In a population-based screening for type 2 diabetes, although biochemical hypoglycemia is not a common finding on oral glucose tolerance tests (OGTTs), it really exists in some subjects. Previously, it was thought to be potentially linked with an increased type 2 diabetes risk [
So far, there are very little data considering the association between biochemical hypoglycemia during an OGTT and insulin resistance [
Is biochemical hypoglycemia on an OGTT associated with insulin resistance in a Chinese population? The aim of this study was to find the association between biochemical hypoglycemia on a 2-hour screening OGTT and insulin resistance in a Chinese population.
Subjects of this study were sampled from the China National Diabetes and Metabolic Disorders Study that was conducted during 2007 and 2008 [
The anthropometrics and biochemical measurements were described previously as follows [
Biochemical hypoglycemia was defined as OGTT 2-hour glucose (Glu120) < 3.0 mmol/l. The cut point of 3.0 mmol/l for blood glucose was recommended by the definition of hypoglycemia according to American Diabetes Association (ADA) guidelines.
The Matsuda’s insulin sensitivity index (Matsuda ISI) was calculated as Stumvoll’s early Stumvoll’s late
All statistical analyses were performed using SPSS software, version 19.0. Data were expressed as mean ± SD (standard deviation) or as median with upper and lower quartiles. Continuous variables were compared by using a one-way analysis of variance (ANOVA) test, while a frequency of dichotomous variables was performed by
In total, 26,606 participants were included in this study analysis. Of these participants, 141 of them were diagnosed with biochemical hypoglycemia on a 2-hour OGTT with a prevalence 0.53%, and 26,465 of them were diagnosed with NGT. Patients with IGR, as well as diabetes, were excluded from this analysis.
The average age of the patients with biochemical hypoglycemia was 42.5 ± 13.1 years, and the average BMI of the patients was 22.8 ± 3.2 kg/m2. The average FPG, Glu30 level, and Glu120 level was 4.62 ± 0.65 mmol/l, 7.82 ± 2.07 mmol/l, and 2.92 ± 0.07 mmol/l, respectively. The average fasting insulin secretion, Ins30 level, and Ins120 level was 7.4 ± 4.6
Compared to participants with NGT, participants with biochemical hypoglycemia showed older age, lower BMI, and lower levels of fasting glucose, Glu30, and Glu120. In terms of insulin secretion, participants with biochemical hypoglycemia showed higher secretion levels of Ins30 but lower secretion levels of Ins120. Details are shown in Tables
Phenotypes of biochemical hypoglycemia compared with NGT.
PPG ≤ 3.0 mmol/l | PPG > 3.0 mmol/l |
|
|
---|---|---|---|
|
141 | 26,465 | — |
Sex (male percent) | 55% | 39% | ≤0.001 |
Age (years) | 42.5 ± 13.1 | 40.2 ± 12.7 | 0.03 |
SBP (mmHg) | 119.3 ± 17.9 | 118.7 ± 15.4 | 0.66 |
DBP (mmHg) | 77.2 ± 11.1 | 77.4 ± 12.3 | 0.84 |
BMI (kg/m2) | 22.8 ± 3.2 | 23.6 ± 3.6 | 0.01 |
Weight (kg) | 62.0 ± 11.5 | 62.0 ± 11.3 | 0.96 |
Waist (cm) | 79.5 ± 9.8 | 80.3 ± 10.5 | 0.36 |
WHR | 0.84 ± 0.07 | 0.84 ± 0.09 | 0.79 |
FPG (mmol/l) | 4.62 ± 0.65 | 4.95 ± 0.55 | ≤0.001 |
Glu30 (mmol/l) | 7.82 ± 2.07 | 8.09 ± 1.74 | 0.06 |
Glu120 (mmol/l) | 2.92 ± 0.07 | 5.69 ± 1.10 | ≤0.001 |
SUA ( |
200.5 ± 148. | 260.0 ± 117.7 | ≤0.001 |
CHO (mmol/l) | 4.60 ± 0.96 | 4.69 ± 0.91 | 0.25 |
TG (mmol/l) | 1.41 ± 0.98 | 1.36 ± 0.94 | 0.55 |
HDL-C (mmol/l) | 1.34 ± 0.34 | 1.36 ± 0.36 | 0.52 |
LDL-C (mmol/l) | 2.68 ± 0.83 | 2.67 ± 0.77 | 0.90 |
Characteristics of insulin secretion and insulin resistance of biochemical hypoglycemia compared with NGT.
Glu120 ≤ 3.0 mmol/l |
Glu120 > 3.0 mmol/l |
|
|
---|---|---|---|
|
141 | 26,465 | — |
Fasting insulin ( |
7.4 (5.0–8.9) | 8.1 (5.0–9.5) | 0.18 |
Ins30 ( |
53.0 (27.6–64.9) | 47.9 (22.8–60.1) | 0.12 |
Ins120 ( |
16.3 (7.5–17.7) | 32.4 (14.6–39.7) | ≤0.001 |
Masuda ISI | 10.72 (7.44–12.94) | 8.64 (5.55–10.79) | ≤0.001 |
Stumvoll early | 1338.4 (1162.6–1466.3) | 1012.3 (792.1–1180.1) | ≤0.001 |
Stumvoll late | 340.8 (310.8–354.8) | 260.7 (212.3–294.8) | ≤0.001 |
Compared to participants with NGT, participants with biochemical hypoglycemia showed higher levels of Matsuda ISI, which may indicate lower levels of insulin resistance. In terms of
Association of Masuda ISI, Stumvoll early index, and Stumvoll late index, with biochemical hypoglycemia in screened patients by multivariate logistic regression analysis.
