The normal function of the growth hormone (GH)/insulin-like growth factor system is essential for growth and development . Puberty is a period that is characterized by complex hormonal and metabolic changes. Growth hormone and insulin growth factors levels reach the highest values in life in order to allow growth spurt and anabolism of muscle mass; insulin levels increase for both anabolic purposes and also to achieve reduced insulin sensitivity; sex hormones – oestrogens in females and testosterone in males – affect lipids levels so that they resemble the adult pattern [2,3].
Prior to the introduction of highly active antiretroviral therapy (HAART), the GH/insulin-like growth factor (IGF)-1 axis was examined in HIV-infected children who failed to thrive. Decreased GH secretion was reported, but primary GH deficiency was identified only occasionally . Moreover, the circadian secretory pattern of GH and stimulated GH levels were normal in the great majority of children with HIV infection [5–7]. IGF-1 has variably been reported as normal  or reduced ; however in these studies the influence of nutritional status on IGF-1 level was not clearly evaluated. Lastly, Frost et al. suggested that enhanced proteolysis of IGF-1 binding protein 3 (IGFBP-3) could be associated with failure to thrive. Thus, HIV-infected children with growth failure showed reduced levels of IGF-1 and IGFBP-3/ternary complex as compared with healthy control children .
Recently, a complex of signs and symptoms (commonly referred to as lipodystrophy), which include alterations in fat distribution and metabolism, has been described in HAART-treated HIV-infected adults. Changes in fat distribution include lipoatrophy of the subcutaneous fat compartment and/or excess visceral fat accumulation. Metabolic abnormalities include insulin resistance, hypercholesterolaemia and hypertriglyceridaemia and other alterations of lipoprotein plasma levels [9–11]. The development of this syndrome during childhood has not been widely investigated; however, the available data suggest that it may occur with features overlapping those reported in adults. Peripheral lipoatrophy and truncal lipohypertrophy have been described in antiretroviral drug-treated children in two studies conducted through objective anthropometric measurements (skin-fold thickness and dual-energy X-ray-absorptiometry) [12,13]. In a case–control study, using dual X-ray-photon absorptiometry (DXA), we demonstrated that peripheral lipoatrophy associated with an increased trunk fat accumulation is a common feature of HAART-treated children. Furthermore, an excess of visceral fat accumulation, assessed by magnetic resonance imaging (MRI), is detectable in children with clinical signs of lipodystrophy . Abnormalities of glucose homeostasis and serum lipid levels were also described in children. Elevated levels of cholesterol, low-density lipoprotein, triglycerides and a decline in insulin sensitivity have been observed in children receiving antiretroviral therapy including a protease inhibitor [15,16].
A reduced GH response to a variety of stimuli is a characteristic feature of obesity in non HIV-infected adults and children . GH dynamics have been assessed recently in antiretroviral-treated HIV-infected adults with and without fat redistribution. That report showed reduced mean GH concentration, basal GH concentration and GH pulse amplitude in subjects with fat redistribution; in addition increased visceral adiposity was the best predictor of reduced GH concentration in this population .
The GH/IGF-1 system has not been yet assessed in HIV-infected children with lipodystrophy. The aim of our study was to characterize the GH profile and the IGF system in HAART-treated adolescents with and without excess visceral fat accumulation and to assess correlation between hormonal and metabolic parameters.
Materials and methods
Twenty-five vertically HIV-infected adolescent, 10 with excess accumulation of visceral fat (V+) and 15 without excess accumulation of visceral fat (V−) followed in the Paediatric Clinic at L. Sacco Hospital, University of Milan, Italy, were included in the study. All subjects were Caucasian and had completed pubertal development according to Tanner's criteria .
None of the study subjects had clinically apparent renal, cardiac or other intercurrent diseases at the time of evaluation. None of the subjects had previously received or were currently treated with testosterone, GH, anabolic hormones, anti-diabetic agents, megestrol acetate, corticosteroids or any other hormone or drug known to affect GH.
Informed consent was obtained from the parents or legal guardians of each patient before participation. The child's assent was also obtained when appropriate. The study was approved by the ethical committee of the L. Sacco Hospital.
