BackgroundVolatile glucose levels after gastrectomy induce dumping syndrome, which adversely affects patient quality of life. This study aimed to evaluate the glycemic variability of proximal gastrectomy with double-tract reconstruction (PGDTR) as a function-preserving procedure. MethodsThis study used a continuous glucose monitoring (CGM) system to record glycemic profiles of patients who underwent PGDTR or total gastrectomy (TG) and compared them. Moreover, this study evaluated postgastrectomy syndrome, including dumping symptoms, between the PGDTR and TG groups using the 37-item Postgastrectomy Syndrome Assessment Scale (PGSAS-37) questionnaire. ResultsOf note, 44 patients underwent PGDTR, and 42 patients underwent TG, which included more advanced cases. CGM results showed that the SD, relative SD, and maximum drop in glucose level between 30 min and 2 h after a meal were smaller in the PGDTR group than in the TG group (14.81 vs 22.40 mg/dL [P < .001], 0.14 vs 0.20 mg/dL [P < .001], and 42.06 vs 117.67 mg/dL [P < .001], respectively). For nocturnal glucose levels, SD and percentage time below the range were smaller in the PGDTR group than in the TG group (11.76 vs 15.16 mg/dL [P = .005] and 11.25% vs 35.27% [P < .001]). The PGDTR group generally performed better than the TG group on all the PGSAS-37 questionnaire items. Patients in the PGDTR group without food inflow into the remnant stomach showed similar CGM results as those in the TG group but with stronger dumping symptoms. ConclusionFood inflow into the remnant stomach is essential for PGDTR to be a function-preserving procedure as it leads to the control of dumping symptoms and lower glucose level spikes.