GLUCOSE TOLERANCE STATUS is traditionally defined

Transcription

1 X/05/$15.00/0 The Journal of Clinical Endocrinology & Metabolism 90(2): Printed in U.S.A. Copyright 2005 by The Endocrine Society doi: /jc The Normal Glucose Tolerance Continuum in Obese Youth: Evidence for Impairment in -Cell Function Independent of Insulin Resistance Catherine W. Yeckel, Sara E. Taksali, James Dziura, Ram Weiss, Tania S. Burgert, Robert S. Sherwin, William V. Tamborlane, and Sonia Caprio Department of Pediatrics (C.W.Y., S.E.T., R.W., T.S.B., W.V.T., S.C.), The General Clinical Research Center (J.D.), and the Department of Internal Medicine (R.S.S.) of Yale University School of Medicine, New Haven, Connecticut Normal glucose tolerance is expressed over a wide range of glucose concentrations. Whether there is a continuum of risk for developing type 2 diabetes mellitus even when the 2-h plasma glucose is still within this normal range is uncertain. Oral glucose tolerance tests were performed in 407 obese normal glucose tolerance youth (4 20 yr) to examine the relationship between variations in 2-h plasma glucose levels and -cell responsiveness. Individuals were grouped by 2-h plasma glucose levels as follows: 1) less than 100 mg/dl, 2) mg/dl, and 3) mg/dl. Subsequent analysis stratified each 2-h plasma level by insulin sensitivity index. Increased 2-h glucose level was associated with a progressive GLUCOSE TOLERANCE STATUS is traditionally defined by the broad categories of normal and impaired glucose tolerance (NGT and IGT), as well as type 2 diabetes mellitus (T2DM), based on the 2-h plasma glucose concentration during an oral glucose tolerance test (OGTT). The growing population of obese youth has triggered the alarm for possible progression to the metabolic syndrome, T2DM, and their attendant complications at an early age (1). It is therefore important to identify the clinical indications of the high-risk profile as early as possible in the course of the disease. Previous work using the glucose clamp techniques revealed that, compared with normal-weight youth, obese children and adolescents exhibit a severe reduction in peripheral insulin sensitivity in response to a standard hyperinsulinemic stimulus and greater insulin secretion in response to a standard hyperglycemic stimulus (2 4). In addition, we recently validated the Matsuda and DeFronzo (5) whole-body insulin sensitivity index (WBISI) in obese youth, enabling us to show that in a large mixed-ethnic cohort, there is a strong overlap in insulin resistance between many of the NGT and IGT children and adolescents (6). The major distinguishing feature between the severely insulin-resistant IGT and NGT First Published Online November 2, 2004 Abbreviations: BMI, Body mass index; GENNID, Genetics of Non- Insulin-Dependent Diabetes; IGI, insulinogenic index; IGT, impaired glucose tolerance; NGT, normal glucose tolerance; OGTT, oral glucose tolerance test; SAM, San Antonio metabolism; WBISI, whole-body insulin sensitivity index. JCEM is published monthly by The Endocrine Society ( endo-society.org), the foremost professional society serving the endocrine community. 747 increase in glucose between 0 and 30 min (P < 0.05). The (0 30 min) insulin did not vary significantly across levels, thus resulting in a decreased insulinogenic index (P < 0.02). This pattern was observed at every level of insulin sensitivity (P < 0.02). These data translated to an unfavorable (leftward) shift in the insulin feedback system for increasing 2-h glucose level (P < 0.005). Increased 2-h plasma glucose within the range of normal glucose tolerance in obese youth is associated with a specific impairment of -cell responsiveness distinct from the deterioration of insulin sensitivity. (J Clin Endocrinol Metab 90: , 2005) youth was the failure of the -cell in the IGT group to increase the early insulin response to the glucose load. Similarly, the loss of -cell compensation has been shown in the progression from the NGT to the prediabetic state in adults (7, 8). By convention, NGT is defined as a plasma glucose level less than 140 mg/dl after a standard oral glucose load. In our multiethnic population of obese youth with NGT, 2-h glucose levels ranged between 60 and 139 mg/dl. Insulin sensitivity and early insulin responses to the glucose load also varied over a large range in these subjects (6). These observations led us to question whether increases in 2-h plasma glucose levels even in the normal range are primarily related to -cell dysfunction or insulin resistance. To examine this question, we first divided our large cohort of obese youth with NGT into three strata based on 2-h plasma glucose concentrations. This stratification allowed us to examine the impact of altered insulin responsiveness on the 2-h glucose level in obese children and adolescents in general. Within each stratum, we subdivided the subjects into moderate, low, and very low insulin sensitivity groups to control for the large magnitude changes in overall insulin response with increasing insulin resistance. Importantly, we examined our large cohort of obese youth to see whether even discrete changes in 2-h glucose within the normal range could be detected with the insulin feedback model of glucose tolerance. Subjects and Methods Study population The Yale Pathophysiology of Type 2 Diabetes in Obese Youth Study is a long-term project that is examining early alterations in glucose

