Abstract
Background
Diurnal glucose fluctuations are increased in prediabetes and might be affected by specific dietary patterns.
Objectives
The present study assessed the relationship between glycemic variability (GV) and dietary regimen in people with normal glucose tolerance (NGT) and impaired glucose tolerance (IGT).
Methods
Forty-one NGT (mean age: 45.0 ± 9.0 y, mean BMI: 32.0 ± 7.0 kg/m2) and 53 IGT (mean age: 48.4 ± 11.2 y, mean BMI: 31.3 ± 5.9 kg/m2) subjects were enrolled in this cross-sectional study. The FreeStyleLibre Pro sensor was used for 14 d, and several parameters of GV were calculated. The participants were provided with a diet diary to record all meals. ANOVA analysis, Pearson correlation, and stepwise forward regression were performed.
Results
Despite no difference in diet patterns between the 2 groups, GV parameters were higher in IGT than in NGT. GV worsened with an increase in overall daily carbohydrate and refined grain consumption and improved with the increase in whole grain intake in IGT. GV parameters were positively related [r = 0.14–0.53; all P < 0.02 for SD, continuous overall net glycemic action 1 (CONGA1), J-index, lability index (LI), glycemic risk assessment diabetes equation, M-value, and mean absolute glucose (MAG)], and low blood glucose index (LBGI) inversely (r = −0.37, P = 0.006) related to the total percentage of carbohydrate, but not to the distribution of carbohydrate between the main meals in the IGT group. A negative relationship existed between total protein consumption and GV indices (r = −0.27 to −0.52; P < 0.05 for SD, CONGA1, J-index, LI, M-value, and MAG). The total EI was related to GV parameters (r = 0.27–0.32; P < 0.05 for CONGA1, J-index, LI, and M-value; and r = −0.30, P = 0.028 for LBGI).
Conclusions
The primary outcome results showed that insulin sensitivity, calories, and carbohydrate content are predictors of GV in individuals with IGT. Overall, the secondary analyses suggested that carbohydrate and daily consumption of refined grains might be associated with higher GV, whereas whole grains and daily protein intake were related to lower GV in people with IGT.
Keywords: prediabetes, glycemic variability, dietary patterns, insulin sensitivity, normal glucose tolerance
Introduction
Glycemic variability (GV) defines intra- and interday glucose excursions, including hypoglycemic and hyperglycemic episodes [1]. GV has been suggested as an independent risk factor for developing chronic diabetes complications [2,3]. In people with type 2 diabetes mellitus (T2DM), GV is associated with cardiovascular disease risk even in the presence of well-controlled HbA1c [4]. GV develops early in the natural history of diabetes. Increased GV has already been found in people with both categories of prediabetes: impaired fasting glucose (IFG) and impaired glucose tolerance (IGT) [5]. These individuals also have a premature risk of diabetic microvascular disease [6]. The nature of GV at the early stage of diabetes is poorly understood, and most likely, multiple factors contribute to generating excessive glucose fluctuations throughout the day. Impairments in insulin secretion and sensitivity, along with lifestyle and nutrition habits, might play a key role in GV in people with IFG and IGT, whereas medications are unlikely to be involved because they are rarely used in these populations.
Available data for people with diabetes suggest that the proportion of macronutrients in the daily food regimen substantially impacts GV. For example, a low-carbohydrate, high-fat diet has been shown to improve GV in both T2DM [7,8] and type 1 diabetes mellitus (T1DM) [[9], [10], [11]] and increase the time range in the latter [[9], [10], [11]]. The daily proportion of macronutrients and their distribution between meals also seem to affect glucose excursions: a lower amount of carbohydrate at breakfast has been demonstrated to reduce postprandial hyperglycemic peaks during the day in T2DM [12]. Finally, the GI has also been found to affect GV [13,14].
Recent data suggest that even the order of food intake can impact glucose fluctuations. Ingesting proteins before carbohydrate has been reported to reduce GV in people with T2DM [15] and prediabetes, including IFG and IGT [16], although data are limited for the latter.
The aim of the current study was to investigate the role of dietary regimens on GV in people with normal and IGT. We hypothesized that people with normal glucose tolerance (NGT) and IGT might have different dietary patterns and that the quantity of macronutrients and energy consumption per day and their ratio in each main meal might be independently related to GV indices even at the early stage of dysglycemia. We tested our hypothesis using flash glucose monitoring (FGM) to derive GV parameters while carefully recording food intake within a daily routine diet.
Methods
Participants
We recruited 94 consecutive subjects (25 men and 69 women), with a mean age of 46.9 ± 10.4 y and a mean BMI of 31.6 ± 6.3 kg/m2, from those referred to the Division of Diabetology at the Department of Endocrinology, Medical University of Sofia, from August 2019 to August 2021, for being considered at high-risk for T2DM. People with a known diagnosis of T1DM or T2DM, or using glucose-lowering or other medications known to affect glucose metabolism, with serious comorbidities including kidney, liver, cardiovascular disease, and recent acute illness, were excluded. The study was approved by the Ethics Committee of the Medical University of Sofia, and written informed consent was obtained from all participants before entering the study.
All participants presented at the Division of Diabetology at 08:00, after an overnight fast, for the collection of medical history, anthropometric measurements for BMI calculation, and waist circumference measurements. Body composition was assessed by bioimpedance analysis (InBody 720 device; BioSpace), visceral fat area (VFA) was calculated as cm2, and total body fat (TBF) was expressed in %. After collecting baseline information, a peripheral vein was cannulated, and blood samples were collected for baseline (−20, −10, 0 min) measurements of plasma glucose and hormone concentrations, followed by the ingestion of a 75 g oral glucose load. Blood samples were then collected at 10, 20, 30, 45, 60, 90, and 120 min. Plasma glucose (hexokinase method), insulin, and C-peptide concentrations (electrochemiluminescence method) were measured at each time point. HbA1c was assessed by an National Glycohemoglobin Standardization Program (NGSP)-certified method.
Upon completion of all baseline measurements, all participants were given an FGM with a blinded sensor (FreeStyle Libre Pro, Abbot GmbH & KG). Data recording started after a 1-h warming-up period and continued for the next 14 d. All participants were carefully instructed on using the sensor and completing a daily food consumption diary (Supplemental Table 1). They were also advised to adhere to their usual diet regimen and to record each main meal and snack daily in detail, including time, type, and amount of food ingested. During the period to be recorded, all participants were also requested to have not >30 min of moderate physical activity per day. At the end of this period, all participants again attended the Division of Diabetology to upload FGM data into a computer system for subsequent analysis.
Calculations and statistical analysis
Based on the OGTT results, the study population was divided into 2 groups according to American Diabetes Association (ADA) criteria [17]: NGT: 9 men and 16 women with a mean age of 45.0 ± 9.0 y and a mean BMI of 32.0 ± 7.0 kg/m2 and IGT: 32 men and 37 women with a mean age of 48.4 ± 11.2 y and a mean BMI of 31.3 ± 5.9 kg/m2.
The AUC during the OGTT for glucose, insulin, and C-peptide concentrations was calculated using the trapezoidal rule. The HOMA-IR was calculated according to Matthews et al. [18].
The data from the initial 24-h period was not used to assess GV parameters due to instability. The mean time of sensor use was 13.7 d, or 98% of the whole period. The glucose concentration (mean ± SD, mmol/L), CV, continuous overall net glycemic action 1 (CONGA1) [19], lability index (LI) [20], J-index [21], low blood glucose index (LBGI) [22], high blood glucose index (HBGI) [22], glycemic risk assessment in diabetes equation (GRADE) [23], mean absolute glucose (MAG) [24], M-value [25], and mean amplitude of glycemic excursions (MAGE) [26] were calculated using EasyGV version 9.0.R2, available online (https://innovation.ox.ac.uk/licence-details/glycaemic-variability-calculator-easygv/).
Total daily macronutrients, EI, and their distribution between meals were calculated from diet diaries. The participants were instructed to record individual main meals and snacks in detail and accurately record meal composition and quantity. A Bulgarian website (www.BB-Team.org) was used for different variable calculations. Its database encompasses the nutritional composition of >10,000 foods and products exclusively from the national Bulgarian cuisine, which appeared convenient for the studied cohort. The website provides information on the calories and various macronutrient content in 100 g of a dish. Using this tool, the following parameters were calculated: carbohydrate_total (%), proteins_total (%), fats_total (%),carbohydrate_breakfast (%), proteins_breakfast (%), fats_breakfast (%),carbohydrate_lunch (%), proteins_lunch (%), fats_lunch (%),carbohydrate_dinner (%), proteins_dinner (%), and fats_dinner (%), reflecting the quantity of carbohydrate, proteins, and total fats per day and the quantity of carbohydrate, proteins, and fats at breakfast, lunch, and dinner, measured in grams and then calculated as a proportion in percentage. For EI, the amount of calories was calculated for the whole day (total EI, kcal), as well as for each main meal (breakfast EI, kcal; lunch EI, kcal; dinner EI, kcal), using the software. The calculations were performed separately for each recorded day, and the mean value was calculated and used for subsequent analyses. Days with >1 main meal skipped and >2 skipped snacks were excluded from calculations because these days were accepted as outliers.
