Search

To investigate cardiac effects of excess and deficiency of growth hormone (GH) 20 acromegalic subjects and 11 adult patients with GH deficiency were studied by means of a non invasive method, the Doppler echocardiography. The results obtained in the group of patients were compared with those of 2 groups of 20 and 12 normal subjects, respectively, age and sex matched. The age of the acromegalic patients ranged from 20 to 62 years. Nineteen patients were considered to have active acromegaly at the time of the study. Mean duration of disease since treatment was 12 +/- 5 years (range 5 to 24 years). The age of GH deficient adults ranged from 21 to 33 years. All these patients have been treated with extractive GH over 9 years and the therapy withdrawal was performed at least 3 years before entering the study. In the acromegalic patient group, a subgroup including 9 patients with mild to moderate hypertension was considered. All subjects gave informed consent and the study protocol was approved by the Ethical Committee of the Medical School of Naples. Right ventricular free wall thickness resulted significantly increased in acromegalic patients (8 +/- 2 versus 4 +/- 1 mm; p0.001). Left ventricular mass index was augmented both in the whole group and in the subgroup of normotensive acromegalics as compared with normals (134 +/- 33 and 115 +/- 20 versus 80 +/- 18 g.m-2; p0.01). Ejection phase indices were normal in patient group, while impaired left and right ventricular diastolic filling was found.(ABSTRACT TRUNCATED AT 250 WORDS)

The beta-endorphin-like-immunoreactivity (beta-ELI) has been evaluated both in plasma and in cerebrospinal fluid (CSF) in 30 patients during trans-sphenoidal surgery. Blood and liquoral samples were collected in five conditions: (1) "reference", (2) "pain", (3) "analgesia", (4) "end", and (5) "24th hour". A significant rise of both plasma and liquoral beta-ELI levels (p less than 0.00001 and p less than 0.08, respectively) when compared to basal ones occurred following the painful stimulation due to the divarication of the nasal mucosa by speculum. A significant decrease (p less than 0.01) was noticed for plasma concentrations at the third sample followed by a new significant increase at the end of the operation, (p less than 0.05 when compared to the third sample and p less than 0.01 when compared to the reference sample). In CSF, beta-ELI levels decreased at the third sample (p less than 0.01 when compared to the painful levels) and at the end of surgery (p less than 0.01, p less than 0.01 and p less than 0.05 vs first, second and third samples, respectively). Twenty-four hours after surgery either plasma and liquoral beta-ELI levels decreased (p less than 0.05). The modifications of the opiatergic system after acute painful stimuli should be, hence, characterized by an early rise followed by a progressive decrease of beta-ELI concentrations. The increase of plasma beta-ELI levels, at the end of surgery, could be due to pituitary manipulation with massive release in the peripheral blood.

To investigate cardiac effects of excess and deficiency of growth hormone (GH) we studied twenty acromegalic subjects and eleven adult patients with GH deficiency by means of a non invasive method, the Doppler echocardiography. The results obtained in the group of patients were compared with those of two groups of twenty and eleven normal subjects, respectively, age and sex matched. The age of the acromegalic patients ranged from 20 to 62 years. Nineteen patients were considered to have active acromegaly at the time of the study. Mean duration of disease since treatment was 12 +/- 5 years (range 5 to 24 years). The age of GH deficient adults ranged from 21 to 33 years. All these patients have been treated with extractive GH over nine years and the therapy withdrawal was performed at least three years before entering the study. In the group of acromegalic patients, a subgroup including nine patients with mild to moderate hypertension was considered. All subjects gave informed consent and the study protocol was approved by the Ethical Committee of the Medical School of Naples. Right ventricular free wall thickness resulted significantly increased in acromegalic patients (8 +/- 2 vs 4 +/- 1 mm; p0.001). Left ventricular mass index was augmented both in the whole group and in the subgroups of normotensive and hypertensive acromegalics as compared with normals (134 +/- 33 p0.001, 115 +/- 20 p0.01 and 156 +/- 31 p0.001 vs 80 +/- 18 g.m-2). Ejection phase indices were normal in patient group, while impaired left and right ventricular diastolic filling was found.(ABSTRACT TRUNCATED AT 250 WORDS)