2 748 J Clin Endocrinol Metab, February 2005, 90(2): Yeckel et al. -Cell Impairment in NGT Obese Youth TABLE 1. Demographic characteristics of obese youth Testing procedures: OGTT 2-h glucose level (mg/dl) N Male/female 35/60 62/111 58/81 White/Black/Hispanic 35/37/23 73/50/50 68/39/32 Age (yr) BMI (kg/m 2 ) BMI Z score h glucose Demographic parameters for age, BMI, and BMI Z score (age and gender adjusted) are presented as means SD. Mean SD for 2-h glucose are also provided. Systeme International unit for glucose is mmol/liter (conversion factor, ). metabolism in a large mixed-ethnic cohort of obese children and adolescents. The Human Investigations Committee of Yale University School of Medicine approved the study protocol. Written parental consent and child assent were obtained before the study. The volunteers were recruited from our Pediatric Weight Management Clinic. To be eligible for participation, subjects had to be obese ( 95th percentile for age and gender) and free of chronic medical conditions not related to the metabolic syndrome or diabetes. Based on their OGTT results (see methods below), 407 subjects were divided into three strata by their 2-h glucose concentrations: 1) less than 100 mg/dl ( 5.55 mmol/liter), 2) mg/dl ( mmol/ liter), and 3) mg/dl ( mmol/liter). Demographic data appear in Table 1 by category of 2-h glucose level. For the purpose of controlling for insulin sensitivity, each level of 2-h glucose was further stratified by insulin sensitivity (WBISI). Levels were based on the previously used tertile values (6) [very low, less than 1.44 (n 127); low, (n 138); moderate, more than 2.25 (n 142) (Table 2)] to allow direct comparisons of data with those previously published. (Readers are referred to this publication for consideration of major differences across insulin sensitivity tertiles.) Subjects were studied in the General Clinical Research Center of Yale University School of Medicine at 0800 h after a 10- to 12-h overnight fast. A standard (1.75 g/kg body weight, up to 75 g) OGTT was performed in all children and adolescents to establish glucose tolerance status. After the local application of a topical anesthetic cream containing 2.5% lidocaine and 2.5% prilocaine, one antecubital iv catheter was inserted for blood sampling and maintained patent by a normal saline drip. Two baseline samples were then obtained at 15 and 0 min for measurements of plasma glucose, insulin, and C-peptide. Thereafter, the flavored glucose (Orangedex; Custom Laboratories, Baltimore, MD) was given orally, and blood samples were obtained every 30 min for 180 min for the measurements of plasma glucose, insulin, and C-peptide. Biochemical analyses Plasma glucose was determined using a glucose analyzer by the glucose oxidase method (Beckman Instruments, Brea, CA). Plasma insulin was measured by the Linco RIA, which has less than 1% crossreactivity with C-peptide and proinsulin. Plasma C-peptide was assayed with a Linco RIA kit. Calculations Index of insulin sensitivity from the OGTT. The composite WBISI is based on mean values of insulin ( U/ml) and glucose (mg/dl) obtained from the OGTT and the corresponding fasting values, as originally described by Matsuda and DeFronzo (5). We have recently validated this index for use in obese children and adolescents (6). WBISI 10,000 (fasting glucose fasting insulin)(mean glucose mean insulin) TABLE 2. Demographic characteristics of obese youth stratified by insulin sensitivity and 2-h glucose level Insulin sensitivity 2-h glucose level (mg/dl) Very low (WBISI, 1.44) N Male/female 7/12 14/35 25/34 White/Black/Hispanic 8/7/4 15/19/15 29/14/16 Age (yr) BMI (kg/m 2 ) BMI Z score h glucose WBISI a Low (WBISI, ) N Male/female 12/14 27/42 21/22 White/Black/Hispanic 7/12/7 32/19/18 21/15/7 Age (yr) BMI (kg/m 2 ) BMI Z score h glucose WBISI Moderate (WBISI, 2.25) N Male/female 16/34 21/34 12/25 White/Black/Hispanic 20/18/12 26/12/17 18/10/9 Age (yr) BMI (kg/m 2 ) BMI Z score h glucose WBISI Demographic parameters for age, BMI, and BMI Z score (age and gender adjusted) are presented as means SD. Mean SD for 2-h plasma glucose are also provided. Systeme International unit for glucose is mmol/liter (conversion factor, ). a Significantly different from mg/dl group (P 0.05).

