Clinical and Investigative Medicine

 

Factors affecting glucose tolerance in hereditary hemochromatosis

Irene M. Hramiak, MD
Diane T. Finegood, PhD
Paul C. Adams, MD

Clin Invest Med 1997;20(2):110-8.

[résumé]

Drs. Hramiak and Adams are from the Department of Medicine, University of Western Ontario, London, Ont., and Dr. Finegood is from the Departments of Medicine and Physiology, University of Alberta, Edmonton, Alta.

(Original manuscript submitted May 14, 1996; received in revised form Jan. 29, 1997; accepted Jan. 29, 1997)

Reprint requests to: Dr. Irene M. Hramiak, London Health Sciences Centre, University Campus, Room 5 OP 18, 339 Windermere Rd., London ON N6A 5A5; fax 519 663-3676; irene.hramiak@lhsc.on.ca

Contents
Abstract

Objective: To determine insulin action and insulin secretory function in patients with hemochromatosis, and to find evidence for or against hypothesized pathogenetic mechanisms for the abnormal glucose metabolism associated with hemochromatosis. These mechanisms include decreased ß-cell secretion of insulin due to iron overload, insulin resistance and genetic factors.

Design: Prospective in vivo study.

Participants: Seventeen subjects with hemochromatosis, of whom 4 had cirrhosis but not diabetes mellitus, 6 had diabetes mellitus and 7 had neither; 10 controls.

Interventions: Insulin sensitivity and insulin secretion were determined during an intravenous glucose tolerance test. Insulin secretion was measured as the acute insulin response to glucose (AIRg). Insulin sensitivity (SI) was quantified with the minimal-model method. Of the patients with hemochromatosis, 14 agreed to undergo a second metabolic study after treatment with venous phlebotomy.

Results: All subjects with hereditary hemochromatosis had impaired glucose tolerance as measured by Kg (rate of glucose disappearance). Subjects who were free of both diabetes mellitus and liver cirrhosis had a normal SI and a decreased AIRg. In these subjects, phlebotomy treatment normalized serum ferritin levels, increased AIRg by 35% and normalized glucose tolerance (Kg). Subjects with hemochromatosis and cirrhosis had a reduced SI and maintained a normal insulin secretion. Phlebotomy treatment did not change these parameters. Subjects with hemochromatosis and diabetes mellitus had both a reduced SI and AIRg, and these parameters were unaffected by phlebotomy treatment.

Conclusions: These results suggest that iron overload can impair insulin secretion and glucose tolerance early in hereditary hemochromatosis, before cirrhosis occurs. Phlebotomy treatment can reverse these defects. Impaired glucose tolerance resulting from insulin resistance in subjects with cirrhosis or diabetes mellitus is not affected by phlebotomy treatment.

Résumé

Conception : Déterminer l'action de l'insuline et la fonction de sécrétion d'insuline chez les patients atteints d'hémochromatose afin de trouver des données probantes démontrant l'existence ou l'inexistence de mécanismes pathogénétiques hypothétiques pour le métabolisme anormal du glucose lié à l'hémochromatose. Ces mécanismes comprennent une diminution de la sécrétion d'insuline par les cellules ß causée par une surcharge ferrique, une résistance à l'insuline et des facteurs génétiques.

Conception : Étude prospective in vivo.

Participants : Dix-sept personnes atteintes d'hémochromatose, dont 4 étaient atteintes de cirrhose mais non de diabète sucré, 6 avaient le diabète sucré et 7 n'avaient ni l'un ni l'autre de ces problèmes; 10 témoins.

Interventions : On a déterminé l'insulinosensibilité et la sécrétion d'insuline au cours d'une épreuve d'hyperglycémie provoquée par voie intraveineuse. On a mesuré la sécrétion d'insuline sous forme de réponse insulinémique aiguë au glucose (RIAg). On a quantifié l'insuloinsensibilité au moyen de la méthode du modèle minimal. Parmi les patients atteints d'hémochromatose, 14 ont consenti à subir une deuxième analyse métabolique après traitement par phlébotomie veineuse.

