INTRODUCTION: Obstructive sleep apnea (OSA) is characterized by a complete or partial obstruction of the upper airway, along with hypoxemia, microarousals, and sleep fragmentation. Compelling evidence has clarified a bidirectional correlation between OSA and diabetes mellitus (DM). This paper was to assess the link between OSA and DM via meta-analysis, consisting of type 1 diabetes mellitus (T1DM) and type 2 diabetes mellitus (T2DM). MATERIALS AND METHODS: Four databases (PubMed, Cochrane Library, Embase, and CNKI) were screened from inception to March 2024 for observational studies of OSA and DM, including case-control studies and cohort studies. Bidirectional associations between OSA and DM were analyzed, consisting of T1DM and T2DM. Random-effect models were employed to determine the pooled odds ratio (OR) and 95% confidence intervals (CIs) to compare prevalence. Traditional subgroup analyses were implemented. Review Manager 5.3 and Stata 16.0 were utilized for data analyses. RESULTS: Thirty-five studies were enrolled, including 12 prospective cohort studies, 4 retrospective cohort studies, and 19 case-control studies. DM prevalence was notably higher in OSA patients than in non-OSA patients (OR: 2.29, 95% CI: 1.93-2.72), and OSA prevalence was notably higher in DM patients than in non-DM patients (OR: 2.12, 95% CI: 1.73-2.60). Subgroup analysis uncovered that DM prevalence in the OSA population was more significant in the group <
50 years (OR: 3.28, 95% CI: 2.20-4.89) and slightly decreased in the group >
50 years (OR: 1.82, 95% CI: 1.38-2.40). CONCLUSIONS: The meta-analysis reveals a bidirectional link between OSA and DM.