3 Yeckel et al. -Cell Impairment in NGT Obese Youth J Clin Endocrinol Metab, February 2005, 90(2): FIG. 1. Glucose, insulin, and C-peptide concentrations during the OGTT for each level of 2-h glucose (left) and (0 30 min) for glucose, insulin, and C-peptide (right). Data are expressed as least-squares means and 95% confidence intervals (adjusted for age, gender, race/ethnicity, and BMI). The early glucose response was significantly greater with increasing 2-h glucose level (*, P compared with lowest 2-h glucose;, P 0.05 for moderate to high 2-h glucose comparison). No difference between groups was observed for early insulin or C-peptide. Systeme International unit for glucose is mmol/ liter (conversion factor, ); for insulin, pmol/liter (conversion factor, 6.0). -cell function. The insulinogenic index (IGI), a commonly used index of -cell function (9), was calculated from the OGTT data: IGI insulin (0 30) in U/ml divided by the glucose (0 30) in mg/dl. In children and adolescents, the IGI is positively correlated with the first-phase insulin response from the hyperglycemic clamp (r 0.68; P 0.001; Caprio, S., unpublished data). Alternatively, we replaced the change in insulin with the change in C-peptide (pmol/liter) divided by the glucose (mmol/liter). Likewise, the area under the curve for glucose, insulin, and C-peptide were used to represent the incremental response for the entire OGTT. These area-under-the-curve values were obtained by the trapezoid rule. Insulin feedback system. In this analysis, we use both the hyperbolic insulin feedback curves, which demonstrate the relationship between insulin sensitivity and acute insulin response (10), and the single positional mean data, referred to here as the disposition index. We previously demonstrated using the IGI and WBISI parameters that obese NGT and IGT youth followed the classic insulin feedback model (10, 11) with the IGT youth shifted significantly to the left of the NGT youth (6). We therefore represent the disposition index here as the product of the IGI and WBISI. Statistical analysis Group comparisons were made using analysis of covariance. All least-squares means with corresponding 95% confidence intervals and significance tests were estimated with adjustment for sex, race/ethnicity (White, Black, and Hispanic), age, and body mass index (BMI). Planned contrasts with Bonferroni adjustment for multiple comparisons were performed to compare across NGT categories or to compare across NGT categories within insulin sensitivity strata. Where appropriate, geometric means are presented for variables that were logarithmically transformed to meet analysis assumptions. Summary hyperbolic feedback curves were compared using linear regression in which the logarithm of the IGI was modeled as a function of the logarithm of the WBISI. The effect of NGT category on these curves was examined by evaluating the main effect of NGT category as well as the interaction of NGT category with WBISI. This analysis was also adjusted for sex, race/ethnicity, age, and BMI. Comparisons of demographic variables were made by 2 analysis and ANOVA as appropriate. Results NGT cohort divided by 2-h glucose concentration Table 1 provides the demographic characteristics for participants and mean 2-h glucose values based on 2-h plasma glucose stratification. Although the absolute level of BMI was significantly different between the lowest and moderate 2-h glucose categories, BMI Z scores were not different between groups.

