]oirrritrl oj' lriterritrl Metlicitie 1992 : 232 : 447-452
The bone mineral density in acquired growth hormone deficiency correlates with circulating levels of insulin-like growth factor I A. G. JOHANSSON, P. BURMAN, K. WESTERMARK & S. I,JUNGHAI,I, Frorti tlie 1leptrrtrrierit 01Iriterrial Medicine. University Hospital. Uppsola. Sumfen
Abstract. Johansson AG. Burman P. Westermark K. Ljunghall S (Department of Internal Medicine, University Hospital. Uppsala, Sweden). The bone mineral density in acquired growth hormone deficiency correlates with circulating levels of insulin-like growth factor I. lorrrtial of Internal Medicine 1992 : 232 : 447-452.
Insulin-like growth factor I (IGF-I) is an important anabolic factor for osteoblasts in vitro. Low plasma levels of IGF-I have been observed in young men with osteoporosis. In the present study, we have studied bone mineral density (BMD) and the circulating levels of IGF-I and growth hormone (GH) in adults with acquired CH deficiency. f3MI) was determined by dual-energy x-ray absorptiometry in 1 7 men and 12 women (age 27-54 years). Spinal BMD was positively correlated with the plasma levels of IGF-I ( r = 0.43, P = 0.01 9), with the median of GH values obtained by repeated sampling a t night ( r = 0.43. I-' = 0.001 9), and with the peak of GH values during GHKH provocation test (r = 0.49. P = 0.039). The total BMD was positively related to plasma IGF-I and median of GH values. but not to peak GH by GHKH provocation. In a multiplc regression analysis model, IGF-I. peak GH by GHKH provocation test and duration of GH deficiency explained 49%)of the variation in spinal BMD. As compared to healthy controls, total, but not spinal. bone mass was lower in men with GH deficiency, but no clinical symptoms of osteoporisis were observed. The positive relationships between BMD and circulating IGF-I and other indices of GH secretion suggest that IGF-I has an endocrine effect on bone mass. Keywords: bone mineral density, growth hormone, growth hormone deficiency, insulinlike growth factor I .
lntroduction Insulin-like growth factor I (IGF-I, Somatomedin-C) is known to mediate actions of growth hormone (GH). and to have endocrine as well as paracrine and autocrine effects [ l l . The plasma levels of IGF-I are low in GH deliciency 121. IGF-I and GH affect skeletal metabolism in vivo as well as in vitro r3-51. IGF-I stimulates collagen synthesis and osteoblast proliferation [6], while GH has been reported to act via stimulation of IGF-I synthesis in the bone tissue r71. Human osteoblastlike cells have the ability to synthesize IGF-I, a capacity that is increased by GH [S]. The IGF-I synthesis in bone it1 vitro is also influenced by other hormones important for skeletal metabolism, e.g. parathyroid hormone, calcitriol, cortisol and oestrogen [S-ll].
Osteopenia has been reported in GH-deficient children [12], and in adults with GH deficiency of childhood onset [13], but bone mass in adult-onset GH deficiency has not been thoroughly investigated. We have previously reported on low IGF-I levels in male patients with idiopathic osteoporosis [I 41. In the present study, our aim was to investigate the relationship between BMD and circulating IGF-I in adult patients with acquired CH deficiency.
Patients and methods A total of 1 7 men and 1 2 women (mean age ( k S D ) 4 4 . 9 k 7 . 4 years, range 27-57 years) with GH deficiency were included in the study ('l'able 1).All patients had a GH peak below 3 pg I-' in response to insulin-induced hypoglycaemia (blood glucose d 2.2 mmol I-'). The known duration of GH de-
44 7
448
A. G. JOHANSSON et d.
‘I’ahle 1 . Patients with acquired growth hormone tlclicicncy Patient no.
Sex (M/l:)
Weight (kg)
Length (cm)
Ilu ration (years)
h7
168
93
I73
42
98
47 33
98 79
I82 1x1
34 2h 19
50 36 38 42 42
97 85
Age (years)
Iliagnosis
Sex steroid substitution
I3MIl
(g cni-?)
~
1
M
-7
M M M
3 4 5
M
6
M
7 X
M M
9
M
10
M M
46 44
177 I76
9
C.A
9
CIJ
cont. cont. intermitt.
1 149 1 031
I 105 I 1x8 I219
cont. intermitt. cont.
-/>
178
7
93 92
1 x9
C.A C.A C.A
cont. cont.
176
6 6
78
177
5
C.A
174 18 0 190
3
C.A C.A
27
91 89 73
cont. cont. cont.
3
trau niii
cont.
41 49 3h
83 X2 76
175
3 1
C.A
cont. cont.
57 48
26 24 20
C.A C.1’ C.A
16
CU
I6
empty sella
12 8
C.A C.A C.A
M M 1\11
I5
M
1 (7
17
M M
18
I:
19 20 21
F 1;
51 44 50
I!
55
22 23 24 25
1: 1: 1: 1:
43 50 54
26 27 28 29
1: 1: i:
apoplexia CIJ C.A
10 10
1 13X 0 925 I 304 I 053 I 096
cont.
180
11 12 13 14
I:
C.1’ C.1’
51 54 40
49
57 54 36 41
I72 168
157
hI 78
166 163 I66
65 73 85 x3
158 168 165 179
64
I65 I59 1 f> 1 162
6I
60
-13
3
*
6 2 1
* *
apoplcxia idiopathic
C.A
empty sella C.1’
C.A
cont.
0 9x9 I 128 0 940 I 000
1087
cont.
0 945 I 003
intermitt. never intermitt. postnienop.
0 920 1 024 0 958 0 927
cont. never postnienop. cont. postmenop. postnienop. cont. cont.