Based on the overall carbohydrate intake, the participants were subdivided into 3 groups: low (40–99 g daily, moderate (100–200 g daily), and high carbohydrate consumption (>200 g daily) [27]. Qualitative analysis of different types of carbohydrate consumed was also performed to identify participants with no consumption, at least once weekly consumption, and at least once daily consumption. No consumption was defined as the absence of a certain type of carbohydrate for the reported period. At least once weekly consumption was defined as consuming a certain type of carbohydrate every week but less than once daily. At least once daily consumption was defined as consuming a certain type of carbohydrate at least once daily. We calculated the frequency of different carbohydrate consumption based on the data from the diet dairies. The participants were also subdivided into 2 groups based on having or skipping breakfast according to their diet diaries.
The primary outcomes were defined as the difference between the NGT and IGT groups in total daily macronutrients and energy consumption and their distribution in the main meals and GV parameters. The secondary outcomes were defined as the difference in GV indices between the subgroups according to carbohydrate consumption, intake frequency of various types of carbohydrate, and having or skipping breakfast.
Statistical analyses were performed using SPSS version 23.0 (IBM Corporation). The normal distribution of the data was tested by the Kolmogorov-Smirnov test, and variables with a skewed distribution were analyzed after log transformation. Differences across groups were tested using descriptive and single-factor dispersion ANOVA with post hoc multiple comparisons using the Bonferroni test. Pearson correlation analysis was performed to determine the relationships between diet patterns and GV indices. Partial correlation analysis was performed to control the estimated relationship between protein consumption and GV indices for carbohydrate and calorie intake as confounding variables. Stepwise forward regression was performed to test the independence of the observed relationships. All variables are shown as mean ± SD or median and interquartile range, depending on their distribution. A P value (2-tailed) of <0.05 was considered statistically significant.
Results
Table 1 shows the main anthropometric parameters, clinical characteristics, and diet patterns of the studied population. There was no difference between the NGT and IGT groups for age, BMI, waist circumference, VFA, and TBF. As expected, the AUCglucose (1200 compared with 895, P < 0.0001) and HbA1c (5.6 compared with 5.4, P = 0.032) were higher in IGT than in NGT. Similarly, the AUCinsulin (11320 compared with 8240, P = 0.037) and HOMA-IR (3.3 compared with 2.2, P = 0.034) were higher in IGT than in NGT. Several GV parameters, including CV (20 compared with 16, P < 0.0001), SD (1.0 compared with 0.8, P < 0.0001), J-index (10.7 compared with 9.6, P = 0.013), HBGI (1.3 compared with 0.8, P = 0.001), LI (0.9 compared with 0.7, P = 0.003), and MAGE (2.5 compared with 2.0, P = 0.016) significantly increased in IGT compared with the NGT group. There was no difference in daily calorie intake and macronutrient patterns between the 2 groups (Table 1).
TABLE 1.
Main characteristics of the groups according to glucose tolerance with NGT and IGT
| Parameter | NGT (n = 41) | IGT (n = 53) | P value |
|---|---|---|---|
| Age (y) | 45.0 ± 9.0 | 48.4 ± 11.2 | 0.118 |
| BMI (kg/m2) | 32.0 ± 7.0 | 31.3 ± 5.9 | 0.596 |
| Waist (cm) | 102 ± 16 | 103 ± 15 | 0.836 |
| Visceral fat area (cm2) | 155 ± 59 | 155 ± 49 | 0.959 |
| Total body fat (%) | 38 ± 11 | 38 ± 8 | 0.981 |
| HOMA-IR | 2.2 (1.2–3.8) | 3.3 (1.6–5.3) | 0.034 |
| AUC_glucose (mmol/L × 120 min) | 895 ± 153 | 1200 ± 175 | <0.0001 |
| AUC_insulin (μIU/mL × 120 min) | 8240 (5690–13360) | 11320 (7720–14920) | 0.037 |
| AUC_C-peptide (nmol/L × 120 min) | 414 ± 181 | 451 ± 158 | 0.286 |
| HbA1c (%) | 5.4 ± 0.3 | 5.6 ± 0.3 | 0.032 |
| CV (%) | 16 (15–19) | 20 (17–23) | <0.0001 |
| Mean interstitial glucose (mmol/L) | 4.7 (4.5–4.9) | 4.8 (4.4–5.2) | 0.154 |
| SD (mmol/L) | 0.8 ± 0.1 | 1.0 ± 0.2 | <0.0001 |
| CONGA1 | 4.2 ± 0.3 | 4.3 ± 0.6 | 0.178 |
| LI | 0.7 (0.5–0.9) | 0.9 (0.7–1.4) | 0.003 |
| J-index | 9.6 (8.9–10.3) | 10.7 (8.9–13.4) | 0.013 |
| LBGI | 4.1 (3.4–5.2) | 4.2 (2.7–6.3) | 0.490 |
| HBGI | 0.8 ± 0.5 | 1.3 ± 0.7 | 0.001 |
| GRADE | 0.6 (0.3–0.8) | 0.8 (0.4–1.4) | 0.121 |
| MAGE | 2.0 (1.7–2.3) | 2.5 (2.1–3.0) | 0.016 |
| M-value | 6.2 (5.0–8.6) | 6.2 (3.3–10.3) | 0.460 |
| MAG | 1.0 ± 0.2 | 1.0 ± 0.2 | 0.491 |
| CHO_total (%) | 50 (20–30) | 40 (35–55) | 0.271 |
| CHO_breakfast (%) | 30 (20–35) | 25 (5–35) | 0.356 |
| CHO_lunch (%) | 35 (30–45) | 35 (30–40) | 0.851 |
| CHO_dinner (%) | 35 (23–48) | 40 (28–53) | 0.138 |
| Sweets | |||
| No consumption | 17.1% (7) | 18.9% (10) | 0.324 |
| At least once weekly consumption | 61.0% (25) | 66.0% (35) | |
| At least once daily consumption | 22.0% (9) | 15.1% (8) | |
| Pastry | |||
| No consumption | 22.0% (9) | 13.2% (7) | 0.368 |
| At least once weekly consumption | 56.1% (23) | 77.4% (41) | |
| At least once daily consumption | 22.0% (9) | 9.4% (5) | |
| Whole grains | |||
| No consumption | 26.8% (11) | 17.0% (9) | 0.126 |
| At least once weekly consumption | 43.9% (18) | 39.6% (21) | |
| At least once daily consumption | 29.3% (12) | 43.4% (23) | |
| Refined grains | |||
| No consumption | 7.3% (3) | 15.1% (8) | 0.651 |
| At least once weekly consumption | 63.4% (26) | 56.6% (30) | |
| At least once daily consumption | 29.3% (12) | 28.3% (15) | |
| Proteins_total (%) | 25 (20–30) | 30 (23–35) | 0.361 |
| Proteins_breakfast (%) | 15 (10–30) | 15 (5–25) | 0.366 |
| Proteins_lunch (%) | 40 (30–50) | 40 (35–50) | 0.717 |
| Proteins_dinner (%) | 45 (30–50) | 40 (30–50) | 0.948 |
| Fat_total (%) | 25 (20–30) | 25 (20–30) | 0.906 |
| Fat_breakfast (%) | 20 (8–20) | 20 (10–35) | 0.307 |
| Fat_lunch (%) | 35 (30–50) | 35 (30–45) | 0.948 |
| Fat_dinner (%) | 40 (30–50) | 40 (33–50) | 0.812 |
| Total EI, kcal/d | 1600 (1250–1925) | 1500 (1300–1800) | 0.604 |
| Breakfast EI, kcal | 350 (200–525) | 300 (75–500) | 0.546 |
| Lunch EI, kcal | 600 (425–750) | 600 (450–700) | 0.781 |
| Dinner EI, kcal | 650 (500–850) | 600 (525–750) | 0.874 |
Data are presented as means ± SD or median (25th–75th percentile) depending on their distribution. P < 0.05 is statistically significant. CONGA1, continuous overall net glycemic action; GRADE, glycemic risk assessment in diabetes equation; HBGI, high blood glucose index; IGT, impaired glucose tolerance; kcal, kilocalories; LBGI, low blood glucose index; LI, liability index; MAG, mean absolute glucose; MAGE, mean amplitude of glycemic excursions; NGT, normal glucose tolerance.
When daily carbohydrate consumption was taken into consideration (Table 2), the following parameters were significantly increased only in IGT with moderate and high consumption compared with low-carbohydrate intake: SD (P = 0.035 for low- compared with high-carbohydrate group, P < 0.0001 for low- compared with moderate-carbohydrate group), CONGA1 (P = 0.022 for low-carbohydrate compared with high-carbohydrate group, P = 0.001 for low- compared with moderate-carbohydrate group), J-index (P = 0.002 for low-carbohydrate compared with high-carbohydrate group, P < 0.0001 for low- compared with moderate-carbohydrate group), LBGI (P = 0.039 for low-carbohydrate compared with high-carbohydrate group, P = 0.004 low- compared with moderate-carbohydrate group), LI (P = 0.010 for low-carbohydrate compared with high-carbohydrate group, P < 0.0001 low- compared with moderate-carbohydrate group), GRADE (P = 0.016 low-carbohydrate compared with moderate-carbohydrate group), and M-value (P = 0.017 for low-carbohydrate compared with high-carbohydrate group, P = 0.002 for low-carbohydrate compared with moderate-carbohydrate group). Only MAG was increased when comparing low-carbohydrate with the other groups in both NGT (P = 0.021 compared with high) and IGT (P = 0.002 compared with high, P = 0.001 compared with moderate). In IGT but not in NGT, CONGA1 (P = 0.004 for at least once daily compared with no consumption, P = 0.012 for at least once daily compared with at least once weekly), J-index (P = 0.009 for at least once daily compared with no consumption, P = 0.019 for at least once daily compared with at least once weekly), LBGI (P = 0.001 for at least once daily compared with no consumption, P = 0.008 for at least once daily compared with at least once weekly), M-value (P < 0.0001 for at least once daily compared with no consumption, P = 0.005 for at least once daily compared with at least once weekly), and MAG (P = 0.025 for at least once daily compared with no consumption) increased with an increase in the overall amount of refined grain consumption (Table 3). In the same participants, CONGA1 (P = 0.009 for at least once daily compared with no consumption, P = 0.029 for at least once daily compared with at least once weekly), J-index (P = 0.035 for at least once daily compared with no consumption, P = 0.039 for at least once daily compared with at least once weekly), LBGI (P = 0.004 for at least once daily compared with no consumption, P = 0.022 for at least once daily compared with at least once weekly), and M-value (P = 0.004 for at least once daily compared with no consumption, P = 0.022 for at least once daily compared with at least once weekly) improved with an increase in the frequency of whole grain intake (Table 4). The amount of sweets and pastry did not affect the GV parameters in both groups (Supplemental Tables 1 and 2).