4 750 J Clin Endocrinol Metab, February 2005, 90(2): Yeckel et al. -Cell Impairment in NGT Obese Youth OGTT data (Fig. 1) The OGTT temporal data for glucose, insulin, and C-peptide responses are shown in Fig. 1, whereas the early changes in each of these parameters appear alongside. All statistical comparisons were made after adjustment for age, sex, ethnicity, and BMI. The group means for the early (0 30 min) glucose excursion were 36 (33 40), 46 (43 49), and 51 (48 54) mg/dl (translating to changes of 1.99, 2.55, and 2.83 mmol/ liter, respectively). Significant differences existed between the less than 100 mg/dl and mg/dl 2-h glucose groups (P 0.001), the less than 100 mg/dl to mg/dl groups (P 0.001), and the mg/dl and mg/dl groups (P 0.05). Importantly, insulin and C-peptide responses did not vary between groups. Insulin feedback system (Fig. 2) To evaluate -cell function, we examined the ratio of the initial change (0 30 min) in insulin (or C-peptide) to that of plasma glucose (insulinogenic index). The IGI tended to decrease (P 0.06) from the lowest to the moderate level of 2-h glucose and significantly decreased from the lowest to the highest 2-h glucose category (P 0.02). Insulin sensitivity (WBISI) also significantly decreased as 2-h glucose level increased. All data were adjusted for age, gender, BMI, and race/ethnicity. Consequently, above a glucose level of 100 mg/dl, the disposition index showed a decrement (P 0.01 for all comparisons). We found that both IGI and insulin sensitivity components of the insulin feedback system changed across 2-h glucose categories. Further inspection of the cohort, however, revealed a wide range of insulin sensitivities within each category of 2-h glucose. We therefore examined the same parameters with the added control of insulin sensitivity to determine whether the same deterioration in the insulin feedback system would persist at each level of insulin sensitivity. NGT cohort by 2-h glucose concentration with the additional stratification by insulin sensitivity The demographic breakdown for the more complex stratification appears in Table 2. Age, sex, ethnicity, and BMI and BMI Z scores did not vary significantly within each insulin sensitivity stratum. Each level of insulin sensitivity had obese youth corresponding to each of the different levels of 2-h glucose, although the less than 100 mg/dl category had a much lower representation in the very low insulin sensitivity stratum. OGTT data (Table 3) Fasting (0 min), early glucose, insulin, and C-peptide responses (0 30 min) from the OGTT are presented in Table 3. Fasting values did not differ significantly within each level of insulin sensitivity. However, the early increase in glucose (0 30 min) was significantly greater between the highest and lowest 2-h glucose categories in each insulin sensitivity stratum (P 0.01 for all comparisons). In addition, the 30- min glucose increased from the lowest to moderate 2-h glucose categories in the most insulin-sensitive stratum (P 0.006). The initial increase in glucose was accompanied by an FIG. 2. Relationship between -cell function (IGI) (top) and insulin sensitivity (WBISI) (middle) as a function of 2-h glucose category. Data are expressed as least-squares means and 95% confidence intervals (adjusted for age, gender, race/ethnicity, and BMI). For comparisons made between the lowest and more elevated 2-h glucose levels: *, P 0.02;, P 0.05 for moderate to high 2-h glucose levels. The net response on the insulin feedback system was a decrease in disposition index (bottom) at each level of increasing 2-h glucose category (*, P 0.01 for all comparisons). increased insulin (or C-peptide) response within each insulin sensitivity stratum. It is important to note that, although not significant, the general trend was a decrease in early insulin, not an increase, as would be expected if the insulin feedback system were attempting to compensate for the higher glucose levels. Insulin feedback system (Fig. 3) The stratification by level of insulin sensitivity accomplished the overall control for level of WBISI; however, there were slight decreases in values as 2-h glucose increased for each strata of insulin sensitivity (Table 2). As shown in Fig. 3, the calculated IGI was markedly reduced between the less

