Effects of exercise and physical fitness on the pituitary-thyroid axis and on prolactin secretion in male runners

The effects of acute exercise and thyrotropin-releasing hormone on the pituitary-thyroid axis were examined in men placed into three well-defined categories of physical fitness. There were 20 sedentary men, 22 joggers (running four to 20 miles per week) and 18 marathoners (running 30 to 100 miles per week) who participated. During treadmill exercise, the mean VO2 max differed among all groups, being 38.5, 45.0, and 60.3 mL/kg . min in the sedentary, jogger, and marathon groups, respectively. Serum was obtained before, immediately after, and one hour after exercise for measurement of thyroxine (T4), triiodothyronine (T3), reverse T3, thyrotropin (TSH), and prolactin. Basal values of all hormones did not differ among the groups. Maximal short-term treadmill exercise produced no change in serum T4, T3, reverse T3, or TSH. Prolactin rose significantly by a similar amount in all three subject groups. On a separate day, ten individuals from each group received thyrotropin releasing hormone (TRH; 500 micrograms IV). Neither the peak TSH response nor the total TSH secreted during two hours after TRH differed among groups. The mean total prolactin secretion in the joggers and marathoners was 48% and 45% greater, respectively, than in the sedentary men. Five subjects in each group also underwent a TRH test immediately postexercise. Similar to the results on the nonexercise day, the integrated TSH response to TRH was similar in all three groups, whereas the integrated PRL response to TRH was increased by 52% and 78% in the two conditioned groups. Post-TRH sera from one subject in each group were fractionated on a Sephadex G-100 column.(ABSTRACT TRUNCATED AT 250 WORDS)     

3'-Monoiodothyronine degradation in rat liver homogenate: enzyme characteristics and documentation of deiodination by high-pressure liquid chromatography

Characteristics of 3'-monoiodothyronine (3'-T1) degradation were examined in vitro in rat tissue homogenates. In rat liver homogenates, 3'-T1 degradation was optimal at pH 7.4, and was dependent upon time, temperature, and tissue concentration. The Michaeli's constant (Km) = 0.84 mumol/L. 3'-T1 degradation was enhanced by dithiothreitol and inhibited by propylthiouracil, sodium ipodate, ANS, and sodium azide but not by methimazole. Animals that fasted for three days had significant reductions in both hepatic T4 to T3 conversion (199 +/- 12 v 116 +/- 12 pg T3 generated/mg protein; P less than 0.001) and 3'-T1 degradation (588 +/- 31 v 148 +/- 53 pg 3'-T1 degraded/mg protein; P less than 0.001). To document that 3'-T1 degradation was occurring by deiodination, both liver and kidney homogenates were incubated with 125I-3'-T1 (approximately 3 microCi; 13.1 nmol/L). The reaction products were separated on a reverse-phase high pressure liquid chromatography (HPLC) column. In both tissues an iodide peak was generated, and no other radiolabeled peaks appeared except for 125I-3'-T1. These data suggest that 3'-T1 is metabolized by phenolic-ring monodeiodination and is enzymic in nature.     

Acromegaly and the heart. An echocardiographic study.

Twenty-seven patients with acromegaly had echocardiograms performed to delineate the ventricular septum, left ventricular posterior wall and mitral valve. Left ventricular function was assessed by calculating the systolic internal dimensional shortening of the left ventricle. Six patients met the criteria for asymmetric septal hypertrophy and eight had concentric left ventricular hypertrophy. The remaining 13 patients were categorized as "normal," although six had septal measurements greater than 11 mm. The group with asymmetric septal hypertrophy had significantly greater percentage of internal dimensional shortening during systole than either the normal group (p less than 0.05) or the group with left ventricular hypertrophy (p less than 0.01). Initial mean growth hormone levels were considerably higher in the group with left ventricular hypertrophy than in the normal group (93 versus 34 ng/ml). Thus, echocardiographic abnormalities are common in acromegaly, and patients with asymmetric septal hypertrophy and acromegaly appear to have significantly increased ventricular ejection. Many of the patients with left ventricular hypertrophy have no evidence of clinical cardiovascular disease, and their left ventricular hypertrophy may be related to higher initial growth hormone levels.     