TABLE 2.
Glucose variability parameters in NGT and IGT groups subdivided according to the overall consumption of carbohydrate per day with low, moderate, and high consumption
| NGT |
Low (n = 7) | Moderate (n = 24) | High (n = 10) | P value |
|---|---|---|---|---|
| Parameter | ||||
| AUC_glucose (mmol/L × 120 min) | 951 ± 159 | 894 ± 144 | 859 ± 176 | ≥0.05 |
| HbA1c (%) | 5.4 ± 0.3 | 5.5 ± 0.3 | 5.4 ± 0.5 | ≥0.05 |
| CV (%) | 16 (15–17) | 17 (15–19) | 16 (14–18) | ≥0.05 |
| Mean interstitial glucose (mmol/L) | 4.7 (3.9–5.3) | 4.6 (4.4–4.8) | 4.7 (4.6–4.9) | ≥0.05 |
| SD (mmol/L) | 0.7 ± 0.1 | 0.8 ± 0.1 | 0.8 ± 0.1 | ≥0.05 |
| CONGA1 | 4.2 ± 0.6 | 4.1 ± 0.3 | 4.3 ± 0.1 | ≥0.05 |
| LI | 0.6 (0.4–0.9) | 0.7 (0.5–0.9) | 0.7 (0.5–1.0) | ≥0.05 |
| J-index | 9.7 (6.7–11.3) | 9.3 (8.8–10.1) | 9.9 (9.1–10.2) | ≥0.05 |
| LBGI | 4.2 (1.9–9.3) | 4.4 (3.8–6.3) | 3.7 (3.3–4.1) | ≥0.05 |
| HBGI | 0.9 ± 0.5 | 0.8 ± 0.6 | 0.5 ± 0.4 | ≥0.05 |
| GRADE | 0.6 (0.2–3.4) | 0.6 (0.5–1.2) | 0.4 (0.2–0.6) | ≥0.05 |
| MAGE | 2.0 (1.6–2.3) | 2.1 (1.8–2.4) | 1.9 (1.6–2.3) | ≥0.05 |
| M-value | 6.2 (2.1–17.5) | 6.5 (5.5–10.7) | 5.4 (4.7–6.2) | ≥0.05 |
| MAG | 0.8 ± 0.21 | 1.0 ± 0.2 | 1.1 ± 0.1 | 0.021 vs. high |
| IGT | ||||
| Parameter | Low (n = 15) | Moderate (n = 27) | High (n = 11) | P value |
| AUC_glucose (mmol/L × 120 min) | 1160 ± 123 | 1250 ± 196 | 1150 ± 162 | ≥0.05 |
| HbA1c (%) | 5.4 ± 0.3 | 5.6 ± 0.3 | 5.7 ± 0.3 | ≥0.05 |
| CV (%) | 20 (14–21) | 20 (18–24) | 20 (17–23) | ≥0.05 |
| Mean interstitial glucose (mmol/L) | 4.3 (3.9–4.5)1,2 | 5.0 (4.6–5.3) | 4.9 (4.4–5.7) | 0.0041 vs. high <0.00012 vs. moderate |
| SD (mmol/L) | 0.8 ± 0.22 | 1.1 ± 0.2 | 1.0 ± 0.2 | 0.0351 vs. high <0.00012 vs. Moderate |
| CONGA1 | 3.9 ± 0.51,2 | 4.5 ± 0.5 | 4.4 ± 0.6 | 0.0221 vs. high 0.0012 vs. moderate |
| LI | 0.7 (0.4–1.0)1,2 | 1.1 (0.8–1.6) | 1.0 (0.7–1.5) | 0.0101 vs. high <0.00012 vs. moderate |
| J-index | 8.9 (6.6–9.6)1,2 | 11.3 (10.4–13.9) | 11.1 (8.9–14.4) | 0.0021 vs. high <0.00012 vs. moderate |
| LBGI | 6.5 (4.9–9.2)1,2 | 3.5 (2.4–4.9) | 3.7 (1.9–5.6) | 0.0391 vs. high 0.0042 vs. moderate |
| HBGI | 1.0 ± 0.8 | 1.4 ± 0.6 | 1.3 ± 1.0 | >0.05 |
| GRADE | 1.4 (0.8–2.8)2 | 0.6 (0.4–1.1) | 0.6 (0.4–1.1) | 0.0162 vs. moderate |
| MAGE | 2.1 (1.9–2.6) | 2.7 (2.3–3.4) | 2.6 (2.1–3.0) | >0.05 |
| M-value | 5.2 (2.1–8.9)1,2 | 4.7 (3.0–7.6) | 11.6 (7.4–17.6) | 0.0171 vs. high 0.0022 vs. moderate |
| MAG | 0.8 ± 0.21,2 | 1.1 ± 0.2 | 1.1 ± 0.2 | 0.0021 vs. high 0.0012 vs. moderate |
High: >200 g daily; moderate: 100–200 g daily; low: 40–99 g daily. Data are presented as means ± SD or median (25th–75th percentile) depending on their distribution. P < 0.05 is statistically significant. CONGA1, continuous overall net glycemic action; GRADE, glycemic risk assessment in diabetes equation; HBGI, high blood glucose index; IGT, impaired glucose tolerance; LBGI, low blood glucose index; LI, liability index; MAG, mean absolute glucose; MAGE, mean amplitude of glycemic excursions; NGT, normal glucose tolerance. 1 p low vs high; 2 p low vs moderate.
TABLE 3.
Glucose variability parameters in NGT and IGT groups subdivided into no consumption, at least once weekly, and at least once daily consumption subgroups of refined grains
| NGT |
No consumption (n = 3) | At least once weekly (n = 26) | At least once daily (n = 12) | P value |
|---|---|---|---|---|
| Parameter | ||||
| AUC_glucose (mmol/L × 120 min) | 862 ± 121 | 915 ± 169 | 659 ± 123 | ≥0.05 |
| HbA1c (%) | 5.4 ± 0.2 | 5.4 ± 0.4 | 5.4 ± 0.3 | ≥0.05 |
| CV (%) | 19 (12–19) | 17 (15–19) | 16 (14–17) | ≥0.05 |
| Mean interstitial glucose (mmol/L) | 5.1 (4.3–5.1) | 4.7 (4.5–4.8) | 4.6 (4.5–4.8) | ≥0.05 |
| SD (mmol/L) | 0.7 ± 0.1 | 0.8 ± 0.1 | 0.7 ± 0.1 | ≥0.05 |
| CONGA1 | 4.5 ± 0.6 | 4.1 ± 0.3 | 4.2 ± 0.3 | ≥0.05 |
| LI | 0.7 (0.4–0.7) | 0.7 (0.5–1.0) | 0.6 (0.5–0.9) | ≥0.05 |
| J-index | 11.2 (8.3–11.2) | 9.6 (9.0–10.1) | 9.2 (8.9–10.3) | ≥0.05 |
| LBGI | 2.5 (1.5–2.5) | 4.3 (3.6–5.1) | 4.0 (3.4–5.3) | ≥0.05 |
| HBGI | 1.4 ± 0.2 | 0.8 ± 0.5 | 0.7 ± 0.5 | ≥0.05 |
| GRADE | 0.2 (0.1–0.2) | 0.6 (0.4–0.8) | 0.6 (0.3–0.8) | ≥0.05 |
| MAGE | 2.1 (1.6–2.1) | 2.1 (1.7–2.3) | 1.9 (1.7–2.2) | ≥0.05 |
| M-value | 3.1 (1.8–3.1) | 6.3 (5.4–8.5) | 5.8 (4.9–8.9) | ≥0.05 |
| MAG | 0.8 ± 0.2 | 1.0 ± 0.2 | 1.0 ± 0.2 | ≥0.05 |
| IGT | ||||
| Parameter | No consumption (n = 8) | At least once weekly (n = 30) | At least once daily (n = 15) | P value |
| AUC_glucose (mmol/L × 120 min) | 1260 ± 227 | 1190 ± 163 | 1210 ± 172 | ≥0.05 |
| HbA1c (%) | 5.4 ± 0.3 | 5.6 ± 0.3 | 5.7 ± 0.3 | ≥0.05 |
| CV (%) | 21 (15–29) | 21 (17–23) | 20 (17–21) | ≥0.05 |
| Mean interstitial glucose (mmol/L) | 4.4 (3.9–5.0) | 4.7 (4.3–5.0) | 5.2 (4.9–5.8)1,2 | 0.0021 vs. no consumption 0.0102 vs. at least once weekly |
| SD (mmol/L) | 1.0 ± 0.4 | 0.9 ± 0.2 | 1.0 ± 0.2 | ≥0.05 |
| CONGA1 | 3.9 ± 0.5 | 4.2 ± 0.5 | 4.7 ± 0.51,2 | 0.0041 vs. no consumption 0.0122 vs. at least once weekly |
| LI | 0.8 (0.3–1.5) | 0.8 (0.7–1.2) | 1.1 (0.9–1.6) | ≥0.05 |
| J-index | 9.6 (6.7–13.1) | 10.1 (8.7–11.4) | 13.1 (11.1–15.3)1,2 | 0.0091 vs. no consumption 0.0192 vs. at least once weekly |
| LBGI | 6.6 (4.8–9.0) | 4.6 (2.8–6.6) | 2.8 (1.6–3.9)1,2 | 0.0011 vs. no consumption 0.0082 vs. at least once weekly |
| HBGI | 1.5 ± 1.1 | 1.3 ± 0.8 | 1.1 ± 0.3 | ≥0.05 |
| GRADE | 1.5 (0.6–2.7) | 0.8 (0.3–1.4) | 0.6 (0.4–0.8) | ≥0.05 |
| MAGE | 3.1 (1.9–4.1) | 2.4 (2.0–2.7)1 | 2.6 (2.4–3.0) | 0.0451 vs. no consumption |
| M-value | 4.0 (1.7–5.4) | 7.1 (3.7–11.8) | 11.3 (7.6–17.0)1,2 | <0.00011 vs. no consumption 0.0052 vs. at least once weekly |
| MAG | 0.9 ± 0.3 | 1.0 ± 0.2 | 1.1 ± 0.21 | 0.0251 vs. no consumption |
Data are presented as means ± SD or median (25th–75th percentile) depending on their distribution. P < 0.05 is statistically significant. CONGA1, continuous overall net glycemic action; GRADE, glycemic risk assessment in diabetes equation; HBGI, high blood glucose index; IGT, impaired glucose tolerance; LBGI, low blood glucose index; LI, liability index; MAG, mean absolute glucose; MAGE, mean amplitude of glycemic excursions; NGT, normal glucose tolerance. 1 p at least once daily vs no consumption; 2 p at least once daily vs at least once weekly.
