| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Original Studies |
Divisions of Metabolism, Endocrinology, and Nutrition, and Gerontology and Geriatric Medicine (R.S.S.), Department of Medicine, Veterans Affairs Puget Sound Health Care System, Harborview Medical Center (R.S.S.), and University of Washington, Seattle, Washington 98108
Address all correspondence and requests for reprints to: Michael E. Røder, M.D., Department of Endocrinology F, University of Copenhagen, Hillerød Sygehus, Helsevej 2, DK-3400 Hillerød, Denmark. address: mir{at}dadlnet.dk
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Twenty-six healthy older (mean ± SEM age, 67 ± 1 yr) and 22 younger (28 ± 1 yr) subjects with similar body mass indexes (27.9 ± 0.6 vs. 26.3 ± 1.0 kg/m2) were studied. PI was measured by a RIA recognizing both intact PI and its conversion intermediates. The insulin sensitivity index (SI) was quantified using the minimal model, and B cell function was measured as fasting insulin levels, the acute insulin response to glucose (AIRglucose), and as the acute insulin response to arginine at maximal glycemic potentiation (AIRmax). B cell function was also adjusted for SI based on the known hyperbolic relationship between these two variables.
Older and younger subjects had similar fasting glucose (5.3 ± 0.1 vs. 5.2 ± 0.1 mmol/L), IRI (83 ± 8 vs. 76 ± 9 pmol/L), PI (8.9 ± 0.8 vs. 10.6 ± 2.0 pmol/L), and PI/IRI ratio (12.3 ± 1.3% vs. 13.9 ± 1.6%; all P = NS) despite a 50% reduction of insulin sensitivity (SI, 1.94 ± 0.21 vs. 3.88 ± 0.38 x 10-5 min-1/pmol·L; P < 0.001) and in B cell function [SI x fasting IRI, 139 ± 18 vs. 244 ± 24 x 10-5 (P < 0.001); SI x AIRglucose, 0.75 ± 0.13 vs. 1.70 ± 0.15 x 10-2 min-1 (P < 0.001); SI x AIRmax, 3.63 ± 0.53 vs. 6.81 ± 0.70 x 10-2 min-1 (P < 0.001)] in the older subjects.
These findings suggest that the B cell dysfunction in older subjects is not associated with disproportionate proinsulinemia. However, in older subjects the B cell response to the insulin resistance of aging is reduced whether measured as fasting levels of PI or IRI or as the acute response to secretagogues. Thus, when examined in terms of the degree of insulin sensitivity, the lower fasting IRI levels in older subjects suggest that the utility of fasting insulin levels as a surrogate measure of insulin resistance in older individuals may be limited.
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
It has been well established that immunoreactive insulin (IRI) and proinsulin (PI) levels increase with insulin resistance. In healthy subjects, this alteration occurs in parallel, so that the PI/IRI ratio (PI/IRI) does not change. However, in type 2 diabetes, changes in IRI and PI levels are disproportionate, so that an elevated PI/IRI results (12, 13, 14, 15, 16, 17). It has also been demonstrated that an elevated fasting PI/IRI predicts the subsequent development of type 2 diabetes (18, 19, 20, 21), with this change being present 5 (18) or even 20 (20) yr before development of the clinical syndrome. It has previously been suggested that an elevated PI and/or PI/IRI is an early indicator of B cell secretory insufficiency (22, 23), and recently, we demonstrated that the elevated PI/IRI in type 2 diabetes reflects the degree of impaired B cell secretory capacity (24). As aging is associated with many of the secretory alterations observed in type 2 diabetes, albeit to a lesser degree, it is uncertain whether the response of fasting PI and IRI to the aging-associated insulin resistance is parallel as in young individuals, or disproportionate, more akin to the changes observed in individuals with or at high risk of developing type 2 diabetes.
To assess this, we measured fasting PI and IRI levels, insulin sensitivity, and B cell function both as the acute ability of glucose to directly stimulate the pancreatic B cells to secrete insulin (AIRglucose) and as the ability of glucose to potentiate the B cell insulin response to arginine (AIRmax) in groups of apparently healthy young and older subjects. Using these data, we determined whether aging is associated with a disproportionate increase in PI relative to IRI and whether the relationship between insulin sensitivity and fasting PI and IRI levels was similar in young and older subjects. This analysis also allowed us to determine whether these fasting measures could be used as surrogate measures of insulin resistance in older subjects as they are in younger individuals.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Forty-eight healthy male subjects without any history of
diabetes, cardiac, hepatic, or kidney disease were recruited from the
general population. Two age groups were selected: a group of 22 younger
subjects and a group of 26 older subjects. The mean ±
SEM age (range) was 28.0 ± 1.1 (20–38) and 66.8
± 1.0 (60–82) yr, respectively. Body mass indexes and fasting plasma
glucose levels were not significantly different between the two groups
(Table 1
). All subjects had fasting
plasma glucose levels below 7.0 mmol/L.
|
The study protocol was approved by the human subjects review committee of the University of Washington, and written informed consent was obtained from all subjects before commencement of the study.