I053 0 889
1 508 12x2 0 x9x 1 I63 1000 I 034
’I’he causes of GH deliciency were. if not stated otherwise. pituitary tumours treated with surgery and/or radiation. C.1’ =
craniopharyngioni~i,CIJ = Cushing’s disease. C . A = cromophobic adenonia. The ctiuse of hypopituitarism could not be established in patient no. 17. All patients. except patients nos 3 and 24 who were only deficient in GH and FSH/I,H. were adequately substituted with thyroxine. glucocorticoids and. if necessary. with desmopressin. Some of the patients did not receive sex steroid substitution despite lack of gonadotropins. Intermitt. = intermittently. cont. = continuously substituted with sex steroids during the known period of hypopituitarisni. Patients nos 22. 24. 26 and 27 were postmenopausal at the onset of hypopituitarism. Iluration = duration of CH deliciency. The asterisks indicate that the duration was not possible to establish. 13MIl = bone niineral density in the spine measured by dual energy x-ray absorptiometry of the total body.
liciency was 11.2 1-9.1 years (range 1-34 years). A total of 1 7 patients had been treated for pituitary tumours with radiation and/or surgery. In one patient, panhypopituitarism had developed after a skull, trauma. while the aetiology of pituitary insufficiency could not be established in one patient. All but two patients suffered from panhypopituitarism. and were treated with thyroxine and glucocorticoids, and their urinary cortisol and serum T:, and free T, levels were within the normal ranges. Seven patients also received a vasopressin-analogue, desmopressin. Patients nos 1 9 and 24 had partial hypopituitarism and were only deficient in GH and FSH/LH. All of the
men and five of the women were substituted with sex hormones at the time of the study, but not all of them had had continuous substitution despite panhypopituitarism. Six of the women had never received sex steroids, and four of them were already postmenopausal at the onset of gonadotropin deficiency (Table 1). None of the patients had any known fragility fractures. BMD of the total body was measured by dual energy X-ray abosorptiometry (equipment DPX-I, from Lunar, USA). BMD of the lumbar region (BMD spine) was obtained by regional analysis of the total body measurement. BMD of the total body was also
BONE MASS A N D IGF-I IN ACQUIRED GH DEFICIENCY
nieasured in 29 healthy men, members of the hospital staff. selected to match the male patients with GH deliciency regarding age. IGF-I concentrations in plasma were measured by direct radioimmunoassay using a commercial kit from the Nichols Institute (San Capistrano, CA. USA). The reference ranges were 0.34-1.9 U ml-' for men and 0.45-2.2 U ml-' for women. Repeated measurements of GH were made using a vacuum pump with samples drawn every 20 min between 22.00 and 04.00 hours. GH was analysed by radioimmunoassay. The median of the GH values in each patient was used in the calculations. GHRH (Groliberin, Kabi, Sweden) provocation test was done in 2 7 patients. 1 pg kg-' body weight of GHKH was injected intravenously and GH samples were drawn every 1 5 min for 9 0 min. the highest value being used in the calculations. The statistical package STATVIEW SE+ v 1.03 from Abacus Concepts, Inc. (Berkeley, CA, USA) was used for all statistical calculations, including Student's ttest, Pearson's correlation coefficients and multiple regression analysis. All tests were two-tailed and P < 0.05 was considered to be statistically significant.
449
I .6
(a) total body
-
..
E0.032
I .5 I .4
b) spine e0.093
..
N
'
Em
1.3
0
Y
0 1.2
I
m
1.1
I .o 0.9
Patients Controls Patients Controls Fig. 1 . Bone mineral density (BMI))of (a) the total body and (b) the spine in 1 7 male patients with acquired GH deticiency (0) and in 29 age-matched healthy men
(m).
I
0.7 1
1 I
0.6
I
0
0
0
Heslrl t s
individual BMD spine values are given in Table 1. Compared to the normal healthy male controls of the same ages, the male patients had significantly lower BMD of the total body ( 1 . 2 0 f 0 . 0 8 vs. 1 . 2 6 f 0 . 0 8 , P = 0.032; Fig. 1) but not in the spine (1.08+0.11 vs. 1 .1 4 + 0 .1 2 . P = 0.093; Fig. 1). The female patients were compared to age- and weight-matched reference material provided by the manufacturer, and the median Z score for BMD spine was -0.75 (95% confidence interval: - 1.11-0.34) while it was -0.12 ( 9 5 % confidence interval: -0.89-0.6) for total body BMD. The plasma levels of IGF-I were markedly lower than normal. and amounted to 0 . 2 7 k 0 . 1 U mi-' in male patients and 0.32 0.2 U ml-' in female patients. The median GH value obtained by repeated sampling at night was 0 . 5 4 k 0 . 2 1 pg I-', and the highest GH value in any patient was 1.5 pg I-'. The mean peak GH responses to GHRH provocation tests w a s 2 . 2 f 1 . 4 p g l - I f o r t h e m e n a n d 1 . 8 f 1 . 3 pg1-l for the women, clearly below the reported values in these age groups [15]. There was a direct correlation between BMD of the spine and plasma IGF-I (r = 0.43, P = 0.019; Fig.
0.8 0.9
1.0
1.1
1.2
1.3
1.4
1.5
1.6
BMD of spine (g cm-')
Fig. 2. Relationship between bone mineral density (BMD) of the spine and plasma levels of insulin-like growth factor I (IGF-I) in patients with acquired GH deficiency. ( 0 )= men. ( 0 )= women.
2). BMD was also positively related to the median of GH values obtained by repeated sampling at night ( r = 0.43, P = 0.019; Fig. 3) and to the peak of GH during GHRH provocation (r = 0.40, P = 0.039 : Fig. 4). For total body BMD, positive correlations were found with IGF-I (r = 0.44, P = 0.018) and with the median of GH values at night (r = 0.40, P = 0.032) but not with the peak of GH by GHRH (r = 0.23, P = 0.25) (individual data not shown). A multiple regression analysis was performed including age, sex, weight, duration of GH deficiency, as well as median and stimulated GH levels. Circulating IGF-I appeared to be the major factor that explained variations in spinal bone mass (Table 2). Forty-nine per cent of the variation in spinal BMD was explained by IGF-I, peak, GH value and duration of GH deficiency.
