Serum thyrotropin response to thyrotropin-releasing hormone and free thyroid hormone indices in patients with familial thyroxine-binding globulin deficiency.

The response in serum thyrotropin (TSH) to synthetic thyrotropin-releasing hormone (TRH) as well as serum free thyroxine index (FT4I) and free triiodothyronine index (FT3I) was investigated in six patients with familial thyroxine-binding-globulin (TBG) deficiency. The total serum thyroxine (T4) and triiodothyronine (T3) concentrations were significantly decreased, compared with those of normal subjects (3.4 +/- 0.9 microgram/dl, mean +/- SD. vs. 9.0 +/- 1.5 microgram/dl, p less than 0.01 and 87 +/- 27 ng/dl vs. 153 +/- 37 ng/dl, p less than 0.01, respectively). FT4I was lower than the normal range in all but one (5.3 +/- 1.5 vs. 8.9 +/- 1.6, p less than 0.01), whereas FT3I was all in the normal range and of no significant difference from the normal control (132 +/- 22 vs. 148 +/- 25). Serum TSH concentrations in TBG deficiency were all in the normal range (1.0-4.2 muU/ml) and the maximum TSH increments following TRH 500 microgram iv were 8.9 +/- 2.0 muU/ml and of no significant difference from the normal control (10.2 +/- 4.5 muU/ml). These results indicate that the euthyroid state in familial TBG deficiency is more clearly defined by TRH-test and the normal response to TRH in familial TBG deficiency is presumably under the control of the serum free T3 level rather than the serum free T4 level.     

Reduced serum free thyroxine concentration in postmenopausal women receiving oestrogen treatment.

Thyroid hormone state was assessed in a group of postmenopausal women who had received long term treatment with oestrogen. Serum concentrations of total thyroxine, triiodothyronine, and thyroxine binding globulin were raised compared with those in a control group given placebo; serum concentrations of thyroid stimulating hormone did not differ between the groups. Oestrogen treatment resulted in a significant decrease in the serum free thyroxine concentration and in the ratio of thyroxine to thyroxine binding globulin, which supports the view that oestrogen is the causative factor of the physiological reduction in free thyroid hormone during pregnancy.     

Serum 5-androstene-3 beta,17 beta-diol sulphate and 5 alpha-androstane-3 alpha,17 beta-diol sulphate in hirsute females with polycystic ovarian disease

Serum sulphates of 5-androstene-3 beta,17 beta-diol (5-ADIOL-S), 5 alpha-androstane-3 alpha,17 beta-diol (3 alpha-DIOL-S) and dehydroepiandrosterone (DHEA-S), unconjugated androstene-dione (AD) and testosterone (T), sex hormone binding globulin (SHBG), free androgen index (FAI), 17 alpha-hydroxyprogesterone (17OHP), luteinising hormone (LH) and follicle stimulating hormone (FSH) were measured by specific radioimmunoassay in 28 hirsute women with polycystic ovarian disease (PCO) and in normal women (n = 73). Mean levels of steroids measured were significantly elevated, and SHBG significantly depressed, in the women with PCO with values (mean +/- SE) for 5-ADIOL-S (516 +/- 51 vs 267 +/- 10 nmol/l), 3 alpha-DIOL-S (130 +/- 9 vs 52 +/- 2 nmol/l), DHEA-S (7.3 +/- 0.5 vs 4.4 +/- 0.2 mumol/l), AD (11.3 +/- 1.1 vs 3.4 +/- 0.2 nmol/l), T (3.3 +/- 0.2 vs 1.5 +/- 0.1 nmol/l) and 17OHP (5.1 +/- 0.8 vs 2.8 +/- 0.2 nmol/l). SHBG levels were 31 +/- 2.9 vs 65 +/- 2.5 nmol/l, and the free androgen index [100 x T (nmol/l) divided by (SHBG nmol/l)] was 12.5 +/- 1.4 vs 2.4 +/- 0.1. The mean LH to FSH ratio was also elevated at 2.8 +/- 0.3. These studies suggest that the measurement of 5-ADIOL-S and DHEA-S may indicate adrenal gland involvement in PCO while 3 alpha-DIOL-S appears to be a reflection of peripheral androgen metabolism. A comprehensive biochemical profile of PCO should thus include the analysis of these sulphoconjugates as well as unconjugated steroids.     