5 Yeckel et al. -Cell Impairment in NGT Obese Youth J Clin Endocrinol Metab, February 2005, 90(2): TABLE 3. OGTT characteristics of obese youth stratified by insulin sensitivity and 2-h glucose level Insulin sensitivity than 100 mg/dl and mg/dl 2-h glucose categories in each strata of insulin sensitivity (P 0.05 for all comparisons). The most insulin-sensitive group also had a significant decrease in IGI from the lowest to moderate 2-h glucose categories (P 0.002). The combined effect on the insulin feedback system was to decrease the disposition index between the less than 100 and mg/dl 2-h glucose categories similarly at every level of insulin sensitivity (P 0.01 for all comparisons). Likewise, the lowest and moderate 2-h glucose categories were different in the most insulinsensitive stratum (P 0.001). Summary hyperbolic feedback curves constructed from all participants according to 2-h glucose category appear in Fig. 4. Curves were compared by linear regression using the logarithm of the IGI modeled as a function of the logarithm of the WBISI. A significant main effect of 2-h glucose category was observed (P 0.001), indicating a leftward shift in the insulin feedback curves as 2-h glucose increased in obese youth: group comparisons less than 100 mg/dl to mg/dl, P 0.001; less than 100 mg/dl to mg/dl, P 0.001; and mg/dl to mg/dl, P Discussion During an OGTT, 2-h glucose levels below and above 140 mg/dl are defined by current criteria as normal and impaired glucose tolerance, respectively (12). We previously showed that 2-h plasma glucose levels in obese children and adolescents with IGT differ from the corresponding values in the obese youth with NGT because of a failure to overcome 2-h glucose (mg/dl) Very low (WBISI, 1.44) N min Glucose (mg/dl) 87 (84 91) 90 (88 92) 92 (90 94) Insulin ( U/ml) 49 (41 59) 45 (41 49) 46 (41 50) C-peptide (pmol/l) 1321 ( ) 1336 ( ) 1429 ( ) 0 30 min Glucose 38.5 ( ) b 48.6 ( ) 54.1 ( ) Insulin 290 ( ) 258 ( ) 251 ( ) C-peptide 3509 ( ) 2622 ( ) 2766 ( ) Low (WBISI, ) N min Glucose 89 (85 92) 89 (87 91) 91 (89 94) Insulin 28 (26 31) 27 (26 29) 26 (24 27) C-peptide 942 ( ) 994 ( ) 1049 ( ) 0 30 min Glucose 36.6 ( ) b 44.6 ( ) 51.5 ( ) Insulin 183 ( ) 159 ( ) 146 ( ) C-peptide 2475 ( ) 2236 ( ) 2283 ( ) Moderate (WBISI, 2.25) N min Glucose 86 (85 88) 87 (86 89) 90 (87 93) Insulin 17 (16 19) 15 (14 16) 16 (14 18) C-peptide 738 ( ) 704 ( ) 812 ( ) 0 30 min Glucose 33.9 ( ) a,b 48.3 ( ) 48.8 ( ) Insulin 106 (90 125) 94 (77 115) 81 (63 103) C-peptide 1826 ( ) 1806 ( ) 1824 ( ) Data are expressed as least-squares means (95% confidence intervals). Data are adjusted for age, BMI, gender, and race/ethnicity. Systeme International unit for glucose is mmol/liter (conversion factor, ); for insulin, pmol/liter (conversion factor, 6.0). a Significantly different from mg/dl 2-h glucose category (P 0.05). b Significantly different from mg/dl 2-h glucose category (P 0.05). insulin resistance by an appropriate insulin response during the first 30 min after the oral glucose load (6). These data revealed a significant leftward shift in the hyperbolic plot using the IGI for insulin secretion and WBISI for insulin sensitivity. Similar results have been reported for obese IGT adults (7, 13). Even more severe impairments in early insulin responses to oral or iv glucose have been reported in adults and children with T2DM (8, 14, 15). In this study we examined the importance of the early insulin response on 2-h glucose levels at the other end of the glucose tolerance spectrum, namely, obese youth with NGT. The use of the OGTT in this investigation allowed us to examine a large number of subjects both to facilitate the subdivision by NGT status and to further stratify by the level of insulin sensitivity. It is noteworthy that the two NGT groups with higher 2-h glucose levels were on average more insulin resistant than the group with the lowest glucose levels. At first glance, the 2-h glucose category data suggest that worsening insulin resistance may contribute to early deterioration in oral glucose tolerance. On further analysis, however, insulin secretion, as reflected by the IGI, appears to have a strong impact on glucose tolerance (albeit still normal), irrespective of insulin sensitivity. As illustrated in Fig. 3, at any level of insulin sensitivity (WBISI), the IGI was lowest in the group with the highest 2-h glucose level. The same result is obtained if C-peptide replaces insulin in the equation. These data provide support for the notion that even obese youth with NGT who are more insulin sensitive can

