Response of blood glucose and GLP-1 to different food temperature in normal subject and patients with type 2 diabetes

The study addresses whether the ambient temperature of ingested glucose/food influences postprandial glycemia and glucose-regulating hormones in healthy individuals and in patients with type 2 diabetes mellitus (T2DM). While meal timing and frequency have recognized roles in glycemic control, food temperature has received less attention despite cultural variation in consumption of hot versus cold foods and drinks. Prior observations suggest temperature may affect gastric emptying and neural pathways (for example, heat-activated channels such as TRPV1), potentially altering glycemic and hormonal responses. The purpose was to compare blood glucose and glucose-responsive hormones during OGTTs conducted with hot versus cold glucose solutions in normal subjects and newly diagnosed untreated T2DM patients, and to assess daily glycemic profiles under hot versus room-temperature meals using CGM.

Previous work has linked gastric emptying rates to postprandial glycemia, with faster emptying correlating with higher glucose early after ingestion and inverse relations later. Beverage or meal temperature can modulate gastric emptying and gastrointestinal motility/electrical activity, with both cold (≈4–8 °C) and hot (≈50 °C) liquids slowing emptying relative to neutral (≈37 °C), and cold drinks taking longer to equilibrate to body temperature. Animal studies indicated early-phase insulin release may be temperature-sensitive. Environmental temperature also affects OGTT responses, with higher ambient heat associated with elevated glucose and insulin versus cold environments, potentially through changes in glucose uptake (e.g., shivering, sympathetic activation). Taste and sweetness perception vary with temperature; cooling reduces perceived sweetness of glucose and may attenuate sweet taste receptor (TRPM5/STR) signaling, which has been implicated in modulating GLP-1 and other gut hormone secretion, though evidence is mixed. These findings suggest plausible pathways by which ingestate temperature could influence glycemic and incretin responses, but direct human endocrine comparisons with controlled ingestate temperatures had not been reported.

Design: Crossover, self-controlled study conducted in a controlled facility (ambient 22 °C). Participants: 19 healthy normal subjects and 22 newly diagnosed, untreated T2DM patients (June 2017–October 2019, Nanjing First Hospital). Inclusion criteria: ages 18–60 years; BMI 18.0–28.0 kg/m². Normals: no history of illness; fasting plasma glucose <6.1 mmol/L and 2-h glucose <7.8 mmol/L after 75-g OGTT (WHO 1999). T2DM: newly diagnosed by WHO criteria; HbA1c <86 mmol/mol (10.0%). Exclusion criteria: use/history of hypoglycemic agents; abnormal liver/kidney function; systemic corticosteroid use within 3 months; recent infections/acute events; pregnancy. Ethics: Institutional approval; informed consent; trial registration ChiCTR-OOC-17011643. Randomization and interventions: After overnight fast (>10 h), participants were randomly assigned on day 1 at 08:00 to either a hot or cold OGTT, then crossed over the next day to the opposite temperature OGTT. OGTT preparation: 75 g anhydrous glucose in 300 mL water. Cold solution: cooled at 4 °C; hot solution: heated at 55 °C. Final administered temperatures verified: cold 6–8 °C; hot ~50 °C. Blood sampling: venous samples at 0, 5, 10, 30, 60, and 120 min for glucose, insulin, GLP-1 (7–36/7–37 amide), GIP, and cortisol. Post-OGTT diet: After hot OGTT, participants consumed hot foods/drinks (>42 °C; meals 45–55 °C) for lunch and dinner (11:00–12:00; 17:00–18:00), identical composition/calories both days; after cold OGTT, foods/drinks at room temperature (20–24 °C). Meals consumed within 30 min; subjects avoided opposite-temperature items. Measurements and assays: Plasma glucose by glucose oxidase (Modular E170, Roche); HbA1c by HPLC (Bio-Rad); insulin by chemiluminescent immunometric assay (Roche; ref 2.3–11.6 mU/L); cortisol by radioimmunoassay; GLP-1 and GIP by ELISA (USCN LIFE), in tubes with protease and DPP-IV inhibitors (intra-assay CV <10%, inter-assay CV <12%). CGM: Medtronic sensor inserted anterior abdomen one day before first OGTT; capillary glucose measured four times daily for calibration; interstitial glucose recorded every 5 min for 3 consecutive days; sensor removed 24 h after second OGTT. Participants maintained usual activities without strenuous exercise; those with capillary glucose ≥20.0 mmol/L at any time were removed. CGM outcomes: 24-h mean glucose (MBG), standard deviation of MBG (SDBG), coefficient of variation (CV), 24-h largest amplitude of glycemic excursion (LAGE), and time in range (TIR). Statistics: Normality assessed; data as mean ± SEM or median (IQR). Repeated-measures ANOVA compared hot vs cold OGTT responses; if significant (P<0.05), Bonferroni tests assessed time-point differences. Paired t-tests compared CGM metrics within subjects (hot vs cold days). Between-group differences by unpaired t-test or Mann–Whitney U. Two-sided alpha 0.05. Stepwise linear regression tested associations of ΔIAUC (hot minus cold) for glucose/GLP-1 with age, sex, BMI, HbA1c, and temperature difference. Sample size: Pre-study in six normals showed AUC glucose 241.67±89.37 (cold) vs 280.58±86.32 (hot); ≥17 subjects required for 80% power at α=0.05 (PASS).