Influence of chronic thyroxine treatment on plasma hormone and metabolite concentrations and on responses to insulin, glucagon and thyrotrophin releasing hormone in adult sheep

Four adult sheep fed twice daily were given daily subcutaneous injections of saline for four weeks, followed by a similar period of daily L-thyroxine (T4) injection (1 mg/day). T4 treatment increased basal plasma concentrations of T4, triiodothyronine (T3), insulin and glucose, together with T3-uptake and the free thyroxine index, while cholesterol and urea concentrations decreased. T4 treatment reduced the rise in prolactin levels after the morning meal. Thyrotrophin releasing hormone (TRH) injection increased plasma T3 only in the control period and T3-uptake only in the T4 treatment period. T4 treatment did not affect the prolactin response to TRH injection or the insulin and glucose responses to glucagon injection. The increase in insulin concentrations after insulin injection and the secondary hyperglycaemia following initial insulin-induced hypoglycaemia were reduced by T4 treatment.     

Are biochemical tests of thyroid function of any value in monitoring patients receiving thyroxine replacement?

To establish their role in monitoring patients receiving thyroxine replacement biochemical tests of thyroid function were performed in 148 hypothyroid patients studied prospectively. Measurements of serum concentrations of total thyroxine, analogue free thyroxine, total triiodothyronine, analogue free triiodothyronine, and thyroid stimulating hormone, made with a sensitive immunoradiometric assay, did not, except in patients with gross abnormalities, distinguish euthyroid patients from those who were receiving inadequate or excessive replacement. These measurements are therefore of little, if any, value in monitoring patients receiving thyroxine replacement. To stop doing thyroid function tests in these cases would result in considerable savings nationally in the cost of reagents in laboratories using commercial kits.     

Possible role of adrenomedullin in the regulation of Fontan circulation: mature form of plasma adrenomedullin is extracted in the lung in patients with Fontan procedure

OBJECTIVE: We investigated the pathophysiological significance of molecular forms of adrenomedullin (AM) in patients after the Fontan procedure. METHODS: Plasma concentrations of mature AM (AM-m), an active form, glycine-extended AM (AM-Gly), an inactive form, and total AM (AM-T: AM-m+AM-Gly) were measured by specific immunoradiometric assay in the femoral vein, pulmonary artery and femoral artery of 29 consecutive patients after the Fontan procedure. The eleven patients who had history of Kawasaki disease and have normal coronary and hemodynamics served as control. RESULTS: Patients who underwent Fontan procedure had significantly higher venous concentrations of AM-T, AM-Gly, and AM-m than age-matched normal controls (AM-T, 12.0+/-3.3 vs. 9.6+/-2.0; AM-Gly, 10.4+/-3.0 vs. 8.5+/-1.6; AM-m, 1.6+/-0.7 vs. 1.0+/-0.6 pmol/l, each p<0.05). In patients with Fontan procedure, there were no differences in plasma AM-T, AM-Gly or AM-m levels between the femoral vein and pulmonary artery, however, there was a significant step-down in the AM-m levels, but not in plasma AM-T or AM-Gly levels, between the pulmonary artery and femoral artery (1.3+/-0.6 to 1.0+/-0.6, p<0.05). The venous concentrations of AM-m correlated negatively with systemic blood flow (cardiac output) (r=-0.46, p<0.05). CONCLUSIONS: Results suggest that in Fontan circulation plasma AM-m is increased in parallel with those of AM-T and AM-Gly and that AM-m is extracted in the lung. Extracted AM-m may be involved in the regulation of pulmonary arterial tonus, although further studies are necessary to elucidate the exact role of AM in Fontan circulation.     

Effects of cortisol on lipolysis and regional interstitial glycerol levels in humans

Cortisol's effects on lipid metabolism are controversial and may involve stimulation of both lipolysis and lipogenesis. This study was undertaken to define the role of physiological hypercortisolemia on systemic and regional lipolysis in humans. We investigated seven healthy young male volunteers after an overnight fast on two occasions by means of microdialysis and palmitate turnover in a placebo-controlled manner with a pancreatic pituitary clamp involving inhibition with somatostatin and substitution of growth hormone, glucagon, and insulin at basal levels. Hydrocortisone infusion increased circulating concentrations of cortisol (888 +/- 12 vs. 245 +/- 7 nmol/l). Interstitial glycerol concentrations rose in parallel in abdominal (327 +/- 35 vs. 156 +/- 30 micromol/l; P = 0.05) and femoral (178 +/- 28 vs. 91 +/- 22 micromol/l; P = 0.02) adipose tissue. Systemic [(3)H]palmitate turnover increased (165 +/- 17 vs. 92 +/- 24 micromol/min; P = 0.01). Levels of insulin, glucagon, and growth hormone were comparable. In conclusion, the present study unmistakably shows that cortisol in physiological concentrations is a potent stimulus of lipolysis and that this effect prevails equally in both femoral and abdominal adipose tissue.     

