The relative efficacy of sensory modalities in driving insulin release, salivary secretion, appetite, and food intake: A systematic review and meta-analysis

Background Food sensory stimulation may influence eating behavior, but the effects of different stimulation patterns have not been systematically compared. Objective To quantify the effects of food sensory stimulation on insulin release, salivary secretion, appetite, and food intake, and compare the efficacy of those sensory modalities. Methods We systematically searched five databases (PubMed, Web of Science, Cochrane Library, CINAHL, and Embase) for controlled human studies comparing visual, olfactory, gustatory, tactile, modified sham-feeding, or multisensory stimulation. Studies without an eligible comparator or relevant outcome were excluded. RCTs and non-RCTs were synthesized separately using random-effects meta-analysis when compatible numerical data were available and otherwise summarized qualitatively. Results We included 22 reports comprising 24 studies (6 RCTs, 18 non-RCTs) and 1008 adults (413 men, 595 women). Participants were predominantly healthy; two cohorts were women with obesity. The oldest cohort had a mean age of 72.4 years. RCTs showed increased salivary secretion (SMD = 1.04, 95% CI [0.64, 1.45], P < 0.0001) but no clear overall effect on food intake (SMD = 0.32, 95% CI [−0.25, 0.88], P = 0.2742). Insulin release and appetite were each assessed in one RCT and synthesized narratively. Non-RCTs indicated increases in insulin release (SMD = 0.95, 95% CI [0.24, 1.66], P = 0.0088), salivary secretion (SMD = 0.78, 95% CI [0.38, 1.18], P = 0.0002), appetite (SMD = 0.54, 95% CI [0.29, 0.79], P < 0.0001), and food intake (SMD = 0.19, 95% CI [0.02, 0.35], P = 0.0282). Conclusion RCT evidence suggests that food sensory stimulation increases salivary secretion but has uncertain effect on food intake. Non-RCT evidence indicates potential benefits across all four outcomes, although evidence for insulin release and appetite is limited. Further well-designed RCTs are needed to confirm these effects and clarify differences between sensory modalities.