| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Second Department of Internal Medicine (G.D., P.M., V.L., E.B., S.A.R.), Research Institute and Diabetes Center, Athens University Medical School, 12462 Haidari, Greece; Hellenic National Diabetes Center (E.M., S.A.R.), 106 75 Athens, Greece; and Department of Endocrinology, "Elena Venizelou" (E.K.) and "Evangelismos" (M.T., N.T.) Hospitals, Athens, Greece
Address all correspondence and requests for reprints to: George Dimitriadis, M.D., DPhil, Second Department of Internal Medicine, Research Institute and Diabetes Center, Athens University, "Attikon" University Hospital, 1 Rimini Street, GR-12462 Haidari, Greece. E-mail: gdim{at}internet.gr.
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Methods: To investigate this, a meal was given to 12 hyperthyroid (HR) and 10 euthyroid (EU) subjects. Blood was withdrawn for 360 min from veins draining the anterior abdominal sc AD and from the radial artery. Blood flow was measured with 133Xe. Lipoprotein lipase (LPL) was calculated as triglyceride flux across AD, and AD-lipolysis was calculated as glycerol flux minus LPL.
Results: Both groups displayed comparable postprandial glucose levels, with the HR having higher insulin levels than the EU. In AD of HR vs. EU: 1) blood flow was increased [area under curve 0–360 min (milliliters per 100 milliliters of tissue); 1746 ± 208 vs. 1344 ± 102, P = 0.001], but glucose uptake was normal [area under curve 0–360 min (micromoles per 100 milliliters of tissue); 501 ± 114 vs. 368 ± 48]; 2) fasting rates of lipolysis (nanomoles per minute per 100 milliliters of tissue; 329 ± 75 vs. 89 ± 22, P = 0.02) and nonesterified fatty acid (NEFA) release (nanomoles per minute per 100 milliliters of tissue; 841 ± 146 vs. 316 ± 97, P = 0.01), and plasma NEFA levels (micromoles per liter; 623 ± 50 vs. 454 ± 57, P = 0.03) were increased, but were all rapidly suppressed to levels similar to those in EU after the increase in plasma insulin levels after the meal; and 3) LPL was not stimulated by insulin.
Conclusions: In hyperthyroidism, AD lipolysis and glucose uptake are resistant to insulin. The defect in lipolysis is manifested in the fasting state, whereas postprandially this rate is rapidly suppressed to normal. This may relieve tissues from the burden of NEFAs after the meal, thus facilitating muscle glucose disposal by insulin.
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
In the adipose tissue, insulin increases the rates of glucose disposal, increases the expression of lipoprotein lipase (LPL), and decreases the rates of lipolysis. The latter effects contribute to the buffering of lipid fluxes in the circulation, therefore, sparing the rest of the body from elevated levels of nonesterified fatty acids (NEFAs). This favors the increase of glucose use in muscle and the suppression of endogenous glucose production (3, 4). There are no studies to investigate these effects of insulin on adipose tissue in hyperthyroidism and in particular in vivo.
In adipocytes isolated from hyperthyroid (HR) patients and examined in vitro, insulin-stimulated glucose uptake has been found to be either decreased (5) or normal (6). In HR subjects, the activity of LPL has been estimated in plasma after heparin injection and found to be normal (7), increased (8), or decreased (9, 10). Finally, measurements of plasma NEFAs in HR subjects have shown that their levels are elevated but they are readily suppressed to normal even at physiological concentrations of insulin (2, 11, 12, 13, 14, 15), suggesting that the suppression of adipose tissue lipolysis by insulin is not impaired.
This study was undertaken to investigate insulin effects on glucose uptake, lipolysis, LPL action, and NEFA fluxes across the abdominal sc adipose tissue in subjects with hyperthyroidism, after the consumption of a mixed meal. The meal creates a metabolic environment that permits the interaction of insulin and substrates to be investigated under conditions that are as close to physiological as possible.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Twelve newly diagnosed female HR patients were studied before initiation of treatment [age, 37 ± 4 yr; body mass index, 24 ± 1 kg/m2; T3, 334 ± 44 ng/dl (5.13 ± 0.7 nmol/liter); TSH, undetectable] and compared with 10 female euthyroid subjects [EU; age, 35 ± 4 yr; body mass index, 24 ± 0.6 kg/m2; T3, 112 ± 8 ng/dl (1.7 ± 0.1 nmol/liter); TSH, 1.1 ± 0.07 µU/ml]. From the medical history, the duration of hyperthyroidism was estimated to be about 1 month. The study was approved by the hospital ethics committee and subjects gave informed consent.
