Abstract
AIMS/HYPOTHESIS: Functional variants in the TAS2R38 bitter taste receptor influence bitter taste perception and inform dietary and lifestyle behaviors that impact glucose homeostasis. Experimental data suggest that TAS2R38 receptors also mediate postprandial GLP-1 secretions from intestinal L-cells, though human study data are limited. To further establish the role of TAS2R38 in glucose homeostasis in humans, we tested whether functional variants conferring greater TAS2R38 sensitivity associate with lower blood glucose, particularly in the postprandial state, independent of dietary- and lifestyle-mediated effects of TAS2R38 on glucose levels. METHODS: We analyzed participants without type 2 diabetes of European ancestry in the UK Biobank. We used known functional variants in TAS2R38 to assign canonical haplotypes (AVI, PAV) and diplotypes conferring low (AVI/AVI nontasters), moderate (AVI/PAV tasters) or high (PAV/PAV supertasters) TAS2R38 receptor sensitivity and bitter taste perception. Linear models were used to quantify the associations of TAS2R38 diplotype with random glucose, and glucose levels over various time windows spanning postprandial and fasting states, adjusting for demographics and BMI, then sequentially for dietary and lifestyle factors. We used variants in other bitter taste receptors (TAS2R14 and TAS2R19), related to similar dietary and lifestyle behaviors as TAS2R38 but without hypothesized roles in glucose metabolism, to serve as negative controls. RESULTS: Among 218,688 individuals, 34%, 49% and 18% were AVI/AVI nontasters, AVI/PAV tasters and PAV/PAV supertasters, respectively. In BMI-adjusted models, each additional PAV haplotype associated with decreases in random glucose levels (beta [95% CI] = -0.07 [-0.13, -0.01] mg/dL; P = 0.021). In analyses stratified by fasting time, associations were only significant in the 0-2 hr 'postprandial' window (-0.24 [-0.39, -0.08] mg/dL per PAV haplotype; P = 0.003). These associations replicated (at the variant-level) in a published GWAS meta-analysis of 2-hr OGTT glucose and persisted after further adjustment for dietary and lifestyle behaviors (adjusted beta = -0.22; P = 0.004). Variants in TAS2R14 and TAS2R19 were related to similar behavioral traits as TAS2R38 but were not associated with 0-2 hr glucose, supporting behaviorally-independent effects of TAS2R38 diplotypes on 0-2 hr glucose. CONCLUSIONS/INTERPRETATION: Functional variants conferring greater TAS2R38 receptor sensitivity were associated with lower glucose levels in the postprandial state. These findings align with experimental evidence supporting a functional role for TAS2R38 in postprandial glycemia, reinforcing it as a potential target for type 2 diabetes prevention and treatment.