Insulin Metabolism in Hypothyroidism Jayendra H. Shah, M.B., B.S., M.C.P.S., George S. Motto, M.D., Elaine Papagiannes, M.S., and Gerald A. Williams, M.D., Chicago, Illinois SUMMARY Insulin resistance has been invoked to explain the glucose intolerance obeserved in hypothyroid patients. This possibility was studied by determining fractional and metabolic clearances of intravenously administered porcine crystalline insulin (0.1 U./kg.) and its effect on plasma glucose concentration in ten hypothyroid patients, ten normal subjects, and six treated euthyroid patients. Following administration of porcine insulin, serum immunoreactive insulin concentrations during the period of observation were similar in hypothyroid patients, in normal control subjects, and in treated euthyroid patients. Similarly, no significant differences in the mean half-life, distribution space, or fractional and metabolic clearances of insulin were observed among any of the three groups. In response to insulin administration, plasma glucose concentrations declined to the nadir of 36 ± 4, 43 ± 3, and 38 ± 4 mg. per 100 ml. in hypothyroid patients, normal control subjects, and treated euthyroid patients, respectively. Thereafter, plasma glucose steadily increased and approached the baseline value at ninety minutes in normal subjects and treated euthyroid patients. In contrast, the plasma glucose values remained significantly lower than the baseline for the rest of the procedure in hypothyroid patients. The present study demonstrates that there is no evidence of resistance to the action of insulin in hypothyroid patients. The observation of prolonged hypoglycemic action of exogenously administered insulin in hypothyroid patients might in fact suggest increased sensitivity to insulin action. These findings indicate that glucose intolerance of the hypothyroid state is not characterized by insulin resistance. DIABETES 24:922-25, October, 1975.
The frequent occurrence of abnormal carbohydrate tolerance in hypothyroid patients has been reported. 1 ' 2 Insulin secretion in response to oral3 or intravenous 2 glucose stimulus has been reported to be elevated in hypothyroid subjects, with subsequent decline of insulin values on restoration of the euthyroid state. 3 This observation of elevated levels of insulin has resulted in a hypothesis of insulin resistance to explain glucose intolerance of the hypothyroid state. 2 From the Section of Endocrinology and Metabolism, Departments of Medicine and Nuclear Medicine, Veterans Administration West Side Hospital and University of Illinois College of Medicine, Chicago, Illinois. Address reprint requests to Dr. Shah, VA West Side Hospital (M.P. 172), P.O. Box 8195, Chicago, Illinois 60680. Accepted for publication July 9, 1975. 922
In contrast to this hypothesis of insulin resistance, our previous study 4 clearly demonstrated diminished early insulin secretory response to intravenous glucose in hypothyroid patients. Following treatment and attainment of the euthyroid state, the same patients portrayed a significantly increased insulin response associated with a marked improvement in the glucose tolerance. Although this study strongly suggested the presence of hypoinsulinemia in the hypothyroid state, the possibility of coexisting insulin resistance could not be excluded entirely. The present study was undertaken to examine the possibility of insulin resistance in hypothyroid patients. In the experiments to be described, insulin resistance was determined from fractional and metabolic clearances of intravenously administered crystalline porcine insulin and its effects on peripheral plasma glucose levels. SUBJECTS AND METHODS Three groups of subjects were studied: 1. Ten hypothyroid patients, eight male and two female, were investigated soon after establishing the diagnosis of primary hypothyroidism. Clinical diagnosis of hypothyroidism was substantiated by the laboratory data, such as decrease in serum thyroxin, triiodothyronine resin uptake, and radioactive iodine uptake of the thyroid gland. None of the patients was known to have diabetes or any other complicating disease. Only one patient had a positive family history of diabetes. Three of these ten hypothyroid patients were greater than 10 per cent above ideal body weight (Metropolitan Life Insurance tables). 2. The second group was comprised of ten normal subjects without clinical or chemical evidence of hypothyroidism. These subjects were of age and weight comparable to the hypothyroid patients so that three normal subjects were also greater than 10 per cent above ideal body weight. None of these subjects had a personal or a family history of diabetes or other known illness at the time of the study. 3. Six hypothyroid patients from group 1 underDIABETES, VOL. 2 4 , NO. 10
JAYENDRA H. SHAH, M.B., B.S., M.C.P.S., AND ASSOCIATES