Experimental protocol
All subjects were admitted to the hospital at 0700 h after an overnight fast and had the radial artery (A) and two veins (V, a contralateral antecubital vein draining deep forearm tissues and a superficial abdominal vein) catheterized, following previously published guidelines (16, 17, 18).
A meal was given (730 Kcal, 50% carbohydrate, 40% fat, 10% protein; fat content was 33 g, of which 35% was saturated, 54% monounsaturated, and 11% polyunsaturated) at least 1 h after catheter insertion and consumed within 20 min (17).
Blood samples were withdrawn from the three sites before the meal (at –30 and 0 min) and at 30- to 60-min intervals for 360 min thereafter for measurements of insulin (Linco Research, St. Charles, MO), glucose (Yellow Springs Instruments, Yellow Springs, OH), glycerol, triglycerides (corrected for free glycerol), and NEFA (Roche Diagnostics, Mannheim, Germany). The NEFA assay was validated by measuring dilutions of palmitate and found to be accurate for concentrations of 50 µmol/liter.
Blood flow (BF) in the adipose tissue was measured with 133Xe (2 MBq dissolved in sterile saline; DuPont, MDS Nordion, Belgium) and in the forearm with mercury strain-gauge plethysmography (Hokanson). Two minutes before taking a sample from the forearm vein, a cuff was inflated around the wrist to occlude hand blood flow; before measurements, a cool fan was used to minimize contamination with superficial blood (16, 17).
Although radiolabeled tracers have been used for determining NEFA fluxes (3), the AV-difference technique (16) was considered adequate for the present study for the following reasons: 1) it can determine the effects of insulin on all major metabolic currencies of adipose tissue, NEFA, triglycerides, and glycerol and allows the assessment of LPL action and rates of adipose tissue lipolysis (19); 2) its apparent disadvantage (application is restricted to one particular depot), does not preclude its use in this study because, in lean females, the depot of visceral adipose tissue is expected to be small and, therefore, the major contributor to the oversupply of NEFAs to the tissues after the meal is the sc fat (20, 21); in hyperthyroidism, the sc adipose tissue of upper and lower body contributes equally to the excessive rate of lipolysis (15).
AV-difference measurements, although ideally undertaken under steady-state conditions (22), have been suggested to be valid also in non-steady-state situations because of the high fractional turnover of the substrates measured (3, 23).
The values obtained from the two preprandial samples were averaged to give a "0 time-value". Because blood flow was used in the calculation of fluxes, the plasma levels (P) of metabolites were converted to whole blood (B) by using fractional hematocrit (Ht): B = P(1 – Ht) for NEFA or triglycerides and B = P(1 – 0.3 Ht) for glucose.
Insulin sensitivity in the pre- and postprandial states was calculated by homeostasis model of assessment (HOMA; measures insulin resistance) and "Matsuda index" (measures insulin sensitivity), respectively.
The net flux of glucose or NEFA was calculated as (A – V)glucose or (V – A)NEFA and multiplied by blood flow to give absolute values. Glucose fractional extraction was calculated as (A – V)glucose/Aglucose (this is independent of blood flow; the results are expressed as the percentage of glucose disposal per minute).
Reesterification of NEFA within the adipose tissue was calculated on the assumption that hydrolysis of each triglyceride molecule releases one molecule of glycerol and three molecules of NEFA as follows: 3[(V – A)glycerol(BF)] – [(V – A)NEFA(BF)].
The net inward flow of NEFA from capillaries to adipose tissue (transcapillary flux) was calculated as [3(A – V)triglycerides – (V – A)NEFA] x (BF); this represents the net movement of NEFA across the capillary wall (negative values indicate outward flow).
The relative rates of LPL and adipose tissue lipolysis were calculated assuming that triglyceride extraction from blood represents LPL action: LPL = (A – V)triglycerides(BF), and adipose tissue lipolysis = [(V – A)glycerol(BF)] – LPL (expressed in glycerol concentration units).
Triglyceride clearance across the adipose tissue was calculated as: [(A – V)triglycerides(BF)]/(Atriglycerides).
