Background <p>The Cardiac Metabolic Index (CMI) has recently gained attention as a new tool for assessing visceral fat. However, the link between CMI and obstructive sleep apnea (OSA) remains unclear. This study aimed to explore the relationship between CMI and OSA risk and to examine the potential statistical mediation role of diabetes in this connection.</p> Methods <p>The data for this study were obtained from the National Health and Nutrition Examination Survey (NHANES). Weighted multiple regression analysis was used to explore the correlation between CMI and OSA. Restricted cubic spline (RCS) curves were plotted to assess nonlinear associations, and threshold effect analysis was performed to enhance the translational value of the results. Statistical mediation analysis was used to assess the role of diabetes as a hypothesised mediator in the relationship between CMI and OSA. Subgroup analyses and interaction tests confirmed the robustness of the results.</p> Results <p>Among the 5,818 participants included in the study, 2,941 were diagnosed with obstructive sleep apnoea (OSA). Fully adjusted logistic regression analysis revealed a significant positive correlation between CMI and OSA risk. When CMI was below 0.41, each 1-unit increase in CMI was associated with a 16.48-fold increase in OSA risk. The association between CMI levels and OSA was more pronounced in men, individuals without cardiovascular disease, those with moderate to severe depressive symptoms, and participants with obesity. Statistical mediation analysis indicated that diabetes accounted for 35.7% of the total association between CMI and OSA.</p> Conclusion <p>This study indicates that there is a significant non-linear relationship between CMI levels and the prevalence of obstructive sleep apnoea (OSA), with diabetes playing a partial statistical mediation role between the two.</p>