Age Median GH n intercept r2 P-value
0.50
0.09
24 0.56 0.58 0.029
1.42
0.93
0.12
Sex
1.62
1.31
2.78
0.82
t-value
-0.004
0.004
Weight
-0.004
0.07
Peak GH by GHRH
Duration
0.19
IGF-I
B
-0.005
0.12
0.004
-0.004
0.08
0.26
B
24 0.58 0.57 0.01 5
0.99
1.44
1.68
1.29
3.31
1.34
t-value
0.07
0.003
-0.004
0.07
0.34
B 0.07
0.36
B
24 0.73 0.52 0.0059
0.97
1.35
3.23
2.05
t-value
-0.005
0.08
0.38
/3
24 0.87 0.49 0.003 1
1.76
3.49
2.20
t-value
3
1
O O
0.002
-0.004
+
@ O
24 0.52 0.54 0.0095
1.04
1.41
1.23
3.16
1.90
t-value
0.04
0.44
B
27 0.86 0.34 0.0075
2.32
2.52
t-value 0.46
B
29 0.94 0.19 0.019
2.49
t-value
Table 2. Partial correlation coefficients (8) for determinants of bone mineral density of the spine (BMD). as determined by multiple linear regression analysis. incorporating age. weight, sex, median of GH values obtained by repeated sampling during 6 h at night. duration of growth hormone deficiency, GH peak in response to GHRH provocation test. and plasma levels of IGF-I as the independent variables
BONE MASS AND IGF-I I N ACQUIKED GH DEFICIENCY
well as seruni levels of IGF-I decline with age and lowered physical fitness [23, 241, and previous studies have shown a positive relationship between BMD and circulating IGF-I in healthy subjects, although dependent of age [25] and physical fitness 1261. The liver has been reported to be the main contributor to plasma IGF-I [2 71. However, Bautista et ol. [28] suggested, on the basis of their results from a comparison of skeletal and serum concentrations of TGF-I and IGF-I1 in different vertebrates as well as in adult vs. neonatal mice, that bone might be a major source of the circulating somatomedins. Previous investigations on IGF-I production in different tissues have revealed variability in the sensitivity to GH stimulation [29-301. Irl vitro, bone cells respond to various hormones besides GH with increased synthesis of IGF-I [ S - l l ] . If bone is a substantial contributor to the plasma pool of IGF-I, and the bone production of IGF-I is less GH-dependent than the liver production, this could explain the correlation between bone mass and IGF-I in the GH-deficient patients. Oestrogen is known to be important for bone mass in women of all ages [31]. and male hypogonadism is a known cause of secondary osteoporosis [32]. Although only some patients had had continuous sex hormone substitution (‘l’able l), there was a significant correlation between IGF-I and BMD. However, ocstrogen has been found to influence circulating IGF-I levels [33, 341, and post-menopausal women have lower circulating IGF-I levels than premenopausal women [34]. suggesting that thc influence of oestrogens on bone mass is to some extent due to circulating IGF-r. The importance of IGF-I per se was further emphasized by the multivariate analysis (Table 2), where the residual secretion of GH and the duration of GH deficiency also appeared to contribute to the variation in 13MD. The importance of GH for bone mineralization during childhood has been studied previously [ l a , 131. but BMD in patients with GH deliciency of adult onset is not known. In this study, the BMD of the total body was significantly lower in men with GH deficiency than in healthy subjects (Fig. 1).whereas there was no statistical difference for the female group. In the circulation, most of the IGF-I is bound to the large GH dependent binding protein complex, IGFBP3 1351. In GH deficiency, ICFBP-3 levels are low. while the concentrations of the smaller binding-
451
proteins (IGFBP-1 and - 2 ) are high [36]. IGFBP-1 and -2, but not IGFBP-3, can cross the capillary wall [37] and transport IGF-I into the tissues. In states with low IGFBP-3 and high IGFBP-1 and -2 levels, proportionally more IGF-I would be bound to IGFBP1 and -2, and be transported into the tissues than in healthy subjects with similar TGF-t levels. This might be a compensatory mechanism which preserves bone mass in these patients. In summary, our findings suggest that IGF-I, as well as the postulated autocrine/paracrine mechanisms of action, also has endocrine effects on bone mass.
Acknowledgements The technical assistance of KN Eva-Britt Borgestig is gratefully acknowledged.
References I Daughaday WH. Kotwein P. Insulin-like growth factors I and 11. Peptide. messenger ribonucleic acid and gene structures. serum. and tissue concentrations. Eridocr Kev 1989 : 10: 68-9 1 . 2 Hall K. Brandt J, Enberg G. Fryklund L. lmmunoreactive sometomedin A in human serum. / Cliri I:’nrlocrinol Metah 1979: 48: 271-8. 3 Banard R. Ng KW. Martin Tj. Waters Mj. Growth hormone (GH) receptors in clonal osteoblast-like cells mediate a niitogenic response to GH. Eiidocriiiolog!/ 1991 : 128: 1459-64. 4 Canalis E. Effect of insulin-like growth factor I on DNA and prctein synthesis in cultured rat calvariae. / C h i lrivest 1980: 66: 709-1 9. 5 Spencer E. I h C. Si E. Howard G. Iri vivo actions of insulin-like growth factor -I (IGF-I) on bone formation and resorption in rats. Hone 1991 : 12: 21-6. 6 Hock IM, Centrella M . Canalis 1.; Insulin-likc growth factor I has independent effects on bone matrix formation and cell replication. Ertdocriiiolog!j 1988 : 122 : 254-60. 7 I!rnst M. Froesch EK. Growth hormone dependent stimulation ofosteoblast-like cells in serum-free cultures via local synthesis of insulin-like growth factor 1. Bioch!rft Hioplt!/s Kes Cot?tiittiif 1988: 151: 142-7. 8 Chenu C . Valentin-Opran A , Chavassieux P. Saez S. Meunier PJ. Delmas PD. Insulin like growth factor I hormonal regulation by growth hormone and by 1,25(OH)2D3 and activity on human osteoblast-like cells in short-term cultures. Bone 1990: 1 1 : 81-6. 9 Canalis E. Centrella M, 13urch W, McCarthy TIJ. Insulin-like growth factor 1 mediates selective anabolic effects of parathyroid hormone in bone cultures. / Clirt Irtvest 1989 : 8 3 : 60-5. 1 0 Gray T K . Mohan S. Linkhart TA. Haylink Dj. Estradiol stimulates iit vitro the secretion of insulin-like growth factors by the clonal osteoblastic cell line, UMRlO6. Hioclteiit Hiopltys Kcs Comnrtirt 1989: 158: 407-12. 1 1 McCarthy TI,. Ccntrella M. Canalis 1;. Cortisol inhibits the