Hypocalcemia with osteoblastic metastases in a patient with prostate carcinoma: A cause of secondary hyperparathyroidism

A 68 year old man with prostatic carcinoma and extensive painful osteoblastic metastases was discovered to have hypocalcemia (serum calcium 7.1 mg/dl) without evidence of hypoalbuminemia, renal failure or malabsorption. Baseline studies revealed hypocalciuria (24 hour urine calcium <5 mg/day), normal serum phosphate (3.4 mg/dl), low tubular reabsorption of phosphate (68 percent), undetectable serum calcitonin, normal serum 25-hydroxyvitamin D, slightly elevated serum parathyroid hormone level and increased urinary cyclic AMP (8.87 μmol/g creatinine). These studies were compatible with secondary hyperparathyroidism. The intravenous administration of parathyroid extract produced no further change in urinary phosphate but a 25-fold increase in nephrogenous cyclic AMP. Three days administration of intramuscular parathyroid extract slowly and temporarily restored serum calcium to normal levels while increasing urinary cyclic AMP and phosphate. Chemotherapy with cyclophosphamide and 5-fluorouracil rendered the patient free of pain while reducing serum acid and alkaline phosphatase levels and restoring serum total and ionized calcium and urinary cyclic AMP excretion to normal.     

Recent developments in thyroid hormone metabolism: interpretation and significance of measurements of reverse T3, 3,3'T2, and thyroglobulin

The development of specific radioimmunoassays has allowed measurements of 3,3',5'-triiodothyronine (reverse T3), 3,3'-diiodothyronine (3,3'T2), and thyroglobulin to be performed in serum and in various biological fluids both in normal and in altered states of thyroidal economy. The physiology, kinetics, and metabolic actions of reverse T3, 3,3'T2 and thyroglobulin are reviewed. Presently, it appears that reverse T3 and 3,3'T2 measurements in amniotic fluid and cord serum may potentially be useful in diagnosing fetal or neonatal thyroid dysfunction, and serum thyroglobulin measurements appear to be important as a measure of the efficacy of treatment of patients with differentiated thyroid carcinoma.     

Glucose modulation of alterations in serum iodothyronine concentrations induced by fasting.

In order to investigate the process by which dietary composition may regulate T4 conversion to T3 and reverse T3, iodothyronine levels were measured in the sera of seven obese subjects during consecutive study periods. These study periods included the ingestion of an approximate weight-maintaining diet (40% carbohydrate, 40% fat, 20% protein) during a control period of 4 days, a fast of 7 days thereafter, and then a 5-day period of glucose ingestion (50 g/day) only. The mean (±SE) serum T3 concentration was 117 ± 8 ng/dl on day 4 of the control period, and gradually decreased to 66 ± 11 ng/dl (p < 0.01) on the last day of fasting. The subsquent administration of glucose was associated with an increase in the mean serum T3 level to 94 ± 10 ng/dl (p < 0.01). Mean (±SE) serum levels of reverse T3 varied reciprocally and were 52 ± 9 ng/dl, 82 ± 12 ng/dl (p < 0.005), and 65 ± 9 ng/dl (compared to fasting, p < 0.05) during the fed and fasting states and during glucose administration, respectively. Furthermore, employing a similar protocol in a different group of subjects, serum sampled during the administration of 100 g of fructose orally during days 8–12 of fasting also was associated with an increase in mean serum T3 and a decrease in mean serum reverse T3, as compared to values obtained on day 6 or day 7 of fasting (T3: 83 ± 6 ng/dl, fasting vs. 111 ± 10, fructose (p < 0.05); rT3: 56 ± 9, fasting vs. 42 ± 6 ng/dl, fructose (p < 0.025)). Serum T4 concentrations were not significantly altered in any study period either during glucose or fructose ingestion. Despite the decrement in serum T3 levels observed during fasting, the mean peak TSH in response to TRH stimulation in a group of 15 obese subjects was decreased during fasting as compared to the fed state (8.1 ± 1.2 μU/ml, fast vs. 12.8 ± 2.0 μU/ml, fed). These observations suggest that both glucose and fructose are capable of modulating serum T3 and reverse T3 levels and that administration of these hexoses in doses of only 100–200 g/day for 5 days may be effective in altering T4 degradative pathways. Furthermore, despite the decreased serum T3 levels, the TSH response to TRH stimulation is decreased, paradoxically, during fasting.     

