| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
From the Clinical Research Centers |
Departments of Pediatrics, Internal Medicine, and the General Clinical Research Center (N.A., S.C., T.W.J., R.H., D.S., R.S.S., W.V.T.), Yale University School of Medicine, New Haven, Connecticut 06510; and the Lilly Research Laboratories (J.H.), Eli Lilly & Co., Indianapolis, Indiana 46202
Address correspondence and requests for reprints: Dr. W. V. Tamborlane, Section of Pediatric Endocrinology, Yale University School of Medicine, 333 Cedar Street, New Haven, Connecticut 06510. E-mail: william.tamborlane{at}yale.edu
| Abstract |
|---|
|
|
|---|
Results: After interruption of the basal insulin infusion, plasma free insulin levels fell from 60 ± 12.0 pmol/L to 10.8 ± 4.2 pmol/L, and plasma glucose rose from 5.6 ± 0.4 mmol/L to 14.8 ± 1.2 mmol/L (P < 0.01). During insulin withdrawal, IGFBP-1 increased by more than 6-fold (from 32 ± 8 to 205 ± 17 ng/mL, P < 0.001), IGFBP-3 increased significantly (from 2631 ± 118 to 3053 ± 101 ng/mL, P < 0.001), and total IGF-1 levels declined modestly (from 226 ± 33 to 182 ± 26 ng/mL, P < 0.001). In contrast, free IGF-1 concentrations (0.72 ± 0.22 ng/mL at baseline) were markedly suppressed during insulin withdrawal to values below the detection limit of the assay (0.08 ng/mL) in 15 of the 17 patients (P < 0.001). Circulating plasma leptin declined markedly in females from 20 ± 3 ng/mL to 11 ± 2 ng/mL (P < 0.0001) and in males from 10 ± 2 ng/mL to 7 ± 2 ng/mL (P < 0.02). Within 2 h of insulin replacement, the changes in circulating concentrations of IGFBP-1 and IGFBP-3 were partially reversed, and free IGF-1 levels rebounded to 0.54 ± 0.22 ng/mL (P < 0.1 vs. insulin withdrawal). Growth hormone, glucagon, and IGFBP-2 levels did not change significantly throughout the study. Despite the rapid restoration of plasma insulin and substrate levels, circulating leptin levels continued to fall in the 2-h period after insulin replacement in both females and males. The marked reduction in circulating free IGF-1 after insulin withdrawal and its increase after insulin administration suggest that acute changes in IGFBP concentrations induced by insulin are important regulators of IGF-1 bioavailability in patients with type 1 diabetes. In both males and females, the rapid induction of severe insulin deficiency is associated with a consistent fall in plasma leptin levels.
| Introduction |
|---|
|
|
|---|
The influence of insulin on leptin production has also been extensively studied. In vitro studies in rodents have shown that insulin enhances leptin gene transcription at physiological concentrations (16). Conversely, insulin deficiency provoked by streptozotocin down-regulates leptin messenger RNA, and this suppression is rapidly reversed by insulin (17). In contrast to the data in rodents, in vivo studies in humans have not found any direct short-term effects of insulin on circulating leptin concentrations (18, 19, 20, 21, 22, 23). Boden et al. (24) reported, however, that 72 h of hyperinsulinemia stimulated leptin release in humans. Moreover, a substantial fall in plasma leptin levels in healthy human subjects was observed after only 24 h of fasting, a phenomenon that is reversed by refeeding (25). Because circulating insulin is known to fall during starvation, it has been suggested that insulin might mediate these effects on circulating leptin. If this hypothesis is true, insulins acute effects on leptin levels may become even more evident in conditions of severe insulin deficiency.