6 752 J Clin Endocrinol Metab, February 2005, 90(2): Yeckel et al. -Cell Impairment in NGT Obese Youth FIG. 3. -Cell function and disposition index as a function of insulin sensitivity level. Data are expressed as least-squares means and 95% confidence intervals (adjusted for age, gender, race/ethnicity, and BMI). All groups showed a similar pattern of reduction with increasing 2-h glucose irrespective of insulin sensitivity, for the IGI (top) between the less than 100 mg/dl and mg/dl glucose categories (*, P 0.05 for all comparisons). The disposition index (bottom) revealed a decrement from less than 100 to mg/dl categories in all insulin sensitivity categories (*, P 0.01 for all comparisons). have a perturbed -cell response to a normal physiological stimulus (glucose ingestion). Interestingly, this pattern of -cell dysfunction could be replicated if fasting insulin tertiles were used in place of the WBISI strata (highest insulin tertile corresponding to lowest WBISI tertile). Furthermore, these data emphasize that it is important to know both insulin secretion and insulin sensitivity information, a topic recently reviewed by Ahren and Pacini (16). Obese youth typically have a large increase in their firstphase insulin response (hyperglycemic clamp technique) compared with nonobese controls because of their increased insulin resistance (2, 4). The insulin response to the hyperglycemic clamp can be further augmented by ingestion of glucose (17). In this investigation all subjects were obese youth; therefore, when viewed as a single group, they manifested a large insulin response and were in general insulin resistant. These data fit within the normal context of the insulin feedback system, which represents the response from the -cells to secrete more insulin as a function of decreased peripheral insulin sensitivity (10, 11). The large-magnitude insulin response between insulin sensitivity strata, observed here and previously (6), highlight the extent of adaptation to FIG. 4. Summary hyperbolic feedback curves constructed from all participants grouped by 2-h glucose category. Feedback curves were compared using linear regression in which the logarithm of the IGI was modeled as a function of the logarithm of the WBISI. There was a significant leftward shift (toward IGT) in the insulin feedback curve for increasing 2-h glucose category. This leftward shift remained significant after adjustment for age, gender, race/ethnicity, and BMI. Group comparisons: less than 100 mg/dl to mg/dl, P 0.001; less than 100 mg/dl to mg/dl, P 0.001; and mg/dl to mg/dl, P ambient insulin resistance and, therefore, the need to control for this overriding factor. The decrease in disposition index with increasing 2-h glucose concentration was primarily accounted for by the inability of the -cells to respond promptly and adequately during the oral glucose challenge. These observations are in accordance with the recent data from both the San Antonio metabolism (SAM) study in lean and obese adults (primarily Mexican-American) (18) and the American Diabetes Association Genetics of Non-Insulin-Dependent Diabetes (GENNID) study in first-degree relatives of type 2 diabetes of mixed ethnic background (19). In the SAM study, the investigators observed a similar decrease in -cell response at levels of 2-h glucose corresponding to more than 100 mg/dl. When their data were adjusted for differences in insulin sensitivity, they determined that the decreased -cell response was occurring irrespective of obesity. In the GENNID study, Jensen and colleagues (19) demonstrated a reduced IGI as glucose tolerance worsened when adjusted for the homeostasis assessment model for insulin resistance. This finding was also observed within the NGT participants based on a higher than median glucose response to the OGTT. A low disposition index has been identified in a prospective study in adults to have predictive value for worsening glucose tolerance (20). Similar to our findings, the GENNID study data also showed perturbed -cell function, adjusted for insulin resistance, across all ethnic groups. Our population of youth represent a growing population of obese children and adolescents (1) at high risk for abnormal glucose tolerance (21). Given both the diverse age and body composition represented in the adult populations from both the SAM and GENNID studies, it is likely that the same pattern of -cell dysfunction would exist even for nonobese youth, e.g. firstdegree relatives of T2DM. However, screening of pediatric populations in general has been very limited.

