Both deteriorated insulin sensitivity (SI) and impaired insulin secretion are recognized as 2 of the foremost forms of pathophysiology for type 2 diabetes (T2DM) [
Two phases of insulin secretion are widely known: the first-phase insulin secretion (FPIS) and the second-phase insulin secretion (SPIS). The FPIS is normally secreted by the
Numerous studies have documented the link between metabolic syndrome (MetS), the clustering of hyperglycemia, hypertension, obesity, and dyslipidemia, and an elevated risk of developing cardiovascular disease and diabetes [
In total, 186 participants were enrolled in this study. Either they were self-referred or health professionals had referred them for diabetes screening. They had no history of diabetes and, therefore, they took no diabetes medications at the time of the study. They were defined as having a normal glucose tolerance (NGT), PreDM, and T2DM according to the criteria published by the American Diabetes Association in 2012 [
Frequently sampled intravenous glucose tolerance test (FSIGT): all tests were performed at the clinical research center. On the day of the study visit, after a 12 h overnight fast, one catheter was placed on both arms of each participant. A bolus of 10% glucose water (0.3 g/kg) was given. Another bolus of regular human insulin (Novo Nordisk Pharmaceutical, Princeton) 0.05 units/kg was injected 20 min after the glucose load. Blood samples for plasma glucose and insulin levels were collected at 0 min, 2 min, 4 min, 8 min, 19 min, 22 min, 30 min, 40 min, 50 min, 70 min, 100 min, and 180 min. The data were inputted into a Bergman minimal model [
The calculations of HOMA-IR and HOMA-
Plasma was separated within 1 h of blood withdrawal and stored at −30°C until the analysis. Plasma glucose was measured using a glucose analyzer by employing an oxidase method (YSI Model 203, Scientific Division, Yellow Spring Instrument Company, Inc., Yellow Spring, OH, USA). Plasma insulin was assayed using a commercial solid phase radioimmunoassay technique (Coat-A-Count insulin kit, Diagnostic Products Corporation, Los Angeles, CA, USA) with intra- and interassay coefficients of variance of 3.3% and 2.5%, respectively. Serum TG was measured using the Fuji Dri-Chem 3000 analyzer (Fuji Photo Film Corporation, Minato-Ku, Tokyo, Japan) by employing the dry multilayer analytical slide method. The serum HDL-C concentration was determined using the enzymatic cholesterol assay method after dextran sulfate precipitation.
The data were tested for normal distribution by using the Kolmogorov-Smirnov test and for the homogeneity of variances by using the Levene test. Continuous variables were expressed as mean ± standard deviation. Among the data, FPIS, FPI, SI, and DI were not normally distributed and were logarithmically transformed. An independent
These equations were subsequently used to calculate the FPIS among the remaining 25% of participants. The correlation between the calculated FPIS and measured FPIS was measured using Pearson’s
All statistical analyses were performed using the SPSS software system, version 13.0 (SPSS Inc., Chicago, IL, USA). All
In the study, 140 and 46 participants were classified into the study group and the external validation group, respectively. Table
Demographic data of the study and external validation groups.
Study group | Ext. val. group |
|
|
---|---|---|---|
|
140 | 46 | |
Sex (male/female) | 69/71 | 25/21 | 0.552 |
Age (y) | 50.7 ± 13.5 | 50.8 ± 14.7 | 0.910 |
Body mass index (kg/m2) | 25.2 ± 3.9 | 25.8 ± 5.1 | 0.366 |
Systolic blood pressure (mmHg) | 121.5 ± 13.0 | 118.3 ± 16.2 | 0.186 |
Diastolic blood pressure (mmHg) | 76.3 ± 8.0 | 73.4 ± 7.7 | 0.076 |
Triglyceride (mmol/L) | 1.3 ± 0.6 | 1.5 ± 0.6 | 0.112 |
HDL-C (mmol/L) | 1.1 ± 0.3 | 1.1 ± 0.4 | 0.350 |
Fasting plasma glucose (mmol/L) | 7.8 ± 2.8 | 7.5 ± 3.0 | 0.540 |
Fasting plasma insulin (pmol/L) | 30.5 (12.2–61.3) | 27.6 (14.4–62.0) | 0.353 |
First-phase insulin secretion ( |
115.0 (23.5–426.4) | 114.9 (24.4–430.4) | 0.822 |
Insulin sensitivity (10−4⋅min−1⋅pmol−1⋅L−1) | 1.274 (0.5–3.4) | 1.6 (0.3–3.3) | 0.501 |
Disposition index | 87.8 9 (20.3–900.8) | 95.3 (17.0–408.7) | 0.830 |
Glucose effectiveness (10−2⋅dL⋅min−1⋅kg−1) | 0.016 ± 0.010 | 0.015 ± 0.010 | 0.314 |
HOMA-IR | 1.7 (0.6–3.1) | 1.7 (0.7–3.1) | 0.615 |
HOMA- |
22.1 (6.9–83.0) | 20.0 (8.5–92.0) | 0.334 |
Data are expressed as mean ± SD or median (interquartile range). Ext. val. group: external validation group.
