The Study to Investigate the Contribution of Basal and Postprandial Glucose to Overall Hyperglycemia in T2DM

Background The current evidence shows that hyperglycemia is one of the important cause of atherosclerosis。In the DCCT study[1], comparing with the conventional therapy group (HbA1c<9％), the risk of retinopathy, kidney disease and neuropathy was significantly reduced in the intensive therapy group (HbA1c<7％). The EDIC study[2] shows that there is a reduction(42% p=0.02) in the incidence of cardiovascular event in the intensive therapy group and the risk of nonfatal myocardial infarction, stroke, cardiovascular death reduced by 57%. (p=0.02). Based on those evidence, we could tell that a continued cardiovascular benefit after early intensive glucose control was evident among T1DM patients. Similarly, the UKPDS study[3,4] demonstrated that in the newly diagnosed T2DM patients, the incidence of the microvascular complication was 25% lower and the cardiovascular complication (include fatal and nonfatal MI) was 16% lower in the intensive glucose control group which is coincident with the conclusion from EDIC study. The blood glucose of the diabetes patients are consist of 3 parts：normal basal blood glucose ,basal hyperglycemia and postprandial hyperglycemia which is a further increase based on the basal hyperglycemia[5]. HbA1c is the standard indicator for glucose metabolism which is determined by Basal blood glucose (BBG) and post prandial glucose (PPG) levels. HbA1c is a risk factor for diabetes vascular lesions has become the consensus and UKPDS study shows that increasing HbA1c and fasting blood glucose levels maight be associated with the regression of beta cell function , basal hyperglycemia is a major cause of diabetic cardiovascular complications[6]. Most previous studies focused on HbA1c and FPG level to determine the control of blood glucose[7,8] .But now the relationship between PPG increasing and diabetes complications are gradually valued. Evidence shows PPG and HbA1c is main predictor of cardiovascular events and all-cause mortality in T2DM[9-11] . The contribution of basal and post-prandial blood glucose to overall glycaemia is one of point to investigate glucose metabolism impaired. Wenhui Li [12]et al found that the relationship between fasting, post-absorption blood glucose and HbA1c level is more closely than with PPG. Especially the blood glucose at 8:00 AM is closely related with HbA1c (r=0.84), The A1c-Derived Average Glucose study (ADAG) [13]also got the similar conclusion. However, some other study shows the relationship between PPG and HbA1c is much more closer [14]. The main reason that lead to this argument is the lack of an accepted accurate method to assess the contribution of basal and post-prandial blood glucose to overall glycaemia, and the contribution is associated with the choice of therapeutic strategy.

Currently, few studies have been done to investigate the newly diagnosed or treatment-naive type 2 diabetes, Only Peter R, et al. found that the contribution of fasting hyperglycemia derived from a standardized meal test to excess hyperglycemia increase as glycaemia control deteriorates, becoming dominant with an HbA1c in excess of 7.0%. While in the T2DM patients treated by OAD, different researchers have concluded differences, such as Monnier et al[5,16] found that in the T2DM patients treated by OAD, if HbA1c ≤7%, the relative contribution of PPG was around 69.7%, but this proportion decreased gradually accompany with the increasing HbA1c. When HbA1c≥10.2%, the contribution of PPG was only 30.5%; Kikuchi et al[17] concluded the similar result with Monnier in the T2DM patients of Japan. Riddle et al[18] found that when HbA1c≤8%, PPG contributes more to HbA1c, with the deterioration of blood glucose, the relative contribution of FPG increased to 70%.

However, blood glucose fluctuation is affected by many factors such as disease duration, sex, diet, food cooking methods and race[19]. Our previous study found compared with NGT, the intra-day blood glucose fluctuation was similar in IGR patients, but it had already occurred. In the newly diagnosed type 2 diabetic patients, the inter-day and intra-day blood glucose fluctuation was significantly increased, besides, the effect of different ratio of carbohydrate in the diet was different[20-24] For Asian yellow, there are less research and the conclusions are not consistent. Japanese scholars Kikuchi [17]obtained the similar conclusion with Monnier, but the study done by Taiwanese researchers [25] show that in the good blood glucose control patients, the relative contribution of PPG was up to 70% , but with the blood glucose control deterioration, FPG and PPG to its effect is similar, about 50% each.

Until now, the research on the contribution of the BBG and PPG to HbA1c is less, and the results are different, May be different from the research methods and the study population. At present, all the research use 6.1mmol/L (WHO criteria)or 5.6mmol/L(ADA criteria) as normal FBG level (WHO criteria) to calculate the fasting or postprandial hyperglycemia, not according to the blood glucose fluctuation curve of NGT population, so that the BBG and PPG contribution of HbA1c might be underestimated or overestimated. In addition, Peter and Monnier used MTT method which would cause difference result from the T2DM patients in the real world .Peter, Monnier, Riddle and Kikuchi monitored the blood glucose by collecting pre or postprandial blood sample frequently or SMBG. The contribution of the postprandial glucose to HbA1c is underestimated. As the continuous glucose monitoring system (CGMS) is becoming more and more widely used, it is considered as a better method to evaluate the contribution of BBG and PPG to HbA1c because of the minimizing bias which is resulted of research methodology[25]. There was no any China mainland data showing the contribution of BBG and PBG to HbA1c in T2DM patients treated with OADs using the CGM method. Therefore this study is aimed to investigate the contribution of BBG and PBG to HbA1c in Chinese T2DM patients treated with OADs using CGMS. It's expected to generate evidence to support the concept of individualized therapy when patients are uncontrolled by OADs.