All subjects were studied on two separate occasions. On 1 day an iv glucose tolerance test (IVGTT) was performed, and on the other day an arginine stimulation test was performed. The 2 study days were separated by no more than 2 weeks. All studies were performed at the Special Studies Unit at the Veterans Affairs Puget Sound Health Care System in Seattle. Studies commenced between 0700–0900 h after a minimum of a 10-h fast, and subjects remained supine throughout the studies. Blood samples were drawn from a catheter placed in a forearm vein in one arm, and iv solutions were administered through a catheter in the opposite forearm. The arm from which the blood samples were drawn was wrapped in a heating pad to arterialize the samples.
Insulin sensitivity and glucose tolerance
The insulin sensitivity index (SI) was calculated using Bergman’s minimal model of glucose kinetics (25) and provides an estimate of the ability of insulin to enhance the effect of hyperglycemia on glucose disposal. This measure is obtained from modeling of glucose and insulin data from a frequently sampled, tolbutamide-modified IVGTT (26). In brief, the study comprised baseline sampling, injection of iv glucose (11.4 g/m2) at 0 min, and injection of tolbutamide (125 mg/m2; Orinase, Upjohn, Kalamazoo, MI) at 20 min. In total, 34 blood samples were drawn between 0 and 240 min for glucose and insulin measurements. The intrasubject coefficient of variation for SI at our institution is 17% (27). The glucose disappearance constant (Kg) was calculated as the slope of the least squares regression line relating the natural logarithm of the glucose concentration to time between 10 and 19 min of the IVGTT and provides a measure of iv glucose tolerance.
B cell function tests
Fasting insulin immunoreactivity was calculated as the average of two basal measures. The acute insulin responses to both glucose and a nonglucose secretagogue were quantified. The acute insulin response to glucose (AIRglucose) was calculated as the mean increment above basal from 2–10 min after the glucose bolus administered in the IVGTT. Assessment of the potentiating effect of glucose on the B cell response to a nonglucose secretagogue was performed by elevating the glucose levels to 25–30 mmol/L using a variable rate iv infusion of 10% dextrose. At 45 min, 5 g arginine (Ajinomoto, Los Angeles, CA) were administered iv. Samples for insulin measurements were taken at 2, 3, 4, and 5 min after arginine injection. The glucose level of 25–30 mmol/L was chosen to obtain the maximal glycemic potentiation of the insulin response to arginine (AIRmax), a measure of B cell secretory capacity (28). AIRmax was calculated as the mean increment (2–5 min) above the mean of two prestimulus values.
Assays
Plasma glucose concentrations were measured using a glucose oxidase method (Beckman Coulter, Inc., Palo Alto, CA). Plasma PI and insulin were measured by RIAs (29, 30). The PI RIA measures both intact PI as well as PI conversion intermediates with 100% efficiency (29) and has intra- and interassay coefficients of variation of 10% and 14%, respectively. The insulin RIA detects all insulin-like molecules, as the antibody used in this assay cross-reacts 100% with intact PI and its conversion intermediates (29, 30). The assay has intra- and interassay coefficients of variation of 5% and 8%, respectively. Thus, IRI refers to both insulin and PI-like molecules.
