Urinary C-Peptide Secretion(UCPS): Can It be A New Index to Reect Insulin Sensitivity and A Predictor in Pregnant Women?

Background: In order to assess the correlation between pregnant women's urinary C-peptide secretion(UCPS) and insulin sensitivity, and to examine the predictors of gestational diabetes. Methods: we recruited 166 women between 20 and 28 weeks of gestation. Their height and weight were measured to calculate the body mass index (BMI). 75g OGTT was carried out, so as to detect the serum glucose, serum insulin and C-peptide levels. Besides, the homeostasis model assessment of insulin resistance (HOMA-IR), homeostasis model assessment of β cell function (HOMA-β), Matsuda index, and area under the serum C-peptide curve (CPauc) were calculated. Additionally, the fasting urine specimen and all urine samples within 2 hours after OGTT were collected to determine the urinary C-peptide and urine volume to calculate UCPS. Results: We found that Fasting serum insulin (Fins), Fasting serum C-peptide (FCP), BMI, OGTT 2h UCPS(UCPS120), CPauc, Matsuda index and HOMA-IR of GDM group were higher than NGT group. FCP was correlated with UCPS0 (r=0.234, p=0.002) and HOMA-β(r=0.251, p=0.001). UCPS120 was correlated with CPauc(r=0.176 p<0.001), Matsuda index(r=-0.362, p<0.001) and HOMA-IR(r=0.336, p<0.001). The combination of BMI and UCPS120 was better than other index in predicting GDM, with a sensitivity of 72.0% and specicity of 70.7%. Conclusion: UCPS120 had the potential to be a new index to reect insulin sensitivity in pregnant women. To screening better than other index.


Background
Gestational diabetes mellitus (GDM) is a global health problem, which is mainly caused by impaired insulin action and β-cell dysfunction [1]. With the progress of pregnancy, maternal insulin sensitivity decreases by about 60 percent [2], and the increasing insulin antagonistic hormone will lead to maternal hyperinsulinemia [3]. In clinical practice, measuring insulin secretion is relatively straightforward using HOMA-β and CPauc. Matsuda index and HOMA-IR are commonly used to evaluate the insulin sensitivity [4,5].
To diagnose GDM and evaluate the islet function of pregnant women, we usually use the serum glucose, insulin and C-peptide detected in three or more points during OGTT. There are a number of clinical studies to explore relatively simpli ed indicators. Urinary C peptide is a non-invasive, easily operative, and highly acceptable index [6]. Studies have shown that Urinary C peptide creatinine ratio (UCPCR) is a practical and non-invasive method to assess insulin secretion [7,8]. But urine creatinine is easily affected by muscle content, which may lead to a higher UCPCR in women [9]. So our study rst attempted to use postprandial urinary C-peptide secretion(UCPS) to assess islet function in pregnant women between 20 and 28 weeks of gestation, and to investigate whether it can be used as an indicator for GDM screening.

Objects
Pregnant Han nationality women at 20-28 gestational weeks (according to Chinese Diabetes Society 2017 guidelines) undergoing routine checks in the outpatient department of our hospital from January 2017 to December 2017 were selected. All of them were preliminarily screened for the 75g OGTT to collect the fasting, 1 h and 2h plasma as well as urine specimens. All participants were then divided into GDM group and NGT group according to the serum glucose level. Those previously diagnosed with diabetes, combined with acute infection, taking glucocorticoids within the past 2 weeks, and combined with other chronic disease were excluded. This study was approved by Ethics Committee, and all enrolled participants had signed the informed consent.

Methods
1 General clinical data collection: the age, family history of diabetes, history of GDM, previous medical history, gestational week (w), height (m), and current weight (kg) of all enrolled pregnant women were recorded. Meanwhile, the current body mass index (BMI kg/m 2 ) was calculated.
2 Blood specimen collection and determination: for all enrolled pregnant women, 5ml fasting blood was collected at 8:00 in the morning after an 8h overnight fast, so as to determine the FBG, blood lipid, liver and kidney function, glycated hemoglobin (HbA1c), fasting insulin (FIns), and fasting C-peptide (FCP).
After fasting blood collection, 75g oral glucose tolerance test (OGTT) was taken, the venous blood was collected at 1h and 2h respectively, so as to determine the serum glucose, serum insulin and serum Cpeptide at corresponding time points.
3 Urine specimen collection and determination: the fasting urine (the participants should passed their overnight rst void urine before OGTT), as well as urine at OGTT 2h (all urine should be collected until 2h after taking the glucose) were collected. No food was allowed during the OGTT, while water was allowed.
The urine volume should be recorded each time, and 10 ml urine should be retained, which was transferred to the 5 ml frozen tubes and preserved in the refrigerator at -80℃ until detection.