452
12
13
14
15
16
17
18
19
20
21
22
23
24
25
A. G. J O H A N S S O N et nl.
synthesis of insulin-like growth factor-I in skeletal cells. Erihcririolo{j~/1 9 9 0 : 126: 1569-75. Shore KM. Chesney KW. Maxess KB. Rose PG. Bargnian GJ. Bone mineral status in growth hormone deficiency. / Pedintr 1980: 96: 393-6. Kaufman J-M.. Taelman P. Vermeulen A. Vandeweghe M. Bone mineral status in growth hormone-delicient males with isolated and multiple pituitary deficiencies of childhood onset. / Clin Eridocrirrol Metab 1992: 74: 118-23. 1,junghall S. Johansson AC. Burman P. Kiimpe 0. Lindh E. Karlsson FA. Low plasma levels of IGF-I in men with idiopathic osteoporosis. / Iriterri Met1 1992: 232: 59-64. Lang I. Schernthaner G . Pietschmann P. KurA R. Stephenson IM, Temp1 H. Effects of sex and age on growth hormone response to growth hormone-releasing hormone in healthy individuals. / Cliri Endocririol Metnb 1 9 8 7 : 6 5 : 535-40. 13rixen K. Nielsen HK. Mosekilde I,. Flyvbjerg A. A short course of recombinant human growth hormone treatment stimulates osteoblasts and activates bone remodeling in normal human volunteers. / Hone Miner Kes 1990: 5 : 609-18. van der Veen EA. Netelenbos JC. Growth hormone (replacement) therapy in adults: bone and calcium metabolism. Horni Hes 1990: 33 (Suppl. 4): 65-8. I’arfitt AM. Growth hormone and adult bone remodelling. Clirt Eridocrirrol 1 9 9 1 : 35 : 467-70. Johansen JS, Pedersen SA. Jorgensen JOL et ol. Effects of growth hormone (GH) on plasma bone Gla protein in GH deficient adults. / Cliri Errdocririol Metrrb 1990: 70: 91 6-9. lensen LT. Jorgenson JOL. Kisteli J, Christiansen JS, Lorenxen 1. Type I and III procollagen propeptides in growth hormonedeficient patients: effects of increasing doses of GH. Actn Ertdocririol (Copenh) 199 1 : 124: 278-82. Kudman D. Feller AC. Hoskote SN et a/. Effects of human growth hormone in men over 60 years old. N I h g I / Med 1 9 9 0 : 323: 1-6. Zamboni G. Antoniazzi F. Kadetti G . Musemeci C. Tat0 L. IXects of two different regimens of recombinant human growth hormone therapy on the bone mineral density of patients with growth hormone deficiency. / Pedintr 199 1 : 119: 483-5. Kiggs 131,. Wahner HW. Seeman E e l a/. Changes in bone mineral density of the proximal femur and spine with aging. / Clin Irivest 1 9 8 2 : 70: 716-23. Poehlman ET. Copeland KC. Influence of physical activity on insulin-like growth factor-l in healthy younger and older men. / Cliri Ertdocririol Metob 1990: 71 : 1468-73. Bennett AE. Wahner HW. Kiggs BL. Hinta KL. Insulin-like growth factors I and II: aging and bone density in women. / Clin Eridocrinol Metab 1 9 8 4 : 59: 701-4.
26 Kelly PJ. Eisman ]A, Stuart MC. Pocock NA. Sambrook PN, Gwinn TH. Somatomedin-C. physical fitness. and bone density. / Cliri Ertdocriiid Metnb 1990: 70: 7 18-23. 2 7 Il’Ercole AJ. Stiles AD. Underwood LI:. Tissue concentrations of somatomadin C: further evidence for multiple sites of synthesis and paracrine or autocrine mechanisms of action. I’roc Not1 Acrtrl Sci U S A 1984: 81 : 935-9. 28 Bautista CM. Mohan S. Baylink DJ. Insulin-like growth factors I and I I are present in the skeletal tissues of ten vertebrates. Metabolisrrr 1990: 39: 96-100. 29 Murphy LJ. Friescn HG. Ilifferential effects of estrogen and growth hormone on uterine and hepatic insulin-like growth factor I gene expression in the ovariectomizcd hypophysectomized rat. Eiidocrirtolog!/ 1988 : 122: 325-32. 30 Grant AL. Helferich WG. Kramer SA. Merkel KA. Bcrgcn WG. Administration of growth hormone to pigs alters the relative amount of insulin-like growth factor-l mKNA in liver and skeletal muscle. / Eridocrirrol 1 9 9 1 : 130: 3 3 1-8. 31 Richelson I S . Wahner HW. Melton. Lj I l l . Kiggs 131,. Relative contributions of aging and estrogen deficiency to postmenopausal bone loss. N Erigl/ Med 1984: 31 1 : 1273-5. 32 Jackson JA. Kleerekoper M. Osteoporosis in men : diagnosis, pathophysiology. and prevention. Medicirre 1990: 6 9 : 137-52. 33 Frohlander N. von Schoula B. Growth hormone and somatoniedin C during postmenopausal replacenient therapy with oestrogen alone and in combination with an antioestrogen. Mrituritrts 1 9 8 8 : 9 : 297-302. 34 Weissberger A]. Ho KKY. Lazarus I,. Contrasting effects of oral and transdermal routes of estrogen replacement therapy on 24-hour growth hormone (GH) secretion. insulin-like growth factor I. and GH-Binding protein in postmcnopsusal women. / Cliri Eridocririol Mefrtb 199 1 : 72 : 3 74-8 I . 35 Hintx KL. Liu I:. Demonstration of specific plasma protein binding sites for somatomedin. / Cliri Ertclocririol Mettrh 1977: 45: 988-95. 36 Hardoin S. Gourmelen M. Noguica P et ctl. Molecular forms of serum insulin-like growth factor (IGI~)-bindingproteins in man : relationships with growth hormone and IGPs and physiological significance. / Cliri f3idocririol Metctb 19x9 : 69: 1291 -30 1 . 37 Binoux M , Hossenlopp 1’. Insulin-like growth f k t o r (IGF) and IGF binding proteins: comparison of human serum and lymph. / Cliri Eridocrirtol Metnb 1988 : 6 7 : 509-1 4. Received 2 7 April 1992. accepted 1 5 July 1992. Corresporiderice: Professor Sverker Ljunghall. Ilcpartment of Internal Medicine. University Hospital, S-75 1 8 5 Uppsala. Sweden.
The bone mineral density in acquired growth hormone deficiency correlates with circulating levels of insulin-like growth factor I A. G. JOHANSSON, P. BURMAN, K. WESTERMARK & S. I,JUNGHAI,I, Frorti tlie 1leptrrtrrierit 01Iriterrial Medicine. University Hospital. Uppsola. Sumfen
Abstract. Johansson AG. Burman P. Westermark K. Ljunghall S (Department of Internal Medicine, University Hospital. Uppsala, Sweden). The bone mineral density in acquired growth hormone deficiency correlates with circulating levels of insulin-like growth factor I. lorrrtial of Internal Medicine 1992 : 232 : 447-452.
Insulin-like growth factor I (IGF-I) is an important anabolic factor for osteoblasts in vitro. Low plasma levels of IGF-I have been observed in young men with osteoporosis. In the present study, we have studied bone mineral density (BMD) and the circulating levels of IGF-I and growth hormone (GH) in adults with acquired CH deficiency. f3MI) was determined by dual-energy x-ray absorptiometry in 1 7 men and 12 women (age 27-54 years). Spinal BMD was positively correlated with the plasma levels of IGF-I ( r = 0.43, P = 0.01 9), with the median of GH values obtained by repeated sampling a t night ( r = 0.43. I-' = 0.001 9), and with the peak of GH values during GHKH provocation test (r = 0.49. P = 0.039). The total BMD was positively related to plasma IGF-I and median of GH values. but not to peak GH by GHKH provocation. In a multiplc regression analysis model, IGF-I. peak GH by GHKH provocation test and duration of GH deficiency explained 49%)of the variation in spinal BMD. As compared to healthy controls, total, but not spinal. bone mass was lower in men with GH deficiency, but no clinical symptoms of osteoporisis were observed. The positive relationships between BMD and circulating IGF-I and other indices of GH secretion suggest that IGF-I has an endocrine effect on bone mass. Keywords: bone mineral density, growth hormone, growth hormone deficiency, insulinlike growth factor I .
lntroduction Insulin-like growth factor I (IGF-I, Somatomedin-C) is known to mediate actions of growth hormone (GH). and to have endocrine as well as paracrine and autocrine effects [ l l . The plasma levels of IGF-I are low in GH deliciency 121. IGF-I and GH affect skeletal metabolism in vivo as well as in vitro r3-51. IGF-I stimulates collagen synthesis and osteoblast proliferation [6], while GH has been reported to act via stimulation of IGF-I synthesis in the bone tissue r71. Human osteoblastlike cells have the ability to synthesize IGF-I, a capacity that is increased by GH [S]. The IGF-I synthesis in bone it1 vitro is also influenced by other hormones important for skeletal metabolism, e.g. parathyroid hormone, calcitriol, cortisol and oestrogen [S-ll].
Osteopenia has been reported in GH-deficient children [12], and in adults with GH deficiency of childhood onset [13], but bone mass in adult-onset GH deficiency has not been thoroughly investigated. We have previously reported on low IGF-I levels in male patients with idiopathic osteoporosis [I 41. In the present study, our aim was to investigate the relationship between BMD and circulating IGF-I in adult patients with acquired CH deficiency.
Patients and methods A total of 1 7 men and 1 2 women (mean age ( k S D ) 4 4 . 9 k 7 . 4 years, range 27-57 years) with GH deficiency were included in the study ('l'able 1).All patients had a GH peak below 3 pg I-' in response to insulin-induced hypoglycaemia (blood glucose d 2.2 mmol I-'). The known duration of GH de-
44 7
448
A. G. JOHANSSON et d.
‘I’ahle 1 . Patients with acquired growth hormone tlclicicncy Patient no.
Sex (M/l:)
Weight (kg)
Length (cm)
Ilu ration (years)
h7
168
93
I73
42
98
47 33
98 79
I82 1x1
34 2h 19
50 36 38 42 42
97 85
Age (years)
Iliagnosis
Sex steroid substitution
I3MIl
(g cni-?)
~
1
M
-7
M M M
3 4 5
M
6
M
7 X
M M
9
M
10
M M
46 44
177 I76
9
C.A
9
CIJ
cont. cont. intermitt.
1 149 1 031
I 105 I 1x8 I219
cont. intermitt. cont.
-/>
178
7
93 92
1 x9
C.A C.A C.A
cont. cont.
176
6 6
78
177
5
C.A
174 18 0 190
3
C.A C.A
27
91 89 73
cont. cont. cont.
3
trau niii
cont.
41 49 3h
83 X2 76
175
3 1
C.A
cont. cont.
57 48
26 24 20
C.A C.1’ C.A
16
CU
I6
empty sella
12 8
C.A C.A C.A
M M 1\11
I5
M
1 (7
17
M M
18
I:
19 20 21
F 1;
51 44 50
I!
55
22 23 24 25
1: 1: 1: 1:
43 50 54
26 27 28 29
1: 1: i:
apoplexia CIJ C.A
10 10
1 13X 0 925 I 304 I 053 I 096
cont.
180
11 12 13 14
I:
C.1’ C.1’
51 54 40
49
57 54 36 41
I72 168
157
hI 78
166 163 I66
65 73 85 x3
158 168 165 179
64
I65 I59 1 f> 1 162
6I
60
-13
3
*
6 2 1
* *
apoplcxia idiopathic
C.A
empty sella C.1’
C.A
cont.
0 9x9 I 128 0 940 I 000
1087
cont.
0 945 I 003
intermitt. never intermitt. postnienop.
0 920 1 024 0 958 0 927
cont. never postnienop. cont. postmenop. postnienop. cont. cont.
I053 0 889
1 508 12x2 0 x9x 1 I63 1000 I 034
’I’he causes of GH deliciency were. if not stated otherwise. pituitary tumours treated with surgery and/or radiation. C.1’ =
craniopharyngioni~i,CIJ = Cushing’s disease. C . A = cromophobic adenonia. The ctiuse of hypopituitarism could not be established in patient no. 17. All patients. except patients nos 3 and 24 who were only deficient in GH and FSH/I,H. were adequately substituted with thyroxine. glucocorticoids and. if necessary. with desmopressin. Some of the patients did not receive sex steroid substitution despite lack of gonadotropins. Intermitt. = intermittently. cont. = continuously substituted with sex steroids during the known period of hypopituitarisni. Patients nos 22. 24. 26 and 27 were postmenopausal at the onset of hypopituitarism. Iluration = duration of CH deliciency. The asterisks indicate that the duration was not possible to establish. 13MIl = bone niineral density in the spine measured by dual energy x-ray absorptiometry of the total body.
liciency was 11.2 1-9.1 years (range 1-34 years). A total of 1 7 patients had been treated for pituitary tumours with radiation and/or surgery. In one patient, panhypopituitarism had developed after a skull, trauma. while the aetiology of pituitary insufficiency could not be established in one patient. All but two patients suffered from panhypopituitarism. and were treated with thyroxine and glucocorticoids, and their urinary cortisol and serum T:, and free T, levels were within the normal ranges. Seven patients also received a vasopressin-analogue, desmopressin. Patients nos 1 9 and 24 had partial hypopituitarism and were only deficient in GH and FSH/LH. All of the
men and five of the women were substituted with sex hormones at the time of the study, but not all of them had had continuous substitution despite panhypopituitarism. Six of the women had never received sex steroids, and four of them were already postmenopausal at the onset of gonadotropin deficiency (Table 1). None of the patients had any known fragility fractures. BMD of the total body was measured by dual energy X-ray abosorptiometry (equipment DPX-I, from Lunar, USA). BMD of the lumbar region (BMD spine) was obtained by regional analysis of the total body measurement. BMD of the total body was also
BONE MASS A N D IGF-I IN ACQUIRED GH DEFICIENCY
nieasured in 29 healthy men, members of the hospital staff. selected to match the male patients with GH deliciency regarding age. IGF-I concentrations in plasma were measured by direct radioimmunoassay using a commercial kit from the Nichols Institute (San Capistrano, CA. USA). The reference ranges were 0.34-1.9 U ml-' for men and 0.45-2.2 U ml-' for women. Repeated measurements of GH were made using a vacuum pump with samples drawn every 20 min between 22.00 and 04.00 hours. GH was analysed by radioimmunoassay. The median of the GH values in each patient was used in the calculations. GHRH (Groliberin, Kabi, Sweden) provocation test was done in 2 7 patients. 1 pg kg-' body weight of GHKH was injected intravenously and GH samples were drawn every 1 5 min for 9 0 min. the highest value being used in the calculations. The statistical package STATVIEW SE+ v 1.03 from Abacus Concepts, Inc. (Berkeley, CA, USA) was used for all statistical calculations, including Student's ttest, Pearson's correlation coefficients and multiple regression analysis. All tests were two-tailed and P < 0.05 was considered to be statistically significant.
449
I .6
(a) total body
-
..
E0.032
I .5 I .4
b) spine e0.093
..
N
'
Em
1.3
0
Y
0 1.2
I
m
1.1
I .o 0.9
Patients Controls Patients Controls Fig. 1 . Bone mineral density (BMI))of (a) the total body and (b) the spine in 1 7 male patients with acquired GH deticiency (0) and in 29 age-matched healthy men
(m).
I
0.7 1
1 I
0.6
I
0
0
0
Heslrl t s
individual BMD spine values are given in Table 1. Compared to the normal healthy male controls of the same ages, the male patients had significantly lower BMD of the total body ( 1 . 2 0 f 0 . 0 8 vs. 1 . 2 6 f 0 . 0 8 , P = 0.032; Fig. 1) but not in the spine (1.08+0.11 vs. 1 .1 4 + 0 .1 2 . P = 0.093; Fig. 1). The female patients were compared to age- and weight-matched reference material provided by the manufacturer, and the median Z score for BMD spine was -0.75 (95% confidence interval: - 1.11-0.34) while it was -0.12 ( 9 5 % confidence interval: -0.89-0.6) for total body BMD. The plasma levels of IGF-I were markedly lower than normal. and amounted to 0 . 2 7 k 0 . 1 U mi-' in male patients and 0.32 0.2 U ml-' in female patients. The median GH value obtained by repeated sampling at night was 0 . 5 4 k 0 . 2 1 pg I-', and the highest GH value in any patient was 1.5 pg I-'. The mean peak GH responses to GHRH provocation tests w a s 2 . 2 f 1 . 4 p g l - I f o r t h e m e n a n d 1 . 8 f 1 . 3 pg1-l for the women, clearly below the reported values in these age groups [15]. There was a direct correlation between BMD of the spine and plasma IGF-I (r = 0.43, P = 0.019; Fig.
0.8 0.9
1.0
1.1
1.2
1.3
1.4
1.5
1.6
BMD of spine (g cm-')
Fig. 2. Relationship between bone mineral density (BMD) of the spine and plasma levels of insulin-like growth factor I (IGF-I) in patients with acquired GH deficiency. ( 0 )= men. ( 0 )= women.
2). BMD was also positively related to the median of GH values obtained by repeated sampling at night ( r = 0.43, P = 0.019; Fig. 3) and to the peak of GH during GHRH provocation (r = 0.40, P = 0.039 : Fig. 4). For total body BMD, positive correlations were found with IGF-I (r = 0.44, P = 0.018) and with the median of GH values at night (r = 0.40, P = 0.032) but not with the peak of GH by GHRH (r = 0.23, P = 0.25) (individual data not shown). A multiple regression analysis was performed including age, sex, weight, duration of GH deficiency, as well as median and stimulated GH levels. Circulating IGF-I appeared to be the major factor that explained variations in spinal bone mass (Table 2). Forty-nine per cent of the variation in spinal BMD was explained by IGF-I, peak, GH value and duration of GH deficiency.
Age Median GH n intercept r2 P-value
0.50
0.09
24 0.56 0.58 0.029
1.42
0.93
0.12
Sex
1.62
1.31
2.78
0.82
t-value
-0.004
0.004
Weight
-0.004
0.07
Peak GH by GHRH
Duration
0.19
IGF-I
B
-0.005
0.12
0.004
-0.004
0.08
0.26
B
24 0.58 0.57 0.01 5
0.99
1.44
1.68
1.29
3.31
1.34
t-value
0.07
0.003
-0.004
0.07
0.34
B 0.07
0.36
B
24 0.73 0.52 0.0059
0.97
1.35
3.23
2.05
t-value
-0.005
0.08
0.38
/3
24 0.87 0.49 0.003 1
1.76
3.49
2.20
t-value
3
1
O O
0.002
-0.004
+
@ O
24 0.52 0.54 0.0095
1.04
1.41
1.23
3.16
1.90
t-value
0.04
0.44
B
27 0.86 0.34 0.0075
2.32
2.52
t-value 0.46
B
29 0.94 0.19 0.019
2.49
t-value
Table 2. Partial correlation coefficients (8) for determinants of bone mineral density of the spine (BMD). as determined by multiple linear regression analysis. incorporating age. weight, sex, median of GH values obtained by repeated sampling during 6 h at night. duration of growth hormone deficiency, GH peak in response to GHRH provocation test. and plasma levels of IGF-I as the independent variables
BONE MASS AND IGF-I I N ACQUIKED GH DEFICIENCY
well as seruni levels of IGF-I decline with age and lowered physical fitness [23, 241, and previous studies have shown a positive relationship between BMD and circulating IGF-I in healthy subjects, although dependent of age [25] and physical fitness 1261. The liver has been reported to be the main contributor to plasma IGF-I [2 71. However, Bautista et ol. [28] suggested, on the basis of their results from a comparison of skeletal and serum concentrations of TGF-I and IGF-I1 in different vertebrates as well as in adult vs. neonatal mice, that bone might be a major source of the circulating somatomedins. Previous investigations on IGF-I production in different tissues have revealed variability in the sensitivity to GH stimulation [29-301. Irl vitro, bone cells respond to various hormones besides GH with increased synthesis of IGF-I [ S - l l ] . If bone is a substantial contributor to the plasma pool of IGF-I, and the bone production of IGF-I is less GH-dependent than the liver production, this could explain the correlation between bone mass and IGF-I in the GH-deficient patients. Oestrogen is known to be important for bone mass in women of all ages [31]. and male hypogonadism is a known cause of secondary osteoporosis [32]. Although only some patients had had continuous sex hormone substitution (‘l’able l), there was a significant correlation between IGF-I and BMD. However, ocstrogen has been found to influence circulating IGF-I levels [33, 341, and post-menopausal women have lower circulating IGF-I levels than premenopausal women [34]. suggesting that thc influence of oestrogens on bone mass is to some extent due to circulating IGF-r. The importance of IGF-I per se was further emphasized by the multivariate analysis (Table 2), where the residual secretion of GH and the duration of GH deficiency also appeared to contribute to the variation in 13MD. The importance of GH for bone mineralization during childhood has been studied previously [ l a , 131. but BMD in patients with GH deliciency of adult onset is not known. In this study, the BMD of the total body was significantly lower in men with GH deficiency than in healthy subjects (Fig. 1).whereas there was no statistical difference for the female group. In the circulation, most of the IGF-I is bound to the large GH dependent binding protein complex, IGFBP3 1351. In GH deficiency, ICFBP-3 levels are low. while the concentrations of the smaller binding-
451
proteins (IGFBP-1 and - 2 ) are high [36]. IGFBP-1 and -2, but not IGFBP-3, can cross the capillary wall [37] and transport IGF-I into the tissues. In states with low IGFBP-3 and high IGFBP-1 and -2 levels, proportionally more IGF-I would be bound to IGFBP1 and -2, and be transported into the tissues than in healthy subjects with similar TGF-t levels. This might be a compensatory mechanism which preserves bone mass in these patients. In summary, our findings suggest that IGF-I, as well as the postulated autocrine/paracrine mechanisms of action, also has endocrine effects on bone mass.
Acknowledgements The technical assistance of KN Eva-Britt Borgestig is gratefully acknowledged.
References I Daughaday WH. Kotwein P. Insulin-like growth factors I and 11. Peptide. messenger ribonucleic acid and gene structures. serum. and tissue concentrations. Eridocr Kev 1989 : 10: 68-9 1 . 2 Hall K. Brandt J, Enberg G. Fryklund L. lmmunoreactive sometomedin A in human serum. / Cliri I:’nrlocrinol Metah 1979: 48: 271-8. 3 Banard R. Ng KW. Martin Tj. Waters Mj. Growth hormone (GH) receptors in clonal osteoblast-like cells mediate a niitogenic response to GH. Eiidocriiiolog!/ 1991 : 128: 1459-64. 4 Canalis E. Effect of insulin-like growth factor I on DNA and prctein synthesis in cultured rat calvariae. / C h i lrivest 1980: 66: 709-1 9. 5 Spencer E. I h C. Si E. Howard G. Iri vivo actions of insulin-like growth factor -I (IGF-I) on bone formation and resorption in rats. Hone 1991 : 12: 21-6. 6 Hock IM, Centrella M . Canalis 1.; Insulin-likc growth factor I has independent effects on bone matrix formation and cell replication. Ertdocriiiolog!j 1988 : 122 : 254-60. 7 I!rnst M. Froesch EK. Growth hormone dependent stimulation ofosteoblast-like cells in serum-free cultures via local synthesis of insulin-like growth factor 1. Bioch!rft Hioplt!/s Kes Cot?tiittiif 1988: 151: 142-7. 8 Chenu C . Valentin-Opran A , Chavassieux P. Saez S. Meunier PJ. Delmas PD. Insulin like growth factor I hormonal regulation by growth hormone and by 1,25(OH)2D3 and activity on human osteoblast-like cells in short-term cultures. Bone 1990: 1 1 : 81-6. 9 Canalis E. Centrella M, 13urch W, McCarthy TIJ. Insulin-like growth factor 1 mediates selective anabolic effects of parathyroid hormone in bone cultures. / Clirt Irtvest 1989 : 8 3 : 60-5. 1 0 Gray T K . Mohan S. Linkhart TA. Haylink Dj. Estradiol stimulates iit vitro the secretion of insulin-like growth factors by the clonal osteoblastic cell line, UMRlO6. Hioclteiit Hiopltys Kcs Comnrtirt 1989: 158: 407-12. 1 1 McCarthy TI,. Ccntrella M. Canalis 1;. Cortisol inhibits the
452
12
13
14
15
16
17
18
19
20
21
22
23
24
25
A. G. J O H A N S S O N et nl.
synthesis of insulin-like growth factor-I in skeletal cells. Erihcririolo{j~/1 9 9 0 : 126: 1569-75. Shore KM. Chesney KW. Maxess KB. Rose PG. Bargnian GJ. Bone mineral status in growth hormone deficiency. / Pedintr 1980: 96: 393-6. Kaufman J-M.. Taelman P. Vermeulen A. Vandeweghe M. Bone mineral status in growth hormone-delicient males with isolated and multiple pituitary deficiencies of childhood onset. / Clin Eridocrirrol Metab 1992: 74: 118-23. 1,junghall S. Johansson AC. Burman P. Kiimpe 0. Lindh E. Karlsson FA. Low plasma levels of IGF-I in men with idiopathic osteoporosis. / Iriterri Met1 1992: 232: 59-64. Lang I. Schernthaner G . Pietschmann P. KurA R. Stephenson IM, Temp1 H. Effects of sex and age on growth hormone response to growth hormone-releasing hormone in healthy individuals. / Cliri Endocririol Metnb 1 9 8 7 : 6 5 : 535-40. 13rixen K. Nielsen HK. Mosekilde I,. Flyvbjerg A. A short course of recombinant human growth hormone treatment stimulates osteoblasts and activates bone remodeling in normal human volunteers. / Hone Miner Kes 1990: 5 : 609-18. van der Veen EA. Netelenbos JC. Growth hormone (replacement) therapy in adults: bone and calcium metabolism. Horni Hes 1990: 33 (Suppl. 4): 65-8. I’arfitt AM. Growth hormone and adult bone remodelling. Clirt Eridocrirrol 1 9 9 1 : 35 : 467-70. Johansen JS, Pedersen SA. Jorgensen JOL et ol. Effects of growth hormone (GH) on plasma bone Gla protein in GH deficient adults. / Cliri Errdocririol Metrrb 1990: 70: 91 6-9. lensen LT. Jorgenson JOL. Kisteli J, Christiansen JS, Lorenxen 1. Type I and III procollagen propeptides in growth hormonedeficient patients: effects of increasing doses of GH. Actn Ertdocririol (Copenh) 199 1 : 124: 278-82. Kudman D. Feller AC. Hoskote SN et a/. Effects of human growth hormone in men over 60 years old. N I h g I / Med 1 9 9 0 : 323: 1-6. Zamboni G. Antoniazzi F. Kadetti G . Musemeci C. Tat0 L. IXects of two different regimens of recombinant human growth hormone therapy on the bone mineral density of patients with growth hormone deficiency. / Pedintr 199 1 : 119: 483-5. Kiggs 131,. Wahner HW. Seeman E e l a/. Changes in bone mineral density of the proximal femur and spine with aging. / Clin Irivest 1 9 8 2 : 70: 716-23. Poehlman ET. Copeland KC. Influence of physical activity on insulin-like growth factor-l in healthy younger and older men. / Cliri Ertdocririol Metob 1990: 71 : 1468-73. Bennett AE. Wahner HW. Kiggs BL. Hinta KL. Insulin-like growth factors I and II: aging and bone density in women. / Clin Eridocrinol Metab 1 9 8 4 : 59: 701-4.
26 Kelly PJ. Eisman ]A, Stuart MC. Pocock NA. Sambrook PN, Gwinn TH. Somatomedin-C. physical fitness. and bone density. / Cliri Ertdocriiid Metnb 1990: 70: 7 18-23. 2 7 Il’Ercole AJ. Stiles AD. Underwood LI:. Tissue concentrations of somatomadin C: further evidence for multiple sites of synthesis and paracrine or autocrine mechanisms of action. I’roc Not1 Acrtrl Sci U S A 1984: 81 : 935-9. 28 Bautista CM. Mohan S. Baylink DJ. Insulin-like growth factors I and I I are present in the skeletal tissues of ten vertebrates. Metabolisrrr 1990: 39: 96-100. 29 Murphy LJ. Friescn HG. Ilifferential effects of estrogen and growth hormone on uterine and hepatic insulin-like growth factor I gene expression in the ovariectomizcd hypophysectomized rat. Eiidocrirtolog!/ 1988 : 122: 325-32. 30 Grant AL. Helferich WG. Kramer SA. Merkel KA. Bcrgcn WG. Administration of growth hormone to pigs alters the relative amount of insulin-like growth factor-l mKNA in liver and skeletal muscle. / Eridocrirrol 1 9 9 1 : 130: 3 3 1-8. 31 Richelson I S . Wahner HW. Melton. Lj I l l . Kiggs 131,. Relative contributions of aging and estrogen deficiency to postmenopausal bone loss. N Erigl/ Med 1984: 31 1 : 1273-5. 32 Jackson JA. Kleerekoper M. Osteoporosis in men : diagnosis, pathophysiology. and prevention. Medicirre 1990: 6 9 : 137-52. 33 Frohlander N. von Schoula B. Growth hormone and somatoniedin C during postmenopausal replacenient therapy with oestrogen alone and in combination with an antioestrogen. Mrituritrts 1 9 8 8 : 9 : 297-302. 34 Weissberger A]. Ho KKY. Lazarus I,. Contrasting effects of oral and transdermal routes of estrogen replacement therapy on 24-hour growth hormone (GH) secretion. insulin-like growth factor I. and GH-Binding protein in postmcnopsusal women. / Cliri Eridocririol Mefrtb 199 1 : 72 : 3 74-8 I . 35 Hintx KL. Liu I:. Demonstration of specific plasma protein binding sites for somatomedin. / Cliri Ertclocririol Mettrh 1977: 45: 988-95. 36 Hardoin S. Gourmelen M. Noguica P et ctl. Molecular forms of serum insulin-like growth factor (IGI~)-bindingproteins in man : relationships with growth hormone and IGPs and physiological significance. / Cliri f3idocririol Metctb 19x9 : 69: 1291 -30 1 . 37 Binoux M , Hossenlopp 1’. Insulin-like growth f k t o r (IGF) and IGF binding proteins: comparison of human serum and lymph. / Cliri Eridocrirtol Metnb 1988 : 6 7 : 509-1 4. Received 2 7 April 1992. accepted 1 5 July 1992. Corresporiderice: Professor Sverker Ljunghall. Ilcpartment of Internal Medicine. University Hospital, S-75 1 8 5 Uppsala. Sweden.