Hormonal and metabolic parameters assessment
Blood was drawn after an overnight fast, 12 h after the evening dose and before the morning dose of antiretroviral treatment. Blood samples for serum measurements of IGF-1, IGFBP-3, insulin and GH levels were taken in all children at 08.00 h, immediately before a combined stimulation testing with semorelin acetate [intravenous administration of recombinant human (rh) GH 1–29, 1 μg/kg] (GEREF, Serono Pharma, Unterschleißheim, Germany) and arginine [intravenous administration of arginine hydrocloride, 0.5 g/kg (maximum dose 15 g)] with GH sampling at 30, 45, 60, 90 and 120 min after stimuli infusion.
Samples for glucose, cholesterol, low-density lipoprotein (LDL), high-density lipoprotein (HDL), triglycerides, free fatty acids and nitric oxide determinations were also collected at 08.00 h.
Anthropometric measurements and body composition analysis
All subjects enrolled in the study underwent physical examinations to obtain anthropometric measures, according to standard reference manual . Body weight was measured to the nearest 0.1 kg on a balance beam scale (Seca, Hamburg, Germany) and height was measured to the nearest millimeter using a wall-mounted stadiometer (Holtain Ltd, Crosswell, UK). Body mass index (BMI) was then calculated as weight/height (in kg/m2).
Whole body composition was assessed with a DXA scanner (Lunar DPX-L, Lunar Radiation Corp., Madison, Wisconsin, USA), equipped with specific paediatric software (version 1.5e). Scans were performed as described previously  with subjects in the supine position, the children not requiring sedation. The entire body was scanned, beginning at the top of the head. Mean measurement time was 20 min; radiation exposure was < 8 μSv. Body fat was expressed in kg and as a percentage of body weight; lean mass was expressed in kg; the ratio between fat and lean mass was then calculated. Three-compartment analyses were performed in the arms, trunk and legs. Daily quality-assurance tests were performed according to the manufacturer's directions. All scans were performed and analysed by the same operator (S.M.). The precision of the instrument was 0.7% for fat mass and 0.9% for lean mass in normal-weight paediatric subjects. Anthropometry and DXA were performed on the same day.
Evaluation of visceral fat content by MRI
Intra-abdominal fat content was measured by MRI. Patients were imaged on a Philips Gyroscan ACS-NT 1.5 T (Philips Medical Systems, Best, the Netherlands). The slice passing through the umbilicus (at the fourth lumbar vertebra) was used to estimate intra-abdominal fat. A manual trackball with visual control was used to limit adipose tissue areas, and then areas were automatically computed as described previously . Intra-abdominal adipose tissue (IAT) area was expressed in cm2. The same trained operator performed the MRI analysis (P.B.) and he was kept blinded to the patients’ details.
GH concentrations were measured by a two-site radioimmunometric assay (Corning Inc., Nichols Institute Diagnostics, San Juan Capistrano, California, USA). Serum IGF-1 was measured, after an acid–alcohol extraction, using a RIA kit (Corning, Inc., Nichols Institute Diagnostics). Serum insulin levels were measured using a Microparticle Enzyme Immunoassay (MEIA, IMX, Abbott Laboratories, Diagnostics Division, Abbott Park, USA). Serum concentrations of IGF BP-3 have been measured using a commercially available kit (IGFBP-3 ELISA, Mediagnost, Reutlingen, Germany). Any samples that did not require extraction were diluted 1 : 200. Nitric oxide levels were evaluated through the measurement of metabolic end products, i.e., nitrite and nitrate, using enzymatic catalysis coupled with Griess reaction, as reported previously . Free fatty acids (FFA) were measured using a spectrophotometric method adapted on an automated centrifugal analyser, COBAS FARA II (Roche, Basel, Switzerland) as reported . Glucose, cholesterol (total, LDL, HDL) and triglyceride levels were determined using standard laboratory assays.
Normal distribution of the variables has been evaluated by the Shapiro normality test. Normally distributed variables are described as mean (± SD), while not normally distributed variables are shown as median (range).
All of the analyses were performed at the α = 0.05 level, and were two-tailed. Comparisons between the V+ and V− groups were performed using either the Student's t test or the Wilcoxon non-parametric test. The distribution of boys and girls among the groups has been evaluated by the χ2 test.
Multivariate analyses have been used to compare the body composition and biochemical variables between the two groups of patients, to simultaneously correct for the diverse distribution of boys and girls among the groups, and for the corresponding differences in anthropometric measurements. In all the models the independent confounding variables were sex and weight, and the two groups were expressed as a dichotomous variable.
Relationships between variables have been analysed by Spearman rank test as to a null hypothesis σ = 0.
Statistical analyses were performed using the JMP-IN statistical software (SAS Institute, Cary, North Carolina, USA) on a Macintosh computer (Apple Computer Inc., Cupertino, California, USA).
The clinical, demographics and growth hormone features of the 10 V+ and 15 V− HIV-infected adolescents included in this study are shown in Table 1.
Eight out the 10 V+ subjects showed clinical signs of fat loss (cheeks, buttocks and limbs) and fat accumulation (abdominal girth) associated with breast enlargement in four cases and the dorsocervical fat pad in one case. Among the 15 V− subjects, clinical signs of fat loss were detectable in five cases and signs of fat loss combined with fat accumulation were detectable in one case.
All patients were on a long-lasting HAART regimen with two nucleoside analogue inhibitors (stavudine and lamivudine) and one protease inhibitor (nelfinavir in one, ritonavir in three and indinavir in 22 cases, respectively).
The 10 V+ and 15 V− subjects did not differ in age (13.8 versus 14.8 years), weight (49.4 versus 51.9 kg), height (155.1 versus 162.3 cm), or BMI (20.2 versus 19.5 kg/m2). The height and growth velocity of all patients was within the 10th and 90th centiles on the Tanner growth chart. Despite a higher proportion of boys in the V− as compared with the V+ group (8/7 versus 3/7), sex distribution was not statistically different (χ2 = 1.35; P = 0.24).
Previous exposure to nucleoside reverse trascriptase inhibitors (33 versus 48 months) and current exposure to HAART (54.5 versus 55 months) were similar between the two groups.
Impaired GH secretion and IGF-1 levels occur in HAART-treated adolescents with excess visceral fat accumulation
The area under the curve of the GH response (GH-AUC0–120 min) to rhGH and arginine stimulation was significantly reduced in V+ (16.39 μg⋅1h/L) as compared with V− subjects (31.58 μg⋅h/L) (P = 0.002). Similarly, significantly lower IGF-1 levels (P = 0.030) were detected in V+ (384.2 ng/ml) than in V− subjects (515.0 ng/ml), even after adjusting for confounding variables (sex, weight). On the contrary, serum concentrations of IGFBP-3 were similar between V+ (4.3 μg/ml) and V− (4.7 μg/ml) subjects.
Fat mass and insulin levels are increased in HAART-treated adolescents with excess visceral fat accumulation
By design, the IAT content, assessed by MRI, was higher in V+ (median, 75; range, 50–135 cm2) than in V− (median, 28; range, 10–40 cm2) subjects (P < 0.0001).
As shown in Table 2, whole and regional body lean and fat mass, determined by DXA were significantly different between the two groups, except for fat mass of the legs. These differences were also found after correcting for confounding variables (weight and sex). In particular, V+ subjects showed lower lean mass as a total (P = 0.025) and at the arms (P = 0.024) and legs (P = 0.008) and higher fat mass as a total (P = 0.0038), as a percentage (P = 0.0014) and at the arms (P = 0.028) and trunk (P < 0.0001). Moreover, V+ subjects showed a lower – but not significantly so – limbs : trunk fat ratio and a significantly higher fat mass : lean mass percentage (P = 0.0021).
As shown in Table 3, fasting lipid profile (cholesterol total, LDL, HDL and triglycerides), fasting glucose as well as fasting FFA and nitric oxide levels did not differ between the two groups of patients. On the contrary, fasting insulin level was significantly higher in V+ (17.8 μIU/ml) than in V− (10.5 μIU/ml) subjects (P = 0.01).
Finally, we found that GH-AUC0–120 negatively correlated with IAT content (σ = –0.36; P = 0,03) and insulin levels (σ = –0.54; P = 0,001).
Our data demonstrate a reduced GH response to rhGH and arginine stimulation testing, associated with a reduced IGF-1 concentration in adolescents with an abnormally high amount of visceral fat. A pattern of impaired synchronous release of GH and IGF-1 is novel and seems to indicate true impaired GH dynamics in HAART-treated adolescents with central adiposity.
The extent of visceral fat content was measured by MRI. In adults [25,26], and recently also in studies on children [27,28], MRI showed great accuracy in quantifying IAT in patients affected by body composition abnormalities. Subjects with central obesity were defined on the basis of an IAT content > 41 cm2. This cut-off value was generated from the mean (21 cm2) and SD (10 cm2) values observed in healthy non-obese children and adolescents previously described [14,27]. To increase the sensitivity to detect visceral adiposity, the cut-off value was set to the mean + 2 SD.
Age, sex and pubertal status are confounding factors that must be considered in the interpretation of GH dynamics . Age is a major negative statistical determinant of GH production rate and half-life; however, the comparable age of subjects with and without excess of IAT ruled out an age-effect on the GH response observed in our study. Sex differences in the pattern of GH secretion are not clear-cut in humans, but there are noticeable quantitative differences with serum GH concentration, being higher in women than in men. The female : male ratio was not significantly different in subjects with and without excess IAT but the female gender was relatively higher (60% versus 47%) in the group with excess IAT. Puberty strongly affects GH secretion during childhood; to avoid a possible confounding role of this factor, only subjects showing complete pubertal development were included in our study.
Impaired GH dynamics (reduced mean overnight GH concentration, reduced basal GH concentration and GH pulse amplitude) associated with normal IGF-1 levels was recently demonstrated in lipodystrophic HAART-treated adults with increased visceral adiposity. In addition, visceral fat was the most significant predictor of GH secretion in these subjects . Using specific peak GH stimulatory cut-off values of 3.0 μg/l to rhGH and 5.0 μg/l to arginine testing, this study detected a 33–48% prevalence of deficient GH response among lipodystrophic subjects.
GH status is conventionally assessed by provocative stimuli including arginine, clonidine, insulin and rhGH . The threshold level used to define a normal GH response to a single provocative test is set at 10 ng/ml in children and varies from 3 to 5 ng/ml in adults . The use of rhGH in combination with substances that act via inhibition of endogenous somatostatin, such as pyridostigmine and arginine induces a much greater GH response and the combined test (rhGH and arginine) shows much reduced inter- and intra-individual variability .
In a recent comparative study of the reliability of various GH provocative tests, Ghigo et al. found that only rhGH combined with arginine or pyridostigmine may be able to clearly differentiate normal children from patients with GH deficiency and the minimal cut-off for a normal GH response is higher than 19 ng/ml .
In our study, a peak GH response of < 20 ng/ml on arginine and rhGH testing was found in 60% of adolescents with increased visceral adiposity, suggesting a higher prevalence of GH abnormality than in adults. Furthermore IGF-1 levels were significantly lower in these subjects; this datum suggests a more severe impairment of the GH axis in lipodystrophic adolescent subjects as compared with lipodystrophic adults.
The pattern of reduced GH secretion and IGF-1 levels is a new finding in the HIV-infected population. Previous studies in HIV-infected adults and children with wasting syndrome and/or active disease showed a classic picture of GH resistance, characterized by reduced IGF-1 and normal or increased GH levels [6,8,31,32]. On the contrary, a true GH deficiency was described only occasionally in a few children with severe growth retardation [4,5]. Our data are completely different from those described previously in wasting subjects, as we found GH impairment (and not resistance) in well-nourished and adolescents with normal grown.
The subjects included in our study showed height-for-age, height velocity and height-for-target-height all within the normal range, and all had completed puberty. Nevertheless we do not know how long the GH deficiency lasted; the GH abnormalities observed did not appear to have impaired linear growth, but they were associated – as they are in adults – with changes in body composition and metabolism. Subjects with excess IAT showed significantly higher trunk fat and significantly lower lean mass (total and regional) than those without excess IAT. This body composition pattern is a characteristic feature of adults as well as of children with GH deficiency [33,34]. After reaching final height, GH acts physiologically as an anabolic agent for lean mass and as a lipolitic factor for visceral fat — its deficiency leads to visceral fat accumulation and progressive impairment of muscle mass. Metabolic consequences may involve lipids and insulin metabolism . We found a significant increase in insulin levels in adolescents with an excess of IAT, suggesting a marked degree of insulin resistance, and a significant negative correlation between GH-AUC0–120 min and insulin level. It is known that protease inhibitors may interfere with glucose receptors directly, leading to glucose metabolism abnormalities; however our data support also a role of GH deficiency on worsening insulin resistance .
Our data do not suggest which variable is the first to be affected — GH or visceral adiposity: the GH/fat cycle is well known to have a reciprocal regulatory effects .
The combination of excess visceral fat, insulin resistance and dyslipidaemia observed in HIV-related lipodystrophy is similar to the metabolic syndrome, or ‘syndrome X', and suggests potentially increased risks of type 2 diabetes mellitus, cardiovascular disease and stroke.
Recent studies have shown the effectiveness of rhGH therapy on reducing visceral adiposity in obese adults  and in lipodystrophic HAART-treated HIV- infected adults [20,21]. These data suggest a new therapeutic approach in order to improve quality of life and to reduce cardiovascular risk. This is the first reported evidence of abnormal GH physiology in HAART-treated children with visceral fat accumulation and outlines a possible role of rhGH therapy to improve body composition and metabolic parameters also in childhood.
We thank E. Paola Sandoli for excellent technical assistance.
Sponsorship: Supported by Grant No. 30D.84 from Istituto Superiore di Sanità - Programma nazionale di ricerca sull'AIDS, 2000.
1.Rosenfeld RG, Wilson DM, Lee PDK, Hintz RL. Insulin-like growth factor-I and II in evaluation of growth retardation.J Pediatr
2.Roemmich JN, Clark PA, Lusk M, Friel A, Weltman A, Epstein LH, et al
. Pubertal alterations in growth and body composition. VI. Pubertal insulin resistance: Relation to adiposity, body fat distribution and hormone release.Int J Obesity
3.Mauras N, Rogol AD, Haymond MW, Veldhuis JD. Sex Steroids, growth hormone, insulin-like growth factor-1: neuroendocrine and metabolic regulation in puberty.Horm Res
4.Jospe N, Powell KR. Growth hormone deficiency in a 8 year-old girl with Human Immunodeficiency Virus Infection.Pediatrics
5.Laue L, Pizzo PA, Butler K, Cutler GBJ. Growth and neuroendocrine dysfunction in children with acquired immunodeficiency syndrome.J Pediatr
6.Rondanelli, MG, Caselli D, Maccabruni A, Maghnie M, Bacchella L, DeStefano A, et al
. Involvement of hormonal circadian secretion in the growth of HIV-infected children.AIDS
7.Lepage P, Van de Perre P, Van Vliet G, Nsenugumuremyi F, Van Goethem C, Kestelyn P, et al
. Clinical and endocrinologic manifestations in perinatally human immunodeficiency virus type. 1-infected children aged 5 years or older.Am J Dis Child
8.Frost RA, Nachman SA, Lang CH, Gelato MC. Proteolysis of insulin-like growth factor-binding protein-3 in human immunodeficiency virus-positive children who fail to thrive.J Clin Endocrinol Metab
9.Safrin S, Grunfeld C. Fat distribution and metabolic changes in patients with HIV infection.AIDS
10.John M, Nolan D, Mallal S. Antiretroviral therapy and the lipodystrophy syndrome.Antivir Ther
11.Chen D, Misra A, Garg A. Lipodystrophy in human immunodeficiency virus-infected patients.J Clin Endocrinol Metab
12.Jaquet D, Levine M, Ortega-Rodriguez E, Faye A, Polak M, Vilmer E, et al
. Clinical and metabolic presentation of the lipodystrophic syndrome in HIV-infected children.AIDS
13.Arpadi SM, Cuff PA, Horlick M, Wang J, Kotler DP. Lipodystrophy in HIV-infected children is associated with high viral load and low CD4+-lymphocyte count and CD4+-lymphocyte percentage at baseline and use of protease inhibitors and stavudine.J Acquir Immune Def Syndr
14.Brambilla P, Bricalli D, Sala N, Renzetti F, Manzoni P, Vanzulli A, et al
. Highly active antiretroviral-treated HIV-infected children show fat distribution changes even in absence of lipodystrophy.AIDS
15.Melvin AJ, Lennon S, Mohan KM, Purnell JQ. Metabolic abnormalities in HIV type 1-infected children treated and not treated with protease inhibitors.AIDS Res Hum Retroviruses
16.Bitnun A, Sochett E, Babyn P, Arneson C, Holowka S, Forbes J et al
. Insulin sensitivity, serum lipids and abdominal adipose tissue distribution in protease inhibitor-treated and protease inhibitor-naive HIV-infected children.Antivir Ther
17.Müller EE, Locatelli V, Cocchi D. Neuroendocrine control of growth hormone secretion.Physiol Rev
18.Rietschel P, Hadigan C, Corcoran C, Stanley T, Neubauer G, Gertner J, et al
. Assessment of growth hormone dynamics in human immunodeficiency virus-related lipodystrophy.J Clin Endocrinol Metab
19.Tanner JM, Whitehouse RH. Clinical longitudinal standards for height, weight, height velocity and weight velocity and stages of puberty.Arch Dis Child
20.Lohman TG, Roche AF, Martorell R. Anthropometrical Standardization Reference Manual.
Edited by Lohman TG, Roche AF and Martorell R. Champaign IL: Human Kinetics Books’ 1988.
21.Brambilla P, Bosio L, Manzoni P, Pietrobelli A, Beccaria L, Chiumello G. Peculiar body composition in patients with Prader-Labhart-Willi syndrome
. Am J Clin Nutr
22.Van der Kooy K, Seidell JC. Techniques for the measurement of visceral fat: a practical guide.Int J Obesity
23.Verdon CP, Burto BA, Prior RL. Sample pretreatment with nitrate reductase and glucose-6-phosphate dehydrogenase quantitatively reduces nitrate while avoiding interference by NADP+; when the Griess reaction is used to assay for nitrite.Anal Biochem
24.Monti LD, Brambilla P, Stefani I, Caumo A, Magni F, Poma R, et al
. Insulin regulation of glucose turnover and lipid levels in obese children with fasting normoinsulinaemia.Diabetologia
25.Seidel JC, Bakker CJG, Van der Kooy K. Imaging techniques for measuring adipose tissue distribution. A comparison between computed tomography and 1,5 T magnetic resonance.Am J Clin Nutr
26.Fowler PA, Fuller MF, Glasbey CA, Foster MA, Cameron GG, Naughan RJ, et al
. Total and subcutaneous adipose tissue in women: the measurement of distribution and accurate prediction of quantity by using magnetic resonance imaging.Am J Clin Nutr
27.Brambilla P, Manzoni P, Sironi S, Simone P, Del Maschio A, di Natale B, et al
. Peripheral and abdominal adiposity in childhood obesity.Int J Obes
28.Fox K, Peters D, Armstrong N, Scarpe P, Bell M. Abdominal fat deposition in 11 years old children.Int J Obes
29.Shalet SM, Toogood A, Rahim A, Brennan BMD. The diagnosis of growth hormone deficiency in children and adults.Endocrine Rev
30.Ghigo E, Bellone J, Aimaretti G. Reliability of provocative tests to assess growth hormone secretory status. Study in 472 normally growing children.J Clin Endocrinol Metab
31.Geffner M, Yed DY, Landaw EM, Scott LM, Stiehm ER, Bryson YJ, et al
. In vitro IGF1, GH, insulin resistance occurs in symptomatic HIV infected children.Pediatr Res
32.Greenspoon S, Corcoran C, Lee K, Burrows B, Hubbard J, Katznelslon L, et al
. Loss of lean body and muscle mass correlates with androgen levels in hypogonadal men with acquired immunodeficiency syndrome and wasting.J Clin Endocrinol Metab
33.Binnerts A, Deuremberg P, Swart GR, Wilson JHP, Lamberts SWJ. Body composition in growth hormone deficient adults.Am J Clin Nutr
34.Leger J, Carel C, Legrand I, Paulsen A, Hassen M, Czernichow P. Magnetic resonance imagimg evaluation of adipose tissue and muscle tissue mass in children with growth hormone deficiency, Turner's syndrome and intrauterine growth retardation during the first year of treatment with GH.J Clin Endocrinol Metab
35.Murata H, Hruz PW, Mueckler M. Indinavir inhibits the glucose transporter isoform Glut4 at physiologic concentrations.AIDS
36.Gertner JM. Growth hormone actions on fat distribution and metabolism.Horm Res