Diagnostic strategy for growth hormone deficiency: relevance of IGF-1 determination as a screening test.

BACKGROUND: Adult growth hormone (GH) deficiency must be diagnosed before prescribing therapeutic recombinant human GH. We studied the clinical relevance of a diagnostic strategy for growth hormone deficiency (GHD) using IGF-1 determination as a first step. METHODS: In 2000 and 2001, we tested 142 adult patients with hypothalamo-pituitary disorders for somatotropic function using Insulin Tolerance Test (ITT), the reference test for the diagnosis of GHD, with concomitant Insulin-like growth factor-1 (IGF-1) determination, a marker of somatotropic function. Patients were classified as GHD (peak GH concentration<3 ng/ml with the ITT) or normal. SETTING: Monocenter prospective study in a tertiary referral center. RESULTS: GHD was diagnosed in 61 subjects. Using a ROC curve, a threshold IGF-1 concentration of 175 ng/ml yielded a negative predictive value of 89+/-5%. A diagnostic strategy with IGF-1 determination as the first step followed by ITT for patients with an IGF-1 concentration below 175 ng/ml missed five of the 61 GHD patients, avoided 46/142 ITT and reduced the cost of diagnosis by 15%. CONCLUSION: We propose the use of a strategy consisting of IGF-1 determination followed, if below 175 ng/ml by confirmatory ITT to diagnose GHD in adults.

Maternal history of type 2 diabetes is associated with diabetic nephropathy in type 1 diabetic patients.

AIMS: Insulin resistance is a key feature of type 2 diabetes. It is also involved in the development and progression of microvascular complications. We analysed the relationship between parental history of diabetes, insulin resistance and diabetic nephropathy (DN) and assessed the specific maternal and paternal influences of history of type 2 diabetes on DN in type 1 diabetic offspring. METHODS: We recorded information regarding family history of type 2 diabetes and of cardiovascular disease in 160 consecutive, unrelated type 1 diabetic patients. Insulin resistance was assessed using a validated estimation of the glucose disposal rate (eGDR). RESULTS: Type 1 diabetic patients with a maternal history of type 2 diabetes were more likely to be insulin-resistant (P=0.043) and to have renal complications (P=0.0041) than those from the reference group (without parental history of diabetes), while patients with a paternal history were not different from those from the reference group, regarding eGDR and DN. Time to development of abnormal albuminuria was significantly affected by maternal history of type 2 diabetes (log-rank=12.66; P=0.0004) and by familial history of premature cardiovascular disease (log-rank=5.48; P=0.0234). In multivariate analysis, a maternal history of type 2 diabetes was independently associated with nephropathy after adjustment for sex, diabetes duration and familial history of premature cardiovascular disease. CONCLUSION: Maternal history of type 2 diabetes is independently associated with DN in type 1 diabetic patients. This might suggest the transmission of a maternal trait related to microvascular complications, raising the hypothesis of imprinted genes predisposing to diabetic renal disease.

Effect of dietary calcium intake on weight gain in type 2 diabetic patients following initiation of insulin therapy.

OBJECTIVES: This pilot study analyses weight gain in type 2 diabetic patients at initiation of insulin therapy, according to daily calcium intake. METHODS: Type 2 diabetic patients consecutively admitted for initiation of insulin therapy were studied between January and March 2004 in a monocenter study. Dietary intake was assessed by a 7-day food history before insulin treatment (initial visit) and 4 to 6 months later (final visit). RESULTS: Thirty-one patients were studied (18 males and 13 females; mean age 62+/-9 years, with diabetes duration 14+/-10 years). Weight significantly increased between initial and final visits (81.9+/-16.2 vs. 84.8+/-17.8 kg; P=0.0272). Median weight gain was 2.4 kg (IQR: -1.15 to +5.27 kg). Waist circumference increased by 2 cm (IQR: 0 to +4 cm). There was no difference between weight change and tertile of calcium intake adjusted on energy intake. We did not find any correlation between weight change and total calcium intake (Rho=0.186; P=0.3165) or dairy calcium intake (Rho=0191; P=0.3040). Similarly, we did not find any correlation between waist circumference change and total calcium intake (Rho=0.324; P=0.1205) or dairy calcium intake (Rho=0.285; P=0.0755). CONCLUSION: We found no relation between total or dairy calcium intake and weight change during initiation of insulin therapy in type 2 diabetic patients. Dietary calcium intake does probably not play a major role on insulin-induced body weight gain.

Determination of HbA1c concentrations in patients with acute myocardial infarction: comparison of the DCA 2000 device with the HPLC method.