TABLE 4.
Glucose variability parameters in NGT and IGT groups subdivided into no consumption, at least once weekly, and at least once daily consumption subgroups of whole grains
| NGT |
No consumption (n = 11) | At least once weekly (n = 18) | At least once daily (n = 12) | P value |
|---|---|---|---|---|
| Parameter | ||||
| AUC_glucose (mmol/L × 120 min) | 867 ± 136 | 892 ± 162 | 925 ± 163 | ≥0.05 |
| HbA1c (%) | 5.4 ± 0.3 | 5.4 ± 0.4 | 5.5 ± 0.3 | ≥0.05 |
| CV (%) | 16 (14–19) | 17 (14–19) | 16 (16–19) | ≥0.05 |
| Mean interstitial glucose (mmol/L) | 4.7 (4.5–5.0) | 4.6 (4.5–4.8) | 4.7 (4.6–4.9) | ≥0.05 |
| SD (mmol/L) | 0.8 ± 0.1 | 0.8 ± 0.1 | 0.8 ± 0.2 | ≥0.05 |
| CONGA1 | 4.2 ± 0.3 | 4.2 ± 0.3 | 4.2 ± 0.4 | ≥0.05 |
| LI | 0.6 (0.5–0.8) | 0.6 (0.5–0.8) | 0.9 (0.4–1.0) | ≥0.05 |
| J-index | 9.2 (8.8–10.6) | 9.6 (8.8–10.1) | 9.8 (9.1–10.4) | ≥0.05 |
| LBGI | 4.0 (3.0–5.5) | 4.3 (3.5–5.8) | 4.0 (3.5–4.9) | ≥0.05 |
| HBGI | 0.8 ± 0.6 | 0.8 ± 0.5 | 0.7 ± 0.4 | ≥0.05 |
| GRADE | 0.6 (0.2–0.8) | 0.6 (0.3–0.9) | 0.6 (0.4–0.7) | ≥0.05 |
| MAGE | 1.9 (1.6–2.4) | 2.0 (1.7–2.2) | 2.3 (1.6–2.4) | ≥0.05 |
| M-value | 5.8 (4.3–9.2) | 6.5 (5.3–10.2) | 6.0 (4.9–7.9) | ≥0.05 |
| MAG | 1.0 ± 0.2 | 0.9 ± 0.2 | 1.0 ± 0.2 | ≥0.05 |
| IGT | ||||
| Parameter | No consumption (n = 9) | At least once weekly (n = 21) | At least once daily (n = 23) | P value |
| AUC_glucose (mmol/L × 120 min) | 1220 ± 150 | 1180 ± 170 | 1220 ± 193 | ≥0.05 |
| HbA1c (%) | 5.6 ± 0.3 | 5.5 ± 0.3 | 5.6 ± 0.3 | ≥0.05 |
| CV (%) | 18 (17–24) | 20 (16–21) | 20 (17–23) | ≥0.05 |
| Mean interstitial glucose (mmol/L) | 5.7 (4.8–5.8) | 4.8 (4.4–5.1)1 | 4.7 (4.2–5.1)2 | 0.0172 vs. no consumption 0.0341 vs. no consumption |
| SD (mmol/L) | 1.1 ± 0.2 | 0.9 ± 0.2 | 1.0 ± 0.3 | ≥0.05 |
| CONGA1 | 4.8 ± 0.6 | 4.2 ± 0.51 | 4.2 ± 0.52 | 0.0092 vs. no consumption 0.0291 vs. no consumption |
| LI | 0.9 (0.4–1.0) | 1.1 (0.8–1.7) | 0.9 (0.6–1.2) | ≥0.05 |
| J-index | 14.4 (11.0–15.6) | 10.4 (8.8–11.6)2 | 10.4 (8.7–11.3)1 | 0.0352 vs. no consumption 0.0391 vs. no consumption |
| LBGI | 2.2 (1.2–4.5) | 3.9 (2.9–5.8)1 | 4.9 (2.8–8.0)2 | 0.0042 vs. no consumption 0.0221 vs. no consumption |
| HBGI | 1.2 ± 0.4 | 1.1 ± 0.6 | 1.4 ± 0.9 | ≥0.05 |
| GRADE | 0.6 (0.5–1.0) | 0.8 (0.37–1.05) | 1.1 (0.6–1.8) | ≥0.05 |
| MAGE | 2.7 (2.2–3.3) | 2.5 (2.0–2.7) | 2.5 (2.2–3.1) | ≥0.05 |
| M-value | 8.2 (4.0–14.4) | 5.5 (4.1–9.5)1 | 2.4 (1.1–6.6)2 | 0.0042 vs. no consumption 0.0221 vs. no consumption |
| MAG | 1.1 ± 0.2 | 1.0 ± 0.2 | 1.0 ± 0.2 | |
Data are presented as means ± SD or median (25th–75th percentile) depending on their distribution. P < 0.05 is statistically significant. CONGA1, continuous overall net glycemic action; GRADE, glycemic risk assessment in diabetes equation; HBGI, high blood glucose index; IGT, impaired glucose tolerance; LBGI, low blood glucose index; LI, liability index; MAG, mean absolute glucose; MAGE, mean amplitude of glycemic excursions; NGT, normal glucose tolerance. Grades of significance: p2 is lower than p1.
For the correlation analysis, a positive relationship was observed for SD, CONGA1, J-index, LI, GRADE, M-value, and MAG (r = 0.14–0.53; all P < 0.02) and an inverse relationship for LBGI (r = −0.37, P = 0.006) and the total percentage of carbohydrate in the IGT group. No association was found for the distribution of carbohydrate (Table 5). The IGT group showed a negative relationship between total protein consumption and GV indices (r = −0.27 to −0.52; all P < 0.05 for SD, CONGA1, J-index, LI, M-value, and MAG; Table 6) as well as between total calorie intake and GV parameters (r = 0.27–0.32; all P < 0.05 for CONGA1, J-index, LI, and M-value; and r = −0.30, P = 0.028 for LBGI; Table 7). However, after controlling for carbohydrate and calorie intake, the observed relationship between protein intake and GV became nonsignificant. The distribution of calories, proteins, fats, and the total daily fat intake did not appear to affect GV in both groups (TABLE 6, TABLE 7 and Supplemental Table 3).
TABLE 5.