Résultats : Tous les sujets atteints d'hémochromatose héréditaire avaient une déficience de la tolérance au glucose mesurée par le coefficient Kg (taux d'élimination du glucose). Les sujets qui n'avaient pas de diabète sucré ni de cirrhose du foie avaient une insuloinsensibilité normale et une RIAg réduite. Chez ces sujets, le traitement par phlébotomie a normalisé les taux de ferritine sérique, augmenté de 35 % la RIAg et normalisé la tolérance au glucose (Kg). Les sujets atteints d'hémochromatose et de cirrhose avaient une insuloinsensibilité réduite et ont maintenu une sécrétion d'insuline normale. Le traitement par phlébotomie n'a pas modifié ces paramètres. Les sujets atteints d'hémochromatose et de diabète sucré avaient à la fois une insuloinsensibilité réduite et une RIAg réduite et le traitement par phlébotomie n'a pas modifié ces paramètres.

Conclusions : Ces résultats indiquent qu'une surcharge en fer peut nuire à la sécrétion d'insuline et à la tolérance au glucose tôt dans les cas d'hémochromatose héréditaire, avant l'apparition de la cirrhose. Le traitement par phlébotomie peut inverser ces défauts. La phlébotomie n'affecte pas la déficience de la tolérance au glucose découlant d'une résistance à l'insuline chez les sujets atteints de cirrhose ou de diabète sucré.

[ Top of document ]

Introduction

Diabetes mellitus is a well-recognized feature of hereditary hemochromatosis; however, its cause has not been clearly established. Defects in both insulin secretory function and insulin action, the primary components of glucose tolerance, have been implicated. Impaired insulin secretion due to iron deposition in the ß-cells of the pancreas, decreased insulin action due to liver cirrhosis and genetic factors have been suggested as pathogenic mechanisms for the abnormal metabolism of glucose in hereditary hemochromatosis.[1,2]

Normally, insulin resistance is compensated for in part by an increase in insulin secretion, to maintain a constant level of glucose tolerance.[3] In subjects with liver cirrhosis due to other causes, such as alcoholism, insulin resistance occurs.[4] Despite some elevation of insulin levels in subjects with cirrhosis, compensation is incomplete and the resulting glucose tolerance is abnormal.[5,6] The pathogenesis of the glucose intolerance found in subjects with hereditary hemochromatosis but without cirrhosis is less clear. Plasma insulin levels increase in response to oral ingestion of glucose in these subjects.[7] One recent report suggests that impaired degradation of insulin due to hepatic iron overload may play a role.[5] However, increased insulin levels also may be compensating for a decrease in insulin action or insulin resistance. No direct assessment of insulin action in these subjects has been conducted.

Although the mechanisms by which iron overload may affect insulin secretion or insulin action are unknown, it has been suggested that treatment of the iron overload may improve glucose tolerance. In several studies, more than 40% of subjects with hereditary hemochromatosis were able to discontinue insulin therapy, or decrease the amount of insulin needed, after venous phlebotomy.[5,8,9] This suggests that a component of the glucose intolerance associated with hereditary hemochromatosis may be reversible.

In this study, we specifically determined insulin action and insulin secretory function in subjects with hereditary hemochromatosis and differing levels of liver function, including both nondiabetic and diabetic subjects, to determine which factors affect glucose tolerance and to find evidence for or against the hypothesized mechanisms causing diabetes mellitus in patients with hemochromatosis.

[ Top of document ]

Methods

We recruited 27 patients: 17 with hemochromatosis and 10 controls. The diagnosis of hemochromatosis was based on the clinical history, physical examination, serum ferritin level, transferrin saturation, liver biopsy and hepatic iron concentration. None of the subjects had a history of alcoholism or iron-loading anemia. All subjects with hemochromatosis had a liver biopsy, and liver biopsy specimens were classified as normal or cirrhotic by a single pathologist. Subjects with hemochromatosis were placed into groups according to whether they had cirrhosis, diabetes mellitus (with or without cirrhosis), or neither.

After the diagnosis was established, we obtained informed consent from each subject to carry out metabolic studies. This study was approved by the University of Western Ontario Review Board for Health Sciences Research Involving Human Subjects.

The clinical characteristics of the subjects with hemochromatosis and the normal subjects studied are shown in Table 1. The diagnosis of diabetes mellitus was based on the National Diabetes Data Group criteria.[10] Four of the 6 diabetic subjects with hereditary hemochromatosis also had cirrhosis and were receiving exogenous insulin (mean dose 0.3 U/kg per day). The remaining 2 diabetic subjects with hereditary hemochromatosis did not have cirrhosis and were managed with a diabetic diet alone. All diabetic subjects with hereditary hemochromatosis regularly conducted capillary glucose self-monitoring at home.

