Mean plasma glucose levels in patients with Huntington disease (HD) and control participants during the oral glucose tolerance test (OGTT). Error bars represent SEM. There were no significant differences between the groups at any point during the test. To convert glucose to millimoles per liter, multiply by 0.0555.
Evaluation of insulin sensitivity in patients with Huntington disease (HD) and control participants using 2 different methods: the homeostasis model assessment–insulin resistance (HOMA-IR) index (A) and the insulin sensitivity index (Si), determined during minimal model analysis (B). Values are given as means. Error bars represent SEM. The HOMA-IR index, describing insulin resistance as a reciprocal value of insulin sensitivity, was significantly higher, whereas the Si was significantly lower, in patients with HD.
Analysis of early-phase insulin secretion in patients with Huntington disease (HD) and control participants using 2 different methods: the acute insulin response (AIR) during the intravenous glucose tolerance test (A) and the insulinogenic index (Δ plasma insulin [PI] 30/Δ plasma glucose [PG] 30) during the oral glucose tolerance test (B). Values are given as means. Error bars represent SEM. The AIR and insulinogenic index values were significantly lower in patients with HD.
Lalić NM, Marić J, Svetel M, Jotić A, Stefanova E, Lalić K, Dragašević N, Miličić T, Lukić L, Kostić VS. Glucose Homeostasis in Huntington DiseaseAbnormalities in Insulin Sensitivity and Early-Phase Insulin Secretion. Arch Neurol. 2008;65(4):476-480. doi:10.1001/archneur.65.4.476
Patients with Huntington disease (HD) develop diabetes mellitus more often than do matched healthy controls. Recent studies in neurodegenerative diseases suggested that insulin resistance constitutes a metabolic stressor that interacts with a preexisting neurobiological template to induce a given disorder.
To investigate possible changes in insulin sensitivity and secretion, major determinants of glucose homeostasis, in a group of consecutive normoglycemic patients with HD.
Twenty-nine untreated, nondiabetic patients with HD and 22 control participants matched by age, sex, and socioeconomic background.
Main Outcome Measures
Glucose tolerance, assessed by means of the glucose curve during oral glucose challenge; insulin sensitivity, assessed using homeostasis model assessment and minimal model analysis based on frequent sampling of plasma glucose and plasma insulin during the intravenous glucose tolerance test; and insulin secretion, determined by means of the acute insulin response and the insulinogenic index.
The evaluation of insulin sensitivity using homeostasis model assessment demonstrated higher homeostasis model assessment insulin resistance indices, and a lower sensitivity index when the minimal model approach was used, in patients with HD compared with controls (P = .03 and P = .003, respectively). In the assessment of early-phase insulin secretion, the acute insulin response and the insulinogenic index were lower in patients with HD compared with controls (P = .02). The number of CAG repeats correlated significantly only with acute insulin response (P = .003).
Besides impairment in insulin secretion capacity, a simultaneous decrease in insulin sensitivity, with an increase in the insulin resistance level, was found in normoglycemic patients with HD compared with controls. These data imply that progression of the insulin secretion defect in HD may lead to a failure to compensate for insulin resistance.
Huntington disease (HD) is an autosomal dominant progressive neurodegenerative disorder characterized by involuntary movements, especially chorea; frank dementia; behavioral and psychiatric disturbances; and functional disability. It is caused by an unstable expanded CAG trinucleotide repeat in exon 1 on the huntingtin gene (HD [OMIM 143100) on the tip of the short arm of chromosome 4.1
Patients with HD develop diabetes mellitus (DM) approximately 7 times more often than matched control participants.2 Podolsky and Leopold3 suggested that decreased insulin secretion might be a possible explanation. Recent studies4 in neurodegenerative diseases suggested that insulin resistance constitutes a metabolic stressor that interacts with a preexisting neurobiologic template to induce a given disorder. Studies5 with a transgenic HD mouse model that expresses a portion of the human HD gene have shown that these animals develop impaired glucose tolerance and eventually become diabetic, associated with weight loss.
