The aim of the present study was to assess the level of glycaemic control by the measurement of 24 h blood glucose profiles and standard blood analyses under identical nutritional and physical activity conditions in patients with Type II diabetes and healthy normoglycaemic controls. A total of 11 male patients with Type II diabetes and 11 healthy matched controls participated in a 24 h CGMS (continuous subcutaneous glucose-monitoring system) assessment trial under strictly standardized dietary and physical activity conditions. In addition, fasting plasma glucose, insulin and HbA1c (glycated haemoglobin) concentrations were measured, and an OGTT (oral glucose tolerance test) was performed to calculate indices of whole-body insulin sensitivity, oral glucose tolerance and/or glycaemic control. In the healthy control group, hyperglycaemia (blood glucose concentration >10 mmol/l) was hardly present (2±1% or 0.4±0.2/24 h). However, in the patients with Type II diabetes, hyperglycaemia was experienced for as much as 55±7% of the time (13±2 h over 24 h) while using the same standardized diet. Breakfast-related hyperglycaemia contributed most (46±7%; P<0.01 as determined by ANOVA) to the total amount of hyperglycaemia and postprandial glycaemic instability. In the diabetes patients, blood HbA1c content correlated well with the duration of hyperglycaemia and the postprandial glucose responses (P<0.05). In conclusion, CGMS determinations show that standard measurements of glycaemic control underestimate the amount of hyperglycaemia prevalent during real-life conditions in Type II diabetes. Given the macro- and micro-vascular damage caused by postprandial hyperglycaemia, CGMS provides an excellent tool to evaluate alternative therapeutic strategies to reduce hyperglycaemic blood glucose excursions.
- continuous overall net glycaemic action (CONGA)
- continuous subcutaneous glucose-monitoring system (CGMS)
- glycated haemoglobin (HbA1c)
- Type II diabetes
Over the last 15 years, improvements in microdialysis and biosensor technology have enabled clinicians to reliably monitor plasma and/or interstitial glucose concentrations in an ambulatory and continuous way [1,2]. These, so-called, CGMSs (continuous subcutaneous glucose-monitoring systems) have proven quite useful in optimizing individual exogenous insulin administration in diabetes patients , since they provide information on ambulatory postprandial  and/or nocturnal glucose excursions . Moreover, it has been shown in both children and adults with type I diabetes that average 24 h blood glucose concentrations strongly correlate with HbA1c (glycated haemoglobin) concentrations [6,7]. However, the inter- and intra-individual day-to-day variation in glycaemic load, meal composition  and daily physical activity  can complicate therapeutic decision-making based on these 24 h blood glucose profiles [10,11]. Therefore, in order to compare CGMS results between normoglycaemic and diabetic subjects, standardization of both diet  and physical activity  is essential.
Epidemiological studies and preliminary intervention studies have shown that postprandial hyperglycaemia is a direct and independent risk factor for the development of cardiovascular disease . Importantly, the rapid postprandial increase in blood glucose concentrations or ‘hyperglycaemic spikes’ seem to be even more relevant to the onset of cardiovascular complications than merely elevated FPG (fasting plasma glucose) . Therefore therapeutic targets should be aimed at reducing postprandial blood glucose excursions. Although scientific studies on the prevalence of hyperglycaemic spikes in Type II diabetes are still scarce , recommendations on proper glycaemic control have been redefined recently [14,15].
To define abnormal postprandial blood glucose excursions and relate this to the pathogenesis of diabetic vascular complications, it is important to have more detailed information on normal postprandial blood glucose profiles in a non-insulin resistant population under exactly the same dietary ambulatory conditions. Therefore, in the present study, we investigated 24 h blood glucose profiles in patients with Type II diabetes on oral blood-glucose-lowering medication and healthy normoglycaemic controls under strictly standardized, but free-living, conditions. As such, the present study provides a frame of reference for future studies on the role of real-life postprandial hyperglycaemia in the pathogenesis of diabetic complications.