Severe insulin deficiency is the hallmark of diabetic ketoacidosis; however, increments in other hormones, such as growth hormone, glucagon, and catecholamines, facilitate the profound disturbances in glucose, lipid, and protein metabolism accompanying the disorder. Whether changes in the concentration of plasma free IGF-1 and leptin occur with sufficient rapidity to also have an impact on the rate of metabolic decompensation has not been investigated. To address this issue, we used the paradigm of acute insulin withdrawal to assess the effects of acute insulin deficiency and replacement on two separate systems: 1) the IGFBP axis, and 2) circulating leptin levels.
| Subjects and Methods |
|---|
|
|
|---|
We studied 17 patients with type 1 diabetes who were being treated with continuous subcutaneous insulin infusion (CSII). Clinical characteristics of the patients were previously reported elsewhere to compare the effects of discontinuation of CSII between patients treated with regular and lispro insulin (26). Nine patients (5 males/4 females) were receiving buffered regular insulin (Velosulin, Novo Nordisk, Bagsvaerd, Denmark), and 8 (4 males/4 females) were treated with insulin lispro (Humalog, Eli Lilly & Co., Indianapolis, IN). Their mean age was 31 ± 10 yr (range 1551 yr), they had an HbA1c of 7.6% (range 6.19.5%; nondiabetic range 4.36.1%). The mean body mass index (BMI) in the females with IDDM was 24 ± 1 kg/m2 (range 16.728.3 kg/m2) and in the males 29.1 kg/m2 (range 2335 kg/m2). The increased BMI in some of the men was responsible for the elevated plasma leptin levels observed in the present study (10 ± 1 ng/mL).
None of the patients had evidence of residual insulin secretion as assessed by measurement of fasting C-peptide (<0.2 pmol/L). Subjects were excluded if they had evidence of impaired renal function, macrovascular complications of diabetes, or other chronic disease.
The protocol was approved by the Yale Human Investigation Committee, and each subject gave written informed consent to the study.
Protocol
The study protocol was divided into two phases: a period of acute insulin withdrawal of up to 8 h, followed by a further 2-h period of acute insulin replacement. All subjects were admitted to the Yale Clinical Research Center in the afternoon of the first day of the study. An iv catheter was inserted into an antecubital vein for blood sampling. The regimen of CSII was continued at the usual basal rate, with the usual presupper and bedtime insulin boluses to achieve a plasma glucose of 60150 mg/dL immediately before the interruption of CSII. The basal insulin infusion was stopped at 3 a.m. In 12 subjects, the insulin withdrawal phase of the study was terminated after 8 h or earlier if either of the following was noted: plasma glucose was greater than 350 mg/dL or urine dipstick showed moderate ketonuria.
After the period of insulin withdrawal, a single injection of either regular or insulin lispro (randomly allocated irrespective of usual insulin therapy) was given subcutaneously in a dose of 0.2 units/kg body weight into the subcutaneous tissue of the anterior abdominal wall. After this injection, blood sampling was continued for an additional 2 h.
Blood samples for the measurement of glucose, insulin, IGF-1, glucagon, free IGFBP-1 and -3, and growth hormone, epinephrine and norepinephrine were obtained every 30 min. In the diabetic subjects, samples for measurement of plasma leptin and free fatty acid (FFA) were obtained at baseline, at 360 min, and at the end of the insulin withdrawal phase, and hourly after the subcutaneous insulin injection.