The aim of this study was to investigate possible changes in insulin sensitivity (ie, the ability of insulin to exert its biological actions in an individual, whereas insulin resistance was defined as a decrease in insulin sensitivity and signified a defect in the biological action of insulin) and secretion, major determinants of glucose homeostasis, in a group of consecutive normoglycemic patients with HD. The analysis of insulin sensitivity comprised evaluation of the hepatic and peripheral components of insulin sensitivity, and the analysis of insulin secretion was focused on its early phase, the potential failure of which was expected to be the first step in glucose intolerance development in normoglycemic patients.6,7
After providing informed consent, 29 consecutive untreated patients with clinically and genetically verified HD were included in the study from families registered at the Department of Movement Disorders, Institute of Neurology, Clinical Center of Serbia (Table). The study was approved by the ethics committee of the University Clinical Center of Serbia. The inclusion criterion was adult-onset HD confirmed by DNA analysis, as previously described.8 Patients with intestinal absorption problems or previously diagnosed DM or patients who were confined to a bed or wheelchair were excluded from the study. The control group (n = 22) included individuals matched by age (±2 years), sex, and socioeconomic background who were recruited from the hospital staff and spouses of the department's patients.
Only nondiabetic patients and controls were included, which was verified by using a 2-hour 75-g oral glucose tolerance test (OGTT) (Figure 1) and the World Health Organization criteria for the diagnosis of DM during recruitment (DM was defined as a 2-hour plasma glucose [PG] level ≥200 mg/dL [to convert to millimoles per liter, multiply by 0.0555]).9 In patients with HD and controls, body mass index was calculated as weight in kilograms divided by height in meters squared.10
Metabolic investigations (performed at the Center for Metabolic Disorders in Diabetes of the Institute of Endocrinology in the same manner as performed in DM research in non-HD subjects) included (1) glucose tolerance assessment based on the blood glucose curve during OGTT; (2) insulin sensitivity assessment using 2 different methods: homeostasis model assessment (HOMA) based on basal PG and plasma insulin (PI) levels,6,11 primarily evaluating hepatic insulin sensitivity,11 and the minimal model analysis based on frequent sampling of PG and PI during the intravenous GTT (IVGTT),12 determining predominantly peripheral insulin sensitivity6,12; and (3) insulin secretion evaluation using 2 different methods for estimating early-phase insulin secretion: the acute insulin response (AIR), determined during the IVGTT, evaluating the response to intravenous glucose challenge,13 and the insulinogenic index, determined during the OGTT, evaluating the response to oral glucose challenge.14 In all the participants, the OGTT preceded the IVGTT by 4 to 7 days.
The OGTT was used to evaluate the glucose tolerance status and to determine the insulinogenic index, as a measure of insulin secretion, from the values of PG and PI obtained during the test after the glucose load. The OGTT was exerted after a 12-hour overnight fast, with a 75-g oral glucose load in the form of 50% glucose administered in 3 minutes. Blood samples for the determination of PG and PI levels were collected immediately before and 30, 60, 90, and 120 minutes after the glucose load. The insulinogenic index was determined during the OGTT as the ratio of the increment of PI to that of PG determined 30 minutes after a glucose load, that is, ΔPI (0-30 minutes)/ΔPG (0-30 minutes).14
The IVGTT was performed after a 12-hour overnight fast by infusing 0.3 g/kg of 50% glucose in 30 seconds. Blood samples for the testing of PG and PI were drawn immediately before and at multiple intervals after the intravenous glucose load. The PG and PI values obtained from the IVGTT were used to evaluate insulin sensitivity by means of minimal model analysis and to analyze early-phase insulin secretion by determining AIR (described in the third paragraph following).12,13
Minimal model analysis of insulin sensitivity was performed from the frequently sampled PG and PI levels during the IVGTT.12 Briefly, after the injection of 0.3 g/kg body weight of glucose, the blood samples were collected immediately before and 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 12, 14, 16, 20, 23, 24, 25, 27, 30, 40, 50, 60, 70, 80, 90, 100, 120, 160, and 180 minutes after intravenous glucose stimulation. Insulin was injected as a continuous infusion, 4 mU/kg/min, between minutes 20 and 25 to optimize the conditions for calculations of the insulin sensitivity index from the declining curve of PG and PI.12 The sensitivity index was calculated from the results of PG and PI levels by means of computerized minimal model analysis using the MINMOD program.
The HOMA analysis of insulin sensitivity was implemented using fasting PG and PI samples during the IVGTT. The HOMA insulin resistance (HOMA-IR) index, describing insulin resistance as a reciprocal value of insulin sensitivity, was calculated from the following equation: (PG × PI)/22.5.11
The AIR was determined during the IVGTT as the average increment of PI higher than basal values for samples obtained in the first 10 minutes of the IVGTT.13 The PG level was determined by means of the glucose oxidase method using a glucose analyzer (Beckman Instruments, Fullerton, California). The PI level was tested by radioimmunoassay using double antibody kits (INEP, Zemun, Serbia).