Calculations and statistics
All data are expressed as the mean ± SEM. Comparisons were performed using Student’s two-tailed t tests for unpaired data. The coefficient of determination (r2) represents the proportion of the total variation in the dependent variable that is explained by the fitted regression. Based on our previous observation (31), a hyperbolic relationship between SI and B cell function was assumed, such that the product of these variables is a constant. This hyperbolic relationship is nonlinear, heteroscedastic, and non-Gaussian. To allow analysis by nonweighted linear least squares regression, SI and B cell function measures were log transformed. Comparisons of the SI/B cell relationship between young and old groups was made using unpaired t tests on the y-intercept mean and SD. This analysis was also performed by bootstrap modeling using 100,000 iterations, and results were the same for all comparisons. Finally, a B cell function index that is adjusted for insulin sensitivity was obtained by calculating SI x fasting IRI, SI x AIRglucose, and SI x AIRmax. For all comparisons, P < 0.05 was considered significant.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
. The findings illustrate that the older cohort had characteristics
typical of this age group. Thus, the SI was
significantly reduced in the older subjects, being approximately 50%
that in the younger subjects (P < 0.001). However, B
cell function measured as fasting IRI, AIRglucose, or AIRmax did not
increase appropriately in the older subjects to compensate for the
degree of insulin resistance. This lack of adequate compensation was
highlighted by the significantly lower products between
SI and fasting IRI (43% lower), AIRglucose
(56% lower), and AIRmax (47% lower) in the older than in the younger
subjects (all P < 0.001). In accordance with these
observations, the glucose disappearance rate (Kg)
was significantly decreased in the older subjects (P <
0.001). In keeping with our previous work (31), we found a highly significant relationship in the younger subjects between SI and AIRglucose, which was described by a hyperbolic function (r = -0.83; P < 0.001). A similar finding was made when AIRmax was evaluated relative to SI in the younger subjects (r = -0.66; P < 0.001). In the older subjects, the relationships were not statistically significant.
To determine whether aging is associated with a disproportionate
increase in PI relative to IRI, we measured fasting PI and IRI levels
and calculated the PI/IRI ratio. The individual values of fasting PI,
IRI, and the PI/IRI ratio are shown in Fig. 1, A–C. No significant differences were
observed between the two subject groups in any of these parameters.
Fasting PI, IRI, and PI/IRI ratio in the older compared to the younger
subjects were 8.9 ± 0.8 vs. 10.6 ± 2.0 pmol/L,
83.4 ± 8.1 vs. 75.5 ± 9.3 pmol/L, and 12.3
± 1.3% vs. 13.9 ± 1.6%, respectively (all
P = NS).
|
, A and B,
when we examined the relationship between insulin sensitivity and
fasting PI and IRI levels we found these to be significantly
correlated. As we have previously reported for IRI in individuals under
the age of 45 yr (31), a hyperbolic fit was significant for
SI and fasting IRI in the young and the older
subjects (Fig. 2A). A similar observation was made for the relationship
between SI and fasting PI (Fig. 2B). The degree
of correlation between SI and fasting PI was
similar to that of SI and fasting IRI (r =
-0.47 vs. r = -0.51). The nonlinear nature of this
relationship is supported by a greater hyperbolic than linear
coefficient of determination (young subjects: linear,
r2 = 0.29; hyperbolic, r2 =
0.40). The hyperbolic fits for SI and fasting IRI
and PI were significantly different between the younger and older
subjects (P < 0.001; Fig. 2, A and B). In the younger
subjects the product of SI and IRI was 244
± 24 x 10-5, whereas in the older
subjects SI x IRI was significantly lower at
139 ± 18 x 10-5 (P
< 0.001). The same was true for the product between
SI and PI in younger subjects
(SI x PI: young subjects, 32.5 ± 4.9
x 10-5; older subjects, 14.8 ± 1.2
x 10-5; P < 0.001). Thus, at a
given SI, the fasting IRI and PI levels would be
expected to be lower in the older than in the young subjects.
|
). Furthermore, whether
analyzed in the two groups separately or in all subjects together, no
significant relationships were found between fasting PI/IRI and
AIRglucose (young: r = -0.23; P = 0.3; old:
r = -0.18; P = 0.3), AIRmax (young: r =
0.21; P = 0.3; old: r = -0.26; P
= 0.2), SI x AIRglucose (young: r = -0.27;
P = 0.2; old: r = -0.14; P =
0.4), or SI x AIRmax (young: r = 0.05;
P = 0.8; old: r = -0.14; P =
0.5). |
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Our findings differ from those reported by Duckworth et al. (34), who used an insulin protease method to separate PI from insulin immunoreactivity and found an aging-associated increase in the percentage of the PI-like component after oral glucose ingestion. The observation could be due to clearance differences (35) or methodological difficulties. It may also be explained by a greater impairment of carbohydrate metabolism in the elderly in that study or that some of their subjects were, in fact, prediabetic. In a more recent study of older Japanese subjects, elevated PI and PI/IRI ratios were found (36) despite normal glucose tolerance. A difference in PI levels between different ethnic groups (Asian and Caucasian) with normoglycemia has previously been observed (37) and might explain the difference from our study. Another difference between our study and that in the Japanese subjects is that the PI assay used in the Japanese study (36), in contrast to that which we used (29), does not measure des(31, 32)-PI, which is known to be a major contributor to plasma PI immunoreactivity (38, 39).