All subjects with hemochromatosis were studied as outpatients and had normal activity levels. All patients with cirrhosis were compensated (Childs Class A). All subjects consumed their normal diet; the diabetic subjects were on a weight-maintaining diabetic diet.

After initial metabolic studies were conducted, the subjects with hemochromatosis were treated by venous phlebotomy to remove of 500 mL of blood weekly until their serum ferritin level was approximately 50 µg/L (reference range 15 to 300 µg/L). Insulin secretion and insulin action parameters were subsequently reassessed.

All 17 patients with hereditary hemochromatosis were studied before treatment, at the time of diagnosis. Of these patients, 14 (all 4 of those with cirrhosis, 4 out of 6 of those with diabetes mellitus and 6 out of 7 of those with neither) consented to undergo the second metabolic study after treatment.

Glucose metabolism was quantified with the use of the "minimal-model" method of Bergman,[11] which provides a measure of insulin sensitivity and insulin secretion in a single study. An index of insulin sensitivity (SI) is calculated from a computer analysis of blood glucose and serum insulin levels during a frequently sampled intravenous glucose-tolerance test (FSIGT). The minimal model and glucose clamp have been shown to provide an equivalent measure of insulin sensitivity.[12] The minimal model has also been used to test subjects with diabetes mellitus and cirrhosis.[4,13]

Insulin sensitivity studies

Insulin sensitivity (SI) was determined by FSIGT.[14] All subjects and normal controls were studied after an overnight fast. The 5 subjects treated with insulin had received their last dose of regular insulin (short-acting) at supper the evening before the study. Intravenous catheters were placed in both antecubital veins. One was used for administration of solutions and the other was used for sampling. After basal sampling, a 0.3 g/kg bolus of glucose (50% dextrose) was given intravenously over 1 minute. Samples were drawn every minute until 8 minutes, thereafter every 2 minutes until 16 minutes and once again at 19 minutes. At 20 minutes, 300 mg of tolbutamide was administered intravenously over 30 seconds. This modification has been shown to improve the precision of estimates of SI by the minimal-model method.[12] Subsequent sampling occurred at 22 minutes, 23 minutes, 24 minutes, 25 minutes, 27 minutes, 30 minutes, and thereafter every 10 minutes until 100 minutes, and thereafter every 20 minutes until 180 minutes. In diabetic subjects, tolbutamide is a poor secretagogue. In these subjects we had previously validated a modification of the protocol in which we use an infusion of insulin at a rate of 4 mU × min-1 × kg-1 from 20 to 25 minutes.[13] Blood samples for determining plasma insulin levels were collected into nonheparinized tubes and sodium fluoride was added to samples for determination of plasma glucose levels. Samples for C-peptide were collected into heparinized tubes containing aprotinin. Samples were centrifuged for 15 minutes at 2500 rpm and 4°C and were stored at -20°C until assayed.

Calculations

SI and glucose effectiveness (SG) were determined by using the MINMOD computer program, as previously described.[15] In brief, the minimal model of glucose kinetics describes the relationship between the plasma insulin level and the fall in the plasma glucose level after an intravenous bolus injection of glucose. Insulin is said to enter a remote compartment from which its action occurs. The SG is glucose effectiveness, or the use of glucose at basal insulin levels. The SI index is a measure of the effect of insulin concentrations above the basal level in enhancing glucose disappearance.

The rate of glucose disappearance was calculated as the least-squares slope of the natural logarithm of glucose concentration versus time after glucose injection (from 10 to 19 minutes) and expressed as the percentage of glucose removed per minute. This parameter was called Kg in this study, and although it is calculated over a shorter time interval than that usually used for Kg (10 to 40 minutes) it has been shown to be strongly correlated with the longer Kg.[16]

As a measure of insulin secretion, we used the acute insulin response to glucose (AIRg). This was determined as the mean incremental area in plasma insulin levels above the basal level from 2 to 5 minutes after the glucose injection during the FSIGT.

