Impaired cardiac performance in elderly patients with growth hormone deficiency

Several evidences indicate that GH and/or insulin-like growth factor I (IGF-I) are involved in the regulation of cardiovascular function. In patients with childhood and adulthood-onset GH deficiency (GHD), the impairment of cardiac performance is manifest primarily as a reduction in the left ventricular (LV) mass (LVM), inadequacy of LV ejection fraction both at rest and at peak exercise, and abnormalities of LV diastolic filling. No study has been reported to date in elderly GHD patients that investigated cardiac function. In particular, it is unknown whether cardiac function is modified in accordance with patients' age as a physiological response to aging, as in normal subjects the rate and extent of LV filling are reduced with age. This study was designed to evaluate heart morphology and function, by echocardiography and equilibrium radionuclide angiography, respectively, in rigorously selected elderly patients with GHD but without evidence of other complications able to affect cardiac performance. Eleven patients with hypopituitarism (6 men and 5 women, aged 60-72 yr) and 11 sex- age- and body mass index-matched healthy subjects entered this study. None of the patients and controls presented with or had previously suffered from other concomitant diseases, such as diabetes mellitus, coronary artery diseases, long-standing hypertension, and hyperthyroidism, which could affect cardiac function. All patients had been previously operated on via the transsphenoidal and/or transcranic route for nonfunctioning pituitary adenoma, meningioma, or craniopharyngioma, and 6 of them had been irradiated. Eight patients had FSH/LH insufficiency, 5 had TSH insufficiency, and 6 had ACTH insufficiency, appropriately replaced. All subjects were tested with the combined arginine plus GHRH test showing a GH response below 9 microg/L. No significant difference was found in plasma IGF-I levels (49.2 +/- 8.5 vs. 71.8 +/- 7.5 microg/L) between patients and controls. However, IGF-I levels were lower than the normal range in 8 patients and 3 controls. Interventricular septum thickness (9.1 +/- 0.2 vs. 9.1 +/- 0.2 mm), LV posterior wall thickness (9.1 +/- 0.2 vs. 9.0 +/- 0.2 mm), and LVM after correction for body surface area (97.6 +/- 1.8 vs. 99.9 +/- 1.5 g/m2) were similar in patients and controls. Similarly, the LV ejection fraction at rest was similar in patients and controls (57.1 +/- 2% vs. 63.2 +/- 2.5%; P = NS), and it was normal (> or = 50%) in all controls and in 10 of 11 patients. By contrast, the LV ejection fraction at peak exercise was markedly depressed in elderly GHD patients compared to age-matched controls (51 +/- 2.5% vs. 73.3 +/- 3%; P < 0.001). A normal response (> or = 5% increase compared to basal value) of LV ejection fraction at peak exercise was found in 8 controls (72.7%) and in 2 of 11 patients (18.2%). No difference was found in the peak rate of LV filling, whether peak filling rate was normalized to end-diastolic volume (2.5 +/- 0.2 vs. 2.6 +/- 0.2 end-diastolic volume/s) or stroke volume (4.3 +/- 0.3 vs. 4.0 +/- 0.3 stroke volume/s), between patients and controls. Finally, exercise duration was significantly shorter in elderly GHD patients than in age-matched controls (7.2 +/- 2.1 vs. 9.1 +/- 0.2 min; P < 0.01). In the patient group, the GH peak after arginine plus GHRH test was significantly correlated with the LV ejection fraction at rest (r = 0.822; P < 0.01), whereas IGF-I was significantly correlated with the peak rate of LV filling whether the peak filling rate was normalized to end-diastolic volume (r = -0.863; P < 0.001) or stroke volume (r = -0.616; P < 0.05) or expressed as the ratio of peak filling rate to peak ejection fraction rate (r = -0.736; P < 0.01). Disease duration was significantly correlated with heart rate at peak exercise (r = 0.614; P < 0.05) and with systolic and diastolic blood pressures both at rest (r = 0.745; P < 0.01 and r = 0.650; P < 0.05) and at peak exercise (r = 0.684; P < 0.05 and r =