MATERIALS AND METHODS
A total of 11 long-term diagnosed male patients with Type II diabetes and 11 healthy, age- and BMI-matched, normoglycaemic control subjects were selected to participate in the present study. Subjects' characteristics are shown in Table 1. Exclusion criteria were impaired renal or liver function, severe obesity [BMI (body mass index) >35 kg/m2], cardiac disease, hypertension, diabetic complications and exogenous insulin therapy. All patients with Type II diabetes were treated with oral plasma-glucose-lowering medication [metformin only (n=3), or in combination with sulphonylureas (n=8)]. All medication was continued during the trials. All subjects were informed about the nature and the risks of the experimental procedures before their written informed consent was obtained. The study was approved by the local Medical Ethical Committee.
Before inclusion, all subjects first performed an OGTT (oral glucose tolerance test). Blood-glucose-lowering medication was withheld prior to the screening. After an overnight fast, subjects reported to the laboratory at 08.00 hours. A catheter (Baxter) was inserted into an antecubital vein and a resting blood sample was drawn after which a bolus of 75 g of glucose (dissolved in 250 ml water) was ingested (t=0 min). After the bolus was consumed, blood was sampled every 30 min until t=120 min. Plasma glucose concentrations were measured to determine glucose intolerance and/or Type II diabetes according to the World Health Organization criteria of 1999 . In addition, plasma glucose and insulin concentrations were used to assess insulin sensitivity using the OGIS (oral glucose insulin sensitivity) index for a 2 h OGTT, as described by Mari et al. , and whole-body insulin resistance using HOMA-IR (homoeostasis model assessment insulin resistance index) .
Blood sample analysis
Blood (10 ml) was collected in EDTA-containing tubes and centrifuged at 1000 g and 4 °C for 10 min. Aliquots of plasma were immediately frozen in liquid nitrogen and stored at −80 °C until analyses. Glucose concentrations (Uni Kit III; Roche) were analysed with the COBAS FARA semi-automatic analyser (Roche). Plasma insulin was determined in duplicate by RIA (HI-14K; Linco Research) and the results averaged. To determine HbA1c content, a 3 ml blood sample was collected in EDTA-containing tubes and analysed by HPLC (Bio-Rad Laboratories).
All experimental trials described in the present study are part of a larger project investigating the effects of nutritional interventions to improve glycaemic control in patients with Type II diabetes. On the first day, subjects reported to the laboratory in the afternoon and were instructed about their diet and on the use of the food intake and physical activity diaries. Next, subjects received instructions on the use of the capillary blood sampling method (Glucocard Memory PC; A. Menarini Diagnostics) used for the calibration of the CGMS. All subjects were instructed to measure capillary blood glucose concentrations before every meal. After the subjects were fully instructed, a microdialysis fibre (internal diameter of 0.17 mm and a cut-off mass of 18 kDa; Medica) was inserted into the peri-umbilical region, without anaesthesia, using an 18-gauge Teflon catheter as a guide, as described previously . For the measurements the microdialysis tube was then connected to a portable CGMS (GlucoDay® S; A. Menarini Diagnostics), which consists of a peristaltic pump that pumps Dulbecco's solution at 10 μl/min through the microdialysis fibre. A more detailed description of the device has been published previously . Briefly, the subcutaneous interstitial fluid is taken up by the microdialysis fibre and is transported to the measuring cell. The glucose sensor, consisting of immobilized glucose oxidase, measures the glucose concentration every second and stores an average value every 3 min for a total of 48 h. The entire device weighs approx. 250 g and is worn in a pouch under the subjects' clothes. After the CGMS was checked for proper functioning, subjects were provided with their diet (pre-weighed and packaged meals, drinks and snacks) and were allowed to return home and resume all their normal activities. CGMS data from the second test day (from 07.00–07.00 hours) were used for data analysis. The first period was used to familiarize subjects with the equipment and, therefore, was not used in the data analyses.
