Bone mineral density (BMD) was evaluated by dual energy x ray absorptiometry in 60 adults (33 males, 27 females; aged 50, range 23-76 years) who were growth hormone deficient from various causes for 10-4 (1-31) years. Adult patients who had acquired growth hormone deficiency before completion of puberty had significantly reduced mean (SEM) BMD compared with age matched healthy controls at the lumbar spine: 0-87 (0.09) v 1-20 (0.03) g/cm2, femoral neck: 0-81 (0-06) v 1-08 (0.04) g/cm2, and Ward's triangle: 0-68 (0.07) v 1-04 (0.05) g/cm2. These values were also reduced compared with those of patients who had received human growth hormone during puberty. Untreated growth hormone deficiency when present during puberty results in reduced adult bone density.

Having noted symptomatic osteoporotic vertebral collapse in young adult survivors of childhood malignancy, bone mineral density (BMD) was examined at three sites by dual‐energy X‐ray absorptiometry in 64 patients treated in childhood for intracranial malignancy (group 1; n = 21) or acute leukaemia (group 2; n = 43). Patients in group 1 were selected for growth hormone deficiency (GHD) by auxological and biochemical criteria before the end of puberty (Tanner stage V). Seven patients (six men; mean (± SEM) age at study, 28.0 ± 2.9 years; mean age at diagnosis, 8.7 ± 1.5 years) in this group had been treated with human pituitary growth hormone (GH) for 1–12 years; and 14 patients (nine men; mean age at study, 26.8 ± 1.0 years; mean age at diagnosis, 10.7 ± 1.4 years) had not received GH. Bone densities in group 1 were normal in the GH‐treated patients at the femoral neck (98.4 ± 3.8% of control), lumbar spine (100.4 ± 6.1% of control) and Ward's triangle (101.0 ± 6.1% of control) but markedly reduced in the untreated group (femoral neck, 81.2 ± 2.6% of control (p= 0.002); lumbar spine, 79.1 ± 4.1% of control (p= 0.04); Ward's triangle, 80.1 ± 3.6% of control (p= 0.01)). The majority of patients in group 2 had been treated for acute lymphoblastic leukaemia (ALL) and were in three subgroups. Fifteen (ten men; mean age at study, 22.1 ± 0.8 years; mean age at diagnosis, 5.7 ± 0.8 years) had no auxological evidence of GHD, ten (six men; mean age at study, 18.8 ± 0.7 years; mean age at diagnosis, 6.6 ± 1.2 years) received GH therapy for a mean of 2.6 years (range, 0.5–5.0 years), while 14 patients (three men; mean age at study, 20.9 ± 0.4 years; mean age at diagnosis, 5.1 ± 0.8 years) had GHD but did not receive GH. A small group of four patients (three men; mean age at study, 23.2 ± 2.1 years; mean age at diagnosis, 8.0 ± 2.3 years) who had been treated for acute myeloid leukaemia (AML) in childhood was also studied. The patients with AML had normal bone densities at all three sites (femoral neck, 106 ± 6.1% of control; lumbar spine, 96.5 ± 3.0% of control; Ward's triangle, 110.8 ± 9.3% of control), as did the patients with ALL who did not have GHD (femoral neck, 102.3 ± 2.9% of control; lumbar spine, 98.6 ± 1.7% of control; Ward's triangle, 108.3 ± 3.2% of control). The patients with ALL and GHD not treated with GH had markedly reduced BMD at all three sites (femoral neck, 90.5 ± 2.6% of control; lumbar spine, 88.4 ± 2.5% of control; Ward's triangle, 94.5 ± 3.7% of control), but those treated with GH had a BMD no different from control (femoral neck, 100.6 ± 3.3% of control; lumbar spine, 95.7 ± 4.6% of control; Ward's triangle, 106.2 ± 4.9% of control). It is concluded that GHD during childhood and adolescence predisposes to osteopenia.

The rise in serum IGF I concentration during continuous subcutaneous insulin infusion (CSII) may be a contributory factor in the deterioration of diabetic retinopathy that sometimes occurs during this treatment but the relation of serum levels to the severity of retinopathy has not been previously studied. In twelve non-obese insulin dependent diabetics (age range: 22-41 yrs) with mean +/- SD duration of diabetes: 14.8 +/- 4.7 yrs, serum IGF I concentration, HbA1 and retinopathy score were estimated prospectively over twelve months following the institution of CSII therapy. After four months of treatment, eight patients showed deterioration of retinopathy by at least one level of severity. Serum IGF I concentration rose from a mean +/- SEM of 155 +/- 17.7 micrograms/l at entry to 199 +/- 23.1 micrograms/l at four months and by twelve months had returned to near initial values 163 +/- 17.4 micrograms/l. There was however, no significant correlation between retinopathy score and serum IGF I level by analysis of variance for the whole group, or in the group of diabetics whose retinopathy deteriorated. The rise in IGF I concentration over the first four months and subsequent decline in IGF I values over the next eight months was inversely related to HbA1 concentration (r = -0.58; P less than 0.05). One patient with early ischaemic retinopathy on entry, experienced a marked rise in serum IGF I corresponding to a rapid tightening of glycaemic control. At four months she developed florid proliferative changes requiring panretinal laser therapy.(ABSTRACT TRUNCATED AT 250 WORDS)

No abstract