In the present study we observed that B cell function is reduced in older subjects when the degree of insulin sensitivity is taken into account (decreased SI x AIRglucose and SI x AIRmax). This confirms other reports that the compensatory ability of B cell function is reduced in older subjects (7, 8) and is in keeping with our recent observation that the release of the B cell peptide amylin is reduced in older subjects (40).
In order that our older subjects would be representative of typical nondiabetic subjects in this age category, we examined a cohort selected that did not meet the fasting glucose criteria for diabetes. Thus, it is possible that some subjects would have had impaired glucose tolerance or even mild diabetes based on the 120 min glucose value on an oral glucose tolerance test, and a proportion may have subsequently progressed to clinical diabetes. These subjects may be those with the highest PI/IRI ratios, but as we did not determine oral glucose tolerance status or follow the subjects longitudinally, we were unable to make this determination. However, we did not observe a relationship between either fasting PI or the PI/IRI ratio and iv glucose tolerance measured as Kg (data not shown), suggesting that glucose tolerance was not a determinant of these measures in this cohort.
The product of SI and IRI was significantly lower in the older subjects than in the young, and the hyperbolic fit for the relation between SI and IRI was significantly different between the younger and older subjects. Although it is well known that insulin sensitivity and fasting insulin are inversely related, and this relationship is hyperbolic in nature (31), the present data suggest that although aging does not affect the inverse nature of this relationship, it appears to change the quantitative interaction. Thus, although compensation of fasting IRI does occur in older subjects, the magnitude of this effect on the B cell differs between old and young subjects. Data from studies by others (5, 36, 41) also support an independent effect of age on fasting insulin levels, as fasting IRI in those studies was similar in young and older subjects despite the fact that the older subjects were more insulin resistant (5, 6, 7). This observation of a change in the quantitative nature of the relationship between insulin sensitivity and fasting insulin strongly suggests that caution must be employed when interpreting insulin levels as a surrogate measure of insulin resistance in groups of subjects who differ by age.
As observed with IRI, SI and fasting PI were inversely correlated, compatible with PI levels increasing in compensation for a reduction in insulin sensitivity. This relationship between SI and PI was also hyperbolic in nature. Thus, fasting PI levels seem to reflect the degree of insulin resistance in normoglycemic subjects. This observation is in accord with a recent study demonstrating a relationship between insulin-mediated glucose disposal (measured as the steady state plasma glucose concentration during a somatostatin, insulin, and glucose infusion) and PI in middle-aged subjects with normal and impaired glucose tolerances (16). It is also worth noting that this increase in fasting PI with insulin resistance is in keeping with our previous observations in Pima Indians and Japanese Americans that PI levels increase with increasing obesity (13, 15). In our study of Japanese Americans, we found this relationship between fasting PI levels and obesity whether the latter was determined as body mass index or as intraabdominal fat area. Although body fat distribution is a known determinant of insulin sensitivity, and the degree of central adiposity increases with age (42), we do not believe our findings would have differed if we had selected subjects for the present study based on intraabdominal fat area rather than body mass index.
In contrast to the observation of a hyperbolic relationship between SI and PI, we found no correlation between SI and the PI/IRI ratio (even when younger and older subjects were analyzed separately), suggesting that the processing of proinsulin to insulin is not affected by the variations in insulin demand. Further, we found that the PI/IRI ratio was not different in the older subjects when they were insulin resistant. This observation of no change in the PI/IRI ratio with insulin resistance is in keeping with our previous work and that of others suggesting that increased B cell secretory demand alone is not sufficient to produce disproportionately elevated PI levels. In two of our previous reports, we found a slight decrease in this measure when nicotinic acid was used to induce experimental insulin resistance (43, 44). An unchanged PI/IRI ratio has also been observed in insulin-resistant nonobese subjects (16), in subjects with the severe congenital insulin resistance syndrome (45), and in obese subjects with moderate insulin resistance (17). In keeping with these observations regarding the effect of insulin resistance, a study of a large cohort of subjects observed a weak positive correlation between SI and PI/IRI in middle-aged individuals (46). In contrast, this same group of investigators reported an increased PI/IRI associated with features of the insulin resistance syndrome (hypertension, low high density lipoprotein cholesterol, high triglycerides, and impaired glucose tolerance), but in this instance they did not measure insulin sensitivity directly so it is unclear whether this measure differed (47). If these subjects were,in fact, more insulin resistant, it is possible that the apparent difference in the findings in the two reports by these investigators may be the result of differences in the population samples.
