Abstract
BACKGROUND/AIMS: This study aims to assess whether specific symptoms from the Gastroparesis Cardinal Symptom Index (GCSI) could predict delayed gastric emptying and to establish the prevalence of gastroparesis (GP). METHODS: Diabetic patients were recruited from a multicenter cross-sectional cohort. Each participant underwent esophagogastroduodenoscopy, symptom assessment by GCSI, and solid-meal gastric emptying scintigraphy (GES). GP was defined by GES data showing a T(1/2) > 85 minutes and/or gastric retention > 8% at 3 hours. RESULTS: Among 138 patients, those with delayed GES (21.01%) had a higher incidence of nephropathy (51.72% vs 22.02%, P = 0.002) and lower albumin levels (3.91 g/dL vs 4.18 g/dL, P = 0.003). Cardinal symptoms such as nausea (P = 0.011), retching (P = 0.040), vomiting (P = 0.010), stomach fullness (P = 0.001), fullness after eating (P = 0.025), and loss of appetite (P = 0.039) were more prevalent in patients with abnormal GES. A higher overall GCSI score was found to independently predict delayed gastric emptying (P = 0.028) in multivariate analysis. The area under the receiver operating characteristic curve for the GCSI in predicting GP was 0.672, with an optimal cutoff value of 1.78 (sensitivity 79.31%, specificity 49.54%). The prevalence of GP, defined by both GCSI ≥ 1.78 and abnormal GES, was 16.67% among diabetic patients. Notably, 16 (11.59%) patients had rapid emptying. CONCLUSION: Diabetic patients exhibiting specific cardinal symptoms should be considered for GES evaluation, as the overall GCSI score independently predicts delayed gastric emptying. This study suggests that the GCSI may be useful as a screening tool rather than a diagnostic method for diabetic gastroparesis.