Assays

Plasma C peptide was determined by radioimmunoassay (RIA).[17] The intra-assay variation was 5%, and the interassay variation was 8%. Plasma immunoreactive insulin was measured by RIA with dextran and charcoal separation using a human insulin standard.[18] A Beckman glucose analyzer II (Beckman, Fullerton Calif.) was used to measure plasma glucose levels.

Statistical analysis

Analysis of variance (ANOVA) with the least-significant-difference method was used to compare differences between groups. The effect of iron-depletion therapy on carbohydrate metabolism in the 14 subjects that were studied before and after venous phlebotomy was compared with the paired Student's t-test. Correlation coefficients were calculated for the relationship between serum ferritin level and each of AIRg, SI and SG Results were considered to be significant at a p level of less than 0.05.

[ Top of document ]

Results

The clinical characteristics of all subjects studied before phlebotomy treatment are shown in Table 1. There were no significant differences in body-mass index (BMI) among the groups. Subjects with cirrhosis were the oldest. The glycated hemoglobin (Ghb) was in the normal range (less than 6.5%) for all nondiabetic subjects. Although Ghb was higher in the diabetic subjects, the elevation was not statistically significant. The serum ferritin levels were abnormally high in all subjects with hemochromatosis and were significantly higher in the subjects with cirrhosis than in the other subjects with hereditary hemochromatosis. The liver iron concentration was elevated in all subjects with hereditary hemochromatosis but did not differ among the groups of subjects with hereditary hemochromatosis.

The normal subjects did not have a family history of diabetes mellitus or hereditary hemochromatosis. All subjects with hereditary hemochromatosis were asked about family history, and 11 had a known family history of diabetes mellitus. Only 8 subjects had a known family history of hemochromatosis. There were 6 subjects who had a family history of both diabetes mellitus and hemochromatosis. Of the 6 subjects who had diabetes mellitus, 4 had a known family history of diabetes mellitus alone and the other 2 had a family history of both diabetes mellitus and hereditary hemochromatosis.

Metabolic parameters for all subjects before treatment are shown in Table 2. The diabetic subjects had a significantly increased plasma glucose level the morning of the study. The fasting plasma insulin level and fasting C-peptide levels appeared to be elevated in the subjects with cirrhosis but this elevation was not statistically significant, owing to the large variation among subjects. The ratio of C-peptide to insulin levels in all groups in the fasting state did not differ from that found in the normal subjects. The Kg was significantly (p < 0.05) decreased in all subjects with hereditary hemochromatosis compared with normal subjects.

To assess insulin secretion, the AIRg was calculated from the FSIGT. Fig. 1 shows the average time course of plasma glucose and insulin levels during the FSIGT in subjects with hereditary hemochromatosis and normal subjects. The AIRg was a mean of 300 (standard error of the mean [SEM] 58) pmol/L × minute as compared with the normal mean of 408 (SEM 77) pmol/L × minute. Although the AIRg was smaller than normal, the difference did not achieve statistical significance. The mean incremental value of 409 (SEM 49.8) pmol/L × minute for the subjects with hereditary hemochromatosis and cirrhosis was in the normal range. The diabetic subjects had a significantly lower AIRg than all of the other subjects, with a mean value of 41.0 (SEM 5.2) pmol/L × min (p < 0.0002). This lowered secretion was present in all diabetic subjects, regardless of whether they had cirrhosis. There was no significant correlation between either serum ferritin level or hepatic iron level and AIRg.

The subjects with hereditary hemochromatosis had a normal mean SI when compared with normal subjects (51.3 [SEM 10.1] min-1 per nmol/mL v. 42.6 [SEM 7.8] min-1 per nmol/mL). The time course of change in the insulin and glucose levels during the FSIGT is shown in Fig. 1. The SI was significantly (p < 0.027) reduced in both the subjects with cirrhosis (14.3 [SEM 4.4] min-1 per nmol/mL) (Fig. 2) and those with diabetes mellitus (21.0 [SEM 10.8] min-1 per nmol/mL). The SG was normal in all subjects with hemochromatosis, as shown in Table 2. No correlation was seen between serum ferritin level, hepatic iron concentration and either SI or SG.