Diet and physical activity
All subjects maintained their normal physical activity patterns throughout the entire experimental period. Subjects refrained from heavy physical labour and exercise training for at least 3 days prior to and on the day of the trial. Subjects were asked to keep a comprehensive record of the time spent performing all activities (to the nearest 10 min), including sleeping, eating sitting, standing, watching television, occupational activity and household tasks, as well as information on the duration and relative intensity (e.g. light, moderate etc.) of all structured activities. The rate of energy expenditure for each activity was then determined using the Compendium of Physical Activities . Daily energy expenditure did not differ between groups and averaged 13.6±0.7 and 13.3±0.6 MJ/day in the patients with Type II diabetes and normoglycaemic controls respectively. All meals, snacks and beverages were provided in pre-weighed packages and ingested at pre-determined time points to ensure fully standardized dietary modulation. On the evening prior to the 24 h analysis period, all subjects received the same standardized meal [43.8 kJ/kg of body weight, consisting of 60 En% (energy %) carbohydrate, 28 En% fat and 12 En% protein]. The following day, the subjects were instructed to ingest their designated meals, drinks and snacks at set time points. Throughout this 24 h test period, subjects received a standardized diet (three meals and three snacks per day) representing an energy intake of 121 kJ·kg−1 of body weight·day−1, consisting of 64 En% carbohydrate, 25 En% fat and 11 En% protein. Before and after consuming a meal (i.e. breakfast, lunch and dinner), subjects were asked to obtain a capillary blood glucose sample (Glucocard Memory PC). The following day, the subjects reported back to our laboratory to obtain a non-fasting venous blood glucose measurement and to remove the CGMS. The acquired data were then downloaded from the device to a PC with GlucoDay® software (version 3.0.5). Values from the CGMS were converted into glucose values using the capillary glucose measurements as calibration values.
Statistics and data analyses
Data are means±S.E.M. Glucose responses were calculated as the mean glucose AUC (area under the curve) up to 6 h after each meal. Since the CGMS device provides an average glucose value every 3 min, AUC is expressed as mmol/l×3 min. To quantify and compare the glucose excursions in the control and diabetes population, AUC and the amount of time during which glucose concentrations were above 10.0 mmol/l or below 3.9 mmol/l were calculated. On the first and second study days, FPG was determined from the calibrated CGMS curves 10 min before breakfast and averaged. The non-fasting venous blood glucose measurement was used to calculate the CV (coefficient of variation) of the CGMS data. Relationships between CGMS parameters and standard measurements of insulin sensitivity were calculated using linear regression models.
To assess intra-day glycaemic variability, CONGA (continuous overall net glycaemic action), a novel method recently described by McDonnell et al. , was used. CONGAn has been defined as the S.D. of the differences in glucose concentration using varying time differences of n hours. We used CONGA1, CONGA2 and CONGA4, indicating intra-day glycaemic variability based on 1, 2 and 4 h time differences respectively. In normal non-diabetic subjects, CONGA values vary between 0.4 and 1.2, whereas values above 1.5 indicate glycaemic lability .
Before pooling data from all 22 subjects, homogeneity of regression was tested using ANCOVA (analysis of covariance) in order to exclude significant interactions. Time-dependent variables were tested using repeated-measures ANOVA with a Tukey–Kramer post-hoc test when applicable. For non-time-dependent variables, a Student t test for unpaired observations was applied. Significance was set at the 0.05 level of confidence. All statistical calculations were performed using the SPSS 12.0.1 software package.
Baseline and postprandial blood glucose responses are shown in Table 2. Total 24 h blood glucose responses in both diabetes patients and healthy controls are illustrated in Figure 1. Basal and mean glucose concentrations were significantly greater in the patients with Type II diabetes compared with the normoglycaemic controls (P<0.05, as determined by a Student t test). In the patients with Type II diabetes, the prevalence of hyperglycaemia (>10.0 mmol/l) was 55±7% of the 24 h period. In contrast, in the normoglycaemic controls, the prevalence of hyperglycaemia was 1.6±1%. As such, hyperglycaemia was present for 13.3±1.7 h and 0.38±0.2 h respectively.
The postprandial AUCs >10.0 mmol/l following breakfast, lunch and dinner contribute 46±7%, 29±3% and 11±3% respectively, to the total amount of hyperglycaemia present during the 24 h monitoring period in our patients with Type II diabetes. This breakfast-related hyperglycaemia was significantly greater (P<0.01, as determined by ANOVA) compared with the amount of hyperglycaemia during the evening or night.
