Methods: We performed MEDLINE and EMBASE searches for English-language studies published through November 2005 using terms such as growth hormone and aging. We included randomized controlled trials that compared GH to no-GH therapy or GH with lifestyle interventions (exercise ± diet) to lifestyle interventions-alone that provided GH for at least two weeks to community-dwelling subjects aged ≥ 50 years with a BMI of ≤ 35 kg/m2. We excluded studies that evaluated GH as treatment for a specific illness. Two authors independently reviewed articles and abstracted data. We calculated weighted and standardized mean differences between GH and no-GH therapy on 12 key clinical outcomes including weight, fat mass, lean body mass, serum lipids, bone density, maximal oxygen capacity, and fasting glucose and insulin using a random effects model. We also synthesized data on key adverse events, including edema and new diabetes mellitus.

Results: Our search yielded 3,028 titles. 31 articles describing 19 unique populations met inclusion criteria. 226 GH-treated subjects (110 person-years) completed their respective studies. Subjects were elderly (mean ± SD, 68±6 years) and overweight (BMI, 28±2 kg/m2). Initial GH dose (15±9 mcg/kg/d) and treatment duration (27±9 weeks) varied. In GH-treated compared to non-GH-treated subjects, fat mass decreased (1.9 kg) and lean body mass increased (2.0 kg) (p<0.0001); total cholesterol decreased (0.25 mmol/l, p=0.04), although not significantly after adjusting for changes in body composition. Other outcomes, including bone density and serum lipids, did not change. We found no significant differences in clinical outcomes between shorter (≤ 26 weeks of GH therapy) or longer (> 26 weeks) trials, or between GH-treated subjects who received or did not receive lifestyle interventions. GH-treated subjects had higher rates of edema and arthralgias (p<0.0001) and tended to have worsened glucose control (p=0.1).

Conclusions: The cumulative published experience of GH therapy in the healthy elderly is limited. Although GH may minimally alter body composition, it does not improve other clinically important outcomes and may cause considerable harm. Based on the available evidence, GH cannot be recommended as an anti-aging therapy.