Usefulness of different biochemical markers of the insulin-like growth factor (IGF) family in diagnosing growth hormone excess and deficiency in adults

Paolo Marzullo, Carolina Di Somma, Katherine L. Pratt, Javad Khosravi, Anastasia Diamandis, Gaetano Lombardi, Annamaria Colao, Ron G. Rosenfeld

Research output: Contribution to journalArticlepeer-review

84 Scopus citations


The diagnostic approach to acromegaly and GH deficiency frequently includes measurement of several components of the insulin-like growth factor (IGF) system. IGF-I levels are reported to be good predictors of active and cured acromegaly, but are commonly found within the normal age-adjusted range in adult GH-deficient (GHD) patients. Circulating concentrations of IGF-binding protein-3 (IGFBP-3), acid-labile subunit (ALS), and free IGF-I reflect the GH secretory status, but their diagnostic accuracy is still debated. In this study serum levels of total and free IGF-I, IGFBP-3, ALS, and IGFBP-3-IGF-I and IGFBP-3-ALS complexes were determined in patients previously diagnosed with active (n = 67) or inactive (n = 16) acromegaly and adult GHD (n = 34) and compared with results obtained in 58 healthy controls. In healthy subjects, IGF-I, IGFBP-3, ALS, and both IGFBP-3 complexes declined with age; a correlation was found between IGF-I and IGFBP-3 (r = 0.59; P < 0.001), ALS (r = 0.67; P < 0.001), and free IGF-I (r = 0.40; P < 0.05). Active acromegalic patients showed a significant increase in all parameters tested. IGF-I concentrations were above +2 SD in 100% of patients, whereas slightly lower sensitivities were shown for IGFBP-3 (85%), ALS (88%), and free IGF-I (94%). In this group, I GF-I exhibited a slightly higher correlation with IGFBP-3 (r = 0.83; P <0.001) than with ALS levels (r = 0.78; P < 0.001). In cured acromegalic patients, we observed the normalization of all parameters but free IGF-I levels. Adult GHD patients showed a significant reduction of all hormones. Unlike active acromegalic patients, all parameters had only a modest sensitivity in GHD; suppression below -2 SD was observed in 41% of GHD patients for IGF-I, 47% for IGFBP-3, 32 % for ALS, and 35% for free IGF-I measurements. Previous radiotherapy and GH peak response below 3 μg/L were associated with significantly lower IGF-I, IGFBP-3, and ALS levels. IGF-I levels were significantly correlated to ALS (r = 0.68; P < 0.001) and IGFBP-3 (r = 0.64; P < 0.001) as well as with free IGF-I (r = 0.67; P < 0.001) levels. By multiple regression analysis, the number of anterior pituitary hormones impaired was the most predictive indicator of IGF-I, IGFBP-3, and free IGF-I levels in GHD patients; conversely, the GH peak response better anticipated ALS concentrations. The pattern of IGFBP-3 complexes paralleled previous hormonal findings. In active acromegalic patients, IGFBP-3-IGF-I levels were 5.4-fold higher than in controls and were above +2 SD in 95% of patients, whereas IGFBP-3-ALS levels were elevated in 15% of cases: On the other hand, both IGFBP-3 complexes were able to predict GHD in only a minority of cases. Taken together, these data support the diagnostic role of IGF-I in acromegaly and suggest that free IGF-I and the IGFBP-3-IGF-I complex can assist diagnostic strategies in this condition. All markers are of limited predictive value in adult GHD, as hormonal values are commonly found within the normal limits. In these patients, low IGFBP-3 and IGF-I concentrations can add further clinical information on the residual GH activity.

Original languageEnglish (US)
Pages (from-to)3001-3008
Number of pages8
JournalJournal of Clinical Endocrinology and Metabolism
Issue number7
StatePublished - 2001
Externally publishedYes

ASJC Scopus subject areas

  • Endocrinology, Diabetes and Metabolism
  • Biochemistry
  • Endocrinology
  • Clinical Biochemistry
  • Biochemistry, medical


Dive into the research topics of 'Usefulness of different biochemical markers of the insulin-like growth factor (IGF) family in diagnosing growth hormone excess and deficiency in adults'. Together they form a unique fingerprint.

Cite this