The effect of T3 and reverse T3 administration on muscle protein catabolism during fasting as measured by 3-methylhistidine excretion.

Since recent studies have indicated that measurement in urine of the amino acid, 3-methylhistidine, accurately reflects the extent of muscle catabolism, and because it has been suggested that thyroid hormones may influence muscle breakdown, especially during fasting, the effect of T3 and reverse T3 (rT3) administration on the excretion of 3-methylhistidine was examined in obese subjects during fasting. The mean (+/- SE) 3-methylhistidine excretion in patients fed an egg protein diet (devoid of meat protein) was 256 +/- 35 mumoles/day and decreased to 190 +/- 14 mumoles/day during fasting. T3 administration (100 microgram/day x 5 days) increased 3-methylhistidine excretion to 304 +/- 37 mumoles/day during its ingestion and to 485 +/- 46 mumoles/day in the T3 posttreatment interval. T3 doses of 10 microgram every 4 hr (q4h) for the first 6 days of fasting also appeared capable of increasing 3-mehis excretion whereas 5 microgram T3 q4h administered during the first 6 days of fasting did not increase 3-mehis excretion. Reverse T3 administration (80 microgram q6h) during fasting was associated with a mean 3-methylhistidine of 130 +/- 13 mumoles/day, a value no higher than in patients fasted alone. These observations suggest that: (1) skeletal muscle catabolism decreases during fasting: and (2) pathophysiologic doses of T3 (60 microgram/day or more), but not reverse T3, enhance muscle catabolism during fasting.     

Coronary artery disease in young patients: arteriographic and clinical review of 40 cases aged 35 and under

Coronary arteriography was performed in 60 patients aged 35 or less with suggested coronary artery disease (CAD). Twenty patients (Group 1) had normal coronary arteries and 40 patients (Group 2) had one or more obstructive lesions. The left anterior descending artery was commonly involved followed by the right coronary and left circumflex arteries. The right coronary artery was most commonly completely obstructed. Single-vessel disease (50 per cent or greater obstruction) was found in 60 per cent of the patients, an incidence that is considerably higher than in studies of older patients. A total of 1.6 diseased vessels per patient was present. A hyperlipoproteinemia (HLP) was found in 68 per cent of Group 2 patients. Patients in Group 2 with an HLP had significantly more CAD than Group 2 patients with normal lipoproteins. The incidence of the following clinical features were not significantly different in Groups 1 and 2: typical angina, atypical angina, positive family history, smoking, hypertension, obesity, abnormal electrocardiogram, positive treadmill test, HLP, and diabetes mellitus. A fourth heart sound and a history of a myocardial infarction were significantly common in Group 2. Since almost all of the previously reported cases of myocardial infarction with normal coronary arteries have occurred in young patients, history of a myocardial infarction does not assure the presence of obstructive coronary artery lesions. It is suggested that coronary arteriography is a justifiable procedure in a young patient who presents with a clinical picture that is either compatible with or cannot be clearly distinguished from CAD.     

Normal valine disposal in obese subjects with impaired glucose disposal: Evidence for selective insulin resistance

Insulin has major effects on both glucose and branched chain amino acid metabolism. To determine whether the insulin resistance of obesity equally affects both glucose and branched chain amino acid metabolism, we measured the ability of obese and normal subjects to dispose of intravenous bolus doses of glucose (25 g) or L-valine (4 g). Basal plasma glucose levels were the same in the 18 normal and 17 obese (163 ± 8% of ideal body weight) subjects, but basal plasma insulin levels were higher in the obese group (15 ± 2 vs 6 ± 1 μU/ml; p < 0.001). The obese group had a slower glucose disappearance rate after glucose challenge (0.84 ± 0.06 vs 1.11 ± 0.07 hr^?1; p < 0.01) despite having a greater serum insulin response to the glucose load (26 ± 4 vs 11 ± 1 insulin area units; p < 0.01), confirming insulin resistance. In contrast, disposal of a valine load was the same in normal and obese subjects, as assessed by initial and second phase exponential disappearance rates, metabolic clearance rates of valine, and volumes of distribution. In normal men, disposal rates of glucose and valine after simultaneous administration of both substances were slower than corresponding disposal rates determined when each substance was given alone. We conclude that obese subjects with impaired glucose disposal have normal valine disposal, suggesting that the insulin resistance of obesity can be selective in its effect on different metabolic systems. Glucose and valine also appear to mutually antagonize each other's disposal.     

