Background <p>Diabetic gastroparesis is a chronic complication that increases the risk of perioperative aspiration. Despite adherence to current fasting guidelines, gastric emptying remains difficult to predict in patients with diabetes. However, comparative data on ultrasound-defined gastric contents and residual gastric volume in fasting patients with and without diabetes scheduled for elective surgery remain limited. This knowledge gap provides the rationale for the present study. Therefore, this study aimed to assess gastric contents and volume using gastric ultrasonography in patients with diabetes scheduled for elective surgery and to investigate clinical variables associated with a full stomach, indicative of an increased risk of aspiration.</p> Methods <p>In this prospective observational study, 39 patients with diabetes and 39 control participants, all of whom had fasted for at least 8 h, were enrolled. Preoperative gastric ultrasonography was performed to assess the Perlas grade, antral cross-sectional area (CSA, cm²), and estimated gastric volume (mL). High aspiration risk was defined using ultrasound-based surrogate markers, rather than clinically observed aspiration events, and was identified by the presence of solid or thick fluid contents (Perlas grade 3) or an estimated clear fluid volume exceeding 1.5 mL/kg. Gastric aspirate volume (mL) and pH were subsequently recorded. Among patients with diabetes, the associations of glycemic control (HbA1c), disease duration, and diabetes-related complications (ascertained from pre-existing medical records) on gastric parameters were evaluated. Statistical analyses included the independent-samples t-test or Mann–Whitney U test for group comparisons, chi-square test or Fisher’s exact test, as appropriate for categorical variables, Spearman correlation analysis, and receiver operating characteristic (ROC) curve analysis.</p> Results <p>In the diabetes group, antral CSA was significantly greater than that in the control group (10.91 ± 9.49 cm² vs. 5.11 ± 2.01 cm²; p &lt; 0.001), as were estimated gastric volume (94.21 ± 110.65 mL vs. 41.04 ± 29.78 mL; p = 0.008), aspirated gastric fluid volume (19.67 ± 14.49 mL vs. 8.69 ± 5.68 mL; p &lt; 0.001), Perlas score (median 0 [range 0–3] vs. 0 [range 0–0]; p &lt; 0.001), and gastric pH (2.56 ± 0.54 vs. 2.14 ± 0.85; p = 0.011). A full stomach was identified in 17.94% of patients in the diabetes group (7/39) and in none of the control participants (0/39). Among participants classified as having high aspiration risk, the prevalence of diabetes-related neuropathy and the volume of aspirated gastric fluid were significantly higher. An internal exploratory ROC analysis indicated that a CSA threshold of ≥ 10 cm² may have descriptive value within this cohort (AUC: 0.862; 95% CI: 0.728–0.995, estimated using the DeLong method); however, as the outcome was partly derived from CSA, this threshold is descriptive and should not be interpreted as an independently validated diagnostic cutoff. No significant correlation was observed between HbA1c levels and gastric volume or gastric fluid pH, and no significant association was found between diabetes duration and gastric parameters.</p> Conclusion <p>Despite adherence to standard fasting protocols, patients with diabetes exhibited higher residual gastric volumes and a greater prevalence of a full stomach than control participants. Among patients with diabetes, neuropathy was more prevalent in those classified as having ultrasound-defined high aspiration risk than in those with low aspiration risk (57.1% vs. 9.4%; p = 0.012). No significant correlation was identified between HbA1c levels and gastric volume, and no statistically significant association was found between glycemic control and ultrasound-defined aspiration risk. Therefore, these findings should be interpreted as exploratory rather than conclusive. Preoperative bedside gastric ultrasonography may provide additional information for aspiration risk assessment in selected patients with diabetes, but these findings require confirmation in larger studies.</p> Trial registration <p>The trial was registered at ClinicalTrials.gov (NCT06195488 first posted January 8, 2024). As participant enrollment commenced on November 1, 2022, prior to the registration date, this trial was retrospectively registered.</p>