The aim of the current study was to evaluate the effectiveness of a long-term treatment with sodium valproate in 19 patients with Cushing's disease. Before therapy beginning, the patients were subjected to acute test with 600 mg sodium valproate. Then, they were subjected to a 3-month therapy with sodium valproate at the dose of 600 mg/day before surgery (presurgical study). The 7 patients not surgically cured were subjected again to a 3-month therapy with sodium valproate at the dose of 600 mg/day after surgery (postsurgical study). Circulating ACTH and cortisol and urinary free cortisol levels were evaluated before and monthly after the beginning of the therapy. A decrease of plasma ACTH and serum cortisol levels greater than 50% of baseline was considered as positive response to acute test whereas the normalization of plasma ACTH, serum cortisol and urinary free cortisol levels and the clinical remission were considered as positive response to the long-term treatment. At acute test, 8 patients were considered responders and 11 patients non-responders. In no patient plasma ACTH, serum cortisol and urinary free cortisol were normalized during the long-term treatment. Urinary free cortisol levels significantly decreased (483.2 +/- 33.8 vs 699.4 +/- 67.0 micrograms/24 h), whereas plasma ACTH (302.8 +/- 17.7 vs 183.3 +/- 25.0 ng/l) and serum cortisol (466.5 +/- 23.2 vs 356.7 +/- 19.6 micrograms/l) significantly increased during sodium valproate administration in the 19 patients enrolled in the presurgical study. Plasma ACTH (247.7 +/- 22.3 vs 168.6 +/- 15.0 ng/l), serum cortisol (387.4 +/- 35.8 vs 282.0 +/- 16.0 micrograms/l) and urinary free cortisol (370.9 +/- 70.6 vs 261.3 +/- 37.8 micrograms/24 h) levels significantly increased in the 7 patients enrolled in the postsurgical study. No patient had clinical remission of Cushing's disease. In conclusion, the current study showed that long-term therapy with sodium valproate is not useful in the therapeutic management of Cushing's disease neither as alternative nor as adjunctive therapy to surgery.

Complex and bidirectional relationships operate between the hypothalamic-pituitary-adrenal (HPA) axis and the immune system (IS) and either in vivo or in vitro evidence supports a physiological role of the HPA axis-IS network. A part of the well-known pharmacological effects of glucocorticoid hormones (GC) as immunodepressive agents, the direct effects of many HPA axis hormones on IS functions are actually documented also in physiologic conditions. Conversely, numerous IS soluble mediators are reported to affect the HPA axis functions at various steps of HPA axis regulation, in both physiologic and pathologic conditions. Stress and aging may represent two paradigmatic conditions to show the relevance of the bidirectional network between HPA axis and IS, as in both HPA activation and IS impairment are frequently coexistent. Finally, in the context of the wide spectrum of HIV-related HPA axis abnormalities, a case of a Cushing's syndrome associated to an acquired immunodeficiency syndrome (AIDS-related complex) in a 24-year-old homosexual drug abuser is reported.

The aim of this study was to evaluate the effect of acute human corticotropin (ACTH)-releasing hormone (CRH) administration (100 micrograms, as i.v. bolus) on tumor necrosis factor-alpha (TNF alpha) levels in the inferior petrosal sinuses and in the peripheral blood of 7 patients with Cushing's disease subjected to diagnostic inferior petrosal sinus sampling. Blood samples for ACTH, beta-endorphin (beta-EPH) and TNF alpha were collected from inferior petrosal sinuses and periphery simultaneously. In addition, TNF alpha concentrations were measured after CRH administration (10 nmol/l, 100 nmol/l and 1 mumol/l) in culture medium from primary cultures obtained in 3 of 7 patients. At baseline, plasma ACTH and beta-EPH levels were significantly higher in the inferior petrosal sinus ipsilateral to the ACTH-secreting adenoma than in the contralateral one and in the periphery (p < 0.001) whereas no significant difference was found as far as serum TNF alpha levels were concerned. CRH administration caused a significant increase of ACTH (p < 0.001), beta-EPH (p < 0.01) and TNF alpha (p < 0.01) levels greater in the ipsilateral inferior petrosal sinus than in the contralateral one and in the periphery. In addition, CRH increased ACTH, beta-EPH and TNF alpha levels in the culture medium of three ACTH-secreting tumors at the doses of 100 nmol/l and 1 mumol/l (greater than 300, 200 and 110% of baseline pretreatment incubation levels, respectively). These data suggest that CRH may increase TNF alpha concentrations in the inferior petrosal sinus ipsilateral to the ACTH-secreting adenoma and in the peripheral blood as well. In addition, it stimulated TNF alpha release both in vivo and in vitro. These findings suggest the possibility that an imbalanced intrapituitary TNF alpha production can be detected in ACTH-secreting adenomas.