In both patients with Type II diabetes and healthy controls, the average CONGA1 values following breakfast were significantly raised compared with the 6 h following lunch and dinner (Table 2; P<0.01, as determined by ANOVA). CONGA1 values were lowest during the night (P<0.01, as determined by ANOVA) and did not differ between groups from 01.00–06.00 hours (P>0.05, as determined by ANOVA; Table 2).
In the present study, the CV between interstitial CGMS glucose values and venous blood glucose was on average 8.0±1.3%.
Correlations between CGMS parameters and our standard measurements for glycaemic control are shown in Table 3. In the patients with Type II diabetes, HbA1c levels correlated well with the average 24 h blood glucose concentrations (R=0.81, P<0.01), the time during which blood glucose levels were >10 mmol/l (R=0.70, P<0.05), and postprandial AUC following lunch (R=0.80, P<0.01) and dinner (R=0.87, P<0.01). In a subgroup of patients with Type II diabetes with apparent acceptable glycaemic control (HbA1c ≤7.0; n=6), hyperglycaemia was present for 46±8% of the day (11.0±1.9 h).
In both the patients with Type II diabetes and control group, mean 24 h and nocturnal blood glucose concentrations correlated strongly with FPG levels (R values were between 0.61–0.86; P<0.05). In both groups, no significant correlations were reported between the 24 h CONGA indices and HbA1c content; however, in the patients with Type II diabetes, a significant correlation was found between postprandial CONGA1 values and AUC in the 6 h following a meal (R=0.47, P<0.01). When pooling the data from both groups, the 24 h CONGAn values correlated significantly with blood HbA1c content (R=0.53–0.66, P<0.01), mean 24 h glucose concentrations (R=0.73–0.77, P<0.001) and, to a lesser extent, with mean FPG concentrations (R=0.50–0.52, P<0.05). Also, a significant correlation was found between postprandial CONGA1 values and AUC 6 h following a meal (R=0.60, P<0.001)
The present study shows that, under normal standardized dietary conditions, patients with Type II diabetes using oral blood-glucose-lowering medication experience a substantial amount of hyperglycaemia for more than 13 h within a 24 h period. This disturbance in blood glucose homoeostasis is predominantly present following breakfast. After comparing 24 h blood glucose profiles between healthy normoglycaemic controls and patients with Type II diabetes under usual medical care by a general practitioner, it seems clear that standard treatment schemes with oral blood-glucose-lowering drugs appear to have insufficient therapeutic strength to normalize postprandial hyperglycaemia. Given the clinical relevance of the hyperglycaemic spikes , CGMS provides an excellent tool to evaluate the level of glycaemic stability in patients with Type II diabetes.
The concept that oral blood-glucose-lowering therapy provides inadequate protection against hyperglycaemia is not new [22–24]. Epidemiological studies and preliminary intervention studies have shown that postprandial hyperglycaemia is a direct and independent risk factor for the development of cardiovascular disease . However, the rapid postprandial increase in blood glucose concentrations seems to be more relevant to the onset of cardiovascular complications than merely elevated FPG concentrations . Therefore more detailed information on 24 h blood glucose profiles in a diabetic state is essential to increase our understanding of the relationship between hyperglycaemia, glucotoxicity and cardiovascular morbidity. In an attempt to assess postprandial glycaemic instability in Type II diabetes, we applied CGMS in patients with diabetes and compared this with blood glucose profiles of normoglycaemic subjects under strict nutritional and exercise standardization, but otherwise free-living conditions. In most of our normoglycaemic subjects, hyperglycaemia or glycaemic instability was not detectable. In contrast, despite healthy dietary conditions and continued use of oral blood-glucose-lowering medication, according to standard primary care  and international guidelines , the patients with Type II diabetes were hyperglycaemic for more than 13 h/day when using exactly the same diet as the normoglycaemic controls. In accordance with previous observations by Monnier et al. , the present study shows that postprandial hyperglycaemia was most prominent following breakfast and less evident during the night. As we provided a healthy balanced diet (43.8 kJ/kg of body weight, consisting of 60 En% carbohydrate, 28 En% fat and 12 En% protein), it could be speculated that the total amount of hyperglycaemia may even be worse under normal unrestricted dietary conditions. The observed levels of hyperglycaemia during the day (13±2 h over 24 h) are unacceptable and probably cause the excess formation of AGEs (advanced glycation end-products) , causing the macro- and micro-vascular damage .
