Respiratory function test showed severe restrictive pulmonary dysfunction (forced vital capacity 890 ml; forced expiratory volume in 1 s 880 ml). Respiratory muscle strength was assessed by measuring maximal static expiratory (PE_max) and inspiratory (PI_max) mouth pressures, both of which were severely impaired (PE_max 57.5 cm H₂O; PI_max 36.2 cm H₂O). Muscle MRI detected no abnormal signal and neuromuscular diseases were not suspected. Sleep apnoea syndrome was not found by apnomonitor test. After NPPV was started, he was discharged in April 2008. He was referred to the Endocrinology Department in June 2008. MRI, as shown in figure 1, revealed pituitary stalk agenesis. Fasting plasma hormone levels were as follows: GH 0.07 ng/ml, insulin-like growth factor 1 59.0 ng/ml, luteinising hormone (LH) undetectable, follicle-stimulating hormone (FSH) undetectable, testosterone 0.13 ng/ml, adrenocorticotropic hormone (ACTH) 22 pg/ml, cortisol 8.9 μg/dl, thyroid-stimulating hormone (TSH) 1.08 μU/ml, F-T3 3.12 pg/ml, F-T4 1.31 ng/dl. While plasma levels of ACTH increased in response to corticotropin-releasing hormone (baseline: 23 pg/ml, at 30 min 56 pg/ml), urinary excretion of free cortisol was low (18.6 μg/day). Plasma levels of GH and LH/FSH did not increase in response to growth hormone-releasing peptide-2 (GHRP) and luteinizing hormone-releasing hormone (LHRH), respectively (data not shown).

After discharge in April 2008, PO₂ ranged between 75.9 and 85.0 mm Hg, and PCO₂ ranged between 46.9 and 47.9 mm Hg until August 2008 when GH substitution started. GH therapy for 6 months increased PO₂ to 93.1 mm Hg while it decreased PCO₂ to 43.5 mm Hg (figure 2). Addition of testosterone or hydrocortisone showed no apparent additional effects on respiratory function. Respiratory muscle strength, assessed by the ratio of PE_max and PI_max, was increased after substitution of hormones (figure 3).

It is suggested that severe hypothyroidism could result in respiratory failure in some cases.^{3 4} The respiratory muscle strength is reportedly weakened in hypothyroidism and the weakness seems to be proportional to the degree of thyroid dysfunction.⁵ However, this is not the case in our patient as the plasma levels of thyroid hormones were within normal ranges with hormone substitution when he was diagnosed as respiratory failure. On the other hand, it is also reported that adult patients with GH deficiency showed moderate impairment of ventilatory function⁶ and that GH replacement therapy reversed that function.⁷ While it has been postulated that respiratory dysfunction in patients with GH deficiency might be due to weakened respiratory muscle strength and a reduction of lung volumes,⁷ there are no reports, to our knowledge, showing that GH deficiency resulted in severe respiratory failure as in our case. In treating patients with hypopituitarism, it is common practice to substitute hydrocortisone or testosterone, if necessary, before starting GH replacement. However, in our case, the patient refused replacement of hydrocortisone or testosterone at first and we substituted GH first. This resulted in an improvement of respiratory function, suggesting that a GH deficiency caused the respiratory dysfunction. As the ratio of PE_max and PI_max, which reflects the respiratory muscle strength,⁷ increased after the hormone therapy, it is possible that the improvement in respiratory function was due to increases in respiratory muscle strength. It is unclear whether deficits of glucocorticoids and androgens could affect respiratory function.³ In our case, the addition of hydrocortisone and testosterone had no apparent effects on respiratory function, although we cannot exclude the possibility that replacement of these hormones contributed to the maintenance of improved muscle strength. While patients with hypopituitarism may not often have respiratory dysfunction, our case report suggests that we should examine pituitary hormones, including GH, when the cause of chronic respiratory dysfunction is not clear.