All subjects underwent treatment with weekly iron-depleting phlebotomy. The metabolic parameters in the 14 subjects who consented to a second metabolic study after treatment are compared in Table 3. The serum ferritin levels were significantly decreased, reaching normal levels. There was no significant change in such parameters as fasting plasma glucose, fasting plasma insulin or basal C-peptide levels in any of the subject groups. Phlebotomy did not change the insulin sensitivity in any of the groups. In the subjects with hereditary hemochromatosis only, both AIRg and Kg increased significantly after phlebotomy. In the subjects with hereditary hemochromatosis and cirrhosis, glucose tolerance and insulin secretion did not change after phlebotomy treatment. In the subjects with hereditary hemochromatosis and diabetes mellitus, all parameters of insulin secretion and action remained low after phlebotomy. No change occurred in the clinical treatment of our diabetic subjects.

[ Top of document ]

Discussion

The incidence of diabetes mellitus in hereditary hemochromatosis has been reported to be as high as 72% and is most common in patients who also have cirrhosis.[9] In this study we found that 6 of the 17 subjects with hereditary hemochromatosis also had diabetes mellitus. We determined that all of our diabetic subjects with hereditary hemochromatosis had a family history of diabetes mellitus, while only 33% of these subjects had a family history of hereditary hemochromatosis. In the subjects who did not have diabetes mellitus, 23% had a family history of diabetes mellitus and 39% had a family history of hereditary hemochromatosis. These data support the concept that a genetic predisposition to diabetes mellitus may be an important factor in determining whether subjects with hereditary hemochromatosis develop the disease. There was no correlation between either hepatic iron concentration or serum ferritin level and diabetes mellitus in our subject group. Similarly, the degree of liver damage did not predict the presence of diabetes mellitus; 4 of our patients with cirrhosis did not have diabetes mellitus, and only 2 of the patients with diabetes mellitus had cirrhosis. This confirms the results from other studies that also found a poor correlation between diabetes mellitus in subjects with hereditary hemochromatosis and the degree of iron overload or liver damage.[6,7]

The subjects with hereditary hemochromatosis who had diabetes mellitus in our study exhibited insulin sensitivity that was 49% of normal before treatment. The insulin secretion response to glucose (AIRg) was also markedly abnormal (8% of normal). These results are typical of subjects who have non-insulin-dependent diabetes mellitus (NIDDM);[19,20] in such subjects, overt diabetes develops as a result of impaired insulin secretion and decreased insulin sensitivity. As in our diabetic subjects with hereditary hemochromatosis, a complete loss of the acute insulin response to intravenously administered glucose or first-phase insulin response to glucose, at fasting blood glucose levels above 6.4 mmol/L, has been reported in these subjects.[21]

This loss of acute insulin response in patients with NIDDM has been shown to be improved by tight metabolic control, which suggests that it is an acquired defect secondary to hyperglycemia and, as such, is partially reversible. Some authors have suggested that the first-phase insulin response is normal in the relatives of subjects with NIDDM,[22] while others have suggested that patients who have had gestational diabetes (and are therefore at risk for NIDDM) as well as first-degree relatives of patients with NIDDM have impaired insulin secretion before the onset of frank diabetes mellitus.[23,24] It is unclear whether the impairment in insulin secretion in the relatives of patients with NIDDM in general is genetically determined. No specific information on hemochromatosis is available.

In previous studies, up to 40% of 115 subjects with hereditary hemochromatosis had an improvement in diabetic symptoms after phlebotomy treatment, as determined by a decrease in exogenous insulin or oral hypoglycemic dose needed.[8,9] We did not observe this improvement in our small group of diabetic subjects. Our patients with hereditary hemochromatosis and diabetes mellitus did not show an improvement in AIRg, SI or Kg after phlebotomy treatment.

In our subjects with liver cirrhosis but without diabetes mellitus, we found a significant reduction in insulin sensitivity. The subjects with hereditary hemochromatosis alone were age-matched with the control subjects, but the subjects with cirrhosis were significantly older. In normal subjects, the impairment in glucose tolerance with advancing age from 40 to 63 years is about 17%. The impairment in glucose tolerance in the group with cirrhosis was 42% compared with normal controls, and was unlikely to be caused by age alone.[25] We have previously shown a correlation between insulin sensitivity and BMI in normal subjects.[26] However, the 10% variation in the BMI of subjects in our study would not explain the variation in SI we observed.