Consistent with previous studies [29–31], our present findings emphasize the need for different types of interventional strategies in patients with Type II diabetes. It should be noted that there was a weak, non-significant, correlation between FPG and the percentage of hyperglycaemia (R=0.50, P>0.05; Table 3). The latter indicates that FPG is unlikely to be of sufficient sensitivity to successfully evaluate new treatment strategies that focus on reducing postprandial hyperglycaemia. For more long-term evaluation purposes, changes in blood HbA1c concentrations have generally been assessed, since blood HbA1c content correlates relatively well with both mean 24 h [32,33] and postprandial [6,7] glucose levels. In accordance, in the present study, we observed strong correlations between HbA1c and mean 24 h glucose and postprandial glucose levels following lunch and dinner (Table 3). It should be mentioned here that, even under clinically acceptable HbA1c levels (i.e. HbA1c ≤7.0 in six out of 11 patients with Type II diabetes), hyperglycaemia can still be unacceptably high at 11±2 h of blood glucose excursion >10 mmol/l per 24 h. Therefore these results extend earlier findings [24,34,35], and strongly suggest that the ability of HbA1c to monitor postprandial hyperglycaemia is debatable. Moreover, the measurement of prospective changes in blood HbA1c content only has sufficient sensitivity to detect changes in glucose homoeostasis during mid- to long-term interventions . Therefore the present study underlines the notion that CGMS is a promising tool when evaluating short-term (<3 months) changes in blood glucose homoeostasis following pharmacological, dietary and/or exercise interventions .
Another benefit of the CGMS approach, which has potential clinical application as well, is the possibility to calculate the level of glycaemic instability in insulin-resistant states. This so-called CONGAn is probably a more appropriate measurement to assess short-term changes in glucose homoeostasis throughout the day . This CGMS measurement reflects the S.D. of the differences in glucose concentration using various time windows . Therefore we determined CONGAn values in both our patients with Type II diabetes and normoglycaemic controls (Table 2). The proposed sensitivity of CGMS to detect subtle variations in glycaemic control was confirmed in our normoglycaemic control group. Interestingly, two of our control subjects appeared to have rather high postprandial CONGA1 values that almost approached values observed in the patients with Type II diabetes (i.e. average postprandial CONGA1 >2.1). These two subjects also showed the highest insulin values during the OGTT and were the only ‘normoglycaemic’ subjects who showed some hyperglycaemia throughout the day (results not shown). Taken together, our results suggest that more advanced CGMS analysis techniques provide promising measurements to assess glycaemic instability in patients with Type II diabetes . Research is warranted to investigate the diagnostic value of CGMS in other diabetes-related populations, such as patients in a pre-diabetic and/or insulin-resistant state.
In conclusion, detailed analyses of 24 h blood glucose profiles show that standard measurements of glycaemic stability grossly underestimate the amount of hyperglycaemia during real-life conditions in patients with Type II diabetes. Given the macro- and micro-vascular damage caused by postprandial hyperglycaemia, CGMS provides an excellent tool to evaluate additional therapeutic strategies more directly to reduce the amount of glycaemic instability and risk of cardiovascular complications in patients with Type II diabetes.
We thank A. Menarini Diagnostics, and Hanneke van Milligen for the technical support. We gratefully acknowledge our volunteers for participating in the experimental trials.
Abbreviations: AUC, area under the curve; CGMS, continuous subcutaneous glucose-monitoring system; CONGA, continuous overall net glycaemic action; CV, coefficient of variation; En%, energy %; FPG, fasting plasma glucose; HbA1c, glycated haemoglobin; HOMA-IR, homoeostasis model assessment insulin resistance index; OGIS, oral glucose insulin sensitivity; OGTT, oral glucose tolerance test
- The Biochemical Society