People with depressive disorder have been reported to experience increased premature mortality and higher prevalence of diabetes mellitus compared with the general population. Diabetes may amplify the risk of premature death from diabetes-related causes, particularly cardiovascular disease (CVD). Earlier work produced mixed findings and was constrained by limitations such as reliance on self-reported depression screening tools that can misclassify subclinical symptoms or diabetes-related distress. The present systematic review and meta-analysis aimed to quantify the risk of mortality and specific diabetes-related complications in people with clinically identified depressive disorder and co-occurring diabetes (the depression-diabetes group) versus people with diabetes only (the diabetes-only group). Outcomes of interest included all-cause mortality, cause-specific mortality where reported, and occurrence of defined diabetes complications.
The authors searched Embase, MEDLINE, PsycInfo and Web of Science from database inception through 20 December 2024. Eligible studies directly compared mortality and complication outcomes between a depression-diabetes group and a diabetes-only group. Results were pooled using random-effects meta-analytic models. The investigators prespecified stratified analyses (subgroup analyses and meta-regression) to explore study-level characteristics including age, sex distribution, study period, geographic region, follow-up duration and the nature of the diabetes sample (for example incident versus prevalent diabetes). The review was registered with PROSPERO (CRD42024595145).
Twenty-six studies from nine geographic regions met inclusion criteria. When pooled, patients with both depressive disorder and diabetes had a higher risk of all-cause mortality compared with diabetes-only patients (relative risk [RR] = 1.30; 95% confidence interval [CI] 1.21–1.39). For cause-specific mortality, the pooled estimate indicated an increased risk of CVD-specific mortality in the depression-diabetes group (RR = 1.15; 95% CI 1.02–1.29). The authors report that elevated mortality risks were present across various regions and persisted over time in longitudinal analyses reported by included studies.
Pooled analyses also showed an increased overall risk of diabetes-related complications in the depression-diabetes group relative to the diabetes-only group (RR = 1.28; 95% CI 1.18–1.40). When examining specific complication types, the meta-analysis identified higher pooled risks for:
By contrast, pooled estimates indicated a lower likelihood of retinopathy in the depression-diabetes group (RR = 0.84; 95% CI 0.76–0.94). Rates were comparable between groups for several complication types: cerebrovascular complications (RR = 1.36; 95% CI 0.99–1.87), nephropathy (RR = 1.09; 95% CI 0.93–1.27) and peripheral vascular complications (RR = 0.97; 95% CI 0.79–1.18).
The elevation in complication risk was especially evident in studies sampling incident diabetes cases, which the authors interpret as potentially reflecting more advanced disease presentation or unstable glycaemic control at diagnosis in the presence of co-occurring depression.
The review included stratified and meta-regression analyses considering study-level moderators such as age, sex, follow-up duration, region and diabetes sample characteristics. Elevated risks for mortality and complications were observed across multiple regions and over time. However, the authors note appreciable between-study heterogeneity for some pooled outcomes and that stratified analyses did not fully account for observed heterogeneity. Specific sources of residual heterogeneity were not completely resolved by the reported subgroup analyses.
The pooled evidence indicates that patients with clinically identified depressive disorder who also have diabetes are at higher risk of premature death and of several diabetes-related complications—notably metabolic and cardiovascular complications—compared with non-depressed peers with diabetes. The authors suggest these patterns are consistent with overall poorer glycaemic control among people with co-occurring depression and diabetes and that this poorer control may contribute to the observed excess morbidity and mortality.
On a clinical level, the findings support the need for comprehensive, multipronged approaches to care for people with comorbid depression and diabetes. The authors recommend individualized risk estimation for diabetes-related outcomes and early, targeted interventions aimed at reducing avoidable physical morbidity and premature mortality in this vulnerable population.
The source notes prior studies were limited by depression ascertainment methods, especially reliance on self-report questionnaires that risked misclassification by capturing subclinical depressive symptoms or diabetes-related distress rather than clinical depressive disorder. Although the meta-analysis pooled data from studies meeting inclusion criteria, residual heterogeneity across studies remained and could not be entirely explained by stratified analyses. Details on study-level covariate adjustments, variation in diagnostic criteria for depression or diabetes, and other methodological differences across the included studies were not fully resolved in the pooled synthesis as reported in the abstract.
This meta-analysis appears in Epidemiology and Psychiatric Sciences (2026) with PMID 42693777 and DOI 10.1017/S2045796026100882. The review was registered with PROSPERO under CRD42024595145.