Published online Jan 14, 2027. doi: 10.3748/wjg.125781
Revised: August 17, 2026
Accepted: October 10, 2026
Published online: January 14, 2027
Processing time: 112 Days and 15.9 Hours
Functional dyspepsia (FD) is associated with impaired gastroduodenal motility, duodenal low-grade inflammation, and epithelial barrier dysfunction. Standardization of massage pressure remains a methodological challenge, and the pressure-response characteristics of robot-assisted massage, particularly in relation to duodenal barrier-associated outcomes, remain insufficiently characterized.
To compare the effects of three standardized robot-assisted massage pressure levels on gastrointestinal motility and duodenal barrier-related outcomes in an FD-like rabbit model and to characterize associated changes in myosin light chain kinase (MLCK)/phosphorylated myosin light chain (p-MLC)/E-cadherin-related molecular markers.
In this randomized controlled animal pressure-response study, 50 New Zealand White rabbits were randomized to control, model, light-pressure, medium-pressure, and heavy-pressure groups (n = 10 each). An FD-like phenotype was induced using sucrose-iodoacetamide administration combined with irregular feeding. Robot-assisted massage was applied at Zhongwan (CV12) and bilateral Tianshu (ST25) once daily for 10 consecutive days. Body weight and food intake, gastrointestinal transit, duodenal histology and ultrastructure, serum biomarkers, quantitative polymerase chain reaction, Western blotting, and immunofluorescence were assessed.
Compared with the control group (C group), the model group (M group) showed lower body weight and food intake and reduced gastric emptying rate (38.76% ± 4.88% vs 74.21% ± 5.67%, P < 0.001) and small intestinal propulsion rate (42.94% ± 4.93% vs 67.56% ± 5.74%, P < 0.001). Among the intervention groups, gastric emptying rate was 55.13% ± 5.21%, 65.84% ± 5.09%, and 34.21% ± 4.35% in the light-pressure group (LP group), the medium-pressure group (MP group), and the heavy-pressure group (HP group), respectively, while small intestinal propulsion rate was 55.06% ± 5.24%, 60.48% ± 5.11%, and 36.05% ± 4.68%, respectively. The LP group and particularly the MP group improved both outcomes compared with the M group (P < 0.001), whereas the HP group was inferior to the M group for small intestinal propulsion rate (P < 0.05). Similar pressure-dependent patterns were observed in mucosal morphology, ultrastructure, inflammatory and barrier-related markers, and MLCK/p-MLC/E-cadherin-associated molecular measurements. The MP group showed the broadest favorable pattern across body weight, food intake, gastrointestinal motility, mucosal morphology and ultrastructure, serum inflammatory and barrier-related markers, and MLCK/p-MLC/E-cadherin-associated molecular measurements, with statistically significant improvements in several quantitative outcomes compared with the M group (P < 0.05). The LP group showed a more limited pattern of improvement, whereas the HP group showed no consistent benefit and several adverse changes. Across the three tested pressure settings, the overall response pattern appeared non-monotonic; however, no formal dose-response model was fitted.
Robot-assisted massage was associated with improved gastrointestinal motility and duodenal barrier-related outcomes in an FD-like rabbit model, with medium pressure producing the most favorable overall response. These changes were accompanied by alterations in MLCK, p-MLC, E-cadherin, and junctional proteins.
Core Tip: This study standardized robot-assisted massage pressure and showed a pressure-response pattern that was not monotonic across the three tested pressure levels in a functional dyspepsia-like rabbit model. Medium-pressure produced the most consistent improvements in body weight, food intake, gastrointestinal motility, duodenal structural integrity, inflammatory and barrier-related markers, and epithelial junction-associated proteins, whereas light-pressure produced less consistent improvements and heavy-pressure showed no consistent benefit, with unfavorable changes in several outcomes. These responses were accompanied by changes in myosin light chain kinase, phosphorylated myosin light chain, E-cadherin, and junctional proteins; however, these expression changes do not establish causal mediation through the myosin light chain kinase/phosphorylated myosin light chain/E-cadherin pathway.