Final height outcome and value of height prediction in boys with constitutional delay in growth and adolescence treated with intramuscular testosterone 125 mg per month for 3 months.

Objectives: Constitutional delay in growth and adolescence (CDGA) is common in boys, some of whom request treatment to accelerate growth and attainment of secondary sexual characteristics. The aims of this study were to confirm that a 3-month course of intramuscular testosterone oenanthate does not impair final height in boys with CDGA, and to determine the accuracy of height prediction in this condition.

Design And Patients: Boys with CDGA who had attended the growth clinic, who were now at or close to final height and who had received either testosterone or declined treatment, were identified by retrospective case note analysis. Bone age assessment was carried out by a single observer, using the RUS (TW2) method of Tanner and Whitehouse.

Measurements: The following auxological data were extracted from the case records: age, bone age, height, pubertal stage, parental heights and predicted final height. All subjects were then measured at age 19 years or greater. The main outcome measures were comparison of final height in treated and untreated boys; final height comparison with mid-parental height and with height prediction [RUS (TW2) method] at initial assessment and at subsequent review.

Results: Sixty-four boys met the inclusion criteria, of whom 41 subjects had received testosterone and 23 were untreated. There were no significant differences between the groups (treated mean/SD vs. untreated mean/SD; P-value) in age (14.3/0.7 vs. 14.0/1.1; 0.13), height (144.7/6.2 vs. 144.2/6.2; 0.79), mid-parental heights (170.4/5.5 vs. 171.1/4.5; 0.59), and bone age (12.0/1.2 vs. 12.3/1.3; 0.36). Final heights in both groups (168.9/6.0 vs. 168.2/3.5; 0.65) were closely related to predicted final heights (170.0/5.0 vs. 168.1/4.1; 0.15) and only slightly less than mid-parental heights. Only three subjects had final heights below the initial height prediction range.

Conclusions: Our data support the hypothesis that this treatment regime does not adversely affect the final height achieved in constitutional delay of growth and adolescence and that height prediction, assessed by a single observer, is a useful and accurate tool.