Objectives Primary objective is to investigate the relative contribution of BBG and PBG to overall glycaemia in T2DM with OAD. The secondary is to investigate the absolute contribution of BBG and PBG to overall glycaemia in T2DM with OAD treatment, investigate the correlation between overall glycaemia exposure and HbA1c and regression equation of HbA1c to BBG and PBG.

Sample size According to MONNIER' research, the contribution of PBG to whole day hyperglycemia use 6.1mmol/L as base line is： HbA1c<7.3%， 69.7%±4.18%(n=58)； HbA1c(7.3-8.4%)，51.34±4.18%(n=58)； HbA1c(8.5-9.2%)，44.18±3.58%(n=58)； HbA1c(9.3-10.2%)，40.60±4.78%(n=58)； HbA1c>10.2%，30.5%±3.58%(n=58)； If there are 5 groups, the mean CSS is 854.94, SD is 33.55,α=0.05.After enter the SAS 8.01 software, every group has 22 sizes and totally 110 cases with 5 groups. So more than 132 cases will be enrolled ( rate of dropout and withdraw is 20%, the cases will need 110+110×0.2=132).

One-Way ANOVA

# Treatments = 5 CSS of Means = 854.94 Standard Deviation = 33.55 Alpha = 0.05

N per Power Group

0.800 17 0.850 19 0.900 22 （SAS 8.01 software）

Because this is exploratory research, so we justified the sample:

T2DM subjects: the primary endpoint is the relative contribution of FBG. The contribution will be calculated by individual subject. A sample size of 60 for each group produces a two-sided 95% confidence interval for relative contribution of FBG with a precision of approximately ±6.5% when the estimated standard deviation is 25% (refer to previous study), so the total subjects of five group is 300.

Normal glycaemic subjects: a sample size of 100 normal glycaemic subjects produces a two-sided 95% confidence interval for AUC with a precision of approximately ±0.08d∙mmol∙L-1when the estimated standard deviation is 0.4d∙mmol∙L-1(refer to previous study).

Number of subjects per treatment arm: N (normal glycaemic group)=100; N (T2DM with OADs)=300

Cases with one of the following reasons should be removed

• Misdiagnos
• No medication or no regular medication
• Irregular diet
• No test recording
• Using any drug which might interfere the glucose metabolism and mislead the result.

Study design This is a single center, two groups, cross sectional study.

• One group consists of normal glycaemic subjects, N=100.
• One group consists of T2DM subjects treated with OADs, N= 300, which will be divided into 5 sub-groups by HbA1c level and each sub-group will include 60 subjects (HbA1c<6.5%, 6.5-7.0%, 7.0-8.0%, 8.0-9.0%, >=9.0%) CGMS will be used to collect blood glucose data. The normal glycaemic subjects will not receive any treatment during this study. The T2DM subjects will maintain their previous OADs treatment and no other new treatment will be added during this study.

Relative and absolute contribution will be calculated by two ways:

• Refer the methods of Monnier and Riddle to calculate the contribution with WHO normal FBG cut point of 6.1 mmol/L
• AUC 24h total high glycaemia=the area above the curve of FBG 6.1 mmol/L
• AUCPPG=(the area above pre-prandial glucose in a 4-h period after each meal)X3 meals
• AUC BBG= AUC 24h total high glycaemia - AUCPPG
• Using the data of the group with normal glycaemic subjects as the basis to calculate the contribution (no need to select a normal FBG cut-off point, therefore avoid the overestimated or underestimated contribution of FBG due to the conflict of different FBG cut-off point in western guideline and China guideline with 5.6 mmol/L and 6.1 mmol/L respectively)
• AUC24h NGT total=the area above the curve of glucose 0 mmol/L
• AUC 24h T2DM total= the area above the curve of glucose 0 mmol/L
• Total high glycaemia= AUC 24h total high glycaemia = AUC 24hT2DM total -AUC 24h NGT total
• AUC PPG==(the area above post-prandial glucose in a 4-h period after each meal)X3 meals in T2DM patients
• AUC BBG= AUC 24h total high glycaemia- AUCPPG
• Relative glucose contribution calculation:
• Relative contribution of BBG= AUC BBG/ AUC 24h total high glycaemia X100%
• Relative contribution of PPG= AUC PPG/ AUC 24h total high glycaemiaX100%
• Absolute glucose contribution calculation:
• Refer Monnier, et al: Absolute PBG contribution= HbA1c×(AUCPBG/AUCtotal)
• Refer Peter, at al: Absolute PBG contribution= [（mean HbA1c-T2DM - mean HbA1c-NGT）×AUCPBG/ AUCtotal]; Absolute BBG contribution= [（mean HbA1c-T2DM - mean HbA1c-NGT）×（AUCtotal-AUCPBG）/ AUCtotal]

The contribution with ADA normal FBG cut point of 5.6mmol/L is the same as above.

Adverse events record and report

• Definition: any adverse medical events which happened during the study regardless of the relationship with Investigational product
• Adverse events information obtain: observed by the physician, reported by subjects. In addition ,the physician should ask the adverse events information in every visit
• Recording：Time, severity, duration, treatment, and outcome of the adverse event
• Standard for determining the severity of adverse events:
• llight: does not affect the normal function of subjects.
• moderate: to a certain extent affect the normal function of subjects.
• severe: significantly affect the normal function of subjects. Timeline
• 9/1/2015----11/1/2015 previous preparation，apply for ethical approval;
• 11/18/2015---7/31/2017 Screen subjects in community , complete the subjects who were eligible for the trial and to wear 72h continuous glucose monitor systerm;
• 8/1/2017---11/1/2016 Supplement experiment omission, organize data and data lock
• 11/1/2017---1/30/2018 Statistical analysis, draw conclusions, the paper writing. Complete the report of the project finishing..