Correlation between glucose variability parameters and the distribution of carbohydrate consumption for the main meals in the NGT and IGT groups
| Glucose variability parameter |
% Total CHO |
% CHO for breakfast |
% CHO for lunch |
% CHO for dinner |
||||
|---|---|---|---|---|---|---|---|---|
| NGT | R | P | r | P | r | P | r | P |
| AUC_glucose (mmol/L × 120 min) | −0.17 | 0.283 | −0.11 | 0.482 | 0.07 | 0.655 | 0.05 | 0.741 |
| LnCV (%) | 0.08 | 0.633 | −0.19 | 0.232 | 0.22 | 0.176 | −0.04 | 0.807 |
| LnMean interstitial glucose (mmol/L) | −0.02 | 0.913 | −0.01 | 0.940 | 0.25 | 0.123 | −0.22 | 0.160 |
| SD (mmol/L) | 0.17 | 0.298 | −0.09 | 0.593 | 0.24 | 0.137 | −0.16 | 0.320 |
| CONGA1 | −0.12 | 0.470 | −0.07 | 0.645 | 0.22 | 0.177 | −0.14 | 0.397 |
| LnLI | 0.22 | 0.169 | 0.05 | 0.357 | 0.16 | 0.319 | −0.21 | 0.196 |
| LnJ-index | 0.04 | 0.812 | −0.04 | 0.812 | 0.29 | 0.064 | −0.25 | 0.118 |
| LnLBGI | 0.11 | 0.485 | 0.05 | 0.765 | −0.20 | 0.217 | 0.14 | 0.378 |
| HBGI | −0.22 | 0.163 | −0.07 | 0.686 | 0.01 | 0.964 | 0.05 | 0.745 |
| LnGRADE | 0.06 | 0.695 | −0.01 | 0.993 | −0.19 | 0.248 | 0.17 | 0.288 |
| LnMAGE | 0.18 | 0.272 | 0.01 | 0.935 | 0.15 | 0.348 | −0.16 | 0.314 |
| lnM-value | 0.11 | 0.513 | 0.05 | 0.774 | −0.21 | 0.178 | 0.16 | 0.320 |
| MAG |
0.38 |
0.016 |
0.19 |
0.248 |
0.24 |
0.136 |
−0.40 |
0.009 |
| IGT |
R |
P |
r |
P |
r |
P |
r |
P |
| AUC_glucose (mmol/L × 120 min) | 0.01 | 0.933 | 0.08 | 0.573 | 0.01 | 0.962 | −0.06 | 0.545 |
| lnCV (%) | 0.14 | 0.333 | −0.01 | 0.970 | −0.04 | 0.786 | 0.02 | 0.880 |
| lnMean interstitial glucose (mmol/L) | 0.47 | <0.0001 | 0.07 | 0.644 | 0.01 | 0.969 | −0.07 | 0.605 |
| SD (mmol/L) | 0.33 | 0.017 | 0.04 | 0.754 | −0.05 | 0.741 | −0.02 | 0.903 |
| CONGA1 | 0.40 | 0.003 | 0.04 | 0.754 | 0.02 | 0.888 | −0.06 | 0.655 |
| lnLI | 0.43 | 0.001 | 0.07 | 0.630 | 0.01 | 0.954 | −0.09 | 0.518 |
| lnJ-index | 0.48 | <0.0001 | 0.07 | 0.638 | −0.01 | 0.948 | −0.07 | 0.644 |
| lnLBGI | −0.37 | 0.006 | −0.05 | 0.702 | −0.02 | 0.873 | 0.07 | 0.599 |
| HBGI | 0.14 | 0.322 | 0.02 | 0.887 | −0.06 | 0.680 | 0.01 | 0.976 |
| lnGRADE | 0.41 | 0.002 | −0.10 | 0.484 | 0.08 | 0.558 | 0.05 | 0.751 |
| lnMAGE | 0.06 | 0.673 | 0.17 | 0.228 | −0.36 | 0.008 | 0.14 | 0.321 |
| lnM-value | 0.14 | 0.002 | −0.06 | 0.678 | −0.03 | 0.830 | 0.09 | 0.545 |
| MAG | 0.53 | <0.0001 | 0.03 | 0.839 | 0.03 | 0.857 | −0.06 | 0.673 |
Data are presented as means ± SD or median (25th–75th percentile) depending on their distribution. P < 0.05 is statistically significant. CONGA1, continuous overall net glycemic action; GRADE, glycemic risk assessment in diabetes equation; HBGI, high blood glucose index; IGT, impaired glucose tolerance; LBGI, low blood glucose index; LI, liability index; MAG, mean absolute glucose; MAGE, mean amplitude of glycemic excursions; NGT, normal glucose tolerance.
TABLE 6.
Correlation between glucose variability parameters and the distribution of protein consumption for the main meals in NGT and IGT groups
| Glucose variability parameter |
% Total proteins |
% Proteins for breakfast |
% Proteins for lunch |
% Proteins for dinner |
||||
|---|---|---|---|---|---|---|---|---|
| NGT | r | P | r | p | r | p | r | p |
| AUC_glucose (mmol/L × 120 min) | 0.06 | 0.707 | 0.05 | 0.767 | −0.20 | 0.223 | 0.17 | 0.279 |
| lnCV (%) | −0.22 | 0.165 | 0.08 | 0.623 | 0.05 | 0.736 | −0.10 | 0.542 |
| LnMean interstitial glucose (mmol/L) | −0.09 | 0.588 | −0.07 | 0.648 | 0.15 | 0.354 | −0.09 | 0.592 |
| SD (mmol/L) | −0.28 | 0.079 | 0.15 | 0.341 | 0.05 | 0.744 | −0.16 | 0.323 |
| CONGA1 | 0.01 | 0.939 | −0.17 | 0.298 | 0.16 | 0.319 | −0.02 | 0.901 |
| LnLI | −0.32 | 0.044 | 0.21 | 0.186 | −0.03 | 0.853 | −0.14 | 0.381 |
| LnJ-index | −0.17 | 0.291 | −0.01 | 0.933 | 0.15 | 0.362 | −0.13 | 0.431 |
| LnLBGI | −0.01 | 0.980 | 0.14 | 0.368 | −0.13 | 0.436 | 0.01 | 0.947 |
| HBGI | 0.11 | 0.510 | 0.06 | 0.716 | −0.12 | 0.441 | 0.08 | 0.605 |
| LnGRADE | −0.01 | 0.994 | 0.08 | 0.614 | −0.10 | 0.548 | 0.03 | 0.841 |
| LnMAGE | −0.30 | 0.053 | 0.13 | 0.409 | 0.03 | 0.840 | −0.14 | 0.381 |
| LnM-value | 0.01 | 0.968 | 0.14 | 0.394 | −0.13 | 0.425 | 0.02 | 0.925 |
| MAG |
−0.41 |
0.007 |
−0.41 |
0.009 |
0.01 |
0.974 |
−0.33 |
0.034 |
| IGT |
r |
P |
r |
P |
r |
P |
r |
P |
| AUC_glucose (mmol/L × 120 min) | −0.08 | 0.566 | −0.04 | 0.788 | 0.08 | 0.563 | 0.02 | 0.895 |
| LnCV (%) | −0.16 | 0.247 | 0.05 | 0.700 | −0.05 | 0.718 | 0.11 | 0.439 |
| LnMean interstitial glucose (mmol/L) | −0.35 | 0.011 | 0.32 | 0.020 | −0.14 | 0.327 | −0.06 | 0.650 |
| SD (mmol/L) | −0.30 | 0.027 | 0.23 | 0.094 | −0.11 | 0.418 | 0.02 | 0.872 |
| CONGA1 | −0.27 | 0.050 | 0.31 | 0.025 | −0.11 | 0.448 | −0.10 | 0.493 |
| LnLI | −0.39 | 0.004 | 0.21 | 0.123 | −0.08 | 0.561 | 0.03 | 0.860 |
| LnJ-index | −0.37 | 0.007 | 0.32 | 0.020 | −0.15 | 0.299 | −0.04 | 0.782 |
| LnLBGI | 0.24 | 0.086 | −0.27 | 0.052 | 0.06 | 0.673 | 0.12 | 0.413 |
| HBGI | −0.07 | 0.620 | 0.03 | 0.858 | 0.04 | 0.767 | −0.01 | 0.924 |
| LnGRADE | 0.23 | 0.098 | −0.07 | 0.622 | 0.14 | 0.315 | 0.10 | 0.472 |
| LnMAGE | −0.07 | 0.598 | 0.01 | 0.963 | 0.03 | 0.845 | 0.01 | 0.969 |
| LnM-value | −0.28 | 0.039 | −0.29 | 0.036 | 0.09 | 0.546 | 0.11 | 0.452 |
| MAG | −0.52 | <0.0001 | 0.18 | 0.211 | −0.09 | 0.536 | 0.02 | 0.906 |
Data are presented as means ± SD or median (25th–75th percentile) depending on their distribution. P < 0.05 is statistically significant. CONGA1, continuous overall net glycemic action; GRADE, glycemic risk assessment in diabetes equation; HBGI, high blood glucose index; IGT, impaired glucose tolerance; LBGI, low blood glucose index; LI, liability index; MAG, mean absolute glucose; MAGE, mean amplitude of glycemic excursions; NGT, normal glucose tolerance.
TABLE 7.
