Estimates of insulin secretory function in apparently healthy volunteers vary as a function of how the relevant variables are quantified.
More recently, it has been suggested that to understand the physiological significance of values of HOMA-[beta], it is necessary to relate them to a relevant measure of insulin resistance (4). Homeostasis model assessment of insulin resistance (HOMA-IR) is often used for this purpose, as it is also based on a mathematical formula involving FPG and FPI concentrations (1). HOMA-IR provides a surrogate estimate of insulin action that is determined under basal conditions and accounts for no more than one-third of the variance of an estimate of insulin action based on the ability of insulin to limit the increase in plasma glucose concentration in response to continuous glucose infusion in apparently healthy individuals (5, 6). In addition, because HOMA-[beta] is an estimate of insulin secretory function based on only FPG and FPI concentrations, we thought it would prove interesting to compare HOMA-[beta] with the day-long plasma insulin response to a test mixed meal, a variable that takes into account more than just FPG and FPI.
To address these issues, we determined HOMAIR, HOMA-[beta], the ability of insulin to limit the change in plasma glucose concentration that occurs in response to a constant infusion of glucose, and the plasma glucose and insulin responses to test mixed meals in 285 apparently healthy individuals whose FPG was <7.0 mmol/L (126 mg/dL).
Materials and Methods
The experimental population consisted of 285 volunteers (180 women and 105 men) who had responded to newspaper advertisements describing our interest in the role of insulin resistance in human disease. Participants were apparently healthy, with FPG concentrations <7.0 mmol/L (126 mg/dL), and without known vascular disease. In addition, subjects had a normal medical history and physical examination and were free of hematologic, liver, or kidney disease. Stanford University's Human Subjects committee approved the study protocol, and all subjects gave written informed consent for participation.
The ability of insulin to modulate the change in plasma glucose concentration in response to a constant infusion of glucose was quantified by a modified version (7) of the insulin suppression test (IST) as described and validated by our research group (8, 9). After an overnight fast, an intravenous catheter was placed in each participant's arm, 1 for the 3-h infusion of octreotide (0.27 /[micro]g/[m.sup.2]/min), insulin (32 mU/[m.sup.2]/ min), and glucose (267 mg/[m.sup.2]/min), and 1 for drawing blood. Plasma glucose and insulin concentrations were measured every 10 min from 150 to 180 min and averaged to obtain the steady-state plasma glucose (SSPG) and insulin (SSPI) concentrations. Because SSPI concentrations are comparable in all individuals, and glucose infusion is identical, the SSPG concentration provides an estimate of the ability of insulin to limit the magnitude of the glucose concentration that occurs in response to the infused glucose load: the greater the SSPG concentration, the more "resistant" the individual is to the action of insulin. We have previously shown that estimates of degree of insulin resistance with the IST correlate highly (r > 0.9) with measures of insulin action obtained using the hyperinsulinemic, euglycemic clamp method (9).
On another morning after an overnight fast, volunteers were admitted to the General Clinical Research Center for day-long measurements of plasma glucose and insulin concentrations before and after test breakfast and lunch meals (10, 11). Each meal had, as percent of daily calories, 15% protein, 42% carbohydrate, and 43% fat. Breakfast was at 0800 (20% of estimated daily caloric intake) and lunch was at noon (40% of estimated daily caloric intake). Blood was drawn at hourly intervals from 0800 to 1600, before and after the meals.
Data are presented as mean (SD). We used fasting plasma insulin (mU/L) and glucose (mmol/L) concentrations to calculate HOMA-[beta] [(20 X insulin)/(glucose - 3.5)] and HOMA-IR (insulin X glucose/22.5) as described by Matthews et al. (1). We calculated the total integrated day-long plasma glucose [glucose area under the curve (AUC)] and insulin (insulin AUC) responses using the trapezoidal rule. Statistical calculations used SPSS Statistics 17.0. We evaluated relationships between variables of interest by calculating Pearson correlation coefficients and compared correlations using the Meng-Rosenthal-Rubin method (12) with Fisher's r-to-z transformation and a 2-tailed significance test.
The demographic and metabolic characteristics of the study population are given in Table 1. Of note, the group tended to be overweight/obese, with a mean body mass index of 30.4 kg/[m.sup.2]. All of the experimental variables varied by at least 2-fold, and of particular importance to this analysis was the wide range of values for insulin action (SSPG concentration and HOMA-IR varied by approximately 10-fold).
The relationship between estimates of insulin action provided by calculating HOMA-IR and those obtained by determining the SSPG concentration during the IST is illustrated in Fig. 1. These data demonstrate that although the 2 methods to quantify insulin action are significantly correlated (r = 0.47, P < 0.001), the magnitude of the relationship is relatively modest.