HDL-C: high-density lipoprotein cholesterol; HOMA-IR and HOMA-
Demographic data of normal glucose tolerance, prediabetes, and diabetes groups.
Normal glucose tolerance | Prediabetes | Diabetes | |
---|---|---|---|
|
51 | 40 | 95 |
Age (y) | 42.5 ± 17.22,3 | 54.4 ± 11.91 | 53.7 ± 10.31 |
Body mass index (kg/m2) | 26.1 ± 5.9 | 24.9 ± 3.1 | 25.1 ± 3.5 |
Systolic blood pressure (mmHg) | 118.1 ± 10.9 | 121.0 ± 14.7 | 121.9 ± 14.9 |
Diastolic blood pressure (mmHg) | 74.0 ± 6.8 | 76.1 ± 8.5 | 76.4 ± 8.5 |
Triglyceride (mmol/L) | 1.2 ± 0.6 | 1.4 ± 0.6 | 1.4 ± 0.6 |
HDL-C (mmol/L) | 1.1 ± 0.3 | 1.1 ± 0.3 | 1.1 ± 0.3 |
Fasting plasma glucose (mmol/L) | 4.6 ± 0.52,3 | 6.4 ± 0.41,3 | 9.9 ± 2.21,2 |
Fasting plasma insulin (pmol/L) | 49.5 (9.3–81.1) | 25.5 (7.5–61.7) | 23.0 (14.4–44.3) |
First-phase insulin secretion ( |
517.5 (183.0–5144.7)2,3 | 123.6 (35.5–390.7)1 | 37.8 (11.6–158.3)1 |
Insulin sensitivity (10−4⋅min−1⋅pmol−1⋅L−1) | 0.8 (0.2–3.2) | 1.9 (0.6–4.4) | 1.4 (0.6–2.9) |
Disposition index | 893.9 (240.0–2447.1)2,3 | 54.8 (21.7–894.6)1 | 40.7 (8.3–182.5)1 |
Glucose effectiveness (10−2⋅dL⋅min−1⋅kg−1) | 0.020 ± 0.0102,3 | 0.014 ± 0.0081 | 0.014 ± 0.0101 |
HOMA-IR | 1.7 (0.4–2.7) | 1.3 (0.3–3.1) | 1.7 (0.8–3.3) |
HOMA- |
134.0 (27.9–352.4)2,3 | 29.0 (10.4–72.6)1,3 | 13.0 (6.2–26.4)1,2 |
Data are expressed as mean ± SD or median (interquartile range). HDL-C: high-density lipoprotein cholesterol; HOMA-IR and HOMA-
To identify the parameters that contribute most to the FPIS, the correlations between the FPIS and different parameters were evaluated; the results are shown in Table
Pearson correlation between the clinical parameters and log (first-phase insulin secretion) in the study group.
