Growth hormone (GH) replacement in adults with severe GH deficiency (GHD) has well established beneficial effects on the cardiovascular risk profile. Whether this is associated with reduced mortality is still unclear. The present study addresses this question by analysing mortality data in KIMS (Pfizer international metabolic database) and is based on 9610 patients (47% females), accumulating 37882 patient-years of follow-up. There were 257 recorded deaths against 255.1 expected. Overall all-cause mortality rate was not different from the population rate (standardized mortality ratio (SMR) 1.01, 95% CI 0.891.14). However, this ratio varied significantly with attained age at follow-up (<45 years: SMR 4.35, 4569 years: SMR 0.84, ≥70 years: SMR 0.64). Female gender, craniopharyngioma or malignancy as underlying disease, pre-treatment with radiotherapy, and diabetes insipidus were associated with a significantly increased mortality rate. No significant association was seen for age at onset of pituitary disease, number of pituitary deficits in addition to GHD, as well as the presence of ACTH-, TSH-, or LH/FSH-deficiency. IGF-I SDS levels at one year of GH replacement therapy, reflecting the adequacy of GH treatment, were significantly associated with mortality (P<0.0001). Each unit increase in IGF-I SDS was associated with a 20.7% decrease of SMR. Poisson regression analyses identified gender, attained age, etiology of GHD, pre-treatment radiotherapy, diabetes insipidus and IGF-I SDS after 1 year of GH treatment as independent significant factors influencing all-cause mortality. Overall, deaths from cardiovascular and cerebrovascular disease and malignancies were most common; their rate, however, was not different from the normal population rate. In conclusion, the significant association between IGF-I SDS levels during GH treatment and mortality rates provides the first direct evidence for a beneficial effect of GH replacement on mortality in GHD patients.
03 - 07 May 2008
European Society of Endocrinology