This study was designed to assess systolic wall stress and ventricular function in patients with deranged growth hormone secretion, in an attempt to elucidate the mechanisms of growth hormone interaction with heart performance.A case-control study.Thirty patients with active acromegaly, free of diabetes mellitus and coronary artery disease, and 25 subjects with congenital growth hormone deficiency were studied. Twelve growth hormone-deficient subjects were reevaluated after 12 months of recombinant human growth hormone therapy. Two groups of 30 normal subjects each were used as controls for the acromegalic and growth hormone-deficient patients, respectively.In the acromegalics, end-systolic wall stress was reduced (-20%; P0.01) due to ventricular wall thickening (+ 26%; P0.001), whereas cardiac output was significantly increased (+ 20%; P0.01). The velocity of fibre shortening was unchanged. In growth hormone-deficient subjects, end-systolic wall stress was markedly increased (+ 38%; P0.001) due to a significant reduction of ventricular wall thickness (- 28%; P0.001), whereas cardiac output was significantly decreased (-44%; P0.001). Replacement therapy with recombinant human growth hormone produced a partial correction of wall thickness and stress. Consequently, systolic performance and cardiac output improved significantly.This study demonstrates that growth hormone plays a role in the control of cardiac wall stress and performance through a mechanism mediated by the effect of growth hormone on myocardial tissue growth. The data may have therapeutic implications in cardiac diseases that lead to heart failure.

To investigate left and right ventricular involvement in acromegaly, 20 patients were studied by Doppler echocardiography. Nine of them had systemic hypertension. Right ventricular free wall thickness was significantly increased in acromegalic patients (8 +/- 2 vs 4 +/- 1 mm; P < 0.001). Left ventricular mass index was augmented both in the whole group and in the subgroup of normotensive acromegalics, as compared with normals (134 +/- 33 and 115 +/- 20 vs 80 +/- 18 g.m-2; P < 0.01). Ejection phase indices were normal in the patient group, while impaired left and right ventricular diastolic filling was found. In fact isovolumic relaxation time was prolonged (118 +/- 21 vs 78 +/- 12 ms; P < 0.001), ratio of early to late mitral (0.9 +/- 0.3 vs 1.8 +/- 0.5; P < 0.001) and tricuspid (1.0 +/- 0.2 vs 1.4 +/- 0.3; P < 0.001) flow velocities were significantly decreased as compared with controls. Superior vena cava flowmetry was also abnormal showing a marked decrease of diastolic filling wave and, consequently, of the ratio between peak diastolic and peak systolic flow velocity. No significant differences were observed between normotensive and hypertensive acromegalics, except for left ventricular mass index (115 +/- 20 vs 156 +/- 31 g.m-2; P < 0.01). These findings indicate that abnormal diastolic filling patterns of transmitral, transtricuspid, and superior vena cava flowmetry suggesting 'impaired relaxation' associated with increased left and right ventricular mass, frequently occur in acromegaly.

The clinical effects of CV 205-502, a potent and non-ergot-derived dopamine agonist, were investigated in 24 selected patients with hyperprolactinemia previously treated with standard oral bromocriptine, the slow-release oral form of bromocriptine (BRC-SRO) and/or the long-acting injectable form of bromocriptine (BRC-LAR); 14 were chosen because of their resistance to treatment and ten because they were intolerant of the different forms of bromocriptine. A macroprolactinoma was present in seven patients and a microprolactinoma in ten, whereas seven had no radiological images of a pituitary tumor and were classified as having non-tumoral hyperprolactinemia. All the 24 patients were treated with CV 205-502 at a daily dose of 0.075-0.6 mg for 3-12 months. All the patients had gonadal dysfunction and galactorrhea. Basal serum prolactin values ranged from 70 to 1677 ng/ml. CV 205-502 was effective in 11 of the 14 patients resistant bromocriptine, BRC-SRO and BRC-LAR; serum prolactin levels became normal within 6 months and a tumor shrinkage was obtained in five of the seven macroprolactinomas. In general, the drug was effective and well tolerated. Only three patients (two resistant and one intolerant) manifested nausea, vomiting and postural hypotension. In conclusion, this study shows that CV 205-502 is effective in bromocriptine-resistant hyperprolactinemic patients. Furthermore, CV 205-502 has insignificant and tolerable side-effects in patients intolerant of bromocriptine. CV 205-502 can, therefore, be considered a useful and effective drug, and an interesting therapeutic alternative to the ergot-derived dopamine-agonist drugs in use today.