Correlation between glucose variability parameters and the distribution of energy consumption for the main meals in NGT and IGT groups
| Glucose variability parameter |
Total EI, kcal |
Breakfast EI, % of total kcal |
Lunch EI, % of total kcal |
Dinner EI, % of total kcal |
||||
|---|---|---|---|---|---|---|---|---|
| NGT | r | P | r | P | r | P | r | P |
| AUC_glucose (mmol/L × 120 min) | −0.09 | 0.597 | −0.05 | 0.745 | 0.07 | 0.665 | −0.04 | 0.819 |
| LnCV (%) | −0.25 | 0.110 | −0.12 | 0.455 | −0.24 | 0.139 | −0.29 | 0.065 |
| LnMean interstitial glucose (mmol/L) | 0.09 | 0.570 | 0.12 | 0.460 | 0.13 | 0.435 | 0.16 | 0.320 |
| SD (mmol/L) | −0.12 | 0.448 | −0.10 | 0.527 | −0.13 | 0.433 | −0.17 | 0.297 |
| CONGA1 | 0.05 | 0.766 | 0.14 | 0.396 | 0.17 | 0.277 | 0.17 | 0.295 |
| LnLI | −0.03 | 0.838 | −0.13 | 0.421 | −0.08 | 0.624 | −0.11 | 0.478 |
| LnJ-index | 0.04 | 0.799 | 0.07 | 0.648 | 0.08 | 0.630 | 0.09 | 0.573 |
| LnLBGI | −0.07 | 0.659 | −0.15 | 0.353 | −0.20 | 0.206 | −0.20 | 0.215 |
| HBGI | 0.02 | 0.925 | 0.08 | 0.632 | 0.27 | 0.089 | 0.17 | 0.301 |
| LnGRADE | −0.16 | 0.331 | −0.17 | 0.294 | −0.17 | 0.301 | −0.23 | 0.142 |
| LnMAGE | −0.10 | 0.536 | −0.16 | 0.333 | −0.10 | 0.552 | −0.17 | 0.295 |
| LnM-value | −0.07 | 0.667 | −0.15 | 0.343 | −0.18 | 0.266 | −0.19 | 0.239 |
| MAG |
0.16 |
0.314 |
−0.01 |
0.979 |
−0.22 |
0.172 |
−0.02 |
0.909 |
| IGT |
r |
P |
r |
P |
r |
P |
r |
P |
| AUC_glucose (mmol/L × 120 min) | 0.14 | 0.306 | 0.08 | 0.587 | 0.08 | 0.578 | 0.19 | 0.173 |
| LnCV (%) | 0.07 | 0.623 | 0.06 | 0.676 | 0.08 | 0.573 | 0.14 | 0.321 |
| LnMean interstitial glucose (mmol/L) | 0.31 | 0.022 | −0.01 | 0.953 | 0.07 | 0.623 | 0.28 | 0.040 |
| SD (mmol/L) | 0.23 | 0.096 | 0.05 | 0.702 | 0.12 | 0.410 | 0.28 | 0.039 |
| CONGA1 | 0.29 | 0.033 | 0.01 | 0.973 | 0.06 | 0.688 | 0.27 | 0.056 |
| LnLI | 0.27 | 0.048 | 0.02 | 0.878 | 0.06 | 0.693 | 0.26 | 0.062 |
| LnJ-index | 0.32 | 0.019 | 0.01 | 0.972 | 0.08 | 0.560 | 0.30 | 0.027 |
| LnLBGI | −0.30 | 0.028 | 0.01 | 0.931 | −0.02 | 0.898 | −0.23 | 0.093 |
| HBGI | 0.07 | 0.636 | 0.19 | 0.185 | 0.02 | 0.886 | 0.19 | 0.184 |
| LnGRADE | −0.20 | 0.157 | 0.11 | 0.437 | −0.09 | 0.547 | −0.16 | 0.268 |
| LnMAGE | −0.03 | 0.847 | −0.13 | 0.341 | −0.02 | 0.903 | −0.09 | 0.506 |
| LnM-value | 0.31 | 0.025 | 0.01 | 0.925 | −0.04 | 0.790 | −0.25 | 0.071 |
| MAG | 0.17 | 0.232 | 0.04 | 0.757 | 0.13 | 0.370 | 0.26 | 0.061 |
Data are presented as means ± SD or median (25th–75th percentile) depending on their distribution. P < 0.05 is statistically significant. CONGA1, continuous overall net glycemic action; GRADE, glycemic risk assessment in diabetes equation; HBGI, high blood glucose index; IGT, impaired glucose tolerance; kcal, kilocalories; LBGI, low blood glucose index; LI, liability index; MAG, mean absolute glucose; MAGE, mean amplitude of glycemic excursions; NGT, normal glucose tolerance.
In the stepwise multiple regression analysis, the total daily carbohydrate, calorie consumption, and HOMA-IR emerged as independent contributors to GV in IGT. The log-transformed percentage of total carbohydrate intake accounted for ∼18% of the variances for SD and LBGI and ≤34% for MAG and J-index. The log-transformed percentage of total calorie intake accounted for ∼26% for LBGI and ≤46% for J-index. The log-transformed HOMA-IR accounted for ∼12%, 34%, and 47% of the variances for CV, SD, and LI, respectively (Table 8).
TABLE 8.
Multivariate regression analysis for the predictive HOMA-IR value and main diet parameters for glucose variability in the IGT group
| Stepwise forward regression explanatory variables | Regression coefficient (β) | SEM | P value | Coefficient of determination (R2) | 95% CI |
|---|---|---|---|---|---|
| SD | |||||
| CHO_total %1 | 0.227 | 0.061 | 0.001 | 0.175 | 0.104, 0.350 |
| Total EI, kcal | 0.375 | 0.124 | 0.004 | 0.271 | 0.125, 0.625 |
| HOMA-IR | −0.070 | 0.030 | 0.024 | 0.344 | −0.129, −0.010 |
| CV | |||||
| HOMA-IR | −0.078 | 0.029 | 0.011 | 0.119 | −0.137, −0.018 |
| CONGA1 | |||||
| CHO_total %1 | 0.603 | 0.149 | <0.0001 | 0.215 | 0.303, 0.903 |
| Total EI, kcal | 0.766 | 0.299 | 0.014 | 0.306 | 0.165, 1.367 |
| J-index | |||||
| CHO_total %1 | 0.381 | 0.067 | <0.0001 | 0.337 | 0.247, 0.515 |
| Total EI, kcal | 0.443 | 0.134 | 0.002 | 0.456 | 0.174, 0.712 |
| LI | |||||
| CHO_total %1 | 0.692 | 0.130 | <0.0001 | 0.296 | 0.430, 0.954 |
| Total EI, kcal | 0.845 | 0.265 | 0.002 | 0.381 | 0.312, 1.377 |
| HOMA-IR | −0.177 | 0.064 | 0.008 | 0.465 | −0.305, −0.049 |
| LBGI | |||||
| CHO_total %1 | −0.664 | 0.186 | 0.001 | 0.181 | −1.038, −0.291 |
| Total EI, kcal | −0.831 | 0.373 | 0.031 | 0.255 | −1.580, −0.081 |
| GRADE | |||||
| CHO_total %1 | 1.032 | 0.257 | <0.0001 | 0.241 | 1.547, 0.517 |
| M-value | |||||
| CHO_total %1 | −0.896 | 0.227 | <0.0001 | 0.210 | −1.353, −0.439 |
| Total EI, kcal | −1.097 | 0.456 | 0.020 | 0.292 | −2.013, 0.181 |
| MAG | |||||
| CHO_total %1 | 0.292 | 0.054 | <0.0001 | 0.342 | 0.185, 0.398 |
| Total EI, kcal | 0.268 | 0.106 | 0.015 | 0.417 | 0.015, 0.481 |
P < 0.05 is statistically significant. Some variables, including glucose variability, diet parameters, and HOMA-IR, were log-transformed for normal distribution. CONGA1, continuous overall net glycemic action; GRADE, glycemic risk assessment in diabetes equation; IGT, impaired glucose tolerance; kcal, kilocalories; LBGI, low blood glucose index; LI, liability index; MAG, mean absolute glucose.
Variables used in the regression analysis: CHO_total %, Proteins_total %, Total EI, kcal, HOMA-IR, and HOMA-beta.
Dividing the participants according to having or skipping breakfast did not affect the GV indices in both groups (Supplemental Table 4).
Discussion
Based on a 14-d FGM, the current study showed that people with IGT have greater daily GV than those with NGT. These results contribute to the limited data currently available [5]. Moreover, our study shows multiple associations between GV parameters and total calorie intake with qualitative aspects of food consumption.
Our results suggest that GV occurs and is affected by nutritional habits in IGT but not in NGT subjects. Individuals with IGT are considered at an increased risk of diabetes because they may already have impaired insulin secretion and resistance, the main pathogenetic mechanisms accounting for alterations in glucose homeostasis. Therefore, we hypothesize that nutrition interacts with the underlying defects in glucose metabolism in determining GV. Our prior studies [5,28] and others [29,30] showed that daily GV is already impaired in individuals with mild impairment of glucose tolerance and that the degree of GV is inversely related to insulin secretion and action. Prior studies have also suggested a potential interaction between these 2 parameters in determining the degree of GV [31,32]. Our results reinforce such an interaction because parameters of insulin sensitivity (HOMA-IR), calories, and carbohydrate content were found to be independent predictors of several GV parameters (Table 8).
Although nutrition is an important component of diabetes care, the role of the specific macronutrient composition on the glucose profile is still unclear. For instance, there is no consensus on the exact range of daily carbohydrate consumption in T2DM management. As for the latter, we found that low total carbohydrate intake (that is, <99 g) [27] was associated with milder GV in IGT subjects. This observation is in line with studies showing that a low-carbohydrate diet regimen can limit GV and improve HbA1c in people with T2DM [7,8]. Similar data are available for individuals with T1DM in whom a low-carbohydrate intake has been shown to increase the time range of blood glucose concentrations and reduce hypoglycemic episodes and GV [9,10,33].