[FIGURE 1 OMITTED]
Fig. 2, A and B compare results when insulin secretory function, as estimated from HOMA-[beta], is related to the coexisting degree of insulin resistance, as estimated by either HOMA-IR or SSPG concentration. The measurements of HOMA-IR and HOMA-[beta] are highly correlated (r = 0.74, P < 0.001) in this group of healthy individuals (Fig. 2A), a result that is not surprising considering that both estimates are derived from values of FPG and FPI concentrations. In contrast, there is a much weaker relationship (r = 0.22, P < 0.001) between degree of insulin resistance, as measured by SSPG concentration, and insulin secretory function as estimated by HOMA-[beta]. These 2 correlations were statistically different from each other (P < 0.001).
[FIGURE 2 OMITTED]
Given the discrepancy between evaluations of insulin secretory function when viewed in relationship to the 2 methods for estimating insulin action, we were interested in what would happen if the total day-long insulin response to test mixed meals (insulin AUC) were viewed in the context of the HOMA-IR and SSPG concentration. The results of this comparison (Fig. 2, C and D) show that when insulin secretory function is estimated by insulin AUC in response to mixed meals, it does not seem to matter if it is related to HOMA-IR (r = 0.67, P < 0.001) or SSPG concentration (r = 0.59, P < 0.001).
Pancreatic [beta]-cell function is often evaluated by relating it to the coexisting plasma glucose concentration, in addition to degree of insulin resistance. We followed this approach by relating the total integrated day-long plasma glucose concentration (glucose AUC) in response to mixed meals and either HOMA-[beta] (Fig. 3A) or the day-long insulin response (Fig. 3B). These data demonstrate that apparently healthy individuals are able to maintain plasma glucose concentrations within a relatively narrow range and maintain glucose homeostasis by varying their insulin secretory response several fold. Furthermore, the results do not differ substantially depending on how the insulin secretory response is being assessed.
[FIGURE 3 OMITTED]
Finally, we evaluated the relationship of HOMA-[beta] and the insulin AUC to the coexisting FPG concentration (Fig. 3, C and D). In this instance, the physiological response of the [beta]-cell does seem to vary as a function of how it is estimated. Thus, Fig. 3C is consistent with the conclusion that as FPG concentration increases throughout the nondiabetic range, insulin secretory function, as estimated by HOMA-[beta], declines (r = -0.33, P < 0.001). In contrast, as shown in Fig. 3D, insulin AUC increases modestly as FPG concentrations increase (r = 0.23, P < 0.001). These 2 correlations are statistically different from each other (P < 0.001).
The results of this study emphasize that differences in experimental technique confound attempts to evaluate insulin secretory function, whether or not it is related to estimates of insulin action or level of glycemia. For example, discordant findings are observed when insulin secretory function, as assessed by HOMA-/, is considered in light of 2 different estimates of insulin action, HOMA-IR or SSPG concentration. The more insulin resistant the subject, as assessed by HOMA-IR (a mathematical estimate of insulin action), the greater the insulin secretory response as estimated by HOMA-[beta] (r = 0.74) (Fig. 2A). In contrast, the magnitude of the relationship between insulin action and secretion is markedly attenuated (r = 0.22) when viewed in terms of SSPG concentration (Fig. 2B).
The strong relationship between HOMA-IR and HOMA-[beta] suggests that the pancreatic [beta]-cells of these individuals compensate effectively for increasing degrees of insulin resistance. In contrast, the relationship between SSPG concentration and HOMA-[beta] indicates that, in these same patients, increases in insulin secretory function do not effectively compensate for increasing degrees of insulin resistance.
Which of these conclusions regarding insulin secretory function in our group of apparently healthy individuals is the right one?
This issue gets more confounded when we compare the relationship between a very different estimate of insulin secretory function and the 2 measures of insulin action (SSPG concentration and HOMA-IR). To avoid the emphasis on fasting values, we chose the total day-long integrated insulin response to test mixed meals as an index of insulin secretory function (insulin AUC). The results of these comparisons (Fig. 2, C and D) demonstrate that reasonably comparable relationships are seen when insulin secretory function as estimated by insulin AUC is viewed in light of increasing degrees of insulin resistance as assessed by either HOMA-IR (r = 0.67, P < 0.001) or SSPG concentration (r = 0.59, P < 0.001). The observation--that [beta]-cell function, as estimated by insulin AUC, correlates with the 2 disparate measures of insulin action to a comparable degree--does not necessarily mean that insulin AUC, an estimate based on postprandial insulin measurements, provides a "better" estimate of overall insulin secretory function than HOMA-[beta], an estimate based on a mathematical formula involving fasting glucose and insulin concentrations. Rather, it emphasizes how conclusions concerning insulin secretory function vary with the methods used to assess it.