Acute effects of metabolic hormones in swine

Many metabolic hormones (glucagon, hydrocortisone, corticosterone, TSH, thyroxine and triiodothyronine) did not stimulate porcine adipose tissue lipolysis in vitro. Growth hormone and ACTH stimulated lipolysis at high concentrations, in the presence of theophylline. Insulin inhibited lipolysis. Infusion of metabolic hormones with measurement of plasma free fatty acid and glycerol concentrations, purportedly indicative of in vivo lipolysis, indicated that glucagon and somatotropin had no effect, adrenocorticotropin increased and insulin depressed plasma concentrations of the metabolites. Overall, the in vitro predicts the in vivo response. There were exceptions, e.g. adrenocorticotropin moderately increased plasma metabolites but had little effect in vitro.     

Relevance of increased serum thyroxine concentrations associated with normal serum triiodothyronine values in hypothyroid patients receiving thyroxine: a case for "tissue thyrotoxicosis".

Fifteen patients receiving standard thyroxine replacement therapy (100-200 micrograms daily) for primary hypothyroidism and who had persistently raised free thyroxine concentrations in their serum were investigated to see whether the dose being given was too high. In addition to the usual thyroid hormone assays systolic time intervals (which indicate left ventricular contractility) were calculated as accurate reflectors of tissue thyroid activity. All patients showed the expected increased free and total thyroxine concentrations; but mean total and free concentrations of triiodothyronine were normal, while reverse triiodothyronine values were raised. Mean systolic time intervals were significantly reduced as compared with normal and fell within the thyrotoxic range. Seven patients subsequently had their doses of thyroxine reduced by 50 micrograms daily and were reinvestigated one month later. All showed significant falls in circulating thyroxine and triiodothyronine concentrations and an increase in mean systolic time intervals to the normal range. In patients receiving thyroxine replacement therapy for primary hypothyroidism a raised serum thyroxine concentration may indicate tissue thyrotoxicosis and should prompt a reduction of the thyroxine dose.     

PRF activity of VIP in vitro (author's transl)

The effect of VIP on prolactin secretion from incubated rat hemipituitaries was characterized. Under these conditions, the secretion of GH, LH, FSH, ACTH was not affected, indicating that the effect of VIP is hormone specific. The stimulation of prolactin was dose-dependent, with an apparent affinity of VIP of 10.9 +/- 3.1 nM and a maximal stimulation of 57.7 +/- 4.2%. Secretin, a structurally related peptide, was also active at higher concentrations, whereas another partial analogue, glucagon, was ineffective. Furthermore, VIP does not act through pituitary DA receptors since alpha-flupentixol, a potent dopaminergic antagonist, does not block the stimulation of prolactin secretion by VIP. In addition, stimulation by VIP and TRH was additive. Naloxone and met-enkephalin were ineffective on the VIP effect on prolactin release. In contrast, SRIF seems to inhibit the VIP stimulation of prolactin release. Our data suggest that VIP, which was found in the hypothalamo-hypophyseal blood at concentrations of the same order of magnitude as that found to stimulate PRL in vitro, could be a physiological PRF.     

Neonatal screening for congenital hypothyroidism by measurement of plasma thyroxine and thyroid stimulating hormone concentrations.

Neonatal screening for congenital hypothyroidism was introduced in the City of Birmingham in 1980 by measuring concentrations of both thyroid stimulating hormone and thyroxine in plasma. Over two years 30 108 babies were tested. Thirty one babies were recalled because of thyroid stimulating hormone concentrations greater than 40 mU/l, of whom 12 were treated with replacement thyroxine. Six babies were found to have low thyroxine concentrations because of reduced thyroxine binding globulin and five raised thyroxine values because of increased thyroxine binding globulin. As a result of this study screening was continued with measurement of thyroid stimulating hormone only as the primary test for congenital hypothyroidism, the thyroxine value being measured only when the concentration of thyroid stimulating hormone exceeded 20 mU/l.     