Statistical analysis
Statistical analysis was performed using SPSS 13.0. A p value of 0.05 was considered statistically signi cant. Measurement data were expressed as mean ± standard deviation, and inter-group comparison was carried out using t test. For non-normal distribution data, Mann-Whitney U test was employed for comparison. The Pearson correlation coe cient and linear regression models were adopted for correlation analysis. Receiver operator characteristic curve (ROC) was used to assess the predictive ability for GDM of each index. (Table 1): 166 pregnant women were enrolled into this study, which were all Han nationality. 41 were diagnosed with GDM, 21 with family history of diabetes, accounting for 24.7% and 12.7% of all participants, respectively. There were no signi cant differences in age, gestational age and family history of diabetes between the two groups.  (Figure 1A, 1B).
3.6. To predict the GDM, we try to use the simple and non-invasive index, so BMI, UCPS120, HOMA-IR and Matsuda index are put into the model. The combination of BMI and UCPS120 was better than other index, with a sensitivity of 72.0% and speci city of 70.7% (Table 2, Figure 4).

Discussion
UCPS is a new index adopted in this study, which re ects the total secretion of C-peptide in urine within a certain period of time, while different from UCPCR. In our study, UCPS0 was associated with FCP and HOMA-β, but the r value was low. In other study, they asked all the women to pass their overnight rst void urine and collect the second void urine. We didn't emphasize to collect the second void urine, it may resulted in some discrepancy. UCPS120 of OGTT was correlated with CPauc, Matsuda index and HOMA-IR, and the Matsuda index had the strongest correlation. Previous study showed that Matsuda index had stronger correlation with hyperinsulinaemic-euglycaemic clamp in pregnant women [10]. Although UCPCR120 only had weak correlation with Matsuda index, due to the small sample size, it still had the potential to be a new index to re ect insulin sensitivity in pregnant women.
Comparing the indexes of the two groups, it was found that the BMI, UCPS120, Fins, FCP, CPauc, Matsuda index and HOMA-IR of the GDM group were signi cantly higher than the NGT group. In all of these indexes, the calculation of CPauc, Matsuda index and HOMA-IR requires more parameters. The determination of these parameters requires multiple blood tests. while BMI and UCPS120 are noninvasive ones. HOMA-IR and Matsuda index had better predictions for GDM in previous studies [11,12].
Recently researches showed that the pregnant women with high BMI before pregnancy had signi cantly high risk in GDM [13]. And pregnancy high BMI may be the independent risk factor for GDM [14]. We hope to use effective, simple and best non-invasive index to predict the GDM. So we choose BMI, UCPS120, HOMA-IR and Matsuda index. ROC curve analysis showed that to predict the GDM, the combination of BMI and UCPS120 was better than other index, with a sensitivity of 72.0% and speci city of 70.7%. Many hospitals in china now advocate 75 g OGTT for all pregnant women, but multiple blood collection during pregnancy may cause problems for some women. If non-invasive tests can be used to screen out highrisk groups, and then OGTT tests for high-risk groups, it may help to reduce the psychological pressure of pregnant women, but also to save medical resources.
There are still some limitations in this study, all women included were in the second trimester of pregnancy (20-28 weeks), and the sample size was small. The conclusion needs to be expanded to verify the sample size. Secondly, the women included in this study were all with normal renal function, and the effect of different renal function status on this index could not be determined. There is increasing evidence that lifestyle and therapeutic intervention after 18 weeks of pregnancy in at-risk women has little effect on preventing both GDM and fetal macrosomia [15][16][17][18]. Further studies on a larger scale are needed, such as women with different renal function states in the early stage of pregnancy, to evaluate whether UCPS can be applied to women in early pregnancy to facilitate early screening of high-risk populations and early intervention.

Conclusion
UCPS120 had the potential to be a new index to re ect insulin sensitivity in pregnant women. To screening the GDM, the combination of BMI and UCPS120 was better than other index.   Linear regression model for UCPS120 and CPauc, Y=2.968+0.003X, r=0.176, p=0.012. CPauc: area under the serum C-peptide curve C peptide, UCPS120: OGTT 2h urinary C peptide secretion.