The results were evaluated using a statistical software program (SPSS 10.0; SPSS Inc, Chicago, Illinois). Before statistical analysis, the normal distribution and homogeneity of the variances were evaluated using the Kolmogorov-Smirnov test. Comparisons of the mean values between groups were performed using the unpaired t test for continuous variables or the Mann-Whitney test for the insulinogenic index as a variable found to be without normal distribution. Spearman correlations were calculated to determine the relationship between the variables of insulin sensitivity and insulin secretion capacity and genetic features of HD. A P < .05 was considered statistically significant.
The anthropometric and metabolic characteristics of patients with HD and controls are given in the Table. Only insulin levels were significantly higher in patients with HD compared with controls (P = .005). The PG levels during the OGTT did not differ between patients with HD and controls at any point after the oral glucose load (Figure 1).
Evaluation of insulin sensitivity using HOMA demonstrated that the mean (SEM) HOMA-IR index, describing insulin resistance as a reciprocal value of insulin sensitivity, was significantly higher in patients with HD (2.68 [0.33]) compared with controls (1.71 [0.18], P = .03) (Figure 2A). Use of the minimal model approach revealed that the mean (SEM) sensitivity index was significantly lower in patients with HD compared with controls (3.49 [0.30] vs 6.76 [0.90] min−1/mU/L × 104, P = .003) (Figure 2B).
The mean (SEM) AIR values were significantly lower in patients with HD compared with controls (14.69 [1.60] vs 25.71 [2.20] mU/L, P = .02) (Figure 3A). Determination of the insulinogenic index revealed that its mean (SEM) values were significantly lower in patients with HD (7.47 [1.90]) compared with controls (20.9 [4.50], P = .02) (Figure 3B). The mean (SEM) number of CAG repeats in patients with HD was 45.6 (7.3) (range, 40-73). The number of CAG repeats correlated only with the AIR (r = 0.17, P = .003), and no correlation was observed for the HOMA-IR index (r = −0.32, P = .10), the sensitivity index (r = 0.29, P = .54), or the insulinogenic index (r = −0.22, P = .27).
The main finding of this study was the impairment in insulin sensitivity and insulin secretion in normoglycemic patients with HD. Insulin sensitivity was analyzed using 2 different methodological approaches, HOMA and minimal model analysis, widely used in the evaluation of insulin sensitivity in healthy subjects15 and in patients with DM16 and in other diseases presumably caused by impaired insulin sensitivity.17,18 In these studies, HOMA was found to evaluate primarily hepatic insulin sensitivity,6 whereas minimal model analysis detected insulin sensitivity on the level of peripheral tissues.19 Thus, we found a decrease in insulin sensitivity in patients with HD on the level of hepatic (Figure 2A) and peripheral (Figure 2B) tissues.
The present data of the simultaneous presence of insulin resistance and decreased early-phase β-cell secretion were similar to the changes observed in the state of increased risk for type 2 DM.20 In the prediabetic state and in type 2 DM, the pivotal role in further glucose metabolism impairments was ascribed to the deterioration of insulin secretion capacity.20 However, the insulin resistance found in our patients was associated with lower-than-normal body mass index and showed no association with lipid abnormalities (Table), in contrast to patients with pre-DM and the metabolic syndrome who exhibit obesity, low high-density lipoprotein cholesterol levels, and high triglyceride levels.6
Analysis of early-phase insulin secretion demonstrated the decrease in insulin secretion capacity after intravenous glucose challenge (AIR) (Figure 3A) and after a more potent oral glucose challenge that involved the cephalic phase of hormone release (Figure 3B).14,21 Only a decrease in AIR weakly correlated with the number of CAG repeats, suggesting that at least the early phase of insulin response to intravenous challenge may be affected by the polymorphism of the mutated HD gene. However, lack of a large CAG expansion range in the present study could be partly responsible for the observed general lack of correlations (except for the AIR).
In the R6/2 transgenic mouse model of HD, which expresses exon 1 of the human HD gene containing 150 CAG repeats, mice developed glycosuria and glucose intolerance from age 9 weeks, and by 14 weeks of age more than 70% of them had developed DM.21 The histopathologic hallmark of HD, intranuclear inclusions, is also present in pancreatic cells.22,23 In R6/2 mice, the onset of DM followed the formation of ubiquitin-positive huntingtin inclusions in pancreatic cells,21 suggesting that inclusions might be involved in the development of hyperglycemia and type 1 DM. Indeed, plasmid vaccination against mutant huntingtin dramatically improved the diabetic phenotype.24 The accumulation of such insoluble nuclear protein aggregates in islets was temporally associated with selective impairment of expression of transcriptional regulatory proteins essential for glucose-responsive insulin gene expression.23 More recently, Björkqvist et al25 identified 2 separate pathologic processes that are responsible for DM in R6/2 mice: deficient β-cell mass and a loss of insulin-containing secretory granules, abrogating stimulated hormone secretion. In the same animal model, Hunt and Morton21 attempted pharmacologic treatment of DM and found that mice responded acutely to glyburide (which induces exocytosis of insulin) but not to rosiglitazone maleate (which induces sensitization to insulin). Therefore, they suggested that DM in R6/2 mice was caused by an impairment in insulin release rather than by an impairment in insulin sensitivity.