Failure of 3,3'-diiodothyronine administration to alter TSH and prolactin responses to TRH stimulation.

In order to determine whether elevations in serum 3,3'-diiodothyronine (3,3'T2) concentrations influence the hypothalamic-pituitary--thyroid axis, thyrotropin (TSH) and prolactin responses to thyrotropin-releasing hormone (TRH) were assessed in five patients both prior to and during 3,3'T2 administration. Mean (+/- SE) peak TSH responses to TRH were 168 +/- 64 microU/ml during 3,3'T2 administration and 168 +/- 65 muU/ml during 3,3'T2 administration. Mean basal and peak prolactin concentrations after TRH were 6 +/- 3 ng/ml and 54 +/- 26 ng/ml, whereas during 3,3'T2 administration the basal and peak prolactin levels were 6 +/- 2 ng/ml and 55 +/- 28 ng/ml, respectively. Hypothyroid rats administered triiodothyronine (10 migrogram b.i.d.) for 5 days had a mean TSH response to TRH stimulation of 0.051 +/- 0.003 mU/ml, whereas rats to whom saline or 3,3'T2 (50 microgram b.i.d.) had been given for the same time interval had mean TRH-induced TSH responses of 1.127 +/- 0.179 mU/ml and 1.324 +/- 0.286 mU/ml, respectively. None of the TSH or prolactin responses to TRH, in either human or rat studies, were apparently altered by 3,3'T2. These observations suggest that elevation of serum 3,3'T2 levels are not associated with alterations in the hypothalamic--pituitary--thyroid axis in the experimental systems employed.     

Androgen receptor abnormalities in identical twins with oligospermia. Clinical and biochemical studies

Identical twin brothers presented with oligospermia, small testes, normal male phenotypes, elevated serum luteinizing hormone levels, and normal or elevated serum testosterone levels. Both men had low to low-normal cytosol androgen receptor binding capacity in cultured fibroblasts from pubic skin biopsy specimens. Qualitative abnormalities of cellular androgen receptors were suggested by low-normal or low nuclear androgen uptake in fibroblasts from both brothers as well as abnormal thermolability and subnormal molybdate stabilization of androgen receptors from one brother. In vivo androgen sensitivity was assessed in one twin following administration of testosterone or the non-aromatizable androgen fluoxymesterone. Fluoxymesterone suppressed serum luteinizing hormone and serum testosterone/estradiol-binding globulin, and although testosterone suppressed both serum luteinizing hormone and serum follicle-stimulating hormone, the suppression of serum luteinizing hormone by testosterone was subnormal. Both subjects showed marked exaggeration of the serum 17-hydroxyprogesterone increase after administration of human chorionic gonadotropin, despite normal serum testosterone increases, suggesting a block in testicular 17,20-desmolase, which converts 17-hydroxyprogesterone to testosterone. These studies suggest that oligospermia and block of the enzyme 17,20-desmolase may be the earliest manifestations of androgen resistance, and the finding of the syndrome of oligospermia, normal male phenotype, and androgen receptor abnormalities in identical twins indicates a genetic etiology of this disorder.     

Adult meningoencephalitis caused by herpesvirus hominis type 2

Herpesvirus hominis (HVH) type 2 was isolated from cerebrospinal fluid of a 19 year old man whose meningoencephalitis was associated with marked lymphocytic pleocytosis (as high as 2,080 cells/mm³), elevated cerebrospinal fluid protein (280 mg/ 100 ml) and hypoglycorrachia. Serologic studies showed significant antiherpes antibody increase during convalescence. The pathobiologic features of central nervous system infection with herpes simplex subtypes are discussed.     