Results are presented as mean ± SEM of plasma levels or integrated postprandial responses (areas under curve, AUC, from the start of the meal to 360 min). Repeated measures ANOVA (SPSS, Chertsey, UK) was applied to evaluate differences between groups across time. Grouping variable was considered as fixed and group-to-time interaction was evaluated; no random effects factors were used. Post hoc analysis was applied to test for the group effect on the investigated variables at specific time points (Bonferroni-rule). Differences between time points within groups were tested with paired t test.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Fasting plasma glucose and triglyceride levels were not altered by hyperthyroidism, whereas levels of insulin and NEFAs were increased (P = 0.04) (Fig. 1
).
|
). Postprandial changes in plasma NEFAs showed a similar pattern in both HR and EU: their levels decreased by 60% and 40%, respectively, within 60–90 min, remained suppressed until 180 min, and then gradually returned to baseline [levels at 300 (P = 0.01) and 360 min (P = 0.002) were higher in HR; Fig. 1C]. Postprandial plasma triglycerides increased steadily in both groups, and by 240 min, reached values 1.5- to 2-fold higher than their fasting levels. Thereafter, plasma triglycerides began to decrease in HR, whereas in EU they remained elevated (P = 0.04) (Fig. 1D). The HOMA index was 2.2 ± 0.5 in HR and 1.0 ± 0.2 in EU (P = 0.02) and the Matsuda index 5.2 ± 0.8 in HR and 8.1 ± 0.7 in EU (P = 0.02), suggesting insulin resistance in the fasting and postprandial state, respectively (these go opposite directions because HOMA measures insulin resistance but Matsuda measures insulin sensitivity) (25).
Blood flow
In HR, elevated fasting blood flow rates in the adipose tissue were unaltered by eating. The differences from EU were significant only at baseline (P = 0.01), at 30 min (P = 0.009), and late postprandial periods (P = 0.03) (Fig. 2A).
|
Glucose uptake
At 0 and 30 min after the meal, net glucose uptake by the adipose tissue was higher in HR (P = 0.03), but its total uptake (AUC0–360 501 ± 114 vs. 368 ± 48 µmol/100 ml of tissue in EU) and fractional extraction in the postprandial period (Fig. 2B) were not significantly different between the two groups. Postprandial glucose uptake was dominated by that occurring 0–120 min after the start of eating, during the time when arterial insulin and glucose levels rose rapidly (Fig. 1, A and B).
The net forearm uptake of glucose in HR (AUC0–360 673 ± 143 µmol/100 ml of tissue) and EU (AUC0–360 826 ± 157 µmol/100 ml of tissue) was similar at all time points after meal ingestion. In contrast, fractional glucose uptake after the meal was 50% lower in HR (4.5 ± 0.7% per minute) vs. EU (9 ± 1% per minute, P = 0.01).
Lipid fluxes
In HR, the fasting net release of NEFAs from the adipose tissue was higher (P = 0.04), but was rapidly suppressed to normal after the meal. These rates remained suppressed until 180 min and then progressively increased to baseline values by 360 min (P = 0.03 vs. EU; Fig. 3A).
|
There were no differences in adipose tissue NEFA reesterification rates between HR and EU (AUC0–360 101 ± 27 and 115 ± 28 µmol/100 ml of tissue, respectively).
The net release of glycerol was higher in HR in the fasting state (314 ± 47 vs. 135 ± 35 nmol/min/100 ml of tissue in EU, P = 0.005) but not after the meal (AUC0–360 67 ± 8 vs. 63 ± 12 µmol/100 ml of tissue in EU).
Veno-arterial differences of circulating NEFA and glycerol levels across the adipose tissue were similar (in both fasting and postprandial states) in HR (AUC0–360 25 ± 5 and 15 ± 2 mmol/liter·min for NEFA and glycerol, respectively) and EU (AUC0–360 21 ± 1 and 17 ± 3 mmol/liter·min for NEFA and glycerol, respectively).
The net movement of NEFAs from capillaries to adipocytes (transcapillary flux) is shown in Fig. 3B. In the fasting state, there was a net release from the adipose tissue into the circulation, that was higher in HR than in EU (P = 0.03). Postprandialy, there was a net uptake of NEFAs into the adipose tissue, which was not different between the two groups. After 240 min, NEFAs in HR gradually started to flow out of the adipose tissue into capillaries as in the fasting state (P = 0.0001 for differences at 300–360 min).
Hyperthyroidism increased fasting rates of lipolysis in the adipose tissue (P = 0.02 vs. EU). This was completely suppressed within the first 60 min after the beginning of the meal. After 240 min, adipose tissue lipolysis in HR was gradually increased to levels observed in the fasting state (P = 0.001 for differences from EU at 300–360 min; Fig. 3C).