Measurement and analysis
Plasma glucose levels were measured by the glucose oxidase method with a Beckman Coulter, Inc. glucose analyzer (Beckman Coulter, Inc., Brea, CA). Plasma total IGF-1 was measured by acid-ethanol precipitation (Nichols Institute Diagnostics, San Juan Capistrano, CA). Plasma free IGF-1, IGFBP-1, -2, and -3 were measured by a two-site immunoradiometric assay (Diagnostic Systems Laboratories, Webster, TX), as described by Takada et al. (27). Recently, there has been a great interest in the measurement of free IGF-1, which, theoretically, is the biologically active fraction. Various methods have been used to measure the free (or freely dissociated) IGF fraction. We used a two-site immunoradiometric assay kit (Diagnostic Systems Laboratories) that is highly sensitive and is used as a direct assay to measure free IGF-1 fraction. As described in detail by Juul et al. (28), this immunoradiometric assay is a noncompetitive assay in which the analyte is sandwiched between two antibodies. Free IGF-1 values below the detection limit of the assay (<0.08 ng/mL) were assigned a value of 0.0. The intraassay coefficients of variation were 10.6% for IGFBP-1, and 6.1% for IGFBP-3; the interassay coefficients of variation were 8% for free IGF-1 10% for IGFBP-1 and 16% for IGFBP-3. Plasma leptin levels were measured in duplicate using a double antibody radioimmunoassay (Human Leptin RIA kit, Linco Research, Inc., St. Charles, MO). The limit of sensitivity for human leptin assay was 0.5 ng/mL, the intra- and interassay assay coefficients of variation were 6% and 7%, respectively.
Plasma FFA was assayed by a colorimetric method. Insulin, GH, and glucagon were measured by double antibody RIA. Plasma catecholamines were determined by radioenzymatic method (Upjohn, Kalamazoo, MI). Except where noted, all data are expressed as mean ± SE. Changes between baseline and withdrawal values and between withdrawal and replacement values of plasma glucose, insulin, IGF-1, BP-13, free IGF-1, FFA, and plasma leptin were analyzed by repeated measured analysis of variance (ANOVA), performed with a single factor to compare responses over time (Systat v.5, SPSS, Inc., Chicago, IL). A paired t test was used to localize effects found in the initial set of repeated measures analysis. Whereas the type of insulin being administered or withdrawn did not influence the magnitude or direction of changes in IGF-1, free IGF-1, IGFBP-13, leptin, ketone, and FFA levels during either phase of the study, data from all subjects were combined for the analyses in this paper. However, as leptin levels are higher in females than males, changes in leptin concentration during the study are presented according to gender.
| Results |
|---|
|
|
|---|
After the interruption of subcutaneous insulin infusion (CSII),
insulin levels steadily decreased to a nadir concentration of 12 pmol/L
(1.8 ± 0.4 µU/mL). As a consequence of the fall in plasma
insulin, plasma glucose increased gradually from 5.5 ± 1
mM to 15 ± 1 mM at the end of the insulin
withdrawal phase. Similarly, plasma ß-hydroxybutyrate levels rose
from 0.09 ± 0.03 to 0.97 ± 0.07 mmol/L, and FFA levels rose
from 337 ± 45 to 1439 ± 133 µM,
P < 0.001, as indicated in Table 1
.
|
Serum GH concentrations fell slightly (but not significantly) during
both phases of the study (Table 1
). Plasma glucagon and epinephrine
levels did not change during either phase of the study, only plasma
norepinephrine increased significantly (by 40%) during the insulin
withdrawal phase (P < 0.05) (Table 1
).
Changes in the IGF-1/IGFBP system during insulin withdrawal and replacement
IGFBP-2 levels (560 ± 45 ng/mL at baseline) did not change
significantly during insulin withdrawal (591 ± 40 ng/mL) or
insulin replacement (583 ± 44 ng/mL). In contrast, as shown in
Fig. 1
, IGFBP-1 levels rose
markedly during insulin withdrawal to peak values that were more than
6-fold higher than basal levels (from 32 ± 8 to 205 ± 17
ng/mL, P < 0.001). Even though GH levels did not
change, IGFBP-3 also increased from 2631 ± 118 to 3053 ±
101 ng/mL (P < 0.001) during insulin withdrawal. The
rise in IGFBPs during insulin withdrawal was accompanied by a modest
decline in total IGF-1 levels (from 226 ± 33 to 182 ± 26
ng/mL, P < 0.01) and a marked suppression of
circulating free IGF-1 levels (P < 0.001) (Fig. 1
). Free IGF-1 levels, which averaged
0.72 ± 0.22 ng/mL at baseline, fell to values below the level of
detection of the assay in 15 of the 17 subjects at the end of the
insulin withdrawal phase. Free IGF-1 values were also markedly
suppressed in the other 2 subjects (from 2.54 to 0.21 ng/mL and from
1.36 to 0.31 ng/mL), who were adolescent girls.
