Josina M. Rijkelijkhuizen

Relationship Between A1C and Glucose Levels in the General Dutch Population: The New Hoorn Study

OBJECTIVE To investigate the relationship among A1C, fasting plasma glucose (FPG), and 2-h postload plasma glucose in the Dutch general population and to evaluate the results of using A1C for screening and diagnosis of diabetes. RESEARCH DESIGN AND METHODS In 2006–2007, 2,753 participants of the New Hoorn Study, aged 40–65 years, who were randomly selected from the population of Hoorn, the Netherlands, underwent an oral glucose tolerance test (OGTT). Glucose status (normal glucose metabolism [NGM], intermediate hyperglycemia, newly diagnosed diabetes, and known diabetes) was defined by the 2006 World Health Organization criteria. Spearman correlations were used to investigate the agreement between markers of hyperglycemia, and a receiver operating characteristic (ROC) curve was calculated to evaluate the use of A1C to identify newly diagnosed diabetes. RESULTS In the total population, the correlations between fasting plasma glucose and A1C and between 2-h postload plasma glucose and A1C were 0.46 and 0.33, respectively. In patients with known diabetes, these correlations were 0.71 and 0.79. An A1C level of ≥5.8%, representing 12% of the population, had the highest combination of sensitivity (72%) and specificity (91%) for identifying newly diagnosed diabetes. This cutoff point would identify 72% of the patients with newly diagnosed diabetes and include 30% of the individuals with intermediate hyperglycemia. CONCLUSIONS In patients with known diabetes, correlations between glucose and A1C are strong; however, moderate correlations were found in the general population. In addition, based on the diagnostic properties of A1C defined by ROC curve analysis, the advantage of A1C compared with OGTT for the diagnosis of diabetes is limited.

High risk of cardiovascular mortality in individuals with impaired fasting glucose is explained by conversion to diabetes - The Hoorn Study

OBJECTIVE—To optimize identification of future diabetic patients, the American Diabetes Association (ADA) introduced criteria for impaired fasting glucose (IFG) in 1997 (IFG 6.1 mmol/l [IFG6.1]) and lowered the threshold from 6.1 to 5.6 mmol/l (IFG5.6) in 2003. Our aim was to assess the consequences of lowering the IFG cutoff on the risk of cardiovascular disease (CVD) mortality and to evaluate whether this risk is explained by a conversion to type 2 diabetes within 6.4 years. RESEARCH DESIGN AND METHODS—In a population-based cohort, the Hoorn Study, plasma glucose was determined in 1989 and 1996 (n = 1,428). Subjects were classified in 1989 according to 1997 and 2003 ADA criteria. Subjects with IFG in 1989 were further classified according to diabetes status in 1996. Hazard ratios for CVD mortality (n = 81) in the period 1996–2005 were adjusted for age and sex. RESULTS—Subjects with IFG6.1, but not IFG5.6, had a significantly higher CVD mortality risk than normal fasting glucose (NFG) subjects. Subjects who converted from IFG to diabetes (IFG6.1: 42%; IFG5.6: 21%) had a more than twofold risk of CVD mortality (IFG6.1: 2.47 [1.17–5.19]; IFG5.6: 2.14 [1.12–4.10]) than subjects with NFG. IFG subjects who did not develop diabetes did not have significantly higher CVD mortality risks (IFG6.1: 1.50 [0.72–3.15]; IFG5.6: 1.15 [0.69–1.93]). CONCLUSIONS—The lower cutoff for IFG (ADA 2003 criteria) results in a category of IFG that no longer represents a high-risk state of CVD. Furthermore, only subjects who convert from IFG to diabetes have a high risk of CVD mortality.

Effects of meal size and composition on incretin, alpha-cell, and beta-cell responses

The incretins glucagon-like peptide-1 (GLP-1) and glucose-dependent insulinotropic polypeptide (GIP) regulate postprandial insulin release from the beta-cells. We investigated the effects of 3 standardized meals with different caloric and nutritional content in terms of postprandial glucose, insulin, glucagon, and incretin responses. In a randomized crossover study, 18 subjects with type 2 diabetes mellitus and 6 healthy volunteers underwent three 4-hour meal tolerance tests (small carbohydrate [CH]-rich meal, large CH-rich meal, and fat-rich meal). Non-model-based and model-based estimates of beta-cell function and incremental areas under the curve of glucose, insulin, C-peptide, glucagon, GLP-1, and GIP were calculated. Mixed models and Friedman tests were used to test for differences in meal responses. The large CH-rich meal and fat-rich meal resulted in a slightly larger insulin response as compared with the small CH-rich meal and led to a slightly shorter period of hyperglycemia, but only in healthy subjects. Model-based insulin secretion estimates did not show pronounced differences between meals. Both in healthy individuals and in those with diabetes, more CH resulted in higher GLP-1 release. In contrast with the other meals, GIP release was still rising 2 hours after the fat-rich meal. The initial glucagon response was stimulated by the large CH-rich meal, whereas the fat-rich meal induced a late glucagon response. Fat preferentially stimulates GIP secretion, whereas CH stimulates GLP-1 secretion. Differences in meal size and composition led to differences in insulin and incretin responses but not to differences in postprandial glucose levels of the well-controlled patients with diabetes.

