: 2.6; 95% CrI: 0.84-4.35;SUCRA = 80.53%), life review (MD:1.92; 95% CrI:0.71-3.14; SUCRA = 65.62%), cognitive behavioral therapy (MD: 1.27; 95% CrI: 0.23-2.38; SUCRA = 45.4%), aerobic (MD: 1.84; 95% CrI: 0.39-3.36; SUCRA = 63%) and resistance training (MD: 1.72; 95% CrI: 0.06-3.42; SUCRA = 59.24%). Network meta-regression showed that there were no statistically significant subgroup effects. Conclusions: Physical activity and psychotherapy demonstrated statistically significant superiority over conventional treatment. Music and life review therapy proved the most promising individual interventions. However, conclusions are limited by the lack of sufficient sample size and consensus regarding intervention categories and so an adequately powered study is necessary to consolidate these findings.