Similar serum concentrations of thyroid hormones in two geographically separate populations on disparate iodine intake.

Serum thyroid hormones were measured in Montreal, Canada (urinary iodine 446 +/- 164 micrograms/day) and Zagreb, Yugoslavia (urinary iodine 108 +/- 32 microgram/day). The serum concentrations of thyroxine and triiodothyronine in the two populations were almost identical. We conclude that dietary iodine, within accepted normal limits, is not a factor in determining serum thyroid hormone levels. The wide differences in reported serum triiodothyronine concentrations are related to methodological problems.     

Oral thyrotropin-releasing hormone (TRH) test in acromegaly

The effects of 40 mg oral and 200 microgram intravenous TRH were studied in patients with active acromegaly. Administration of oral TRH to each of 14 acromegalics resulted in more pronounced TSH response in all patients and more pronounced response of triiodothyronine in most of them (delta max TSh after oral TRh 36.4 +/- 10.0 (SEM) mU/l vs. delta max TSH after i.v. TRH 7.7 +/- 1.5 mU/l, P less than 0.05; delta max T3 after oral TRH 0.88 +/- 0.24 nmol/vs. delta max T3 after i.v. TRH 0.23 +/- 0.06 nmol/l, P less than 0.05). Oral TRH elicited unimpaired TSH response even in those acromegalics where the TSH response to i.v. TRH was absent or blunted. In contrast to TSH stimulation, oral TRH did not elicit positive paradoxical growth hormone response in any of 8 patients with absent stimulation after i.v. TRH. In 7 growth hormone responders to TRH stimulation the oral TRH-induced growth hormone response was insignificantly lower than that after i.v. TRH (delta max GH after oral TRH 65.4 +/- 28.1 microgram/l vs. delta max GH after i.v. TRH 87.7 +/- 25.6 microgram/l, P greater than 0.05). In 7 acromegalics 200 microgram i.v. TRH represented a stronger stimulus for prolactin release than 40 mg oral TRH (delta max PRL after i.v. TRH 19.6 +/- 3.22 microgram/, delta max PRL after oral TRH 11.1 +/- 2.02 microgram/, P less than 0.05). Conclusion: In acromegalics 40 mg oral TRH stimulation is useful in the evaluation of the function of pituitary thyrotrophs because it shows more pronounced effect than 200 microgram TRH intravenously. No advantage of oral TRH stimulation was seen in the assessment of prolactin stimulation and paradoxical growth hormone responses.     

The pituitary-thyroid axis in acromegaly

The pituitary-thyroid axis of 12 acromegalic patients was evaluated by measurement of the serum concentrations (total and free) of thyroxine (T4), triiodothyronine (T3) and reverse T3 (rT3) and thyrotropin (TSH), growth hormone (GH) and prolactin (PRL) before and after iv stimulation with thyrotropin releasing hormone (TRH). Using an ultrasensitive method of TSH measurement (IRMA) basal serum TSH levels of the patients (0.76, 0.07-1.90 mIU/l) were found slightly, but significantly (P less than 0.01), lower than in 40 healthy controls (1.40, 0.41-2.50 mIU/l). The total T4 levels (TT4) were also reduced (84, 69-106 nmol/l vs 100, 72-156 nmol/l, P less than 0.01) and significantly correlated (P less than 0.02, R = 0.69) to the TSH response to TRH, suggesting a slight central hypothyroidism. The acromegalics had, however, normal serum levels of TT3 (1.79, 1.23-2.52 nmol/l vs 1.74, 0.78-2.84 nmol/l, P greater than 0.10), but significantly decreased levels of TrT3 (0.173, 0.077-0.430 nmol/l vs 0.368, 0.154-0.584 nmol/l, P less than 0.01) compared to the controls. The serum concentration of the free iodothyronines (FT4, FT3, FrT3) showed similar differences between acromegalics and normal controls. All the acromegalics showed a rise of serum TSH, GH and PRL after TRH. Positive correlation (P less than 0.05, R = 0.59) was found between the TSH and GH responses, but not between these two parameters and the PRL response to TRH. These findings may be explained by the existence of a central suppression of the TSH and GH secretion in acromegalic subjects, possibly exerted by somatostatin. Euthyroidism might be maintained by an increased extrathyroidal conversion of T4 to T3.     