Nature of suppressed TSH secretion during undernutrition: Effect of fasting and refeeding on TSH responses to prolonged TRH infusions

TSH responses to 4-hr continuous TRH infusions of approximately 0.8 μg/min were assessed during feeding (1500 Kcal), fasting, and refeeding (1500 Kcal) intervals in 9 euthyroid obese subjects. The total area under the TSH response curve was 1854 ± 322 μU/ml · 4-hr during feeding, decreased to 1359 ± 199 μU/ml · 4-hr (p < 0.01) on the 10th day of fasting, and remained low, being 1405 ± 185 μU/ml · 4-hr, despite refeeding a 1500 Kcal diet (40% carbohydrate, 40% fat, 20% protein) for 5 days. Baseline serum T3 concentrations were 167 ± 11 ng/dl during feeding, 86 ± 8 ng/dl during fasting, and 119 ± 12 ng/dl during refeeding. The observed decreases in TSH release appeared to correlate with decreased biologic action on the thyroid gland since the net rise in T3 during the infusion was less in fasting and refeeding than in the control (fed) period. Basal serum rT3 levels were 42 ± 5 ng/dl during feeding, rose as expected to 56 ± 5 ng/dl during fasting (p < 0.005), and were completely restored to normal during refeeding (36 ± 5 ng/dl). These data suggest that: (1) TSH responsiveness to prolonged TRH infusion is diminished during fasting and does not return to control (fed) values despite 5 days of refeeding a 1500 Kcal diet; (2) net T3 increases observed during the TRH infusion are greater in the fed period than in the fasting or refeeding periods; and (3) 5 days of refeeding a 1500 Kcal diet (40% carbohydrate, 40% fat, 20% protein) did not return the T3 to its original fed value whereas rT3 was completely restored to control values. Lastly, since the TSH response was lower both during the early and late phases of the infusion, the decrease in ΔTSH to a bolus of TRH during fasting appears to represent one manifestation of a more general suppression of TSH neogenesis associated with caloric deprivation.     

Alternate C3 pathway activation in pneumococcal glomerulonephritis

Glomerulonephritis following pneumococcal infection has been observed, but possible immunopathologic mechanisms have not been adequately explored. Multiple serologic studies as well as light, immunofluorescence and electron microscopic evaluation of kidney biopsy tissue from a 4 year old girl with pneumococcal glomerulonephritis were performed. Clinical studies at the onset of disease showed normal serum C3 and C4 levels (third and fourth components of complement) with progression to selective C3 hypocomplementemia from days 2 to 58. A serum factor capable of breaking down C3 in normal human serum was present during the period of maximum C3 hypocomplementemia. Renal glomerular histology revealed a mesangial proliferative glomerulonephritis. Glomerular bound C3 and type 14 pneumococcal antigen were associated with similar, but less extensive, deposits of properdin. Minimal immunoglobulin M (IgM) and C4 were seen, but immunoglobulin G (IgG) and fibrinogen were absent. Ultra-structurally, subepithellal “humps” and intramembranous electron dense deposits were noted. It is hypothesized that the pneumococcal polysaccharide can activate the alternate complement pathway and may be responsible for a limited course of glomerulonephritis.     

Ventricular function curves from the cardiac response to angiographic contrast: a sensitive detector of ventricular dysfunction in coronary artery disease

Fifty cardiac catheterizations were performed in 44 patients undergoing evaluation for coronary artery disease. Ventricular function curves (VFC) were constructed by plotting the left ventricular end-diastolic pressure (LVEDP) and stroke work index (SWI) before and three to four minutes after a standard left ventricular angiogram. In an attempt to delineate the mechanism that produces changes in the post-angiogram LVEDP and SWI, 13 individuals (Group A) with no evidence of cardiac disease were compared to 14 patients with severe coronary artery disease (Group B). Cardiac output and LVEDP increased in both groups after angiogram. The increase in cardiac output was less and the increase in LVEDP greater in Group B. V_max. did not change significantly after angiogram in either group. Other measures of contractility ($\text{dp}\text{dt}$ max., peak Vce, and dp/dt/40 mm. developed pressure) changed appropriately for the large changes in preload seen after angiogram.Thirty-seven studies in patients with coronary artery disease demonstrated that VFC obtained from the cardiac response to contrast are more sensitive than resting LVEDP or ejection fraction in detecting left ventricular abnormality.VFC can be obtained from the ventricular response to angiographic contrast because of the increase in venous return produced by the hyperosmotic effect of contrast. Depressed curves occur in patients with coronary artery disease because of their stiff ventricles and not because of depression of myocardial contractility.     

Parameters of thyroid function in patients with active acromegaly.