BACKGROUND: Recent studies suggest that HbA1c is an important predictor of the glycometabolic state of patients admitted for acute myocardial infarction (AMI). OBJECTIVE: We aimed at comparing the results of HbA1c concentrations obtained by 2 different methods in patients with AMI. RESEARCH DESIGN AND METHODS: In a first study, HbA1c was measured in all patients consecutively hospitalized for AMI, during a 6 month period using the HPLC method and the DCA 2000 device in the biochemistry laboratory. In a second study, HbA1c measured by the DCA 2000 device in the intensive care unit was compared with HbA1c determined by HPLC in the biochemistry laboratory in a similar sample of patients. In patients without personal history of diabetes, those patients with HbA1c > 6.5% (HPLC method), were classified as possible diabetes. RESULTS: A total of 146 patients were included (119 males, 27 females; mean age: 63 +/- 15 years). Twenty-seven of the patients had a personal history of diabetes. HbA1c determined by 2 techniques were highly correlated (r = 0.939, P < 0.0001). The mean of the differences (Bland and Altman analysis) was 0.4 +/- 0.3%. Compared with the HPLC method, the sensitivity of DCA 2000 device for the detection of possible diabetes was 81.8 +/- 11.6 and the specificity was 99.1 +/- 0.9%. The diagnostic accuracy of DCA method was 97.5 +/- 1.4%. In the second study, the HbA1c concentrations of 21 additional subjects, determined in an intensive care unit, were not different from the first 21 patients of the first study. CONCLUSIONS: HbA1c can be effectively determined using the DCA 2000 device. This method is reliable and easy to be implemented in an intensive care unit.

Prognostic value of admission plasma glucose and HbA in acute myocardial infarction.

OBJECTIVE: Stress hyperglycaemia increases the risk of mortality after acute myocardial infarction in diabetic and in non-diabetic patients. We aimed to determine the contribution of admission plasma glucose and HbA(1c) on post-acute myocardial infarction prognosis. PATIENTS AND METHODS: Admission plasma glucose and HbA(1c) were simultaneously measured in all patients consecutively hospitalized for acute myocardial infarction. Patient survival was measured on 5 and 28 days after admission. Patients were defined as having 'previously diagnosed diabetes' (personal history of diabetes defined using ADA 1997 criteria), 'no diabetes', those without previously diagnosed diabetes and HbA(1c) below 6.5%, or 'possible diabetes', i.e. those without previously diagnosed diabetes and HbA(1c) above 6.5%. RESULTS: Of the 146 patients included, four had died by day 5 and 14 by day 28. Admission plasma glucose was higher in patients who had died by day 28 (11.7 +/- 5.8 vs. 8.0 +/- 3.3 mmol/l, P = 0.002), whereas HbA(1c) was not (6.4 +/- 1.9 vs. 6.1 +/- 0.8%, NS). Admission plasma glucose was significantly higher in those who had died by day 28 after adjustment on HbA(1c). A multivariate analysis, including sex, age and heart failure prior to acute myocardial infarction, showed that admission plasma glucose concentration was an independent predictor of survival after acute myocardial infarction. Twenty-seven of the patients had previously diagnosed diabetes and 119 had no history of diabetes. Eleven were found to have possible diabetes. Admission plasma glucose was significantly higher in previously diagnosed diabetes (11.1 +/- 5.6) than in the other groups: 7.7 +/- 2.9 in non-diabetes, 8.2 +/- 2.1 in possible diabetes (P < 0.0001). The relationship between HbA(1c)-adjusted admission plasma glucose and mortality after acute myocardial infarction was also found in the non-diabetes group. CONCLUSIONS: Admission plasma glucose, even after adjustment on HbA(1c), is a prognostic factor associated with mortality after acute myocardial infarction. Acute rather than the chronic pre-existing glycometabolic state accounts for the prognosis after acute myocardial infarction.

Insulin effects on acetate metabolism.

Acetate metabolism was studied in patients with insulin resistance. To evaluate the interaction between glucose and acetate metabolism, we measured acetate and glucose turnover with a hyperinsulinemic euglycemic clamp (hot clamp) in obese and diabetic patients with insulin resistance (n = 8) and in a control group with normal insulin sensitivity (n = 6). At baseline, acetate turnover and plasma concentrations were similar between the two groups (group means: 4.3 +/- 0.4 micromol x kg-1 x min-1 and 128.2 +/- 11.1 micromol/l). Acetate concentrations decreased in both groups with hyperinsulinemia but were significantly lower in the insulin-resistant group (20% vs. 12%, P < 0.05). After the hot clamp treatment, acetate turnover increased for the two groups and was higher in the group with normal insulin sensitivity: 8.1 +/- 0.7 vs. 5.5 +/- 0.5 micromol x kg-1 x min-1 (P < 0.001). No change related to insulin action was observed in either group in the percentage of acetate oxidation. This was approximately 70% of overall utilization at baseline and during the clamp. No correlation between glucose and acetate utilization was observed. Our results support the hypothesis that, like glucose metabolism, acetate metabolism is sensitive to insulin.