Hyperthyroidism did not alter fasting LPL action in the adipose tissue; however, this did not rise after the meal as in the EU (P = 0.001 for differences at 240–360 min; Fig. 3D).
Triglyceride clearance across the adipose tissue was lower in HR (AUC0–360 34 ± 10 ml/100 ml of tissue) vs. EU (AUC0–360 140 ± 34 ml/100 ml of tissue, P = 0.01).
In the forearm of HR, calculation of NEFA transcapillary flux showed an increased movement from the capillaries into muscle tissues in the fasting state (1.1 ± 0.3 vs. 0.18 ± 0.08 µmol/min/100 ml of tissue in EU, P = 0.03), but not after the meal (AUC0–360 107 ± 50 vs. 130 ± 24 µmol/100 ml of tissue in EU).
LPL action in the forearm of HR (AUC0–360 23 ± 6 µmol/100 ml of tissue) was not significantly different from that of EU (AUC0–360 40 ± 9 µmol/100 ml of tissue).
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The effect of insulin on blood flow is an important component of its stimulation of glucose uptake (28). The possibility that the increase of blood flow in the HR subjects masked the defect in insulin-stimulated glucose disposal at the tissue level was examined by calculating fractional glucose extraction. This was not impaired in the adipose tissue, but was markedly decreased in the forearm muscles in the presence of hyperinsulinemia. The results support the importance of blood flow in maintaining normal or even increased rates of glucose disposal in peripheral tissues in the HR state, despite the defects in intracellular pathways of insulin-stimulated glucose use (1, 2, 5, 27).
Fasting adipose tissue blood flow (29) in our HR subjects was higher than that in the EU, but there was no further increase after the meal when plasma insulin levels were elevated. Because the effects of insulin on blood flow in this tissue are mediated by sympathetic activation (30), our findings may be explained by reports demonstrating that hyperthyroidism increases sympathetic response (31, 32). This is already maximal in the fasting state (Fig. 2A).
In the EU state, the buffering of NEFA flux by insulin is regulated by the suppression of lipolysis and NEFA release from the adipose tissue via a decrease in the activity of hormone-sensitive lipase (HSL), and the increase of triglyceride clearance through an increase in the activity of LPL and capture of the released NEFAs by adipocytes and muscle (3, 4). Interestingly, a new lipolytic enzyme "adipose triglyceride lipase" was recently described in adipocytes. This catalyzes the initial step in the hydrolysis of stored triglycerides in coordination with HSL (33).
This is the first study to examine changes in plasma triglyceride levels after a meal in hyperthyroidism. The pattern was unusual because fasting levels were not significantly different from those in EU (7, 8, 9, 10), but late postprandial levels were decreased. These changes are consistent with a higher triglyceride turnover (9).
The late postprandial lowering of plasma triglycerides was not secondary to an increased rate of removal by the two major tissues expressing LPL, adipose tissue, and muscle, because postprandial LPL action was low or unchanged in these tissues. However, because LPL was calculated and not measured, an increased sensitivity of this enzyme to VLDL cannot be excluded: in human adipose tissue, thyroxine was shown to increase LPL activity measured in vitro (34). The possibility that increased adipose tissue blood flow was responsible for the late postprandial drop of triglycerides was also unlikely because this can only be achieved through increased LPL action and triglyceride clearance (24, 26), and yet both were blunted in our HR subjects [in these calculations, increased blood flow is taken into consideration (see Subjects and Methods)]. Could increased triglyceride removal by the liver account for the postprandial triglyceride reductions? Experiments in humans have suggested that hyperthyroidism enhances the capacity of the liver for whole particle uptake of the remnants of triglyceride-rich lipoproteins (7), but this may only partly explain our results, because the liver primarily removes remnant particles which are low in triglycerides (35). An alternative possibility might be that the VLDL-triglyceride production by the liver was suppressed in the HR subjects in the late postprandial period. This is suggested by experiments in rats made hyperthyroid, showing that the output of VLDL-triglycerides from the isolated-perfused liver was decreased (36). However, in patients with hyperthyroidism, hepatic VLDL-triglyceride synthesis and release were either increased [owing mostly to an increased delivery of NEFAs to the liver (9, 36)] or normal (10).