In order to determine if acromegaly per se may be associated with abnormalities in thyroidal economy, serum thyroxine-binding globulin (TBG), resin T3 uptake, total and free T4, T3, and reverse T3 concentrations were measured in 21 patients with active acromegaly. Mean (+/- SE) total T4, T3, and reverse T3 levels were 7.1 +/- 0.2 microgram/dl, 111 +/- 4 ng/dl, and 45 +/- 2 ng/dl, respectively, and the mean TBG concentration was 3.6 +/- 0.2 mg/dl. Similarly, mean free T4, T3, and reverse T3 concentrations were 2.4 +/- 0.09 ng/dl, 383 +/- 22 pg/dl, and 118 +/- 7 pg/dl, respectively. None of these values is significantly different from normal and the thyrotropin response to thyrotropin-releasing hormone was also normal. In contrast to several earlier reports, these data suggest that parameters of thyroid function are generally normal in patients with active acromegaly.     

Hemiagenesis of the thyroid gland.

Three patients with hemiagenesis of the typhoid gland are described. One was clinically euthyroid, whereas the other two were more unusual in that one had coincident Graves' disease with thyrotoxicosis, and one had primary myxodema. In all three cases diagnosis of hemiagenesis was established by the administration of thyroid-stimulating hormone (TSH). The literature on hemiagenetic thyroid glands with and without associated thyroid disease is reviewed. Although the anomaly is uncommon, awareness and recogniton of its existence may clarify an otherwise puzzling clinical thyroid evaluation, and thus possible avert an unnecessary surgical procedure.     

Skeletal responsiveness in pseudohypoparathyroidism. A spectrum of clinical disease

Three cases of pseudohypoparathyroidism with roentgenographic evidence of hyperparathyroid bone disease are described. Renal resistance to exogenous parathyroid hormone (PTH), the hallmark of pseudohypoparathyroidism, was documented by markedly blunted or absent urinary phosphate and cyclic AMP responses to parathyroid extract. At the time of diagnosis all patients were hypocalcemic and hyperphosphatemic with elevated serum alkaline phosphatase levels and subperiosteal resorption noted on skeletal films. Bone biopsy in one patient revealed a histologic appearance consistent with hyperparathyroidism. Serum PTH levels, measured in two patients while they were hypocalcemic, were elevated. None of the patients had short stature, brachydactyly, subcutaneous calcification or mental deficiency. These cases are compared to the 15 well-documented cases previously reported. The presently available information on pseudohypoparathyroidism indicates a variable skeletal response to PTH mediated by several factors extrinsic to bone and suggests that pseudohypoparathyroidism with hyperparathyroid bone disease is one extreme of a clinical spectrum of skeletal responsiveness to PTH. This disorder is part of an expanding clinical picture which makes pseudohypoparathyroidism a diagnostic consideration in any patient with unexplained hypocalcemia, hyperphosphatemia, elevated alkaline phosphatase levels or metabolic bone disease.     

Testicular dysfunction in untreated Hodgkin's disease

Gonadal function was examined in 19 young men with Hodgkin's disease before therapy and compared with that of 11 men with other malignancies, 13 men with primary testicular failure, and 19 normal men of similar age. Total (p less than 0.01) and free (p less than 0.05) testosterone levels were decreased in Hodgkin's disease. In those with advanced (stage III + IV) and symptomatic (B), Hodgkin's disease serum testosterone levels were indistinguishable from those in primary testicular failure, yet serum levels of luteinizing hormone were normal. Moreover, the acute response of serum testosterone to exogenous human chorionic gonadotropin (HCG) was significantly greater in Hodgkin's disease than in primary testicular failure (p less than 0.03). These data and the finding that basal serum follicle-stimulating hormone levels are significantly lower than normal in Hodgkin's disease (p less than 0.05) suggest that the cause of pretreatment hypogonadism in Hodgkin's disease is not simple primary testicular failure. Total sperm count was decreased in 40 percent of men with Hodgkin's disease but in none of the men with other malignancies (p less than 0.05), suggesting specific seminiferous tubular dysfunction in Hodgkin's disease. However, motility was abnormal in 69 percent of men with Hodgkin's disease and 60 percent of those with other malignancies, suggesting that this is a nonspecific effect of cancer. Serum prolactin levels were significantly higher than normal in Hodgkin's disease (p less than 0.05) but not in other malignancies. Our findings suggests that the cause of testicular dysfunction that is present before treatment of Hodgkin's disease is complex, perhaps involving both pituitary and gonadal abnormalities.