7 Yeckel et al. -Cell Impairment in NGT Obese Youth J Clin Endocrinol Metab, February 2005, 90(2): Although the mechanism underlying the hyperbolic relationship between insulin secretion and insulin resistance is still unknown, several factors have been proposed to explain the inadequate -cell response in subjects with altered glucose tolerance: -cell mass, glucose sensitivity, and rate sensitivity. One hypothesis regarding the response of the pancreas to insulin resistance is that -cell mass increases as an adaptive mechanism to increase overall insulin secretion, thus overcoming peripheral insulin resistance. For example, Butler and colleagues (22) demonstrated that a high -cell mass is evident in obese nondiabetic individuals, whereas it is reduced in patients with T2DM. A finding consistent with the potential for increased -cell mass is the large adaptive hyperinsulinemia observed across insulin sensitivity strata (we observed greater than a 2.5-fold increase in insulin response over the range of insulin sensitivity levels in these obese youth). This adaptive hyperinsulinemia was particularly noteworthy given the apparent inability of individuals within a given insulin sensitivity level to augment insulin secretion to accommodate elevated glucose concentrations (increased 2-h glucose category). This observation is consistent with results obtained from using multiple methods for quantifying insulin secretion (23). Two of the factors, glucose sensitivity and rate sensitivity (response to change in glucose), are components of the IGI used here as the surrogate for -cell function. Modeling performed by Ferrannini and colleagues (24) recently demonstrated that individuals with IGT (vs. normal controls) had reduced glucose sensitivity, whereas rate sensitivity was unaltered. Because we observed no change or a decrease in the initial insulin response, despite a wide range of resulting glucose concentrations across the NGT spectrum, the -cell sensitivity to glucose may be impaired even in young obese NGT individuals. Controlling for the overall magnitude of the insulin response by examining the data by level of insulin sensitivity served to further strengthen the general interpretation that the deficient -cell response is consistent with impaired glucose sensitivity. The net result from our data would indicate that insulin resistance is likely a necessary component for large-magnitude changes in insulin secretory response to a glucose load, whatever the mechanism for adaptive plasticity within the -cells to increase insulin secretion. This concept fits within the general framework of the hyperbolic feedback curves (10, 16). We identified a systematic leftward shift in each hyperbolic curve (Fig. 4) as glucose increased across 2-h glucose categories toward IGT. This demonstrates that the hyperbolic model for glucose tolerance can distinguish even finer levels of perturbation in glucose homeostasis besides the traditional broad categories of NGT, IGT, and T2DM. In summary, increased 2-h glucose concentrations during an OGTT in NGT obese youth reflect the apparent inability of the pancreatic -cells to fully compensate for early increases in glucose. We observed the same pattern of dysfunction at every level of insulin sensitivity. Furthermore, we determined that the insulin feedback curves were sensitive enough to identify differences in 2-h glucose level even in the NGT range. Consequently, in obese children and adolescents, the transition from NGT to IGT and ultimately T2DM more likely represents a gradual deterioration in glucosestimulated insulin response rather than a threshold effect or an all-or-none phenomenon. Acknowledgments We are particularly grateful to all the children and adolescents who participated in the OGTT studies. We thank the nursing staff, especially Karin Allen and Melinda Lopes, and the research staff of the Yale University School of Medicine General Clinical Research Center and Core Laboratory for their excellent care of subjects and professional assistance during the studies. Received June 30, Accepted October 25, Address all correspondence and requests for reprints to: Dr. Sonia Caprio, Yale University School of Medicine, Department of Pediatrics, P.O. Box , New Haven, Connecticut Sonia. Caprio@yale.edu. This work was supported by National Institutes of Health (NIH) Grants RO1-HD40787 and RO1-HD28016 (S.C.), the Stephen I. Morse Pediatric Diabetes Research Fund (R.W.), and NIH Grants MO1- RR00125 and MO1-RR S.C. is a recipient of a K24 HD Award for Patient-Oriented Research. References 1. Goran MI, Ball GD, Cruz ML 2003 