|
Leptin levels during insulin withdrawal and replacement
Changes in plasma leptin concentrations in females and males are
illustrated in Fig. 2
. Although at the start of the study basal plasma
leptin levels were significantly higher in females than males
(P < 0.001), during insulin withdrawal, plasma leptin
levels declined markedly in both groups. At the end of the insulin
withdrawal, plasma leptin fell by 40% in females and 30% in the
males, reaching a mean of 11 ± 2 ng/mL (P <
0.0001 vs. basal) in the females and in males a mean of
7 ± 2 ng/mL in the males (P < 0.02
vs. basal). Despite the rapid restoration of plasma insulin
and substrate levels, circulating leptin levels continued to fall in
the 2-h period after insulin replacement in both females and males.
|
| Discussion |
|---|
|
|
|---|
Although the changes in total IGF-1 levels were modest during both phases of the study, circulating free IGF-1 (the biologically active form of the peptide) fell to virtually undetectable levels during insulin withdrawal in association with the increases in IGFBP-1 and IGFBP-3. Similarly, after insulin replacement, free IGF-1 concentrations rebounded rapidly toward basal values. The fact that free IGF-1 rebounded to near normal levels with only a slight decrease in IGFBP-1 suggests that IGF-1 was entering the circulation. These observations support the contention that acute changes in IGFBP concentrations induced by insulin are important regulators of IGF-1 bioavailability. In patients with diabetes, increased affinity of IGFBP-1 for IGF-1 may also contribute to reduced bioavailability of free IGF-1 (28, 29).
The hyperglycemic and ketogenic effects of GH, glucagon, and catecholamines in combination with insulin deficiency have long been implicated in accelerating the development of diabetic ketoacidosis. Nevertheless, the acute interruption of the basal insulin infusion in our patients led to a prompt rise in plasma glucose and ß-hydroxybutyrate levels in the absence of changes in GH, glucagon, or epinephrine concentrations and a only modest increase in plasma norepinephrine levels. In contrast, there were striking and rapid alterations in free IGF-1 levels after insulin withdrawal. Thus, severe free IGF-1 deficiency may have magnified the early phases of metabolic decompensation caused by insulin withdrawal in our patients. The lack of changes in serum GH levels despite marked decreases in the free IGF-1 concentrations are of interest. It is possible that the kinetics of the well-known IGF-1-GH feedback axis is sluggish and cannot quickly adjust to the acute changes observed in this short term study.
Unlike IGF-1, much less is known about the regulation of circulating leptin concentrations in humans. Insulin does appear to act as a modulator of leptin concentration. However, previous studies suggested that the effect of insulin was not immediate. In humans, prolonged (72 h) hyperinsulinemia was required to increase leptin levels in a dose-dependent fashion (24), whereas no acute effect was seen (23). Plasma leptin levels also respond slowly to fasting in nondiabetic subjects and start to decrease only after 12 h (25). In the current study, we employed a model of accelerated starvation, namely CSII-treated, C-peptide negative subjects with type 1 diabetes who were withdrawn from insulin. As expected, basal leptin concentrations were higher in females than males (23). Nevertheless, in both males and females, the rapid induction of severe insulin deficiency was associated with a consistent acute fall in plasma leptin levels. These changes in leptin concentrations are similar to those described during prolonged fasting in animals and humans and could not be accounted for by normal diurnal fluctuations of plasma leptin as there is little variation in leptin concentrations overnight in healthy controls (31). Recent studies suggest that IGF-1 may have a direct inhibitory effect on leptin messenger RNA synthesis in cultured rat adipocytes (32). Although the effects of IGF-1 on leptin expression are yet to be confirmed in humans, IGF-1 lowered leptin levels significantly in patients with chronic renal insufficiency (33). This effect may be secondary to the suppressive effects of IGF-1 on insulin secretion. In the presence of insulin deficiency, we found a parallel fall in both IGF-1, free IGF-1, and leptin levels. From this study, it is not possible to establish a connection between these two systems, which are both highly modulated by circulating insulin levels.