Reduced serum acylated ghrelin levels in patients with hyperthyroidism

BACKGROUND AND OBJECTIVE: Recent studies have revealed that circulating ghrelin levels seem to play a role in energy homeostasis. The effect of hyperthyroidism on ghrelin levels is not fully known. METHODS: Serum levels of ghrelin and its relationship with insulin resistance were evaluated in 48 patients with hyperthyroidism and 43 euthyroid healthy controls. Thyroid hormones, insulin, glucose, ghrelin levels and lipid parameters were measured in all subjects. Insulin sensitivity was determined using the homeostasis model assessment. RESULTS: Serum ghrelin levels were significantly decreased in hyperthyroid patients than in controls (32.5 +/- 23.3 vs. 54.1 +/- 35.5 pg/ml, p < 0.001). Circulating ghrelin levels significantly correlated with age (r = -0.26, p = 0.01), fasting glucose (r = -0.21, p = 0.01), free triiodothyronine (r = -0.18, p = 0.04), free thyroxine (r = -0.23, p = 0.02) and thyroid stimulating hormone (r = 0.21, p = 0.04), but not with blood pressure, body mass index, lipid parameters, insulin and homeostasis model assessment (p > 0.05). Multiple regression analysis revealed glucose level to be the most important predictor of circulating ghrelin level. CONCLUSION: These results indicate that hyperthyroidism has effect on serum ghrelin levels. Further studies are needed for the exact mechanism.     

Role of glucagon in the splanchnic and systemic hemodynamic changes induced by portal-systemic blood shunting

Portal-systemic blood shunting is often accompanied by hyperglucagonemia and hemodynamic changes. To determine this causal relation, splanchnic and systemic hemodynamics (radioactive microspheres) and plasma glucagon levels (radioimmunoassay) were assessed in conditions of total portal-systemic shunting in portacaval-shunted (PCS) rats and in sham-operated (SO) normal rats. To compare these results, another hemodynamic study was undertaken basally and during glucagon infusion in nonoperated normal rats. PCS rats showed a threefold greater plasma glucagon concentration than SO animals (924 +/- 134 vs. 309 +/- 18 pg/ml, p less than 0.01), and they developed a hyperdynamic splanchnic circulation with higher portal venous inflow than SO rats (8.29 +/- 1.1 vs. 5.09 +/- 0.4 ml/min/100 g, p less than 0.05). Infusion of a pharmacological dose of glucagon in normal rats increased portal venous inflow (from 4.92 +/- 0.33 to 6.24 +/- 0.48 ml/min/100 g, p less than 0.05) so as to imply this hormone in the development of the hyperdynamic splanchnic circulation in conditions of portal-systemic shunting. However, the discrepancies in systemic hemodynamics between PCS and glucagon-infused rats may be a result of the different plasma glucagon levels reached in the two groups.     

A novel melatonin antagonist, N-(2,4-dinitrophenyl)-5-methoxytryptamine neutralizes some effects of melatonin in the female Syrian hamster

In this present study we evaluated the ability of a recently synthesized melatonin antagonist, N-(2,4-dinitrophenyl)-5-methoxytryptamine (ML-23), to antagonize the effects of afternoon injections of melatonin on the reproductive and thyroid axes in the female Syrian hamster. Thirty-six animals were divided into four groups and treated daily for 13 weeks with an afternoon injection of melatonin (25 micrograms/injection) or saline diluent. ML-23 was given via the drinking water to both melatonin- and saline-treated groups. The experiment was continued until 78% of melatonin-treated animals exhibited acyclicity. The results show that ML-23 partially reversed the effects of melatonin on pituitary follicle-stimulating hormone concentrations but was without effect on the decreased pituitary and plasma prolactin concentrations induced by melatonin treatment. Furthermore, ML-23 antagonized the effects of melatonin on plasma thyroxine levels and significantly increased plasma triiodothyronine concentrations and the free triiodothyronine index when used in combination with melatonin. The decrease in ovarian weight and plasma estradiol, but not progesterone, obtained with melatonin treatment also was reversed by ML-23. Our data suggest that ML-23 prevents the effects of melatonin treatment on ovarian weight, pituitary follicle-stimulating hormone levels, plasma estradiol, and thyroxine concentrations in the female Syrian hamster. Since ML-23 did not prevent the effects of melatonin on pituitary weight, plasma luteinizing hormone and prolactin, and pituitary prolactin concentrations, the actions of ML-23 may involve only peripheral sites of action of melatonin. Alternatively, the dose of ML-23 may not have been optimal to prevent all of the central effects of the indoleamine.     

