Quantitative assessment of the thickness of the buccal subcutaneous fat using computed tomography to detect diabetes mellitus

Aim: Diabetes mellitus (DM) is associated with a variety of complications, including retinopathy, nephropathy, neuropathy, and cardiovascular disease. However, detecting DM based on head and neck imaging findings remains challenging. Therefore, in this study, we aimed to quantitatively evaluate the thickness of the buccal subcutaneous fat using computed tomography (CT) for the purpose of detecting DM. Methods: This retrospective study included 240 patients (91 males, 149 females; mean age, 59.5±12.6 years; range, 17–86 years). DM was defined based on the diagnostic criteria of the American Diabetes Association, with glycosylated hemoglobin (HbA1c) levels ≥ 6.0%. Patients with type 1 DM, those with tumor-related lesions in the head and neck fat pads, and those with CT images showing metal artifacts were excluded from this study. Buccal subcutaneous fat was measured manually at the point closest to the masseter muscle in the axial slice with the thicker buccal subcutaneous fat. Spearman's correlation coefficient was used to examine the correlation between buccal subcutaneous fat thickness and HbA1c levels. Receiver operating characteristic (ROC) curves were used to determine the cut-off value for detecting DM. Results: A positive correlation was observed between the buccal subcutaneous fat thickness and HbA1c levels. ROC analysis revealed that a buccal subcutaneous fat thickness of ≥ 10.875 mm indicated DM, demonstrating high sensitivity (82.2%). Conclusions: Our results demonstrate the relationship between the buccal subcutaneous fat thickness and HbA1c levels. Thus, DM can be detected based on the buccal subcutaneous fat thickness on CT images.