Growth Hormone Should Be Used Only for Approved Indications

We are writing to express our strong disagreement with the editorial by Sonksen (1), who commented on a review of the syndrome of adult GH deficiency by Melmed (2). The review focused on adult GH deficiency in the context of patients with pituitary disease, it noted known diagnostic confounders such as obesity and aging, and it emphasized that most of the studies that have measured improvement in response to GH therapy have primarily involved subjects who have multiple pituitary hormone deficits and structural evidence of hypothalamic/pituitary disease, thereby making it difficult to conclude that patients with isolated, idiopathic GH deficiency and no structural changes will derive the same benefit. In his editorial, Sonksen criticizes Melmed's recommendation that strict GH stimulation testing criteria be used to establish the diagnosis of isolated GH deficiency and expresses the viewpoint that a decision to initiate GH treatment can be based solely on clinical criteria. Sonksen then deviates significantly from a discussion of the issues in that review and expresses opinions that support the unsubstantiated use of GH therapy in normal elderly patients and the use of GH to enhance athletic performance.

First, the concept that GH should be administered therapeutically to elderly adults without structural evidence of hypothalamic/pituitary disease is completely unsubstantiated, with most studies suggesting that the side effects outweigh any purported benefits (3). Sonksen argues that there is a clear rationale for prescribing GH for the prevention of frailty. He also states that GH is relatively safe in the elderly because side effects are only noted when very high doses were used. However, in the published trials that he cites, relatively high doses were used and a significant number of side effects occurred. Sonksen concludes that the physiological decline in GH in the elderly can be used as the basis for predicting a positive response to therapy. He argues that the decision whether or not to treat the elderly is best made by a family doctor, but the clinical criteria on which a physician would base this decision are not delineated. He abjures biochemical testing to make the diagnosis, which he claims is just a “semantic issue.”

The studies cited by Sonksen do not conclude that this question has been answered definitively. Sonksen repeatedly cites the study by Rudman et al (4), omitting the fact that no functional benefits were observed in the participants in that study. Similarly, the study by Papadakis et al (5) also failed to detect any functional benefits, but reported significant side effects in the elderly treated with exogenous GH in a placebo-controlled fashion. Similarly, the report by Blackman et al (6) did not show clear benefit. The study of Giannoulis et al (7) evaluated patients for 6 months who received GH alone or GH plus T. GH was administered in a dose that increased IGF-1 2-fold, and clearly some of the subjects were receiving doses that were greater than those required for physiological replacement. Maximal aerobic capacity was used as a measure of physical performance. Importantly, only the group that received GH plus T had a significant increase. GH alone had no effect on these exercise parameters. Therefore, this study certainly does not definitively answer the question as to whether this would be an acceptable long-term treatment in this population. In a study by Sattler et al (8) in which patients received GH plus T, IGF-1 increased 2.5-fold. However, although strength testing improved in the subjects receiving both hormones, there was no difference in the degree of improvement compared to subjects receiving T alone (8). Therefore, there is no evidence that GH administration had any added benefit. The study by Weissberger et al (9) analyzed the effect of GH in a dose of 16 μg/kg, which is clearly a pharmacological dose, in elderly subjects in a catabolic state. The subjects were undergoing hip surgery and did have a modest improvement in strength. However, although three functional tests were conducted, only one showed significant improvement. It is important to point out that this effect was only achieved using pharmacological doses of GH that induced side effects and increased IGF-1 to supraphysiological levels. Edema developed in 13 of 17 patients, and joint pain developed in nine of 17 patients. IGF-1 was greater than the upper limit of normal in 11 of 17 patients. Importantly, this study investigated the role of GH in accelerating recovery from surgery, and the results may not be applicable to noncatabolic elderly patients. We conclude that the studies cited by Sonksen do not support the concept that GH administration to elderly subjects who are secreting relatively low amounts of GH improves functional performance. It is also clear that in elderly subjects, whether they are GH deficient or not, the potential for toxicity is high. Furthermore, prospective determination of what is a normal IGF-1 response for subjects who are not clearly GH deficient is quite problematic. It is certainly possible that increasing IGF-1 levels to the upper limit of the normal range may induce side effects in some individuals, and it is not clear what degree of incremental IGF-1 response is required to achieve efficacy. Therefore, GH therapy in the frail elderly in the absence of hypothalamic-pituitary disease and documented GH deficiency by acceptable stimulation-testing methods cannot be recommended at the present time. We are concerned that this editorial will provide a justification for “antiaging” doctors to prescribe GH inappropriately as an antiaging tonic, a practice that is proscribed by law.

Second, Sonksen's editorial indirectly endorses the concept that administration of GH in the hope of gaining competitive athletic advantage is justified. It is clear that when GH is administered in pharmacological doses to normal subjects, there is often a change in body composition, with decreased fat and, in some trials, an increase in lean body mass. In some studies this has been shown to result in a slight improvement in physical performance, although in most studies improved performance has not been demonstrated (10). Two of the studies conducted in athletes that he cites (11, 12) concluded that the effect on performance was unclear. Graham et al (13) in particular state that “the question whether GH enhances athletic performance has not been definitively answered.” Furthermore, in those studies, GH was administered in pharmacological doses, and therefore the results cannot be directly compared to patients with GH deficiency in whom the goal is “physiologic replacement therapy.” Finally, exercising lactate levels were higher in subjects who received GH in two studies, a finding that is associated with decreased exercise stamina (10). Thus, whereas Sonksen admonishes us that athletes “know” what works to improve performance, carefully controlled studies have not shown substantial ergonomic improvements with GH supplementation. We are concerned that Sonksen's editorial will give credibility to an illegal practice.

In summary, administration of GH to patients with known hypothalamic/pituitary disease and multiple pituitary hormone deficits has been shown to have therapeutic benefit and safety in trials that have followed patients for periods as long as 15 years (14). However, the benefit that can be attained in patients with age-based lower levels of GH (so-called “somatopause”), particularly in those patients with no known structural hypothalamic/pituitary disease, has not been established and cannot be supported. We strongly believe that this recommendation is not scientifically based. Finally, the use of GH administration in athletes, an illegal practice, has little scientific or ethical justification, and support of this ban by the endocrine community must remain.