went repeat studies following thyroid hormone replacement therapy and approximately six to eight months after attaining the euthyroid state. All subjects were on a regular diet and consumed at least 250 gm. of carbohydrate daily for three to five days prior to each test. Studies were performed in a quiet procedure room between 7 and 9 a.m. after an overnight fast. A scalp vein needle was placed in the antecubital vein and was kept patent with physiologic saline during a baseline period of at least thirty minutes. Crystalline porcine insulin, 0.1 unit per kilogram of body weight, was rapidly administered intravenously through a superficial arm vein of the opposite extremity. Venous blood was obtained via the scalp vein needle at three-minute intervals between ten and forty minutes and at ten-minute intervals between forty and ninety minutes. Plasma glucose was immediately measured on a glucose analyzer (glucose oxidase method), and serum was frozen at — 20° C. for subsequent immunoreactive insulin (IRI) determination. Serum IRI was determined by a modification of the radioimmunoassay method of Herbert and associates.5 Serum samples of each patient, obtained during the hypothyroid and the subsequent euthyroid state, were analyzed in the same radioimmunoassay. Fractional clearance rate of insulin (Kp was calculated for each patient by the method of the least squares,6 using the natural logarithms of actual insulin levels between ten and forty minutes. The resulting value multiplied by 100 expresses the rate of fall of serum insulin in per cent per minute. Mixing and distribution of injected insulin was expected to occur within ten to fifteen minutes. An insulin level of 600 /LIU. per milliliter was considered the end point of mixing, assuming an extracellular fluid volume between 15 and 20 per cent of body weight. Therefore insulin values above 600 /u,U. per milliliter were not utilized in calculating Kj. Insulin levels approached a near-baseline value at forty minutes, and therefore the values obtained after forty minutes were disregarded for the calculation of fractional clearance of insulin. Distribution space (DS) of the injected insulin was determined by the formula7 DS = Q/I 0 -If; where DS is distribution space in milliliters of injected insulin, Q is the amount of insulin injected, I o the theoretical value of insulin at zero time (obtained by extrapolating to zero time the slope of disappearance after equilibration, i.e. the same slope used to determine Kj), and If the baseline fasting concentration of serum insulin. The metabolic clearance rate (MCR) is the volume OCTOBER, 1975
of blood completely and irreversibly cleared of insulin per minute. It was calculated by the formula8 MCR = DS X K[. Since the metabolic clearance rate and distribution space are related to body mass, the results reported here are expressed as a function of body weight in kilograms. All calculations were done on a programmable calculator, and statistical analysis was performed by applying the Student / test and/or by the analysis of variance. RESULTS Insulin Responses in Ten Hypothyroid Patients, Ten Normal Subjects, and Six Treated Euthyroid Patients Following the intravenous administration of porcine insulin, peak levels of serum IRI occurred at or before ten minutes in all subjects. The IRI levels thereafter fell very rapidly in an exponential manner, reaching near baseline values at forty-five minutes and fluctuating minimally thereafter (figure 1). Mean serum IRI values at the different time intervals were slightly higher in the hypothyroid group than in the normal control group and in the treated euthyroid group. However, the differences were not statistically significant. The mean IRI level of less than 600 /AU. per milliliter was observed by thirteen minutes in normal subjects and in treated euthyroid patients but not until sixteen minutes in hypothyroid subjects, suggesting that the distribution of injected insulin in the extracellular fluid compartment was somewhat delayed in hypothyroid patients. Mean half-life (Ti/2), K[, DS, and MCR of insulin in the hypothyroid patients, in the normal subjects, and in the treated euthyroid patients are summarized in table 1. When compared by the analysis of variance, the differences in these parameters were not statistically significant among any of the three groups of subjects. Glucose Responses in Ten Hypothyroid Patients, Ten Normal Subjects, and Six Treated Euthyroid Patients Mean plasma glucose values following the administration of insulin in ten hypothyroid patients, ten normal subjects, and six treated euthyroid patients are shown in figure 2. Mean baseline plasma glucose of 84 ± 2 mg. per 100 ml. in hypothyroid patients was significantly lower (p j HYPOTHYROID N = IO 4 EUTHYROID N=6
500-
200100cc
~
50 H
I0- 1 0
10 20 30 40 50 60 70 80 90 MINUTES
FIG. 1. Mean serum IRI responses following intravenous administration of porcine insulin (0.1 U./kg.) in hypothyroid patients, normal control subjects, and treated euthyroid patients. Vertical line represents standard error.