Another important component of the diet is the type and frequency of carbohydrate consumed. Estimation of the GI and derived glycemic load exerts a beneficial effect on postprandial glucose excursion and GV [13,14]. In our study, more frequent consumption of refined grains with a higher GI was associated with worse GV parameters in the IGT group, whereas an inverse correlation was observed with the consumption of lower GI whole grains.
We have also investigated whether protein and fat could contribute to GV. We found an inverse relationship between total protein consumption and GV indices, confirming prior observations in T2DM [33,34] and IGT individuals, with the possible mechanism having a more pronounced effect on GLP-1 secretion from L cells [35,36]. However, after repeating the analysis controlling for carbohydrate and calorie consumption, there was no longer a significant relationship, suggesting that the amount of carbohydrate and calorie intake are confounding factors that neutralize the beneficial effect of proteins.
Interestingly, the sequence of macronutrient intake during the meal may also influence postprandial glucose excursions. Thus, the fat and protein intake before carbohydrate has been found to reduce postmeal glucose peaks in people with T2DM [15] and IGT [16] through a delaying effect on gastric emptying and enhanced insulin secretion [37,38]. However, this does not seem to be the case in our study because the specific distribution of proteins between meals did not significantly impact GV indices, although the explanation for this discrepancy is not readily apparent.
Another important dietary pattern is the quantity and distribution of the main meals. Breakfast is of particular interest because insulin resistance and increased hepatic glucose production are more pronounced in the morning [39,40]. Previous work suggests that breakfast replacement can improve GV in newly diagnosed T2DM [41] and reduce postmeal glycemic peak [42]. However, when we compared GV indices between those having compared with those skipping breakfast, we found no difference between the NGT or IGT groups. These conflicting observations might be due to the specific population included in our study, notably NGT individuals at the very early stages of dysglycemia that may have prevented full disruption of the homeostatic system as it occurs in the presence of overt diabetes. In line with this interpretation is the observation that low-carbohydrate breakfast ameliorates GV and reduces AUC for glucose after breakfast without affecting glucose excursions after lunch or dinner [12,43].
Our findings are based on food diaries, relying on self-reported information. This may have introduced some degree of inaccuracy, though a great effort was made to properly instruct all participants. They were also instructed not to exceed 30 min of moderate physical activity daily, introducing some potential variability. However, available data in physically active people with T1DM do not support an association between various aspects of physical activity (type, weekly time, frequency, and intensity) and HbA1c concentrations [27]. Regarding the statistical analysis, there is a possibility of type 1 error given the number of statistical tests performed, although a Bonferroni correction was applied to the post hoc comparisons. Because of the cross-sectional study design, we cannot claim a causal relationship between the investigated parameters. Our study has some strengths, including a full 14-d FGM use, allowing a direct assessment of the impact of each main meal on GV in everyday routine.
In conclusion, our primary outcomes showed that insulin sensitivity, calories, and carbohydrate content are predictors of GV in individuals with IGT. The secondary analyses imply that overall carbohydrate and daily consumption of refined grains are the main dietary patterns that might be associated with higher GV, whereas whole grains and daily protein intake are likely related to lower GV in people with IGT but not in those with NGT. The optimal macronutrient composition in people at risk of diabetes is still debated. However, our results can offer further insights into designing a more efficacious dietary intervention, including the suggestion to increase the protein/carbohydrate ratio and restrict total EI. These findings could serve as a basis for further investigations to elucidate the most efficacious set of dietary recommendations for T2DM prevention in the high-risk population with IGT.
Acknowledgments
We are grateful to our colleagues who generously contributed to patients’ recruitment in this study. We especially thank all subjects who volunteered to participate. Finally, we thank our nurses and laboratory technicians for their superb support in performing the study procedures and laboratory analyses.
The authors’ responsibilities were as follows—RD and SDP: designed the study; RD and NC: collected the data; RD: analyzed the results and wrote the article in consultation with SDP and TT; and all authors: read and approved the final manuscript.
Footnotes
Supplementary data to this article can be found online at https://doi.org/10.1016/j.tjnut.2023.03.007.
Data availability
The data that support the findings of this study are available from the corresponding author upon reasonable request.
Funding
This study was funded by the European Foundation for the Study of Diabetes (EFSD) Mentorship Program 2019–2021 and National Science Fund, financial support for basic research projects 2019, contract КП-06 Н 33/7 from 13.12.2019.
Author disclosures
The authors report no conflicts of interest.
Appendix A. Supplementary data
The following is the Supplementary data to this article:
References
- 1.Umpierrez G.E., Kovatchev B.P. Glycemic variability: how to measure and its clinical implication for type 2 diabetes. Am. J. Med. Sci. 2018;356(6):518–527. doi: 10.1016/j.amjms.2018.09.010. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Frontoni S., Bartolo P.D., Avogaro A., Bosi E., Paolisso G., Ceriello A. Glucose variability: an emerging target for the treatment of diabetes mellitus. Diabetes Res. Clin. Pract. 2013;102(2):86–95. doi: 10.1016/j.diabres.2013.09.007. [DOI] [PubMed] [Google Scholar]
- 3.J. Fleischer, S. Lebech Cichosz, P. Hoeyem, E. Laugesen, P. Loegstrup Poulsen, J. Sandahl Christiansen, et al., Glycemic variability is associated with reduced cardiac autonomic modulation in women with type 2 diabetes, Diabetes Care 38 (4) (2015) 682–688. 10.2337/dc14-0654. [DOI] [PubMed]
- 4.Tang X., Li S., Wang Y., Wang M., Yin Q., Mu P., et al. Glycemic variability evaluated by continuous glucose monitoring system is associated with the 10-y cardiovascular risk of diabetic patients with well-controlled HbA1c. Clin. Chim. Acta. 2016;461:146–150. doi: 10.1016/j.cca.2016.08.004. [DOI] [PubMed] [Google Scholar]
- 5.Chakarova N., Dimova R., Grozeva G., Tankova T. Assessment of glucose variability in subjects with prediabetes. Diabetes Res. Clin. Pract. 2019;151:56–64. doi: 10.1016/j.diabres.2019.03.038. [DOI] [PubMed] [Google Scholar]
- 6.Baranowska-Jurkun A., Matuszewski W., Bandurska-Stankiewicz E. Chronic microvascular complications in prediabetic states—an overview. J. Clin. Med. 2020;9(10):3289. doi: 10.3390/jcm9103289. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Mori Y., Ohta T., Yokoyama J., Utsunomiya K. Effects of low-carbohydrate/high-monounsaturated fatty acid liquid diets on diurnal glucose variability and insulin dose in type 2 diabetes patients on tube feeding who require insulin therapy. Diabetes Technol. Ther. 2013;15(9):762–767. doi: 10.1089/dia.2013.0066. [DOI] [PubMed] [Google Scholar]
- 8.Blaychfeld-Magnazi M., Reshef N., Zornitzki T., Madar Z., Knobler H. The effect of a low-carbohydrate high-fat diet and ethnicity on daily glucose profile in type 2 diabetes determined by continuous glucose monitoring. Eur. J. Nutr. 2020;59(5):1929–1936. doi: 10.1007/s00394-019-02043-z. [DOI] [PubMed] [Google Scholar]
- 9.Ranjan A., Schmidt S., Damm-Frydenberg C., Holst J.J., Madsbad S., Nørgaard K. Short-term effects of a low carbohydrate diet on glycaemic variables and cardiovascular risk markers in patients with type 1 diabetes: a randomized open-label crossover trial, Diabetes Obes. Metab. 2017;19(10):1479–1484. doi: 10.1111/dom.12953. [DOI] [PubMed] [Google Scholar]
- 10.Ahola A.J., Forsblom C., Harjutsalo V., Groop P.H., FinnDiane Study Group Dietary carbohydrate intake and cardio-metabolic risk factors in type 1 diabetes. Diabetes Res. Clin. Pract. 2019;155:107818. doi: 10.1016/j.diabres.2019.107818. [DOI] [PubMed] [Google Scholar]
- 11.Eiswirth M., Clark E., Diamond M. Low carbohydrate diet and improved glycaemic control in a patient with type one diabetes. Endocrinol. Diabetes Metab. Case Rep. 2018;2018 doi: 10.1530/EDM-18-0002. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Chang C.R., Francois M.E., Little J.P. Restricting carbohydrates at breakfast is sufficient to reduce 24-hour exposure to postprandial hyperglycemia and improve glycemic variability. Am. J. Clin. Nutr. 2019;109(5):1302–1309. doi: 10.1093/ajcn/nqy261. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.C.J. Henry, B. Kaur, R.Y.C. Quek, S.G. Camps, A low glycaemic index diet incorporating isomaltulose is associated with lower glycaemic response and variability, and promotes fat oxidation in Asians, Nutrients 9 (5) (2017) 5. 10.3390/nu9050473 [DOI] [PMC free article] [PubMed]