In conclusion, insulin resistance and reduced compensation in B cell function was observed in this group of older nonhyperglycemic subjects, but disproportionate hyperproinsulinemia, as found in type 2 diabetes, was not observed. Furthermore, the fasting insulin level was not a comparable index of insulin sensitivity when this group of older subjects was compared with younger subjects with comparable fasting glucose levels and body adiposity.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received August 3, 1999.
Revised January 19, 2000.
Accepted February 15, 2000.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  J. M. Burns, J. E. Donnelly, H. S. Anderson, M. S. Mayo, L. Spencer-Gardner, G. Thomas, B. B. Cronk, Z. Haddad, D. Klima, D. Hansen, et al. Peripheral insulin and brain structure in early Alzheimer disease Neurology, September 11, 2007; 69(11): 1094 - 1104. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. R. Gambert and S. Pinkstaff Emerging Epidemic: Diabetes in Older Adults: Demography, Economic Impact, and Pathophysiology Diabetes Spectr, October 1, 2006; 19(4): 221 - 228. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Maedler, D. M. Schumann, F. Schulthess, J. Oberholzer, D. Bosco, T. Berney, and M. Y. Donath Aging Correlates With Decreased {beta}-Cell Proliferative Capacity and Enhanced Sensitivity to Apoptosis: A Potential Role for Fas and Pancreatic Duodenal Homeobox-1 Diabetes, September 1, 2006; 55(9): 2455 - 2462. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Alonso, R. Fernandez, M. Moreno, P. Ordonez, H. Gonzalez-Pardo, N. M. Conejo, F. Diaz, and C. Gonzalez Positive Effects of 17{beta}-Estradiol on Insulin Sensitivity in Aged Ovariectomized Female Rats. J. Gerontol. A Biol. Sci. Med. Sci., May 1, 2006; 61(5): 419 - 426. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. A. Smith, L. E. Porter, N. Biswas, and M. I. Freed Rosiglitazone, But Not Glyburide, Reduces Circulating Proinsulin and the Proinsulin:Insulin Ratio in Type 2 Diabetes J. Clin. Endocrinol. Metab., December 1, 2004; 89(12): 6048 - 6053. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. E. Roder and S. E. Kahn Suppression of Beta-Cell Secretion by Somatostatin Does Not Fully Reverse the Disproportionate Proinsulinemia of Type 2 Diabetes Diabetes, December 1, 2004; 53(suppl_3): S22 - S25. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. M. Utzschneider, D. B. Carr, R. L. Hull, K. Kodama, J. B. Shofer, B. M. Retzlaff, R. H. Knopp, and S. E. Kahn Impact of Intra-Abdominal Fat and Age on Insulin Sensitivity and {beta}-Cell Function Diabetes, November 1, 2004; 53(11): 2867 - 2872. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. M. Utzschneider, D. B. Carr, S. M. Barsness, S. E. Kahn, and R. S. Schwartz Diet-Induced Weight Loss Is Associated with an Improvement in {beta}-Cell Function in Older Men J. Clin. Endocrinol. Metab., June 1, 2004; 89(6): 2704 - 2710. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. Bacha, R. Saad, N. Gungor, and S. A. Arslanian Adiponectin in Youth: Response to Stefan et al. Diabetes Care, June 1, 2004; 27(6): 1520 - 1521. [Full Text]
|
|
|
|
|
|
N. Stefan, M. Stumvoll, H.-U. Haring, and A. Fritsche Adiponectin in Youth: Response to Bacha et al. Diabetes Care, June 1, 2004; 27(6): 1519 - 1520. [Full Text]
|
|
|
|
 |
|
 M. E. Doyle and J. M. Egan Pharmacological Agents That Directly Modulate Insulin Secretion Pharmacol. Rev., March 1, 2003; 55(1): 105 - 131. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. M. Chang and J. B. Halter Aging and insulin secretion Am J Physiol Endocrinol Metab, January 1, 2003; 284(1): E7 - E12. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Fritsche, A. Madaus, N. Stefan, O. Tschritter, E. Maerker, A. Teigeler, H. Haring, and M. Stumvoll Relationships Among Age, Proinsulin Conversion, and {beta}-Cell Function in Nondiabetic Humans Diabetes, February 1, 2002; 51(90001): S234 - 239. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
) subjects. The
relationships shown are nonlinear functions [young: r = -0.70;
P < 0.001 (solid line); old: r
= -0.45; P < 0.05 (dashed line)].
The fitted functions for young vs. old subjects are
significantly different (P < 0.001) based on an
analysis of the y-intercept mean and SD
after log transformation of the data. B, Fasting PI concentrations as a
function of the SI in younger (•) and older (