Growth Hormone after Oral Glucose Overload: Revision of Reference Values in Normal Subjects

The evaluation of growth hormone (GH) secretion continues to be important in acromegaly and the nadir GH (n-GH) level in the oral glucose tolerance test (OGTT) is the gold standard for the demonstration of secretory autonomy of this hormone. n-GH levels < 1 mu g/L are defined as normal suppression but, using current assays, n-GH < 1 mu g/L is detected in patients with untreated acromegaly and this value seems to be much lower in normal subjects. The objective of the present study was to evaluate n-GH levels in the OGTT in normal subjects using three different assays (GH ICMA Immulite; GH IRMA DSL and GH IFMA AutoDelfia). Two-hundred apparently healthy subjects (120 women) ranging in age from 18 to 70 years and with a BMI >= 18.5 and <= 27 kg/m(2), who used no medications and presented normal glycemia, blood count, albumin, creatinine, TSH, SGOT, SGPT and bilirubin were studied. Serum samples were obtained before and 30,60,90 and 120 min after oral administration of 75 g glucose. The test was repeated after 4 weeks in 157 participants, with the same protocol being used in 79 and 78 receiving an overload of 100 g glucose. n-GH cut-off values (97.5th percentile) were higher in women than in men (GH-IFMA: 0.30 versus 0.11 mu g/L; GH-ICMA: 0.60 versus 0.25 mu g/L; GH-IRMA: 0.20 versus 0.10 mu g/L, respectively). No correlation was observed between n-GH and age or BMI. A difference was only observed when comparing women 35 years (n = 40) versus 35 years (n = 80), with higher values in the former (n-GH cut-off in this subgroup: GH-IFMA 0.40 versus 0.26 mu g/L, GH-ICMA 0.74 versus 0.50 mu g/L, GH-IRMA 0.25 versus 0.15 mu g/L). A good correlation was observed between the assays (r = = 0.9-0.96), however, the highest values were always obtained with the Immulite assay. Test repetition with 75 g oral glucose showed a variation in n-GH < 10.2% (GH-IFMA), < 13.4% (GH-ICMA) and < 11% (GH-IRMA) in 95% of the subjects. This variation was similar when the test was repeated with 100 g glucose. A good correlation was observed between n-GH in the first and second test (r = 0.83-0.92). We suggest the following n-GH reference values: for men, 0.14 mu g/L for the GH IRMA DSL and GH IFMA AutolDelfia kits and 0.25 mu g/L for the GH ICMA Immulite kit; for women, 0.25 mu g/L, 0.40 mu g/L and 0.70 mu g/L, respectively. (Arq Bras Endocrinol Metab 2008; 52/7:1139-1144)