Recent evidence suggests that GH and insulin growth factor-I (IGF-I) play a role in adrenal steroidogenesis. On the other hand, it has been shown that ACTH stimulates IGF-I secretion by cultured fasciculata adrenal cells. Aim of the present study was to investigate the influence of GH administration on adrenal steroids and IGF-I responsiveness to ACTH in children affected with isolated GH deficiency. Ten children (seven males and three females, 5-10 yr old) affected with isolated GH deficiency underwent a synthetic ACTH 1-17 test before and after administration of human recombinant GH at a dose of 4 IU/day sc for 10 days. After the therapy, no significant differences were detected in the responses of cortisol, dehydroepiandrosterone-sulfate, androstenedione, and 17-hydroxyprogesterone to ACTH 1-17, whereas an increased 11-deoxycortisol responsiveness to ACTH 1-17 was noted (P less than 0.005). Surprisingly, IGF-I significantly increased in response to ACTH 1-17 after short-term rGH administration (P less than 0.006). In conclusion, our data indicate that in isolated GH deficiency a short-term GH therapy does not substantially modify the adrenal responsiveness to exogenous ACTH, even if an increased 11-deoxycortisol and an induced IGF-I responsiveness to ACTH were observed.

Aim Since growth hormone (GH) has proven beneficial in experimental heart failure, and the natural history of Duchenne muscular dystrophy (DMD) and Becker muscular dystrophy (BMD) is frequently complicated by the development of dilated cardiomyopathy, we administered GH to six patients with DMD and 10 with BMD, with the evidence of cardiac involvement. Methods and results Patients were randomized to receive for 3 months either placebo or recombinant human GH, in a double-blind fashion. In GH-treated patients, left ventricular (LV) mass increased by 16% in BMD and by 29% in DMD (both \batchmode \documentclass[fleqn,10pt,legalpaper]{article} \usepackage{amssymb} \usepackage{amsfonts} \usepackage{amsmath} \pagestyle{empty} \begin{document} \(p{

Background —The reason why patients with growth hormone (GH) deficiency (GHD) are at increased risk for premature cardiovascular death is still unclear. Although a variety of vascular risk factors have been identified in GHD, little is known regarding vascular reactivity and its contribution to premature arteriosclerosis. Methods and Results —We assessed vascular function in 7 childhood-onset, GH-deficient nontreated patients (age 22±3 years, body mass index [BMI] 25±1 kg/m 2 ) and 10 healthy subjects (age 24±0.4 years, BMI 22±1 kg/m 2 ) by using strain gauge plethysmography to measure forearm blood flow in response to vasodilatory agents. The increase in forearm blood flow to intrabrachial infusion of the endothelium-dependent vasodilator acetylcholine was significantly lower in GH-deficient nontreated patients than in control subjects ( P P P P 2 ) who were receiving stable GH replacement therapy. In these patients, the response to both endothelium-dependent and -independent vasodilators, as well as forearm nitrite and cGMP, release was not different from that observed in normal subjects. Peak hyperemic response to 5-minute forearm ischemia was significantly reduced in GH-deficient nontreated patients (17.2±2.6 mL · dL −1 · min −1 , P −1 · min −1 ) compared with normal subjects (29.5±3.2 mL · dL −1 · min −1 ). Conclusions —The data support the concept that GH plays an important role in the maintenance of a normal vascular function in humans.

