Meta-analysis finds inconsistent diabetes mortality link in COPD exacerbations

Photo by Kate / Unsplash

Frontiers in Medicine Published May 13, 2026 Medically reviewed May 13, 2026 DOI ↗ By Dr. Amelia Tan, PhD · Internal Medicine & Chronic Disease

Key Takeaway

Consider that diabetes may not be an independent mortality risk in AECOPD based on inconsistent meta-analysis findings.

This is a meta-analysis of 23,883 participants investigating the association between diabetes mellitus and mortality in patients with acute exacerbation of chronic obstructive pulmonary disease (AECOPD). The authors synthesized findings from various study designs, including case-control and cohort studies.

In a univariate model, diabetes was associated with an increased mortality risk (HR = 1.45, 95% CI 1.19 to 1.77, P < 0.0001). However, in a multivariate analysis, no significant association was observed (HR = 1.50, 95% CI 0.45 to 5.00, P = 0.513). The association was significant in case-control studies and in studies with a sample size greater than 1000, but not in cohort studies or smaller studies.

The authors note inconsistent associations between diabetes and mortality in AECOPD patients. They acknowledge that associations may be confounded by methodological factors and study design. Limitations include the lack of reported follow-up duration and the absence of safety data.

The authors state that current evidence does not support diabetes as an independent risk factor for mortality in AECOPD. They suggest the analysis provides an evidence-based foundation for clinical risk stratification and individualized intervention, but this should be interpreted cautiously given the inconsistent findings.

Study Details

Study typeMeta analysis

EvidenceLevel 1

PublishedMay 2026

View Original Abstract ↓

BackgroundAs one of the most prevalent comorbidities in chronic obstructive pulmonary disease (COPD), diabetes mellitus has drawn growing attention for its potential impact on clinical outcomes in acute exacerbation of COPD (AECOPD). However, the association between diabetes and mortality risk in AECOPD remains controversial. This meta-analysis synthesized available evidence to quantify the effect of diabetes on mortality among AECOPD patients, aiming to provide an evidence-based foundation for clinical risk stratification and individualized intervention.MethodsWe systematically searched PubMed, Embase, Cochrane Library, Web of Science, CNKI, and Wanfang databases up to March 2026 for cohort or case-control studies investigating the association between diabetes and mortality in AECOPD patients. Two researchers independently conducted literature screening, data extraction and quality evaluation. The Newcastle-Ottawa Scale (NOS) was used to assess risk of bias. Meta-analysis was performed using Stata 14.0 software. The hazard rations (HR) was used to combine the effect size, and corresponding 95% confidence interval (CI) was computed.ResultsA total of 9 studies were included, involving 23,883 participants. In the univariate model, diabetes increased the mortality risk of patients with AECOPD (HR = 1.45, 95%CI (1.19,1.77), P < 0.0001), whereas no significant association was observed in multivariate analysis (HR = 1.50, 95% CI (0.45, 5.00), P = 0.513). Subgroup analysis showed that diabetes increased mortality risk in hospitalized AECOPD patients but not in those admitted to intensive care units. A significant association was found in case-control studies but not in cohort studies, and in studies with a sample size > 1000 but not in smaller studies.ConclusionsThis meta-analysis demonstrates inconsistent associations between diabetes and mortality in AECOPD patients, which may be confounded by methodological factors and study design. Current evidence does not support diabetes as an independent risk factor for mortality in AECOPD. Further well-designed, adequately powered prospective cohort studies with rigorous adjustment for confounders are warranted to clarify the true prognostic impact of diabetes on AECOPD patients.