- 14.Camps S.G., Kaur B., Quek R.Y.C., Henry C.J. Does the ingestion of a 24 hour low glycaemic index Asian mixed meal diet improve glycaemic response and promote fat oxidation? A controlled, randomized cross-over study. Nutr. J. 2017;16(1):43. doi: 10.1186/s12937-017-0258-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.D. Tricò, E. Filice, S. Trifirò, A. Natali, Manipulating the sequence of food ingestion improves glycemic control in type 2 diabetic patients under free-living conditions, Nutr. Diabetes. 6 (8) (2016) e226. 10.1038/nutd.2016.33. [DOI] [PMC free article] [PubMed]
- 16.Shukla A.P., Dickison M., Coughlin N., Karan A., Mauer E., Truong W., et al. The impact of food order on postprandial glycaemic excursions in prediabetes, Diabetes Obes. Metab. 2019;21(2):377–381. doi: 10.1111/dom.13503. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.American Diabetes Association Standards of medical care in diabetes-2019. Diabetes Care. 2019;42(Suppl 1):S13–S28. doi: 10.2337/dc19-S002. [DOI] [PubMed] [Google Scholar]
- 18.Matthews D.R., Hosker J.P., Rudenski A.S., Naylor B.A., Treacher D.F., Turner R.C. Homeostasis model assessment: insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia. 1985;28(7):412–419. doi: 10.1007/BF00280883. [DOI] [PubMed] [Google Scholar]
- 19.McDonnell C.M., Donath S.M., Vidmar S.I., Werther G.A., Cameron F.J. A novel approach to continuous glucose analysis utilizing glycemic variation. Diabetes Technol. Ther. 2005;7(2):253–263. doi: 10.1089/dia.2005.7.253. [DOI] [PubMed] [Google Scholar]
- 20.E.A. Ryan, T. Shandro, K. Green, B.W. Paty, P.A. Senior, D. Bigam, et al., Assessment of the severity of hypoglycemia and glycemic lability in type 1 diabetic subjects undergoing islet transplantation, Diabetes 53 (4) (2004) 955–962. 10.2337/diabetes.53.4.955. [DOI] [PubMed]
- 21.Wójcicki J.M. J”-index. A new proposition of the assessment of current glucose control in diabetic patients. Horm. Metab. Res. 1995;27(1):41–42. doi: 10.1055/s-2007-979906. [DOI] [PubMed] [Google Scholar]
- 22.Kovatchev B.P., Cox D.J., Kumar A., Gonder-Frederick L., Clarke W.L. Algorithmic evaluation of metabolic control and risk of severe hypoglycemia in type 1 and type 2 diabetes using self-monitoring blood glucose data. Diabetes Technol. Ther. 2003;5(5):817–828. doi: 10.1089/152091503322527021. [DOI] [PubMed] [Google Scholar]
- 23.Hill N.R., Hindmarsh P.C., Stevens R.J., Stratton I.M., Levy J.C., Matthews D.R. A method for assessing quality of control from glucose profiles. Diabet. Med. 2007;24(7):753–758. doi: 10.1111/j.1464-5491.2007.02119.x. [DOI] [PubMed] [Google Scholar]
- 24.Hermanides J., Vriesendorp T.M., Bosman R.J., Zandstra D.F., Hoekstra J.B., Devries J.H. Glucose variability is associated with intensive care unit mortality. Crit. Care Med. 2010;38(3):838–842. doi: 10.1097/CCM.0b013e3181cc4be9. [DOI] [PubMed] [Google Scholar]
- 25.Schlichtkrull J., Munck O., Jersild M. The M-valve, an index of blood-sugar control in diabetics. Acta Med. Scand. 1965;177:95–102. doi: 10.1111/j.0954-6820.1965.tb01810.x. [DOI] [PubMed] [Google Scholar]
- 26.F.J. Service, G.D. Molnar, J.W. Rosevear, E. Ackerman, L.C. Gatewood, W.F. Taylor, Mean amplitude of glycemic excursions, a measure of diabetic instability, Diabetes 19 (9) (1970) 644–655. 10.2337/diab.19.9.644. [DOI] [PubMed]
- 27.Colberg S.R., Kannane J., Diawara N. Physical activity, dietary patterns, and glycemic management in active individuals with type 1 diabetes: an online survey. Int. J. Environ. Res. Public Health. 2021;18(17):9332. doi: 10.3390/ijerph18179332. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Dimova R., Chakarova N., Grozeva G., Kirilov G., Tankova T. The relationship between glucose variability and insulin sensitivity and oxidative stress in subjects with prediabetes. Diabetes Res. Clin. Pract. 2019;158:107911. doi: 10.1016/j.diabres.2019.107911. [DOI] [PubMed] [Google Scholar]
- 29.Hanefeld M., Sulk S., Helbig M., Thomas A., Köhler C. Differences in glycemic variability between normoglycemic and prediabetic subjects. J. Diabetes Sci. Technol. 2014;8(2):286–290. doi: 10.1177/1932296814522739. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Kaya A., Koçyiğit C., Çatlı G., Özkan E.B., Dündar B.N. The relationship between glycemic variability and inflammatory markers in obese children with insulin resistance and metabolic syndrome. J. Clin. Res. Pediatr. Endocrinol. 2017;9(3):202–207. doi: 10.4274/jcrpe.4031. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Fernström M., Fernberg U., Hurtig-Wennlöf A. Insulin resistance (HOMA-IR) and body fat (%) are associated to low intake of fruit and vegetables in Swedish, young adults: the cross-sectional lifestyle, biomarkers and atherosclerosis study. BMC Nutr. 2019;5:15. doi: 10.1186/s40795-019-0279-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Castro-Quezada I., Flores-Guillén E., Núñez-Ortega P.E., Irecta-Nájera C.A., Sánchez-Chino X.M., Mendez-Flores O.G., et al. Dietary carbohydrates and insulin resistance in adolescents from marginalized areas of Chiapas, México. Nutrients. 2019;11(12):3066. doi: 10.3390/nu11123066. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Lin Y.H., Huang Y.Y., Chen H.Y., Hsieh S.H., Sun J.H., Chen S.T., et al. Impact of carbohydrate on glucose variability in patients with type 1 diabetes assessed through professional continuous glucose monitoring: a retrospective study. Diabetes Ther. 2019;10(6):2289–2304. doi: 10.1007/s13300-019-00707-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Thomsen M.N., Skytte M.J., Astrup A., Deacon C.F., Holst J.J., Madsbad S., et al. The clinical effects of a carbohydrate-reduced high-protein diet on glycaemic variability in metformin-treated patients with type 2 diabetes mellitus: a randomised controlled study. Clin. Nutr. ESPEN. 2020;39:46–52. doi: 10.1016/j.clnesp.2020.07.002. [DOI] [PubMed] [Google Scholar]
- 35.Orskov C., Jeppesen J., Madsbad S., Holst J.J. Proglucagon products in plasma of noninsulin-dependent diabetics and nondiabetic controls in the fasting state and after oral glucose and intravenous arginine. J. Clin. Invest. 1991;87(2):415–423. doi: 10.1172/JCI115012. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Olivos D.R., McGrath L.E., Turner C.A., Montaubin O., Mietlicki-Baase E.G., Hayes M.R. Intraduodenal milk protein concentrate augments the glycemic and food intake suppressive effects of DPP-IV inhibition. Am. J. Physiol. Regul. Integr. Comp. Physiol. 306. 2014;3:R157–R163. doi: 10.1152/ajpregu.00358.2013. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Jakubowicz D., Froy O., Ahrén B., Boaz M., Landau Z., Bar-Dayan Y., et al. Incretin, insulinotropic and glucose-lowering effects of whey protein pre-load in type 2 diabetes: a randomised clinical trial. Diabetologia. 2014;57(9):1807–1811. doi: 10.1007/s00125-014-3305-x. [DOI] [PubMed] [Google Scholar]
- 38.J. Ma, J.E. Stevens, K. Cukier, A.F. Maddox, J.M. Wishart, K.L. Jones, et al., Effects of a protein preload on gastric emptying, glycemia, and gut hormones after a carbohydrate meal in diet-controlled type 2 diabetes, Diabetes Care 32 (9) (2009) 1600–1602. 10.2337/dc09-0723. [DOI] [PMC free article] [PubMed]
- 39.L. Monnier, C. Colette, S. Dejager, D. Owens, Magnitude of the dawn phenomenon and its impact on the overall glucose exposure in type 2 diabetes: is this of concern?, Diabetes Care 36 (12) (2013) 4057–4062. 10.2337/dc12-2127. [DOI] [PMC free article] [PubMed]
- 40.DeFronzo R.A. From the triumvirate to the ominous octet: a new paradigm for the treatment of type 2 diabetes mellitus. Diabetes. 2009;58(4):773–795. doi: 10.2337/db09-9028. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Peng J., Lu J., Ma X., Ying L., Lu W., Zhu W., et al. Breakfast replacement with a liquid formula improves glycaemic variability in patients with type 2 diabetes: a randomised clinical trial. Br. J. Nutr. 2019;121(5):560–566. doi: 10.1017/S0007114518003628. [DOI] [PubMed] [Google Scholar]
- 42.V.A. Mustad, R.A. Hegazi, D.S. Hustead, E.S. Budiman, R. Rueda, K. Maki, et al., Use of a diabetes-specific nutritional shake to replace a daily breakfast and afternoon snack improves glycemic responses assessed by continuous glucose monitoring in people with type 2 diabetes: a randomized clinical pilot study, BMJ Open Diabetes Res. Care. 8 (1) (2020) e001258. 10.1136/bmjdrc-2020-001258. [DOI] [PMC free article] [PubMed]
- 43.Clark C.A., Gardiner J., McBurney M.I., Anderson S., Weatherspoon L.J., Henry D.N., et al. Effects of breakfast meal composition on second meal metabolic responses in adults with type 2 diabetes mellitus. Eur. J. Clin. Nutr. 2006;60(9):1122–1129. doi: 10.1038/sj.ejcn.1602427. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The data that support the findings of this study are available from the corresponding author upon reasonable request.