Patients with HD have an increased prevalence of DM and have pathologic GTT findings.3 Podolsky and Leopold3 performed OGTTs and intravenous arginine tolerance tests in 14 patients with HD and found that 50% of them had abnormal glucose tolerance characterized by hyperglycemia and hyperinsulinemia. The etiology of such abnormalities is unknown.
Patients with HD are significantly underweight on a population basis, despite the fact that they tend to have adequate caloric intake.26 It is unlikely that energy expenditure caused by involuntary movements explains this feature because the patients were significantly underweight even at the time of their first neurologic examination, when they had only minimal chorea,27 suggesting that the low body mass index in HD could be related to a general metabolic derangement.28,29 Similarly, most animal models of HD are associated with major loss of weight.30 Recently, Duan et al31 found that paroxetine, a serotonin reuptake inhibitor, improves survival in a transgenic mouse model of HD while inhibiting weight loss. This drug increases concentrations in brain-derived growth factor, which in turn seems to improve otherwise impaired insulin production and responsiveness in these mice. The same effect on brain-derived growth factor was also found with food restriction, which paradoxically slowed disease progression and weight loss in HD transgenic mice.30
Yamamoto et al32 demonstrated that activation of insulin receptor substrate 2, a scaffolding protein that mediates the signaling cascades of growth factors, such as insulin, leads to macroautophagy-mediated clearance of the accumulated polyglutamine proteins, suggesting that an impairment in insulin regulatory mechanisms also affects the insulin signaling pathways important for the pathogenesis of HD. Crocker et al33 conducted DNA microarray analysis of striatal gene expression in symptomatic transgenic R6/2 mice, and their bioinformatic analysis pointed to insulin and neuroinflammatory mediators as potential key regulators of a variety of differentially expressed genes in the transgenic mice and, consequently, in HD.
Neurons share similarities with insulin-producing pancreatic islet cells, possibly owing to the islet's evolution from an ancestral insulin-producing neuron.34 Therefore, the pathogenetic mechanisms in the pancreatic islets may be relevant for the pathogenesis of neurodegeneration in HD.
Data from transgenic HD animals suggested that DM occurs in mice as a result of impairment of the release of insulin, whereas insulin resistance is not a primary factor in the DM occurring in these animals.21,25 Opposite of experimental findings, in our normoglycemic patients with HD, we found, besides impairment in insulin secretion capacity, a decrease in insulin sensitivity compared with controls. These data imply that progression of the insulin secretion defect may lead to a failure to compensate for insulin resistance, but further analyses are necessary with longer follow-up to elucidate (1) the progression of observed impairments and (2) whether the improvement in insulin sensitivity may have a beneficial effect on HD phenotype.
Correspondence: Vladimir S. Kostić, MD, PhD, Institute of Neurology, Clinical Center of Serbia, Ul. Dr Subotića 6, 11000 Belgrade, Serbia (firstname.lastname@example.org).
Accepted for Publication: November 28, 2007.
Author Contributions: All authors had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: N. M. Lalić, Stefanova, and Kostić. Acquisition of data: N. M. Lalić, Marić, Svetel, Jotić, K. Lalić, Dragašević, Miličić, and Kostić. Analysis and interpretation of data: N. M. Lalić, Marić, Lukić, and Kostić. Drafting of the manuscript: N. M. Lalić, Marić, Svetel, Stefanova, Dragašević, and Kostić. Critical revision of the manuscript for important intellectual content: N. M. Lalić, Jotić, K. Lalić, Miličić, Lukić, and Kostić. Statistical analysis: N. M. Lalić, Svetel, Stefanova, Lukić, and Kostić. Obtained funding: Kostić. Administrative, technical, and material support: Dragašević, Miličić, and Kostić. Study supervision: N. M. Lalić, Marić, Jotić, K. Lalić, and Kostić.
Financial Disclosure: None reported.
Funding/Support: This work was funded by projects 145057Đ and 145089 from the Ministry of Science, Republic of Serbia.
Additional Contributions: Richard Bergman, MD, PhD, of the University of Southern California, Los Angeles, provided the MINMOD program.