Obesity and risk of type 2 diabetes and cardiovascular disease in children and adolescents. J Clin Endocrinol Metab 88: Caprio S, Bronson M, Sherwin RS, Rife F, Tamborlane WV 1996 Co-existence of severe insulin resistance and hyperinsulinaemia in pre-adolescent obese children. Diabetologia 39: Monti LD, Brambilla P, Stefani I, Caumo A, Magni F, Poma R, Tomasini L, Agostini G, Galli-Kienle M, Cobelli C, Chiumello G, Pozza G 1995 Insulin regulation of glucose turnover and lipid levels in obese children with fasting normoinsulinaemia. Diabetologia 38: Uwaifo GI, Parikh SJ, Keil M, Elberg J, Chin J, Yanovski JA 2002 Comparison of insulin sensitivity, clearance, and secretion estimates using euglycemic and hyperglycemic clamps in children. J Clin Endocrinol Metab 87: Matsuda M, DeFronzo RA 1999 Insulin sensitivity indices obtained from oral glucose tolerance testing: comparison with the euglycemic insulin clamp. Diabetes Care 22: Yeckel CW, Weiss R, Dziura J, Taksali SE, Dufour S, Burgert TS, Tamborlane WV, Caprio S 2004 Validation of insulin sensitivity indices from oral glucose tolerance test parameters in obese children and adolescents. J Clin Endocrinol Metab 89: Kahn SE 2003 The relative contributions of insulin resistance and -cell dysfunction to the pathophysiology of type 2 diabetes. Diabetologia 46: DeFronzo RA, Bonadonna RC, Ferrannini E 1992 Pathogenesis of NIDDM: a balanced overview. Diabetes Care 15: Phillips DI, Clark PM, Hales CN, Osmond C 1994 Understanding oral glucose tolerance: comparison of glucose or insulin measurements during the oral glucose tolerance test with specific measurements of insulin resistance and insulin secretion. Diabet Med 11: Kahn SE, Prigeon RL, McCulloch DK, Boyko EJ, Bergman RN, Schwartz MW, Neifing JL, Ward WK, Beard JC, Palmer JP, Porte Jr D 1993 Quantification of the relationship between insulin sensitivity and -cell function in human subjects: evidence for a hyperbolic function. Diabetes 42: Bergman RN 1989 Lilly lecture Toward physiological understanding of glucose tolerance: minimal-model approach. Diabetes 38: Report of the Expert Committee on the Diagnosis and Classification of Diabetes Mellitus. Diabetes Care 23(Suppl 1):S4 S Larsson H, Ahren B 1996 Failure to adequately adapt reduced insulin sensitivity with increased insulin secretion in women with impaired glucose tolerance. Diabetologia 39: Libman I, Arslanian S 2003 Type 2 diabetes in childhood: the American perspective. Horm Res 59(Suppl 1): Pratley RE, Weyer C 2002 Progression from IGT to type 2 diabetes mellitus: the central role of impaired early insulin secretion. Curr Diab Rep 2: Ahren B, Pacini G 2004 Importance of quantifying insulin secretion in relation to insulin sensitivity to accurately assess -cell function in clinical studies. Eur J Endocrinol 150: Heptulla RA, Tamborlane WV, Cavaghan M, Bronson M, Limb C, Ma YZ, Sherwin RS, Caprio S 2000 Augmentation of alimentary insulin secretion despite similar gastric inhibitory peptide (GIP) responses in juvenile obesity. Pediatr Res 47: Gastaldelli A, Ferrannini E, Miyazaki Y, Matsuda M, DeFronzo RA Cell dysfunction and glucose intolerance: results from the San Antonio metabolism (SAM) study. Diabetologia 47:31 39

8 754 J Clin Endocrinol Metab, February 2005, 90(2): Yeckel et al. -Cell Impairment in NGT Obese Youth 19. Jensen CC, Cnop M, Hull RL, Fujimoto WY, Kahn SE Cell function is a major contributor to oral glucose tolerance in high-risk relatives of four ethnic groups in the U.S. Diabetes 51: Larsson H, Ahren B 2000 Islet dysfunction in insulin resistance involves impaired insulin secretion and increased glucagon secretion in postmenopausal women with impaired glucose tolerance. Diabetes Care 23: Sinha R, Fisch G, Teague B, Tamborlane WV, Banyas B, Allen K, Savoye M, Rieger V, Taksali S, Barbetta G, Sherwin RS, Caprio S 2002 Prevalence of impaired glucose tolerance among children and adolescents with marked obesity. N Engl J Med 346: Butler AE, Janson J, Bonner-Weir S, Ritzel R, Rizza RA, Butler PC Cell deficit and increased -cell apoptosis in humans with type 2 diabetes. Diabetes 52: Ahren B, Larsson H 2002 Quantification of insulin secretion in relation to insulin sensitivity in nondiabetic postmenopausal women. Diabetes 51(Suppl 1):S202 S Ferrannini E, Gastaldelli A, Miyazaki Y, Matsuda M, Pettiti M, Natali A, Mari A, DeFronzo RA 2003 Predominant role of reduced -cell sensitivity to glucose over insulin resistance in impaired glucose tolerance. Diabetologia 46: JCEM is published monthly by The Endocrine Society ( the foremost professional society serving the endocrine community.