Although LPL may contribute to the NEFA pool, the majority of NEFA appearance after a meal derives from lipolysis of stored triglycerides (3, 4, 37). Rates of lipolysis and NEFA release in the adipose tissue of HR subjects were both increased in the fasting and late postprandial state, but were rapidly suppressed to normal shortly after the beginning of the meal (Fig. 3, A and C). The results suggest that hyperthyroidism induces resistance of lipolysis to insulin, which is evident at low (basal) levels of insulin. This rate is rapidly suppressed when insulin is increased after the meal. It is noteworthy that lipolysis is extremely sensitive to insulin; the half-maximal suppression in EU humans in vivo (38, 39) or in isolated adipocytes in vitro (40) ranges from 2–11 µU/ml.
Because the veno-arterial differences of plasma NEFAs across the adipose tissue in the HR subjects were similar in the two groups, the fluctuations in the rates of lipolysis may be due to those in blood flow and not to a decreased sensitivity of HSL and adipose triglyceride lipase to insulin. However, because this suggestion is based on calculated and not measured rates, the establishment that increased blood flow is the mechanism for increasing adipose tissue lipolysis in hyperthyroidism, would require experiments with local manipulation of blood flow (30).
Our results correspond well with observations in adipocytes isolated from HR patients and incubated in vitro. The suppression of glycerol release was less at insulin levels less than 10 µU/ml, but was normal at levels between 10–100 µU/ml (40). Our results also agree with a study in HR patients in vivo, examining insulin effects on glycerol release from the sc adipose tissue, using microdialysis and euglycemic-hyperinsulinemic clamps (15): basic glycerol release from the adipose tissue was higher in the HR than in controls, but was suppressed to the same extent (50%) in both groups when insulin was infused, suggesting that lipolysis was resistant to basal levels of insulin, but responded normally when these levels were increased.
The significance of the changes in lipid fluxes in the HR subjects becomes apparent from the transcapillary flow of NEFA. In the fasting state, due to insulin resistance, there is an increased outflow from the adipose tissue into the capillaries—necessary to stimulate gluconeogenesis (1, 2) and provide NEFAs for oxidation in other tissues (such as muscle, see Results)—which, however, quickly subsides after the meal, to facilitate the disposal of glucose by the insulin-resistant muscle (2, 27, 41). This ensures the preferential use of glucose when available and helps to preserve fat stores. These findings are supported by previous experiments with indirect calorimetry in HR patients showing increased whole-body lipid oxidation in the fasting and late postprandial states, and carbohydrate oxidation shortly after the meal (12, 42).
Excessive postabsorptive mobilization and oxidation of lipids and protein previously have been recognized as a prime abnormality in thyrotoxicosis, required to cover increased energy needs (41, 43); indeed, the adipose tissue of HR subjects was generally more catabolic than that of the EU (transcapillary flux was more negative in the fasting state and only 5 h after the meal).
In conclusion, our study showed that, in hyperthyroidism, adipose tissue lipolysis and glucose uptake are resistant to insulin. The defect in lipolysis is manifested in the fasting state, whereas shortly after the meal, this rate is suppressed to levels similar to those in EU subjects by the elevated concentrations of insulin. These changes may be required to relieve tissues from the burden of NEFA surplus after the meal, thus facilitating glucose disposal by insulin.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Abbreviations: A, Artery; AD, adipose tissue; AUC, area under curve; HOMA, homeostasis model of assessment; HR, hyperthyroid; HSL, hormone-sensitive lipase; LPL, lipoprotein lipase; NEFA, nonesterified fatty acid; V, vein.
Received May 3, 2005.
Accepted December 19, 2005.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  G. Dimitriadis, E. Maratou, E. Boutati, A. Kollias, K. Tsegka, M. Alevizaki, M. Peppa, S. A Raptis, and D. J Hadjidakis IGF-I increases the recruitment of GLUT4 and GLUT3 glucose transporters on cell surface in hyperthyroidism Eur. J. Endocrinol., March 1, 2008; 158(3): 361 - 366. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. Ortega, J. Koska, N. Pannacciulli, J. C Bunt, and J. Krakoff Free triiodothyronine plasma concentrations are positively associated with insulin secretion in euthyroid individuals Eur. J. Endocrinol., February 1, 2008; 158(2): 217 - 221. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Dimitriadis, P. Mitrou, V. Lambadiari, E. Boutati, E. Maratou, D. B. Panagiotakos, E. Koukkou, M. Tzanela, N. Thalassinos, and S. A. Raptis Insulin Action in Adipose Tissue and Muscle in Hypothyroidism J. Clin. Endocrinol. Metab., December 1, 2006; 91(12): 4930 - 4937. [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 |