Total and free thyroid hormone concentrations in patients receiving maintenance replacement treatment with thyroxine.

Total and free serum concentrations of thyroxine and triiodothyronine were measured in 122 subjects with hypothyroidism who were clinically well while receiving conventional replacement treatment with thyroxine. In a third of patients concentrations of total and free thyroxine were raised, often considerably; nevertheless concentrations of total and free triiodothyronine were usually normal. Though significant correlations were obtained between total triiodothyronine concentrations and total thyroxine concentrations (p less than 0.001) and between the triiodothyronine concentrations and free thyroxine concentrations (p less than 0.001) the slope of the line of the regression equation describing these correlations was small, hence large increases in both total and free thyroxine concentrations were accompanied by only modest increases in total and free triiodothyronine concentrations. The presence of total or free thyroxine concentrations above normal in patients taking thyroxine therefore are not necessarily of clinical consequence. In the assessment of adequacy of replacement treatment with thyroxine the most logical combination of in vitro thyroid function test results may be a normal thyrotrophin concentration and normal free triiodothyronine concentration.     

Demonstration of a dawn phenomenon in normal adolescents

To ascertain whether the dawn phenomenon occurs in normal adolescents and, if so, to determine its mechanism, we measured nocturnal plasma glucose, insulin, glucagon, growth hormone, cortisol, and adrenocorticotropic hormone (ACTH) levels between 01.00 and 08.00 h in 10 healthy adolescents. The prehepatic insulin secretion rate was calculated based on C peptide levels. The metabolic clearance rate of insulin (MCRI) was calculated as the ratio of mean insulin secretion rate to mean insulin concentration. There was no change in plasma glucose, insulin, and glucagon between 01.00-04.00 and 05.00-08.00 h (paired t test). The MCRI was higher at 05.00-08.00 h compared to 01.00-04.00 h (9.30 +/- 1.50 vs. 4.87 +/- 1.11 ml.kg-1.min-1; p = 0.008). The prehepatic insulin secretion increased at 05.00-08.00 h relative to 01.00-04.00 h (1.1 +/- 0.2 vs. 0.6 +/- 0.1 pmol.kg-1.min-1; p = 0.013). Similarly, cortisol and ACTH levels were higher at 05.00-08.00 versus 01.00-04.00 h (323 +/- 33 vs. 102 +/- 22 nmol/l, p less than 0.001; 3.6 +/- 0.5 vs. 1.8 +/- 0.4 pmol/l, p = 0.006, respectively). Growth hormone was higher at 01.00-04.00 versus 05.00-08.00 h (7.6 +/- 1.2 and 3.0 +/- 0.9 microgram/l; p = 0.019). ACTH correlated with MCRI (r = 0.66; p = 0.002) and prehepatic insulin secretion (r = 0.75; p less than 0.01).(ABSTRACT TRUNCATED AT 250 WORDS)     

Influence of thyroid function on serum bone Gla protein

The serum BGP level was assayed in patients with hyperthyroidism (untreated and remittent cases) and hypothyroidism. The mean serum BGP concentration was 9.7 +/- 0.90 ng/ml in 30 patients with untreated hyperthyroidism which was significantly higher than the 2.7 +/- 0.38 ng/ml in 15 remittent patients and 1.3 +/- 0.31 ng/ml in 13 patients with hypothyroidism (p less than 0.001, p less than 0.001). Serum BGP had a significant positive correlation with the concentrations of free triiodothyronine and alkaline phosphatase in the serum, while it had a significant negative correlation with serum PTH. In the patients with hypothyroidism, serum BGP increased significantly in parallel with increases in serum free triiodothyronine with thyroxine therapy. In the patients with hyperthyroidism, serum free triiodothyronine decreased significantly after the first month of methimazole treatment, and fluctuated within the normal range after two months. Serum alkaline phosphatase and BGP did not show significant changes during the first six months of treatment, although they were eventually reduced significantly at the end of one year. These results suggest that thyroid hormone directly stimulates the synthesis and secretion of BGP in existent osteoblasts and also acts on the bone remodeling cycle, therapy accelerating the rate of bone formation; the latter action may occur over a long period.