Insulin resistance has been described in all subjects with cirrhosis, regardless of their glucose tolerance.[27] In undifferentiated cirrhosis, insulin resistance has been attributed to the liver[4] as well as to peripheral factors.[27] Our subjects maintained a normal range of insulin, C-peptide and AIRg levels. However, to maintain normal glucose tolerance, increased insulin secretion would be required. The lack of improvement in glucose tolerance, insulin secretion and insulin sensitivity after phlebotomy treatment may indicate that these changes are not related to iron overload but to the degree of liver impairment, which is unchanged by the depletion of iron.

In studies of patients with iron overload secondary to thalassemia, insulin resistance was described in patients who had degrees of iron overload (as assessed by serum ferritin level) similar to those of our patients with cirrhosis.[28,29] However, the patients with thalassemia could compensate for this insulin resistance with a marked increase in insulin secretion. Our patients with hemochromatosis and insulin resistance may not have had increased insulin secretion because of (1) the effects of liver cirrhosis, (2) older age and, concomitantly, longer duration of iron overload, or (3) a possible genetic predisposition to NIDDM.

Glucose tolerance is the result of both insulin secretion and insulin action. A decrease can occur with either factor. If such a decrease is accompanied by a compensatory increase in the other factor, no net effect in glucose tolerance results. In patients with hemochromatosis, the decrease in glucose tolerance or Kg is the result of a change in insulin secretion, since their insulin sensitivity is normal. We quantified only the first 5 minutes of the acute insulin response in our measurement of AIRg. This is an imprecise assessment of the entirety of insulin secretion. After phlebotomy treatment, a 25% increase in AIRg was seen and the Kg was normalized. It has been suggested that iron deposition on transferrin receptors is a selective process occurring in the ß-cells of the pancreas.[30] Our results suggest that, in modest iron overload, a subtle defect in insulin secretion is the only defect in glucose tolerance. However, with more severe iron overload and liver cirrhosis, insulin resistance is the predominant defect affecting metabolic control. In patients with cirrhosis and hemochromatosis, although their AIRg is normal, one would expect a compensatory increase in insulin secretion if glucose tolerance were to be maintained in the normal range. Diabetes mellitus in the patients with hemochromatosis was characterized by a degree of insulin resistance similar to that seen in the patients with cirrhosis, but a lack of the acute insulin response to glucose. Studies are needed to identify the relative importance of genetic factors versus iron overload and deposition on this loss of first-phase response.

[ Top of document ]

Acknowledgements

We would like to thank Alison Cardiff for her technical assistance with these studies. We acknowledge the secretarial support of Wanda McBeth in the preparation of the manuscript.

[ Top of document ]