The aim of the current study was to evaluate bone metabolism and mass before and after 2 years of GH replacement therapy in adults with childhood or adulthood onset GH deficiency. Thirty-six adults with GH deficiency, 18 with childhood onset, 18 with adulthood onset GH deficiency and 28 sex-, age-, height- and weight-matched healthy subjects entered the study. Biochemical indexes of bone turnover such as serum osteocalcin, serum carboxyterminal telopeptide of type-I procollagen, urinary hydroxyproline/creatinine and deoxypyridinoline/creatinine, of soft tissue formation such as aminoterminal propeptide of type-III and bone mineral density were evaluated. Childhood onset GH deficient patients had significantly decreased bone (osteocalcin: 2.5+/-1.3 vs 6.6+/-4.8 mcg/l, p

OBJECTIVE: The impairment of heart structure and function in adults with childhood onset GH deficiency has been recently described. However, previous echocardiographic studies have reported no differences in cardiac mass and function between adulthood onset GH deficient patients and healthy subjects. DESIGN: The aim of this study was to evaluate cardiac performance in adult patients with childhood and adulthood onset GH deficiency, using equilibrium radionuclide angiography, a method more accurate than echocardiography. PATIENTS: Eleven patients with childhood onset GH deficiency, 9 patients with adulthood onset GH deficiency and 12 age-, gender-, height- and weight-matched healthy subjects entered the study. MEASUREMENTS: All the study population underwent equilibrium radionuclide angiography at rest and during physical exercise. RESULTS: Both childhood and adulthood onset GH deficient patients had an impaired left ventricular systolic performance both at rest (ejection fraction was 55 +/- 6%, 55 +/- 10% and 66 +/- 6% in childhood and adulthood onset GH deficient patients and control group, respectively; P < 0.0001) and during physical exercise (ejection fraction was 54 +/- 9% in childhood onset GH deficient patients, 53 +/- 9% in adulthood onset GH deficient patients and 76 +/- 7% in normal subjects; P < 0.0001). Peak ejection rate was 3.2 +/- 0.8 end-diastolic volume/second, 3.0 +/- 0.6 end-diastolic volume/second and 3.9 +/- 0.8 end-diastolic volume/ second in childhood and adulthood onset GH deficient patients and control group, respectively (P < 0.01). Exercise-induced changes in end-systolic volume were increased in both groups of patients compared with healthy subjects. In contrast, exercise-induced end-diastolic volume changes were not different between GH deficient patients and controls. Resting peak filling rate was 2.6 +/- 0.7 end-diastolic volume/second, 2.5 +/- 0.7 end-diastolic volume/ second and 3.1 +/- 0.3 end-diastolic volume/second in the 2 groups of patients and healthy subjects, respectively (P < 0.05). Reduced exercise tolerance in all patients, as shown by the significantly lower values of peak workload (P < 0.0001), peak rate-pressure product (P < 0.01) and exercise duration (P < 0.0001) was observed. CONCLUSION: Patients affected by GH deficiency have left ventricular systolic dysfunction at rest and during physical exercise, suggesting that GH plays a physiological role in maintaining normal cardiac performance in humans. Furthermore, no difference between childhood and adulthood onset GH deficient patients was found indicating that both group of patients have an impairment of cardiac function.

GABAergic drugs affect PRL secretion in both rat and man. Sodium valproate (SV) inhibits GABA transaminase so increasing the endogenous GABAergic tone. The aim of this study was to evaluate the effects of SV at low and high doses on PRL release in healthy subjects and hyperprolactinemic patients. Fifteen patients with prolactinomas, 8 patients with non-tumoral hyperprolactinemia and 10 healthy subjects were studied: in non consecutive days, all subjects received placebo and SV at the dose of 400 and 800 mg po. Serum PRL levels were assessed 30, 15 and 5 min before and every 30 min for 4 hours after administration. SV at the dose of 400 mg induced a significant decrease of serum PRL in healthy subjects (p < 0.05), whereas no effect was noted in both tumoral and non-tumoral hyperprolactinemia. The administration of 800 mg SV induced a significant decrease of PRL levels in healthy subjects and in patients with non-tumoral hyperprolactinemia (p < 0.05). Conversely, in prolactinomas a paradoxical increase of serum PRL concentration (p < 0.05) was observed 120 min after the administration of the drug. These data confirm the inhibitory activity of SV on PRL release in healthy subjects, and suggest the existence of a partial resistance to GABA in non-tumoral hyperprolactinemia. In prolactinomas, the paradoxical PRL increase after high dose of SV suggests the existence of a complete pituitary resistance to GABA. This finding might be explained by the appearance of the stimulatory effect of GABA at hypothalamic level that could have been unmasked by the lack of pituitary GABA effects on adenomatous lactotrophs.

Two cases of Langerhans cell histiocytosis (LCH) expressing as Hand-Schuller-Christian syndrome with diabetes insipidus, hyperprolactinemia and empty sella are here reported. Up-to-date this four-fold association is lacking in world literature and it is here discussed in the light if LCH is a cancer or the clinical expression of an immunologic disorder.

