Adults with severe growth hormone deficiency have more fat and less lean tissue than their peers, an adverse lipid profile and reduced wellbeing, and hypopituitary adults receiving conventional replacement without growth hormone were found to have excess cardiovascular mortality (Rosén and Bengtsson, 1990). The first placebo-controlled trials of recombinant growth hormone, published in 1989, showed that treatment increased lean mass, reduced fat mass and improved metabolic measures (Salomon et al., 1989; Jørgensen et al., 1989). Replacement has been licensed for adults in many countries since the 1990s.

Those trials borrowed their dosing from children. A child's dose is scaled to body size because the aim is growth and the child is growing. An adult is not, and the argument of this article is that scaling the adult dose to weight has neither a physiological rationale nor a good safety record. Most specialists already titrate doses individually, and guidelines endorse it. The position is restated here because weight-based thinking persists in product labeling, in reimbursement rules and in the design of some trials, and because new long-acting preparations have reopened the question of how doses are chosen.

The side effects of growth hormone in adults are mostly those of sodium and water retention: peripheral edema, arthralgia, myalgia, paresthesia and carpal tunnel syndrome. They are dose related. In the early experience they were frequent enough to lead many participants to reduce or stop treatment, and analysis of who was affected showed that older and heavier patients, and those whose deficiency began in adult life, were the most susceptible (Holmes and Shalet, 1995). The plain reading of that experience is that the doses were too high for many of the people who received them, and that the excess was not spread evenly: it fell on exactly those patients for whom a weight-based calculation produced the largest number.

The pattern is what physiology predicts. Spontaneous growth hormone secretion declines steadily through adult life, and it is suppressed in obesity; age and relative adiposity are independent negative determinants of secretory burst frequency and amplitude (Iranmanesh et al., 1991). A rule that multiplies a fixed quantity by body weight therefore runs opposite to nature. It assigns the largest dose to an obese man of seventy, who if healthy would secrete very little, and the smallest to a lean woman of twenty-five, who would secrete a good deal.

The argument rests on three observations. Requirements vary widely between patients, the variation is partly predictable from characteristics other than weight, and a titrated regimen has been shown to work. Figure 1 gathers the factors that bear on the dose an individual needs, and Table 1 states the direction in which each acts.

A view held this widely should still be tested against the arguments on the other side. Three deserve an answer.

For clinicians the practical message is already familiar: start low, increase slowly, and let age, sex, estrogen route, adiposity and glucose tolerance shape expectations of where the dose will settle. A change in estrogen therapy, including a switch between oral and transdermal routes, is a reason to recheck IGF-I, as are substantial weight change and the passage of years, since a dose that suited a patient at forty may be excessive at sixty-five. For those who write formularies and reimbursement criteria, dose limits expressed per kilogram should be replaced by limits framed around titration. For trialists, weight-based arms should no longer be used as comparators in adults.

The research agenda is the larger matter. Individualized dosing in its present form means titration to a biochemical target that has never been tested against hard outcomes. Whether aiming for the middle or the upper part of the IGF-I range yields better results for body composition, quality of life, fracture or cardiovascular events is unknown. So is the best approach in older adults, in whom benefit is least certain and side effects most frequent. Individualization should be understood as a direction of travel, with the targets still to be validated.

Dosing adult growth hormone replacement by body weight was a historical accident with predictable consequences. Secretion and sensitivity depend on age, sex, estrogen exposure and adiposity in ways that weight does not track and partly inverts. Titration from a low starting dose, guided by clinical response and IGF-I, gives comparable benefit with fewer adverse effects and is recommended by every major guideline. Its tools are imperfect and its targets rest on consensus. The remedy is to improve them, and the evidence offers no reason to go back.

To be declared by the author(s) before publication.

To be declared by the author(s) before publication.