Variables |
|
|
---|---|---|
Age | −0.398 | 0.000 |
Body mass index | 0.264 | 0.002 |
Systolic blood pressure | −0.044 | 0.623 |
Diastolic blood pressure | 0.030 | 0.740 |
Triglyceride | −0.064 | 0.463 |
HDL-C | −0.190 | 0.034 |
Fasting plasma glucose | −0.475 | 0.000 |
Log (FPI) | 0.382 | 0.000 |
Log (insulin sensitivity) | −0.184 | 0.035 |
Log (HOMA-IR) | 0.231 | 0.006 |
Log (HOMA- |
0.551 | 0.000 |
HDL-C: high-density lipoprotein cholesterol; FPI: fasting plasma insulin; HOMA-IR and HOMA-
Only MetS components were used in multiple linear regression analysis. Three of them were selected from regression analysis, and the equation was built and is shown as log (FPIS) = 1.477 − 0.119 × FPG + 0.079 × BMI − 0.523 × HDL-C (standard coefficients are shown in Table
Multiple linear regression of the associated factors with log (first-phase insulin secretion) in the 2 equations.
Variables | MetS components |
MetS components + FPI |
---|---|---|
Fasting plasma glucose | −0.386 (0.000) | −0.415 (0.000) |
Body mass index | 0.361 (0.000) | 0.269 (0.001) |
HDL-C | −0.181 (0.028) | −0.177 (0.017) |
Log (FPI) | — | 0.288 (0.005) |
Beta: standardized coefficients; MetS: metabolic syndrome; HDL-C: high-density lipoprotein cholesterol; FPI: fasting plasma insulin.
The correlation between the calculated first-phase insulin secretion and measured first-phase insulin secretion by using metabolic syndrome components in the external validation group.
Because the FPI is also considered a surrogate for the FPIS, it was also added to multiple linear regression analysis to build a second equation and, unlike the first, 4 factors were selected, and the following equation was formulated: log (FPIS) = 1.532 − 0.127 × FPG + 0.059 × BMI − 0.511 × HDL-C + 0.375 × log (FPI). The difference of predicting power of FPIS between the first and the second equation was determined using hierarchical multiple regression method. The
The correlation between the calculated FPIS and the measured FPIS in the external group was also evaluated, and the results are shown in Figure
The correlation between the calculated first-phase insulin secretion and measured first-phase insulin secretion by using metabolic syndrome components and fasting plasma insulin in the external validation group.
In this study, we built an equation by using routine clinical measurements and MetS components to predict the FPIS in participants with different levels of glucose tolerance. Because of the tight correlation between FPI and FPIS, FPI was also added into analysis to build a second equation to improve predictive accuracy. To verify our results, external validation was also performed. Although previous studies have been done to predict the FPIS, most of them enrolled only nondiabetic participants [
The FPIS is the immediately releasable stored insulin in
In our study, the FPG, BMI, HDL-C, and FPI were selected among all other factors and inputted into multiple linear regression analysis. Because the predominant function of
Following the FPG, the BMI was the second most critical factor inputted in the model. The results were not surprising because the evidence has shown that people with a higher BMI would have a better
Because the lower HDL-C is associated with IR [
The FPI, which is not a routinely used measurement, is associated with
The
To the best of our knowledge, the current study is the first to formulate an equation for estimating the FPIS by using the MetS components and the FPI level. However, our study has limitations. First, the body fat content and its distribution, which were known to be associated with IR and the
In conclusion, by using the MetS components, the FPIS could be predicted with reliable accuracy (
Insulin sensitivity
Type 2 diabetes
Insulin resistance
First-phase insulin secretion
Second-phase insulin secretion
Prediabetes
Metabolic syndrome
Body mass index
Triglyceride
High-density lipoprotein cholesterol
Fasting plasma glucose
Fasting plasma insulin
Normal glucose tolerance
Frequently sampled intravenous glucose tolerance test
Acute insulin response after glucose load
Glucose effectiveness
Disposition index
Homeostasis model assessment of insulin resistance
Homeostasis model assessment of the
The authors have no conflict of interests.
Chang-Hsun Hsieh analyzed the data. Jiunn-diann Lin wrote the paper. Chung-Ze Wu and Yen-Lin Chen reviewed and edited the paper. Dee Pei contributed to the discussion and edited the paper. Wei-Cheng Lian, Chun-Hsien Hsu, An-Tsz Hseih, and Chuan Chieh Liu analyzed the data and contributed to the discussion.
The authors thank all participants of the study.