Odds ratio | 95% CI |
|
|
---|---|---|---|
Fasting insulin | 0.959 | 0.915–1.006 | 0.09 |
Ins30 | 1.002 | 0.998–1.005 | 0.41 |
Ins120 | 0.927 | 0.911–0.943 | ≤0.001 |
Masuda ISI | 1.075 | 1.047–1.104 | ≤0.001 |
Stumvoll early index | 1.002 | 1.001–1.002 | ≤0.001 |
Stumvoll late index | 1.007 | 1.006–1.008 | ≤0.001 |
Adjusted by gender, age, BMI, serum uric acid level; Ins30: serum insulin level at 30 minutes after an OGTT; Ins120: serum insulin level at 120 minutes after an OGTT; Masuda ISI: the Matsuda’s insulin sensitivity index.
The insulin secretion of early stage and insulin sensitivity levels between participants with biochemical postprandial hypoglycemia and NGT.
The insulin secretion of late stage and insulin sensitivity levels between participants with biochemical postprandial hypoglycemia and NGT.
Biochemical hypoglycemia is a relatively rare finding on OGTTs; therefore, little attention has been paid to postchallenge biochemical hypoglycemia compared to hyperglycemia. According to this cross-sectional analysis based on such a large Chinese population, the prevalence of postchallenge biochemical hypoglycemia was 0.53%. Results from this study indicated that, compared to participants with NGT, participants with biochemical hypoglycemia had higher levels of insulin sensitivity, as well as higher levels of
The insulin sensitivity was evaluated as the level of Matsuda ISI, a standard figure for evaluating insulin resistance. Studies focusing on the insulin resistance and dysglycemia suggested that, compared to participants with NGT, patients with IGT or impaired fasting glycemia (IFG) had higher levels of HOMA-IR, which indicated that patients with postchallenge hyperglycemia tended to be more insulin resistant. On the other hand, the previous studies on postchallenge hypoglycemia also suggested that, compared to participants with NGT, patients with postchallenge hypoglycemia had higher levels of HOMA-IR and might be associated subsequently with IGT and diabetes. Arii et al. [
On the contrary, Tamburrano et al. [
This study was the first epidemiology study in such a large Chinese population that focused on postprandial hypoglycemia. As mentioned above, there were only a few previous studies focusing on reactive postprandial hypoglycemia, and few patients were included in each study. The recent research in a United Kingdom multiethnic population included more than 6000 participants, which is a significant number of patients; however, in that study, the researchers could only calculate HOMA-IR and QUICKI as the indexes for insulin resistance because only fasting and postprandial 2-hour glucose levels and insulin levels were recorded. In our study, with the recorded levels of both glucose and insulin at the time of fasting, also OGTT 30 minutes and 2 hours, a number of indexes were calculated to evaluate the level of insulin resistance as Matsuda ISI, and insulin secretion as the early and late stages of insulin secretion, Stumvoll early and late secretion indexes.
Of course, as a study, there were some limitations. First, we calculated different surrogate indexes of insulin sensitivity as well as insulin secretion, but it was well known that the gold standard should be represented by clamp-derived indexes. Unfortunately, in this large sample epidemiology study, the clamp test could not be made and these indexes could not be collected. Second, in terms of insulin secretion and sensitivity, there are very strong ethnic differences; however, just because this study was based on the data from the cross-sectional study in China, not in other ethnicities, we may not answer this ethnicity difference well.
With the whole evaluation, we confirmed that in a Chinese population, biochemical postprandial hypoglycemia was associated with higher levels of both insulin sensitivity and insulin secretion.
The funding agencies had no role in the study design, data collection or analysis, decision to publish, or preparation of the manuscript. There are no other disclosures to report for this research.
Linong Ji has received fees for lecture presentations and for consulting from Abbott, AstraZeneca, Bristol-Myers Squibb, Merck, Metabasis, Novartis, Eli Lilly, Roche, Sanofi-Aventis, and Takeda.
Xueyao Han and Xiaoling Cai conceived and designed the protocol; Xianghai Zhou, Lingli Zhou, and Simin Zhang performed the data extraction; Xiaoling Cai and Xueyao Han performed the statistical analyses; and Xiaoling Cai and Linong Ji prepared the manuscript outlines and drafts. All of the authors contributed to the critical revision of the manuscript drafts and gave final approval for submission of the manuscript content.
The authors thank all of the participants in the China National Diabetes and Metabolic Disorders Study group. This study was supported by the National High-Technology Research and Development Program of China (863 Program 2012AA02A509).