Although the fall in plasma leptin levels coincided with that of plasma insulin, insulin deficiency per se may not have directly mediated the suppression of leptin concentrations. Plasma glucose, FFA, and blood ß-hydroxybutyrate levels increased when insulin and leptin levels fell. It is conceivable that these metabolic alterations, especially increased lipolysis, may have provided the signal to decrease leptin levels. Plasma leptin continued to fall during the 2 h after insulin injection, despite the rise in insulin and at least partial correction of the metabolic abnormalities. The duration of this phase of the study was likely too short to see a rebound increase in leptin concentrations.
While the suppression of free IGF-1 may have accelerated the metabolic decompensation observed in our patients, reductions in plasma leptin might serve a protective role. In rodents, administration of leptin increases energy expenditure and fat breakdown and suppresses food intake (34). Moreover, leptin directly alters lipid partitioning in skeletal muscle specifically by increasing muscle fatty oxidation and decreasing incorporation of fatty acid into TAG (35). Consequently, it is intriguing to speculate that the fall in plasma leptin that we observed may have slowed the rate of acceleration of lipolysis and ketogenesis during insulin withdrawal. Moreover, such reductions in plasma leptin might contribute to two of the classic warning symptoms of diabetes, increased appetite and hyperphagia.
| Acknowledgments |
|---|
| Footnotes |
|---|
Received December 30, 1998.
Revised April 1, 1999.
Accepted April 7, 1999.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
C. H. Lang, T. C. Vary, and R. A. Frost Acute in Vivo Elevation of Insulin-Like Growth Factor (IGF) Binding Protein-1 Decreases Plasma Free IGF-I and Muscle Protein Synthesis Endocrinology, September 1, 2003; 144(9): 3922 - 3933. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. S. Sandhu, D. B. Dunger, and E. L. Giovannucci Insulin, Insulin-Like Growth Factor-I (IGF-I), IGF Binding Proteins, Their Biologic Interactions, and Colorectal Cancer J Natl Cancer Inst, July 3, 2002; 94(13): 972 - 980. [Abstract] [Full Text] [PDF] |
||||
![]() |
L. E. L. Katz, D. D. DeLeon, H. Zhao, and A. F. Jawad Free and Total Insulin-Like Growth Factor (IGF)-I Levels Decline during Fasting: Relationships with Insulin and IGF-Binding Protein-1 J. Clin. Endocrinol. Metab., June 1, 2002; 87(6): 2978 - 2983. [Abstract] [Full Text] [PDF] |
||||
![]() |
P. Stattin, S. Söderberg, G. Hallmans, A. Bylund, R. Kaaks, U.-H. Stenman, A. Bergh, and T. Olsson Leptin Is Associated with Increased Prostate Cancer Risk: A Nested Case-Referent Study J. Clin. Endocrinol. Metab., March 1, 2001; 86(3): 1341 - 1345. [Abstract] [Full Text] |
||||
![]() |
R. Heptulla, A. Smitten, B. Teague, W. V. Tamborlane, Y.-Z. Ma, and S. Caprio Temporal Patterns of Circulating Leptin Levels in Lean and Obese Adolescents: Relationships to Insulin, Growth Hormone, and Free Fatty Acids Rhythmicity J. Clin. Endocrinol. Metab., January 1, 2001; 86(1): 90 - 96. [Abstract] [Full Text] |
||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |