ORIGINAL ARTICLE

Safety and specificity of the growth hormone suppression testin patients with diabetes

Pedro Weslley Rosario • Maria Regina Calsolari

Received: 12 February 2014 / Accepted: 26 April 2014

� Springer Science+Business Media New York 2014

Abstract The purpose of this study was to evaluate the

safety of the oral glucose tolerance test (OGTT) and its

capacity to suppress growth hormone (GH) in diabetic

patients without acromegaly. A total of 135 diabetic patients

submitted to the OGTT for GH suppression were studied.

The following selection criteria were applied: age between

20 and 70 years; body mass index C18.5 and B27 kg/m2;

absence of kidney, liver, or thyroid disease; no use of

estrogens, androgens, corticosteroids, or levothyroxine.

Adequate suppression of GH was defined as a nadir below the

cut-off established for a sample of 200 normoglycemic

subjects (\0.25 lg/L for men, \0.74 lg/L for premeno-

pausal women, and\0.5 lg/L for postmenopausal women).

Acromegaly was diagnosed in five patients. Among the 130

diabetic patients without known pituitary disease or a clinical

suspicion of acromegaly, 95.5 % of men, 94 % of pre-

menopausal women, and 96.6 % of postmenopausal women

presented adequate GH suppression (vs 97.5 % of normo-

glycemic controls). In all patients without acromegaly, the

lowest GH levels (nadir) were achieved after the adminis-

tration of glucose and not during baseline measurement.

None of the patients had acute complications [ketoacidosis,

hyperosmolar state, and symptomatic marked hyperglyce-

mia ([300 mg/dL)] on the day of the test and up to 3 days

thereafter. We demonstrated the safety of the OGTT and its

capacity to suppress GH in diabetic patients without

acromegaly. In addition, we suggest the adoption of a pro-

tocol to prevent possible risks of the OGTT in patients with

diabetes.

Keywords Growth hormone � Suppression test �Diabetes � Acromegaly

Introduction

The oral glucose tolerance test (OGTT) for growth hormone

(GH) suppression continues to be recommended for the

diagnosis of acromegaly, at least in subjects without an

exuberant phenotype and/or only slightly elevated IGF-1,

and for the evaluation of disease activity in patients treated

only by surgery [1]. Many authors do not perform the OGTT

in patients with diabetes because of safety reasons and low

specificity (i.e., the absence of GH suppression should be

common in diabetic patients without acromegaly). This

restriction is even found in some consensuses: (i) ‘‘extreme

caution should be exercised when employing this test in

patients with glucose intolerance’’ [2]; (ii) ‘‘diabetic patients

should not be submitted to the OGTT’’ [3]; (iii) ‘‘OGTT must

not be performed in patients with overt diabetes’’ [4]; (iv)

‘‘apart from an obvious lack of control, in a diabetic who is

only weakly hyperglycemic, OGTT is best avoided’’ [5];

(v) ‘‘glucose loading should be avoided if glucose is higher

than 140 mg/dL since complete GH suppression would be

expected in the presence of hyperglycemia of this magni-

tude’’ [6]; and (vi) ‘‘OGTT is unreliable in diabetes where

loss of normal GH suppression may be seen’’ [7].

Little is known about the safety of the OGTT and its

capacity to suppress GH, measured by sensitive assays, in

diabetic patients without acromegaly [8, 9]. This infor-

mation is important since diabetes is common among

P. W. Rosario (&)

Postgraduation Program, Instituto de Ensino e Pesquisa da Santa

Casa de Belo Horizonte, Rua Domingos Vieira, 590, Santa

Efigenia, Belo Horizonte, MG CEP 30150-240, Brazil

e-mail: pedrowsrosario@gmail.com

P. W. Rosario � M. R. Calsolari

Endocrinology Service, Santa Casa de Belo Horizonte,

Belo Horizonte, MG, Brazil

123

Endocrine

DOI 10.1007/s12020-014-0282-2

patients with excess GH production. There are studies

reporting the results of the OGTT for diabetic patients in

whom acromegaly was confirmed or for known acrome-

galic patients submitted to different treatment modalities,

as recently reported by Reimondo et al. [10], without

information of specificity (i.e., in patients without acro-

megaly). Even these studies provide no information about

the safety of the OGTT in diabetic patients.

Marked hyperglycemia can interfere not only with GH

concentrations, but also with IGF-1 levels (underestimated)

[3–6]. Therefore, laboratory assessment of acromegaly

(GH and IGF-1) should be performed after reasonable

metabolic compensation.

The objective of the present study was to evaluate the

safety of the OGTT and its capacity to suppress GH in

diabetic patients without acromegaly (specificity).

Patients and methods

A total of 135 diabetic patients (known diagnosis) sub-

mitted to the OGTT for GH suppression due to the fol-

lowing reasons were studied: (i) 41 patients without known

pituitary disease or a clinical suspicion investigated in a

study on acromegaly screening [11] and with elevated IGF-

1; (ii) two patients without a clinical suspicion of acro-

megaly investigated because of pituitary adenoma and

elevated IGF-1; (iii) two patients with a clinical suspicion

of acromegaly; and (iv) 90 patients without known pituitary

disease or a clinical suspicion of GH hypersecretion

selected for the study. The study was approved by the

Ethics Committee of our institution.

Adequate suppression of GH was defined as a nadir below

the cut-off established for a sample of 200 normoglycemic

subjects, corresponding to the 97.5th percentile of the results

[\0.25 lg/L for men (n = 80), \0.74 lg/L for premeno-

pausal women (n = 70), and\0.5 lg/L for postmenopausal

women (n = 50)] [12]. To guarantee the reliability of the

comparison, in addition to the same GH assay, the same

selection criteria were applied [12]: (i) age between 20 and

70 years (excluding pregnant women); (ii) body mass index

(BMI) C18.5 and B27 kg/m2; (iii) absence of kidney, liver,

or thyroid disease (these conditions were ruled out based on

anamnesis and serum measurement of creatinine, alanine

transaminase, aspartate transaminase, alkaline phosphatase,

c-glutamyl-transferase, bilirubins, and TSH); and (iv) no use

of estrogens, androgens, corticosteroids, or levothyroxine.

The protocol for the OGTT shown in Table 1 was

followed in all the patients

GH was measured with a chemiluminescence assay (Im-

mulite, Diagnostic Products Corporation, Los Angeles, CA).

The standard provided by the kit was calibrated against the

World Health Organization (WHO) 2nd International

Standard (IS) 98/574. The results are expressed as lg/L.

Data were presented as median and range. The Fisher’s

exact test or X2 test was used to detect differences in the

proportions between groups. The Spearman correlation test

was used to analyze the correlation of plasma glucose and

HbA1c levels with nadir GH. A P value \ 0.05 was con-

sidered to be statistically significant.

Results

Seventy men and 65 women, ranging in age from 22 to

67 years (median 45 years), with a BMI of 20.2–27 kg/m2

(median 24.8 kg/m2), and duration of diabetes of

2–10 years (median 6 years), were studied. All of the

patients used oral hypoglycemic agents and 54 also used

insulin. HbA1c levels ranged from 6.6 to 9.2 % (median

7.5 %). Plasma glucose levels before the OGTT (fasting)

ranged from 86 to 160 mg/dL (median 120 mg/dL) and

were C126 mg/dL in 60 patients. Acromegaly was diag-

nosed in five patients (Table 2).

Among the 130 diabetic patients without known pituitary

disease or a clinical suspicion of acromegaly, 95.5 % of men

(n = 67), 94 % of premenopausal women (n = 33), and

96.6 % of postmenopausal women (n = 30) presented ade-

quate GH suppression (versus, by definition, 97.5 % of

normoglycemic controls [12]). Nadir GH \ 0.4 lg/L [1, 3,

5–7] was seen in all men (controls and diabetics), in 68.6 %

of normoglycemic women versus 63.6 % of premenopausal

diabetic women, and in 84 % of controls versus 80 % of

postmenopausal diabetic women. All normoglycemic sub-

jects [12], as well as men and postmenopausal women with

Table 1 Protocol for the oral glucose tolerance test

After the patient presented fasting capillary blood glucose

B160 mg/dL for two consecutive days, the test was scheduled

for the next day

Plasma glucose was obtained on the day of the test, at 7:30 am,

after a fast of approximately 10 h. Only if glucose B160 mg/dL

was the test initiated. In the case of patients using insulin, the

dose of slow-acting insulin was administered normally

Time 0: measurement of GH followed by the oral administration

of 75 g anhydrous glucose. Measurement of GH after 30, 60, 90,

and 120 min. The patient remained at rest during the test

After the last collection, capillary blood glucose was measured and

correction with ultra-rapid-acting insulin was performed if

[180 mg/dL. The patient was advised to take his usual diabetes

medication and to measure capillary glucose throughout the day

(before and after meals and at bedtime), correcting with ultra-

rapid-acting insulin if [180 mg/dL

On the day after the test, the patient returned to his usual

medications and orientations

Endocrine

123

diabetes, had nadir GH \ 1 lg/L. One of the 33 premeno-

pausal women with diabetes (3 %) had nadir GH [ 1 lg/L

(1.25 lg/L) and mildly elevated IGF-1 [1.3 9 ULN (upper

limit of normal)]. In this patient, acromegaly was excluded

based on the absence of any clinical suspicion and adenoma

by magnetic resonance imaging and because she developed

no symptoms and IGF-1 was 0.95 9 ULN after 1 year.

In all the patients without acromegaly (n = 130), including

patients with glucose levels[126 mg/dL immediately before

the OGTT (n = 60), the lowest GH levels (nadir) were

achieved after the administration of glucose and not during

baseline measurement.

No correlation was observed between plasma glucose or

HbA1c, obtained immediately before the test and nadir

GH.

Following the protocol shown in Table 1, none of the

patients had acute complications [ketoacidosis, hyperos-

molar state, and symptomatic marked hyperglycemia

([300 mg/dL)] on the day of the test and up to 3 days

thereafter. Glucose levels at the end of the test ranged from

122 to 272 mg/dL (median 172 mg/dL) and were

[180 mg/dL (need for correction with ultra-rapid insulin)

in 55 patients (40.7 %).

The characteristics and results of the diabetic patients

without acromegaly are shown in Table 3.

Discussion

Despite the sensitivity and specificity of serum IGF-1 for

the diagnosis of acromegaly and for the definition of dis-

ease activity after therapy, the GH suppression test con-

tinues to be useful and recommended in some situations

[1]. Since many patients with acromegaly have associated

diabetes, determination of the safety and specificity of this

test in diabetic patients, which was the objective of this

study, is important. Although there are other tests for the

diagnosis of acromegaly and/or evaluation of disease

activity, such as TRH [13] and galanin [14], these tests are

only considered when the OGTT cannot be performed,

since normative data and information about the accuracy of

these alternative tests are limited [13].

Few studies have evaluated GH suppression during an

OGTT in diabetic patients (without acromegaly). In addi-

tion, these studies involved a small number of patients,

were conducted before the current requirements for GH

Table 2 Data of the five

patients with acromegaly

ULN upper limit of normal

range, MRI magnetic resonance

imaginga Immunohistochemistry

confirmed GH-positive pituitary

adenoma

Gender Age

(years)

Reason for investigation IGF-1

(9 ULN)

Nadir

GH (lg/L)

MRI Management

Female 50 Participating in a

study on acromegaly

screening

1.8 0.9 Macroadenoma Surgerya

Male 64 Pituitary adenoma 1.6 1.1 Macroadenoma Surgerya

Female 44 Pituitary adenoma 1.5 1.4 Macroadenoma Surgerya

Male 38 Clinical suspicion

of acromegaly

2.5 10.2 Macroadenoma Octreotide

Male 52 Clinical suspicion

of acromegaly

2 8.1 Microadenoma Surgerya

Table 3 Data of the diabetic

patients without acromegaly

BMI body mass index, OGTT

oral glucose tolerance testa Adequate suppression of GH

was defined as a nadir below the

cut-off established for a sample

of normoglycemic subjects,

corresponding to the 97.5th

percentile of the results

Results Men (n = 67) Premenopausal

women (n = 33)

Postmenopausal

women (n = 30)

Age [range (median, years)] 28–67 (43) 26–50 (36) 53–67 (59)

BMI [range (median, kg/m2)] 23.5–27 (25.5) 20.2–26 (24) 22–27 (24.6)

Duration of diabetes [range (median, years)] 2–8 (5) 3–10 (6) 2–10 (6)

HbA1c [range (median, %)] 7–9 (8) 6.7–9.2 (7.6) 6.6–8.5 (7.4)

Plasma glucose before the OGTT

[range (median, mg/dL)]

92–160 (124) 88–160 (120) 86–156 (118)

Plasma glucose before the OGTT [ 126 mg/dL 30 (44.8 %) 15 (45.4 %) 13 (43.3 %)

Nadir GH [range (median, lg/L)] 0.05–0.3 (0.12) 0.06–1.25 (0.3) 0.05–0.68 (0.23)

Nadir GH \ 0.4 lg/L 67 (100 %) 21 (63.6 %) 24 (80 %)

Adequate suppression of GHa 64 (95.5 %) 31 (94 %) 29 (96.6 %)

Glucose levels at the end of the test

[range (median, mg/dL)]

122–272 (168) 125–268 (176) 124–260 (164)

Glucose levels [180 mg/dL at the

end of the test

26 (39 %) 14 (42.4 %) 12 (40 %)

Endocrine

123

assays, and did not evaluate the safety of the test [8, 9]. The

results of the present study suggest that the OGTT is able to

adequately suppress GH in diabetic patients without acro-

megaly, at least when the patient is reasonably compen-

sated, without the observation of a significant difference

compared to normoglycemic subjects (high specificity).

Therefore, the lack of GH suppression should be consid-

ered even in diabetic patients. We call attention to the fact

that only one premenopausal woman without acromegaly

had nadir GH [ 1 lg/L and none of the men had a nadir

[0.4 lg/L.

The exact mechanism whereby a glucose overload sup-

presses GH secretion is still unclear, but seems to be related

to an increase in somatostatinergic tone and does not

involve the inhibition of GHRH release [15]. In contrast to

the hypothesis that hyperglycemia itself is able to suppress

GH secretion [6], we observed that even in the 60 patients

with glucose levels [126 mg/dL before the administration

of glucose, GH suppression only occurred after its admin-

istration. In fact, spontaneous secretion of GH seems to be

increased in diabetic patients and has a direct correlation

with HbA1c [16]. In addition, it is well established that GH

suppression occurs even in subjects who do not present

hyperglycemia during the OGTT [12]. Therefore, the rapid

increase in glucose or even other mechanisms associated

with oral glucose overload (e.g., the release of intestinal

hormones) seem to be more important for GH suppression

than hyperglycemia itself. Serial measurement of GH at the

same time as the OGTT, but without administration of

glucose, may show how much of the reduction in GH

concentrations observed during the OGTT would be

attributed to spontaneous secretion of this hormone or to the

suppressive effect of oral glucose [8, 17].

The strengths of the present study are the number of

patients studied, the use of a sensitive GH assay in

accordance with current recommendations, comparison

with a normoglycemic control group, and the adoption of a

protocol designed to minimize possible risks of the OGTT.

In conclusion, we demonstrated the safety of the OGTT

and its capacity to suppress GH in diabetic patients without

acromegaly. In addition, we suggest the adoption of a

protocol (Table 1) to prevent possible risks of the OGTT in

patients with known diabetes.

Disclosure There is no potential conflict of interest relevant to this

article.

References

1. A. Giustina, P. Chanson, M.D. Bronstein, A. Klibanski, S.

Lamberts, F.F. Casanueva, P. Trainer, E. Ghigo, K. Ho, S. Mel-

med, A consensus on criteria for cure of acromegaly. J. Clin.

Endocrinol. Metab. 95, 3141–3148 (2010)

2. A. Barkan, M.D. Bronstein, O.D. Bruno, A. Cob, A.L. Espinosa-

de-los-Monteros, M.R. Gadelha, G. Garavito, M. Guitelman, R.

Mangupli, M. Mercado, L. Portocarrero, M. Sheppard, Manage-

ment of acromegaly in Latin America: expert panel recommen-

dations. Pituitary 13, 168–175 (2010)

3. L. Vieira Neto, J. Abucham, L.A. Araujo, C.L. Boguszewski,

M.D. Bronstein, M. Czepielewski, R.S. Jallad, N.R. Musolino,

L.A. Naves, A. Ribeiro-Oliveira Jr, L. Vilar, M. Faria, M.R.

Gadelha, Recommendations of Neuroendocrinology Department

from Brazilian Society of Endocrinology and Metabolism for

diagnosis and treatment of acromegaly in Brazil. Arq. Bras.

Endocrinol. Metab. 55, 91–105 (2011)

4. R. Cozzi, R. Baldelli, A. Colao, G. Lasio, M. Zini, R. Attanasio,

AME position statement on clinical management of acromegaly.

J. Endocrinol. Invest. 32, 2–25 (2009)

5. P. Chanson, J. Bertherat, A. Beckers, H. Bihan, T. Brue, P. Caron,

O. Chabre, M. Cogne, C. Cortet-Rudelli, B. Delemer, H. Dufour,

R. Gaillard, M. Gueydan, I. Morange, J.C. Souberbielle, A.

Tabarin, French consensus on the management of acromegaly.

Ann. Endocrinol. 70, 92–106 (2009)

6. S. Arellano, P. Aguilar, B. Domınguez, A.L. Espinosa-de-Los-

Monteros, B. Gonzalez Virla, E. Sosa, M. Mercado, G. Guinto, I.

Martınez, E. Hernandez, A. Reza, L. Portocarrero, A. Vergara,

F.J. Velazquez, E. Ramırez, Segundo consenso nacional de

acromegalia: recomendaciones para su diagnostico, tratamiento y

seguimiento. Rev. Endocrinol. Nutr. 15, S7–S16 (2007)

7. S. Melmed, F. Casanueva, F. Cavagnini, P. Chanson, L.A.

Frohman, R. Gaillard, E. Ghigo, K. Ho, P. Jaquet, D. Kleinberg,

S. Lamberts, E. Laws, G. Lombardi, M.C. Sheppard, M. Thorner,

M.L. Vance, J.A. Wass, A. Giustina, Consensus statement:

medical management of acromegaly. Eur. J. Endocrinol. 153,

737–740 (2005)

8. N. Hattori, A. Shimatsu, Y. Kato, H. Koshiyama, Y. Ishikawa, H.

Assadian, T. Tanoh, M. Nagao, H. Imura, Growth hormone

responses to oral glucose loading measured by highly sensitive

enzyme immunoassay in normal subjects and patients with glu-

cose intolerance and acromegaly. J. Clin. Endocrinol. Metab. 70,

771–776 (1990)

9. M.J. Kayath, E.M. Russo, S.A. Dib, J.G. Vieira, Do impaired

glucose tolerance and diabetes mellitus interfere with the inter-

pretation of the growth hormone response to the oral glucose

tolerance test? Braz. J. Med. Biol. Res. 25, 449–455 (1992)

10. G. Reimondo, M. Bondanelli, M.R. Ambrosio, F. Grimaldi, B.

Zaggia, M.C. Zatelli, B. Allasino, F. Laino, E. Aroasio, A. Ter-

mine, P. Conton, A. Paoletta, E. Demenis, E.D. Uberti, M.

Terzolo, Growth hormone values after an oral glucose load do not

add clinically useful information in patients with acromegaly on

long-term somatostatin receptor ligand treatment. Endocrine 45,

122–127 (2014)

11. P.W. Rosario, Frequency of acromegaly in adults with diabetes or

glucose intolerance and estimated prevalence in the general

population. Pituitary 14, 217–221 (2011)

12. P.W. Rosario, M. Furtado, Growth hormone after oral glucose

overload: revision of reference values in normal subjects. Arq.

Bras. Endocrinol. Metab. 52, 1139–1144 (2008)

13. L. Cazabat, J.C. Souberbielle, P. Chanson, Dynamic tests for the

diagnosis and assessment of treatment efficacy in acromegaly.

Pituitary 11, 129–139 (2008)

14. G. Mazziotti, S. Bonadonna, M. Doga, I. Patelli, C. Gazzaruso,

S.B. Solerte, E. De Menis, A. Giustina, Biochemical evaluation

of patients with active acromegaly and type 2 diabetes mellitus:

efficacy and safety of the galanin test. Neuroendocrinology 88,

299–304 (2008)

15. C. Park, I. Yang, J. Woo, S. Kim, J. Kim, Y. Kim, S. Park, Acute

hyperglycemia and activation of the beta-adrenergic system

exhibit synergistic inhibitory actions on growth hormone (GH)

Endocrine

123

releasing hormone-induced GH release. Eur. J. Endocrinol. 148,

635–640 (2003)

16. E. Franek, F. Schaefer, K. Bergis, R. Feneberg, E. Ritz, Abnormal

pulsatile secretion of growth hormone in non-insulin-dependent

diabetes mellitus. Clin. Endocrinol. 47, 471–478 (1997)

17. S. Grottoli, P. Razzore, D. Gaia, M. Gasperi, M. Giusti, A. Colao,

E. Ciccarelli, V. Gasco, E. Martino, E. Ghigo, F. Camanni,

Three-hour spontaneous GH secretion profile is as reliable as oral

glucose tolerance test for the diagnosis of acromegaly. J. Endo-

crinol. Invest. 26, 123–127 (2003)

Endocrine

123