References

  1. Finch CS, Finch CA. Idiopathic hemochromatosis an iron storage disease. Medicine 1955;34:381.
  2. Saddi R, Feingold J. Idiopathic hemochromatosis: autosomal recessive disease. Rev Fr Etud Clin Biol 1969;14:238.
  3. Kahn ST, Prigeon RL, McCulloch DK, Boyko EJ, Bergman RN, Schwartz MW, et al. Quantification of the relationship between insulin sensitivity and ß-cell function in human subjects. Diabetes 1993;42:1663-72.
  4. Kruszynska YT, Harry DS, Bergman RN, McIntyre N. Insulin sensitivity, insulin secretion and glucose effectiveness in diabetic and non-diabetic cirrhotic patients. Diabetologia 1993;36:121-8.
  5. Stremmel W, Niederau C, Berger M, Kley H-K, Kruskemper H-L, Strohmeyer G. Abnormalities in estrogen, androgen, and insulin metabolism in idiopathic hemochromatosis. Ann N Y Acad Sci 1988;526:209-23.
  6. Magnusson J, Tranberg KG. Impaired early insulin response to intravenous glucose in alcoholic liver cirrhosis. Scand J Gastroenterol 1987;22:301-7.
  7. Bierens de Haan B, Scherrer JR, Stauffacher W, Pometta D. Iron excess, early glucose intolerance and impaired insulin secretion in idiopathic hemochromatosis. Eur J Clin Invest 1973;3:179-87.
  8. Dymock IW, Cassar J, Pyke DA, Oakley WG, William R. Observations on the pathogenesis, complications and treatment of diabetes in 115 cases of hemochromatosis. Am J Med 1972;52:203-10.
  9. Niederau C, Fischer R, Purschel A, Stremmel W, Haussinger D, Strohmeyer G. Long-term survival in patients with hereditary hemochromatosis. Gastroenterology 1996;110:1107-19.
  10. National Diabetes Data Group. Classification and diagnosis of diabetes mellitus and other categories of glucose intolerance. Diabetes 1979;28:1039-57.
  11. Bergman RN. Toward physiological understanding of glucose tolerance. Minimal model approach. Diabetes 1989;38:1512-27.
  12. Bergman RN, Prager R, Volund A, Olefsky JM. Equivalence of the insulin sensitivity index in man derived by the minimal model method and the euglycemic glucose clamp. J Clin Invest 1987;79:790-800.
  13. Finegood DT, Hramiak IM, Dupre J. A modified protocol for estimation of insulin sensitivity with the minimal-model of glucose kinetics in patients with insulin-dependent diabetes. J Clin Endocrinol Metab 1990;70:1538-49.
  14. Bergman RN, Ider YZ, Bowden CR, Cobelli C. Quantitative estimations of insulin sensitivity. Am J Physicians 1979;236:E667-77.
  15. Pacini G, Bergman RN. MINMOD: a computer program to calculate insulin sensitivity and pancreatic responsitivity from the frequently sampled intravenous glucose tolerance test. Comput Methods Programs Biomed 1986;23:113-22.
  16. Lee A, Ader M, Bray G, Bergman R. Diurnal variation in glucose tolerance. Diabetes 1992;41:750-9.
  17. Heding L. Radioimmunological determination of human C-peptide in serum. Diabetologia 1975;2:541-8.
  18. Kuzuya H, Blix PM, Horwitz DL, Steiner DF, Rubenstein AH. Determination of free and total insulin and C-peptide in insulin-treated diabetics. Diabetes 1977;26:22-9.
  19. Reaven GM. Role of insulin resistance in human disease. Diabetes 1988;37:1595-607.
  20. DeFronzo RA. The triumvirate, B-cell, muscle, liver: a collusion responsible for NIDDM. Diabetes 1988;37:
    667-87.
  21. Brunzell JD, Robertson RP, Lerner RL. Relationships between fasting plasma glucose levels and insulin secretion during intravenous glucose tolerance tests. J Clin Endocrinol Metab 1976;42:222-9.
  22. Martin B, Warram J, Krolewski A, Bergman R, Soeldner J, Kahn R. Role of glucose and insulin resistance in development of type 2 diabetes mellitus: results of a 25-year follow-up study. Diabetologia 1992;340:925-9.
  23. Eriksson J, Kallunki A, Ekstrand A, Saloranta C, Widen E, Schalin C, Groop L. Early metabolic defects in persons at increased risk for non-insulin dependent diabetes mellitus. N Engl J Med 1989:321:337-43.
  24. Ward WK, Johnston C, Beard C, Benedetti T, Halter JB, Porte D. Insulin resistance and impaired insulin secretion in subjects with histories of gestational diabetes mellitus. Diabetes 1985;34:861-9.
  25. Defronzo RA. Glucose intolerance and aging. Diabetes Care 1981;4:493-501.
  26. Feldman RD, Hramiak IM, Finegood DT, Behme MT. Parallel regulation of the local vascular and systemic metabolic effects of insulin. J Clin Endocrinol Metab 1995;80(5):1556-9.
  27. Petrides AS, Groop L, Riely CA, Defronzo RA. Effect of physiologic hyperinsulinemia on glucose and lipid metabolism in cirrhosis. J Clin Invest 1991;88:561-70.
  28. Dmochowski K, Finegood DT, Francombe W, Tyler B, Zinman B. Factors determining glucose tolerance in patients with thalassemia major. J Clin Endocrinol Metab 1993;77:478-83.
  29. Merkel PA, Simonson DC, Amiel SA, Plewe G, Sherwin RS, Pearson HA, et al. Insulin resistance and hyperinsulinemia in patients with thalassemia major treated by hypertransfusion. N Engl J Med 1988;318:
    809-14.
  30. Rahier J, Loozen S, Goebbels RM, Abraham M. The haemochromatotic human pancreas: a quantitative immunohistochemical and ultrastructural study. Diabetologia 1987;30:5-12.
| CIM: April 1997 / MCE : avril 1997 |
CMA Webspinners / >