minutes. In hypothyroid patients plasma glucose levels between fifty and ninety minutes following insulin administration were significantly lower (p
The frequent occurrence of abnormal carbohydrate tolerance in hypothyroid patients has been reported. 1 ' 2 Insulin secretion in response to oral3 or intravenous 2 glucose stimulus has been reported to be elevated in hypothyroid subjects, with subsequent decline of insulin values on restoration of the euthyroid state. 3 This observation of elevated levels of insulin has resulted in a hypothesis of insulin resistance to explain glucose intolerance of the hypothyroid state. 2 From the Section of Endocrinology and Metabolism, Departments of Medicine and Nuclear Medicine, Veterans Administration West Side Hospital and University of Illinois College of Medicine, Chicago, Illinois. Address reprint requests to Dr. Shah, VA West Side Hospital (M.P. 172), P.O. Box 8195, Chicago, Illinois 60680. Accepted for publication July 9, 1975. 922
In contrast to this hypothesis of insulin resistance, our previous study 4 clearly demonstrated diminished early insulin secretory response to intravenous glucose in hypothyroid patients. Following treatment and attainment of the euthyroid state, the same patients portrayed a significantly increased insulin response associated with a marked improvement in the glucose tolerance. Although this study strongly suggested the presence of hypoinsulinemia in the hypothyroid state, the possibility of coexisting insulin resistance could not be excluded entirely. The present study was undertaken to examine the possibility of insulin resistance in hypothyroid patients. In the experiments to be described, insulin resistance was determined from fractional and metabolic clearances of intravenously administered crystalline porcine insulin and its effects on peripheral plasma glucose levels. SUBJECTS AND METHODS Three groups of subjects were studied: 1. Ten hypothyroid patients, eight male and two female, were investigated soon after establishing the diagnosis of primary hypothyroidism. Clinical diagnosis of hypothyroidism was substantiated by the laboratory data, such as decrease in serum thyroxin, triiodothyronine resin uptake, and radioactive iodine uptake of the thyroid gland. None of the patients was known to have diabetes or any other complicating disease. Only one patient had a positive family history of diabetes. Three of these ten hypothyroid patients were greater than 10 per cent above ideal body weight (Metropolitan Life Insurance tables). 2. The second group was comprised of ten normal subjects without clinical or chemical evidence of hypothyroidism. These subjects were of age and weight comparable to the hypothyroid patients so that three normal subjects were also greater than 10 per cent above ideal body weight. None of these subjects had a personal or a family history of diabetes or other known illness at the time of the study. 3. Six hypothyroid patients from group 1 underDIABETES, VOL. 2 4 , NO. 10
JAYENDRA H. SHAH, M.B., B.S., M.C.P.S., AND ASSOCIATES
went repeat studies following thyroid hormone replacement therapy and approximately six to eight months after attaining the euthyroid state. All subjects were on a regular diet and consumed at least 250 gm. of carbohydrate daily for three to five days prior to each test. Studies were performed in a quiet procedure room between 7 and 9 a.m. after an overnight fast. A scalp vein needle was placed in the antecubital vein and was kept patent with physiologic saline during a baseline period of at least thirty minutes. Crystalline porcine insulin, 0.1 unit per kilogram of body weight, was rapidly administered intravenously through a superficial arm vein of the opposite extremity. Venous blood was obtained via the scalp vein needle at three-minute intervals between ten and forty minutes and at ten-minute intervals between forty and ninety minutes. Plasma glucose was immediately measured on a glucose analyzer (glucose oxidase method), and serum was frozen at — 20° C. for subsequent immunoreactive insulin (IRI) determination. Serum IRI was determined by a modification of the