- OGTT glycemia: Mean blood glucose was higher with hot versus cold OGTT in both groups (Normals: 6.79±0.14 vs 6.33±0.15 mmol/L, P=0.002; T2DM: 11.88±0.56 vs 11.35±0.51 mmol/L, P=0.007). Time-point differences: in normals, glucose higher at 5, 10, 30, and 120 min; in T2DM, higher at 5 min (all P<0.05). - Insulin and GLP-1: In normals, insulin and GLP-1 were higher with hot OGTT (Insulin: 46.73±3.90 vs 40.34±3.00 mU/L, P=0.028; GLP-1: 19.19±1.84 vs 16.99±1.34 pmol/L, P=0.047). Significant time-point increases: insulin at 5 and 10 min (P=0.002, P=0.027); GLP-1 at 10 and 30 min (P=0.037, P=0.023). In T2DM, there were no significant differences in insulin (34.81±6.03 vs 32.33±4.71 mU/L, P=0.317) or GLP-1 (15.40±0.89 vs 14.46±1.04 pmol/L, P=0.219). - GIP and cortisol: No significant temperature-dependent differences in either group (Normals: GIP 68.48±5.43 vs 72.05±6.77 pmol/L, P=0.138; Cortisol 104.8±8.67 vs 113.96±4.78 nmol/L, P=0.220. T2DM: GIP 50.74±5.04 vs 52.98±5.03 pmol/L, P=0.327; Cortisol 149.17±12.70 vs 144.47±9.19 nmol/L, P=0.619). - Incremental responses: Percentage increments showed higher glucose and GLP-1 with hot vs cold OGTT in normals (Incremental glucose: 129.17±2.53% vs 122.46±2.21%, P=0.015; Incremental GLP-1: 130.20±5.86% vs 111.36±7.56%, P=0.009). Percentage insulin changes were not significantly different in either group. - CGM outcomes: Hot-meal day vs cold-meal day—Normals: MBG 6.11±0.13 vs 5.84±0.11 mmol/L (P=0.021); SDBG 0.59±0.06 vs 0.48±0.05 mmol/L (P=0.043); no significant differences in CV, TIR, LAGE. T2DM: MBG 8.88±0.39 vs 8.46±0.38 mmol/L (P=0.022); SDBG, CV, TIR, LAGE not significantly different. - Regression: Across individuals, ΔIAUC (hot–cold) for glucose or GLP-1 was not significantly associated with age, sex, BMI, HbA1c, or the temperature difference between OGTTs.

The findings demonstrate that ingestate temperature modestly but consistently elevates post-challenge glycemia, with accompanying increases in insulin and GLP-1 in healthy individuals but not in newly diagnosed, untreated T2DM. This addresses the research question by showing that hotter glucose ingestion accelerates or enhances glycemic and incretin responses in normal physiology, whereas T2DM exhibits a blunted or absent incretin and insulin temperature effect, despite higher glycemia. Mechanistically, differences may reflect temperature-dependent gastric emptying kinetics—cold liquids equilibrate more slowly to body temperature and may slow early gastric emptying more than hot liquids, attenuating early glucose appearance and GLP-1 release. Environmental temperature literature suggests cold exposure increases peripheral glucose uptake (e.g., shivering, sympathetic activation), potentially contributing to lower glycemia with cold conditions. Temperature also modulates sweetness perception and sweet taste receptor signaling in the gut and taste buds, which could influence GLP-1 secretion. Clinically, while the glycemic differences are small and unlikely to affect OGTT-based diagnosis materially, they suggest fasting glucose may be a more temperature-agnostic diagnostic measure and that standardization of ingestate temperature could reduce variability in incretin studies. CGM data extend the OGTT findings to daily life: hot meals modestly increase average glucose in both groups, and increase short-term variability (SDBG) in normals. The lack of effect on variability metrics in T2DM is consistent with a dampened incretin/insulin response to temperature in diabetes.

Hot versus cold ingestion of glucose increases post-OGTT blood glucose in both healthy individuals and patients with T2DM, with significant increases in insulin and GLP-1 observed only in healthy subjects. Over a day, hot meals modestly raise average glucose, and increase short-term variability in normals. These results highlight ingestate temperature as a modifiable factor influencing glycemic and incretin responses, with effects diminished in T2DM. Future research should include a neutral/body-temperature control OGTT to benchmark effects, determine the optimal food temperature for glycemic control and GLP-1 secretion, and use intragastric administration to isolate gastrointestinal from orosensory influences. Longer-term studies are needed to assess clinical relevance for dietary guidance and diabetes management.

- Lack of a third OGTT at neutral/body or room temperature as a control, omitted to avoid potential severe hyperglycemia with three consecutive OGTTs in T2DM. - The optimal food/drink temperature for glycemic control and GLP-1 secretion was not determined. - Potential orosensory and salivary gland effects of temperature could not be excluded; intragastric administration would be needed to isolate these effects. - Short-term, single-center study with modest sample sizes; findings pertain to newly diagnosed, untreated T2DM and may not generalize to all populations.