Classical and model-based estimates of beta-cell function during a mixed meal vs. an OGTT in a population-based cohort

This study compared classical and model-based beta-cell responses during an oral glucose tolerance test (OGTT) and a meal tolerance test (MTT) in a population-based cohort. Individuals with normal glucose metabolism (NGM, n=161), impaired glucose metabolism (IGM, n=19) and type 2 diabetes mellitus (DM, n=20) underwent a 75 g-OGTT and an MTT (75 g carbohydrates, 50 g fat, 24 g proteins). Classical estimates of beta-cell function (insulinogenic index and the ratio of areas under insulin and glucose curves) were calculated. Mathematical modelling was used to determine beta-cell glucose sensitivity, rate sensitivity and potentiation. Insulin sensitivity was characterized by three surrogate estimates. Both classical and model-based estimates of beta-cell function were higher during MTT than during OGTT (P<0.05). Regarding the model-based parameters, especially beta-cell sensitivity was increased following MTT as compared with OGTT (P<0.05). Both during OGTT and MTT, across most parameters describing beta-cell function, the largest reduction in beta-cell response occurred between IGM and DM, while the largest reduction in insulin sensitivity occurred between NGM and IGM. We conclude that beta-cell response is stronger after a mixed meal than after an OGTT with equal carbohydrate quantity, both for classical and model-based parameters. The higher response was mostly explained by higher beta-cell sensitivity during the meal.

HbA1c is an independent predictor of non-fatal cardiovascular disease in a Caucasian population without diabetes: a 10-year follow-up of the Hoorn Study

AIMS To investigate the associations of HbA1c, fasting glucose, and postload plasma glucose with 10-year fatal and non-fatal cardiovascular disease (CVD) and all-cause mortality in Caucasian individuals between 50 and 75 years of age without diabetes. METHOD AND RESULTS The 10-year risk of all-cause mortality and CVD in relation to HbA1cand glucose levels was assessed with Cox survival analysis in 1674 non-diabetic individuals of a population-based cohort (Hoorn Study). Analyses were stratified according to sex and adjustments were made for age and known CVD risk factors. After full adjustment, HbA1c levels ≥ 6.0% were significantly associated with an increased risk of non-fatal CVD compared with the lowest category of HbA1c (≤ 5.1%) in women [hazards ratio (HR) 2.27 (1.24-4.14)]. In addition, HbA1c as a continuous variable was significantly related to non-fatal CVD in both men [HR 1.40 (1.01-1.95)] and women [HR 2.41 (1.51-3.83)]. The relationships of HbA1 c with fatal CVD and all-cause mortality were explained by traditional CVD risk factors in both the sexes, along with the associations between fasting or postload plasma glucose and any of the outcome measures. CONCLUSION In Caucasian men and especially in women between 50 and 75 years of age who are without diabetes, high HbA1c levels are associated with increased risk of future non-fatal CVD, independent of other CVD risk factors.

Preserved GLP-1 and exaggerated GIP secretion in type 2 diabetes and relationships with triglycerides and ALT