radioimmunoassay method of Herbert and associates.5 Serum samples of each patient, obtained during the hypothyroid and the subsequent euthyroid state, were analyzed in the same radioimmunoassay. Fractional clearance rate of insulin (Kp was calculated for each patient by the method of the least squares,6 using the natural logarithms of actual insulin levels between ten and forty minutes. The resulting value multiplied by 100 expresses the rate of fall of serum insulin in per cent per minute. Mixing and distribution of injected insulin was expected to occur within ten to fifteen minutes. An insulin level of 600 /LIU. per milliliter was considered the end point of mixing, assuming an extracellular fluid volume between 15 and 20 per cent of body weight. Therefore insulin values above 600 /u,U. per milliliter were not utilized in calculating Kj. Insulin levels approached a near-baseline value at forty minutes, and therefore the values obtained after forty minutes were disregarded for the calculation of fractional clearance of insulin. Distribution space (DS) of the injected insulin was determined by the formula7 DS = Q/I 0 -If; where DS is distribution space in milliliters of injected insulin, Q is the amount of insulin injected, I o the theoretical value of insulin at zero time (obtained by extrapolating to zero time the slope of disappearance after equilibration, i.e. the same slope used to determine Kj), and If the baseline fasting concentration of serum insulin. The metabolic clearance rate (MCR) is the volume OCTOBER, 1975
of blood completely and irreversibly cleared of insulin per minute. It was calculated by the formula8 MCR = DS X K[. Since the metabolic clearance rate and distribution space are related to body mass, the results reported here are expressed as a function of body weight in kilograms. All calculations were done on a programmable calculator, and statistical analysis was performed by applying the Student / test and/or by the analysis of variance. RESULTS Insulin Responses in Ten Hypothyroid Patients, Ten Normal Subjects, and Six Treated Euthyroid Patients Following the intravenous administration of porcine insulin, peak levels of serum IRI occurred at or before ten minutes in all subjects. The IRI levels thereafter fell very rapidly in an exponential manner, reaching near baseline values at forty-five minutes and fluctuating minimally thereafter (figure 1). Mean serum IRI values at the different time intervals were slightly higher in the hypothyroid group than in the normal control group and in the treated euthyroid group. However, the differences were not statistically significant. The mean IRI level of less than 600 /AU. per milliliter was observed by thirteen minutes in normal subjects and in treated euthyroid patients but not until sixteen minutes in hypothyroid subjects, suggesting that the distribution of injected insulin in the extracellular fluid compartment was somewhat delayed in hypothyroid patients. Mean half-life (Ti/2), K[, DS, and MCR of insulin in the hypothyroid patients, in the normal subjects, and in the treated euthyroid patients are summarized in table 1. When compared by the analysis of variance, the differences in these parameters were not statistically significant among any of the three groups of subjects. Glucose Responses in Ten Hypothyroid Patients, Ten Normal Subjects, and Six Treated Euthyroid Patients Mean plasma glucose values following the administration of insulin in ten hypothyroid patients, ten normal subjects, and six treated euthyroid patients are shown in figure 2. Mean baseline plasma glucose of 84 ± 2 mg. per 100 ml. in hypothyroid patients was significantly lower (p j HYPOTHYROID N = IO 4 EUTHYROID N=6
500-
200100cc
~
50 H
I0- 1 0
10 20 30 40 50 60 70 80 90 MINUTES
FIG. 1. Mean serum IRI responses following intravenous administration of porcine insulin (0.1 U./kg.) in hypothyroid patients, normal control subjects, and treated euthyroid patients. Vertical line represents standard error.
minutes. In hypothyroid patients plasma glucose levels between fifty and ninety minutes following insulin administration were significantly lower (p
