Effect of interventions for major depressive disorder and significant depressive symptoms in patients with diabetes mellitus: a systematic review and meta-analysis

Comorbid depression in diabetes is highly prevalent, negatively impacting well-being and diabetes control. How depression in diabetes is best treated is unknown. This systematic review and meta-analysis aims to establish the effectiveness of existing anti-depressant therapies in diabetes. PubMed, Psycinfo, Embase and Cochrane library. Study eligibility criteria, participants, interventions: randomized controlled trials (RCTs) evaluating the outcome of treatment by psychotherapy, pharmacotherapy or collaborative care of depression in persons with Type 1 and Type 2 diabetes mellitus. risk of bias assessment; data extraction. Synthesis methods: data synthesis, random model meta analysis and publication bias analysis. Meta analysis of 14 RCTs with a total of 1724 patients show that treatment is effective in terms of reduction of depressive symptoms: -0.512; 95% CI -0.633 to -0.390. The combined effect of all interventions on clinical impact is moderate, -0.370; 95% CI -0.470 to -0.271; it is large for psychotherapeutic interventions that are often combined with diabetes self management: -0.581; 95% CI -0.770 to -0.391, n=310 and moderate for pharmacological treatment: -0.467; 95% CI -0.665 to -0.270, n=281. Delivery of collaborative care, which provided a stepped care intervention with a choice of starting with psychotherapy or pharmacotherapy, to a primary care population, yielded an effect size of -0.292; 95% CI -0.429 to -0.155, n=1133; indicating the effect size that can be attained on a population scale. Pharmacotherapy and collaborative care aimed at and succeeded in the reduction of depressive symptoms but, apart from sertraline, had no effect on glycemic control. amongst others, the number of RCTs is small. The treatment of depression in people with diabetes is a necessary step, but improvement of the general medical condition including glycemic control is likely to require simultaneous attention to both conditions. Further research is neede