Insulin secretory function is also often evaluated in relationship to the plasma glucose concentration (13), and we have taken that approach to relate HOMA-[beta] and insulin AUC to 2 measures of glycemia. The results again emphasize that conclusions regarding [beta]-cell function vary dramatically with how the relevant variables are defined. If total day-long integrated glucose response to mixed meals (glucose AUC) is used as the estimate of glycemia, the apparently healthy population we studied were able to increase their insulin secretory response to mixed meals by approximately 6-fold (Fig. 3, A and B), irrespective of how it was measured, thereby maintaining the glucose AUC within a narrow range (approximately 1.5-fold). When HOMA-[beta] and insulin AUC are examined in relationship to FPG concentration, however, insulin secretory function as assessed by HOMA-[beta] declines (Fig. 3C) as FPG concentrations increase (r = -0.33, P < 0.001), whereas values of insulin AUC increase (Fig. 3D) as FPG increases (r = 0.23, P < 0.001). These striking differences are particularly relevant in the context of the recent article by Tabak et al. (3), in which these authors interpreted measurements of HOMA-[beta] to conclude that the development of significant fasting hyperglycemia is preceded by a period of declining insulin secretory function, the relationship seen in Fig. 3C. In contrast, ifinsulin AUC is used as a measure of the compensatory response to increasing levels of glycemia, the data in Fig. 3D lead to an entirely different conclusion.
The goal of our study was not to decide how best to evaluate insulin secretory function in apparently healthy subjects. We wished to emphasize how important it is to address this issue in the future by demonstrating that physiological insights into insulin secretory function in these individuals will vary with differences in the indices used. To do this, we accepted the premise that estimates of insulin secretory function should be viewed in the context of either insulin resistance or degree of glycemia. Our results indicate that conclusions regarding [beta]-cell function in nondiabetic individuals vary depending on (1) the method to assess insulin secretory function and (2) which measure of insulin action or glycemia is used as the independent variable. This dilemma is not unique to the methods we used to assess insulin action and secretion and glycemia, but will extend to other experimental approaches to evaluating insulin secretory function. One example of the many approaches that have been used for this purpose is to take the acute insulin response to intravenous glucose as the measure of insulin secretion and relate that value to insulin sensitivity as estimated by the frequently sampled intravenous glucose tolerance test (14). Another approach has been to take the graded glucose infusion method pioneered by Polonsky and colleagues (15) to quantify insulin secretion rate and relate it to quantification ofinsulin action with the IST (16). Thus, stating that the estimate of insulin secretion used in a study has been viewed in relationship to a relevant physiological variable, e.g., insulin resistance and/or degree of glycemia, does not ensure that a meaningful measure of pancreatic [beta]-cell function has been obtained. In this context, it seems best to conclude by suggesting that it is incumbent on both (1) the investigator publishing studies on insulin secretory function to clearly state the potential limitations of the measurements used to quantify the independent and dependent variables and (2) the reader to evaluate the quality of the methods used and be aware of the impact that these choices can have on interpretation of the data.
Author Contributions: All authors confirmed they have contributed to the intellectual content of this paper and have met the following 3 requirements: (a) significant contributions to the conception and design, acquisition of data, or analysis and interpretation of data; (b) drafting or revising the article for intellectual content; and (c) final approval of the published article.
Authors' Disclosures or Potential Conflicts of Interest: No authors declared any potential conflicts of interest.
Role of Sponsor: The funding organizations played no role in the design of study, choice of enrolled patients, review and interpretation of data, or preparation or approval of manuscript.
Acknowledgments: We are indebted to Karen Johnson for her assistance in preparation of the accompanying figures.
(1.) Matthews DR, Hosker JP, Rudenski AS, Naylor BA, Treacher DF, Turner RC. Homeostasis model assessment: insulin resistance and [beta]-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia 1985;28:412-9.
(2.) U.K. Prospective Diabetes Study Group. U.K. Prospective Diabetes Study 16. Overview of 6 years' therapy of type II diabetes: a progressive disease. Diabetes 1995;44:1249-58.
(3.) Tabak AG, Jokela M, Akbaraly TN, Brunner EJ, Kivimaki M, Witte DR. Trajectories of glycaemia, insulin sensitivity, and insulin secretion before diagnosis of type 2 diabetes: an analysis from the Whitehall II study. Lancet 2009;373:2215-21.
(4.) Wallace TM, Levy JC, Matthews DR. Use and abuse of HOMA modeling. Diabetes Care 2004; 27:1487-95.