OBJECTIVE To i) compare incretin responses to oral glucose and mixed meal of diabetic patients with the normoglycaemic population and ii) to investigate whether incretin responses are associated with hypertriglyceridaemia and alanine aminotransferase (ALT) as liver fat marker. DESIGN A population-based study. METHODS A total of 163 persons with normal glucose metabolism (NGM), 20 with intermediate hyperglycaemia and 20 with type 2 diabetes aged 40-65 years participated. Participants received a mixed meal and oral glucose load on separate occasions. Glucagon-like peptide 1 (GLP-1), glucose-dependent insulinotropic polypeptide (GIP) and glucagon profiles were analysed as total area under the curve (tAUC) and incremental area under the curve. RESULTS In diabetic patients compared with persons with NGM, we found increased GLP-1 secretion (tAUC per hour) following oral glucose (23.2 pmol/l (95% CI 17.7-28.7) vs 18.0 (95% CI 16.9-19.1), P<0.05) but not after the mixed meal. GIP secretion among diabetic patients was increased on both occasions (82.9 pmol/l (55.9-109.8) vs 47.1 (43.8-50.4) for oral glucose and 130.6 (92.5-168.7) vs 83.2 (77.5-88.9) for mixed meal, both P<0.05). After oral glucose, GLP-1 (tAUC per hour) was inversely related to fasting triglycerides. GIP (tAUC per hour) was positively related to fasting and postprandial triglycerides. Higher fasting GIP levels were related to higher fasting and postprandial triglyceride levels and ALT. CONCLUSION This study confirms that in type 2 diabetes, GLP-1 secretion is generally preserved and that GIP secretion is exaggerated. The mechanism underlying the divergent associations of GLP-1 and GIP metabolism with fat metabolism and liver fat accumulation warrants further study.

Hepatic fat is not associated with β-cell function or postprandial free fatty acid response

We evaluated the association of hepatic fat with beta-cell function estimated from the oral glucose tolerance test. In addition, we tested the hypothesis that postprandial free fatty acid (FFA) suppression after a meal tolerance test (MTT) is linked to hepatic fat. Individuals with normal glucose metabolism (NGM; n = 10 with low and n = 10 with high insulin secretion, matched for insulin sensitivity and sex), impaired glucose metabolism (IGM; n = 14), and type 2 diabetes mellitus (DM; n = 14) underwent a 75-g oral glucose tolerance test and MTT. beta-Cell function estimates were calculated from C-peptide using a mathematical model. Liver fat was quantified by proton magnetic resonance ((1)H-MR) spectroscopy. Area under the curve (AUC) of triglycerides (TG) and FFA responses during MTT represented postprandial lipid responses. Linear regression models were adjusted for age, sex, and additionally for insulin sensitivity for IGM/DM subjects. Liver fat content was equal for the NGM groups with low and high insulin secretion: 4.5% (2.6-6.0) (median, interquartile range) and 4.9% (2.3-7.8), respectively; liver fat percentages of IGM and diabetic subjects were significantly higher: 11.2 (6.7-21.1) and 10.0 (7.8-24.5). Liver fat showed a fairly strong, significant negative association with insulin sensitivity, but was not associated with beta-cell function. Significant associations of liver fat with fasting TG and AUC(TG) were shown in the total study population and in IGM/DM subjects separately. No relationship existed between fasting FFA or AUC(FFA) and liver fat. We conclude that fat accumulation in the liver is tightly linked to insulin sensitivity but not to beta-cell function. Furthermore, liver fat is associated with circulating TG levels, but not with FFA concentrations.

Postprandial metabolic responses to mixed versus liquid meal tests in healthy men and men with type 2 diabetes.

AIMS Compare metabolic responses after mixed versus liquid meals of similar caloric/nutritional content in healthy and type 2 diabetes (T2D) subjects. METHODS Ten healthy men and 10 men with T2D received mixed and liquid meals after an overnight fast. Classical (insulinogenic index; insulin/glucose areas under curves, AUC(insulin)/AUC(glucose)) and model-based (beta-cell glucose sensitivity; rate sensitivity; potentiation factor ratio, PFR) beta-cell function estimates were calculated. Between-meal differences in glucose, insulin, C-peptide, triglyceride (TG), beta-cell function and oral glucose insulin sensitivity (OGIS) and between-meal correlations for beta-cell function and OGIS were evaluated. RESULTS Among healthy subjects, beta-cell function and OGIS were similar between meals. C-peptide (p=0.03), insulin (p=0.002), AUC(insulin)/AUC(glucose) (p=0.004) and insulin secretion (p=0.04) were higher after the liquid meal. Among T2D subjects, glucose, insulin, C-peptide, beta-cell function, and OGIS were similar. PFR was higher (p=0.004) and TG increased more slowly (p=0.002) after the liquid meal. OGIS and beta-cell function were correlated during both meals in both groups (r=0.66-0.98), except incremental AUC(insulin)/AUC(glucose), rate sensitivity, and, in healthy subjects, PFR. CONCLUSIONS Metabolic responses after mixed or liquid meals of similar content were highly correlated in T2D and healthy subjects. In T2D, the liquid meal produced beta-cell function estimates generally similar to the mixed meal.