This study was designed to assess right and left ventricular function in patients with active acromegaly. To this end, 26 acromegalic patients (9 of whom had arterial hypertension) and 15 normal subjects of comparable age and sex distribution were studied by radionuclide angiography at rest and during supine bicycle-ergometer exercise and echocardiography. At rest, the filling rates of left (-19%; P0.005) and right ventricle (-32%; P0.001) were significantly reduced in acromegalic patients, whereas right and left ventricle ejection fractions (EFs) were normal. During physical exercise, EF was considerably lower in the acromegalic patients than in controls. This was true for both left (61 +/- 11% vs. 75 +/- 8%; P0.001) and right ventricle (45 +/- 13 vs. 58 +/- 11%; P0.002). In as many as 73% of patients, EF increased less than 5%, thus fulfilling the criteria for impaired cardiac performance. Left ventricular mass index was 60% greater in acromegalics than in controls (P0.001). A significant difference in left ventricular mass index was also present when normotensive acromegalic patients were compared with controls (P0.001). No significant difference in the indices of systolic and diastolic function was observed between the subgroups of normotensive and hypertensive acromegalics, either at rest or during exercise. The data demonstrate that in uncomplicated acromegaly, besides cardiac hypertrophy, there are also important alterations of systolic and diastolic function of both ventricles, leading to a significant impairment of cardiac performance.

It is known that women suffering from temporal lobe epilepsy may frequently present reproductive endocrine disorders (REDs). We hypothesized that a high occurrence of REDs could be found also in primary generalized epilepsy (PGE), and therefore investigated the hormonal and ovarian echographic profiles in 20 PGE female patients of reproductive age. Fourteen reported normal menstrual cycles, while 6 complained of longstanding menstrual irregularities. All but three patients were receiving antiepileptic drug (AED) therapy. In all subjects, the basal levels of gonadotropins, prolactin, and gonadal steroids were assayed. The response of luteinizing hormone (LH) to gonadotropin-releasing hormone was also investigated and ovarian ultrasonographic findings were evaluated. In five of six patients with menstrual problems (25% of the group), a well-defined RED was diagnosed (polycystic ovarian disease in three cases and hypothalamic ovarian failure in two). The 14 patients with normal menstrual cycles showed an elevation of mean basal follicle-stimulating hormone and prolactin, and a blunting of mean LH response. Our results suggest that a high occurrence of REDs may be found also in PGE. We hypothesize that a neurotransmitter dysfunction might be the common pathogenetic mechanism resulting in both REDs and PGE. The hormonal alterations observed in the patients with normal menstrual cycles seem to support our hypothesis. Previous data seem to rule out a possible AED effect accounting for the hormonal findings observed in our series. However, further studies are needed to confirm our preliminary results.

Cardiac performance was investigated by radionuclide angiography in 11 patients with childhood-onset growth hormone (GH) deficiency and in 12 control subjects. Both at rest and during maximal physical exercise, systolic function was markedly depressed in GH-deficient patients. Ejection fraction rose from 66 +/- 6 to 76 +/- 7% during exercise in control subjects, whereas in GH-deficient patients it remained unchanged or even decreased (55 +/- 6 and 54 +/- 9% at rest and after exercise, respectively; P < 0.01 vs. controls). Cardiac index was significantly lower in GH-deficient patients than in controls, both at rest (2.7 +/- 0.6 vs. 3.7 +/- 0.5 l.min-1.m-2; P < 0.001) and during exercise (8 +/- 1.2 vs. 10 +/- 1.5 l.min-1.m-2; P < 0.01). Five GH-deficient patients were treated with recombinant human (rh) GH for 6 mo at a dose of 0.05 IU.kg-1.day-1. Cardiac index at rest improved from 2.8 +/- 0.6 to 3.3 +/- 0.8 l.min-1.m-2 (P < 0.01) after rhGH. Also, cardiac index response to exercise improved markedly and became similar to that of controls (7.5 +/- 1.2 and 10.1 +/- 1.1 l.min-1.m-2 before and after rhGH, respectively; P < 0.005). Exercise tolerance was impaired in GH-deficient patients and was restored by rhGH treatment. The data support the hypothesis that GH plays an important role in the maintenance of a normal cardiac performance in humans.