(5.) Yeni-Komshian H, Carantoni M, Abbasi F, Reaven GM. Relationship between several surrogate estimates of insulin resistance and quantification of insulin-mediated glucose disposal in 490 healthy, nondiabetic volunteers. Diabetes Care 2000;23: 171-5.
(6.) Kim SH, Abbasi F, Reaven, GM. Impact of degree of obesity on surrogate estimates of insulin resistance. Diabetes Care 2004;27:1998-2002.
(7.) Pei D, Jones CNO, Bhargava R, Chen Y-DI, Reaven GM. Evaluation of octreotide to assess insulin-mediated glucose disposal by the insulin suppression test. Diabetologia 1994;37:843-5.
(8.) Shen S-W, Reaven GM, Farquhar JW. Comparison of impedance to insulin-mediated glucose uptake in normal subjects and in subjects with latent diabetes. J Clin Invest 1970;49:2151-60.
(9.) Greenfield MS, Doberne L, Kraemer F, Tobey T, Reaven G. Assessment of insulin resistance with the insulin suppression test and the euglycemic clamp. Diabetes 1981;30:387-92.
(10.) Lamendola C, Abbasi F, Chu JW, Hutchinson H, Cain V, Leary E, et al. Comparative effects of rosuvastatin and gemfibrozil on glucose, insulin, and lipid metabolism in insulin-resistant, nondiabetic patients with combined dyslipidemia. Am J Cardiol 2005;95:189-93.
(11.) Abbasi F, Chen Y-DI, Farin HMF, Lamendola C, Reaven GM. Comparison of three treatment approaches to decreasing cardiovascular disease risk in nondiabetic insulin-resistant dyslipidemic subjects. Am J Cardiol 2008;102:64-9.
(12.) Meng X-L, Rosenthal R, Rubin DB. Comparing correlated correlation coefficients. Psychol Bull 1992;111:172-5.
(13.) Haffner SM, Miettinen H, Gaskill SP, Stern MP. Decreased insulin secretion and increased insulin resistance are independently related to the 7-year risk of NIDDM in Mexican-Americans. Diabetes 1995;44:1386-95.
(14.) Kahn SE The importance of beta-cell failure in the development and progression of type 2 diabetes. J Clin Endocrinol Metab 2001;86:4047-58.
(15.) Byrne MM, Sturis J, Polonsky KS. Insulin secretion and clearance during low-dose graded glucose infusion. Am J Physiol 1995;268:E21-7.
(16.) Jones CN, Pei D, Staris P, Polonsky KS, Chen YD-I, Reaven GM. Alterations in the glucose-stimulated insulin secretory dose-response curve and in insulin clearance in nondiabetic insulin-resistant individuals. J Clin Endocrinol Metab 1997;82: 1834-8.
Barry R. Johns,  * Fahim Abbasi,  and Gerald M. Reaven 
 Department of Medicine, Division of Endocrinology, Gerontology, and Metabolism and Division of Cardiovascular Medicine, Stanford University School of Medicine, Stanford, CA.
 Nonstandard abbreviations: HOMA-[beta], homeostasis model assessment of [beta]-cell function; FPG, fasting plasma glucose; FPI, fasting plasma insulin; HOMA-IR, homeostasis model assessment of insulin resistance; IST, insulin suppression test; SSPG, steady-state plasma glucose; SSPI, steady-state plasma insulin; AUC, area under the curve.
* Address correspondence to this author at: 300 Pasteur Dr., Room S025, Stanford, CA 94305; Fax 650-725-7085, e-mail firstname.lastname@example.org.
Received June 30, 2010; accepted January 21, 2011.
Previously published online at DOI: 10.1373/clinchem.2010.152744
Table 1. Demographic and metabolic characteristics of the study population (105 men, 180 women). Mean (SD) Range Age, years 51 (9) 23-72 Body mass index, kg/[m.sup.2] 30.4 (3.6) 18.8-38.2 FPG, mmol/L 5.36 (0.65) 3.97-6.94 FPI, pmol/L 119.4(67.6) 20.8-388.9 HOMA-IR 2.22 (1.23) 0.38-6.85 SSPG, mmol/L 10.43 (3.94) 2.11-20.70 HOMA-[beta] 141.7 (55.0) 27.2-382.6 Glucose AUC, mmol/L 47.3 (5.7) 35.8-67.1 Insulin AUC, pmol/L 2705 (1697) 413-10 527
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||Endocrinology and Metabolism|
|Author:||Johns, Barry R.; Abbasi, Fahim; Reaven, Gerald M.|
|Date:||Apr 1, 2011|
|Previous Article:||Newborn screening for isovaleric acidemia using tandem mass spectrometry: data from 1.6 million newborns.|
|Next Article:||Short-term treadmill running as a model for studying cell-free DNA kinetics in vivo.|